THE MOLELEKI EXECUTION: A RADICAL PROBLEM OF UNDERSTANDING Vanessa Emma Barolsky A dissertation submitted to the Faculty of Humanities, University of the Witwatersrand, Johannesburg in fulfilment of the requirements for the degree of Doctor of Philosophy (Sociology) 25 January 2010 DECLARATION I declare that this thesis is my own unaided work. It is submitted for the Degree of Doctor of Philosophy in the Faculty of Humanities, University of Witwatersrand, Johannesburg. It has not been submitted before, for any other degree or examination at any other university. -------------------------------------------- Vanessa Emma Barolsky ?.... th day of ??????.??., 20?.. CONTENTS CHAPTER ONE: THE MOLELEKI EXECUTION, A RADICAL PROBLEM OF UNDERSTANDING ................................................................................................................ 1 INTRODUCTION ..................................................................................................................... 1 ?A Finishing Off? .................................................................................................................. 1 The Moleleki Execution, Moleleki Extension 2, 9am, Tuesday 7 December 1993 .............. 1 BEFORE THE EXECUTION.................................................................................................... 3 ?Death Has Arrived? .............................................................................................................. 3 Killing of Civic Activist Bulelwa Zwane .......................................................................... 3 The Call of Bulelwa?s Death, a Lacuna at the Centre of Power ............................................ 5 Residents Gather at Block C .............................................................................................. 6 Killing of SDU leader Blanko ................................................................................................ 7 ?Those people must be attacked?, retaliation for Blanko?s death .......................................... 9 An Order is Issued for Attack by SDU Commander Sugar Ramabele .............................. 9 Abduction of 14 ANCYL members ............................................................................... 9 ?Parents, What Should We Do If People Are Killing People In This Fashion?? ................ 10 ?Presentation? of the abducted youth to the community ................................................. 10 ?Hey, You Boys, What Have You Done?? .......................................................................... 12 Fourteen Young Boys are Incarcerated in a Shack in Block F ........................................ 12 Gathering Addressed by Civic Leader Machinini ............................................................... 14 A RADICAL PROBLEM OF UNDERSTANDING ............................................................... 17 A Mutable Truth .................................................................................................................. 20 The Articulation of Biopolitical and Juridico-Institutional Power, the ?Hidden Nucleus? of Sovereign Power .......................................................................................................... 21 The Moleleki execution, a biopolitical power ............................................................. 23 A Juridico-Political Adjudication of the Moleleki Execution ......................................... 25 The Truth and Reconciliation Commission and the Moleleki execution, an act associated with a political objective ............................................................................ 28 The trial of the Moleleki accused, an association with a common purpose ................ 31 METHODOLOGY .................................................................................................................. 34 Introduction .......................................................................................................................... 34 Genealogy/Archaeology ...................................................................................................... 36 An Encounter with the Moleleki Execution..................................................................... 38 Documentary material .................................................................................................. 39 In-depth interviewing and focus group ........................................................................ 41 CHAPTER OUTLINE ............................................................................................................. 45 CHAPTER TWO: ANALYSING A CONDITION OF EXCEPTION ............................. 52 INTRODUCTION: THE EXCEPTION OF SOUTH AFRICAN VIOLENCE ...................... 52 ii Misrecognising the Nature of the Sovereign Exception ...................................................... 53 A Normative Conception of Sovereignty ........................................................................ 55 A breakdown of state sovereignty................................................................................ 59 A distortion of state sovereignty .................................................................................. 65 Sovereignty and the Exception ............................................................................................ 70 The Structure of the Exception, an Inclusive Exclusion .................................................. 72 ?Pure? Violence, the Object of Politics ........................................................................... 74 Narrating an Original Site of Power .................................................................................... 77 The Subject, Negotiating the Relationship between Cause and Effect ............................ 89 The biopolitical subject ................................................................................................ 93 Homo sacer .................................................................................................................. 95 CHAPTER THREE: A STRUGGLE OF JURIDICO-POLITICAL ORDERING ...... 101 INTRODUCTION ................................................................................................................. 101 Moleleki Extension 2, a Space ........................................................................................... 101 Tying Land to Juridical Order on the East Rand ........................................................... 106 Homo sacer, defining the boundaries of the political community ............................. 111 ?Who Must Rule the Area?? A Struggle of Juridico-Political Ordering ........................... 114 Auctoritas and Potestas .................................................................................................. 124 THE MOLELEKI EXECUTION, A LOCALISATION OF THE EXCEPTION IN THE BODY OF HOMO SACER .................................................................................................... 134 The Body of Homo Sacer .................................................................................................. 134 The Stray Dog ................................................................................................................ 136 A Finishing Off .............................................................................................................. 138 CHAPTER FOUR: A JURIDICO-POLITICAL INVESTIGATION ............................ 145 INTRODUCTION ................................................................................................................. 145 A Mutable Truth, the Juridico-Political Investigation of the Body of Exception .............. 146 Sovereign Police ............................................................................................................ 147 Collecting the ?criminal? body of exception ............................................................. 152 A medico-legal inscription of the body of exception, notarising death ..................... 155 A Juridico-Political Investigation of Death ................................................................... 161 Police investigation, a juridical application ............................................................... 164 TRC investigation of the Moleleki execution, a juridical suspension ....................... 167 CHAPTER FIVE: THE TRUTH AND RECONCILIATION COMMISSION, FIXING THE LIMITS OF THE POLITICAL ................................................................................ 170 INTRODUCTION ................................................................................................................. 170 The TRC, Making an Exception of the Past ...................................................................... 174 Amnesty, Effecting a Relation of Exception ..................................................................... 177 The Juridical Classification of the ?Political? ? An Enigma ............................................. 179 The Political Offence Exception .................................................................................... 180 The Norgaard principles ............................................................................................ 183 The TRC, Establishing an Instrumental Relationship with Violence Committed in Pursuit of Political Objectives ........................................................................................................ 187 Exposing the Relation Between Force and Law ............................................................ 189 iii Adjudicating an instrumental relation with the biopolitical effects of power ........... 190 The Ambiguity of the Political ...................................................................................... 191 A ?political motive?, an adjudication of subjecthood ................................................ 193 CHAPTER SIX: THE MOLELEKI EXECUTION AND THE TRUTH AND RECONCILIATION COMMISSION ............................................................................... 199 INTRODUCTION ................................................................................................................. 199 State of Necessity ............................................................................................................... 201 TRC Adjudication of the Moleleki Execution ................................................................... 204 A Problem of Decidability, Adjudicating the ?Right? to Self-defence ......................... 207 The ?legality? of self-defence .................................................................................... 208 Community defence and political objectives ............................................................. 210 SDUs and the ANC ........................................................................................................ 213 Homo Sacer, Mediating the Boundaries of the Community .......................................... 221 CHAPTER SEVEN: FIXING THE LIMITS OF THE LAW, THE MOLELEKI EXECUTION AND THE COURT TRIAL ....................................................................... 233 INTRODUCTION ................................................................................................................. 233 The Common Purpose Rule ............................................................................................... 235 Applying the Common Purpose Rule to the Moleleki Execution...................................... 240 CHAPTER EIGHT: CONCLUSION, THE EXCEPTION, A RADICAL PROBLEM OF UNDERSTANDING ...................................................................................................... 260 REFERENCES ..................................................................................................................... 270 iv ACKNOWLEDGMENTS Producing this dissertation has been an extraordinary journey. It has been long (as anyone who has known me over the past five years will testify!), financially and emotionally demanding, rewarding and frustrating by turns. I would like to thank, and offer apologies to those close family and friends who on too many occasions were paid less attention than this thesis, which became both a burden, a distraction and an obsession. In particular I would like to thank my younger sister Kathy, who had to put up with my tunnel vision and a thesis with whom she had to sometimes fight for space. You?ve just passed your master?s thesis with flying colours so I can only assume I set a good example! I would also like to thank my aunt Anne Sak, our ?surrogate parent?. Although you grew impatient with the extended process of the dissertation, and asked me how I could possibly be taking so long to write about one thing, you nevertheless remained supportive. I?d like to thank my brother Adam for his support and long crazy skype conversations including babies and dogs. To my partner Ebrahim Fakir, who has been lobbying for an extensive acknowledgement, I can only say how much I appreciate your support, your generous pride in my work, your humour, your presence, your intelligence, our endless debates on politics, religion and more trivial matters, and your kindness. You are in the acknowledgements! I would like to offer many thanks to my supervisor Bridget Kenny. Your astute and timely emotional and intellectual advice and your much needed support made it possible for me to complete this thesis. I really appreciate your contribution. I would also like to thank Belinda Bozzoli for initially encouraging me to convert my master?s thesis into a PHD. I would like to acknowledge the many people, who have contributed to my intellectual development, exposed me to new ideas, challenged my preconceptions and offered much needed encouragement. In particular I would like to acknowledge my former ?boss? at the Truth and Reconciliation Commission Janis Grobbelaar who believed that I had an intellectual contribution to make and wisely advised me to get some more ?paper?. So here is the paper! Finally I would like to thank the residents of Moleleki, particularly those I interviewed for their time and trust in telling me their stories. Thank-you to Jabu Dlamini for assisting in this process. I can only hope that this thesis adds something of value for those still living in Moleleki today, many years after the execution. I dedicate this thesis to my little niece Sophia, the first of a new generation of Barolsky?s. v KATLEHONG TOWNSHIP SOUTH OF JOHANNESBURG1 1 Google Maps South Africa. Retrieved from: http://maps.google.co.za/?utm_campaign=en&utm_medium=ha&utm_source=en-ha-emea-za-bk- gm&utm_term=maps%20south%20africa vi DETAILED MAP OF KATLEHONG TOWNSHIP2 2 Google Maps South Africa. Retrieved from: http://maps.google.co.za/?utm_campaign=en&utm_medium=ha&utm_source=en-ha-emea-za-bk- gm&utm_term=maps%20south%20africa vii MOLELEKI EXTENSION 2 IN KATLEHONG TOWNSHIP3 3 Google Maps South Africa. Retrieved from: http://maps.google.co.za/?utm_campaign=en&utm_medium=ha&utm_source=en-ha-emea-za-bk- gm&utm_term=maps%20south%20africa 1 CHAPTER ONE: THE MOLELEKI EXECUTION, A RADICAL PROBLEM OF UNDERSTANDING Introduction ?A Finishing Off? The Moleleki Execution, Moleleki Extension 2, 9am, Tuesday 7 December 1993 They took us out of the shack and they were kicking us and hitting us. And when we reached the veld?Mr Buthelezi then said we should pray and we prayed?they then untied Sepo?one SDU member gave Sepo a firearm. They then said we should sit down with our backs turned towards them?and then he said to Sepo: shoot your mates and then I heard the sound of gunfire and when the gunfire started everyone of us wanted to shield himself by another one on the ground. And after they stopped firing I heard someone say: finish them off. We were screaming and I was lying folded on top of another man. I was not injured and just kept quiet. And after the word was given: finish them off, I then heard the sound of chopping and I was also chopped, I think four or five times on my head. And after the chopping I heard their footsteps moving away from us. I think they had walked for a distance of about one metre when I heard one of them say: hey there is one of them still alive. And I heard one of them say: take the spear and put the spear through him. I thought that I was being referred to as the one who was still alive and I thought of jumping up but I desisted and decided to just lay there?but then I heard the sound of the spear further to my side. And after they stabbed that person they then walked away? And while I was still there I was lying on top of a certain boy and I saw him lifting his leg time and again. I then untied my hand and untied that boy as well?he could not walk properly. I lifted him up, and walked away with him?And then I ran, I ran and ran and I then became dizzy and it appeared as if I was dreaming?and from then on I don?t know what happened. (The State v. Michael Sonti, 1999, pp. 220-222) Vuyani Tshabalala, breaking down in the courtroom while giving this account of what subsequently came to be known as the ?Katlehong massacre?4, was one of only two young survivors of an atrocity of violence in which seven ANC Youth League (ANCYL) members 4 Referred to in this thesis as the Moleleki execution 2 ranging in age between 14 and 18 and one adult ANC member, 35 year old Alfred Buthelezi, were killed, execution style, on the morning of 7 December 1993 following a night of retaliatory violence. ANCYL leader Isaac Mbitjana Motloung (known as?Wips?) was captured with other ANCYL members but was drowned separately in a nearby stream. The execution took place in the bushveld adjoining the most recently constructed block, Block F, of the still incomplete informal settlement, Moleleki Section Extension 2, recently established on the south east side of Katlehong township on the East Rand in the Witwatersrand region of what is now the Gauteng province. The area of unoccupied land in which the execution took place formed a ?no-man?s? land, a barrier against the adjacent informal settlement of Zonkezizwe, which was by then the almost exclusive political domain of the Inkatha Freedom Party (IFP). The eight executed were among a group of 14 young Moleleki residents who were forcibly abducted during the early hours of the morning of 7 December by adult ANC members, who had organised themselves in a self-defence unit (SDU) operating in Moleleki Extension 2. This SDU was one of many informal defensive structures established in townships on the Witwatersrand following the upsurge of violent political conflict after the deprohibition of political organisations in 1990. Youth League members had initially participated in SDU patrols but after bodies were ?inexplicably? found in the township after night time patrols by adult SDU members, divisions between the SDU and ANC Youth League members deepened to overt hostility. 3 Before the Execution ?Death Has Arrived? Killing of Civic Activist Bulelwa Zwane The execution itself was the culmination of a series of events which occurred during a 24- hour period after 26-year-old Theodora Bulelwa Zwane, local ANC party official and secretary of the civic association in Block F of Moleleki Extension 2, was shot and killed at about 7.30pm as she and a companion walked in a street in Block F of Moleleki Extension 2 (The State v. Michael Sonti, 1999, p. 1766). Bulelwa was claimed as ?friend? and associate by both the ANC Youth League and the SDU in Moleleki. However, the night before Bulelwa?s death, she had attended a meeting in Block F called by Alfred Philemon Buthelezi (who would later die in the execution with his son, Thokazani), to discuss setting up a rival SDU in this block. Witnesses recall Bulelwa questioning the necessity for a ?new? SDU in Block F when an SDU had already been established, which was patrolling the block. Friend Nombuniso Ndwane recalls: ?Bulelwa said or asked if it was necessary to form a new SDU, she asked if this was not going to cause conflict? (TRC, 1998a, p. 558). According to SDU member Zola Sonti, who also attended the meeting: ?It was concluded that because she was a woman she should just remain silent? (TRC, 1998a, p. 289). Bulelwa?s murder, the following evening, sealed her silence forever and precipitated the chain of events, an unravelling of time and order, that would indeed, as she predicted, culminate not only in conflict but in the horror of the execution. As Njebe Ndondolo, Moleleki Extension 2 SDU commander, told SDU ?operator?5 Michael Armoed before taking him to the site of Bulelwa?s body, ?death has arrived? (TRC, 1998a, p. 384). 5 SDU members who carried guns were called ?operators?. 4 It took the police until 8am the following morning to come and collect Bulelwa?s bullet- riddled body, which lay for almost thirteen hours on the streets of Block F. During this thirteen hour period, through the night and the early hours of the morning, a series of events unfolded in the sleepless township that would eventually culminate in the Moleleki execution shortly after the removal of Bulelwa?s corpse by the police. Police photographs of the scene show a young woman lying on her back half-covered by a blanket, spent cartridges scattered around her and a crowd of what SDU member Themba Mtshali estimated to be approximately 100 people still gathered several feet from the body. Themba recalls the mood of the crowd around Bulelwa?s body as ?tense? and remembers that ?the community was discussing crime and violence? (TRC, 1998a, p. 545). Themba was one of several SDU members who arrived at Bulelwa?s body at approximately 9.30pm after being informed of her killing. The SDU members were joined by ANCYL leader 18-year-old Isaac Mbitjana Motloung (known as ?Wips?), who arrived with a group of ANCYL members whom he had informed of Bulelwa?s death. In his prior conversations with ANCYL members as he escorted them to the site of the killing he had already implicated SDU operator Oscar Motlokwa in the killing of Bulelwa (Mokoena, 1995; Moloi, 1995). Also at the scene of Bulelwa?s death were two Block F civic association members, George Samson Mokwena, who had chaired the meeting the previous evening where Bulelwa had been told to keep quiet, and Alfred Buthelezi, who had initiated the meeting to establish a new SDU structure in Block F. Tensions between the groupings were already evident at the site of Bulelwa?s killing, as Alfred Buthelezi told SDU 5 member Temba Mtshali, when he arrived at Bulelwa?s body: ?We don?t want a person wearing a blanket here?6 (TRC, 1998a, p. 544). The Call of Bulelwa?s Death, a Lacuna at the Centre of Power What Bulelwa?s death made explicit was a lacuna at the centre of power in the township, which created the context in which, within 12 hours, the execution could take place. On the one hand there was a complete juridico-institutional absence, as the apartheid state had substantively retreated from violence-racked townships, merely arriving to collect the corpses of this conflict. In the absence of the institutional ordering of the state, the township had sought to establish its own forms of governance through the structures of the African National Congress (ANC), the South African National Civic Organisation (SANCO) and the ANC Youth League (ANCYL). An ostensibly politically ?neutral? SDU was established to defend the township against the violence that ?infested? (TRC, 1998a, p. 11) the trouble-torn townships of the East Rand. However, Moleleki was a new township, only established in November 1992, one year before the execution, and its structures of internal governance remained fragile, fluid, its leadership contested. The modernist vision of the ANCYL and the civic association, which foresaw a representative, juridical political authority, clashed with the biopolitical conception of political authority of the SDU, a conception of the political in which authority is embodied, invested in the person rather than in formal office (Agamben, 2005, p. 82). Tensions rose as the SDU and the ANCYL, allied with adult civic leaders, struggled for the terms on which political authority would operate in Moleleki, clashing around these two visions of the ?political?, which evoked disputations in terms of gender, 6 A reference to the blankets worn by SDU members, perceived to be evidence of their Xhosa ethnic identity. 6 generation and ethnicity and manifested themselves in conflicts around a range of issues, that formed the backdrop to the violence that eventually led to the execution. It was in this lacuna of juridico-political authority and in the environment of political extremis which ensued after Bulelwa?s death that the ?community?, organised not as representative political authority, but simply ?embodied? in gatherings at sites of violence, which became the source of political authority in Moleleki. This political authority, however, in its constant fluidity of embodiment at various sites of death, contained an intrinsic ambiguity and constituted a deeply unstable political referent. Who in fact was the ?community? who could authorise action in response to the violence of that terrible night? Where was it located in the porous boundaries between public and private space evoked as residents gathered on the streets of Moleleki in response to the deaths of that night? Residents Gather at Block C The chaos of the events which were to unfold after Bulelwa?s death evidenced this deep ambiguity at the heart of power and the fluidity of the transformations that took place in this context, marking some citizens as life which deserved to be killed, or tenuously retrieving them from this threshold of death (Agamben, 1998, p. 82). Members of the Moleleki community came out onto the streets of the township throughout the course of the night to gather primarily at the body of Bulelwa in Block F and subsequently at the body of Moleleki Extension 2 deputy SDU commander Malusi Jackson Kiyana (known as Blanko), who was shot dead by ANCYL members in retaliation for Bulelwa?s death. According to civic and youth leader Lethusang Rikaba, the source of authority for this latter killing was a 7 spontaneous gathering of people at approximately 9pm at a Spaza7 shop in Block C of the township, as local residents were alerted to Bulelwa?s death by the cries of her friend and fellow civic member, Alfred Buthelezi. The gathering was addressed by Lethusang: The residents gathered there because of Buthelezi?s scream. They asked me of saying ?why do we hear the scream???That is why I addressed them?. I tried to calm down Mr Buthelezi first so that we?ll be able to hear directly from him about the screaming. That is when Mr Buthelezi, after he?s been calmed down, he said Bulelwa has been killed (TRC, 1998a, p. 685). It was in this context, which was ?not a meeting, a formal meeting, it was because of the call, we met because of the call of Bulelwa?s death? (TRC, 1998a, p. 668) that a group of fifty to seventy residents who were ?emotionally charged ? within those members of the community, the directive came that enough is enough, that we should apprehend these people? (p. 681) and ?bring them to the community? (p. 668). Killing of SDU leader Blanko No alleged perpetrators were in fact ?brought to the community?. In the early hours of the morning, approximately seven hours after Bulelwa?s death, a group of approximately 40 ANCYL members, accompanied by residents of Moleleki Extension 2, some of whom were armed (TRC, 1998a, p. 671), sought to ?hunt down? (The State v. Michael Sonti, 1999, p. 1766) the perpetrators of Bulelwa?s killing, purportedly SDU members operating in Block E of the township. Lethusang Rikaba emphasises: ?As the community, we left, we searched for members of the SDU, we did not find some of them until we arrived at Blanko? (TRC, 1998a, p. 572, own emphasis). 7 Spaza shop [?sp??z?], n, (Business / Commerce) South African slang, a small informal shop in a township, often run from a private house. [from slang, dummy, camouflaged]. (Spaza shop. Retrieved from The Free Dictionary: http://www.thefreedictionary.com/spaza+shop) 8 Approximately seventy or eighty (TRC, 1998a, p. 677) residents gathered outside Blanko?s home in Block E of the township as ANCYL chairperson ?Wips? Motloung and Lethusang Rikaba, as well as several other armed ANCYL members, forcibly entered his house at approximately 2am. They purportedly intended to take Blanko to ?account to the community? at a local meeting place in Moleleki but encountered resistance from Blanko. Lethusang took an AK-47 from a Youth League member accompanying him and fired two shots at Blanko. Two other Youth League members opened fire with 9mm pistols. Blanko died instantly (TRC, 1998a, p. 675). Goods from the Spaza shop that Blanko operated from his home were taken by some younger Youth League members, which when later found in their possession by SDU members provided ?evidence? of their criminality and involvement in Blanko?s death and became the basis for their later abduction and execution. After the Spaza shop was looted, Blanko?s shack was set alight. The remains of his charred body were found in a sitting position on a chair, propped outside against the ruins of his shack. SDU members guarding Bulelwa?s body in the adjacent Block F heard the gunfire in Block E and saw the flames rising from Blanko?s burning shack. They rushed to the scene. ?We found Blanko?s shack burning fiercely. We were quite many in number and we tried to extinguish the fire? (TRC, 1998a, p. 553). SDU members arriving on the scene were informed by Blanko?s neighbour and fellow SDU member Oscar Motlokwa that he had identified ANCYL leader Wips Motloung and Lethusang Rikaba as Blanko?s killers. His identification would later lead to the abduction and killing of Wips Motloung and the attempted abduction of Lethusang Rikaba. 9 ?Those people must be attacked?, retaliation for Blanko?s death The site of Blanko?s death was to create the context in which Katlehong deputy SDU commander Moeketsi William Ramabele (known as ?Sugar?) would issue the instructions that would lead to the forcible abduction of 14 ANCYL members who were then incarcerated in a shack on the edge of the township in Block F. It was in this context too, that Njebe Ndondolo, Moleleki Extension 2 SDU commander, would later issue the final ?order? for the killing of youth in the bushveld adjoining the township. An Order is Issued for Attack by SDU Commander Sugar Ramabele As SDU members congregated at the scene of Blanko?s death to extinguish the flames of his burning shack they were joined by senior SDU member Sugar Ramabele, who told them: ?Those people must be attacked??.We went out hunting them?That decision was taken by Sugar. He had powers?and if a shootout ensues, fine. We went to them through C block, then we met them, then we fought until the early hours of the morning. (TRC, 1998a, p. 153) According to Oscar Motlokwa, ?Now the members of the community even helped with catching some of these people? (TRC, 1998a, p. 157). Abduction of 14 ANCYL members During the following hours, between 3am and 8am, a total of 14 young Moleleki residents, some of whom, like the SDU, had been patrolling the streets of the dark township through the night and some of whom were sleeping together in an ANCYL ?base? in Block A, were abducted by two groups of SDU members who left the site of Blanko?s body in search of the perpetrators of this killing. Most of those abducted and later executed, with the exception of Youth League leader Wips Motloung, had not in fact been involved in the killing of Blanko, 10 as they were either sleeping in the ANCYL base in Block A, or like Vuyani Tshabalala, the survivor whose quote began this chapter, were patrolling in the area. Lethusang Rikaba, who was directly responsible for the shooting of Blanko, was briefly apprehended by SDU members but was released because, according to SDU member Zola Sonti, ?I felt that because the community was still there, we wanted an explanation as to what the origin of this is and where it is headed? (TRC, 1998a, p. 297). Lethusang would never in fact ?account to the community? but would later give his account of the events leading to the execution and his role in the killing of Blanko within the institutional context of the TRC and the courtroom processes, in which five SDU members were finally sentenced to life terms of imprisonment in 1999, six years after the execution took place. ?Parents, What Should We Do If People Are Killing People In This Fashion?? ?Presentation? of the abducted youth to the community Throughout the night and as dawn broke, the abducted youths were openly marched across the township and were taken ?back to the community?, in fact briefly made visible to two large groups of Moleleki residents who remained gathered on the streets of the township throughout the night. One group at the site of Blanko?s killing in Block E had swelled to a gathering of approximately 500 people by 7am, when they were addressed by civic leader ?Machinini?. Most of the youths however were taken to a smaller group of approximately 100 people gathered at the site of Bulelwa?s body in Block F, before being incarcerated from about 5.30am onwards in a shack approximately 100 metres away. The number of youths held in 11 the shack would eventually reach 14 as various groups of armed SDU members arrived with captured youth. Elphina Mugadi, mother of Isaac Mugadi who would die in the execution little more than an hour and a half later, was one of the residents standing at the site of Bulelwa?s body after leaving her home at approximately 6.30am to look for her son who had not returned home the previous evening. A friend had woken her to tell her, ?The children are finished? (Mugadi, 1995, p. 1). She later told the Truth and Reconciliation Commission that ?I understand by the ?term? finished that they had been killed? (Mugadi, 1998). As she stood at Bulelwa?s body among approximately 100 men, women and children, Elphina saw a group of about 20 men, among them SDU members armed with pangas8 and spears, approaching the crowd gathered around Bulelwa?s body. Walking unarmed with the group of men were a number of young ANCYL members, including her 17-year-old son Isaac Mugadi, 18-year-old Mile Simelana, 16-year-old Ditaba Mthembu and 18-year-old Buti Hlatshwayo, who would all soon die in the execution. Fourteen-year-old Jacob Moloi and 15- year-old Charles Matebang Mokoena were also among this group forcibly brought to the site of Bulelwa?s death but were released prior to the execution. SDU leader Njebe Ndondolo rhetorically addressed the gathered crowd, publicly accusing the captured youth of the killing of Bulelwa and Blanko: ?Parents what should we do, you of Moleleki, if people are killing people in this manner or fashion?? According to Elphina, ?nobody answered? (The State v. Michael Sonti, 1999, p. 1088). ANCYL member Vuyani Tshabalala, one of only two survivors of the execution who was also ?presented? at the site of 8 A type of axe. 12 Bulelwa?s body, elucidates: ?We were also not allowed to speak?9 (Tshabalala, 1995, p. 3). The youths were then escorted to a shack 100 metres away from Bulelwa?s body where they were tied to each other with their hands bound behind their backs and their feet bound together. ?Hey, You Boys, What Have You Done?? Fourteen Young Boys are Incarcerated in a Shack in Block F The shack, 100 metres from Bulelwa?s body, was open as it did not have doors nor window glass installed yet, which meant that the boys tied up inside in it were visible to residents gathered around Bulelwa?s body, ?people were arriving and peeping through the [opening for] windows and asking: ?Hey, you boys, what have you done??? (TRC, 1998a, p. 693). Eighteen-year-old Jabulani Nxumalo managed to untie himself and was beginning to untie some of the youths incarcerated with him when he was betrayed by a community onlooker: ?One woman stood by the door, came towards the door and said: ?here, the thugs are trying to escape?? (TRC, 1998a, p. 693). Other members of the Moleleki community who came to the shack where the boys were held were the parents of two children incarcerated there, Tina Mootsi, mother of 15-year-old Itumeleng Edward Mootsi, and Alfred Buthelezi, whose attempts to intervene on behalf of Thokazani, his 14-year-old son, would lead to his own capture and execution. Jabulani Nxumalo, who was released immediately prior to the execution, witnessed Alfred Buthelezi?s attempts to intervene on his son?s behalf: Buthelezi?wanted to know what was going on with his child. There were allegations that he was an Inkatha member and two people in the group grabbed [him]?one of 9 The statement to the police was recorded in Afrikaans. The original read: Ons was ook nie toegelaat om te praat nie. 13 the people chopped Buthelezi on his head with an axe?then Buthelezi and his son were tied together by the neck with a strong rope10 (Nxumalo, 1995, p. 3). Tina Mootsi arrived at the shack where the boys were incarcerated and asked her son what had happened. He replied, ?Mama, I?ve done nothing wrong? (T. Mootsi, personal interview, 2001). However, Tina was not allowed to talk further: his men when he saw me he inquired what are these women doing here?I tried to talk to him, they pushed him away that I don?t talk to him. I left to go and phone his father thinking that as a man they will allow him to talk to him. (Mootsi, 1997, pp. 20- 21) But it was too late. As Tina left she was told she would ?find the children in the veld? (T. Mootsi, personal interview, 2001). Later that day she found her son?s decapitated body among the eight executed. At approximately 8am the Internal Stability Division (ISD)11 arrived to collect the body of Bulelwa and then moved on to collect the body of SDU commander Blanko. While Moleleki residents stood watching the collection of bodies at these two sites, they knew of the incarcerated youths held just metres way, but as Elphina explained, ?some of the parents were afraid [of the police]? (The State v. Michael Sonti, 1999, p. 1094). However, SDU members became concerned that the youth would be discovered by the ISD, so they were moved to a second shack slightly further from Bulelwa?s body. This shack was also incompletely built and the inside was visible to residents. 10 Buthelezi?wou keer wat gaande is rakende sy kind. Daar was bewerings dat hy n Inkatha lid is en het twee persone in die groep vasgegryp waarop een van die persone vir Buthelezi met n byl oor sy hoor gekap het... Daarop was Buthelezi en sy seun met n sterk tou aanmekaar se nekke vasgemaak. 11 The ISD, a specialised police unit, was established in 1992 and was tasked by the outgoing apartheid state with policing the rising levels of politicised violence which accompanied South Africa?s negotiation process between 1990 and 1994. 14 While the youth cowered in the shack, SDU commander Njebe Ndondolo arrived with a list of names of the youth who were to be executed, comprising members of the ANCYL who patrolled the township rather those instrumentally involved in Blanko?s killing. Jabulani Nxumalo was released on this threshold of death when SDU member Michael Armoed looked into the shack and recognised him as his son?s ?playmate? (TRC, 1998a, p. 386). Fourteen-year-old Jacob Moloi and 15-year-old Charles Matebang Mokoena were also released, apparently because their names did not appear on Njebe Ndondolo?s ?list? (Mokoena, 1995, p. 3). Gathering Addressed by Civic Leader Machinini As the crowd of approximately 100 people remained gathered around Bulelwa?s body in Block F throughout the night, another group of people gathered at the site of Blanko?s killing in Block E from 2am onwards. This group of concerned residents grew through the night. By 7am when the gathering was addressed by the civic leader, Machinini, the crowd had swelled to approximately 500 people congregated at the site of Blanko?s still uncollected body. Oscar Motlokwa explains: ?It was a community meeting because the vice chairperson of the zonal civic addressed them, after he arrived many people came? (TRC, 1998a, p. 178). ?He was addressing the issue of Bulelwa and Blanko. He was addressing crime in general in Moleleki section? (TRC, 1998a, p. 179). ?He [Machinini] was addressing the meeting as the leader, he did not take out?an order [to kill the youths] at the meeting? (TRC, 1998a, p. 157). Nevertheless the ?confession? of Alfred Buthelezi?s son, Thokozani, at the meeting regarding the ANCYL?s purported involvement in Bulelwa and Blanko?s deaths, was critical in creating the context in which, according to Oscar, ?the community lost control and?said these children must be killed? (TRC, 1998a, p. 154). After a visibly beaten Thokozani was brought 15 into the meeting, following his violent interrogation by Njebe Ndondolo, he publically cried out to his father Alfred who stood amongst the gathered crowd ?Dad, tell me what you ordered me to do, you sent me? (TRC, 1998a, p. 155). It was in this environment that Moleleki Extension SDU commander, Njebe Ndondolo, acting on the instructions of Katlehong SDU commander, Mzaapindile Ntsingola (known as Manyala12), ?during the process of the meeting? ?but not in the meeting?, ?called us aside as members of the SDU. It is there where he took out an order? (TRC, 1998a, p. 178). As residents flowed in and out of Blanko?s small backyard, SDU members congregated in a small group. ?Ntjebe called the members of the SDU aside and he said Manyala commanded him that these people must be killed and then we were going to fetch them from a shack and then we were going to kill them in an open veld?Ntjebe13 and Manyala14 and Majola15 ordered the death of these people, these criminals? (TRC, 1998a, p. 156, own emphasis). Soon after this order was issued, SDU members returned to the shack where most of the ANCYL members were now incarcerated, on the edge of Block F adjoining the bushveld on the boundary of the township. By this time residents had retreated in fear. ?When we drove them to the veld, there were no people around? (TRC, 1998a, p. 407) and the group were compelled to take their last walk 12 Variously meaning, badness, nastiness, worthlessness, or claw as in clawing at an opponent. (Manyala. 2006. In Merriam-Webster Online dictionary. Retrieved from: http://www.websters-online- dictionary.org/translation/zulu/manyala). Elphina Mugadi describes her encounter with the SDU commander Mzaapindile during the events leading up to the execution where he used the term manyala as both an adjective to describe himself and a verb to describ his actions, ?This man said: ?I am Manyala and everything that I do [are] bad things?...[he]used the word that I am Manyala and what I do is manyala, literally speaking my lord manyala would mean doing just bad things, anything, not caring about what someone is doing?(The State v. Michael Sonti, 1999, p. 1089). 13 Njebe Ndondolo, Moleleki Extension 2 commander. 14 Mzaapindile Ntsingola, Katlehong township SDU commander. 15 Nkosi Kona Majola, Katlehong township deputy SDU commander. 16 accompanied only by their captors, who beat and kicked them as they stumbled along, tied together, while parents, residents and ANCYL members watched furtively from the tall grass and empty shacks adjoining the veld. At the site of the execution those captured were unbound from each other and were told to kneel, as if in prayer, facing away from their executioners. Among the captured was civic leader Tshepo Baloyi. The rope around his neck was untied and he was ordered to start shooting at the abductees. He was later ?released? by SDU commander Njebe Ndondolo and fled the scene to report the execution to a local violence monitoring organisation, Peace Action. The murder charges eventually brought by the Attorney General in 1997 listed the eight dead as: ? Alfred Philemon Buthelezi @ [aka]Alie, 35 years old, residing at Moleleki Extension 2 ? Itumeleng Edward Mootsi, 15 years old, residing at 8183 Moleleki ? Thokozani Buthelezi, 14 years old, residing at Moleleki Extension 2 ? Lucas Hlatswayo @ Buti, 18 years old, residing at 7208 Moleleki ? Isaac Mugadi, 17 years old, residing at 7400 Moleleki ? Ditaba Joseph Mthembu, 16 years old, residing at 6938 Moleleki ? Peter Mavuso Mdishwa, 15 years old, residing at 8155 Moleleki ? Mile Simon Semela (sic), 18 years old, residing at 7215 Moleleki. (The State v. Michael Sonti, 1997, p. 1) The first photograph of the crime scene shows approximately seven policemen standing talking to each other over a number of heaped bodies, lying in the tall bushveld grass. It had taken police more than 24 hours to collect the corpses of the executed boys, by which time their already mutilated bodies had been further maimed by roaming dogs who ate their flesh as they lay piled on top of each other, rapidly decomposing in the summer heat. One of the police?s first actions at the crime scene was to bring drivers from a company that delivered paraffin in Moleleki, who had been repeatedly robbed during deliveries, to ?identify? these now rotting corpses as the ?perpetrators? of the robberies. 17 Close up photographs show a group of young boys still bound together, lying crumpled together at various angles in the tall grass. One boy lies face forwards, his identity obscured, another boy lies curled in a foetal position, another lies on his back, his knees drawn up and his arms twisted upwards, one fist clenched, the other hand reaching toward the sky, other boys lie disturbingly vulnerable, flung upon their backs, their faces a parody of sleep. In the horrible intimacy of a collective death lies another boy, cradled against the shoulder of the headless body of 15-year-old Edward Mootsi, who had been decapitated, his head smashed into several pieces and scattered at some distance from his torso (Dwane-Alpman, 1998, p. 1). Alfred Buthelezi lies separately, a thick rope still tied around his neck, his stomach gaping open, his intestines falling out. In his judgement at the conclusion of the court trial of the SDU members charged with involvement in the execution, Justice Boruchowitz noted the extremity of violence evidenced by these frozen moments of horror captured in the photographs of the crime scene: The depth of the brutality involved in the massacre is evident from the photographs that were handed in as a bundle marked exhibit D2. All of the deceased had been shot and hacked to death and several were disembowelled. One was decapitated. (The State v. Michael Sonti, 1999, p. 1768) A Radical Problem of Understanding How are we to make sense of this extraordinary horror? This litany of death, its ?excess? and it?s heartbreaking wastage of human life? How can its terrible enigma be understood? The Moleleki execution, despite its extraordinary horror, was not an isolated incident of atrocity. During the four years between the opening up of the South African political process in 1990 and the country?s first popular elections in 1994, violent political conflict, frequently 18 profoundly transgressive in character, claimed the lives of approximately 16 000 people. During the preceding five years, between1984-1989, at the height of national uprisings against the apartheid state, no more than a quarter of this number died as a result of political conflict (Coleman, 1998, p. 243). The Moleleki execution itself was one of 35 ?massacres?16 recorded on the East Rand between 1990 and 1994 (Shaw, 1997, p. 33), where approximately 3 000 people lost their lives over a period of four years after the opening up of the South African political process in 1990 (Shaw, 1997, p. 30). The majority of these deaths, including the deaths in Moleleki Extension 2 on the boundary of the township of Katlehong, took place within the Kathorus17 sub-region of the East Rand. Not only did the violence of the 1990s claim more lives than the struggle against the apartheid state during the 1980s but there was a dramatic qualitative change in the forms of violence during this period, forms which appeared both transgressive and excessive, shattering normative boundaries between public and private spaces, including within its orbit women, children, the elderly, the unarmed, the defenceless and the ?innocent?. People were attacked in public places of recreation, transport and worship. Gunmen opened fire on train 16 There were some definitional debates regarding what constituted a ?massacre? in the post-1990 context, with the Human Sciences Research Council defining a massacre as ?an incident in which five or more people are killed in a single attack perpetrated by one group of attackers only? (Minnaar, Keith & Pretorius, 1994). Shaw, in his analysis of violence on the East Rand uses this definition, describing a massacre as ?the killing of five or more people in a single incident? (Shaw, 1997, pp. 32-33). The Human Rights Committee (HRC), another key violence monitoring organisation, simply defined a massacre as ?an incident in which 10 or more people were killed? (Coleman, 1998, p. 262). Clearly such quantitative definitions are infinitely contestable; however, with regard to the Moleleki killings, both the TRC and the court unquestioningly used the term massacre, which was the manner in which residents of Moleleki themselves had defined the incident. 17 Kathorus refers to a collection of three townships, Katlehong, Thokoza and Vosloorus, located in the former sub-region of the East Rand adjacent to the greater metropolitan area of Johannesburg, which have now been incorporated into the municipality of Ekurhuleni, which means ?Place of Peace?. Ekurhuleni is one of South Africa?s most highly industrialised areas and was the site of some of the most extraordinary levels of politicised violence during the four year period preceding South Africa?s first democratic elections. 19 and taxi commuters, people drinking in shebeens18 or gathered at meetings. Residents were violated in the ?ordinary? spaces of their everyday lives, in their homes as they slept, ate or watched television, in the playgrounds of schools, in the neighbourhood street. The violence perpetrated was often extreme in its bodily mutilations. South Africans were shot, stabbed, strangled and burned, their skulls were crushed, babies were hacked to death on mothers? backs, pregnant women were disembowelled (Moser & Clark, 2001, p. 209). Within the South African context, the politicised violence of the pre-election period appeared to constitute what social analyst Andre du Toit framed at the time as ?a radical problem of understanding? (A. du Toit, 1993, p. 2). Contemporary social analysts struggled to articulate the meanings of the violence, to grasp its significations, its rationalities, its forms of intention. Why were people turning to violence at the very moment when a democratic political process appeared for the first time imminent; why were the forms that violence was taking so profoundly transgressive, so spectacularly ?excessive?? The four years between 1990 and 1994 were marked by an apparent paradox at the heart of power, a profound dichotomy between the co-operation and compromise taking place at a national level in the negotiation process, which began in 1990, and the increasingly internalised conflict ravaging many townships. The brutality and desperation of local politics often seemed far removed from the negotiations being held between national political parties. The neat binaries of conflict between the apartheid state and the political opposition now fractured into a diffusion of conflicts between a variety of groupings (initially most visibly 18 Shebeen [she?been ],n, an unlicensed drinking establishment, especially in Ireland, Scotland, and South Africa. (Shebeen. Retrieved from The Free Dictionary: http://www.thefreedictionary.com/shebeen) 20 between the recently de-prohibited ANC and the Inkatha Freedom Party (IFP), which entered the national arena as a formal political party for the first time after 1990), but as the conflict continued over the next four years these early political oppositions diffused and fractured further, the boundaries of conflict becoming more parochial, unstable, unpredictable. What was the nature of the power that emerged in this context of democratic transition? What was the nature of the power in Moleleki where the lines of conflict fractured the boundaries of political party opposition, pitting ANC members against each other and producing the horror of the execution and the mutilation of children? A Mutable Truth This thesis addresses itself to these struggles of comprehension, not in pursuit of an original immutable truth, a secret meaning lying buried with the corpses of the Moleleki atrocity, but instead to attempt to understand the ways in which meaning was made of the execution. Hence it becomes possible to interrogate the construction of knowledge about the execution in terms of a conceptual paradigm in which violence remained an implacable exception to the ?proper? functioning of juridical power. This conception was an understanding of power and of the political, which ?takes law as its model and its code? (Agamben, 1998, p. 5). It was a notion of power where the site of the material enactment of power and the representation of power are conceptually distinct, making possible an ?idealised? juridical realm of power, separate from its site of violent instantiation (Feldman, 1991, p. 235). The interrogation of the construction of knowledge about the Moleleki execution within this juridical conception of power makes it possible to make explicit this construction of knowledge about the execution in terms of a conceptualisation of power, which made possible only a limited field of knowledge about power. This conceptualisation of power, critically, could only conceive of violence in an instrumental relation to a juridical power, as a strategic tool used in the pursuit of ?rationally? conceptualised and articulated political objectives. The violence of the 1990s 21 could not be understood in such terms, its objectives were opaque, its violence disproportionate and its targets random. Therefore the interrogation of the construction of knowledge about the execution, which produced and delimited the ?truth? that could be known about it, makes possible a wider interrogation of the ?truth? about power. In the unpacking of this construction of ?truth? about the execution within a particular ?truth? about power, it becomes possible to ?re-read? the execution in terms of a significantly different conceptualisation of power, a conceptualisation of power which incorporates not only the juridical, but also what Foucault first conceived of as the ?biopolitical?, ?the concrete ways in which power penetrates subjects? very bodies and forms of life? (Foucault cited in Agamben, 1998, p. 5). The biopolitical nature of power has, nonetheless, as Agamben argues, remained concealed and implicit, as was evidenced in the South African context by the struggles of comprehension around the violent conflicts of the 1990s, conflicts which made the biopolitical rather than the juridical nature of power so explicit. Agamben concludes: ?The ?enigmas? that our century has proposed to historical reason and that remain with us (Nazism is only the most disquieting among them) will be solved only on the terrain ? biopolitics ? on which they were formed? (Agamben, 1998, p. 4). The Articulation of Biopolitical and Juridico-Institutional Power, the ?Hidden Nucleus? of Sovereign Power While the ?enigmas? posed to historical reason by events such as the Moleleki execution can only be grasped through a biopolitical understanding of power, this in itself is not enough. In order to function as a whole the juridico-political ?machine? requires an articulation between the two faces of power, the juridico-institutional and the biopolitical, which occurs through what Agamben (1998) calls the structure of exception, which is the mechanism of the relation 22 between ?life? and law. If politics is in fact about the ways in which power penetrates subjects? bodies and lives, and is not simply a juridical politics, then life is in fact the realm of the political and this point of articulation between the biopolitical and juridico-institutional is about establishing a relationship between the juridico-institutional, the sphere we conventionally assume is the domain of the political, and biopolitical life. This point of articulation between juridical and biopolitical power, between life (biopolitical power) and law (juridical power), is also a point of irrevocable ambiguity where what is inside or outside the juridico-political order remains unclear, the boundaries between the two dissolving into each other and the edges of the juridico-institutional order always ambiguous and fragile, neither completely excluding or including the biopolitical life that forms its border. It is this uncertain articulation between juridico-institutional and biopolitical power, through the relation of exception, that Agamben argues is the ?hidden nucleus? (Agamben, 1998, p. 6) of sovereign power. This realm of exception on the boundaries of the juridico- political order is also a space of anomie, a realm of what Benjamin calls ?pure violence?, violence without any relation to law (Benjamin cited in Agamben, 2005, p. 61), violence whose very ?excess?, its normative violations, its ?exceeding? of law and of norm, are in fact the mechanism of the relation of sovereign power. Critical though the articulation between biopolitical and juridico-institutional power through the relation of exception may be, particularly for an exploration of violence, it appears as a ?vanishing point? in the ?entire Western reflection on power? (Agamben, 1998, p. 6). The two faces of power, whose articulation is, for Agamben, the real locus of sovereignty, are instead conflated and obscured within the modern juridical concept of sovereignty in which biopolitical power remains hidden and implicit. 23 This thesis takes its cue from Agamben?s central theoretical problematic, namely to investigate, through the event of the Moleleki execution, this intersection of biopolitical and juridico-institutional power, this ambiguous ?how? of the exception, the ?how? of sovereign power in the exception over two critical historical conjunctures, the moment immediately prior to South Africa?s first popular elections, a significant period of biopolitical power, and the moment of post-apartheid political reconstruction, a period of juridico-institutional power that sought to usurp and contain the biopolitics of the past. The Moleleki execution, an expression of biopolitical power The execution itself was an unequivocal expression of biopolitical power, an inscription of sovereign violence on the bodies of Moleleki citizens that took place in a moment of national juridico-political exception, of suspension of the juridico-political order during the negotiation process, a moment of exception, which signalled, not a ?paradox? in the ?dichotomy? of violence and negotiations, but a condition of exception in which, as Schmitt argues, ?Order must be established for juridical order to make sense? (Agamben, 1998, p. 16). It was this condition of exception, in which the Moleleki execution took place, which created the very conditions of possibility for the juridico-political order of post-apartheid South Africa. At operation in the struggles of juridico-political ordering in Moleleki township during the year after it was established was what Schmitt defines as ?the essence of state sovereignty? (p. 16), namely the sovereign monopoly over the decision on the exception, which makes possible ?the creation and definition of the very space in which the juridico- political order can have validity? (p. 19, own emphasis). If, as Schmitt argues, the essence of sovereign authority is the decision on the exception, it was in the decision on this ambiguous margin, which attempts to trace a threshold, a 24 boundary between inside and outside the juridico-political order, that formed the frontier of conflict between the ANCYL and the SDU in the township of Moleleki Extension 2. The struggle around these boundaries of inclusion and exclusion, boundaries inscribed in territory and bodies, and the struggle to fix these boundaries against the irreducible ambiguity of the exception, shaped the conflict which unfolded in the year after Moleleki Extension 2 was established and finally materialised in the horror of the execution. However, as Agamben argues, the exception, is in fact, ?unlocalizable? ?When our age tried to grant the unlocalizable a permanent and visible localization, the result was the concentration camp? (Agamben, 1998, p. 20). Critically, in Moleleki, which was established in 1992 as a political party homogeneity, an ?ANC township?, the localisation of the exception became not only about a struggle to tie land to juridical order but, as the conflict intensified, became a struggle around the transformation of ?life? itself, if the political is, as Agamben argues, not about a supplement or an attribute to life, but is about life itself. Thus the struggle in Moleleki to define the boundaries of the political community also concerned a struggle around the right of the sovereign to define what was political life with the rights to the protection of this political community, and what was ?bare life?, ?simple natural life?, which falls outside the political community. It is the ambiguous figure of ?homo sacer? (sacred life), ?who may be killed and yet not sacrificed? (Agamben, 1998, p. 8) and who is ?included in the juridical order...solely in the form of its exclusion (that is, of its capacity to be killed)? (Agamben, 1998, p. 8), that embodies the relation of exception that defines the boundaries of the political community. The attempt to irrevocably define the boundaries of the political community in Moleleki, and grant the exception at the boundaries of the political community a permanent and visible location, led ultimately to the atrocity of the execution 25 and the localisation of the exception in the bodies of the Moleleki executed, produced as homo sacer, biopolitical artifacts of sovereign power. A Juridico-Political Adjudication of the Moleleki Execution The 1993 Moleleki execution provides a unique opportunity to explore the intersection of biopolitical and juridico-institutional power as the biopolitical power of the execution became the object of investigation within two separate post-apartheid institutional contexts, the South African Truth and Reconciliation Commission (TRC) and the South African court, in the five years following the country?s first popular elections in 1994. In 1997, four years after the execution had taken place, murder and kidnapping charges were brought by the state against seven Moleleki SDU members allegedly involved in the execution. A year later, after the establishment of the TRC, 13 SDU members applied to the Commission for amnesty for their involvement in the execution on the grounds of political motive. The legal proceedings against the seven facing charges compiled by the Attorney General were suspended while the amnesty application of the SDU members was being heard. However the SDU members were eventually denied amnesty on the grounds that the execution had not been politically motivated, but was a criminally motivated act of revenge for the killing of Moleleki Extension 2 SDU leader Malusi Jackson Keyana in the hours preceding the execution. The legal process around the Moleleki execution then continued in court, with seven SDU members standing trial for their part in the execution. In August 1998, five of the accused 26 were found guilty of participation in the execution on the grounds of a ?common purpose? to ?associate? with a ?common plan? to commit the killings and were given life terms of imprisonment. These processes of juridico-political inscription around the execution reflected another critical struggle of exception, the struggle to establish a relation between anomie and law, to inscribe within the post-apartheid juridico-political order this anomic violence through a relation of exception, ultimately to capture ?pure violence as the extreme political object, as the ?thing? of politics?, a struggle which, Agamben argues, ?is as decisive for Western politics as the struggle for ?pure being? of Western metaphysics? (Agamben, 2005, pp. 59- 60). However, these processes of inscription, both within the TRC and the courtroom proceedings, were structured by an explicitly instrumental understanding of power, an instrumentality which sought to ?capture? violence by making it an instrumental object, a tool of juridico- institutional power, a project which was however confounded by the ambiguity of the relation of exception, through which the juridico-political order implicitly, attempted to effect a relation between juridico-institutional power and the anomic violence of the execution. Instead, the violence of the execution and the adjudication of its biopolitical power by juridical power made explicit a fiction at the heart of the juridico-institutional system, namely, the mythological separation of the origin of power from its effects (Feldman, 1991), which makes possible a conception of an instrumental relation between the cause (of power) that is external to the effects of power. 27 The juridico-institutional order is premised on establishing these instrumental relations between cause and effect. Because they are grounded in a ?mythological? separation between the origin and effect of power and the relation between the two is not merely a logical operation, i.e. an ?internal nexus that allows one to be derived immediately from the other? (Agamben, 2005, p. 40), the tracing of this relation requires in each instance, a practical process. Thus the juridico-institutional order provides, through a process of institutional investigation, an institutional guarantee of the reality of this relation between cause and effect, where the biopolitical act is merely a ?derivative symptom? (Feldman, 1991, p. 3) of an originary source of power. Implicit in these assumed instrumental relations of cause and effect, which the juridico- institutional order attempts to practically establish and adjudicate, is, however, a wholly different and more complex relation, the relation of exception, the point of articulation between biopolitical and juridico-institutional power, a point of infinite mutability and enormous ambiguity. The processes of juridico-political inscription around the Moleleki execution evidenced these systematic institutional processes, which sought and struggled to trace this relation between cause and effect through a relation of exception, when such a relation of ambiguity could in fact only render an inevitably mutable conclusion. This process of inscription began from the moment the bodies of the Moleleki dead were first laid on the tables of mortuary pathologists where they were inscribed in a medical hermeneutics. This medical hermeneutics attempted to translate these wounds into a linguistic narrative of cause and effect, through the reading of the traumata of the body. This enabled the pathologist to ?ventriloquise? for the mute body in order to produce a medico-legal conclusion on the ?cause of death? (Pugliese, 2002, 369). 28 These ?causes?, however, would later be extensively contested in the court room processes around the Moleleki execution. The Truth and Reconciliation Commission and the Moleleki execution, an act associated with a political objective The subsequent processes of the TRC similarly sought to inscribe the Moleleki execution in a relation of instrumentality as the Commission, established in 1996, sought to inaugurate and institutionalise a new juridico-political order in South Africa. The Commission endeavoured to institutionally effect an instrumental relation between acts associated with a political objective, in order to conjure a relation between a juridically defined political motive (essentially a party political motive), and the bloody violence which had been committed in strategic pursuit of these ?political? objectives during the conflicts of the past. In the institutional evocation and adjudication of this instrumental relation, the TRC sought, paradoxically, to sever this relation, to make of this relation an exception, explicitly constituting this violence of the ?past? in an instrumental relation to the ?politics? of the ?past? as a strategic tool used by rational actors in pursuit of ?proportional? ends. However, this relation implicitly indicated a far more complex relation with this past of violent biopolitical power. This was a past that would in fact remain ever present, as an exception that could make possible the constitution and continuation of the present political, critically, a juridico-political present, a relation that is ?cosubstantial?, as Agamben argues, ?with Western politics? (Agamben, 1998, p. 7). If the relation between cause and effect, origin and source of power, is in fact ?mythological? and a practical activity is required to establish this relation, the processes of the TRC and the courtroom each adjudicated the actions of the protagonists of the Moleleki execution on different terms; however, both established the relation between cause and effect through an 29 adjudication of the juridical subjecthood of the Moleleki protagonists, specifically an interrogation of their ?intent? to carry out the actions for which they stood before the juridico-institutional apparatus of the post-apartheid state. The nature of this ?intent? appeared significantly different within the two institutional contexts, namely, the assertion of a political motive in the context of the TRC and a refutation of criminal intent or mens rea in the legal proceedings which followed. However, while the processes appeared different, they were in fact critically co-ordinated, together forming the boundaries of post-apartheid power, for ?the outlaw, the Friedlos, or the convict, [is] historically the symbol of the outside upon whose body and life the boundaries of the political community could be built? (Hansen & Stepputat, 2005, p. 15). In the context of the TRC, the grounds of the adjudication of the Moleleki execution were an interrogation of the political motive of the protagonists of the execution in terms of a conception of the political in which the act, the gross violation of human rights, was merely a derivative symptom of this political motive, an act associated with a political objective, which required an instrumental legitimation of power. In this juridical conception of the political, the strategic interests of the political party form the basis of the political motive, and derived logically from this is the political intent of the subject, whose actions, when adjudicated, merely represent a reflection of this ?external? origin of power. The Moleleki protagonists? failure to demonstrate this instrumental relation between their actions and a juridically conceived political motive, led the TRC to conclude: On the evidence as a whole, we are satisfied that members of the SDU and the ANCYL, who both belonged to the ANC, were embroiled in a conflict over control of Moleleki section. They accused one another of being involved in criminal activities under the pretext of protecting the community. Having regard to the motive for the massacre, the context in which it occurred, in particular, the fact that the attack was not directed at a political opponent, we are 30 satisfied that the killing of the deceased was not an act associated with a political objective as contemplated in the Act. (TRC, 1998c, own emphasis) Thus the Moleleki execution failed the test of political exceptionality, which could have released the Moleleki protagonists from criminal liability for their actions. In their assertions of the ?spontaneity? of their actions and the ?criminality? of their opponents, they were unable to articulate an instrumental relation between their actions and a prior juridically- conceived political intent. Critically, they failed the test of exception because in the paradigm of juridico-political power in terms of which the TRC conducted its processes, ?the political? was the party political. SDU members? evocation of a notion of ?community? in whose defence they purportedly acted, and their articulation of the threat to this ?community? as the figure of the criminal, rather than the party political opponent, meant that the biopolitical nature of power and the sovereign conflict over the territory of Moleleki remained hidden and implicit. This was, however, a conflict, which as Schmitt argues, creates the very conditions of possibility for the juridico-political order (Schmitt cited in Agamben, 1998). However, this boundary of criminality, which, as Buur (2006) has argued, gives voice to polyvalent concerns of social disorder (p. 2), articulated by both the SDU and ANCYL in the proceedings of the TRC and the court trial, was actually an attempt to articulate and materially instantiate the boundaries of sovereign power in Moleleki. The body of the criminal, as the Comaroffs have argued in relation to the post-apartheid state, becomes the ?alibi? against which the integrity of the political ?community? can be substantiated (Comaroff & Comaroff, 2004, p. 4). Inherent in this one decision on the Moleleki execution was an ongoing process of definition of power and the construction of knowledge about power in the immediate post-apartheid 31 period, which sought to circumscribe the boundaries of the post-apartheid political as the juridical political, but which remained implicitly co-ordinated with the biopolitical. Therefore, the struggle of articulation between life and law remained hidden, law?s fragile hold on life assumed, its sovereign violences once again ?inexplicable? in this new context of constitutional power. What then was the nature of this act that took place in a context of exception, of juridical suspension, ?the nature of the human praxis that is wholly delivered over to a juridical void?? (Agamben, 2005, p. 49). The process of adjudication of this human praxis in the context of the TRC had concluded that this action could not be defined in terms of the exception of political violence, which could have exempted the protagonists from legal accountability. How then could its violence now be properly adjudicated? The trial of the Moleleki accused, an association with a common purpose The subsequent legal adjudication of the actions of the Moleleki accused revealed the complexity of the articulation of life to law, the difficulties of welding norm and reality together through a relation of exception. The Moleleki execution exposed the difficulties of the application of the juridical norm established by the processes of the TRC in a trial which attempted to apply this norm to a particular case. As each case ?always involves a plurality of subjects? that requires the pronunciation of a sentence which institutionally guarantees law?s reference to life (Agamben, 2005, p. 40). This struggle to articulate life to law was revealed in particular with regard to the actions of the Moleleki protagonists, which ?betray[ed] the impossibility?of clearly defining the legal consequences? (Agamben, 2005, p. 49) of those acts committed during a context of 32 exception, which Agamben argues, is in fact the context of the political. During the final decade of apartheid this difficulty of welding the law to life became increasingly explicit as the momentum of political resistance to the state rapidly escalated and the apartheid state intensified its efforts to use the law to adjudicate ?the political?, more correctly, to foreclose ?the political? by criminalising it. As Parker argues: Where the threat to the ruler?s law and order is collectively based, it is society itself that needs to be deterred. Western criminal law is inadequate to the task of subjugating a people, because it demands responsibility to be individually determined before the guilty can be punished. Punishing individuals, however, deters individuals, not societies, which is why the security needs of the colonial state everywhere corrupt the values of the rule of law. (Parker, 1996, pp. 101-102) The difficulty of the articulation of this political to law, within a relation of instrumental causality, by providing legal evidence of a causal connection between the death of each of the victims and the actions of each of the accused (a causal requirement usually necessary for criminal liability for murder) was particularly acute in relation to the Moleleki execution as the only ?witnesses? to the actual acts of murder involved in the execution were the young survivors Vuyani Tshabalala and Albert Mangane, who both had their backs towards the accused and hence could not provide reliable identificatory evidence. This meant that Judge Boruchowitz abandoned the attempt to establish this relation of causality in favour of the application of the ?doctrine of common purpose?, which required evidence merely of ?an association? with a ?common purpose? to commit a crime in order to secure a conviction. He found therefore that: there is some evidence that certain of the accused contributed causally to the deaths of the deceased, but that evidence is insufficient to render all of the accused criminally responsible. For that reason the state relies heavily on the doctrine of common purpose as enunciated in S v Mgedezi and others 1989 1 SA 687A at 705F-706B. (The State v. Michael Sonti, 1999, p. 1773) 33 The S v Mgedzi judgement was essentially an elaboration of what became an infamous court judgement, in which six people, who later became internationally known as the ?Sharpeville Six?, who formed part of a large crowd that had marched to the home of the deputy mayor of Lekoa in Sharpeville township19 in 1984 during uprisings in the area at the time, were convicted and eventually sentenced to death in terms of the doctrine of ?common purpose?. There was in fact no evidence linking the accused to the deputy mayor?s death, as Judge Botha of the South African appellate division himself stated ? ?It has not been proved in the case of any of the six accused convicted of murder that their conduct had contributed causally to the death of the deceased? (Parker, 1996, pp. 96-97) ? but the decision was upheld by the Appellate Division20 of the South African courts, which in the face of rising incidents of crowd violence in the context of political resistance to the state simply abandoned the notion of causation in favour of the concept of an ?active association? with a common purpose to commit a crime. In the case of the Moleleki execution, as in many other instances of collective political violence which were adjudicated by the South African courts, the impossibility of applying the critical precepts of Western law, namely a relation of individual causality to a particular crime in the context of the exception, the context of the political, was evidenced in the acknowledgement by Justice Boruchowitz at the conclusion of the trial of SDU members that ?there is evidence that not all offenders have been brought to book? (Sindane, 1999). 19 In the Vaal region of the former Witwatersrand. 20 The Appellate Division, which was renamed the Supreme Court of Appeal of South Africa in 1996, was the highest court of appeal in South Africa at the time of the Sharpeville judgement. (Retrieved from Supreme Court of Appeal of South Africa: http://www.justice.gov.za/sca/index.htm) 34 Although not all the offenders involved in the Moleleki execution were brought to trial and the principle of a causal relation had effectively been suspended through the use of the doctrine of common purpose, in order to make it possible to weld law to life, the final conviction of five of the Moleleki accused (however imperfectly achieved) made it possible for Justice Boruchowitz to publicly reiterate the norm of law: ?This sentence will send a message to the perpetrators to discontinue committing careless, savage and brutal actions? (Sindane, 1999), a ?message? that sought to re-affirm law?s hold on ?savage and brutal? life, a capturing of anomie critically decisive for the Western juridico-political order. Despite the ambiguities of the articulation of the juridico-political to the biopolitical through the relation of exception which the Moleleki execution demonstrated, an articulation which, according to Agamben is, ?founded on an essential fiction? (Agamben, 2005, p. 86) that maintains law in a relation to life, this articulation is critically important, ?as long as the two elements [the juridico-institutional and biopolitical] remain correlated yet conceptually, temporally, and subjectively distinct their dialectic ? though founded on a fiction ? can nevertheless function in some way? (Agamben, 2005, p. 86). Methodology Introduction While one brutal incident of violence stands at the centre of this thesis, the nature of the investigation undertaken here is not primarily an empirical one that seeks to uncover new ?facts? about the Moleleki execution. It is rather an investigation of the construction of knowledge around the ?fact? of violence, the bloody corpses of the Moleleki dead. In this unpacking of the construction of knowledge about the execution within a juridical discourse on power, it becomes possible to re-read the Moleleki execution within a different discourse 35 on power, as an exposure of sovereign exception at the boundaries of the juridico-political order. Thus this dissertation seeks to understand the construction of knowledge around the Moleleki execution as part of a particular construction of knowledge about power in both the period of negotiated transition and the immediate post-apartheid period. During the post-apartheid period this construction of knowledge about power sought to found the sovereignty of the post-apartheid state in terms of a new norm, a norm of human rights, in which violence would be an implacable exception to the proper functioning of juridical power. Within this context, the Moleleki execution was produced as a malleable object in the institutional contexts of the TRC and the courtroom. This thesis, therefore, not only seeks to expose a particular construction of knowledge about power, but is also concerned with the way in which this construction of knowledge about power has produced power in the institutional matrices of the TRC and the courtroom. In this concern, this thesis is informed by a line of investigation articulated by Michel Foucault as pouvoir-savoir, ?power/knowledge? (Gordon, 1980, p. 233), i.e. a refusal ?to separate off knowledge from power? (Rabinow, 1991, p. 7), which led Foucault to explore the way in which this power/ knowledge nexus has led, since the 18th century, to the creation of ?new ?technologies? for the governance of people?, as ?Man? becomes simultaneously a ?subject and object of knowledge? (Gordon, 1980, p. 234). We can see these processes of objectification and subjectification at work in the juridico-political inscription of the Moleleki execution. 36 Genealogy/Archaeology The unpacking of the construction of knowledge about an object of investigation, rather than a direct investigation of the object itself, clearly requires a specific methodological approach. In this line of investigation, this dissertation draws on the genealogical/archaeological methodologies proposed by Foucault, which seek not an ?originary truth? but ?a form of history which can account for the constitution of knowledges, discourses, domains of objects, etc.? (Foucault, 1977a, p. 59). While genealogy seeks the external conditions of the constitution of knowledge, archaeology seeks to expose the internal ?rules of formation? of discourse that constitute the conceptual terrain in which knowledge is formed and produced (Young, 1981, p. 48). Both methodologies work in close association with each other. Critically, in his genealogical method, Foucault opposes himself to a metaphysical conception of ?truth? and instead seeks to historicize the development of the concept of ?truth?, our ?will to truth?, and the various discursive and social practices that have accompanied this pursuit of ?truth? in different historical periods. For Foucault, ?truth? is both a product of power and itself produces ?effects? of power: Truth is a thing of this world: it is produced only by virtue of multiple forms of constraint. And it induces regular effects of power. Each society has its regime of truth, its ?general politics? of truth: that is the types of discourse which it accepts and makes function as true; the mechanisms and instances which enable one to distinguish true and false statements, the means by which each is sanctioned; the techniques and procedures accorded value in the acquisition of truth; the status of those who are charged with saying what counts as true. (Foucault, 1977a, pp. 72-73) In order to expose this ?regime of truth?, Foucault proposes a genealogical method which recovers two types of ?subjugated knowledges? that have been ?disqualified? by ?global totalitarian theories? (Foucault, 1976, p. 80). These subjugated knowledges include ?historical knowledge of struggles? as well as ?popular knowledge?. Archaeology, according 37 to Foucault, provides a method for analysing ?local discursivities?, and ?genealogy? is the ?tactics whereby, on the basis of the descriptions of these local discursivities, the subjected knowledges which were thus released would be brought into play? (Foucault, 1976, p. 85). This dissertation takes its cue from Foucault?s methodology in seeking to locate the production of ?truth? about the Moleleki execution within a particular historical conjuncture, namely the transition to democratic governance, which created the conditions of possibility for the ?truth? about power that the TRC as an institution was established to fix and delineate, and the limits of this ?truth about power?, which were fixed at the boundaries of the law in the courtroom processes where the Moleleki execution was finally adjudicated. This endeavour has been made possible by the deployment of ?subjugated knowledges?, namely an attention to the ?singularity of the event? (Foucault, 1971, p. 76) against assumed teleologies of history and the recovery of the body as the ?inscribed surface of events?, a genealogy that is ?thus situated with the articulation of the body and history? (Foucault, 1971, p. 83). In addition, in recovering ?popular knowledge? of this event through attention to the voices of the various protagonists in the struggles that led to the Moleleki execution, this dissertation has sought not a ?primitive empiricism? (Foucault, 1976, p. 81) but an attention to the dispersion and diffusion of these voices, which refuted the totalising discourses within the institutional contexts of the TRC and the courtroom. Thus this dissertation seeks to utilise these ?subjugated knowledges? to investigate the ways in which the discourse on power within the TRC and the courtroom processes that followed, constrained and produced what could be known about the Moleleki execution, to investigate what Foucault called the ?order of discourse?, the conceptual terrain in which knowledge is 38 formed and produced. (Young, 1981, p. 48) and in this process to make explicit a juridical discourse on power, whose ?discursive rules and categories? were a priori an assumed part of knowledge, ?so fundamental that they remained unvoiced and unthought? (Young, 1981, p. 48). In making explicit what was ?unthought? and assumed, this dissertation seeks to make it possible to think beyond the juridical categories, which have so far constrained our knowledge about power, in order to make explicit the biopolitical nature of the power the Moleleki execution represented. However, this ?biopolitical? nature of power, which fell outside juridical conceptions of power, appeared ?beyond comprehension? in the institutional contexts of the TRC and the courtroom when articulated by SDU members. Both institutions consequently struggled to produce a coherent institutional ?truth? about this incident, its violence remaining essentially an enigma, ?a radical problem of understanding?. An Encounter with the Moleleki Execution I first encountered the incident of violence which is explored here, namely the Moleleki execution, while working as a researcher at a small non-governmental organisation called Peace Action, which was established to monitor and respond to the dramatic escalation of violence in the country during the lead-up to South Africa?s first democratic elections. Unlike other monitoring organisations operating at the time, Peace Action did not primarily rely on press reports of violence, but employed fieldworkers to record first-hand accounts from people who had experienced violence. It also set up a contact line, which was operated by volunteers 24 hours a day, for residents from affected areas to call and request assistance or intervention. In addition, the organisation did a considerable amount of monitoring of the weekly funerals of those who had died during the ongoing conflict. As a researcher at the organisation, I participated in these activities and was, like my co-workers, exposed to a considerable amount of violence and the trauma of those who had lost family, relatives or 39 friends or who had been directly attacked. However, among the welter of violence we all worked with on a daily basis, the Moleleki execution, coming on the eve of the country?s first democratic elections, and so explicitly contesting the primary patterns of violence at the time, namely the fault line of conflict between the IFP and ANC, evoked for me a ?a radical problem of understanding?. This was no doubt produced by the horror of the incident itself and its spectacular brutality. Peace Action had previously worked with the SDU and civic association in Moleleki Extension 2. It was a civic leader from Moleleki that Peace Action had worked with previously, who ?escaped? the execution and first came to report the incident to the organisation. I took his statement. While a researcher at the TRC, I encountered the Moleleki execution again. Here five mothers of the young boys who had died in the execution made statements to the TRC?s Human Rights Committee. Recounting the events surrounding the death of their sons, these mothers clearly understood that the violence had occurred in a ?political context? and they were consequently found by the Human Rights Committee to be victims of ?gross violations of human rights?, i.e. the violations they had suffered had been ?political? in nature. However, when a group of 13 SDU members later applied to the TRC for amnesty, where a more rigid and quasi-legal set of criteria were applied to adjudicate the ?political? nature of their acts, they were denied amnesty, primarily on the basis that their actions had not been ?politically motivated? and therefore were not political in nature. Five of the protagonists were later sentenced to life terms of imprisonment for their role in the execution. Documentary material Because the Moleleki execution has been the subject of adjudication in two institutional contexts, namely the TRC and the courtroom, it has generated a particularly rich body of 40 documentary material for analysis. While the amnesty hearing of the SDU applicants heard evidence from 13 SDU members over seven days and generated more than 700 pages of transcripts, the court proceedings generated approximately 2000 pages of transcript. Read against each other, these transcripts made it possible to trace the successive processes of juridical inscription around the Moleleki execution and the way in which these institutional contexts both constrained and produced what could be known about the execution. In the context of the TRC, the success of SDU members? applications depended on their ?full disclosure? of the incidents of violence in which they had been involved, which meant that that these applicants gave significant amounts of detail about the execution and the circumstances in which it occurred. However, in the context of the court case, the strictly legal nature of the process significantly constrained what could be said about the execution and how it could be said. The fact that admitting involvement in violence could lead the SDU members to be sentenced to jail terms meant that all the SDU members who stood trial denied their involvement in the execution. It was less the falsehood of SDU members? testimony in the courtroom that was of interest, than the way in which the legal discourses utilised in this context, and the alibi defences used by the Moleleki protagonists, both limited and produced a particular construction of knowledge around the execution, which was different to the knowledge produced about the execution, and the way it was constituted as an object, within the context of the TRC. The amnesty processes around the execution were less constrained by a rigid juridical discourse. The TRC Committee?s desire to understand the ?context? in which the Moleleki execution took place, created the opportunity for a more detailed picture of the events surrounding the execution to emerge. On the one hand this provided useful information to 41 build an outline of the details and sequences of events in Moleleki, which made it possible to describe the incident and the struggles around it coherently. However, the primary focus of the analysis of these transcripts was to understand the construction of knowledge around these ?facts? of the execution within this particular institutional context. While the analytic approach adopted was not a traditional ?discourse analysis?, seeking to unpack hidden meaning within the texts, the interactions between the TRC Committee and the SDU members, and the mutual incomprehension which many of these interactions evoked, was particularly valuable in exposing the contradictions and ambiguities created by the attempt to inscribe the Moleleki execution within a juridical discourse on power. Reading the responses of the Moleleki applicants against this dominant discourse made it possible to free this ?subjugated knowledge?, as a tactical strategy to understand the construction of knowledge around the Moleleki execution in terms of a particular genealogy of power. In-depth interviewing and focus group In addition to the analysis of the documentary evidence around the execution, a further research strategy involved in-depth interviews and a focus group conducted with various actors, in an effort to free ?subjugated knowledges? around the execution, in particular the ?popular knowledge? of protagonists in the conflict, which had been constrained by the institutional contexts in which this knowledge had previously been articulated. To this end I held in-depth interviews with four of the SDU members who had been imprisoned for their involvement in the execution during August 2001 at Leeuwkop prison maximum security prison north of Johannesburg. I received permission from the Department of Correctional Services to interview the five SDU members jailed for life terms for the execution. The interviews were conducted after receiving the informed consent of the men to participate in the research process. One SDU member did not give his consent and was consequently 42 excluded. However, what was notable was the willingness of the other four SDU members to participate in the interview process and engage in wide-ranging discussions during protracted interviews of at least two hours, as well as a readiness to show a considerable level of vulnerability and complexity of emotion regarding their actions. Oscar Motlokwa, for instance, while willingly admitting to the killing of young ANCYL members, shed tears of shame and anger regarding his conviction for the murder of his neighbour and colleague, the young civic activist Bulelwa Zwane, for which he denies responsibility. I conducted interviews using a semi-structured interview protocol. During the interview process I was accompanied by a former Peace Action fieldworker, Jabu Dlamini, who was familiar with the Moleleki execution and who has worked for many years on the East Rand. She administered the questionnaire on my behalf in the language respondents felt they were most comfortable with. I sat with Jabu, listening and contributing to the interview process. After most questions, Jabu would translate the responses of the interviewees and I would ask further questions if necessary. There is no doubt that in these processes of translation and retranslation nuances were lost and meaning in some cases may even have been distorted. The problem was further exacerbated by the fact that prison authorities did not allow me to carry a digital recorder into the interviews so that the interviewees? responses could be directly recorded and then translated. Despite these limitations, Jabu Dlamini?s intimate knowledge of the conflict on the East Rand and of the Moleleki incident in particular, helped mitigate these problems, as she was able to probe and develop the themes in the questionnaire based on her own knowledge, in interaction with myself. Basic biographical details of the interviewees revealed that these four SDU members were men in their thirties and forties with established families, who had been engaged in long-term 43 employment prior to the execution, and were not the ?alienated? youth portrayed in most literature portrayed SDU members as. It appeared that is was their social ?rootedness? rather than their social dislocation, as posited by many commentators, which had propelled them to participate in the defensive structures established in South African townships during the 1990s. Other questions elicited information about the township of Moleleki itself, as well as the political conflict in South Africa and within Moleleki at the time of the execution. Further interview questions inquired about the SDU structure in Moleleki, as well as these members? own motivations for participation in SDUs. Questions also sought to elicit a more detailed understanding of the sources of division and contention in Moleleki and interviewees? understandings of violence within this context. Interestingly, what the SDU members said in these interviews did not differ significantly from what had been said in the institutional contexts of the TRC and to a lesser extent, the courtroom process, particularly in terms of their articulation of their role in Moleleki in terms of a concept of ?community defence?. Thus, rather than articulating their role in terms of a party political division, they evoked the biopolitical fracture between bare life and ?political life?, which is substantiated against the boundary of homo sacer, the ?criminal?. However, these discourses on power remained as disqualified knowledge within the institutional contexts in which they had been articulated. In their repetition during the in-depth interviews, it became evident that attention needed to be paid to this ?subjugated knowledge? in pursuit of an understanding of the nature of the conflict in Moleleki and the construction of knowledge about it. In addition to interviews with the SDU members themselves, during 2001 I also held in-depth semi-structured interviews with the mothers of three of the victims of the execution. Two of 44 these interviews took place at the homes of the mothers who still lived in Moleleki Extension 2, while another mother was interviewed at her place of employment as a domestic worker on the East Rand. I was again accompanied by Jabu Dlamini who assisted with translation where this was required. The interviews were recorded and later transcribed. It was clear that these mothers, nearly eight years after the death of their children, were still severely traumatised. We interviewed Edward Mootsi?s mother, Tina Mootsi, at the home where she was employed as a domestic worker. Tina broke down in uncontrollable tears while recounting the events surrounding the death of her fifteen-year-old son, who had been shot and decapitated during the execution. She had received psychological counselling but still struggled to manage the grief of this loss. In general, besides eliciting the mothers? experiences around the loss of their children, these interviews also focused on the women?s perceptions of the sources of conflict in Moleleki, their experience of SDU members in Moleleki prior to the execution as well as their perception of the subsequent amnesty and criminal judicial processes. Particularly significant information to emerge from these interviews was the close ties between some of the relatives of the Moleleki dead and SDU members. Elphina Mugadi, the mother of one of the deceased victims, Isaac Mugadi, described a particularly close association with SDU member Zola Sonti on the basis of a common geographical origin, the Eastern Cape, and a common language, Xhosa, denoting in this closeness the ambiguity of the biopolitical fracture that mediated the boundaries of the ?community? in Moleleki. Finally I also held a focus group interview with members of the ANCYL in Moleleki during 2001, in order to elicit information about the contest for sovereign authority in Moleleki from the perspective of ANCYL members, a perspective which had emerged in a limited fashion 45 within the context of the TRC and the courtroom largely through the testimony of ANCYL and civic leader Lethusang Rikaba. Jabu Dhlamini assisted with translation. The focus group took place in Moleleki Extension 2 where most of the interviewees were still living. The interview was recorded and transcribed. In contrast to the SDU members whom I had interviewed, who presented themselves as mature, articulate, and grounded adults with a firm grasp of their identity and role within the Moleleki community, the group of ANCYL members I met were notable for their youthful appearance and demeanour, hesitant articulation, withdrawn manner and still evident signs of trauma. They appeared to embody marginalisation as young people literally living on the boundaries of the community. They relied significantly on ANCYL leader Lethusang Rikaba for the articulation of their concerns. The intention of the focus group was to acquire basic biographical information about the ANCYL members and explanations as to why they had joined the Youth League. The focus group was also directed to acquiring the perspective of the ANCYL members on the processes that had led to the execution, the division between the ANCYL and the SDU, and the role they believed the SDU had played in Moleleki. It was through this process that it was possible to read against the SDU testimony a different discourse around the conflict, primarily centred on the notion of rights, drawn from the discourse of the ANC and allied organisations. Chapter Outline Chapter 2, ?Analysing a Condition of Exception? seeks to explore the ?radical problem of understanding? posed by incidents of violence such as the Moleleki execution that took place in a context of exception and suspension of juridico-political order. The chapter does this by locating the Moleleki execution within a broader investigation of the juridical terms in which South African analysts attempted to understand the ?exceptionality? of the conflict of the 1990s. These analysts assumed that this state of exception was a consequence of a breakdown 46 of juridico-political order and the political sovereignty of the state, a lacuna in law, rather than an exposure of sovereign exception, at the boundaries of the juridico-political order. The chapter reviews the extensive literature, which the conflict of the period generated, in order to expose the aporia of these analyses, which struggled to recognise the nature of the exception that lay before them as something outside, rather than inside the juridical order. It was these conceptual aporiae which were reproduced in the subsequent institutional contexts of the TRC and the courtroom which formally adjudicated the Moleleki execution, replicating in these contexts the ?radical problem of understanding? encountered by South African analysts who had earlier attempted to interpret the nature of the conflict of the 1990s. Having re-read the violence of the 1990s as an exposure of sovereignty at the boundaries of the juridico-political order, Chapter 3, ?A Struggle of Juridico-Political Ordering?, empirically investigates this ?exposure? of sovereignty during the negotiation period prior to South Africa?s first democratic election in terms of the originary struggles of sovereign exception between the ANCYL and the SDU in Moleleki Extension 2. This struggle which took place during the year preceding the Moleleki execution, sought to tie the land of this newly established township to irrevocable forms of juridico-political ordering, which could putatively form a boundary against the anomie of violence sweeping across the East Rand at the time. However, the boundary which could exclude anomie proved elusive, unlocalisable. It was around this space of exception, which could not be annexed to the juridico-political order of the township that the contest for sovereign authority in Moleleki turned, leading finally to the Moleleki execution as SDU members attempted to localise the exception in the bodies of ANCYL members. The chapter briefly begins to explore the subsequent struggles around the transcription of these bodies, which the SDU had attempted to produce as definitive signs of sovereign exception, in the efforts by the parents of the Moleleki dead to 47 restore their physical integrity through calls for the return of missing body parts and the beginning of a process of juridico-political inscription around these corpses, which sought to start the process of reclaiming the Moleleki dead by re-inscribing the execution within the juridico-political order of the post-apartheid state. Chapter 4, ?A Juridico-Political Investigation?, explores the beginning of these processes of juridico-political inscription, prior to South Africa?s first representative elections under the auspices of the declining apartheid state, as well as the early processes of inscription in the immediate post-apartheid period. Ironically, while the norm of law changed significantly from one period to another, both periods shared an essential contiguity in their efforts of juridico-political inscription, which sought to capture this violence unleashed from law within the law again. These processes of inscription began with the reproduction of the corpses of the Moleleki execution as signs of sovereign exception by the apartheid state. Thus, the corpses of the Moleleki executed were initially identified at the site of the execution as those of ?criminal? hijackers, the embodiment of criminal exception. The reproduction of the Moleleki executed as the embodiment of criminal exception, was followed by a process of medico-legal inscription. While the medico-legal inscription of the bodies of the Moleleki dead, were conducted in anticipation of a juridical investigation, the ongoing context of conflict meant that residents of Moleleki, including the relatives of the deceased were unwilling to work with the police. However, after the election of a democratic government, residents of Moleleki claimed a new juridical subjecthood as citizens of the new state and came forward with information to facilitate the resumption of a police investigation. However, when a number of SDU members applied to the TRC for amnesty for their actions, as the judicial investigation against them was initiated, in the hope that this process would suspend the process of criminal prosecution; a critical ambiguity was evoked around the 48 nature of these actions, i.e. whether legal consequences could be attached to actions that had taken place in a context of exception and therefore whether the Moleleki execution was properly fell under the jurisdiction of the TRC, where legal consequences would be suspended, or the courtroom, where legal consequences would be imposed. Chapter 5, ?The Truth and Reconciliation Commission, Fixing the Limits of the Political?, is a contextual chapter on the TRC, which explores the struggles during the negotiation process and later within the TRC itself, to define the nature of post-apartheid ?political?. The chapter explores the attempt by the nascent post-apartheid juridico-political order to define the juridico-political present in terms of the exception of the biopolitical past. This process initially began with the negotiations prior to the country?s first democratic elections and was later institutionalised within the context of the TRC. The TRC sought to define the sovereignty of the new state in terms of a new norm of law, the norm of ?human rights?, which would putatively make it possible to define the boundaries of the juridico-political present, in terms of the exception of the violence of the past. However, these attempts to redefine the post-apartheid ?political? as the juridico-political, ran aground in terms of a series of theoretical aporiae located within a juridical understanding of the political and its instrumental relation to the exception of the ?past?, which masked the complexity of the political as the site of exception, a realm of ambiguous articulation between biopolitical and juridico-political power at the boundaries of the juridico-political order and the biopolitical nature of the fracture that would constitute the present political through a distinction between bare life and political life. Thus the nature of the ?political? that the TRC was established to articulate, was to remain profoundly elusive, with the Commission acknowledging in its report of 1998 that, ?the political nature of specific acts was hard to define? (TRC, 1998b, p. 82). 49 Chapter 6, ?Fixing the Limits of the Political, the Moleleki Execution and the Truth and Reconciliation Commission?, explores some of these complexities around the adjudication of the political in terms of the post-apartheid processes of juridical inscription around the Moleleki execution within the TRC. These processes of inscription sought to adjudicate the actions of the Moleleki protagonists in terms of an instrumental relation with a prior juridical intention, namely a ?political? motive. However, the TRC struggled to adjudicate the actions of the Moleleki protagonists in these terms as rather than invoking the party political opponent as the object of their actions, SDU members instead invoked the ambiguous figure of homo sacer or the criminal to denote the boundary of the political community. Moreover, they were unable to articulate the motive for their actions in terms of an intention that was instrumentally linked to the programmes and policies of a juridically constituted political party such as the ANC. The chapter documents the struggles of comprehension that were precipitated by the engagement between the Moleleki protagonists and the Commission, as the SDU applicants repeatedly evoked a fluid notion of ?community defence? against a ?criminal? other as the motivation for their actions. However, this reference to a biopolitical ?community? embodied by angry residents at various gatherings, created significant problems of comprehension for the TRC, which was only able to understand ?community? in terms of a fixed juridical representation in an organisation such as the South African National Civic Organisation (SANCO). The failure by SDU members to invoke a juridical authority for their actions and their reference instead to a biopolitical authority led the TRC to reject their application for amnesty on the basis that their actions had been not been ?politically motivated?. 50 Chapter 7, ?Fixing the Limits of the Law, the Moleleki Execution and the Court Trial?, explores the processes of adjudication around the Moleleki execution in the courtroom after the TRC refused amnesty to the SDU applicants. The criminal proceedings which had been suspended for the duration of the amnesty hearings were reinstated, leading to the trial of seven members of the SDU in Moleleki, four of whom had testified at the amnesty hearings. The court proceedings sought to establish the ?criminal? nature of the actions of the Moleleki protagonists and the legal consequences which should be attached to their actions. While criminal law in general requires the demonstration of a direct instrumental relation between the criminal intent of the actor (mens rea) and the actions of this actor, the collective nature of the actions which had led to the Moleleki execution, meant that such an instrumental adjudication simply could not be made. Despite this, in the effort to inscribe this anomie within the juridico-political order, the judge in the Moleleki case used the doctrine of common purpose to attempt to establish the criminality of the actions of the Moleleki protagonists. The common purpose rule, as developed by the period that the Moleleki judgement was delivered, had significantly expanded the basis of liability for murder in contexts where more than one individual was involved, doing away with the requirement that there had to be a causal connection between the acts of each individual and the eventual consequence of murder. The chapter explores the ambiguities evoked by the attempt to inscribe the execution within law in terms of the doctrine of common purpose, which sought a corporate, rather than individual basis of responsibility for the execution. The expanded liability provided for by the common purpose rule meant that despite very little direct causal evidence, five of the accused were found guilty and sentenced to life terms of imprisonment, not in terms of individual liability but in terms of an ?association with? a common purpose to kill the executed youth. 51 This thesis therefore utilises the Moleleki execution as a way to understand the construction of knowledge about power within a juridical discourse on power at two historical junctures, the period preceding the country?s first democratic elections when the Moleleki execution took place, and the immediate post-apartheid period when the Moleleki execution became the subject of significant efforts of juridico-institutional inscription within the contexts of the TRC and the courtroom. In exposing this construction of knowledge about power the thesis seeks to make explicit the biopolitical nature of power, and the ambiguity of its articulation with juridico-political power, in a relation of exception. While this articulation has been conflated under the modern juridical conception of sovereignty, this thesis takes its cue from Agamben to explore this articulation, through the Moleleki execution, as the ?hidden nucleus? of sovereign power. 52 CHAPTER TWO: ANALYSING A CONDITION OF EXCEPTION Introduction: The Exception of South African Violence As was argued in the previous chapter, the struggle to understand the violence of the 1990s and the horror of atrocities such as the Moleleki execution was not so much an empirical problematic, which the most persuasive evidence and argumentation could resolve through the unmasking of a single empirical ?truth?. It instead concerned a construction of knowledge about violence within a juridical conception of power, in which the biopolitical nature of power remained hidden and implicit and violence was conceived as an implacable exception to the ?proper? functioning of juridical power. This chapter therefore endeavours to explore and make explicit these paradoxes of political conception, which traversed the period of violence prior to South Africa?s first democratic elections and the subsequent processes of juridico-political inscription in terms of which this violence was understood in the immediate post- apartheid period of political reconstruction. This chapter explores these paradoxes of political conception through a re-reading of the critical conceptual terms on which the violence of the 1990s and the Moleleki execution itself were understood, namely the terms of the exception. While South African analysts and the juridico-political institutions that attempted to inscribe the violence after 1994 understood this exception in terms of a breakdown of juridico-political ordering, thus as a lacuna in law, an exception to the ?proper? functioning of the juridico-political, the exception instead concerned the very boundaries of the juridico-political order, its initial conditions of constitution at the point of articulation between biopolitical and juridico-institutional power. It constituted ?nothing less than the limit concept of the doctrine of law and the State? (Agamben, 1998, p. 11). 53 Misrecognising the Nature of the Sovereign Exception Thus analysts, while correctly identifying the ?exceptionality? of the violence of the 1990s, fundamentally misrecognised the nature of the exception which lay before them as a juridico- political phenomenon that could be empirically and historically demonstrated (for example Kynoch, 2006). Many authors referred to the major confusion (Beinert, 1992; Bruce & Komane, 1999; Friedman, 1993; Johnston, 1997; Morris & Hindson, 1992; Segal, 1991; Simpson, Mokwena & Segal, 1992; Taylor & Shaw, 1998) of South African politics during this period, to which it was ?very difficult to assign political meaning? (Johnston, 1997, p. 85). They spoke of a power vacuum in this period of interregnum (Brewer, 1994; Friedman, 1993; Johnston, 1997; Sisk, 1995), ?a situation which resembles a Hobbesian natural society more than one of legitimate authority? (Johnston, 1997, p. 86), in which ?the rules of politics are often the rules of war, which are in effect no rules at all? (Sisk, 1995, p. 20). However, in seeing the ?exceptionality? of the period in terms of a breakdown of juridico- political rules, analysts were not able to recognise the real exception before them and the nature of the power that emerged in this exception. Critically, this power concerned the point of intersection between biopolitical and juridico-institutional power, the ?hidden? nucleus of sovereign power. While analysts of the violence at the time assumed that this violence concerned a breakdown in state sovereignty or a distortion of state sovereignty during the period of negotiated transition, what they were in fact attempting to analyse was not the failure of a normatively conceptualised sovereignty instantiated in the state but the exposure of sovereignty, as the form of the state crumbled. In general this threshold remains concealed, it is the dissolution of state structures which makes sovereignty visible: ?As long as the form of State constituted the fundamental horizon of all communal life and the political, religious, juridical, and economic doctrines that 54 sustained this form were still strong, this ?most extreme sphere? could not truly come to light? (Agamben, 1998, pp. 11-12). While the crumbling of state structures may expose this boundary of juridico-political order at particular historical conjunctures, this ?state of exception? is not a temporal phenomenon, a chronological event in time, nor is it a pre-juridical phenomenon, the chaos that precedes order. Instead it is a zone of indistinction and continuous transition that dwells continually within the civil state, simply becoming explicit ?at the moment the City is considered tanquam dissolute, ?as if it were dissolved?? (Agamben, 1998, p. 105). In 1990, as the negotiated South African transition began and the South African state and law hovered on the interregnum of the old and the new, the juridical order was essentially suspended during the negotiation process, the ?City...appeared as if it were dissolved? (Agamben, 1998, p. 105) making explicit the boundaries of the entire juridico-political order. In this historical moment and at this boundary of paradigmatic order, sovereign power revealed itself in an extremity of violence which radically disputed the terms of the juridico- political order, defying comprehension in terms of this order and contesting juridical, political, social or ontological inclusion in this order. The ways in which power emerged at this boundary of juridico-political order therefore disputed inclusion in a juridical conceptual paradigm, which spanned both the period of the violence during the 1990s and the processes of juridico-political inscription around the violence and the Moleleki execution post-1994. The struggles of comprehension of analysts prior to 1994 primarily concerned the attempt to trace an instrumental relation of cause and effect between the violence committed and a prior source of power and origination. The 55 multiplication of causal explanations that emerged during this period spoke to the conceptual aporia embedded within a purely juridical conception of power, which could not recognise the biopolitical nature of power that became so explicit during the 1990s but which, as Agamben argues, has remained implicit in the entire history of Western thinking about sovereignty. This tracing of an instrumental relationship between a prior source of power and an act of violence was substantively premised on a particular understanding of the subject, the sovereign subject of Western discourse, whose actions were conceived as the unitary, voluntary, rationally chosen product of a pre-existing consciousness. However, the violence of the Moleleki execution and the actions of its protagonists implied a different form of subjecthood, an embodied subjecthood in which the sovereign exists as ?living law?, the embodiment of law rather than the subject of law, and homo sacer is the ?bare life? captured by sovereign power (Agamben, 2005, p. 69). In this context, power is not juridical, it is animated directly from the body of the actor and agency is the product of multiple subject positions embedded in the shifting praxis of actors (Feldman, 1991, p. 4) not the instrument of an original source of power but an ?embodied force?, a biopolitical power. A Normative Conception of Sovereignty In South African discourse, the construction of knowledge about the violence of the 1990s was deeply embedded within a genealogy of thought running from Hobbes through to Weber in which the conception of sovereignty is not only a juridical phenomenon but is also fundamentally tied to the form of the state, a conception of sovereignty which Foucault, in his call for a political theory not constructed around the juridical concept of sovereignty, rejected as a basis for an analysis of power (Foucault, 1976). 56 Even more critically, this conception of sovereignty understood the foundation of state power in terms of its ability to monopolise violence within society. For Hobbes, through a social contract citizens give up their right to violent self-defence to the sovereign state or Leviathan, whether democratic or autocratic, to act as guarantor of peace and civility in a world that would otherwise be characterised by a ?condition of war of everyone against everyone? (Woodbridge 1958, p. 270). For Max Weber too, the essential raison d??tre of the state is its ability to monopolise the legitimate means of violence over a specific territory. Therefore, in this conception the legitimacy of the state forms the basis of the legitimacy of its monopoly on violence and its right to employ violence as part of the exercise of state power. As South Africa analyst Pierre du Toit contended, ?security?, the ability to provide for the basic physical security of citizens, forms the ?core of the substance from which a social contract between state and citizen is built. When it [the state] fails to provide security, and is seen to fail, the state?s legitimacy cracks at this core? (P. du Toit, 2001, p. 72). The basis of sovereign state authority as this assumed ability to monopolise the means of violence in society led on the one hand to uncritical assumptions about the capacity of the state, ?as the primary unit for the conduct politics? (P. du Toit, 2001, p. xii), to monopolise power in its entirety, as well as more fundamentally to a misrecognition of the basis of state authority as this monopoly. This misrecognition left South African analysts, when faced with the patent breakdown of this monopoly, with only one avenue of analysis, namely the mechanisms and processes of the restoration of democratic state sovereignty, on the implicit assumption that such a form of state sovereignty could secure the conditions of peace or at least constituted the most fundamental precondition for peace. As analyst Andre du Toit (1993) articulated it, a conception of political violence within a South African master 57 narrative of political struggle, in which the exercise of legitimate juridical state power would end violence and ?projected democracy as the solution to the history of conflict and political violence? (p1). Embedded therefore in these analyses were the teleological conceptions of history, its ?predictable sequences?, which Foucault so strongly critiqued in his essay ?Nietzsche, Genealogy, History? (Foucault, 1971) and the conception of violence within this teleology, which assumed humanity would inevitably move ?from combat to combat until it arrives at universal reciprocity, where the rule of law finally replaces warfare? (Foucault, 1971, p. 85). While acknowledging the complexities of South Africa?s ?bold new experiment? (Giliomee & Schlemmer, 1994) and that ?the reorientation of conflict from bullets to ballots is a long, arduous, and often violence-ridden road? (Sisk, 1995, p. 4), the possibility of this ?transition? from a state of violence to a state of peace was posited as merely a matter of ?reorientation? in which institutional arrangements, the re-establishment of new ?rules of the political game? (Friedman, 1993; Sisk, 1995), could make possible this teleological shift. What was inconceivable in this paradigm was that this state of violence and transition would, as Agamben argues, dwell continuously within the ?City?, violence not displaced but installed in law. The continuation of contemporary forms of violence in the post-1994 context thus compelled analyst Pierre du Toit (2001) to preface his book on South Africa?s ?brittle peace? with the question, ?Why has South Africa, newly democratised, not also become peaceful and free of violence?? (p. xi). These legitimist assumptions about the nature of state authority and the right and ability of a ?legitimate? state to monopolise the means of violence in society, has informed almost all but 58 the most recent analyses of South African violence (Buur, 2005; Hornberger, 2004; Jensen, 2003). Thus in this conception, as South African analyst Pierre du Toit (2001) contends, the ?essential reason of state? is ?expressed in the principle of sovereignty and is executed through the monopoly on force and the prescriptions of law? (p. 4) and simply hinges ?on the credibility of the state?s claim? to a monopoly of force (p. 5, own emphasis), a claim that is ?open to abuse? (p. 5) but not fundamentally misconceived. Therefore, in this paradigm, this state monopoly on the means of violence is the ?first and foremost good public good, which the [democratised] state could offer? (P. du Toit, 2001, p. 4). In this conception the ?turn? to political violence (A. du Toit, 1993, p. 2) of opposition movements such as the ANC, in instrumental pursuit of its political objectives of incorporation in the modern state, was a completely ?rational? and explicable response to the racial exclusions of the South African state and its abuse of its monopoly of violence in repressive defence of its exclusionary policies. Thus, ?if violence was a familiar phenomenon in many different contexts?the violence of apartheid was generally understood as part of the pathology of apartheid?the problem was not so much the violence engendered by [apartheid] but apartheid itself? (A. du Toit, 1993, p. 2). In this vein Brewer argues that ?a political settlement?will ensure that violence loses popular support when directed against a legitimate government? (Brewer, 1994, pp. 5-6). This normative conception of state sovereignty within South African political discourse, what Andre du Toit termed a ?commitment to the basic aims and assumptions of modernisation? (A. du Toit, 1993, p. 20) was shared by the two major antagonists within the South African political struggle, namely the apartheid state and the ANC. As Crais argues, ?the ANC sought access to state power, not its repudiation? (Crais, 2002, p. 143) and ?its critique of the state 59 was less radical than reformist, that all individuals should have access to the state, irrespective of race? (Crais, 2002, p. 143). It was precisely because of its previous exclusion from the state that for the ANC ?inclusion in this modern state came to define the goals of legitimate political action? (Crais, 2002, p. 143) and legitimate political violence. This shared commitment to the aims and assumptions of modernity by the two major protagonists in the political struggle protected the normativity of juridical sovereign state power and eventually created the framework in which political reconciliation between the two antagonists could take place (A. du Toit, 1993, p. 20). This was the discursive framework within which the analysis of the violence of the 1990s took place, conceiving of this violence either as a consequence of an ongoing distortion of state sovereignty contiguous with the distortions of state power under apartheid (Coleman, 1998; Dugard, 2003b; Everatt, 2003; Everatt & Sadek, 1992; Taylor & Shaw, 1998) or as the result of a breakdown of state sovereignty in the context of the negotiated transition and the contestation of the illegitimate form state sovereignty had taken during apartheid (Friedman, 1993; Gotz, 1995; Morris & Hindson, 1992; Olivier, 1994; Simpson & Rauch, 1993; Sisk, 1995; Slabbert, 1992). However, the assumption implicit in all of these analyses was that the restoration of democratic state sovereignty with a legitimate monopoly over the means of violence would address either the distortion or breakdown of state sovereignty. A breakdown of state sovereignty One influential school of thought conceptualised the breakdown of sovereignty during the transition period as a consequence of the breakdown of the ?rules of the political game? and the consequent ?disorientation? of politics during this period (P. du Toit, 2001; Sisk, 1995; Friedman, 1993). Working within the contractarian approach to political thought dominant in 60 South African discourse, which (in a tradition extending from Hobbes? initial vision of government based on a social contract where individuals give up their right to self-defence in return for the protection of the sovereign) saw the basis of social order as the active and voluntary choice of the individual to consent to that order. This individual, sovereign over his or her actions, is subject only to those forms of authority to which they have consented, ?with the context-transcending power to choose [their] ends and purposes? (Markell, 2003, p. 12). The extreme version of this contractarian approach to political thought, using rational choice and game theory derived from economics, analysed the South African transition and the conflict for control of the state as part of a ?metagame? (Sisk, 1995, p. 20; see also Friedman, 1993). Drawing on the analysis of transitions from authoritarian rule by writers such as O?Donnell and Schmitter, the conflict of the period was characterised as the consequence of the breakdown of the rules of the political ?game?. ?South Africa?s potential for instability during the transition? rested on ?the disorientation caused by the sudden change of all the rules of the political game? (Sisk, 1995, p. 190), in which ?actors struggle not just to satisfy their immediate interests and/or interests of those whom they purport to represent, but also to define rules and procedures whose configuration will determine likely winners and losers in the future? (Guelke, 1999, p. 1). The route to democracy was opened by the realisation ?among a core set of elites? (Sisk, 1995, p. 13) that their ?interests? lay with the benefits of a ?positive sum outcome? as opposed to a ?zero-sum? outcome to the conflict and their consequent rational and strategic choice to negotiate a new political order. The restoration of state sovereignty was therefore dependent on the establishment of a ?democratic social contract? between these negotiating elites based on a process of establishing consensus on the ?new rules of the political game?. The violence that 61 dogged the negotiation process was a result of ?slippage? between leaders and their followers, some of whom continued to operate in terms of a ?zero-sum perception of the conflict? (Sisk, 1995, p. 122). Sisk?s emphasis on elite pacting, premised on an entirely instrumental conception of human agency in which the ?interests? of actors are pre-given and unchanging and their interaction during the negotiation process the result of a conscious maximisation of this essentialist interest, was critiqued in particular by Anthony Guelke, who refuted the power of elites to manufacture and impose political consensus from the centre. However the premise of his argument, framed at the beginning of his book by a quote from O? Donnell and Schmitter, whose thinking also influenced Sisk, was fundamentally the same, namely that violence was to be understood in the ?context of conflict over the rules of political competition among all the contending parties?? (Guelke, 1999, p. 183, own emphasis). While authors such as Sisk, Friedman and Guelke primarily interpreted the breakdown of state sovereignty in terms of a breakdown of political rules, there were a range of other authors who conceptualised this ?breakdown? in different terms. Johan Olivier implicitly analysed the ?ethnic? dimension of the conflict, linking resource mobilisation theory (which posits that increased access to scarce resources results in political mobilisation and collective action rather than an aggregate increase in grievances), to the ?competition model of ethnic mobilisation? (Olivier, 1994, p. 27). He argued that ?ethnic mobilisation is the consequence of competition between groups for resources? and that it is the opening up of ?new competitive opportunities? that provokes ethnic mobilisation when ?traditional boundaries? between ethnic and racial groups break down (Olivier, 1994, p. 28). He explained that, The past fifteen years have seen a gradual erosion of the South African government?s apartheid policies which enforced tight boundaries between ethnic and racial groups. 62 This set in motion a process of social change which is affecting South African society as a whole and impacts on social, economic and political relations in South Africa, which in turn have changed the way in which South Africans engage themselves in political behaviour. (Olivier, 1994, p. 28) Simpson and Rauch in 1993 argued a similar point from a different angle, contending that central to an understanding of the political violence of the 1990s was what they referred to as the ?deregulation of social control? that accompanied the ?the deconstruction of formal apartheid? within a context of heightened political competition during the run up to the country?s first democratic elections, which had resulted in ?high levels of social, political, economic and ideological dislocation within?society? (Simpson & Rauch, 1993, p. 1). For Simpson and Rauch, this ?deregulation? was an ?inevitable window period? in which ?social control is utterly deregulated?, in order to create the space for ?consensus based regulatory institutions and ideological formations? (Simpson & Rauch, 1993, p. 2). Therefore while the repressive mechanisms of social control established under apartheid had been under significant challenge since the 1980s and had been undergoing a ?piecemeal? deconstruction for some time, the process of replacing these mechanisms with legitimate and credible alternative structures of authority was slow. This crisis of legitimacy was particularly severe for the security forces and the criminal judicial sector as a whole. While these trends are characteristic of transition processes around the world, Simpson and Rauch argued that ?in the South African context, this dramatic social insecurity, bred of a sense of transitional disintegration, has articulated particularly destructively with the racial, class-based and ethnic identities and hostile stereotypes generated or reinforced by decades of apartheid? (Simpson & Rauch, 1993, p. 2). 63 Graeme Gotz similarly examined this process of ?deregulation? in his analysis of violence in the Vaal21 region from a psychological perspective, in which he argued that political violence was essentially the result of a ?failure of meaning? precipitated by the social and institutional ?disintegration? that characterised the period between 1990 and 1994, which meant that people were left without institutional means to interpret and order their world and were forced to look for new ways to empower themselves and reassert control. He argues that the psychological solution to this problem was found in the identification of an ?other?. This enemy could be ascribed the blame for the conflict while providing a means of overcoming it, through its defeat. The act of violence against the enemy is thus a spontaneous attempt to re- possess a sense of self and re-establish identity. Gotz defines this process as ?territorialisation? because it involved the localisation of forms of identity and the building of extremely parochial, exclusionary associations through groups or camps (Gotz, 1995). On the other hand Morris and Hindson made a very similar argument to Simpson and Rauch, also locating their analysis of the violence in terms of the ?disintegration? of apartheid institutions, which meant that ?racial, ethnic and class antagonisms held in check under classic apartheid?resurfaced in the climate of liberalisation and deracialisation? at the same time as the ability of the state to control the underlying social antagonisms exposed by this ?disintegration? significantly declined as it ?lost social and political control of black urban life? (Morris & Hindson, 1992, p. 164). Therefore, ?the gradual erosion of apartheid institutions and the abandonment of its policies ?led to an escalation of social tensions and increased, not decreased, violence throughout the country? (Morris & Hindson, 1992, p. 155). Like Simpson and Rauch, Morris and Hindson point to the complexities of replacing one 21 The Vaal Triangle is located to the south west of the province of Gauteng, and comprises a triangular area of land formed by the towns of Vereeniging, Vanderbijlpark and Sasolburg ? three small cities that together comprise a substantial urban and industrial complex. 64 social system with another, as ?old elements, ideologies and strategies remain and social forces committed to the previous order still operate?alongside and in conflict with new elements, ideologies and organisational strategies? (Morris & Hindson, 1992, p. 164). However, unlike Simpson and Rauch, Morris and Hindson did not argue that endemic violence was an inevitable consequence of this ?window period?, the interregnum between the old and the new, but situated their argument in terms of ?the particular way in which the state tried to reform apartheid in the 1980s? (Morris & Hindson, 1992, p. 156), which by instituting a policy of ?orderly urbanisation? that offered differential urban residential rights to some Africans, without a concomitant increase in infrastructural development, ?exacerbated the material basis of conflict in black society? (Morris & Hindson, 1992, p. 156) and led to the proliferation of informal settlements on the edges of townships, as well as an influx of men, women and children to hostels on the Reef, which created a new underclass of unemployed dependent on the employed in the hostels. Both these communities were living on the margins of urban society. Squatters had to continually defend their right to keep the land they had occupied, often through violence. The newly arrived hostel residents, without any economic resources, lived an equally tenuous existence. Thus the differential access to urban rights and resources ?facilitated class differentiation? (Morris & Hindson, 1992, p. 157), creating divisions between those who benefited, however, marginally, from the reform process and those who reaped little or no reward from it, thus laying the basis for violent conflict around urban rights which would later be drawn into rising political and ethnic tensions after 1990. 65 John Kane-Berman also located the violence of the 1990s within the context of the decline of state sovereignty, but working within an atavistic notion of ?conflict among black people? and concerned to refute assertions of the state?s primary responsibility for the orchestration of the violence of the 1990s, argued that it was the ANC?s campaign of ungovernability, launched during the 1980s to challenge the apartheid state?s sovereignty and capture state power, that had successfully undermined state authority in townships and set in motion a cycle of violence, largely perpetrated by ?politically fired up?Black youth? (Kane-Berman, 1993, p. 43) deliberately drawn into the conflict by adult leadership, not only prepared to die for the struggle but ?prepared to kill for it? (Kane-Berman, 1993, p. 43). The ?people?s war? initiated by the ANC during the 1980s had two consequences. On the one hand, ?coercive strategies, intensified struggles, assassination campaigns, mass actions and volcanic upsurges of fired-up people? had ?got out of control? (Kane-Berman, 1993, p. 44). In addition, the ?people?s war? had precipitated counter-violence from those who were the targets of the ANC?s strategy of ungovernability, namely councillors, IFP members and other people identified as ?sellouts? and ?informers?. Thus the ANC in challenging the state power, whether legitimate or not, had challenged the state?s monopoly of violence, thus creating conditions of ?lawlessness?, which were perpetuated and exacerbated during the transitional period of the 1990s. A distortion of state sovereignty In contrast to the literature which saw the violence of the 1990s as the consequence of a breakdown of state sovereignty, there was a body of literature which saw this violence in terms of a continuum with the distortion of state sovereignty that was the hallmark of the apartheid state, which had ?throughout the apartheid years, [used] violence strategically?to 66 enforce and uphold ?white? domination? (Taylor & Shaw, 1998, p. 15). As Shaw and Taylor argued, ?it would not be surprising?to expect violence to continue to be used to the advantage of the minority when the future of the country?s power structures was being negotiated? (Taylor & Shaw, 1998, p. 15). These writers analysed the conflict of the 1990s in terms of the theory of low intensity conflict, which had influenced South African security establishment thinking for some time. This was a methodology of warfare utilised from the 1970s in particular by the United States in various international contexts to attempt to ensure ?centrist? solutions to revolutionary challenges to repressive states, through a combination of negotiation and covert action designed to weaken and destabilise opposition movements. These analysts saw the violence of the 1990s in terms of a continuum with a previous apartheid state strategy of low intensity conflict, which had sought to marshal the full range of social forces at the state?s disposal in a ?total strategy? to counter the ?total onslaught? of political opposition (Dugard, 2003a; Everatt & Sadek, 1992; Taylor & Shaw, 1998). The sophistication of these analyses varied considerably, ranging from the early analyses of monitoring organisations such as the Human Rights Commission (HRC) and the Community Agency for Social Enquiry (CASE), which initially utilised the data they collected to correlate the high points of the violence with key developments in the negotiation process. Using the analogy of a ?tap being turned on and off? (Coleman, 1998, p. 168) by the state they argued that ?violence peaks when it is most likely to damage the ANC and dies down most dramatically when it threatens to harm the NP? (cited in Simpson & Rauch, 1993, p. 7). 67 The HRC argued that a ?new total strategy? (Coleman, 1998, p. 166) was being implemented by the state in line with the principles of low intensity conflict and as a continuation of the total strategy of the 1980s, which combined a ?twin track? (Coleman, 1998, p. 174) strategy of negotiation with destabilisation and involved the orchestrated marshalling of all the forces at the state?s disposal, namely, the security forces, ?hit squads? and ?vigilantes?, in a coordinated campaign to foment violence. This analysis of a collection of forces involved in a campaign of destabilisation (although its level of organisation and the extent of senior state complicity remained contested) would come to be widely characterised in terms of the notion of a ?third force?, the ?hidden hand? of the state behind violence of the period (Coleman, 1998, p. 170; Shutte, Liebenberg & Minnaar, 1998). Working within a profoundly functionalist approach to violence, these analyses, by using monitoring data to ascertain the identity of attackers and victims, sought to demonstrate empirically who ?benefited? from the violence and whose ?interests? it was serving (Institute of Criminology, 1991). For example, Everatt argued that an explanation of the violence could be ascertained by using monitoring data to ascertain who were the ?winners? and ?losers? (2003, p. 100). The conceptual aporia of this approach, in which the origin and cause of violence can be unproblematically imputed from its consequences, was demonstrated for example by the Schutte?s analysis of the ?third force?. Using a Weberian understanding of power as ?the probability that one social actor will be able to carry out his will in spite of the resistance of others? and faced with the ?lack of solid evidence? of third force activity, he asks, ?Who benefits from what is going on?? and answers, ?When applied to politics in South Africa, the 68 answer to the question is basically all those involved in political power relations in the country? (Schutte, 1998, p. 19). The weakness of the methodologies used to collect and interpret the data to support the arguments that violence was ?turned on and off? at strategic moments in the negotiation process, or that particular political groupings (characterised substantially as proxies of the state such as the IFP) were primarily responsible for the execution of violence or were its primary victims, meant that these analyses could be relatively easily dismissed on empirical grounds, as they were by organisations such as the South African Institute of Race Relations (Jeffrey, 1992). However, more importantly embedded within such analyses and in common with analyses which understood the violence of the 1990s in terms of a breakdown of state sovereignty, was a normative view of this sovereignty in which the state usurps power in its totality and wields violence as a tool of power at will in an orchestrated strategy to pursue its own essentialist ?interests?. Thus Taylor and Shaw (1998), in an attempt to refine the argument put forward by HRC and CASE, framed their analysis with the question: ?To what extent can the violence be firmly linked to the political will of the state?? (p. 15) The apparently arbitrary and random nature of the violence as the conflict continued after 1990 had presented a significant empirical challenge to advocates of an orchestrated state plan to use violence as a strategic tool in the negotiating process, leading these analysts to increasingly fall back on the explanation that the state had unleashed forces it was now unable to control, even though the violence was now counter-productive and no longer served their ?interests? (Coleman, 1998, p. 205). Taylor and Shaw (1998) argued that, on the 69 contrary, the ?seeming unpredictable nature of the violence? (p. 16) was in fact part of state strategy. Dugard, in support of this analysis, contended that ?the advantage of LIC [Low Intensity Conflict] is that it does not require comprehensive and ongoing orchestration in order to be effective? (2003b, p. 41). Shaw and Taylor thus argue that until 1992 and the Record of Understanding signed between the National Party (NP) and the ANC, the government followed a dual strategy of negotiation and destabilisation which it was hoped would disorganise the ANC sufficiently to allow the NP, as the centre of an alliance of conservative parties, to win a democratic election and ?lock the ANC into a compromise agreement centred on compulsory power-sharing as opposed to majority rule? (Taylor & Shaw, 1998, p. 17). However, following the Boipatong massacre of 1992, the NP began to realise that its dual strategy of destabilisation and negotiation had become counter-productive: ?the state?s game plan was becoming self- defeating and there had to be a rethink of strategic options? (Taylor & Shaw, 1998, p. 23). Taylor and Shaw cite as evidence of the shift in state strategy the fact that from 1992, although the actual number of incidents of violence continued to increase, the nature of the violence shifted significantly. Thus the number of massacres, assassinations carried out by hit squads and train attacks declined substantially from 1992. They argue that these forms of violence constituted classic destabilisation tactics, in line with the principles of ?low-intensity warfare? where the intent is to neutralise the opposition through killing off radical activists, sabotaging organisational attempts to build a support base, and generally creating a climate in which violence becomes self-perpetuating and preys on people, oblivious to whether either their guilt or innocence can be detected. (Taylor & Shaw, 1998, p. 19) It was only after 1992 and the signing of the Record of Understanding that funding and support by the state for violence was finally withdrawn. Taylor and Shaw leave unanswered 70 the question of why violence continued and in fact escalated in the aftermath of the rethinking of state strategy. Thus the conceptual terms in which analysts attempted to understand the violence of the 1990s, which could not separate out sovereignty from the state form and which furthermore understood the basis of this state sovereignty in terms of a putative capacity to exercise a monopoly over the means of violence, led South African analysts into a series of conceptual aporiae that evoked the ?radical problem of understanding? identified by South African analyst Andre du Toit (1993) at the time. For what lay before analysts was an exposure ? not a breakdown or distortion of sovereignty, an exposure which in its irruption of biopolitical power appeared to contest the very basis of the modern juridical conception of sovereignty that in general obscures the articulation of biopolitical and juridico-political power under the rubric of normative state sovereignty, but whose premises now appeared profoundly contested. Sovereignty and the Exception What then was the nature of this ?most extreme sphere?, the threshold of juridico-political order, which became so explicit during the 1990s and which South African analysts struggled so profoundly to analyse? It is the ?sovereign decision on the exception [that] is the originary juridico-political structure on the basis of which what is included in the juridical order and what is excluded acquire their meaning? (Agamben, 1998, p. 19). Critically, Carl Schmitt argues, the sovereign decision on the exception is not about an individual?s decision on these relations of exclusion and inclusion, but is instead a more fundamental process through which what is inside and outside the juridico-political order acquire their meaning, dividing the anomie of life into a structure that law is able to capture. Thus, the sovereign decision on the exception ?is not the expression of the will of a subject hierarchically superior to all others, 71 but represents rather the inscription within the body of the nomos22 of the exteriority that animates it and gives it meaning? (Agamben, 1998, p. 25). These decisions on the boundaries of inclusion and exclusion within the juridico-political order, are, Schmitt argues, ?the essence of State sovereignty, which must therefore be properly juridically defined not as the monopoly to sanction?The decision reveals the essence of State authority most clearly?? (cited in Agamben, 1998, p. 16). Therefore what is at issue in the sovereign exception is, according to Schmitt, ?the very condition of possibility of juridical rule and along with it, the very meaning of State authority? (Schmitt cited in Agamben, 1998, p. 17). Consequently, sovereign exception is not about positive law, its numerous articulations, its ?justices? and ?injustices?, nor is it about the various articulations of the political order, who within the political order is invested with which powers ? the questions which have so far concerned most contemporary analysts. Instead sovereign exception is about how the order comes into being in the first place, its conditions of constitution; it is about the threshold of political and juridical order. Thus, the state of exception is neither external nor internal to the juridical order and the problem of defining it concerns precisely a threshold, or a zone of indifference, where inside and outside [the juridical order] do not exclude each other but rather blur with each other (Agamben, 1998, p. 15). It therefore creates an ?ambiguous, uncertain, borderline fringe, at the intersection of the legal and the political? (Fontana cited in Agamben, 2005, p. 1). 22 Law (Greek: ?????, n?mos; pl. ?????, n?moi). (Nomos. Retrieved from Wikipipedia: http://en.wikipedia.org/wiki/Nomos) 72 It was precisely this ambiguity of the exception, a threshold which, Schmitt argues, ?defies general codification? (cited in Agamben, 1998, p. 15), which precipitated the multiplication of explanations of the conflict of the 1990s. This struggle of ontological inclusion of the exception in the juridico-political and the multiplication of explanations which it evoked led South African analysts in the end to take ?refuge in arguing that?there are ?many truths?, that violence is best understood as multi-faceted? (Taylor and Shaw, 1998, p. 13). The Structure of the Exception, an Inclusive Exclusion These struggles of ontological inclusion were, in part, the consequence of the complexity of the structure of the exception which articulates the biopolitical and juridical, law and life. The nature of this articulation, between the biopolitical and the juridical, in the structure of the exception, is not a simple exclusion, but is instead a type of exclusion where what is excluded is retained in relation to that from which it has been excluded, i.e. it is, ?taken outside and not simply excluded? (Agamben, 1998, pp. 17-18). ?What is excluded, in the exception, maintains itself in relation to the rule in the form of the rule?s suspension? (Agamben, 1998, p. 19, own emphasis). This process, which Agamben calls ?inclusive exclusion?, where that which is excluded is paradoxically included in the juridico-political order through the very process of exclusion, takes place within a context of the suspension of juridico-political order in which the law employs the exception ?as its original means of referring to and encompassing life? (Agamben, 2005, p. 1). Therefore the state of exception is ?not the chaos that precedes order but rather the situation that results from its suspension? (Agamben, 1998, p. 18). Thus the state of exception as a paradigm of government does not concern the collapse of juridico-political order but is ?a suspension not simply of the administration of justice but of 73 the law as such? (Agamben, 2005, p. 41). It is a paradoxical legal institution which ?consists solely in the production of a juridical void? (Agamben, 2005, pp. 41-42) in response to periods of political crisis. These are juridical measures, therefore, which cannot be understood in legal terms and must be understood ?on political not juridico-constitutional grounds? (Agamben, 2005, p. 1). Hence, ?the impossibility (common to both ancient and modern sources) of clearly defining the legal consequences of those acts committed? (Agamben, 2005, p. 49) during this suspension of juridico-political order. The principle of the state of exception is the creation of a situation that makes the application of the norm possible. Agamben identifies that the juridico-political order contains an essential fracture between norm and its application, which in extreme situations can only be filled by means of an exception, the creation of a juridical void. Schmitt therefore distinguishes between two critical elements of law, namely, the norm of law and the decision on its concrete application. The one does not derive automatically from the other. This means that the norm ?can be suspended, without thereby ceasing to remain in force? (Schmitt cited in Agamben, 2005, p. 36). The state of exception in suspending the juridico-political order is the point at which this distinction between norm and application is most explicit, as it separates out the norm of law (the juridico-political constitution) from its actual application by suspending the constitution and allowing it to remain in force without actually being applied. In times of political crisis, the state of exception separates out the norm from its application in order to introduce ?a zone of anomie into the law in order to make the effective regulation of the real possible? (Agamben, 2005, p. 36), i.e. to make the application of the norm, the juridico-political order, possible again. 74 Thus under conditions of what Schmitt identifies as ?commissarial dictatorship?, the constitution is suspended ?in concreto in order to protect its concrete existence? (Schmitt cited in Agamben, 2005, p. 33), i.e. it is suspended but remains in force, making possible the protection of the constitution and the creation of a situation in which law can be applied again. On the other hand, as in the period of negotiated transition in South Africa, in the context of what Schmitt identifies as ?sovereign dictatorship?, this situation is not limited to simply suspending the existing constitution in terms of provisions provided for in the constitution itself; ?rather it aims at creating a state of affairs in which it becomes possible to impose a new constitution? (Agamben, 2005, p. 33). While modern state power may, as argued by Agamben, be constituted by continual evocations of a juridical state of exception, this structure of inclusive exclusion in the relation of exception is in fact the originary principle that first makes possible the space in which the wider juridico-political order can have validity. This anomic space without law, is in fact constitutive of the juridico-political machine as a whole, without it juridico-political order cannot exist. Thus, in its archetypal form, the state of exception is therefore the principle of every juridical localisation, since only the state of exception opens the space in which the determination of a certain juridical order and a particular territory first becomes possible. (Agamben, 1998, p. 19, own emphasis) ?Pure? Violence, the Object of Politics Importantly, it is this fracture between norm and application that makes it possible to distinguish two critical elements of the juridico-political order, what is reflected in Western jurisprudence as the notion of the ?force of law?, which separates out law?s applicability from its formal essence, ?whereby decrees, provisions, and measures that are not formally laws 75 nevertheless acquire their ?force?? (Agamben, 2005, p. 38). Critically, ?in extreme situations ?force of law? floats as an indeterminate element that can be claimed by both the state authority (acting as a commissarial dictatorship) and by a revolutionary organisation (acting as a sovereign dictatorship). The state of exception is an anomic space in which what is at stake is a force of law without law? (Agamben, 2005, pp. 38-39). It was in the contest to claim this force, this indeterminate source of power, radically separated from law, that the violence of the 1990s took place. As Agamben argues, this struggle for anomie seems to be as decisive for Western politics as...?the battle of giants concerning being?, that defines Western metaphysics. Here, pure violence as the extreme political object, as the ?thing? of politics, is the counterpart to pure being?as the ultimate metaphysical stakes; the strategy of exception, which must ensure the relation between anomic violence and law, is as critical as capturing pure being for metaphysics. (Agamben, 2005, pp. 59-60) Importantly, however, Agamben argues that the idea of force of law without law is essentially a fiction, ?through which law attempts to encompass its own absence and to appropriate the state of exception, or at least to assure itself a relation with it? (Agamben, 2005, p. 51).While the idea of force of law as a way of establishing a relation between anomie and law may be essentially a fictive categorisation that attempts to capture something like ?a mystical element?, what is in fact at issue in these categorisations, is, Agamben asserts, ?nothing less than the definition of what Schmitt calls ?the political? (Agamben, 2005, p. 51). It is in defining the modes of relation of this anomie to the law that the meaning of ?the political?, the exception, on the boundaries of the law, can become explicit, ?and perhaps only then will it be possible to answer the question that never ceases to reverberate in the history of Western politics: what does it mean to act politically?? (Agamben, 2005, p. 2). 76 This anomic violence, what Benjamin has called ?pure violence?23, which neither makes nor preserves law, is what is at issue in the struggle over the state of exception and the struggle to inscribe this anomie within the law. Critically, this violence which occurs outside law ? unlike ?mythico-juridical? violence which is always a means to an end, this ?pure? violence, ?is never simply a means ? whether legitimate or illegitimate ? to an end (whether just or unjust)?. It is instead, a ?pure medium?, ?mediality without ends? (Agamben, 2005). Agamben explains, The medium does not owe its purity to any specific intrinsic property that differentiates it from juridical means, but to its relation to them?pure violence is that which does not stand in a relation of means toward an end, but holds itself in relation to its own mediality? pure violence is attested to only as the exposure and deposition of the relation between violence and law.? (Agamben, 2005, p. 62) It was in this context of a deposition of a relation between law and violence that the ?pure? violence of the 1990s was exposed and could only be perceived paradoxically in its very non- relation to law. This violence in non-relation to law was not a means to an end, but instead a pure medium, what Benjamin calls a ?manifestation? (Benjamin cited in Agamben, 2005, p. 62). While the violence that took place within the South African context, as Du Toit (1993) argues, had always, even in attempting to overturn the law, sought to maintain a relation with it, ?pure violence exposes and severs the nexus between law and violence and can thus appear in the end not as violence that governs or executes?but as violence that purely acts and manifests? (Agamben, 2005, p. 62). 23 It is important to note that Benjamin?s conception of ?purity? was about the relation of violence to law, rather than a substantive characteristic of the violent action itself. Benjamin argues, ?The task of a critique of violence can be summarized as that of expounding its relation to law and justice? (cited in Agamben, 2005, p. 61). 77 Narrating an Original Site of Power The multiplication of explanations of the conflict of the 1990s thus concerned not only the ambiguity of the exception itself but the complexity of its relation with the juridico-political order, one of inclusive exclusion, which implied a completely different relation between violence and the juridico-political order than South African theory had previously conceived. Analysts working within a juridical paradigm could only perceive this relation between anomie and the juridical-political in instrumental terms, in which the basis of the state?s authority rested on its ability to monopolise the means of violence, a conception of sovereign state power, in which was embedded ?a long standing philisophico-political tradition according to which violence was to be conceived as instrumental in nature, that is, as a means or implement to put to the service of (political) ends? (Hanssen, 2000, pp. 18-19). Therefore while a number of South African analysts attempted to analyse what they identified as the ?violence-negotiation nexus? (Everatt, 2003; Sisk, 1995), they could only conceive of this ?nexus? in instrumental terms. As Sisk argued, ?political violence was a tactic used by various actors as a ?beyond the table? tool in pursuit of specific political aims? (Sisk, 1995, p. 120, own emphasis). Or as Schutte put it, ?politics is basically a bargaining process and violence is one of the means available to improve one?s position within the process? (Schutte, 1998, p. 14). The metaphor of the ?hidden hand? used by numerous analysts to refer to covert violence (Everatt, 2003; Shutte, Liebenberg & Minnaar, 1998) anthromorphosised this functionalist understanding of violence. However, the anomic violence of the 1990s, which could no longer be seen in instrumental relation to the apartheid state, whose ?functions and purposes? consequently remained oblique precisely because they could not easily be ?derived from? or ?related to? ?primary 78 social processes and political phenomena? or be ?readily harnessed to an intelligible public cause? (A. du Toit, 1993, p. 6) therefore did not fit into the ??master narrative? structuring conventional understandings of political violence? (A. du Toit, 1993, p. 1), namely a ?pre- figured narrative by which violence is appropriated to the domain of language? and given meaning (Thornton, 1995, p. 14). The lack of instrumental rationality, the ?pre-figured narrative? in terms of which political violence had previously been understood in the South African context, led to the struggles of comprehension among South African analysts of the time. This instrumental conception of power which resides within the sovereign state alone relies on a separation between a prior origin of power, as a thing which can be usurped in its totality, and the effects of power (violence), ?which is reduced to a transparent instrument? (Feldman, 1991, p. 3) that simply requires instrumental explanation and rationalisation to trace the relation between cause (power) and effect (violence). This led to an intensive ?metatheoretical? debate at the time of the violence during the 1990s, a conflict over the nature of the conflict (P. du Toit, 2001; Horowitz, 1991), which was in fact, in the absence of an interrogation of the basis of sovereignty as the monopoly over the means of violence, a conflict over the nature of the relation between cause and effect. As Thornton wrote at the time ?the accounts we most often give of violence either imply or state explicitly that violence is the consequence of previous events or that it is the cause of other events? (Thornton, 1995, p. 1). Consequently almost all attempts to ?explain? the violence of the early 1990s, were, as Sue van Zyl argued, additive, each analysis attempting to find the additional ingredients that could explain the relation between a prior cause of power and its violent effects, in terms of a 79 variety of historical, political or sociological accounts, or a ?judicious mixture of each? (Van Zyl, 1990, p. 2), which could explain the nexus between violence and power. This nexus, however, as Agamben argues, is essentially a fiction through which the juridico-political order attempts to establish a relation with the anomie that forms its boundary. Du Toit (1993, p. 3) thus wrote, The very nature and purpose of this proliferation of political violence is intensely controversial: it is hotly disputed on all sides whether this proliferation of political violence should be understood as ?ethnic conflicts with deep cultural and historical roots?, or as a power and ideological struggle between contending political organisations, or as the sinister work of a ?third force? behind the scenes, or as a consequence of poverty, social disruption and the general lack of political authority, or as some combination of all of these. The illusory nature of this search for an original atemporal origin of power was, however, made explicit by the temporality of the conflict itself and the complexities of the identities of its participants. As quickly as analysts developed causal explanations based on the unfolding of ethnic tensions (Horowitz, 1991), collective identities based on political affiliation were claimed by antagonists in struggle. Later further explanations and causes were ?discovered?, adding layers and layers to an analysis that attempted to incorporate both the unfolding trajectory of conflict and the interpretive work of analysts. This would, if all potential causal explanations could be finally exhausted, purportedly ?explain? violence and reveal its ?original? truth. In fact what analysts were doing was describing the fleeting articulation of power in terms of a particular manifestation of ethnicity, generational cleavage or economic division. Thus the attempt to constitute a fictive relation of linear causality was increasingly contested by the unfolding trajectory of the conflict itself, what many analysts referred to as the ?self- perpetuating? nature of violence (Coleman, 1998; Simpson, 1993; Taylor & Shaw, 1998). 80 However, over time, as Feldman argues, ?political action and institutionalised ideology? begin to ?form two discontinuous, significative systems?, as relations of antagonism?carve out autonomous material spheres of effect and affect that diverge from formal political rationalities...political violence is no longer fully anchored in ideological codes and conditions external to the situation of enactment and transaction. Political enactment becomes sedimented with its own local histories that are mapped out on the template of the body. (Feldman, 1991, p. 4) In this vein Marks (1995) found discontinuities between the ?self-identity? (p. 9) of participants in the ANCYL in Diepkloof who had been involved in violent action and the ?identity? of the ANCYL as a social movement organisation. Similarly Segal (1991), in her interviews with hostel residents on the East Rand, noted the discrepancy between the ?perceptions of those involved in political violence, and the coherent explanation of their actions advanced by political organisations? (p. 2). Marks (1995), implicitly assuming an essential contiguity between political agent and institution, understood the discontinuities between the formal political rationalities of the ANC and the actions of ANCYL members in terms of a breakdown of the ?ideological cohesion? (p. 211) of the ANC, which had in the past anchored the actions of its political agents to the formal political programme of the organisation but in the context of the flux of the transition period had been unable or unwilling to maintain this welding of agent to institution. Segal (1991) responded to this theoretical problematic with the truism that ?these explanations of the violence from the people themselves are often contradictory?because there are a range of traditions and ideologies (themselves often contradictory), which may operate simultaneously in an individual?s life? (pp. 2-3). These explanations, which located the emergence of the authoritarian subject as a consequence of a ?breakdown of ideological leadership?, could not recognise the biopolitical nature of the authority they were confronted with. Critically, the nature of the power that 81 emerges in the context of suspension of law, the context of exception, is a type of power that is not invested in formal office but in fact exceeds the rights of formal office, a power that is outside the law and inside every citizen, regardless of formal rank: ?every citizen appears to be invested with a floating and anomalous imperium that resists definition within the terms of the normal order? (Agamben, 2005, p. 43). Agamben argues that it is the conflation within the modern concept of sovereignty of biopolitical authority (auctoritas) and written law (potestas) which has been ?the cause of the philosophical inconstancy in the modern theory of the state? (Fueyo cited in Agamben, 2005, p. 75), as it is premised on an assumed dichotomy between the representation and instantiation of power. While auctoritas attaches itself to the body and ?springs from the person, as something that is constituted through him, lives only in him, and disappears with him? (Heinz cited in Agamben, 2005, p. 82), potestas constitutes the pre-established form of legal office into ?which the individual enters?and which constitutes the source of his power? (Heinz cited in Agamben, 2005, p. 82). Therefore auctoritas, unlike the power exercised through magisterial office, ?seems to imply not so much the voluntary exercise of a right as the actualisation of an impersonal power (potenza) in the very person of the auctor? and ?has nothing to do with representation? (Agamben, 2005, p. 77). Other analysts attempted to hierarchise, to order various causes, triggers, precipitating factors, all inevitably informed by their own intellectual biases, lending more weight to some factors than others, some emphasising ethnicity (Horowitz, 1991) above political affiliation (Aitchison, 2003; Guelke, 1999; Kentridge, 1990; Olivier, 1994), or economic tension (Bonner & Ndima, 1999; Heribert & Moodley, 1992; Morris & Hindson, 1992; Sapire, 1992) above political identity. These hierarchies were, of course, infinitely contestable, on conceptual and empirical grounds. It was impossible to ?prove? the claim that much of the 82 tension was generated by ethnic tension, any more than it was possible to prove the primacy of the ?political? in the conflict. Such proof rested on the demonstration of a causal relationship between power and its effects, which remained both empirically and conceptually elusive. Valuable historiographical work on the conflict (Bonner & Nieftagodien, 2001; Khosa, 1991; Sapire, 1992; Sitas, 1996), which implicitly sought causality in history, added nuance and texture to explanations of the conflict which contested some of the most simplistic causal explanations for violence. For example Bonner and Ndima disputed the media and state characterisation of the conflict as the result of primordial antagonisms between ?Zulus? and ?Xhosas?, pointing out that in the 1940s and 1950s the main axes of ethnic conflict on the Witwatersrand were the South Sotho on one side and either Mpondo or Zulu on the other. There are almost no recorded group clashes between the Zulu and either the Mpondo or the Xhosa. (Bonner & Ndima, 1999, p. 15) Sitas investigated how the differential meanings invested in the concept of ethnicity among hostel residents in the 1980s and the 1990s contributed to the perpetuation of violent conflict: in the earlier period lineage, language, values and expressions that were deemed to be ?ethnic? were used as threads to weave the solidarity of a social movement, unionism. In the 1990s these value components became the defining features of division and strife. (Sitas, 1996, p. 246) Segal provided a valuable corrective to essentialist characterisations of Zulu hostel residents as innately aggressive and ?war-like? by conducting a series of interviews with hostel residents on the East Rand. Read against a historiographical background of the hostel community and the migrant labour system, Segal sought to understand the ?worldview? of these participants in the conflict, their particular ?truth?, by detailing the complexity of their historical and contemporary relationship with township communities and the continued 83 salience of ethnic identity, ?which is not merely crafted from above by an organisation like Inkatha? but ?[is] an easy mobilising agent, and is open to manipulation, precisely because it has a continued reality in popular consciousness? (Segal, 1991, p. 22). Sapire?s (1992) research investigated the significance of Xhosa ethnicity in a range of informal settlements on the Witwatersrand as Xhosa ethnic identity emerged as a significant new source of social tension in the conflict of the 1990s. Sapire documented how there had been an exponential growth of Xhosa immigrants to the Witwatersrand during the mid to late 1980s. Of particular significance was the migration of a number of Xhosa speakers from the former ?homelands? of Ciskei and Transkei to Phola Park informal settlement on the East Rand. Many of these migrants occupied sites vacated by households who moved to the new site and service settlement of Zonkezizwe after 1990 (Sapire, 1992, p. 687). Sapire therefore concludes that, ?the numbers and social impact of Xhosa-speakers rose significantly in 1990 due both to direct immigration from the rural areas and to the significant movement into informal settlements of Xhosa speakers from hostels, backyard shacks and elsewhere within the urban environment? (Sapire, 1992, p. 695). Bonner and Ndima?s (1999) work investigating and demonstrating the ?origins? of the East Rand conflict in a localised economic dispute between rival taxi associations in Katlehong attempted, through the recovery of a hidden historical narrative, to contest the dominant narrative of conflict in this sub-region, a narrative which also implicitly supported a broader thesis about the central role of the IFP as the main ?perpetrator? in the contention of the post- 1990 period. While previous analyses of the violence had dated its ?genesis? on the East Rand to the initiation of a violent political recruitment campaign by the IFP from August 1990, Bonner and Ndima showed how this phase of conflict had been preceded by a violent taxi conflict in Katlehong between a taxi association serving the needs of migrant hostel 84 residents and a rival taxi association, which developed a close association with the urban and politically militant student organisation Congress of South African Students (COSAS). The IFP as a political formation was significantly absent from these processes of contention, therefore Bonner and Ndima were able to add complexity to an analysis which had until then tended to caricature the IFP as the progenitor of all conflict on the East Rand, and the East Rand township community as the passive recipients of this violence. This line of investigation yielded much useful description of the ways in which violence unfolded, the various forms it took. However, despite the putative intention of analysts to discover the ?why? of violence, in fact what they were doing was exploring the continually unfolding, shifting, malleable, how of power, of political conflict. This became a substitute for the why of violence in an analytical paradigm, which, in assuming a juridical conception of power, could not in fact admit this why of violence, the articulation between juridico- institutional and biopolitical power. Trapped in a model of causality which assumes that a relation of causality can be found, analysts remained enmeshed in a debate around the primacy of one causal factor over another. However as analysts increasingly faced the aporia of the causal line of investigation of the violence, as the ?violence developed a momentum of its own?, its ?patterns? changing over time, as ?the effects of earlier factors, in a sense, become new causes? (Schlemmer, 1991, p. 2), a consensus emerged among some analysts about the ?multi-causal? nature of the violence, an explanation that recognised that ?the search for mono-causal explanations is fruitless? (Simpson, 1993, p. 2) but which could not conceptually resolve this theoretical problematic within a paradigm of causality. These analysts instead offered lists of descriptive factors already identified in previous analyses, ranging from ethnic identity, generational 85 tensions, conflict over scarce resources, psycho-social factors, the failure of the state in its duty of protection, or deliberate misuse of its security force, etc., etc., concluding finally that ?none of these descriptions is completely inaccurate. Yet none, on its own, will properly explain this complex situation? (Simpson, 1993, p. 2). Schlemmer, who organised these factors into categories of ?general background conditions?, ?predisposing factors? and ?triggering events?, therefore concluded that ?short-term policies cannot be expected to address the diversity of factors which form ?chains? of causation? and a variety of responses was required at each level of causation, whether to address ?background conditions?, ?predisposing factors? or ?triggering events? (Schlemmer, 1991, p. 2). However, as Feldman argues, drawing on Nietzsche?s critique of the relation between cause and effect and his positing instead of ?a performance theory of power? (Feldman, 1991, p. 3) in which power is the ?simultaneous site of origin and effect? (Feldman, 1991, p. 3), this separation of the origin of power from its effects is as in fact a ?mythic and dramaturgical structure within which actions in time are endowed with a singular atemporal origin? (Feldman, 1991, p. 3). Thornton therefore argues that if we can only know power through its effects, ?it cannot be true to say that ?power? causes violence or that violence is the instrument of power?these relationships are always?constructed? (Thornton, 1995, p. 5). Nevertheless this ?perspectivist illusion? (Feldman, 1991, p. 3) remains deeply embedded in formal political rationalities that assume that ?power distributes itself from some place external to its effects, external to its violence? (p. 2-3) and ?legitimation resides in the construction of a fictive depth? (p. 3), an illusion of cause and effect. This separation of origin and effect makes possible a realm of ideological reasoning and instrumental legitimation of power, ?a counter world of imaginary doubles, agents, and metaphors of 86 moral legitimation ? which function as the originary sites of power, of which the act is but the derivative symptom? (p. 3), drawing ?consciousness away from the concrete material investment in acts and effects that reproduce domination in time and space? (p. 3) which inhibits and denies the ?semantic and material autonomy? of action (p. 3). Thus, violence does not come about as the ?realisation of intentions or the attempt to make certain concepts concrete, violent acts are often meaningfully constituted only at the moment of their commission? (Thornton, 1995, pp. 8-9). The violence of the 1990s, which could not ?be understood in relation to familiar criteria of legitimacy and rationality? (A. du Toit, 1993, p. 2) therefore presented a profound problem of comprehension in terms of the paradigm of legitimation rationality. Segal noted ?the need to impose an official sounding order on the overwhelming confusion and horror of violence, the need to create light where in fact there is none? (1991, p. 2). Van Zyl, recognising the limitations of this instrumental paradigm, argued for a distinction between a ?strategic? violence, which is open to moral or practical criticism but which does not call for explanation as ?rationally chosen action? that is ?self-explanatory? in terms that are available to the actors involved, and violence, which is ?symptomatic?, ?political in origin but not politically motivated? (1990, pp. 1-2, own emphasis). As Johnston noted of the conflict in KwaZulu-Natal: ?despite its clear political overtones, the violence?is relatively ?inarticulate? in that no organized participant has an acknowledged paramilitary strategy? (Johnston, 1997, p. 84). Similarly Segal, who interviewed hostel residents on the East Rand, did not find that participants in the conflict articulated strategic political motivations The interviews conducted reveal a shifting assortment, a ?kaleidoscope? of explanations. The apparent incoherence of the explanations at times indicates that the 87 view of violence as constructed from below, the human face of violence, is far more diffuse and complex than most media or political accounts portray. (Segal, 1991, p. 2) Van Zyl?s (1990) analysis distinguishing the subject?s articulated political motive from the substance of the political offered the possibility of a different reading of the ?political? as the place of the exception, on the boundaries of the juridico-political order. However this was largely lost in a South African conceptual paradigm in which ?the political? was conceived in terms of an instrumental relation between cause and effect within the juridico-political order, namely, an articulated political intent linked directly to an act carried out in furtherance of these stated intentions. Pierre du Toit (2001), analysing the political violence of the 1990s, defined it as ?harm inflicted primarily from political motives? (p. 116, own emphasis), a relation between political intention and violent act that would later be institutionalised within the TRC. Therefore, while, as Feldman argues, the separation between the cause and effect of power is constituted through a ?mythic and dramaturgical structure? through which manifestations of power in time are given a singular atemporal prior origin, violence is in fact ?peculiarly temporal? (Thornton, 1995, p. 2) and profoundly unpredictable; ?the proximate causes of violence are often so complex that it is rarely possible to know exactly or precisely what triggered a violent event? (Thornton, 1995, p. 9). It is instead a case of ?passing epiphenomena? (A. du Toit, 1990, p. 119), which ?necessarily disturbs all structural causal or narrative sequences and continuities? (Thornton, 1995, p. 3) and therefore cannot be properly instrumental. We can know of the possibility or even the probability of violence but in fact we can only really know violence in retrospect (Thornton, 1995), and it is in this retrospective perspective that we construct a relation between a prior cause of power and its violent effects. 88 Therefore, these fleeting manifestations of power are converted, through our own retrospective narrative constructions, into ?metaphorical and metonymical structures?, reconstituting what would otherwise be ?passing epiphenomena as a project of sustained intentionalities? (A. du Toit, 1990, p. 119). The past ? the ?origin? of the effects of power, the violent event ? is retrieved in the present and transformed into ?a social text? (Apter, 1997, p. 13), ?a logic of outcomes? (Apter, 1997, p. 12), which ?seems to emerge from the facts of the narrative, its truth value irresistible? (Apter, 1997, p. 14). The ?inexplicability? of violence, and our ?perplexity? (Thornton, 1995, p. 3) about it is hence rendered ?explicable? (p. 3 ) in an ?explanatory or narrative account that virtually all of the social sciences and history offer? (p. 3), which, however, in its dependence ?on some notion of spatial or temporal continuity? (p. 3), is essentially a fictive construction of a relation between cause and effect. We can, in fact, ?only speculate about causes, which are, in any case, rarely efficient, never unitary, and very often not recoverable from evidence and memory? (p. 5). What we can see and analyse are the consequences of violence, namely the production of the biopolitical body, implicitly acknowledged in South African analyses of the conflict, which relied on the ?numbering of the dead? (Aitchison, 1988) by a variety of monitoring agencies24. However, the signification of these bodies was profoundly contested. It was in the denotation of these bodies by monitoring agencies and other analysts that ?violence comes to have public meaning. Incidents, victims and sometimes perpetrators come to acquire their political labels?Situations of conflict in hostels, squatter settlements, townships or tribal 24 These included most significantly, the Human Rights Commission, the Community Agency for Social Enquiry, the Centre for Adult Education at the University of Natal and the South African Institute of Race Relations. 89 areas are ranked as wholly or partly ?political? by inclusion, qualification and/or omission? (Johnston, 1997, p. 82). However, this process of signification was articulated in South African discourse as a debate about an obfuscated empirical ?truth? that could be resolved through the application of ?objective? methodology. As Horowitz put it, ?the analyst who aims to ameliorate the conflict is confronted with the difficult task of combining relentless pursuit of the truth about the conflict ? and persuasion of the participants that it is the truth? (Horowitz, 1991, p. 30). The task of analysis in this paradigm is to ?unmask? and recover this origin of power, which is also the source of ultimate truth about power25. Therefore the process of inscribing violence within narrative, culture and juridico-political order creates a ?metaphor of causation? (Thornton, 1995, p. 15) that we construct through a conception of intention. It is the intention which welds violence to power in a causal relation: Political violence is violence by virtue?of the violence it intends?a narrative must be constructed which will provide the satisfying explanatory linkage to ?power?. If this linkage?is not constructed, or cannot be constructed, then violence is not ?political?. (Thornton, 1995, p. 16) The Subject, Negotiating the Relationship between Cause and Effect The notion of intention relies on the enlightenment conception of the sovereign subject. And it was this conception of the sovereign subject which established the terms on which the 25 The paradigmatic example of this was a report, Spotlight on disinformation about Violence in South Africa by Anthea Jeffrey of the South African Institute of Race Relations, who in an analysis of the reports of Amnesty International, the International Commission of Jurists and the South African based Human Rights Commission, charged these organisation with, ?disregard?for the rules of evidence and the safeguards of due legal process?(Jeffrey, 1992, p. 9), which meant that the legal principle of audi alteram partem (hear the other side) should be applied in the compilation of research reports, adding that, ?newspaper and other reports are admissible in a court of law only if the author is available to testify as to the truth of the report?(Jeffrey, 1992, p. 17 own emphasis). Implicit in the audi alteram partem principle was a conception of a juridico-political subject rather than a biopolitical subject, who can expressly articulate his or her ?side? of the story. 90 subject would be understood both in the analysis of the political conflict of the 1990s and the subsequent juridico-political inscription of this violence. Analysis of the conflict of the 1990s therefore frequently attempted to mediate the relationship between cause and effect in terms of an implicit assumption of singular subjecthood and instrumental agency among its protagonists. Johnston, reviewing analyses of the conflict, identified this as ?a conceptual orientation towards interpreting the violence in terms of adversarial political groupings and assigning agency to the conscious strategies of political leadership? (Johnston, 1997, p. 83). In this paradigm the task was to discover an essential original subjecthood, the ?true? identity of the subject, which may have become masked or overlaid by other identities, but which the analyst?s patient work could in the end reveal. The paradigmatic example of this notion of the subject was the ?comtotsi? who knowingly masked his or her identity beneath the cloak of political subjecthood, exploiting violent conflict for personal gain in concordance with an original, essential ?criminal? identity, motivated thus by ?selfish/criminal purposes as opposed to political motives? (Minnaar, 1994, p. 1, own emphasis). Minnaar wrote therefore of: the so-called ?comtsotsis? ? those who call themselves ?comrades? (amaqabane), but are in fact ?tsotsis? (thugs/criminals) who hide behind political slogans, use political activities and the name of specific political organisations to gain power or control of a community, but remain just that ? ?comrade criminals?. (Minnaar, 1994, p. 3) Kynoch wrote similarly of ?criminal opportunists?, who ?capitalized on political rivalries to justify their actions on ideological grounds?the term ?com-tstosi? was coined to describe self-appointed comrades who utilized violence for personal gain? (Kynoch, 2005, p. 496). 91 Minnaar implicitly articulated the aporia of these efforts to adjudicate the ?political? nature of conflict through an unmasking of the original subjecthood of its actors as he wrote: it has become very difficult to clearly define what political violence or criminal violence are and whether they are two distinct types of violence or have become interchangeable, i.e. is there any difference between them other than being labelled political or having political motives as opposed to selfish/criminal purposes? (Minnaar, 1994, p. 1) Kynoch on the other hand sought to demonstrate historiographically the ?criminal? dimensions of the ?political? conflict of the 1990s and to investigate the extent to which the violence was ?primarily political? (Kynoch, 2006, p. 8, own emphasis) by looking at the continuities between this period of violence and the ?culture of violence?, which, he argued, ?was already ingrained in township society? (Kynoch, 2005, p. 495). Therefore, ?violent conflicts were frequently instigated by parochial power struggles that acquired larger political dimensions only incidentally when the state, ANC or Inkatha become involved? (Kynoch, 2005, p. 496, own emphasis). Bonner and Nieftagodien similarly critiqued the TRC?s analysis of the conflict on the East Rand during the 1990s, arguing that, as a result of its mandate, it had cast much of the conflict as ?political? in nature, when in fact ?criminal activities may well have been responsible for the largest part of the violence in late 1992 and early 1993? (Bonner & Nieftagodien, 2002, p. 191). This conception of a hierarchy of subjecthood exercised instrumentally also permeated other explanations of the conflict, which, for example, sought its origins in the expression of an essentialist ethnic identity. Horowitz therefore argued that ethnic affiliations are kept in the ?attic? of the mind to be ?dusted off? when the ?home group? is under threat (cited in Shaw, 1997, pp. 47- 48). Bonner and Ndima similarly contended that on the East Rand 92 ?ethnic stereotypes which had lain dormant in individual consciousnesses were now reactivated, collectivised and sharpened? (Bonner & Ndima, 1999, p. 15). Segal notes that in the politically volatile climate of the period the continuities and stability offered by ?older? forms of identity, made ethnic forms of solidarity extremely appealing: ?it is the depth and endurance of ethnic identities, relative to other forms of allegiance, that brings them to the fore in times of crisis? (Segal, 1991, p. 24). This notion of the sovereign self has, however, been the subject of extensive postmodern critique. As Markell writes, influential late twentieth-century critiques of the so-called ?sovereign self??focused on the widespread image of the human being as paradigmatically the owner of private property, with an exclusive right to use and dispose of things under his dominium, which seemed to underwrite mainstream contractarian approaches to political thought; others focused on model of the self as ?unemcumbered? or ?atomistic?, which seemed to treat human beings as somehow existing above and acting independently of their social and historical contexts and bodily matrices. (Markell, 2003, p. 11, original emphasis) Thus, ?contractarian political theories and voluntarist conceptions of the self anchor sovereignty in the notion of choice? (Markell, 2003, p. 12, original emphasis). In opposition to this notion of voluntaristic agency, Markell posits Hannah Arendt?s ?rejections of the aspiration to achieve sovereign agency?, which she calls ?contradictory to the very condition of plurality? (Arendt cited in Markell, p. 13) and to the unpredictability to which that condition gives rise (Markell, 2003, p. 13). For Arendt, the history of Western philosophy is shot through with misguided efforts to escape the condition of non-sovereignty. (Markell, 2003, p. 13). In his critique of the relation between cause and effect within relations of domination, Nietzsche therefore ?disrupts neat, unmediated relations of linearity between institution and 93 agent? (Feldman, 1991, p. 3). Critically, Feldman argues, Nietzsche does this in order to ?arrive at a theory of political subjects? (p. 3). This is a political subject who does not exist prior to action, who is not the ?unified and underlying originator of actions and values? (p. 4), the ?transcendental subject? of the enlightenment. Radically, Nietzsche instead reversed the ?authorial positions between act and agent? (p. 3) and in this overturning, agency becomes ?not the author but the product of doing? (p. 3) and as a result, ?power is embedded in the situated practices of agents? (p. 4), practices which can alter the subject themselves. As a result, Feldman argues, ?political agency becomes the factored product of multiple subject positions. There can be no guarantee of a unified subject, as actors shift from one transactional space to another and from discourse to somatic practice? (Feldman, 1991, p. 4). Because agency is embedded in the praxis of actors themselves, rather than being extrinsic to its site of instantiation within the subject, political agency becomes, not an instrument of an original source of power, but an ?embodied force?, a biopolitical power, in which ?the political subject, particularly the body (is) the locus of manifold material practices? (p. 1). Thus, ?ethnicity and ideology are literally sunk into the material density of the body? (p. 71) which becomes the locus ?for the elimination and purification of political substance? (p. 71) and ?the violent dematerialization of the body as the prescribed site for the lodgement and dislodgement? of the ideologically and ethnically coded corpse (p. 71). The biopolitical subject Thus if subjecthood is literally sunk into the material density of the body, it concerns ?life? itself. Agamben approaches this problematic through the ancient Greek distinction between different forms of ?life?, a word for which the Greeks had no single term. This distinction involved a differentiation between zo?, ?which expressed the simple fact of living common to 94 all living beings (animals, men, or gods)?, and bios, which was ?not at all simple natural life but? a particular way of life? (Agamben, 1998, p. 1); critically bios was politically qualified life. In the classical world, simple natural life, zo?, is excluded from the polis or political life. This distinction, articulated originally by Aristotle as an opposition between the simple fact of living (to z?n) and politically qualified life (to eu z?n), was to become the canonical foundation of the Western political tradition, delineating two conceptually distinct realms, the juridico-political and the organic fact of life itself. It was not until Foucault introduced the concept of biopolitics that the intersection between these two realms was even contemplated. For Foucault, ?the process by which natural life begins to be included in the mechanisms and calculations of State power, and politics turns into biopolitics? (Agamben, 1998, p. 3) was the defining event of modernity. ?For millennia...man remained what he was for Aristotle: a living animal with the additional capacity for political existence; modern man is an animal whose politics calls his existence as a living being into question? (Agamben, 1998, p. 3). Taking up Foucault?s project of examining the intersection between natural life and political life but rejecting his teleology leads Agamben to posit this intersection, the inclusion of ?bare life? in the political realm and ?the production of the biopolitical body? as the ?original activity of sovereign power?, ?the secret tie uniting power and bare life? , which the ?modern State?does nothing other than bring to light? (Agamben, 1998, p. 6). If the inclusion of natural life in political life is not simply a phenomenon of modernity, preceded by a long age in which zo? and bios remained distinct, ?it will be necessary to reconsider the sense of the Aristotelian definition of the polis as the opposition between life (z?n) and good life (eu z?n). The opposition is in fact, at the same time an implication of the 95 first in the second, of bare life in politically qualified life? (Agamben, 1998, p. 7). Thus the political is not a separate realm to natural life, rather, ?what had to be politicized [was] already always bare life? (Agamben, 1998, p. 7). Therefore politics is not about an opposition between life and a ?supplement? or an ?additional capacity? to the simple fact of living (zo?) shared by all living beings, but instead an implication of this fact of living in politically qualified life. Instead of an instrumental relation between zo? and the polis, simple life and a supplement of the political, a different form of relation constitutes the political, critically a relation of exception, a relation which attempts to effect a space between ?bare life? and the ?good life? within life, not in addition to or as an attribution of life. Politics therefore is the realm in which this relation is constituted, it is the terrain in which bare life (zo?) is transformed (not added to or supplemented) into a different form of being, namely ?bios?, the good life. Politics is therefore about a threshold where the ontological relation between living being and political being, between zo? and bios, is articulated. This relation, this threshold of transformation, is maintained in terms of an exception, an inclusive exclusion. Homo sacer Critically, it is the figure of homo sacer (sacred life) that embodies this ontological relation of exception that constitutes the political, a human figure which exists in an ambiguous zone of indistinction between zo? and bios, a liminal status between simple, animal life and political 96 life, analogous with the Germanic figure of the bandit or friedlos (without peace) who when banned from the city becomes a ?wolf-man?, not a piece of animal nature without any relation to law and the city?rather, a threshold of indistinction between animal and man, physis and nomos, exclusion and inclusion: the life of the bandit is the life of the loup garou, the werewolf, who is precisely neither man nor beast, and who dwells paradoxically within both while belonging to neither. (Agamben, 1998, p. 105, original emphasis) It is in the zone of indistinction between the animal and the human, the zone of exception, that Hobbes? ?state of nature? lies when he establishes sovereignty ?by means of reference to the state in which ?man is wolf to men?? (Agamben, 1998, p. 105). This state of nature ?is not a prejudicial condition that is indifferent to the law of the City? (p. 106). The Hobbesian state of nature is the exception, ?a principle internal to the City? (p. 105), ?continually operative in the civil state in the form of sovereign decision?, which appears at the moment the city is considered ?as if it were dissolved? (p. 105). In this state of exception and of nature, ?it is not so much a war of all against all as?a condition in which everyone is bare life (and therefore may be killed with impunity) and a homo sacer for everyone else? (p. 106). Thus all are both sovereign, with the absolute right to decide life and death, as well as bare life, subject to this absolute prerogative over life and death in ?a zone of indistinction and continuous transition between man and beast, nature and culture? (p. 109). A number of South African analysts noted this condition of exception in which bare life and sovereign power enter into a zone of indistinction; however, they could only conceive of this indistinction in terms of the friend?enemy antithesis. Du Toit therefore noted of the period of the violence of the 1990s, ?the distinction between combatant and non-combatant was more ambiguous and that between enemy and ally more uncertain? (P. du Toit, 2001, p. 78). Simpson, Mokwena and Segal also noted the ambiguity of the ?enemy? during this period; 97 ?the legalisation of the previously banned political movements has replaced the externalised enemy with an enemy within? (1991, p. 2). Johnstone wrote of the absence of clearly demarcated roles which often give shape to violent political conflicts. Who are the ?authorities? and who are the ?insurgents?. Which conception of an ?order? (in the sense of a set of values, institutions and material conditions to be defended or challenged) represents the fixed point from which we may plot the orientation of participants? (Johnston, 1997, p. 85) However, what the ?enemy within?, the ?confusion? between protagonist and antagonist denoted, was not a simple relation of inclusion and exclusion but a more originary relation, in which ?the ?estrarity? of the person held in the sovereign ban is more intimate and primary than the extraneousness of the foreigner? (Agamben, 1998, p. 110). The ambiguity of homo sacer, or sacred man, as originally expressed in Roman law is the juxtaposition of two apparently contradictory traits within this one individual who has been judged ?on account of a crime? and stands outside ius divinum (religious law) and therefore may not be sacrificed according to the prescribed ritual practices, and yet on the other hand also stands outside human law (ius humanum) and can therefore be killed with impunity by anyone without this being considered homicide, implying thus a suspension of the application of the law in the unpunishability of this killing. Therefore homo sacer stands in a relation of double exception from both human and divine law, namely, a relation of the ban in which he who has been banned is not, in fact, simply set outside the law and made indifferent to it but rather abandoned by it, that is, exposed and threatened on the threshold in which life and law, outside and inside, become indistinguishable. It is literally not possible to say whether the one who has been banned is outside or inside the juridical order. (Agamben, 1998, p. 29) 98 The relation of the ban connotes the relation of exception, ?the originary structure in which law refers to life and includes it in itself by suspending it?, thus homo sacer stands in a relation of ban to the juridico-political order, a relation of profound ambiguity, in which he who has been banned is both excluded from the community and is subject to ?the command and insignia of the sovereign? (Agamben, 1998, p. 28). As a consequence of this ambiguous relation of the ban, what defines the status of homo sacer is?the violence to which he finds himself exposed. This violence ? the unsanctionable killing that, in his case, anyone may commit ? is classifiable neither as sacrifice nor as homicide, neither as the execution of a condemnation to death nor as sacrilege. Subtracting itself from the sanctioned forms of both human and divine law, this violence opens a sphere of human action that is neither the sphere of sacrum facere nor that of profane action. (Agamben, 1998, pp. 82-83) Thus this is a type of human action which ?takes the form of a zone of indistinction between sacrifice and homicide? (Agamben, 1998, p. 83). At the two extremes of the juridico-political order are two symmetrical figures, the sovereign, ?with respect to whom all men are potentially homines sacri? (sacred men) and homo sacer, ?with respect to whom all men act as sovereigns? (Agamben, 1998, p. 84). The two are joined by the originary juridico-political relation of the ban and an ambiguous human action, the unpunishable killing that excepts itself from both human and divine law. Therefore, the sovereign sphere is the sphere in which it is permitted to kill without committing homicide and without celebrating a sacrifice, and sacred life ? that is, life that may be killed but not sacrificed ? is the life that has been captured in this sphere. (Agamben, 1998, p. 83) Thus homo sacer ?presents the originary figure? of the political relation, ?the originary exclusion through which the political dimension was first constituted? (Agamben, 1998, p. 83) and the ?human victim who may be killed but not sacrificed? is the ?first content of sovereign power? (Agamben, 1998, p. 83). Paradoxically, it is this capturing by sovereign 99 power, in a relation of inclusive exclusion, in which the ?sacredness? of man lies; ?life is sacred only insofar as it is taken into the sovereign exception? (Agamben, 1998, p. 85), but this sacredness also denotes an unconditional capacity to be killed or, as Foucault articulated it, the power of the sovereign to decide life and death. Thus it is through this relation with sovereign power that ?human life is included in the political order in being exposed to an unconditional capacity to be killed? (Agamben, 1998, p. 85). Therefore, ?not simple natural life, but life exposed to death (bare life or sacred life) is the originary political element? (Agamben, 1998, p. 88, original emphasis), thus ?the first foundation of political life is a life that may be killed, which is politicized through its very capacity to be killed? (Agamben, 1998, p. 89) and ?human life is politicized only through an abandonment to an unconditional power of death? (Agamben, 1998, p. 90). This chapter has explored the paradoxes of political conception evoked by the violence of the 1990s, violence which took place in a realm of sovereign exception at the boundaries of the juridico-political order. However, analysts working within a juridical conception of power could not recognise the nature of the power which lay before them. They misrecognised the sovereign exception as a juridical phenomenon which they sought to understand in instrumental terms within the juridico-political order as a simple relation between cause and effect. The multiplicity of causal explanations which this approach generated exposed the aporia of this juridical line of investigation. This chapter instead posits a different conception of the ?political? as the site of sovereign exception at the boundary of the juridico-political order. This is an irrevocably ambiguous realm, which nevertheless creates the conditions of possibility for juridico-political order but which the analysts of the violence of the 1990s struggled to interpret in a juridical paradigm. What the violence of the 1990s exposed, as the form of the state dissolved, was this originary site of the ?political? and the relation of 100 inclusive exclusion through which it is constituted, in the inscription of bare life (homo sacer) within sovereign power. It is these processes of sovereign inscription to which we now turn in an investigation of the originary struggle for juridico-political order in Moleleki Extension 2, which would ultimately lead to the inscription of sovereign power in the bodies of the Moleleki dead. 101 CHAPTER THREE: A STRUGGLE OF JURIDICO-POLITICAL ORDERING Introduction Having re-read the violence of the 1990s in terms of the exception, not as was normatively assumed to be a consequence of a breakdown of juridico-political order and the political sovereignty of the state, but as an exposure of sovereign exception, at the boundaries of the juridico-political order, this chapter empirically investigates this ?exposure? of sovereignty during the negotiation process in terms of the originary struggles of sovereign exception in the small township of Moleleki Extension 2 on the eve of the inauguration of South Africa?s first popularly elected government. This was a struggle for sovereign exception, in a national context of exception, which would create the very conditions of possibility for the juridico- political order of the post-apartheid state. In these struggles of decision on the exception, are revealed, as Schmitt argues (cited in Agamben, 1998, p. 19), the essence of State authority, the principle of sovereignty. Moleleki Extension 2, a Space In 1993, as the old South African state waned and the new state was yet to be inaugurated by the country?s first popular elections in 1994, the informal settlement of Moleleki Extension 2 was being established, on the south east edge of Katlehong township, its western boundary traced by the Natalspruit river, in which the ANCYL leader ?Wips? Motloung, would later be found, drowned, his face partly consumed by the fish that swam in this river. The urgency of the struggle for sovereign exception in Moleleki, on the eve of the country?s first national democratic elections, was the critical effect of an originary struggle for juridical order (Ordnung). This struggle took place as the physical territory of Moleleki Extension 2, 201 hectares of land on portion 141 of the farm Rietfontein was registered as a township in 102 November 1992. This land was therefore included, thirteen months before the Moleleki execution, in the administrative and geographical framework of South Africa?s apartheid townships in terms of the ?Black Communities Development Act? (Schyff, Baylis, Gericke & Druce, 1991), under a nomos of racial exclusion, still in force but in suspension on the eve of the country?s first democratic elections. The pressure for land to house South Africa?s growing urban population since the relaxation of influx controls by the apartheid state in the mid-1980s had been intense. The East Rand and Katlehong in particular, of which Moleleki Extension 2 became part through administrative fiat in November 1992, had been subject to significant urbanisation pressures. By the 1990s, the average stand in Katlehong housed approximately 22-30 people (Mashabela, 1988). Moleleki Extension 2 thus formed part of a local municipal ?Katlehong Outline Plan?, which attempted to address an estimated potential shortfall of 40 000 houses in Katlehong by the year 2000, through a variety of forms of housing and land development (Schyff et al, 1991, p. 2). This small extension to the township of Katlehong therefore sought to attempt to address some of these problems of land and housing through the establishment of Moleleki Extension 2 as what was known as a ?site and service? area, a bare carving out of 200m square pockets of land, without electricity or housing, basic water reticulation providing merely one water point for every 20 erven26 (Schyff et al, 1991, p. 3) . 26 Erf, n pl erven (Engineering / Civil Engineering) South African a plot of land, usually urban, marked off for building purposes [Afrikaans] (Retrieved from The Free Dictionary: http://www.thefreedictionary.com/erven) 103 Thus the administrative registration of Moleleki Extension 2 as a township in November 1992 constituted the township on paper, in administrative edict. However it was the struggle of juridical and territorial ordering between the ANCYL and the SDU in the township which was to substantively constitute this geographical space. This was a space which, like many apartheid townships, was denoted as such, as space, a mere circumference of physical land, an extension, its content, without name, simply areas, blocks, zones. It was in such a conventional South African zone of territorial, juridical and ontological anomie, namely, Blocks A-F of Moleleki Extension 2 that the struggle for sovereign exception took place and culminated in the execution, which attempted to localise the exception in the bodies of ANCYL members. Critically, these struggles of juridico-political order in Moleleki took place within a national context of juridico-political anomie, a context of suspension of juridico-political order. It was in this environment that the township of Moleleki had been administratively constituted in terms of the law of apartheid under the auspices of the then Katlehong City Council, its plans approved by the Transvaal Provincial Administration as consistent with standards entitled ?Ruimtelike standaarde vir Uitlegplanne in Swart dorpe, 1985? (Standards for Spatial Layout in Black townships) and approved as a ?development area? in terms of section 33 of the Black Communities Development Act 4 of 198427 (Schyff et al, 1988, p. 2). 27 In terms of apartheid legislation, which saw black people as permanently excluded from the ?white? national state, confined instead to ethnically designated Bantustan states, this Act provided for, ?the purposeful development of Black communities outside the national states; to amend and consolidate certain laws which apply with reference to such communities; and to provide for matters connected therewith? (Black Communities Development Act 4 of 1984). (Legislation. Retrieved from City of Cape Town: http://www.capetown.gov.za/EN/PLANNINGANDBUILDING/PUBLICATIONS/LANDUSEMANAGEMEN T/Pages/LegislationPart1.aspx) 104 However, during the four years between the initial application for the establishment of the township of Moleleki Extension 2 under the Black Communities Development Act in 1988 and the eventual establishment of the township in 1992, the administrative infrastructure of ?black? local government under which the township had initially been conceived and planned had substantively collapsed in the face of sustained national resistance to apartheid municipal governance from the mid 1980s onwards. A black local authority had been established in Katlehong in 1984 in terms of the Black Local Authorities Act of 1982 (Mashabela, 1988), which attempted to transfer responsibility for local apartheid governance to black municipal authorities and generated widespread national resistance as municipal authorities without access to any income from dormitory townships, increased rents in order to generate revenue for municipal administration. A national uprising against local councils, primarily conducted under the auspices of the United Democratic Front, which had been established in 1983, led to the collapse of many of these local councils in townships around the country after violent protests against the Lekoa Black Local Authority in the Vaal region of the PWV in September 1984 launched what was then known as ?the Vaal uprising? (Chaskalson & Seekings, 1988). From the beginning of the uprising on the Reef in September 1984, Katlehong experienced school and consumer boycotts leading the council to resign in 1985 as protests intensified. A rent and service charge increase introduced at the beginning of January 1986 was suspended indefinitely (Mashabela, 1988). The intensity of this struggle to tie juridico-political order to territory was not only the consequence of an originary processes of geographic establishment as the physical territory of Moleleki was marked out in the bushveld adjoining the Natalspruit river, but was also the consequence of a context of national political extremis inaugurated by the start of the 105 negotiation process, which substantively suspended the old order for the duration of a four year period of negotiations between former political opponents. This period of interregnum, of juridico-political suspension, unleashed what Agamben has called a ?force of law? without law, an anomic violence which was to claim at least 16 000 lives nationally between 1990 to 1994 (Coleman, 1998, p. 243) and which left approximately three thousand people dead on the East Rand (Shaw, 1997, p. 30), the Moleleki executed among them. In the year between the administrative registration of Moleleki Extension 2 as a township in November 1992 and the atrocity of the Moleleki execution on 7 December 1993, as the blocks of the new settlement were marked out, a process of juridico-political ordering took place that sought to tie this new land unequivocally to a hierarchy of authority in the township and in this establishment of juridical and territorial ordering, to constitute an irrevocable boundary against anomie, against the violence sweeping the East Rand in 1993, as the township was being established. Shaw writes: violence on the East Rand was the determinant of conflict on the Reef as a whole for 1993; of the 2 000 deaths in violence in the PWV in 1993 over 1 600 or 80% took place in the industrial towns and townships east of Johannesburg...the East Rand showed a 153% increase in numbers killed (Shaw, 1997, p. 32). This horror of sovereign power, unleashed from law, ?where the community took the law in their own hands? (TRC, 1998a, p. 132), was inscribed in the wounds of more than 544 bodies reported abandoned on the public streets of the Kathorus townships of Katlehong, Thokoza and Vosloorus between July and September 1993. This was a context which was thus ?infested with violence? (TRC, 1998a, p. 11) ?surrounded by violence? (TRC, 1998a, p. 283), ?engulfed in violence? (TRC, 1998a, p. 179), a situation that therefore ?demanded violence? (O. Motlokwa, personal interview, 2001), a state of 106 necessity where ?there was no choice, everybody was new in Moleleki, therefore we had to defend, it was a crisis situation? (M. Sonti, personal interview, 2001). It was against this anomie, this ?engulfing? by violence, that the struggle to constitute a boundary against this violence, in a localization of territory and juridico-political ordering, took place. However, the boundary which could exclude anomie proved elusive, permeable, unlocalisable. It was around this space of anomie, this space of exception, which had not yet been annexed to the juridical political order of the township, that the contest for sovereign authority in Moleleki revolved. The struggle to annex this space to the juridico-political order of the township of Moleleki was to constitute an ongoing anomie in the nomos of the township. It was in such a space of physical and juridical anomie, the geographic ?no-man?s land?, literally the land that belonged to no-one, an area of as yet uninhabited grassland on the edge of the township of Moleleki, that the execution took place. Jabulani Amon Mtwalo, a survivor of the execution, explains, ?We were taken to the last shack in Block F right towards the veld, the last line in Block F?They took us to that shack, it was still under construction it was not complete? (TRC, 1998a, p. 685). Tying Land to Juridical Order on the East Rand Therefore the struggle for sovereign exception in Moleleki has to be located within a wider context of conflict on the East Rand, which by the time Moleleki township was being 107 established at the end of the 1992, had significantly carved up the space of the East Rand into politico-ethnic territorial ?strongholds? under the sovereign authority of either the IFP or the ANC. This violent tying of Ordnung (juridical order) and Ortung (territory) had sought, in a radical localisation of the relation between territory and juridico-political order, to expiate the irreducible ambiguity of the sovereign exception in the townships of the East Rand, in violent inscriptions of this exception in the bodies of the citizens of these townships. Thus during the initial stages of the conflict on the East Rand, hostels and townships were progressively ?cleansed? of any of the social heterogeneity which had previously characterised them. Hostels became bastions of exclusive identity, in particular Zulu ethnic identity, as reciprocal movements of people from hostel and township, frequently under physical coercion, took place. ?Non-Zulus? were driven from hostels at the same time as people identified as ?Zulu? were driven from townships into hostels. These zones of pure identity were physically separated by zones of anti-identity, the ?no-go zone?, across which the battle for hegemony was waged and which marked the boundaries against which solidarity was formulated. A joint report by the Independent Board of Inquiry and Peace Action describes the no-man?s land which formed a buffer zone between the hostel and township on the East Rand: This area is completely deserted. Burnt out houses and empty streets bear testimony to the devastation which violence has wrought on this community. The danger of sniper fire from gunmen hiding in the ruins means that only security forces driving in armoured vehicles are willing to cross this area. (Independent Board of Inquiry & Peace Action, 1993, p. 2) The ?no-man?s land? was the antithesis of space. It was space that could not be used, which excluded uniformly. In it was distilled all the identities and antagonisms of the conflict as 108 well as the absence of these within the space of the no-man?s land. This exclusion represented the antagonists? differences to each other, ?automatically (codifying) the other side of the barricade as an immanent source of transgression? (Lacan cited in Feldman, 1991, p. 35). These processes of territorial reordering precipitated massive geographical movements of people across the spaces of the East Rand during the three years preceding the execution as residents fled persecution and violence, particularly on the boundaries of what were to become the ?no-go? zones, what Feldman refers to as the ?interface?, ?the topographical- ideological boundary sector that physically and symbolically demarcates ethnic communities?from each other? (Feldman, 1991, p. 28). This process of geographical relocation and social reordering constituted the critical spaces of the East Rand conflict, the hostel and township as two axial poles, metonyms for the relations of antagonism between ?politico-ethnic? geographic blocs constituted by and through violence. In the year of the Moleleki execution, these processes of topographical and political reordering of space were still ongoing on the East Rand. In January 1993, a systematic programme of political coercion was undertaken against residents of Phenduka section in Thokoza who failed to actively and visibly assert their association with the IFP, while in Katlehong between May and July later the same year, a violent process of ?ethnic cleansing? was undertaken against Zulu speakers, or a more malleable ?other? which could include residents perceived to have any type of association with the IFP or hostel residents (Independent Board of Inquiry & Peace Action, 1993, p. 23). The majority of these ?others? expelled from Katlehong township were migrants from KwaZulu-Natal living in shacks in Katlehong who were forced to flee to the hostels of Kwesine and Buyafuthi, by now armed and fortified ethnico- political enclaves, effecting, as in similar reciprocal campaigns of arson and intimidation taking place across the East Rand during this period, a radical ?denial of space? (Feldman, 1991) for 109 these violently constituted ?others?, anomalous now in a new spatial homogeneity in the territory of the township. Thus the spatialisation of power which took place in Moleleki Extension 2 during 1993 occurred at the margins of and was part of this wider process of Ordnung and Ortung on the East Rand. Tina Mootsi, whose son, Edward Mootsi, died in the execution, was subjected to these processes of violent territorialisation in the neighbouring informal settlement of Zonkezizwe, which constituted this area as an IFP ?stronghold?. This process helped constitute the boundary against which the homogeneity of Moleleki Extension 2 as an ANC area could be substantiated, a boundary of enmity which the SDU that would later execute Tina?s child would claim to substantiate its sovereignty in the township. Zonkezizwe, the informal settlement on the southern border of Moleleki?s outermost and most contested section, Block F, formed a continuous part of the administrative and geographical municipal plan for the development of the Greater Katlehong area (Schyff et al, 1991, p. 4). However in the contest for sovereign authority which unfolded in Moleleki in the year after its establishment, the geographical boundary with Zonkezizwe, a single line on the edge of the municipal map of Moleleki Extension 2, would constitute a critical boundary evoked by the SDU in Moleleki, against the other of the ?IFP? township, Zonkezizwe, in its struggle for sovereign exception. As an ANCYL member explained, ?They?d [the SDU] say Zonke is 110 attacking. But to tell you the truth there is no enemy that ever came to this section to attack. It was all lies? (ANCYL, focus group, 2001). While the contest between township and hostel stood at the emblematic centre of the bloody internal conflict of the 1990s, Moleleki Extension 2 did not experience direct conflict with the hostels surrounding the township Katlehong, of which it formed part. The geographical distance of Moleleki Extension 2 from the hostels that bordered Katlehong meant that the boundary of enmity in Moleleki was articulated in terms of the physical proximity of the informal settlement, Zonkezizwe, rather than in terms of the hostels which had been implicated in conflict with the older township. The efforts of the SDU in Moleleki Extension 2 to yoke territory to juridical order against the boundary of Zonkezizwe, and to invoke this as an unambiguous ?other?, had been made possible by earlier struggles of sovereign exception in Zonkezizwe, which cleansed its space of political and ethnic diversity through violent contests, which tied an irrevocable ethnic and political identity to this bare informal settlement and constituted it as an IFP ?stronghold?. The space in which the Moleleki execution would eventually take place was the geographic ?no-man?s land?, which formed the physical and metaphorical boundary between Moleleki Extension 2 and Zonkezizwe. Thus the spatialisation of power that took place in Moleleki Extension 2 occurred within a wider national nexus of Ordnung and Ortung, which constituted the township, at its original establishment as a homogeneity, against the boundary of an IFP ?other?. Chairperson Judge 111 Sandile Ngcobo, questioning SDU amnesty applicant Oscar Motlokwa, assumes this spatialisation of power, the de facto sovereignty of the IFP and ANC in hostel and township respectively: CHAIRPERSON: The position is that the townships were then occupied by ANC members in Katlehong, is that right? MR MOTLOKWA: Accept (sic) just one, Zongezizwe. Zongezizwe was inhabited by the IFP but the rest of them were ANC. (TRC, 1998a, p. 135) Critically, however, this spatialisation of power in the territory of Moleleki against the boundary of the IFP ?other? was not the result of a violent reordering within the township as had occurred on the rest of the East Rand, but was instead the result of these wider processes of territorialisation, the geographic inscription of an embodied subjectivity, within the territory of Moleleki, as residents of Katlehong and surrounding townships came to Moleleki Extension 2 as, Zola Sonti argues, ANC members, ?previously living in shacks? (M. Sonti, personal interview, 2001). As SDU member Oscar Motlokwa later affirmed, MR MADASA: Now, was Moleleki from A to F inhabited by the members of the ANC? MR MOTLOKWA: That is correct. (TRC, 1998a, p. 135) Homo sacer, defining the boundaries of the political community It was this putative party political homogeneity in Moleleki Extension 2 which later led the TRC to dispute the ?political motive? of the protagonists in the conflict that unfolded after the township was established. In the context of this political party uniformity SDU and ANCYL members articulated the boundaries of sovereign conflict in Moleleki in terms of the need to defend the ?community? against the ?criminal? other, accusing ?one another of failing to protect the community and of engaging in criminal activities instead? (Truth and Reconciliation Amnesty Hearing 112 Decision, 1998, p. 4). While these mutual assertions of criminality led the TRC to reject the amnesty application of the SDU members, what both SDU and ANCYL members were in fact articulating, in their representation of the boundaries of enmity in terms of the ?criminal? rather than the party political opponent, was the ?originary figure? of the political relation, homo sacer or bare life, whose exclusion defines the boundaries of the political domain. Crucially, homo sacer, one who has been judged ?on account of a crime?, stands in a relation of profound ambiguity to the political juridical order, namely the relation of ban, a relation of double exception from both human and divine law. Abandoned therefore by all law, the killing of homo sacer is considered neither homicide nor sacrilege. Therefore it is the sovereign decision on the exception which defines the relation of ban in which homo sacer finds himself and which traces a threshold between inside and outside the juridico-political order. Paradoxically, it is through this relation of sovereign ban that human life is first politicised through its very capacity to be killed. In the ?state of nature? (Agamben, 1998, p. 106), a principle that dwells continuously within the City but which is exposed when the City appears ?as if it were dissolved? (Agamben, 1998, p. 105), exists a condition in which all are both sovereign, with the absolute power to decide life and death, as well as bare life, subject to this absolute prerogative over life and death. It was in exactly such a context of exception during the period of negotiated transition, a zone of continuous transition between bare life and political life, that the struggles of sovereign exception in Moleleki took place. These ambiguities of the sovereign decision on the exception, articulated through the figure of the criminal, the friedlos (bandit), who exists in a relation of ban from the community, connotes a more originary and intimate relation than a simple relation of inclusion and 113 exclusion, as Oscar Motlokwa explained, ?We were looking for an enemy from outside whereas we had an enemy within, that is why there was this kind of a conflict?28 (TRC, 1998a, p. 167, own emphasis). It was thus this figure of the ?criminal?, the friedlos who usurped and represented the anomie within the township, the anomie of the exception, which even in a wider context of territorial and political rigidification on the East Rand, had not yet been decisively excluded in Moleleki township itself29. As Agamben argues, the ??estrarity? of the person held in the sovereign ban is more intimate and primary that the extraneousness of the foreigner?(Agamben, 1998, p. 110). All the protagonists involved in the struggle for sovereignty in Moleleki lived in close proximity to each other. Civic activist Bulelwa whose murder precipitated the chain of events leading to the Moleleki execution, caught the train to work daily with SDU member Zola Sonti. Elphina Mngadi whose son Isaac was killed in the execution was the neighbour of several SDU members, ?I knew them just as my neighbours but there were others I was close to; and they would come and visit here because I am speaking Xhosa and they were also speaking Xhosa, we were calling each other ?homey??(E. Mugadi, personal interview, 2001). Jabulani Mtwalo explains how he was released moments before the execution because his neighbour, SDU 29 A number of South African authors have noted similar sovereign struggles to define the boundaries of the political community, articulated in terms figure of the ?criminal? who exists not extraneously, but in a relation of inclusive exclusion, on the margins of the political community. Seekings notes for example that ?much collective violence on the part of the ?community? was not directed against targets which were unambiguously part of the system, but rather on the margins of the community?(Seekings cited in Beinert, 1992, p. 483). Marks in her documentation of the role of youth as ?moral defenders of the community? (Marks, 1993, p. 243), notes the struggles to define this boundary of ?community? in the context of the juridico-political suspension of the 1990s as the targets of political violence, ?became anyone or anything that prevented the Charterist social movement from achieving its hegemonic project of a non-racial, democratic South Africa?(Marks, 1993, p. 269). Xaba and Ball?s studies of the violent punishment of ?criminals? in South African townships exposes this punishment, not simply as a juridico-political punishment, but as sovereign punishment, a response to ?treason? against the community, ?Anyone considered guilty of these crimes was perceived to be an ?enemy? of the people?(Xaba, 1995, p. 63). 114 member Michael Armoed, recognised him as his son?s playmate, ?[he said] ?Comrades, this is my boy, he resides right next to my house in Block D??(TRC, 1998a, p. 695). Therefore critically what was at issue in these mutual ascriptions of criminality was a contest of sovereign decision on the exception that would trace the boundaries between political and ?bare life? and define who could be a political subject in the wake of the national deprohibition of the ANC, under whose sovereignty the township of Moleleki Extension 2 had been established. Thus while the contest for sovereign authority in Moleleki was articulated by its citizens as a conflict over territory and criminality, not a party political struggle, what was centrally at issue in these contests was in fact the political party of the ANC. Crucially, on the eve of the country?s first national democratic elections, the content of the democracy which this organisation intended, the extent of political subjecthood which it foresaw, and in the context of the conflicts of Moleleki, would women, ?children? and the ?ethnic? other be part of this political community or would they remain as ?bare life?, essential to but excluded from this community. ?Who Must Rule the Area?? A Struggle of Juridico-Political Ordering These struggles of sovereign decision on the exception, a ?power struggle [over] who must rule the area? (ANCYL, focus group, 2001, own emphasis), which would open the space in which the determination of a local juridical order within a particular territory could become possible, formed the substance of the conflict that took place between the ANCYL and SDU in Moleleki Extension 2 in the year after the establishment of the township. The struggles for sovereignty between the ANCYL and the SDU in Moleleki, the struggles to define the boundaries, the juridico-political substance of this small territory, to unequivocally yoke territory to juridical order and constitute the space in which the juridico-political would 115 function in the township, would finally lead to the violent localisation of the exception in the bodies of nine ANCYL members, in the atrocity of the execution. As the blocks of the new township were being physically laid out from November 1992 onwards, a network of local organisation began to be established, which in the absence of the institutional presence of the state, sought to tie territory to juridical order. At the pinnacle of this rubric of organisation was the local branch of the civic association, which formed part of the national civic organisation at the time, the South African National Civic Association of South Africa (SANCO). SANCO was ostensibly an ?independent? organisation (TRC, 1998a, p. 129), the ?umbrella body? in the township (ANCYL, focus group, 2001) that, ?is a community organisation, it doesn?t matter whether you belong to the IFP or National Party, it belongs to the community? (TRC, 1998a, p. 172, own emphasis). However, from 1991 SANCO was officially given the responsibility by the ANC for the establishment of SDUs, paramilitary defensive structures, which were already rapidly and often chaotically being set up across the country as violence escalated. Critically, it was the right of the civic in Moleleki to exercise its authority over the township as a whole and over the SDUs in particular that became one of the central points of tension in the unfolding conflict over authority in the township. Although SDU members in general agreed that the civic was ?the main body? (M. Sonti, personal interview, 2001), their ?mouth piece? (TRC, 1998a, p. 288) and cite a meeting called by the civic as the main basis of authority to form a SDU in the township, it was the later conflation of the authority of the civic with that of the ANCYL which was to lead to the rejection by the SDU of the authority of both the civic and the ANCYL and their counter- assertion of their right to the decision on the sovereign exception and the articulation of the boundaries of the political community in Moleleki. 116 The election of Lethusang Rikaba, already secretary-general of the local ANCYL, as secretary-general of the civic in approximately August or September of 1993 was a crucial turning point in the contest for sovereign authority in Moleleki. Rikaba argued: ?because the Civic was the umbrella, everybody in the township submitted to the Civic? (TRC, 1998a, p. 663, own emphasis). Crucially, however, it was this ?submission? to the civic closely allied with the ANCYL under the leadership of Lethusang that was at issue in the struggles for sovereign exception in the township. Lethusang explains, ?they [the SDU] said SANCO wants to rule them and yet, they were independent?they did not want to subscribe to us. They said we must subscribe under them? (TRC, 1998a, p. 568). SDU member Michael Nkomo explains the impact that the election of Lethusang to the leadership of the civic had: ?The first executive of the civic was very strong but when the new executive was appointed with Lethusang as the secretary everything changed. [He] was the source of the problem? (M. Nkomo, personal interview, 2001). Importantly, as Oscar Motlokwa argued of Lethusang and his fellow civic leader, Mokwena, ?They were supposed to give proper direction to the youth but they did the opposite? (O. Motlokwa, personal interview, 2001, own emphasis). The conflation of youth and civic authority, authority that was supposed to ?belong to the community?, marked for the SDU a fundamental disputation of the legitimate boundaries of sovereign authority in the township, a boundary which quite explicitly excluded youth from the domain of the political community. Therefore in the six months following the initial patrols of the township at the beginning of 1993, in which ANCYL members participated under the tutelage of the SDU, a profound disputation around the content of sovereign authority in the township began to emerge, which expressed itself in the conflict between the SDU and the ANCYL, increasingly closely allied with the civic association. In this period of juridico-political suspension what was revealed 117 was the tension between biopolitical authority, auctoritas, authority which is embodied and invested in the person, and juridico-political authority, authority exercised through representative office (Agamben, 2005, p. 77), which are normally conflated under the modern juridico-political conception of sovereignty, which assumes the supremacy of potestas (written law). While the ANCYL and the civic association claimed the discourse of rights of the ANC, and in this claiming assumed the primacy of potestas, what was revealed in the conflict that unfolded in Moleleki was the biopolitical nature of the ?Doctrine of the Rights of Man?, which inaugurates the citizen of the modern nation state. These ?inalienable? rights, articulated in written law (potestas) and ostensibly belonging to all human beings by virtue of birth were in fact revealed to be a biopolitical status. It is this biopolitical status, rather than the simple fact of birth in a particular nation, which in fact defines the right to citizenship and the correlate right to participate in the political community (Agamben, 1998). Thus both the ANCYL and the SDU rejected the right of the other to exercise authority in Moleleki, in terms of a conflation of ethnic, generational and gendered markers, a ?reading? of the body as an ideological text that, constructs a conjuncture of clothing, linguistic dialect, facial appearance, corporal comportment, political religious insignia, generalised special movements, and inferred residential linkages into a ?sign system? that coheres into an iconography of the ethnic30 other (Feldman, 1991, pp. 56-57). Because ethnicity and ideology are literally sunk into the material density of the body, in the struggle for the sovereign exception, the body becomes the site for the ?violent dematerialisation? (Feldman, 1991, p. 71) of ideological codes essentialised in the corpus and 30 Note here that the term ethnic is used in the sense of biopolitical marks of identification rather than the more common use of the term as an ascribed cultural identity. 118 the site for the ?elimination and purification of political substance? (p. 71) that took place in the final atrocity of the execution. SDU members bore these ?ethnic? insignia in the blankets that they wore as they patrolled the township31; in the ?Xhosa traditional dresses? (TRC, 1998a, p. 148) they wore when they went to rallies, which led ANCYL members to regard the SDU ?as a Xhosa dominated thing? (TRC, 1998a, p. 148); in the ?magic? they used ? as an ANCYL member explained, ?they are the ones who came with the strategy of ntelezi32 so that they could become the SDUs for this section as a whole? So that is why the SDU was dominated by older people? (ANCYL, focus group, 2001); in the songs they sang33; and in the residential history they shared, as Oscar Motlokwa explained: ?the majority of SDU people come from rural areas?Transkei, Lesotho, Pietersburg?then the youth come from the township? (O. Motlokwa, personal interview, 2001).34 This conflation of biological markers led the ANCYL to dispute the right of the SDU to claim sovereign authority in the township: ?they started saying they were not going to be ruled by the Xhosa speaking people and they [ANCYL] belonged to the township? (TRC, 1998a, p. 31 SDU member Themba Mtshali was told by civic leader Alfred Buthelezi at the site of Bulelwa?s body, ?we don?t want a person wearing a blanket here? (TRC, 1998a, p. 544). 32 Ntelezi is a a herbal preparation which ?protects against misfortune? and is ?sprinkled around home and body boundaries? (Emmet & Butchart, 2000, p. 169). In a study of crime and violence in South Africa, it was found that for a significant proportion of interviewees, ?traditional herbal and spiritual devices constituted an important source of protection and prevention? for residents to, ?keep themselves safe? in an environment of insecurity. One of the key ?herbal and spiritual? protection methods used was Ntelezi, (Emmet & Butchart, 2000, p. 169). 33 Martha Mthembu, whose son, Ditaba Joseph Mthembu died in the execution recounts that after the killing, ?while I was preparing to go to work [I] heard a group of men singing Xhosa songs?(Mthembu, 1998). 34 While most young ANCYL members had been born in urban townships, as most SDU members were older a number had grown up in apartheid South Africa?s rural ?homelands?, coming to urban South Africa to seek work as young adults. This history, despite the fact that most SDU members had spent the larger proportion of their lives living in townships, marked them, for the ANCYL as permanently ?other?. 119 149, own emphasis). While SDU members had in fact come from a range of former ?homelands?35 and some, such as Michael Armoed, were born in townships, according to Sapire (1992), the numbers and social impact of Xhosa-speakers did significantly increase in 1990 due to a variety of forms of migration of Xhosa-speakers, particularly to the informal settlement of Phola Park on the East Rand, (p. 695). At the same time as this migration of Xhosa speakers to Phola Park, there were other new arrivals, particularly former soldiers from war-torn Southern African countries such as Mozambique who were perceived to be playing an important role in both training36 and acquiring weapons for SDUs37. It was in this context that ANCYL members in Moleleki asserted: ?SDUs want to make it like Phola Park, most are Xhosas so they wanted Xhosas to rule Moleleki? (ANCYL, focus group, 2001). What the ANCYL invoked in opposition to the figure of this ethnic other, who had become ?alien and dangerous to those for whom the community had become a community of national democratic subjects? (Chipkin, 2007, p. 135), was the figure of modernity ? free from communal and ?tribal? identities, rural associations, traditional generational hierarchies ? unequivocally secular and urbane. This, critically, was the figure of the National Democratic Revolution (NDR), a politics of nationalism, which had come to dominate civic and youth 35 In terms of the 1970 Bantu Homelands Citizenship Act (National States Citizenship Act) No 26, all black persons were required to become citizens of a purportedly ?sovereign? self-governing territorial authority (homeland) that corresponded to each person?s ethnic designation under the Population Registration Act, thus denying black South Africans the right to claim national citizenship in the ?white? state (Bantu Homelands Citizenship Act (National States Citizenship Act) No 26 1970). (Legislation. Retrieved from South African History Online: http://www.sahistory.org.za/pages/governence-projects/liberation- struggle/legislation_1970s.htm) 36 Police press statements on the SDU in Phola Park alleged first that two Frelimo soldiers had been training SDU members in Phola Park (Potgieter, 1992) and later that, ?Renamo instructors are actively involved in the training and supplying of firearms to self-defence units in Phola Park? (Zwane, 1992). 37 Sally Sealey, researcher for an NGO which investigated and monitored violence on the East Rand, testified at TRC hearings on SDUs in Thokoza, that ?many of the applicants have mentioned that they bought firearms from the Polla [sic] Park informal settlement. In Polla Park there are a number of people that are?illegal immigrants [from] Mozambique and many of [sic] people were involved in selling AK47s and the like to people? (Truth and Reconciliation Amnesty Hearing, 1998, p. 12). 120 organisation on the East Rand as elsewhere during the rise of resistance to apartheid from the mid 1980s onwards. According to Chipkin, NDR invoked a set of biopolitical markers to define the ?authenticity? of the national subject in ways that had on the East Rand increasingly excluded the ?traditional?, Zulu-speaking migrant hostel worker from the orbit of democratic organisation in a similar manner that SDU members were rejected by the ANCYL in terms of a set of biopolitical markers that defined them as the ethnic other. Crucially, what had characterised the politics of NDR over the previous decade was a significant challenging of generational hierarchies, which had seen young people come to dominate and lead urban resistance to apartheid through what Sitas calls the ?politico-cultural web? of the ?comrade? movement, ?involving voluntary (and sometimes coerced) participation, cultural dynamics and a new volatile social identity shaped through mobilisation and conflict? (Sitas, 1992, p. 633). For comrades, struggle was visceral and experiential, ?it is not legitimated in the abstract by a broader movement that was everywhere and nowhere, but also practically? (Sitas, 1992: 640). As SDU member Zola Sonti argued of ANCYL members, they ?would want to act immediately, without discussion [you] must get go ahead from head office therefore there was conflict with the youth? (M. Sonti, personal interview, 2001). Like the signage that delineated SDU members, the boundaries between those who were and weren?t comrades were also delineated through a ?sign system? that cohered into an ?iconography? of the authentic ?comrade?; ?you belong because of the way you sing, the slogans you know, the lineages you have learnt, the way you speak to each other? (Sitas, 1992, p. 636). ANCYL members? rejection of the right of SDU members? to ?rule? in Moleleki, in terms of a set of biopolitical markers, revealed the confusions caused by the conflation of auctoritas 121 and potestas under the modern juridico-political conception of sovereignty, which conceals the ?hidden? nucleus of sovereign power, the articulation between juridico-institutional and biopolitical power. Thus implicit in the claims to sovereign authority of both the ANCYL and the SDU and their rejections of the sovereign authority of the other, was exactly this articulation between the juridico-political and the biopolitical. While ANCYL and civic members claimed their right to sovereign authority in terms of the potestas, representative authority, articulated through the ?Doctrine of the Rights of Man?, the essentially biopolitical nature of these ?rights? to membership of the political community was exposed in their rejection of the claims to sovereign authority of the SDU in terms of a series of biopolitical markers. However, implicit in the very rejection of the right of the SDU to juridico-political ?rule? in Moleleki, while expressed in biopolitical terms, was an assumption that SDUs, despite the fact that they were ostensibly merely a defensive paramilitary formation, did in fact have a wider claim to juridico-political authority in the township, a claim which could thus be contested in juridico-political terms. ANCYL members therefore argued of the SDU that they ?must do things right, if going to rule? (ANCYL, focus group, 2001) and that SDU members had not ?consulted? the ?community? before establishing the SDU, as would have been expected from an organisation established in terms of potestas as a representative authority. However, they expressed their rejection of the SDU?s claims to juridico-political authority in biopolitical terms, arguing that the SDU was established unilaterally by ?Xhosas? who ?selected themselves?; ?we did not know about it, we were never told anything? (ANCYL, focus group 2001). 122 While ANCYL members rejected the authority of the SDU in Moleleki in terms of a conflation of the biopolitical and the juridico-political, SDU members similarly claimed their authority in terms of both potestas and auctoritas. Although SDUs were ostensibly ?independent? paramilitary formations accountable to the ?community?, represented by the civic association, it was quite clear that in the context of the violent conflict which had already parcelled land up on the East Rand in ethnic and political enclaves, the boundaries of the ?community? who could participate in SDU patrols was quite explicitly defined in juridico-political party terms. As Oscar explained, ?people who were allowed to volunteer [in SDUs] were members of the PAC, ANC, members of the civics, not members of the IFP, they were not allowed? (TRC, 1998a, p. 141). In reality, most members of the SDU in Moleleki were supporters of or signed-up members of the ANC38. As SDU member Michael Nkomo explained, he joined the ANC in 1990 in order to ?fight for the rights of people? (M. Nkomo, personal interview, 2001, own emphasis). Moreover SDU members saw themselves as critical allies in the violent struggle for ANC hegemony on the East Rand. Several SDU members argued that SDUs had played a ?key political role? (O. Motlokwa, personal interview, 2001; M. Nkomo, personal interview, 2001) in this struggle for hegemony and understood that while the ANC may have decided that SDU structures should be accountable to civic structures in townships, SDUs were critically the initiative of the ANC and therefore had their tacit, if not explicit, political support. As Oscar Motlokwa argued, ?the mere fact that the ANC made an announcement that [they] should establish [SDUs] and that Kasrils39 accounted for SDUs in Kathorus at the 38 The TRC decision on the amnesty application of 13 SDU members from Moleleki in connection with the Moleleki execution stated, ?The applicants are all members of the ANC?(TRC, 1998c, p. 2). 39 Ronald Kasrils (commonly known as Ronnie Kasrils) (born November 15, 1938) is a lifelong member of the African National Congress (ANC) and one of the founding members of Umkhonto we Sizwe (MK), the ANC?s 123 TRC showed [the ANC?s] involvement? (O. Motlokwa, personal interview, 2001). Not only did SDU members understand their organisation to have the ANC?s political support but they believed that their defensive activities were critical to the ANC?s ability to win a representative election on the East Rand; ?without SDUs the ANC would never have got votes in Kathorus, the IFP would have been ruling? (M. Nkomo, personal interview, 2001). Thus it is clear that SDU members, like ANCYL members, conceptualised their claim to sovereign authority in the township in terms of the ?Rights of Man?, which articulated the ANC?s struggles for popular democracy in South Africa. However, this claim for juridico-political authority based on potestas or written law, which invoked a contractual relationship, was vitally co-ordinated with a claim to biopolitical authority, invoking a relationship of bond. Unlike contractual relationships, which regulate juridico-political exchanges between legal subjects, bonds relationships ?are embodied in inequality ? that of the kind necessarily found between parent and child, lord and serf, man and God? (Van Zyl, 1990, p. 7, own emphasis). While bonds ?constrain the rights of the younger or weaker party?, they offer ?privileges not expected in the contractual form? (Van Zyl, 1990, p. 7), in particular the entitlement to protection and nurture. In return, however, bonds require from the subject obedience and loyalty. Critically, also, ?what distinguishes bonds from contracts... is a right to violence in the interests of nurture? (Van Zyl, 1990, p. 9, own emphasis). armed wing. During the 1990s Kasrils was one of three senior members of MK who set up a unit to assist in the arming of Self Defence Units. Kasrils later applied for amnesty for these activities. (TRC, 2003, p. 272). After 1994 Kasrils served as Deputy Minister of Defence (1994 -1999). He was also appointed Minister of Water Affairs and Forestry (1999- 2004) and Minister for Intelligence Services (2004 ? 2008). (Ronald (Ronnie) Kasrils, Retrieved from South African History Online:http://www.sahistory.org.za/pages/people/bios/kasrils,r.htm) 124 Auctoritas and Potestas The biopolitical authority of SDU members in Moleleki during 1993, in a context of interregnum, of juridico-political suspension, exposed this critical co-ordination between auctoritas, authority founded on the bond relationship and potestas, authority founded on a contractual relationship where neither can fully exist without the other; ?auctoritas seems to act as a force that suspends potestas where it took place and reactivates it where it was no longer in force? (Agamben, 2005, p. 79, own emphasis). Crucially, in a situation analogous with that in the classical Roman state, in an extreme situation such as that of interregnum, where there remained ?no consul or other magistrate? in the city, it is auctoritas that ensures the continuity of power and the continued ?functioning of the Republic under exceptional circumstances? (Agamben, 2005, p. 79). It was in such a ?state of emergency? (ANCYL, focus group, 2001), an anomic space that resulted from the suspension of law, that SDU members claimed the authority to establish a defensive formation in Moleleki in terms of ensuring a continuity of power in this context of juridico-political suspension. As Oscar Motlokwa explained, ?Four people, Manyala, Majola, Mtunzi and Nkosi, started the SDU. [The] Community had nothing to do with how commanders [were] chosen, only these four chose Nchebe as commander, not the rest of the SDU? (O. Motlokwa, personal interview, 2001, own emphasis). Thus, while the SDU was established in Moleleki in response to a general call at a civic association meeting that residents in Moleleki could not expect assistance from the ?central? SDU in Katlehong if they did not have their own defensive structure (M. Nkomo, personal interview, 2001), SDUs were hierarchically established and were not a representatively elected formation. Oscar Motlokwa explains: ?The central civic formed the central structure of the SDU and then the central structure formed the sectional and the zonal structures? (TRC, 1998a, p. 138). 125 Members of SDUs comprised those who volunteered, rather than those who were elected or went through a selection procedure. ?You couldn?t recruit anyone because you had to be ready to die at any time, therefore [you] had to volunteer? (O. Motlokwa, personal interview, 2001, own emphasis). ?Volunteering? was necessitated, ?because of the situation? (M. Sonti, personal interview, 2001), in which you, ?had to choose sides, [you] couldn?t be neutral? (M. Armoed, personal interview, 2001), because ?the IFP attacked everyone, whether SDU or not? (M. Nkomo, personal interview, 2001). Therefore SDU members held their positions, not because of the office they represented but because of the qualities they embodied, they ?had to be ready to make these sacrifices? (O. Motlokwa, personal interview, 2001), to ?risk [their] life? (M. Nkomo, personal interview, 2001). What the context of interregnum reveals, which helps elucidate the nature of the power SDU members exercised, is the structural analogy between private and public law, which requires a ?duality of elements?, auctoritas and potestas (Agamben, 2005, p. 78). In private law, auctoritas is the property of the ?auctor?, the ?pater familias? (p. 76), whose authority (auctoritas) is necessary to ?confer legal validity on the act of a subject who cannot independently bring a legally valid act into being? (p. 76); for example, the ?auctoritas of the father ?authorizes? ? that is, makes valid ? the marriage of the son in potestate? (p. 76). Critically, in a context of exception, when the constitution has been suspended and public offices have been dissolved, public authority returns to the ?fathers?, the pater familias of private law ? in Rome, the consular family. The formula used was ?res public ad patres redit [the republic returns to the fathers]? (p. 79). This authority to ensure the functioning of the Republic under exceptional circumstances is not representative but ?rests immediately with the patres auctores? as ?private citizens? (p. 80). 126 SDU members explained their duty of protection in a context of exception in terms of this contiguity of auctoritas in the private and public domains. All SDU members articulated the basis of their authority as the need to ?defend the community? (M. Armoed, personal interview, 2001; O. Motlokwa, personal interview, 2001; M. Nkomo, personal interview, 2001; M. Sonti, personal interview, 2001) in a time of crisis. This was not a responsibility of participation in representative political office but a responsibility of auctoritas in a time of juridico-political suspension. As Oscar Motlokwa explained: ?in the SDU there were a lot of people who didn?t know anything about [juridical] politics, [they] just had to respond to the situation and defend the community? (O. Motlokwa, personal interview, 2001). This obligation ?as a member of the community? (M. Nkomo, personal interview, 2001) to protect the ?community?, was both a personal and public duty. As Oscar Motlokwa explains, the community used to ?love [him] very much? because [he] didn?t want ?wrong things to be done to the community? (O. Motlokwa, personal interview, 2001, own emphasis). Therefore, ?the community [including his family] always came first? (O. Motlokwa, personal interview, 2001); it was necessary to make sure the community ?were safe? (O. Motlokwa, personal interview, 2001), that they ?were protected? (M. Nkomo, personal interview, 2001), a duty from which SDU members did not ?gain? or ?benefit? personally (M. Armoed, personal interview, 2001; O. Motlokwa, personal interview, 2001; M. Nkomo, personal interview, 2001; M. Sonti, personal interview 2001; TRC, 1998a, p.546), as might be expected from formal representative public office. Crucially, this duty of protection as pater familias in both private and public domains invoked a contest of sovereign authority with the ANCYL in terms of the bond relationship. Thus, it was the betrayal of this bond relationship, and in particular the obligation of 127 protection of the pater familias, in addition to a failure of representivity, which invoked the rejection of the authority of the SDU by the ANCYL. An ANCYL member explained: ?As we were still young we expected the SDU as elder people to protect us when we go to sleep? (ANCYL, focus group, 2001). Elphina Mugadi, whose son Isaac was killed during the execution, articulated the conflict in similar terms: ?We thought that the children were safe under the protection of the SDUs, whereas they were perpetrators, they were the ones who were killing our children? (Mugadi, 1998, own emphasis). Therefore as ANCYL members emphasised, referring to money donated by community members to buy the firearms that would eventually be used in the execution, ?parents collected money for firearms which ended up killing [us]? (ANCYL, focus group, 2001). Agamben has pointed out the essential contiguities between the ?unconditional authority [potestas] of the pater over his sons? (Agamben, 1998, p.87) with sovereign authority, whose first expression in law, as a power to decide life and death, was formulated in terms of this unconditional authority of the father over his son. This power is instantiated in the relation between father and son, ?in the instant in which the father recognizes the son?he acquires the power of life and death over him? (Agamben, 1998, p. 88). This tie between father and son ?is more originary than the tie of positive rule or the tie of the social pact? (Agamben, 1998, p. 90) and indicates in fact an untying in the abandonment of the son to an unconditional power over life and death. Therefore this is not a power which is confined to the private domain but which ?attaches itself to every free male citizen from birth? and hence defines ?the very model of political power in general? in which ?life exposed to death (bare life or sacred life) is the original political element? (Agamben, 1998, p. 88 original emphasis). Thus in Rome, the power of the 128 father, ?patria potestas was felt to a kind of public duty? a ?residual and irreducible sovereignty?? (Agamben, 1998, p. 88), which would allow the father Cassius to drag his son from the rostra, when he was trying to supersede the authority of the senate, the highest political authority in Rome (Agamben, 2005, p. 88). It is this contiguity between the power of the father and the power of the sovereign that Agamben illustrates when he recounts the story of Brutus, who having put his sons to death, ?adopted the Roman people in their place?, transferring the power of death over his sons to the citizenry as a whole, invoking the sinister epithet, ?father of the people?, ?which is reserved in every age to the leaders invested with sovereign authority? (Agamben, 2005, p. 88). In the conflict which unfolded between the ANCYL and the SDU in the township of Moleleki, what were evoked were the ambiguities of the exception which articulates biopolitical and juridico-political authority. In claiming the discourse of rights of the ANC, ANCYL members refuted the authority of the SDU in the township in juridico-political terms, arguing that the SDU had been established unilaterally, without the democratic ?consultation? with the ?community? that would be required of an organisation established in terms of potestas. But this was critically coordinated with a biopolitical claim, namely that this lack of representative consultation was the result of an embodied ?ethnic? identity, which led ?Xhosa? men to select each other for the formation of the SDU. These claims were further articulated with another biopolitical claim, which evoked the relation of the bond and the failure of the duty of the pater to offer protection and guidance to his subjects. As Lethusang Rikaba explained of the ANCYL member?s initial involvement in the patrols of the SDU, ?we wanted to join them because they were elderly people, we wanted them to show us the way? maybe as the Youth we were not walking the right way? (TRC, 1998a, p. 575). 129 Therefore ANCYL members initially participated in patrols of the township under the auspices of the newly formed SDU, which, showing the essential contiguity between the auctoritas of the pater and the sovereign, claimed the authority to ?fetch? young boys from their familial homes for public patrols, evoking the anomic space of exception that existed in the township, where public and private domains enter into a zone of indistinction. Although, as Elphina Mugadi explained, ?they were asking for them politely in the beginning?, over time ?they were not asking for the parents? permission? (E. Mugadi, personal interview, 2001), taking Elphina?s son against his will to participate in patrols. ?I think it was the treason because he said he no longer wanted to patrol? (E. Mugadi, personal interview, 2001, own emphasis). The reason that Isaac, Elphina?s son, no longer wanted to participate in patrols was the violence he allegedly saw being perpetrated by SDU members: my child was always crying. He said they caught one Zulu man who was an IFP member and they stabbed him on the stomach and they drove off with his car. So the children would go but they were scared as they tell us what happened. (E. Mugadi, personal interview, 2001) Critically it was this involvement in violent ?crime? which would lead ANCYL members to contest the sovereign authority of the SDU in Moleleki as a violation of the duty of protection shared and embodied by all ?men? in the township. As one ANCYL member expressed it, ?in the beginning every male was supposed to go out, there was no SDU, so therefore when the whistle blows every male would go out?A father, a son would? all go out? (ANCYL, 2001a, p. 28). In the context of the challenges to generational hierarchies that had taken place under the auspices of the ?comrade? movement, which appealed to a set of ?rights? that ostensibly refuted the biopolitical hierarchies of gender and generation, ANCYL members assumed that they were in fact such ?men?. Burman and Sch?rf write of the anti-apartheid resistance of the 130 1980s and 1990s, ?the battling youth acquired a great deal of prominence and prestige as the ?young lions?, with consequent tacit transfer of authority to them to represent adult interests in one arena of struggle ? the streets? (Burman & Sch?rf, 1990, p. 730). However, in opposition to this assertion of potestas, SDU members evoked the relation of the bond and rejected the right of ?young boys? in the ANCYL to claim the authority to protect the township as ?men?. SDU member Michael Nkomo emphasised, with regard to the establishment of the SDU in Moleleki, that ?it was announced?that men should report to the [SDU] commander, Manyala. Not boys, men. Men with responsibilities, heading households? No women could be part of the SDU? (M. Nkomo, personal interview, 2001). As Carton (2000) points out, ?the mutually reinforcing relationships of generation and gender effectively maintain?patriarchal authority? (p. 6). Therefore just as Bulelwa was told on the eve of her death that as a woman she should remain silent, young ANCYL members who attempted to challenge the authority of the SDU were similarly told that they were ?too young to ask questions? (ANCYL, focus group, 2001) and that ?they must better keep quiet? (TRC, 1998a, p. 570). ANCYL members explained: ?They [SDU members] were untouchables; they did not want us to come and sit down with them to discuss?whatever they are saying is final? (ANCYL, focus group, 2001). However, drawing on the juridical rights articulated by the comrade movement, which urged ?the ideals of democratic decision making, in which they [youth] expected to participate, and to seek a degree of authority in the formulation of ideas and policy? (Burman & Sch?rf, 1990, p. 730), ANCYL members refused the obedience and loyalty of the bond relationship. Elphina concludes: ?The youth were not prepared to do what they [SDU members] told them to do, so there was a fallout? (E. Mugadi, personal interview, 2001). 131 It was in this context that challenges to generational hierarchies provoked challenges to gendered hierarchies, which created another basis for the ANCYL to contest the sovereign authority of the SDU and for the SDU to attempt to re-assert the relationship of the bond. ANCYL and civic leader Lethusang Rikaba argued that SDU members were ?not politically literate, they saw politics as war, women stay in the kitchen, men fight? (ANCYL, focus group, 2001). When, according to SDU member Zola Sonti, girls from the ANCYL joined SDU patrols, the SDU sought to ?consult the community? about this; however, ANCYL members, in a complete overturning of the relation of the bond and the counter-assertion of potestas through the discourse of rights articulated by the ANC, ?went to parents to tell them to release their girls [for patrols] because Peter Mokaba said the ANC would train them? (M. Sonti, personal interview, 2001, own emphasis). Paradoxically, in invoking the ?Doctrine of the Rights of Man?, ANCYL members, although influenced by the ?comrade movement?, which drew in youth of both sexes, in fact invoked a set of biopolitical rights which at their inception excluded women from the political community and which, as Sitas points out in relation to the ?comrade? movement of the 1980s, had in fact included women differentially, away from the frontlines of battle and the male duty of protection. While comrades were drawn from both sexes there was simultaneously ?a military division of labour: young men would be the warriors, and young women their assistants, their support and caring networks, and also the messengers and organisers of the supply lines for battle? (Sitas, 1992, p. 633). Thus, in the months after the SDU began patrols of the township, increasing contention emerged because, as SDU member Michael Nkomo explains, ANCYL members ?didn?t want to be instructed, corrected by adults? (M. Nkomo, personal interview, 2001). This failure to submit to the authority and guidance of the pater familias, purportedly led ANCYL members to engage in criminal activities: 132 during the day [when adult SDU members were at work]?people would be looting Spaza shops? taxis would be shot at, if you were a Zulu speaking person you would be assaulted and they would be saying that you are a member of the IFP. You had this Spaza shop and you would be expected to provide food. Now the commander was trying to bring order to this situation and that is where the division started40(TRC, 1998a, p. 143). Thus, as Oscar Motlokwa explains, ?In the beginning SDUs were joined by adults and young people but over time it was clear that the young were unruly and undisciplined? (personal interview, 2001). In refusing to offer the obedience and loyalty required by the bond relationship with the pater familias, ANCYL members ultimately forfeited sovereign protection, instead provoking a sovereign punishment that in its violent excess would ?eclipse? the crime against the sovereign41: So they told the other ones [ANCYL members who were incarcerated before the execution] that you think you are clever than us your seniors ? We?ve been living for a long time. So I can say that they were beating our guys?because they were saying to them, you think you are wise, you want to [hit] us your seniors. (ANCYL, focus group, 2001) While SDU members accused the ANCYL of being involved in a variety of criminal activities during the day, the SDU in claiming the sovereign decision on the exception, which traces a threshold between inside and outside the juridico-political order, had imposed a township-wide curfew that would attempt, through the violent tying of juridical order to territory, to physically demarcate the boundaries of inclusion and exclusion from the ?community? of Moleleki. During the night within the borders of the township, sovereign 40 ANCYL members argued that SDU members had themselves carried out these acts of criminality (TRC, 1998a, p. 569). 41 Foucault argues of the spectacle of the public execution, ?It is a ceremonial by which a momentarily injured sovereignty is reconstituted. It restores that sovereignty by manifesting it at its most spectacular?(Foucault, 1977, p. 49). Sovereign punishment as an effect of sovereign power thus brought, ?into play, at its extreme point, the dissymmetry between the subject who has dared to violate the law and the all-powerful sovereign?(Foucault, 1977, p. 49) and, ?whose function was to go further than this atrocity [against the sovereign], to master it, to overcome it by an excess that annulled it? (Foucault, 1977: 56). 133 punishment, which evoked the contiguities between the unconditional authority of the pater over his sons and sovereign authority over the ?people?, would be meted out to those who ?disobeyed? instructions (T. Mootsi, personal interview, 2001); ?after a certain time [people] should be in bed?, and there would not be ?shebeens or stokvels, and no lights in the street?42 (O. Motlokwa, personal interview, 2001). An ANCYL member explains: When we call a meeting people would take all their clothes off during the meeting and say look what happened ? they got stripes from sjambokking during the night. And what was said about that is that these people [SDU members] are coming?into the house when you are still asleep ? they find the lights on and they say hey why are you still having your lights on we told you it?s time to go to sleep. So they sjambok them. (ANCYL, focus group, 2001) Critically, it was the ?mysterious deaths at night? (T. Mootsi, personal interview, 2001), ?the bodies that were found in the morning? (ANCYL, focus group, 2001) and the ?inexplicable? (TRC, 1998a, p. 393) violence which occurred under the night-time custodianship of Moleleki by SDU members, which would irrevocably break the bonds of loyalty for young ANCYL members. SDU members had claimed sole domain of the township at night after ?instructing? ANCYL members to ?patrol during the day and they will patrol at night? (ANCYL, focus group, 2001). Despite the night-time patrols of the SDU, people continued to die during the night. As Lethusang Rikaba explains, ?the community was saying we have an SDU, it is patrolling, but people keep on dying? (TRC, 1998a, p. 567). However, the irreducible ambiguities of defining these boundaries and the impossibility of localising the exception in order to irreversibly fix an inside and outside of the ?community?, led irrevocably to sovereign atrocity. Oscar Motlokwa explains, ?we had a sign which we 42 Foucault has noted the conflation of the ?public? and ?private? in the punishment of disobedience to sovereign authority, in which, ?rules and obligations are presented as personal bonds, a breach of which constituted an offence and called for vengeance? but whose breach was simultaneously an act of treason, ?the first sign of rebellion?not in principle different from civil war? (Foucault, 1977, p. 57). 134 used at night?any car which was entering Katlehong, would use a sign of flicking the lights then we?d know that those are members of the community? (TRC, 1998a, p. 131). On the other hand, ?if people don?t use signs they would be killed even if they were members of the community? (O. Motlokwa, personal interview, 2001). Thus, in the attempt by the SDU to localise the exception and define the boundaries of the political community, ANCYL members were constituted as homo sacer, bare life abandoned on the threshold of life and law who ?became enemies in the section? (ANCYL, 2001a, p. 21). The Moleleki Execution, a Localisation of the Exception in the Body of Homo Sacer It is the power of sovereignty, at the boundary of the juridico-political order, to mediate this boundary of life and death, to produce the condition of death as an artifact of power, that is critical to understanding the mute corpses of the Moleleki execution. As Mbembe argues, the ultimate expression of sovereignty resides, to a large degree in the power and the capacity to dictate who may live and who must die. Hence, to kill or to allow to live constitute the limits of sovereignty, its fundamental attributes. To exercise sovereignty is to exercise control over mortality and to define life as the deployment and manifestation of power. (Mbembe, 2003, p. 11) What is produced thus at the limits of sovereign power, is the corporal body, the manifestation of the sovereign?s right to decide who may live and who must die. The production of this biopolitical body, the inclusion of bare life in the political realm, is the starting point for the analysis of the horrifying corpses produced by the Moleleki execution. The Body of Homo Sacer What was radically exposed by the corpses of the Moleleki execution, rotting unattended in the bushveld, consumed as mere flesh by roaming dogs, was the implication of bare life (zo?), ?the simple fact of living common to all living beings (animals, men, or gods)? (Agamben, 1998, p. 1) in political life (bios). This ?bestialization? of man, which Foucault (cited in 135 Agamben, 1998, p. 3) first noted as an effect of modernity, is in fact, ?the secret tie uniting power and bare life? (Agamben, 1998, p. 6). In the production of the biopolitical corpse, therefore, the nature of the ?political? was exposed not as an opposition between bare life and political life, where the ?political? is merely an additional instrumental capacity, which leaves ?life? itself essentially unchanged. Instead the ?political? was revealed as a transformation of life itself, which produces a different genus, a different form of life. The ?political? is the locus of this transformation of life itself, through a relation of exception within life itself, a relation of exception with his own bare life that ?man? achieves critically through language, logos43, where politics ?occupies the threshold on which the relation between the living being and the logos is realized? (Agamben, 1998, p. 8). What was evoked by the Moleleki execution therefore was the extreme locus of ?the political? in the sovereign decision on the exception at the threshold of the juridico-political order. Exposed on this boundary of life and law stands the ambiguous figure of homo sacer, who exists in a relation of ban on a verge of indistinction between ?animal and man, physis and nomos, exclusion and inclusion? (Agamben, 1998, p. 105). In this state of extreme ambiguity and exception, it is the denial of logos by sovereign violence that mediates this boundary of physis and nomos. Without language, the exceptio of bare life within life, which makes possible the good life, is revoked, and what is revealed is merely ?bare life?, sheer corporality. The localization of the exception in the atrocity of the execution irrevocably extinguished logos and sought to permanently fix the transformation of the Moleleki victims 43 The term logos derives from the verb ???? leg? ?to count, tell, say, speak?. The primary meaning of logos is: something said; by implication a subject, topic of discourse, or reasoning. Secondary meanings such as logic, reasoning, etc. derive from the fact that if one is capable of ?????? (infinitive) i.e. speech, then intelligence and reason are assumed (Logos. Retrieved from Wikipedia: http://en.wikipedia.org/wiki/Logos#cite_note-3). 136 from ?man? to ?animal? in order to define the boundary of the ?political? at the boundary of bare life. The Stray Dog In Moleleki, prior to the execution, this mediation of the boundary of life and death took place as both linguistic and material practice (Feldman, 1991, p. 69), in which certain ?indexical terms? refer to ?states of destroyed or altered embodiment? that correlate to ?categories of action and consequent transformations of the body? (Feldman, 1991, p. 69). Prior to the production of the actual corpse as material practice in Moleleki, this linguistic transformation was effected through the metaphor of the dog and in particular the stray dog, a dog who in a condition of displacement and ban, like homo sacer, is exposed and threatened on a threshold of life and law, a threshold of extreme violence. SDU member Michael Sonti articulates the threat of this linguistic and material threshold in the metaphor of the dog, ?There was this word that was saying: ?Here are these dogs, attack gentlemen or attack gents?? (TRC, 1998a, p. 293, own emphasis). In recounting the events immediately prior to the execution, survivor Vuyani Tshabalala evokes the extreme danger of the relation of ban conjured by the metaphor of the stray dog, ?He remembers seeing accused 6 [Langa Michael Nkomo] armed with a long axe upon which was emblazoned the words: ?No peace for Mdlwembe??44 (The State v. Michael Sonti, 1999, p. 1786). However, this relation of ban, this exposure on the threshold of life and law, was a relation of extreme ambiguity in which outside and inside become, as Agamben argues, indistinguishable. These ambiguities of the boundary of the ?community? and the 44 Meaning ?feral dogs ? from the Zulu adage that says once a domestic dog has left the kraal, or homestead, and gone into the bush, it becomes wild and can never be domesticated again? (Marinovich & Silva, 2001, p. 110). 137 tenuousness of the inclusion of Moleleki residents in this community in a context of exception, were reflected in the continual struggles of linguistic transformation and denotation, to which all residents were constantly subject, and which could summarily place them, through the linguistic metaphor of the dog in relation of banishment, vulnerable to a death which was neither homicide nor sacrifice, a death of absolute abjection. As TRC Commission attorney Mr Jonas Sibonyoni asked of amnesty applicant Michael Sonti, ?Mr Sonti, let me interrupt you, the group you met referred to you as dogs but when you use those words, to whom were you referring as dogs?? (TRC, 1998a, p. 293, own emphasis). After the execution, as the corpses now produced as political artifacts lay decomposing under the hot summer sun, SDU members publicly proclaimed this biopolitical transformation, inscribed in the flesh of these corpses as embodied signs but also amplified linguistically in terms of the metaphor of the dog, now shouted like the message of a town crier45, the messenger of sovereign authority to the township of Moleleki. Martha Mthembu, mother of Ditaba Joseph Mthembu, recounts while I was preparing to go to work [I] heard a group of men singing Xhosa songs. They were shouting outside in the street saying, ?Those who know that their dogs (meaning their boys) did not sleep at home they must go look for them in the veld?. (Mthembu, 1998, p. 3) These metaphoric ?dogs? lying in the bushveld, foraged now as mere flesh by the animal genus dog in the 24 hours before their organic bodies were ?collected? by police, marked a 45 The ?town crier?, the public proclaimer of community information, decrees, instructions, and prohibitions plays a critical role in all societies, both contemporary and ancient, in which orality as a means of communication remains important. In medieval Europe, prior to the advent of print media the town crier, shouting aloud in the streets, was the central mechanism through which sovereign authority in a town would communicate public instructions and prohibitions. The etymology of the word ?cry? indicates a very specific link to sovereign authority in ancient Rome, in which, ?ancient folk etymology traces to it to the ?call for the help of the Quirites?, the Roman constabulary?(Cry. In Online Etymology Dictionary. Retrieved from: http://www.etymonline.com/index.php?term=cry). 138 final decomposition of the material and the metaphoric, logos and praxis, in the collapse of the distinction between their flesh and the bare life shared by all living beings. What was exposed in this zone of indistinction between animal and political life was the absolute undecidibility, ?as long as the iustitium46 lasts?, of human action in the state of exception, ?in the extreme case, whether human, bestial, or divine? (Agamben, 2005, p. 50). Elphina Mugadi recounts the indistinguishibility of these spheres at the site of the execution: I took the Bible and went to the place where the children had died. Before I reached the children I stood and prayed because I had already heard that any person that goes to that place may not come back?I went all the way praying saying: ?God I trust you because here is my firearm, I have my firearm??as I was walking I saw two dogs on top of the children?standing on their two legs. When I arrived I said: ?Oh Lord our children have died in a sorrowful manner, that of Jesus Christ, they have been tied up and again they have been given to the dogs?. I tried to hit those dogs, I took away the shawl with which I was beating them and the shawl then fell down to the ground. I tried to chase those dogs away but I could see these dogs? jaws were red, I do not know what these dogs were dragging but I could see blood on these dogs? jaws. (The State v. Michael Sonti, 1999, pp. 1093-1094) A Finishing Off Therefore the metaphor of the dog, in a relation of ban from the human community, materialised the sovereign exception in the unsanctionable killing of homo sacer and finally produced the corpses of the Moleleki dead in a condition of complete abandonment on the edge of township, literally collapsing logos and material practice in the consumption of the bodies of the dead by dogs. Another key metaphor, as both category of performance (verb) and category of embodiment (noun) (Feldman, 1991, p. 68) and ultimately category of ontological being, similarly effected the collapse of logos and praxis in relation to the act of unsanctionable killing itself. The metaphor of ?finishing?, used both as noun and ontological state, ?to be finished?, and verb referring to the act of killing, ?to finish off?, constituted 46 Formal suspension of juridico-political order. 139 ?indexical? terms that also referred to ?states of destroyed or altered embodiment? (Feldman, 1991, p. 69), used and shared by all protagonists in the conflict. On the day of the execution, Selina Motloung returned home to find ?my uncle sitting with other men. I greeted them, asked him how was life, he replied that the children are finished in Zulu, ?Abantwana baphekhe?. I asked him if my son was amongst those children, he replied yes? (Motloung, 1998, p. 1, own emphasis). Oscar Motlokwa in his testimony to the TRC, describing the moment of execution itself ? as Anastasia Mohale, the mother of executed Ditaba Mthembu, fainted in the hearing room (Sebolao, 1998) ? uses these indexical terms to refer very specifically, not simply to the act of killing, but to a particular type of ?finishing?, a decimation and dismembering of the body itself that could literally and figuratively effect a complete effacement, an obliteration. This human action lay in an entirely ambiguous zone between homicide and sacrifice. The mechanism of the killing itself evoked these ambiguities in an initial homicide as the victims were shot at close range. However, the decimation, decapitations and mutilations performed on the wounded bodies of those shot, invoked the form of ritual, rituals of death, which attempt a complete finality, an ontological and physical state in which what is killed is not only the physical body, but the ?essence? that lies within it. The ordinary act of homicide cannot kill this substance, which requires ritual for its final ?death?, its ?finishing off?47. 47 Ball in her study of the ?necklace? method of homicide used primarily during political conflict in South Africa refers to the role that burning had historically played in annihilating more than the physical body itself, ?Yes, there was a reason for the burning of abathakathi [evildoers] like rubbish. It was the shade. If a person was burnt with fire?there could never be a shade of that person. So if a person was umthakathi that person must be destroyed totally (ukuqedwa), nothing remaining (kungasali nokunci)? (quoted in Ball, 1994, p. 7). 140 A TRC Investigation Unit memo on the execution records this investiture of ?magic? and ritual in the execution, which sought, in the dismemberment of the corpses, a restoration of a ?momentarily injured? sovereignty (Foucault, 1977, p. 49). It states: One of the victims Edward Mootsi?s (sic) had been decapitated and his head thereafter chopped into several pieces which were then disposed of some distance from his torso. According to rumours this was done to avoid the head of the deceased returning to his body, as the perpetrators believed him to be a deamon (sic). (Killian, 1998)48 Oscar Motlokwa?s account of the execution thus evokes the ambiguities of the killing of homo sacer, which ?subtracts itself? from both the human and divine and opens a sphere of human action that is ?neither the sphere of sacrum facere nor that of profane action? (Agamben, 1998, pp. 82-83), instead taking place in a zone of un-decidable indistinction between homicide, which seeks merely instrumental death, and sacrifice, which seeks purification and ritual cleansing. Oscar Motlokwa describes the execution, They were shot and they were hacked and stabbed with knives?When we arrived at the veld all of them were ordered to sit down. Ntjebe himself starting shooting. Our commander Ntjebe Ndondolo started shooting them with a 9mm shot and thereafter, with our AK47?s, we shot them. After shooting them, those who had sharp instruments hacked them to finish them off. We shot them Sir, and some of them were still alive and they were finished off with these compasses. The rest of them were shot at their heads and the chest. We were very close to them Sir, close. (TRC, 1998a, p. 41, own emphasis) Vuyani Tshabala, one of two survivors of the execution, describes these same moments of the execution, in which logos was materialised in the attempted ?finishing? of his own body through its dismemberment: then I heard the sound of gunfire and when the gunfire started everyone of us wanted to shield himself by another one on the ground. And after they stopped firing I heard 48 Crais and others (Comaroff & Comaroff, 1997; Geschiere, 1997) have explored the contemporary and historical imbrications of political power and ?magic? in the African context, which Crais argues continues to form a critical part of the contemporary political imagination and everyday discourses of power in the South African context and is, ?often involved in the intricate and competitive politics of power and authority?(Crais, 2002, p. 130). Importantly, the discourse of magic is also a discourse of ?morality?, it is profoundly concerned with the problem of social health, the ?problem of evil?, ?Evil stands in opposition to life and, especially to the ways life should be lived? (Crais, 2002, p. 4). 141 someone say: finish them off? And after the word was given: finish them off, I then heard the sound of chopping and I was also chopped, I think four or five times on my head. (The State v Michael Sonti, 1999, p. 222, own emphasis) Elphina Mugadi, hiding in the nearby bushveld at the moment of the execution, experienced this finishing as an aural amplification, an aural definition of the moment of ?finishing?, a triumph of sovereign exception that would later be broadcast by SDU members to the township of Moleleki: And after they killed them they made a certain noise ? and I asked MaMtolo what is going on, she said in Xhosa [custom] if a person is going to be killed, after he?s been killed they make a certain sign to show that they?re finished. (E. Mugadi, personal interview, 2001) While the ritual of the killing sought a transcendental ?finishing?, its homicide was also an unequivocally political act, expressed again in terms of the metaphor of ?finishing?. Prior to the execution, at the site of the body of slain civic activist Nombulelo Zwane, ?somebody in the SDU group, from the Khumalo Valley SDU, shouted that ?we will finish off the Youth League and nobody will stop us?? (Mugadi cited in Dwane-Alpman, 1998, p. 7). The later decimations of the bodies of the Moleleki executed marked a final moment in the contest for sovereign authority in Moleleki, materially inscribing sovereign exception in a definitive moment of political metaphorization and instrumentation of the body, which transposed ?political values from social totalities to the singularity of the stiff? 49 (Feldman, 1991, p. 70). 49 In Northern Ireland, the term ?cunt? semantically marked the targeted male victim of political assassination, who then in death, in the transposition of the linguistic to the material, becomes a ?stiff? that circulates as an embodied political sign (Feldman, 1991, 69). 142 Critically, however, while the dismemberment of the bodies of the executed sought a transcendental and political ?finishing?, a final localisation of the sovereign exception in the decimated bodies of the children, what this human action instead revealed was the ambiguity of this inscription of sovereign exception in the political encoding of body parts as artefacts of sovereign power and the ultimate reversibility of this encoding of social totalities in body parts. As Feldman argues, the formation of the body as a political artifact permits it to function not only as a representation of wider social totalities but also retrospectively ?as a sympathetic (as in magic) manipulator of those overarching structures? (Feldman, 1991, p. 70). Thus while the localisation of the exception in the bodies of the Moleleki dead had sought to irrevocably represent the triumph of sovereign exception in the inscription of political codes within body parts, the body produced as enchanted political object by the dismemberments of violence in fact produces a struggle around the transcription of the biopolitical corpse produced as a political sign. The pleas of the parents of the Moleleki executed in the proceedings of the TRC and the later courtroom trial for the recovery of the children?s mutilated body parts therefore sought in the restoration of the corporal integrity of these boys, a restoration of their place within the realm of the sacred and the political and in this, finally, a transcendence of the bestiality of their deaths, which could putatively refute the production of these children as artifacts of sovereign power. Tina Mootsi, standing outside the court room during the trial of SDU members, wept repeatedly while telling journalist Dominic Mahlangu, 143 Since 1993 I have been haunted by dreams of my son, whom I was forced to bury without a head. He would have been 18 this year. They must tell me where they took his head to, or what they did with it. (Mahlangu, 1999) Pathologist?s reports reveal, however, that in fact 15-year-old Edward Mootsi?s head, after being hacked from his body, had been battered to pieces. Pathologist Jane Klepp explained: there has been total mutilation of this head, the head is here in small little bits?one can see fragments of the skull and the lacerated brain lying in the vicinity of the body, the head is gone, I mean it has been completely chopped off?and then the head destroyed itself, so it is not just cut off. (The State v. Michael Sonti, 1999, p. 891) The corporal recovery which Edward?s mother Tina pleaded for would thus never in fact happen. Selina Motloung, mother of ANCYL leader ?Wips? Motloung, attending the amnesty application of SDU members, similarly made her forgiveness of those who committed the killing conditional on the recovery of missing body parts, ?I would forgive them if they could give me Vips?s [sic] head, I should say eyes and hair, if they can give me those things, I can forgive them, but if they cannot give me those things, I cannot? (TRC, 1998a, p. 757). However, Wip?s corpse had been left for several days in the Natalspruit river, making impossible the corporal recovery his mother sought. As pathologist Jane Kleep explained, We find that if a person has died in water the fish and crabs move in very quickly and they eat away the lips and eyes and the ears?and give you this gross appearance that we have of his face as demonstrated in photograph 1. (The State v. Michael Sonti, 1999, p. 869) In these parents? attempted recoveries of the body parts of the maimed children and the subsequent processes of juridico-political inscription around the execution, these bodies therefore became not the irrevocable signs of sovereign authority that SDU members had attempted, but instead began to effect a reciprocal political metamorphosis of the ?overarching structures?, the social totalities which SDU members had sought to finally inscribe in the dismembered corpses of the executed. In the wake of the execution, the SDU 144 as a structure of sovereign authority was literally and metaphorically dismembered, its arms confiscated, its structures officially disbanded and finally in the subsequent jailing of some of the protagonists of the execution, a tenuous inscription of this anomic body of sovereignty within the juridico-political order of post-apartheid South Africa was made possible. The bodies of the Moleleki dead thus became part of a process of making feasible not only a local reciprocal political metamorphosis but a wider national political metamorphosis, which would seek to refute the political values that SDU members had attempted to finally inscribe in these bodies produced as political signs by the execution. 145 CHAPTER FOUR: A JURIDICO-POLITICAL INVESTIGATION Introduction This chapter begins the process of examining the juridico-political inscription that followed the production of the Moleleki corpses as the body of exception. These processes of re- inscription traversed both the period of juridico-political exception in which the execution took place and the period subsequent to the inauguration of the post-apartheid state and shared, paradoxically, an essential contiguity in their efforts to recapture this ?pure? violence unleashed from law, within the law again. Thus, although the norm of the law changed significantly from one period to another, the principle of inscription remained essentially the same, immediately evoking in both periods a critical ambiguity around the nature of this human action committed during a context of exception, which it was assumed concerned the breakdown of juridico-political rules, a lacuna in law rather than the exposure of sovereignty at the boundaries of the juridico-political order. This chapter therefore examines the processes that followed immediately after the execution, when the corpses of the young ANCYL members, first produced as signs of sovereign exception by the SDU in Moleleki, were now reproduced as the signs of sovereign exception by the apartheid state. Initially their corpses were identified at the site of the execution as those of ?criminal hijackers? and were claimed by the apartheid state as the embodiment of criminal exception. The subsequent processes of medico-legal inscription, which sought to translate this mass of organic flesh into a ?legible? narrative of cause and effect through post- mortems in order to make subsequent criminal prosecution possible, ran aground in a context of ius belli (legal war). The pressure of multiple corpses produced by this conflict meant that that the identification and ?mapping? of the wounds of the body, which sought to definitively establish a cause of death, remained incoherent and contested. In the immediate post- 146 apartheid period, as the work of the TRC was initiated, a new complexity was introduced around the bodies produced by the Moleleki execution, concerning a contestation regarding the nature of the human action which had led to the production of these bodies, whether ?criminal?, and thus properly the domain of the formal juridical system and hence subject to the legal sanctions it was empowered to impose, or ?political? and therefore eligible for consideration by the TRC Amnesty Committee and the suspension of legal consequence, which this body was authorised to dispense. A Mutable Truth, the Juridico-Political Investigation of the Body of Exception The charge sheet compiled by the Witwatersrand Attorney-General for the Supreme Court tabulates the bodies of the Moleleki deceased in the following way: 13. The deceased died as a result of the wounds as set out hereunder: Count Deceased Cause of Death 1 Mile Simon Simela Bullet wound of left and right jugular vein and trachea 2 Peter Mavuso Mdishwa (sic) Bullet wound of lung 3 Ditaba Joseph Mthembu Bullet wound of Brain, Lung, Liver and intestines 4 Isaac Mgadi Bullet wound of intestines and Aorta 5 Lucas Hlatshwayo Bullet wound of brain. Penetrating incised wounds of lungs 6 Thokozani Buthelezi Bullet wounds of Brain 7 Itumeleng Edward Mootsi Fractured Skull 8 Alfred Philemon Buthelezi Crushed Skull. Lacerated Brain 9 Isaac Mbitjana Motloung Consistent with drowning The deceased were all members of the ANCYL. (The State v. Michael Sonti, 1998, p. 140) In the wake of the execution the apartheid state, which had been strikingly juridically and politically absent from Moleleki as it was in most other South African townships, now 147 engaged in death with these young residents of Moleleki, not as juridical subjects or citizens of a nation state, but as mere bodies. In death, these bodies became the object of a juridico-political process of inscription from the moment their decomposing corpses were first photographed and notarized by local police more than twenty-four hours after the execution. The subsequent post-mortems conducted on these maimed bodies initiated a medico-legal inscription of death as they arrived in the heat of the summer of 1993 in the Germiston mortuary, which was already stacked with the bodies of those who had been killed during this period. The waning juridico-political order of the apartheid state therefore set in motion a process of attempting to capture this anomie in law and in this inscription of anomie to make legible this manifestation of violence, this now undifferentiated mass of organic flesh. This was a process that would pursue these bodies through the next five years across the inauguration of a new political dispensation, in the processes of the TRC and the post-apartheid court room, endeavouring through these processes to finally capture this elusive manifestation for a new law, a new juridico-political order. Sovereign Police During the 1990s, in a context of anomie and of exception, these young people, whom the juridico-political order now attempted to inscribe in death, had been particularly subject to the repressive force of law, without law, ?pure violence?. This ?pure violence? unleashed from law in the state of exception was claimed by both the SDU, as local sovereign power in the context of the execution, and by the security institutions of the state. As Nathan and Philips argued, in the context of exception of the 1990s, ?in the interregnum between apartheid and democracy, the SAP and SADF are not merely instruments of government 148 policy but actors in their own right; they are an unpredictable factor, with the potential to disrupt the transition process? (Nathan & Phillips, 1992, p. 112). Critically, it is the institution of the police which shows the separation of law from the ?force of law? most distinctively. As Benjamin argues: The assertion that the ends of police violence are always identical or even connected to those of general law is entirely untrue. Rather, the ?law? of the police really marks the point at which the state, whether from impotence or because of the immanent connections within any legal system, can no longer guarantee through the legal system the empirical ends that it desires at any price to attain. (Benjamin cited in Agamben, 2000, p. 103,5) Thus, the police, ?contrary to public opinion? ?are not merely an administrative function of law enforcement? (Agamben, 2000, p. 103,5); instead they mark the boundaries of this juridical-political power and evoke the sovereign decision on the exception. The police, therefore, Agamben argues, are always operating within a?state of exception. The rationales of ?public order? and ?security? on which the police have to decide on a case-by-case basis define an area of indistinction between violence and right that is exactly symmetrical to that of sovereignty. (Agamben, 2000, p. 104,4) The Internal Stability Division of the South African Police force represented, perhaps most nakedly, this area of indistinction at the boundaries of the juridico-political order and the entrance of the concept of sovereignty into the figure of the police, in particular as the apartheid state rapidly declined and its legal system proved entirely unable to guarantee the empirical ends of this state, especially the ?government?s frequently stated commitment to maintain law and order in the transition period? (Nathan & Phillips, 1992, p. 123). Thus in March 1992 the South African state, after having withdrawn the South African Defence Force (SADF) from South Africa?s townships under massive internal protest at the end of the 1980s and as part of a wider shift in state security strategy which saw a decline of 149 the dominance of the military in state governance in general (Nathan & Phillips, 1992; Taylor & Shaw, 1998), now responded to the politicised violence sweeping across the country by creating a separate unit of the South African Police (SAP), the Internal Stability Division (ISD), which was in fact the organisational successor to the infamous ?Riot Unit? of the SAP that had violently policed South Africa?s townships for decades. The ISD was thus launched in 1991, specifically tasked with public order functions (Rauch & Storey, 1998, para. 17) and operating essentially as a paramilitary unit; ?although it was never openly stated, the ISD was? designed to be the hard point of the SAP, in a position to counter ?fire with fire? and so serve to contain violence and stabilise conflict torn areas? (Shaw, 1997, pp. 230-231). The establishment of the ISD was part of a wider internal post-1990 SAP strategic initiative to ??depoliticise? the police and re-orientate their focus from ?crimes against the state? to ?crimes against individuals?? (Nathan & Phillips, 1992, p. 116). In the light of this broader strategic alignment within the context of political transition, the Internal Stability Division was established with the ostensible objective of ?defusing political violence and?freeing other police units to ?fight crime?? (Levin, Ngubeni & Simpson, 1994, p. 2). Thus, The creation of the ISD as a specialised public order component separate from the rest of the SAP was premised on naive ideas? that the performance of the ISD would not impact on the image and capacity of the rest of the SAP to carry out other police functions. (Rauch & Storey, 1998, para. 17) Critically, however, as Agamben points out in relation to both the Gulf War and the Nazi extermination of the Jews, both devastating conditions of ius belli that were conceived of and conducted as mere ?police operations?, this ?de-politicisation? of war cannot only be considered to be ?a cynical mystification?, a mere naivety, but is instead about the entrance of sovereignty into the figure of the police, a conjunction premised on the criminalisation of the ?enemy?, ?a process by which the enemy is first of all excluded from civil humanity and 150 branded as criminal; only in a second moment does it become possible and licit to eliminate the enemy by a ?police operation??. Such an operation is not obliged to respect any juridical rule and can thus make no distinctions between the civilian population and soldiers?thereby returning to the most archaic conditions of belligerence? (Agamben, 2000, p. 106,7). This condition of exclusion from civil humanity is the condition of homo sacer. On the East Rand, this condition of ius belli, in which the ISD, purportedly a mere police unit but in fact a critical protagonist in the civil war taking place in the region?s townships, made the civilian population ?criminal?, subjecting its residents to the arbitrary and anomic force of law, loosed from law, as the ISD units which operated in these townships had ?no structural relationship with local police commanders and conducted their activities with no regard to local communities? (Rauch & Storey, 1998, para.18). Repeated allegations that members of the ISD were involved in assault, torture, random shooting, looting and property damage led residents to hate and fear the unit (Peace Action, 1993)50. In Moleleki this was a view shared by all protagonists in the conflict. Civic leader Lethusang Rikaba, questioned by a TRC commissioner on why the police were not called in the wake of the Moleleki execution, reacts with disbelief to this question: ?I think I have alluded to the fact that the situation in Katlehong was to such an extent that co-operation was not possible between the police and the community?I would never dare do that because we did not trust them? (TRC, 1998a, p. 670). SDU member Michael Nkomo similarly states, 50 The report of the TRC comments on the role of the police post 1990: ?Violence?arose from the continued use of lethal force in public order policing. The HRC [Human Rights Committee] estimated that killings by the security forces, primarily in the course of public order policing, numbered 518 between July 1991 and June 1993? (TRC, 1998b, p. 585). 151 ?There was no co-operation with the police at that time, I hope you know the situation that was in place at that time? (TRC, 1998a, p. 438) Therefore on the East Rand, the unit?s conduct led to escalating protests from residents living in a state of necessity and civil conflict, who clearly saw the ISD as another role-player in the conflict, whom they ?accused of deliberately targeting and eliminating the leadership structures of the ANC-aligned SDUs. Furthermore, there were accusations that the ISD was involved in clandestinely supplying hostels with arms and of assisting hostel residents in their attacks? (Shaw, 1997, p. 232). Oscar Motlokwa explains, of the situation in Katlehong, ?if there is a fight between ANC and IFP the police would take the side of the IFP, then we would regard them as our enemy because of their actions? (TRC, 1998a, p. 11). It was the ?youth? in particular, the leadership of previous generations of political resistance to the apartheid state, who in these conditions of ?archaic belligerence? were most extensively subjected to the depredations of the ongoing ?police operation? on the East Rand. Young people were materially essentialised as the body of the ?criminal?, thus the body of the ?enemy? whom it was ?licit to eliminate? as the assassination of numerous youth leaders, defence unit members and former Umkhonto we Sizwe guerrillas during this period made explicit51. Lethusang Rikaba, civic leader in Moleleki, explains the extreme vulnerability of the ?youth?, which left them on the threshold of life and law, subject to the arbitrary violence of the ISD: ?Now for them [the ISD], every young man is a criminal, they don?t choose, that is why the youth ran away from the Stability Unit? (TRC, 1998a, p. 671). Elphina Mugadi, 51 The TRC report states of the period between 1990 and 1994, ?During this period political opponents continued to be killed in circumstances which pointed directly to security force involvement? (TRC of South Africa, 1998, p. 592) and that ?The HRC [Human Rights Committee] recorded large numbers of political assassinations during the early 1990s, the victims of which were largely office-bearers of the newly unbanned ANC, MK members or members of allied organizations?(TRC of South Africa, 1998, p. 594). 152 mother of Isaac Mugadi, one of the ANCYL members killed in the Moleleki execution, recounts: ?There were problems with the ISD, they would pick up children, Isaac was picked up twice and beaten? (E. Mugadi, personal interview, 2001). It was this fear of the ISD which would eventually place the children of the Moleleki execution in a relation of abandonment and danger, as they gathered together on the eve of the execution in an unoccupied shack on the edge of the township, in anticipation of an ?attack? by the ISD, which SDU members had told them was imminent. A number of these ANCYL members were reportedly later captured from this shack by the SDU members who had directed them there. This anomic ?force of law? unleashed by the state on South Africa?s townships in the context of exception, was Moleleki township?s only substantive experience of the state?s security presence during the period prior to popular elections in 1994. As Elphina recounts: ?At that time it was just stability coming here? (E. Mugadi, personal interview, 2001). The unit came in a condition of war, either, as Lethusang Rikaba explains, to ?attack? residents (TRC, 1998a, p. 671) or to collect the bodies of this war, arriving in Moleleki not to intervene in defence of its residents as juridical subjects but merely to collect their corporal bodies. Bulelwa, Blanko and the nine Moleleki executed were all casually ?collected? many hours after their deaths by the ISD, simple material objects to be cleared from the space of Moleleki. Collecting the ?criminal? body of exception Elphina describes the way in which the Moleleki executed were finally collected by the police at least 24 hours after their death: When I arrived I sat down next to the eight corpses or bodies. A policeman then came out and said I am asking for one of the parents, I would like to load these bodies so that the body should not get lost. Some of the parents were afraid. I then stood up and 153 said I am one of the parents, a parent to Isaac. (The State v. Michael Sonti, 1999, pp. 1094-1095, own emphasis) At approximately 11am on the morning of 8 December 1993, Sergeant Andre Hermanus van Vreden, who was stationed with the Unrest and Violent Crimes unit of the SAPS, arrived at the scene where the corpses lay. He had received a call at approximately 10.30am that morning from an ?unknown black man?, informing him that eight bodies were lying in the veld next to Moleleki squatter camp. At 11am when Sergeant van Vreden arrived at the place identified by the anonymous caller a large group of people had already congregated at the site (Van Vreden, 1993). In a final gesture of material essentialisation of the body of the youth as ?criminal?, the mute corpses of these young residents, who could now never stand trial as juridical subjects (some of whom, as legal minors, could never have in fact been engaged as adult juridical subjects), were now in this condition of death, ?judged on account of a crime? (Agamben, 1998, p. 29), as local police brought employees of a paraffin company, which had been repeatedly robbed during deliveries to the township, to the site of the killings to ?identify? these decomposing corpses as the ?perpetrators? of ?crime? in the township. Thus among the crowd at the scene of the killings were three truck drivers working for the Benoni-based52 paraffin company, Royal Paraffin ? Frans Tau, Josia Ngwenya and Herman Mabokela, who had been the victims of these hijackings (Mabokela, 1993; Ngwenya, 1993). 52 Benoni, was one of several ?white? towns which were at the centre of South Africa?s manufacturing economy on the former East Rand, now the Gauteng municipality of Ekurhuleni, which were supplied with labour by a rapidly growing black labour force that settled in what were known as the ?Kathorus? townships, namely Katlehong, Thokoza and Vosloorus. 154 Less than a week before the execution, on 1st December 1993, Frans Tau accompanied by his two assistants, Josia Ngwenya and Herman Mabokela, were stopped by a group of about 17 ?youths? armed with axes and ?kierries?, who hacked open the safe containing the day?s cash, when they tried to deliver paraffin in Moleleki. Tau, along with Joshua and Herman, fled the scene (Tau, 1993, p. 2). After this incident, the owner of Royal Paraffin, Mr Moodley, cancelled all further deliveries to Moleleki until he could ?sort things out with the Civic Association in Katlehong? (Moodley, 1993). On Wednesday 8 December at approximately 8am, Tau received a call from the owner of the Moleleki coal yard, who assured him that he could deliver paraffin safely to Moleleki, informing him that the people who had robbed him were dead and that they had been killed by ?people? from Moleleki (Tau, 1993)53. Later the same day Frans Tau reportedly identified to Sergeant van Zelderen of the Katlehong Unrest and Violent Crimes Unit, four of the bodies of the Moleleki executed as the people who had robbed him (Tau, 1993, p. 3). Also on Wednesday 8 December at about 9am, drivers Josia and Herman arrived at Inyoni Park police station to ask, as they frequently did, for a police escort into Moleleki. They were told that all policemen at the station had already left to investigate a murder case in Moleleki Extension 2. However, not long afterwards, some police returned and asked Josiah and Herman to accompany them to the scene of the murder in order to ascertain whether any of the dead were the ?black males? who had robbed them. Josiah identified two of the bodies 53 Tau?s statement records, ?On Wednesday 93-12-08 I was at work again when the telephone rang, the person said that he was the owner of the Moleleki coal yard. He told me that I can deliver paraffin and I mustn?t be worried because the people who robbed me are dead and further that the people from Moleleki had killed them? (Tau, 1993, p. 3). Original: Op Woensdag 93-12-08 was ek weer by my werk, die telefooon het gelei?die persoon het aan my gese date hy die einaar van die Moleleki kole werf is. Die man het aan my gese dat ek parafien moet kan afbewer en nie moet bekommerd wees nie, aangesien die mense wie my geroof het dood is en verder dat die mense van Moleleki hulle doodgemaak het (Tau, 1993, p. 3). 155 and Herman three of the bodies as those of youths involved in the robbery on 1 December. Both conclude their police statements with the identical sentence: ?Nobody is allowed to robbed (sic) anybody or point fire-arm to anybody? (Mabokela, 1993; Ngwenya, 1993). A medico-legal inscription of the body of exception, notarising death In the wake of the Moleleki execution, which had materialised the exception in the bodies of the executed, a process of juridico-political inscription began which sought to retrieve this anomie for law, by tracing a relation between cause and effect, a relation of juridical instrumentality, which could not recognise the ambiguity of the biopolitical transformation that it in fact sought to elucidate. The medico-legal processes of the post-mortem, through which the apartheid state sought a final ?mapping? of the body of the apartheid subject, was a ?mapping? which had pursued this body in various statutory and discriminatory prohibitions in life and which now pursued this body beyond death. This ?mapping? sought to render this corporal mass ?legible? (Pugliese, 2002, p. 380) in terms of a medico-scientific anatomical ?atlas? (p. 374) which reconstituted this body in terms of regime of medical hermeneutics and medical visuality that would ?translate? (p. 369) this dismembered organic mass into an evidentiary object, making this body with all its wounds and mutilations, legible to law, a medico-legal sign, rendering it thus in a relation to law. Critically, what this anatomical ?reading? (p. 374) of the organic mass of the body makes possible is the inscription of this body within a medico-legal regime of instrumentality that seeks to determine ?the precise cause of death? (p. 369) through a series of instrumental mediations, ?to ascertain whether or not any foreign agent was employed and the mode in 156 which the instrument used (if any) produced the effect? (p. 370), and hence to make it possible to trace a linear relation of cause and effect, which ?discloses the hidden and effaced cause of death? (p. 381) and the agents of its inscription. Thus, in forensic pathology, the practices of visuality and language work hand in hand in making legible the signs of violence and trauma that mark the body of the victim and in transmuting these signs into coherent narratives of cause and effect. (p. 369, own emphasis) This process of transmutation and translation of the traumata of the body into a narrative of cause and effect, moving from the exterior to the interior spaces of the body (p. 368), begins from the moment the body is first encountered by the policing agencies of the state, which, in order to locate the body within a narrative structure which will make possible the disclosure of the precise cause of death, ?orders a vast array of forensic practices, beginning with the photographing of the victim?s body in situ and extending to the presentation of visual evidence?in the courtroom? (p. 370). At 11am on the morning after the Moleleki execution, Sergeant van Vreden of the Unrest and Violent Crimes unit of the SAPS arrived at the scene of the execution and at the discovery of these bodies initiated a process of narrativisation which would track ?when, where and to whom? (Pugliese, 2002, p. 370) control of the body was relinquished, inscribing these bodies in a ?chain of evidence? that would putatively maintain the body ?intact?, in its ?original? form, from ?the site of injury?to the pathologist? (Pugliese, 2002, p. 370), an ?originary state? which in fact, occluded the narrative of the chain of evidence in which the body had been inscribed and which were actually the condition of its intelligibility within each recorded stage of its ?originary condition? (Pugliese, 2002, p. 370). 157 Van Vreden notes the first stage of this condition of originality in his report of the Moleleki bodies in situ at the crime scene, recording that he had found eight black ?men? lying in the veld. Seven were bound together with rope around their hands. One man was lying separately, also bound. Van Vreden comments further that the bodies were ?bundled together? with their heads in a ?southerly direction? (Van Vreden, 1993, p. 1). The dead had a variety of bullet and head wounds. Sergeant van Vreden found 14 spent AK47 cartridges on the scene, which he handed over to Sergeant Hayle from Springs police station. Van Vreden then talked to members of the crowd gathered at the scene. According to his rough sketch of the scene, ?deceased number seven? was identified to him as Isaac Mugadi, ?a 17-year-old Xhosa?, by a ?black woman?, Mugadi?s mother (Van Vreden, 1993, p. 1). ?Deceased number eight? was identified by Paulina Moketsi as Miles from 7215 Moleleki Extension 2. Sergeant van Vreden concludes his report by noting on his sketch that the skull of ?deceased number four? was crushed and pieces of the skull were lying around (Van Vreden, 1993, p. 1). Sergeant Van Vreden, Deputy-Detective Johannes Pieterse and Detective Lance-Sergeant Andre Viljoen, all at the scene of the killing, thus collectively conclude their statements with a standard confirmation of the synchronic originality of the bodies at the moment at which they were found: ?While the bodies were under my supervision they did not acquire any further wounds?54 (Pieterse, 1993; Van Vreden, 1993; Viljoen, 1993), an originality which was confirmed as intact before the bodies were handed over to the next authority in the chain of evidence, Constable Elijahr Mapshoane from the Germiston mortuary (Van Vreden, 1993). 54Terwyl gemelde lyke onder my toesig en beheer?het dit geen verdure beserings opgedoen nie. 158 Constable Mapshoane?s main duties at the time were the collection and delivery of bodies to the Germiston mortuary (The State v. Michael Sonti, 1999, p. 1216). At the site of the bodies, Mapshoane filled out an SAP 180 form stating the place and time at which the bodies were found, the name of the detective who oversaw the bodies until their removal, visible injuries, and the person who received each body for transportation to the Germiston mortuary, namely himself. The sex and race of the bodies were also noted. At the Germiston mortuary, completed SAP 180s were kept in a file for the pathologist and other official reference (The State v. Michael Sonti, 1999, p. 1229). On arrival at the mortuary the bodies were again recorded in the ?condition received?. Each body was ?tagged? and allocated death registration numbers (The State v. Michael Sonti, 1999, p. 203), again inscribing the bodies in a narrative chain of evidence. At 4.31pm the bodies were thus unloaded from the lykswa (hearse), washed down and assigned numbers from the death register, a ?big book? containing the SAP form 183, on which all known details of the dead were recorded as they were ?admitted? to the mortuary (The State v. Michael Sonti, 1999, p. 1231). After ?admittance?, Constable Masidele Benjamin Mbashlehle, a colleague of Elijahr Mapshoane, took photographs of the bodies, for the purposes of examination by a pathologist (The State v. Michael Sonti, 1999, p. 1230). Constable Masidele Mbashlehle, who had received no training or supervision, struggled to carry out his brief and frequently failed to capture on film the wounds which had caused death or injury (The State v. Michael Sonti, 1999, p. 172), contesting the possibility of a synchronic recording of the ?original state? (Pugliese, 2002, p. 370) of the body at the time of receipt by the pathologist in what district surgeon Nienaber described as a condition of war (The State v. Michael Sonti, 1999, p. 195). 159 Doctor Nienaber, who was working as a district surgeon at the Germiston mortuary at the time of the execution, explains the state of necessity that produced this disjuncture between wounds described and wounds photographed: we had violence on the East Rand. We got up to 60 post-mortems between two district surgeons. So we were under big pressure, to do work and I think?the person that takes the photographs did not?look for wounds, before the photographs were taken. (The State v. Michael Sonti, 1999, p. 172) After the bodies had been washed, tagged, numbered and photographed, they were placed above and among bodies already stacked to the ceiling waiting for examination by a district surgeon before burial. Autopsies did not take place in a separate room, district surgeons worked among the corpses. Once the bodies of the Moleleki victims were finally placed on the stainless steel of the pathologist?s table for examination, these now decomposing corpses were brought into a new visibility through the medico-legal gaze of the forensic pathologist, even as they in fact lost their material corporal identity. Thus a series of techniques of visuality established the reality of this material body ?as though, prior to the deployment of a series of medico legal practices, the body is obscured by a shroud that renders it invisible? (Pugliese, 2002, p. 368). This was achieved through the meticulous spatial mapping and documenting of the location and positioning of wounds and a ?detailed analysis of the surface areas of the body? (p. 371). Critically, the bringing into visibility of the body which establishes its evidentiary value must apply logos to this manifestation, must anchor this materiality in language, through the device of the autopsy report, which orients the medico-legal gaze in relation to the visualised object in terms of language, thus rendering ?the organic articulable? (Pugliese, 2002, p. 381) and ?in the context of the law, in which the logos of the word governs all other systems of signification, the visual and the visible must literally become language, or alternatively, must 160 be anchored and explained by language? (Pugliese, 2002, p. 381). This articulation of the silent organic corporality of the body, the bringing of this body into speech and language, enables the corpse to metaphorically ?speak? (p. 369) again, through the forensic pathologist, after a ?rigorous and scientific analysis of the traumata of the body? (p. 369) through which the forensic pathologist ?ventriloquises? this death (p. 369), its causes and its last moments. In this ventriloquisation of death through forensic pathology, a process of juridico-political inscription begins in which ?man?, the animal that has language, separates himself from his own bare life and effects in this distancing a biopolitical transformation between bare life and ?political life?. Thus the post-mortems performed on the Moleleki dead sought to achieve such a transformation and retrieve this flesh made animal for logos and law. However, this retrieval of flesh for law, which attempts to translate the visual data generated by the corporality of the dead body into a linguistic narrative of a relation between cause and effect, remains deeply ambiguous and irrevocably contingent, a juridical ?truth? about the circumstances and causes of death, which in its very statement of causality, questions causality and thus can only remain infinitely contestable. As Didier Coste argues, all narrative discourse, including causal discourse, questions causality as much as it states or suggests it. An attorney defending a criminal case will use this well known device to dissociate verbally the death of the victim from the gesture of the murderer; the defending attorney analyzes the prosecutor?s claim that ?X shot Y dead? into an unfortunate coincidence between ?X shot? and ?Y died?. (Didier Coste cited in Pugliese, 2002, p. 369) The state of ?chaos? (The State v. Michael Sonti, 1999, p. 209) at the Germiston mortuary, in which the bodies of the dead, piled on top of each other, were stored in a condition of ?bare life?, as mere flesh, disputed the possibility of the articulation of a coherent narrative of cause and effect around the Moleleki execution. Under these conditions of war and summer heat in December 1993 the pressure to complete post-mortems before decomposition had wrought 161 irrevocable damage to bodies was intense. The autopsy reports filled in by Nienaber and Bekker reveal that they both averaged 15 minutes per autopsy (Bekker, 1993; Nienaber, 1993). Court records reveal that the quality of their work was erratic. Wounds photographed were not described; wounds described were not photographed; exit and entrance wounds were confused, as were the site or type of wound (The State v. Michael Sonti, 1999, p. 172) and medical terminology was improperly applied. This led senior state pathologist Jane Klepp to conclude, With all due respect the quality of the post-mortems is very bad. You know the terminology, the wounds that are missing, the lack of photographs, but that is just to really explain why some words have been used that shouldn?t be. It is just a very poor post-mortem. (The State v. Michael Sonti, 1999, p. 858) Thus what was demonstrated by these processes was the difficulty ?of locating definitively any cause within a linguistic system of signification? (Pugliese, 2002, p. 388). This is a question concerning causation in both science and law which Pugliese (2002) argues, ?remains fundamentally irresolvable? (p. 388). However empiricism (in science and law) is in fact founded on this fundamental irresolvability. These fundamental ambiguities of cause and effect were made explicit in the struggles over the medico-legal transcription of the bodies of the Moleleki dead, as the ?reading? of these corpses by district surgeons Bekker and Nienaber was extensively contested in court by senior state pathologist Jane Klepp. A Juridico-Political Investigation of Death This ambiguous medico-legal transcription of the organic mass of the body into something putatively, legible and juridical was undertaken in 1993 in anticipation of a further process of legal inscription, namely a subsequent police investigation and court trial, which could ostensibly capture the protagonists of this death and exact a juridical punishment. However, the context of ius belli meant that although a murder docket was opened after the Moleleki 162 execution, the investigation did not proceed further as Moleleki residents, including the relatives of the deceased, were unwilling to work with the agents of apartheid law and make available the information that would have made a police investigation possible. A press report on the trial of the Moleleki accused recounts the conditions of enmity which prevailed on the East Rand at the time: Police collected at least 140 corpses on the streets of East Rand townships over about eight months, during the early 1990s, yet local communities insisted on keeping silent about the circumstances surrounding the deaths. In court yesterday, policemen based in the East Rand townships of Katlehong, Thokoza and Vosloorus (Kathorus) during the violence-ridden early 1990s painted a picture of death, mayhem and distrust between communities and police. (Moya, 1998) Investigating officer Lieutenant du Plessis explains: ?No witnesses could be found who could help with the investigation. No cooperation could be got from the public? 55 (Du Plessis, Unrest and Violent Crime Unit, letter, 4 October, 1999). Consequently the original investigating officer, Inspector Leon van Zelderen, was led to conclude that ?as a result of fruitless previous investigation, the case was prepared for an inquest?56 (The State v. Michael Sonti, 1999, p. 1194). Therefore, following the Moleleki execution, because of the repressive excesses of apartheid law, none of the protagonists to the Moleleki execution, neither survivors, witnesses nor relatives considered using this law to attempt to bring a legal case against the perpetrators. Instead civic leaders in Moleleki ?reported? the matter to SANCO. As civic leader Lethusang Rikaba explained, 55 Geen getuies kon gevind word, wat kon help met die ondersoek nie. Geen samewerking kon van die publiek gekry word nie. 56As gevolg van vrugtelose vooraf ondersoek is die saak voorberei vir die hou van ?n geregtelike nadoodse ondersoek. 163 I think I have alluded to the fact that the situation in Katlehong was to such an extent that co-operation was not possible between the police and the community so it was my duty to inform the bigger organisation which is SANCO or rather the Civic. (The State v. Michael Sonti, 1999, p. 670) In the wake of the execution itself, the ANC, which had already been requested to intervene in escalating tensions between the ANCYL and the SDU in the township, finally sent its representatives to address the atrocious consequences of this battle for sovereignty in the township. Critically, however, this ?intervention? did not involve an engagement with the apartheid law of the state. Thus following the Moleleki execution, Afrika Khumalo, head of the then PWV57 ANC Department of Intelligence Services, was sent by the regional leadership of the ANC to Moleleki to ?intervene? and ?bring the ANCYL and the SDU in Moleleki together? (Dwane-Alpman, 1998, p. 11). Partly by disarming both parties, he sought a path of ?reconciliation? which, rather than exacting individual punishment of the offenders through the judicial system of the apartheid state, attempted the restoration of ?balance in the community? (Crais, 2002, p. 86). Thus Khumalo posited a ?political solution to the problem? (Ndlozi, 1997, p. 3). This ?political? solution, however was not a party political solution but concerned a question of local juridico-political ordering in terms of which ?it was agreed that it [the conflict] would be solved the ?African way? because this was a domestic issue? (O. Motlokwa, personal interview, 2001). Thus after Afrika Khumalo had ensured that the SDU in Moleleki was disarmed, he concluded that the ?situation? had been ?defused? and ?left it with the local leadership? (Ndlozi, 1997, p. 3). 57 The former Pretoria Witwatersrand Vaal (PWV) region of what was then the Transvaal province, which is now know as Gauteng province. 164 Police investigation, a juridical application It was only in 1995 after the inauguration of a new political dispensation that the witnesses, survivors and families of the Moleleki dead claimed a new juridical citizenship through the deployment of the national jurisprudence of the post-apartheid state against the protagonists of the execution. This contested the ?domestic? exercise of ?African? sovereignty that SDU member Oscar Motlokwa articulated and which had in fact claimed the lives of the Moleleki executed. Thus, ?the first statements from witnesses were only provided to the South African Police some two years after the incident? (The State v. Michael Sonti, 1999, p. 1775), as Moleleki citizens sought to ensure that the execution would be juridically investigated and the processes of juridical inscription, which had remained suspended in the condition of exception of the pre-1994 period, would take place. Consequently, during July and August 1995, the police, following allegations made to them in connection with the execution, for the first time actively solicited statements from a number of relatives and witnesses in connection with the Moleleki execution. As a result four suspects were arrested and appeared in court in September 1995. By December 1995 nine men had appeared in the Alberton58 Magistrates Court on charges of murder related to the execution, among them some of the final accused, Zola Michael Nceba Sonti, Siviwe Malcoms Ngam, Michael Armoed and Bethwell Tabiso Ntoma. One of those charged, Bethwell Ntoma, was employed as a community policeman, stationed at the Katlehong police station at the time of his arrest. In January 1996, the case faltered when the accused appeared again at Alberton Magistrates Court before Magistrate Bandjies, who refused to postpone the case further pending instructions from the Attorney-General?s office, which was at the time in 58 Alberton, also part of the industrial complex on the former East Rand, now forms part of the Gauteng municipality of Ekurhuleni. 165 possession of the docket. The case was withdrawn and all the accused were released. Despite this setback, the continued investigation of the case was actively encouraged by the policing agencies of the new juridico-political order, which saw in these processes of juridical investigation an opportunity for the establishment of a new norm of law. Inspector Baumann of the East Rand Murder and Robbery Unit, the third investigating officer assigned to the Moleleki murder docket, in the wake of the withdrawal of the case in January 1996, articulated this concern for the recovery of the norm of law in a juridico-political order, in which for the first time, the Moleleki execution had become a subject of juridical investigation, not an object of local ?custom?. Inspector Baumann explains: When the case against the suspects was withdrawn before I was the investigating officer, the community was angry with the police. The community felt that they had not done their work. After a certain incident and after the arrests then the community began to trust the police. The community then began to now come forward again. They felt that the police now are doing something?It was not like in the past that we must go look for witnesses. (The State v. Michael Sonti, 1999, p. 1272) 59 Thus in 1996, while attending meetings of the recently established Community Policing Forum (CPF), set up in the immediate post-apartheid period in order to bridge divisions between township communities and the South African Police services, Inspector Baumann became party to discussions of ?the Moleleki incident? and during these discussions to information that there were reportedly further witnesses who could testify in relation to the killings. Inspector Baumann worked closely within the CPF with civic leader Lethusang Rikaba, survivor Samson Mokoena and Elphina Mugadi, mother of the one of the executed Moleleki children, who all served as members of the Community Police Forum and 59Toe die saak teruggetrek is teen die verdagtes voordat ek nou die ondersoekbeampte was, was die gemeenskap kwaad vir die polisie. Die gemeenskap het gevoel hulle het nie hulle werk gedoen nie. Na ?n sekere voorval en na die arrestasie toe het die gemeenskap weer begin vertrou vind by die polisie. Die gemeenskap het toe nou weer na vore gekom. Hulle het gevoel die polisie doen nou iets?Dit was nie soos in die verlede dat ons moes gaan soek het na getuies nie. 166 apparently ?assisted Inspector Baumann in regard to the investigation?. The closeness of the relationship between the Inspector and these CPF members eventually lead Justice Boruchowitz to question the credibility of some witnesses in the court trial of SDU members (The State v. Michael Sonti, 1999, p. 1777). In November 1996, despite the initial withdrawal of the case due to repeated postponements, the office of the Attorney-General issued an instruction that ?efforts to re-arrest the accused must proceed? (The State v. Michael Sonti, 1996, p. 1)60 as a result of the continued investigation of Inspector Baumann, which led him to identify further witnesses to the case. In 1997, in response to the instructions of the Attorney-General and more than a year after their release, the accused were rearrested. The seven accused are described by the office of the state prosecutor and the Attorney- General as: 1. Zola Michael Nceba Sonti @ [aka] Ncaba, 39 years old, Zulu speaking, residing at 7080 Moleleki 2. Siviwe Malcolms Ngam @ Manzini, 42 years old, Xhosa speaking, residing at 7056 Moleleki 3. Michael Armoed @ Rooivark, 39 years old, Afrikaans speaking, residing at 6982 Moleleki 4. Betuel Tabiso Ntomo @ Thabiso 30 years old, South Sotho speaking, residing at 7477 Moleleki 5. Petros Mtembu @ Vusi, 31 years old, Sotho speaking, residing at 8470 Moleleki 6. Langa Michael Nkomo @ Michael or Nkomo, 35 years old, Sotho speaking, residing at 8141 Moleleki 7. Oscar Mohale Motolkwa, 34 years old, Pedi speaking, residing at 479 Moleleki/Leeukop Prison (The State v. Michael Sonti, 1997, p. 1) 60 Pogings om die beskulgigdes te herarresteer moet voortgaan. 167 TRC investigation of the Moleleki execution, a juridical suspension However, this renewed process of juridical investigation and inscription, initiated by Moleleki citizens in the wake of the political transition, immediately evoked a critical contest around the nature of this human action committed during the context of a juridical void, in particular a contestation regarding the type of legal consequence which could be attached to such acts, a contest which, as Agamben argues, betrays the difficulty of defining the nature of human action in this juridical void and the ?impossibility? of clearly defining the legal consequences of these actions. This adjudication of the legal consequence of acts committed during the exception hinged critically in the immediate post-apartheid period, on an adjudication by the Amnesty Committee of the TRC regarding whether such acts could be shown to have been committed with ?political? intent and therefore would be eligible for amnesty, including the extinction of the legal consequences that that this entailed. Thus the Moleleki protagonists applied to the TRC for amnesty for their involvement in the execution in May 1997, four months before the state prosecutor recommended to the Attorney-General that 10 charges of murder and 10 charges of kidnapping should be brought against the seven accused in relation to the execution. This precipitated an initial suspension of the formal juridical processes around the execution. However, the ambiguity of this human praxis during a period of iustitium was repeatedly evoked during the course of 1997 in an ongoing contest between the formal judicial system and the amnesty processes of the TRC regarding whose object of adjudication the execution should properly be. Thus, SDU members involved in the execution were re-arrested soon after the investigation unit of the recently established TRC initiated an inquiry into the killing with the intention of persuading the protagonists to voluntarily come forward to the 168 Commission to apply for amnesty. The success of this initiative was primarily the result of the work of TRC investigator George Ndlozi, who had himself been an SDU member in Katlehong. Ndlozi explains: We were very successful in obtaining the information and in convincing perpetrators to apply for amnesty. We were very disturbed to learn that some of the people we spoke to got arrested a week after we spoke to them. This action hampered our investigation as it developed an atmosphere of animosity between us and the SDU in Moleleki. We spoke to some of them and they felt very bitter as they suspect that I might have handed the information to the police about their involvement. (Ndlozi, 1997, p. 4) In fact the re-arrest of the accused in the Moleleki execution had nothing to do with the TRC investigation. Siviwe Malcolms Ngam, accused number two in the subsequent trial, had appeared in the Alberton court in connection with a charge of theft. Captain du Plessis, who was briefly involved in the Moleleki investigation, alerted the current investigating officer, Inspector Baumann, to Ngam?s appearance. Baumann ?booked out? Ngam two nights later and he then reportedly identified five other accused. One of the other accused, Oscar Motlokwa, was in fact already in jail on a charge of murder in connection with an attack on an alleged IFP supporter on a train during 1993. At the time that the state prosecutor was listing witnesses for the state?s case, he was also aware of the pending applications of several of the accused to the TRC for amnesty and attempted to dispute the assertion of political intent on which the amnesty applications of the Moleleki accused hinged, in terms of the apparently transgressive nature of the actions of SDU members, ?It is my respectful submission that there is no foundation for these murders 169 as political?These murders were carried out in an extremely gruesome manner on the lives of young children and a woman?61 (The State v. Michael Sonti, 1997, p. 12). The TRC would ultimately concur with the state prosecutor regarding the nature of these actions as ?non-political?, not on the grounds that the actions had been transgressive but on the basis that SDU members could not articulate a juridical political motive as an instrumental expression of the objectives of a ?known political party?, namely the ANC. Instead they invoked the ambiguous biopolitical concept of the ?community? as the basis of the authority for their actions and the criminal rather than the political party opponent as the object of enmity. After the Truth and Reconciliation process, the execution was passed back to the criminal justice system for a final adjudication regarding its ?criminal? nature. However, in using the doctrine of common purpose for this adjudication, which abandoned the relation of causality usually required for a finding of criminal culpability, the multiple problems of undecidibility around the execution within the TRC, which concerned the attempt to juridically inscribe what had in fact occurred outside the law, were simply exacerbated and reiterated, producing a profoundly mutable ?truth?. 61 Dit is my respekvolle submissie dat hier geen ander grondslag vir hierdie Moorde is, as politiek nie?Hierdie moordes is op ?n uiters grusame wyse uitgevoer op die lewens van jong kinders en ?n vrou. (The woman referred to is Bulelwa Zwane) 170 CHAPTER FIVE: THE TRUTH AND RECONCILIATION COMMISSION, FIXING THE LIMITS OF THE POLITICAL Introduction The negotiated transition between the pre and post-apartheid periods was made possible by the essential continuity of the principle of juridical inscription, namely the capturing of ?pure? violence for law, which was instantiated in terms of the provision, that amnesty would be granted for ?political? offences committed by protagonists on all sides of the conflict. However, during this subsequent historical epoch, the processes of juridical inscription that surrounded the Moleleki execution, and other incidents of biopolitical violence that had occurred during the struggle between the apartheid state and its opponents, were critically concerned with the establishment of the sovereignty of the new state in terms of a new norm of law, the norm of ?human rights?, which sought to define the boundaries of the juridico- political present in terms of the ?exception? of the violence of the past. Paradoxically, however, despite the explicit intention to constitute the present ?political? as the juridico-political, in opposition to the biopolitics of the past, the norm of human rights, ?which is invoked today as an absolutely fundamental right in opposition to sovereign power? (Agamben, 1998), in fact introduces a critical biopolitical principle into the functioning of modern nation states. This apparent paradox becomes explicit if one examines the advent of the modern nation state during the ?great age of revolutions? in the late 18th century, inaugurated by the French Revolution of 1789, which for the first time made the ?people? the immediate bearer of sovereignty. Under the dynastic system, which had preceded this era of the nation state, sovereignty was embodied in the person of the divinely authorised sovereign. However, with the advent of the modern nation state a profound transformation takes place where birth marks not simply the emergence of the subject of a 171 sovereign (who can live anywhere under the dominion of a particular prince), but the emergence of a citizen of a nation state. The instrument of the transformation from subject to citizen is the Declaration of the Rights of Man, which ostensibly makes birth the ?source and bearer of rights? (Agamben, 1998, p. 127). Critically, however, what remains implicit is the fact that these ?inalienable rights? can only be exercised by a citizen of a nation state and are therefore contingent, not on a meta- juridical value but location within a nation state62. Moreover, the ?Rights of Man?, and the protections the declaration affords, also emerge not as a meta-juridical right, but as a biopolitical status, which is acquired within the context of the political community. An examination of the Declaration of Rights itself makes the ?essential fiction? underlying this formulation immediately explicit, as ?at the very moment in which native rights were declared to be inalienable and indefensible, the rights of man in general were divided into active rights and passive rights? (Agamben, 1998, p. 130), where all inhabitants of a society would enjoy ?passive rights?, but ?active rights? ? those rights ?by which society is formed? ? specifically excluded (among others) women, children, foreigners, the insane, minors and those condemned to punishment (Agamben, 1998, p. 130). These people, ?who would not at all contribute to the public establishment? and ?must have no active influence on public matters? would not be citizens (Agamben, 1998, pp. 130-131). Citizenship therefore, is not in fact a right, but a status (Chipkin, 2007, p. 55), which invokes the ?inscription of life in the 62 The hidden difference between birth (allegedly the immediate source of rights) and nation, was made explicit by what Arendt called, ?The Decline of the Nation-State and the End of the Rights of Man?, which produced the figure of the refugee, ?who should have embodied the rights of man par excellence? but in fact signalled the, ?concept?s radical crisis?(Agamben, 1998, p. 126). Confronted with those who had, ?lost all other qualities and specific relationships-except that they were still human? it was instead revealed that, ?in the system of the nation state, the so-called sacred and inalienable rights of man? lack[ed] every protection and reality at the moment in which they [could] no longer take the form of rights belonging to citizens of a state?(Agamben, 1998, p. 126). 172 state order? (Agamben, 1998, p. 129) and whose ?duties and obligations issue not from nature, but from the political community itself? (Chipkin, 2007, p. 55). Thus, in the shifting of sovereignty from the prince to the ?people? through the figure of the citizen of the nation state, a figure that is in fact produced in and through the political community, a fundamental ambiguity emerges, namely, how is this boundary between bare life and political life to be generated? The grounding of sovereignty in the ?people? therefore leads to a ?constant need to redefine the threshold in life that distinguishes and separates what is inside from what is outside? (Agamben, 1998, p. 131) and thus to define the limits of the political community in terms of a distinction between ?authentic life and life lacking every political value? (Agamben, 1998, p. 132), an ?authenticity?, which as a status to be earned rather than a birthright, ?one had to prove oneself worthy and which could therefore always be called into question? (Agamben, 1998, p. 132)63. Thus in the context of the nation state, the determination of who is and isn?t an ?authentic citizen? begins to ?coincide with the highest political task? (Agamben, 1998, p. 130). It was this ?highest political task? that the South African TRC had to grapple with, namely, ?to overcome the worry that the South African people did not actually exist? (Chipkin, 2007, p. 174) and through this delimitation of the ?people? in whom sovereignty would be vested purportedly establish a post-apartheid nationhood, the ?National Unity? on which the founding act of the Commission was premised. 63 Thus for example, ?the mass denaturalization of populations? (Agamben, 1998, p. 132) from the First World War onwards, where, for example, the citizenship of naturalized citizens of the ?enemy? was revoked in France and Belgium, and in Italy where citizenship was revoked from those who had shown themselves to be ?unworthy of Italian citizenship? (Agamben, 1998, p. 132) and finally in Nazi Germany, where Jews were systematically ?denationalized? before being sent to extermination camps as part of the ?Final Solution? (Agamben, 1998, p. 132). 173 In the immediate post-apartheid context, the principle of this constitution was the exception of the past: ?The South African people lacked national marks. It was only really clear who they were not. They were not the South Africans of old: those who had perpetrated and endured the injustices of the past? (Chipkin, 2007, p. 174). Thus the quality of identity that South Africans shared was a common history, more specifically the biopolitics of the past which became the marker of a new type of authenticity, where ?being authentically South African comes to mean sharing the traumas of apartheid and uniting in the subsequent process of ?healing the nation?? (Wilson, 2001, p. 14). It was in establishing an instrumental relation with the violence of the past that the juridico-political present based on the national sovereignty of a post-apartheid ?people? could purportedly be instantiated. Critically, however, the instrumental relation envisaged by the TRC as the means of institutionalising the exception of the violent past as a basis of a common identity and nationhood, in fact disguised the complexity of the relation of exception, through which this relation with the past was in fact enacted. Thus, it was assumed that that the transfer of sovereignty to the ?people? through the country?s first democratic elections was a purely juridical phenomenon, through which all residents of the territory of South Africa would be accorded citizenship rights. This would ostensibly establish a juridical ?political? in opposition to the biopolitics of the past. However, what remained implicit was the biopolitical nature of the relation of exception that would constitute this present ?political? through the biopolitical fracture, in terms of which, ?the human species [is constituted] into a body politic?, through a fundamental split between naked life (people), and political existence (People), exclusion and inclusion, zo? and bios...hence the contradictions and aporias that such a concept [the people] creates every time that it is invoked and brought into play on the political stage. It is what always already is, as well as what has yet to be realized; it is the pure source of identity and yet it has to redefine and purify itself continuously according to exclusion, language, blood and territory. (Agamben, 2000, p. 30. 1-2) 174 At the heart of the TRC?s work therefore was this profound ambiguity, the ambiguity of the relation of exception through which the People could be constituted and in whom national sovereignty could be vested. The TRC, Making an Exception of the Past The TRC was legislatively established in 1995 in terms of the Promotion of National Unity and Reconciliation Act to investigate ?gross violations of human rights? which had taken place in the country between 1 March 1960 and 10 May 1994 and to grant indemnity to those who had committed violent acts associated with a political objective during this period. While it was widely recognised by its architects, critics and advocates alike as a project intended to establish a new basis for the sovereignty of the post-apartheid state, what remained implicit was the ambiguity of the relation of exception that would inaugurate this new sovereignty. A significant body of literature critiqued the ?script of nationhood? (Posel & Simpson, 2002a, p. 9) on which the work of the TRC was so explicitly premised, either criticizing this aspiration toward ?national unity and reconciliation? as a self-serving project of legitimation by the post-apartheid state (Buur, 2002; Wilson, 2001) or noting the various consequences of this nation-building imperative in terms of the limitations it imposed on a range of aspects of the Commission?s ?truth-seeking? work (Kistner, 2003). Various authors observed the ?privileging? (Simpson, 2002, p. 221) of an imperative of ?nation-building? in the work of the TRC, whether in terms of its report, which Posel, for example, argued was less an analytical historiography than a moral narrative, which could make possible the ?imagining? of a new form of national community based on a ?collective memory?, a ?shared history? (Posel, 2002, p. 149); or in terms of the way in which the Commission?s report flattened and obfuscated the complexities of localized conflicts through its emphasis on national political 175 party conflict and its definition of the ?political? in these narrow juridical terms (Bonner & Nieftagodien, 2002); or in terms of the way in which narrative testimony within the Commission was decisively linked to notions of restorative justice, within a wider political and ideological project of ?nation building?, which sidelined ?those truths that could not easily be harnessed to Reconciliation? (Kistner, 2003, p. 11) leading to the ?uncoupling of Truth from Justice?, in order to inscribe it within ?a relation to Reconciliation? (Kirstner, 2003, p.13 ) that de-legitimated calls for retributive justice in the name of ?nation-building?; or in terms of the internal functioning of the Truth Commission, which through the positivist codification of data in its information management system sought to construct a ?new national history? (Buur, 2002, p. 78) that could ?serve as the foundation for a new nationhood? (Buur, 2002, p. 76). Similarly, in a paper entitled National Narrative Versus Local Truths, Van der Merwe noted, in relation to the TRC?s public hearings in the East Rand township of Duduza, the subjugation of competing local ?truths? about the conflict in the township to a ?broader agenda of constructing a national narrative? (Van der Merwe, 2002, p. 216) and the negative impact that this had on the possibility of effecting localized forms of ?reconciliation?, which would have required a more sustained and nuanced engagement with the conflict in the township. Belinda Bozzoli also observed the tensions between a ?nationalist narrative? and the fractured accounts of local community members at the hearings of the human rights violations committee in Alexandra, which, she argued, significantly obfuscated the critical role of the youth in this township: It appears that two different narratives of the events covered by this particular hearing have emerged. First there was the public, nationalist and official ?community? 176 narrative, as already existing in the public sphere, articulated by Obed Bapela and reinforced by other witnesses speaking on ?behalf of the community?. Second, there were the private narratives of the individual witnesses speaking mainly on behalf of themselves and their families. There are several overlaps, but also sharp disjunctures between the two. (Bozzoli, 1998, p. 11) Wilson, who sub-titled his book on the Truth Commission Legitimizing the Post-Apartheid State, saw the South African Truth Commission in a relation of instrumental legitimation to the post-apartheid state, which ?dragooned? human rights ?talk? ?into the service of nation building? and the ?re-imagining? of the nation by constructing new official histories and ?manufacturing legitimacy for key state institutions?, in particular the criminal justice system (Wilson, 2001, p. xvi). Quite correctly identifying the historical connections between human rights and the development of modern nationalism, Wilson however sees these processes in terms of a straightforward project of ideological legitimation, where ?human rights talk? is used by new nation states to bolster their authority and sovereignty by treating the concepts of human rights as ?a full-blown political and ethical philosophy? rather than ?as narrow legal instruments which protect frail individuals from powerful state and societal institutions? (Wilson, 2001, p. 224). The ?appropriation of human rights by nation building discourse? meant that human rights became associated with ?forgiveness, reconciliation and restorative justice?, which deemed ?social stability to be a higher good than the individual right to retributive justice? (Wilson, 2001, p. 26) and ignored much more complex local understandings of justice, which incorporated both retribution, restoration and revenge, as well as an increasing turn in the international global context towards the demand for individual legal accountability. However, what these analyses did not reveal was the fundamental ambiguity on which the entire Truth Commission project hinged. It did not involve a straightforward process of legitimation of something that did not in fact exist a priori, namely the nation state. Instead it 177 concerned a complex process of delineation in terms of the relation of exception, of the people and the People, on which the nationhood of the post-apartheid state could ostensibly be founded. Amnesty, Effecting a Relation of Exception This relation of exception was articulated most critically in a ?Postamble? appended to the Interim Constitution in terms of which the country?s first popular elections would be held and the country governed for a period of five years. This Postamble to the Interim Constitution provided that amnesty would be granted in respect of acts ?associated with a political objective?. The Postamble constituted the legal basis for the TRC and was subsequently included in the Preamble to the Promotion of National Unity and Reconciliation Act No 34 of 1995 (TRC, 1998b, p. 48), which would establish the TRC as an institutional body. By the end of the negotiations process, which led to the finalisation of the Interim Constitution in November 1993, one issue remained outstanding, namely the question of amnesty for those who had committed acts of violence in pursuit of political objectives during the course of the struggle of the previous decades: A compromise was eventually reached only after the finalisation of the rest of the Interim Constitution and was recorded in what became known as the ?Postamble?. This provided that there would be amnesty for politically-motivated offences and that future legislation would provide the criteria and procedures to regulate the process. (TRC, 1998b, p. 52) In the course of the negotiation process that preceded this compromise the state had promulgated the Further Indemnity Act in November 1992 which gave the State President significant powers and discretion, without public process, to grant amnesty to individuals who had committed acts ?with political intent? and whose release might ?promote reconciliation and peaceful solutions? (Van der Merwe, 1999, p. 3). 178 While the secrecy of the amnesty process provided for by this Act was widely contested at the time, it was the process for the granting of amnesty, the suppression of the ?truth? about the crimes of the past in the secrecy of these processes, that was the source of contention, rather than the principle of amnesty itself. The entire negotiation process remained significantly contingent on making an exception of the past of violence, through the granting of amnesty for acts committed during the course of ?past conflicts?. Critically, in the ongoing national debate about the question of amnesty for political offences, as the outgoing state attempted to secure indemnity for its own operatives, the ANC did not reject the possibility of amnesty for operatives of the apartheid state involved in the commission of violence but argued for a qualified amnesty, an amnesty which would rest on the demonstration of an instrumental relationship between this violence and the political of the past, an instrumental relationship to which the TRC would later attempt to give institutional effect. While the question of amnesty for political offences was widely framed as simply a pragmatic issue, a question of ?realpolitik? which needed to be resolved before elections could go ahead, resulting in a ?last minute compromise? concluded behind closed doors by a small coterie of leaders who hammered out a solution to this outstanding issue (Boraine, 2000; Graybill, 2002; Hayner, 2001; Sarkin, 2004; Simpson & Van Zyl, 1995; Tutu, 1999; Wilson, 2001), in fact the Postamble to the Interim Constitution was a far more significant document, which was to establish the terms of exception on which the post-apartheid state was founded by framing the entire Interim Constitution as a ?bridge? between past and future, a ?bridge? which would establish a conjuncture between the biopolitics of the past and the present political. This was a conjuncture that would make possible a ?transcendence? of 179 this past, as an exception that would redefine the ?political? as the juridico-political and establish the basis on which the sovereignty of the new state would be founded. This relation of exception which would constitute the juridico-political of the present in terms of the exception of the biopolitics of the past, was a relation of inclusive exclusion, in which what had been excluded, the gross human rights violations of the past, would in fact remain ever present, producing ?enough truth to demonstrate and exemplify the inequities of the past? (Posel & Simpson, 2002b, p. 151) and ?prevent a repetition of such acts in the future?. Critically, however, the complexity of the relation of exception and the ambiguity of the boundaries that would found the new juridico-political order, remained implicit throughout the negotiation process and in its later institutionalisation in the TRC. It was assumed that the capturing of violence as an exception concerned the tracing of an instrumental relation between a prior origin of power as a thing which could be usurped in its totality and the effects of power (violence) that are reduced to a ?transparent instrument? (Feldman, 1991, p. 3). This would simply require instrumental explanation and rationalisation to trace the relation between cause (power) and effect (violence). By adjudicating this relationship between the commission of violence and a prior political motive as an act associated with a political objective, it was assumed that this relation could be administratively sundered and an exception rendered. The Juridical Classification of the ?Political? ? An Enigma Therefore the process of indemnification which would make possible the commencement of negotiations hinged on a classification of the political that would legislatively establish an 180 instrumental relation between the juridico-political of the past and the perpetration of violence in pursuit of this political, a legislative relation that attempted the constitution of strategic violence as the object, the tool of this ?past? politics. In defining this relation, the process of making an exception of this past violence could putatively be initiated. The first juridico-political act in this process of constituting the exception that would come to define the South African political was the Groote Schuur Minute, an agreement concluded in May 1990 between the then apartheid state and its political antagonists, the ANC, which ?contained and recorded the understandings resulting from the early negotiations between the then government and the ANC? (Sarkin, 2004, p. 37). With this agreement, as the 1998 TRC report explains, the negotiations process began seriously? In terms of the Minute, a working group was established to make recommendations, amongst other things, on a definition of political offences in the South African situation, and to advise on norms and mechanisms to deal with the release of political prisoners. (TRC, 1998b, p. 50, own emphasis) However, the group found that globally such a classification of ?the political? had not yet been attempted and that ?there was no generally accepted definition of a political offence or political prisoners in international law? (TRC, 1998b, p. 50, own emphasis). In this international legislative vacuum the working group turned instead to ?principles of extradition law?to develop guidelines? (TRC, 1998b, p. 50). The Political Offence Exception In turning to international extradition law, the working group were in fact turning to the only formal international legislative attempt to give juridical form and legal consequence to the terrain of human action which constitutes ?the political?. However, these efforts to give legal form to ?the political?, have to date, remained apparently irresolvable. ?There is?no 181 universally recognised definition of political offence in extradition law. Any common crime can be deemed political when it subjectively or objectively affects the existing socio-political order? (Van den Wijngaert cited in Bhargava, 2002, p. 1329). The history of extradition is deeply bound with the protection of state sovereignty. The earliest extradition treaties date back to the 13th century and were concluded between the Egyptian, Chinese, Chaldean and Assyro-Babylonian states (Bassiouni, 1974). Critically, these extradition treaties sought to ensure the return and punishment of citizens of states who had purportedly committed crimes of lese majeste, crimes against the political authority of the monarch or sovereign (Perry, 1988). In the principle of extradition between one sovereign state and another, an exception to the putatively inviolable territorial sovereignty of states was effected which sought to establish a relation between the political and juridical, allowing states to request the juridical extradition of ?political offenders? from the sovereign of the territories to which they had fled. During the 19th century, in the wake of the French revolution and the rise of the concept of inviolable human rights which citizens could invoke against the state, a doctrine of political offence exception developed, which sought to define conditions under which a state could refuse extradition to another state. At this historical conjuncture, the doctrine of political offence exception sought primarily to extend the protection of ?human rights? to the citizens of monarchical states who had committed ?political? offences in the course of struggles to overthrow or change these regimes and had who had subsequently sought asylum in other sovereign states. 182 Critically, the doctrine of political offence exception hinged on a classification of ?the political? that failed to yield a definitive conclusion in more than a century of subsequent international legislative endeavour in the courts of states around the world and in international adjudicative tribunals, leading legal theorists to conclude of the political offence exception that ?it would be Utopian to expect that as complex a problem as this one can be solved on a purely juridical basis? (Van den Wijngaert, 1980, p. 229). The ?complexity? of the problem of political exception, however, exposed, in the multiplication and contradiction of criteria for the legal adjudication of ?the political?, the essential impossibility of inscribing within the juridical what was in fact meta-juridical, a type of human action which occurred fundamentally, as Agamben argues, in a zone of exception on the boundaries of the juridico-political order. The ambiguity evoked by the contradictions of the various laws that sought to classify the political in terms of the political offence exception, were thus not in fact about a lacuna in law, as most legal theorists assumed, but about a boundary of law, it?s very relation to ?life?. Paradoxically, this exception and the ambiguities of this exception were essential to the articulation between anomie and law, between life and law, an articulation which, Agamben argues, is made possible by this very threshold of irrevocable undecidibility. What was at issue in this decision on the ?exception? was, as Schmitt argues, the very condition of possibility of juridical rule. South Africa at the end of the 20th century, with the belated inauguration of a state based on the doctrine of rights in the wake of the decline of the authoritarian apartheid state, ironically used a classification of political exception that had been born at the advent of 19th century 183 European democracy, and which, as it was instantiated in the South African juridical order, was being done away with in the international juridical contexts in which it had initially been given institutional effect64. The Norgaard principles The principles of extradition law which constituted the South African ?political? in terms of the principle of exception, through the doctrine of political offence exception, were grounded in the classification of the political utilised in the United Nations-supervised elections in Namibia in 1989, which became known as the ?Norgaard Principles?. These were ?formulated under the guidance of Professor CA Norgaard, the former President of the European Commission on Human Rights, and [were] applied to guide the process of identifying Namibian political prisoners for release? (TRC, 2003, p. 8). The TRC report explains: The Norgaard Principles were gleaned from a survey of the approaches followed by various state courts in dealing with what is known as the ?political offence exception? in extradition proceedings. In terms of the ?exception?, a state that has been requested to extradite an individual may refuse to do so where the crime for which the extradition is sought is political. It was thus necessary for states to formulate an approach to the question of whether a particular crime amounted to a political offence. The background principles, therefore, recorded the common features of the various states? approaches to the issue. (TRC, 2003, pp. 8-9) 64 The international abandonment of the principle of political offence exception, in the wake of the September 11 2001 attacks on the New York World Trade Centre, represents the collapse of this fictive lacuna, which previously articulated law and anomie in a relation of exception, and evokes in this collapse a permanent state of exception where rule and exception have blurred without articulation. In June 2002 a meeting of G8 Foreign Ministers recommended that in order to combat terrorism countries should, ?exclude or reduce to the greatest possible extent any application of the political offence exception in responding to a request for mutual legal assistance concerning terrorist offences?. (G8 Recommendations on Counter-Terrorism. Retrieved from: http://www.icclr.law.ubc.ca/Site%20Map/compendium/Compendium/Declarations/G8%20Recommendations% 20on%20Counter%20Terrorism%202002.pdf) 184 The Norgaard principles were codified in Section 20 (3) of the TRC Act. The Act suggests that the Amnesty Committee take the following factors into account in determining what constitutes a political crime: ? the motive of the person who committed the act; ? the context in which the act took place; ? the legal and factual nature of the act, including the gravity of the act; ? the object or objective of the act, and in particular whether the act was primarily directed at a political opponent or against private property or individuals; ? whether the act was committed in execution of an order of, or on behalf of, or with the approval of a political organization or the state; ? the relationship between the act and the political objective pursued, and in particular the directness and proximity of the relationship and the proportionality of the act to the objective pursued. (cited in Bhargarva, 2002, p. 1312) Importantly, the definition of ?the political? which Professor Norgaard had developed in the context of the Namibian elections sought to incorporate a number of different principles which had guided various states in giving juridical effect to the principle of political exception. However these efforts to include the full spectrum of ?principles? developed in various legislative contexts around the globe ultimately exacerbated and reiterated the essential undecidibility of this classification. The multiplication of criteria which the Norgaard principles incorporated were not the result of a new comprehensiveness of legal description but reflected instead the struggle to contain in judicial prescription what was in fact on the boundaries of law. This was reflected in the South African context in the ambiguities of the indemnities granted to political offenders in the period preceding South Africa?s first popular elections and most significantly in the amnesty processes of the TRC, whose efforts to grant amnesty to political offenders in terms of the principles of political offence exception developed by Professor Norgaard, revealed, despite significant efforts at judicial investigation and adjudication 185 within the processes of the Amnesty Committee, a multitude of inconsistencies and contradictions of decision (Bhargava, 2002; Foster, Haupt & De Beer, 2005; Sarkin, 2004; Simpson & Van Zyl, 1995). The various principles which Professor Norgaard sought to include in his definition of the political incorporated three central ?tests? which had been utilised in various international legislative contexts to adjudicate the occurrence of ?related? political offences. Importantly, related political offences are internationally legally classified as distinct from purely political offences, which are (i) exclusively directed against the state or a political organization and do not injure private persons, property or interests and (ii) ?are not...accompanied by the commission of common crimes?...Related political offences are common crimes which are considered political offences because a political purpose was being pursued, or because the act had political consequences or was situated in a political context. (Bhargava, 2002, p. 1329) The three tests which have been developed internationally to attempt to determine the occurrence of ?related? political offences in terms of extradition law are the political incidence test, the predominant motive test and the political objective test (Bhargava, 2002, p. 1329). Professor Norgaard, in his classification of ?the political?, in the Namibian context, sought to incorporate aspects of all of these tests. Thus, The political incidence test requires both that there be a political disturbance and that the act is part of and in furtherance of a political struggle. The predominant motive test requires that the crime be connected to a political objective, that the political nature of the crime outweighs the criminal aspect, and that the means used are either the only recourse available or proportionate to the desired political end. The political objective test only examines the specific nature of the act, without regard to the subjective motivation of the individual or whether the desired ends were political. (Bhargava, 2002, p. 1329) The Norgaard principles were initially introduced into South Africa at the inauguration of the negotiation process with the proposals of the working group set up in terms of the Groote 186 Schuur Minute, which recommended that in making decisions on the release of political prisoners ?an adaptation of the Norgaard Principles be used? (TRC, 1998b, p. 51). Critically, the adaptation of the Norgaard Principles which the working group recommended be taken into account in this endeavour, reflected a wide range of criteria, implicitly informed by all three ?tests? referred to above, including: aspects such as motive, context, the nature of the political objective, the legal and factual nature of the offence?the object of the offence and whether the act was committed in the execution of an order and with the approval of the organisation concerned. (TRC, 1998b, p. 51) This multiplicity of criteria were incorporated into South African legislation through the passage of the Indemnity Act of May 1990, which in conjunction with the November 1990 Guidelines for Defining Political Offences in South Africa, enabled individuals who committed political offences before 8 October 1990 to receive temporary immunity or permanent indemnity (Sarkin, 2004, p. 39, own emphasis). Thus the exception of ?political offence? on which the negotiation process ultimately turned was founded on a legislative classification of ?the political?, which in its multiplicity of criteria revealed the irrevocable ambiguity of this political, and the consequent ambiguity of the relation of exception which the negotiation process was attempting to effect. Nevertheless this relation of exception was in fact critical to the negotiation process, which by suspending the application of the law to violence committed in instrumental relation to the ?political? in the form of indemnity for ?political offences?, sought to make a process of transition from past to present possible. 187 The TRC, Establishing an Instrumental Relationship with Violence Committed in Pursuit of Political Objectives The Act establishing the TRC stated that its objectives were to ?promote national unity and reconciliation in a spirit of understanding which transcends the conflicts and divisions of the past?, through two key objectives: (a) establishing as complete a picture as possible of the causes, nature and extent of the gross violations of human rights which were committed during the period from I March 1960 to the cut-off date, including the antecedents, circumstances, factors and context of such violations, as well as the perspectives of the victims and the motives and perspectives of the persons responsible for the commission of the violations, by conducting investigations and holding hearings; (b) facilitating the granting of amnesty to persons who make full disclosure of all the relevant facts relating to acts associated with a political objective and comply with the requirements of this Act. (TRC, 1998b, p. 55) In order to effect these objectives, three committees were established, the Human Rights Violations Committee, the Amnesty Committee and the Committee on Reparation and Rehabilitation. Over its lifetime, the Human Rights Violation Committee would take approximately 21 298 statements from ?victims? of ?gross human rights violations? and hold 76 public hearings where a selection of these ?victims? would recount their stories of violation (Ross, 2003, p. 13). The Amnesty Committee was established as a quasi-judicial institution to adjudicate the applications of those seeking amnesty for offences associated with a political objective. It was to eventually receive over 7000 applications for amnesty, only 1 674 of which were to be successful (Foster, Haupt & De Beer, 2005, p. 12). The Committee on Reparation and Rehabilitation was tasked with the responsibility to, ?provide?measures to be taken in order to grant reparation to victims of gross human rights violations? (TRC, 1998b, p. 171). Critically, in its efforts to promote ?National Unity and Reconciliation? the TRC would not investigate apartheid as a structural phenomenon or a system of power, but would instead 188 establish a very specific type of relation with the past of ?strife, conflict, untold suffering and injustice? (TRC, 1998b, p. 48), namely an instrumental relation with the biopolitical effects of power, the violence committed in pursuit of political objectives during the conflicts of the past. This particular relation with the biopolitical effects of power implicitly informed the entire negotiation process around the question of amnesty prior to the country?s first democratic elections, that is, that some form of indemnity would be necessary for those who had committed acts of violence in pursuit of political objectives. It was merely the terms of this relation, the way in which it would be adjudicated and who had the authority to undertake this adjudication, that were in contention, leading eventually to the prescription in the Postamble to the Interim Constitution that amnesty would be granted for acts associated with a political objective. While the TRC was widely critiqued (Bonner & Nieftagodien, 2002; Mamdani, 1998; Posel, 2002; Posel & Simpson, 2002a; Wilson, 2001) for its narrow focus on the violent effects of apartheid, rather than apartheid itself as a system of structural injustice, and was understood by the TRC itself as the consequence of the ?limitations? of its mandate, which it sought to rectify through a series of institutional hearings, this particular relation with the violent effects of power was in fact critical for establishing the terms of the post-apartheid juridico- political by attempting to capture, through the relation of exception, violence as the object, the ?thing? of ?politics?. The TRC was therefore criticised for establishing an ?equivalence? between the violence committed by the agents of the former state and violence committed by those who resisted the system of apartheid, as a consequence of its focus merely on the effects of violence, which, it was argued, failed to locate violations in historical and political context (Mamdani, 189 1998, Wilson, 2001 ). However this principle of ?equivalence? had in fact been established during the negotiation process, where it was agreed that indemnity would be granted to those who committed acts in pursuit of political objectives on both sides of the political divide. This was not, as commonly assumed, merely the consequence of an expedient ?balancing act? required to negotiate South Africa?s ?delicate? transition, but implied a far more fundamental principle, namely the capturing of violence, perpetrated in whatever context, as the object of the juridico-political, a capturing through the relation of exception, which would maintain law?s relation with anomie. Exposing the Relation Between Force and Law It was ?truth?, the ?uncovering? of gross violations of human rights and the ?full disclosure? required for amnesty, that would mediate this relation of exception and establish a new norm of law in a context where under apartheid the ?rule? and norm of law had been so fundamentally severed from the ?force of law?. The TRC report?s entire mandate chapter is framed by an explicit concern to expose this relation between force and law. It states: ?special attention will be given to the role and contribution of two phenomena or factors in the shaping of this country?s history, namely violence and law, and the relationship between them? (TRC, 1998b, p. 24). Critically, however, as Agamben argues, the idea of a ?relation? between violence (the force of law) and law is essentially a ?fictive categorisation?, which attempts to encompass law?s own absence. Nevertheless this articulation, through the relation of exception between law and anomie, however fragile and ambiguous, is in fact critical to the functioning of the entire juridico-political ?machine?. For law this empty space of the exception is its constitutive dimension (Agamben, 2005, p. 60). It is when the juridico- political rule, which exists only in relation to the exception, for ?the exception does not only confirm the rule; the rule as such lives off the exception alone? (Schmitt cited in Agamben, 190 1998, p. 16), is conflated with the exception in the suspension of juridico-political norms, that you have the type of legal violence which so characterised the apartheid state and which the new state was so concerned to disarticulate65. The TRC report, referring both to the ?violence of the gun? and the ?violence of the law?, stated: ?Violence has been the single most determining factor in South African political history? (TRC, 1998b, p. 40). In exposing the nature of the relation between law and violence, this relation could putatively be sundered, rendering a juridico-political present freed from the stain of violence, Under apartheid, law which should have been a pillar of justice and social stability and a wall against violence and chaos, became instead an agent of injustice and social instability and chaos. Law was systematically reduced?to a continuation of violence through other means. (Asmal, Asmal & Suresh, 1996, p. 75) The TRC through its processes of truth revelation would thus ?lay the foundation for the reestablishment of the rule of law? (Bhargava, 2002, p. 1309) on which the sovereignty of the new state based on ?a monopoly over the legitimate use of force? (TRC, 1998b, p. 88), could be re-established. Adjudicating an instrumental relation with the biopolitical effects of power Critically, however, because the TRC could only understand this relation between law and violence in instrumental terms, what remained hidden was the complexity of the relation of 65 As Benjamin wrote, ?the tradition of the oppressed teaches us that the ?state of exception? in which we live is the rule? (cited in Agamben, 2005, p. 57). It is this conflation between exception and rule which forms the principle of contiguity between the apartheid state under the conditions of the state of emergency and the Nazi regime, both of whom suspended but did not abolish the personal liberties contained in their constitutions, a suspension of juridical political norms, in which parliament became like law, merely an instrument of force, ?the normative aspect of law?thus obliterated and contradicted with impunity by a governmental violence that, while ignoring international law externally and producing a permanent state of exception internally-nevertheless still claims to be applying the law?(Agamben, 1998, p. 132). Thus, ?In this sense, modern totalitarianism can be defined as the establishment, by means of the state of exception, of a legal civil war that allows for the physical elimination not only of political adversaries but of entire categories of citizens who for some reason cannot be integrated into the political system?(Agamben, 2005, p. 2). 191 exception which actually underpinned the processes of the TRC. Rather than the linear relation between law and violence that the TRC sought to instantiate and arbitrate, this was an articulation of profound ambiguity at the limits of the juridico-political order. Nevertheless, it was the sovereign decision on these boundaries of the juridico-political order that the TRC was established to articulate, that would in fact found the sovereignty of the new state. The TRC was thus critically constituted around an attempt to administratively adjudicate the juridico-political in a relation of exception to the biopolitics of the past, defined as the violation of ?bodily integrity rights? (TRC, 1998b, p. 64) perpetrated through the ?gross human rights violations? of ?killing, abduction, torture or severe ill-treatment? committed in pursuit of ?political objectives? during the conflicts of the past (TRC, 1998b, p. 60). As a result, At HRV66 hearings cross-examination ascertained whether the person testifying was a victim of a politically motivated crime and therefore could be designated a ?victim? with the rights to receive reparations. At amnesty hearings, practically the entire thrust of cross examination?was directed towards deciding whether there was a political objective to the violation. (Wilson, 2001, p. 81) The Ambiguity of the Political However, the nature of this ?political? was to remain profoundly elusive throughout the processes of the TRC. As the Commission acknowledged in its report of 1998, ?the political nature of specific acts was hard to define? (TRC, 1998b, p. 82) . The Commission understood the problems it experienced in its adjudication and classification of the political as a pragmatic quandary generated by the ?complexity? of the conflicts of the past and the structural nature of apartheid, which ?affected every sphere of society? (TRC, 1998b, p. 82), a pragmatic question, which, it argued at the end of the TRC process, could most 66 Human Rights Violations Committee. 192 satisfactorily be resolved through a broader classification that would ?extend the ambit of an ?an act associated with a political objective?? (TRC, 2003, p. 91). The Amnesty Committee attempted to negotiate the ?complexity? of the ?political? through a process of quasi-judicial arbitration and hearings, which sought to establish on a balance of probabilities whether a violent act had been committed on behalf of or with the approval of a ?publicly known political organisation? and was directed against ?publicly known political opponents?. In the Human Rights Violation Committee this adjudication of the political took place through a process of codification and forensic verification called the ?information management system?, whose goal was to make it possible for the Commission to make ?findings? on the ?political? nature of the violations reported to it (Buur, 2002; Wilson, 2001). Operating in terms of a juridical understanding of power, where the origin and the violent effects of power are implacably separate, the TRC sought to instrumentally link and adjudicate the ?politicality? of the violent act, by linking it to a ?political objective?, which substantively relied on an adjudication of the ?political motive? of the subject, an interrogation of subjecthood that could putatively offer a definitive reading of the ?political? nature of the violent act. While some strategic acts of violence did indeed render themselves intelligible in these terms, the political motivation of the protagonists easily articulated in terms of a prior motivation to commit a violent act for strategic ends, there were many instances where in fact the authorial position of the agent in relation to the act was far more tenuous, institutionalized ideology and political action not necessarily maintaining a linear relationship which could be easily 193 demonstrated and arbitrated. The violence of the 1990s, which occurred substantially in non- relation to law, exposed the theoretical aporia of this instrumental conception of violence, which could not conceive of the biopolitical nature of the power it was in fact tasked to adjudicate. As a consequence, while over half (5 695) of the total of 9 043 statements on killings received by the TRC?s Human Rights Violations Committee concerned incidents which took place between 1990 and 1994, the TRC?s report conceded that ?the Commission failed to make significant breakthroughs in relation to violence in the 1990s? (TRC, 1998b, p. 206). A ?political motive?, an adjudication of subjecthood The locus of this mediation and the transition from biopolitical past to juridico-political present, was the ?bodiless reality? (Foucault, 1977, p. 17) of the juridical subject and in particular, an adjudication of ?the extent to which the subject?s will was involved in the crime? (p.17). It was these ?shadows lurking behind the case? (p. 17), namely the political motive with which the act of violation was committed that would change the nature of the act itself, ?the common crime?, that became, through an attribution of political motive, a ?political offence?. The Postamble to the Interim Constitution included this principle in its injunction that amnesty would be granted for acts associated with a political objective. The legislation that founded the TRC involved a second move, which translated such acts into gross human rights violations. In terms of the definition of gross human rights violations, ?political motive? became the locus of the adjudication of the ?politicality? of the act, and the authenticity of the subject, both ?victim? and ?perpetrator?, as the intention of the protagonist of violence not only established the authenticity of this subject as a political agent, but in a dialectical move, 194 simultaneously created the authentic ?victim? on the premise that ?it is the intention and action of the perpetrator that creates the conditions of being a victim? (TRC, 1998b, p. 59). This attempt to adjudicate the ?political? in terms of the subjecthood of its protagonists immediately ran aground in terms of multiple conceptual aporiae. The notion of ?political motive? brought to bear the conception of a sovereign individual in an instrumental relation of agency to action. This conception of the sovereign self in an instrumental relation to the ?political? is premised, as Feldman argues, on a mythical separation between the origin and site of enactment of power, conceived in terms of an individual?s prior intention, which is linked directly to action as a manifestation of this ?intent?. This conception masked the biopolitical nature of the fracture that the TRC was in fact established to articulate, namely the boundary between bare life and political life, zo? and bios, which did not concern an instrumental relation, but a relation of exception that would effect a biopolitical transformation of life in life. However, the TRC assumed that the adjudication of the ?political? concerned an attribute or additional capacity ? the political motive, which it struggled so much to systematically arbitrate and which subjects themselves battled to articulate as a ?declared political motive? (Foster, Haupt & De Beer, 2005, p. 5). This search for an articulated political motive as a way in which to judge the ?political? nature of conflict, completely avoided ?the complexities of social causation, where individuals were caught up in structural processes that motivate and constrain their actions, in ways that may not be intelligible to the actors themselves? (Posel & Simpson, 2002a, p. 10). Consequently, of the total of 7 115 applications received by the Amnesty Committee, 3 559 cases were rejected administratively because the Amnesty Committee found no evidence of political motive. Of the 1 626 amnesty cases that took the form of public hearings (Foster, 195 Haupt & De Beer, 2005, p. 24) and within the process of ?verification? of the politicality of victim statements, multiple problems of decidability emerged, what one group of analysts called ?grey areas? (Foster, Haupt & De Beer, 2005, p. 4). Thus many analysts were able to describe these ambiguities, noting, for example, ?the considerable difficulty? presented to the Commission by the challenge of defining a ?political crime? (Bhargava, 2002, p. 1304). However, because they shared the TRC?s juridical conception of ?the political? they were unable to fundamentally elucidate these ambiguities simply stating, for example, that ?little purpose is served here in trying to resolve these issues? (Foster, Haupt, & De Beer, 2005, p. 6) and concluding therefore that the Commission ?failed to provide a viable definition of political crimes? (Bhargava, 2002, p. 1305). These ambiguities and ?inconsistencies? critically concerned the various depletions of subjecthood that were required to instantiate an instrumental relation between a prior ?political? motive and a violent effect, that excluded acts committed for ?personal gain, personal malice, ill-will or spite towards the victim? (TRC, 2003, p. 18). This required subjects to demonstrate a singular motive and intent in instrumental relation to the multiplicity of criteria used by the TRC to define ?the political?. The complexity of the process was further exacerbated by the ambiguity and fluidity of the interpretation of various criteria used to adjudicate ?the political? within the operation of the Amnesty Committee. The adjudications of politicality in the Human Rights Violation Committee were similarly premised on the demonstration of an instrumental relation between act and author, paradigmatically illustrated by the expression, ?who did what to whom?, which referred to the implementation of a positivist data processing methodology developed by the American Association for the Advancement of Science. In order to be able to demonstrate a direct linear 196 relationship between author and act, this approach codified data in terms of a ?controlled vocabulary?, which delineated 48 distinct acts and three categories of persons ? victim, perpetrators and witnesses (Wilson, 2001, p. 40). The work of the data processors was to break down the narratives of the violated into these separate acts and trace their relationship to the three types of subject in order to make it possible to create ?neutral? signs of violation that could be quantified. In the Amnesty Committee this instrumental relation between author and act was mediated by an assumption of direct contiguity between formal political rationalities and the violent act. South African understandings of violence, as Andre du Toit (1993) points out, were significantly premised on the assumption of such relations of instrumental continuity between formal political rationalities and violent acts, which guided the ANC?s understanding of political struggle, in terms of a ?classically liberal, indeed Lockean, politics that ultimately justified violence ? revolution even ? in the name of bourgeois modernist rights? (Crais, 2002, p. 140). These understandings were incorporated into the legislation and operation of the Truth Commission, which specifically premised the adjudication of the political in the Amnesty Committee on both the membership of a ?known political organisation? of the amnesty applicant and the arbitration of whether this subject had acted in terms of the policies and practices of these organisations. Numerous analysts pointed out the inconsistencies and depletions that resulted from this ?party political? adjudication, a ?privileging? of party politics above other criteria included in the terms of the TRC legislation. This created difficulties ?in determining party policies and what acts should be considered under those policies? (Bhargava, 2002, p. 1314), and required the TRC to ?take party political policies at face value? (Wilson, 2001, p. 87). 197 Other authors focused on the way in which the ?privileging? of political party cleavages obfuscated more complex lines of antagonism and did ?more to mystify than to explain continuity and change in patterns of violence? (Simpson, 2002, p. 221), particularly those that underlay ?horizontal? conflict between members of society, where ?criminal? and ?political? motivations were frequently ?intertwined? (Foster, Haupt & De Beer, 2005, p. 12) or the distinctions between them ?hazy? (Foster, Haupt & De Beer, 2005, p. 21). This distinction between political objective and common crime, Wilson argued, ?hinged on a bipolar distinction between public ideology and private belief?, a ?dualism? that was ?fixed and authorised by TRC legislation? (Wilson, 2001, p. 84), leading Simpson to contend that ?competing versions of whether specific acts were criminally or politically motivated, whether they were undertaken for personal reasons, such as revenge or pecuniary gain, or in the name of a known political organisation, were seldom conclusively or consistently adjudicated by the Amnesty Committee? (Simpson, 2002, pp. 236-237). Wilson quite rightly identified these distinctions between the ?political? and the ?criminal? as the critical boundary on which human rights are founded, arguing that ?separating human rights and common crime is the first distinguishing act of all human rights institutions? (Wilson, 2001, p. 81), but he saw this as a purely pragmatic distinction essential to all human rights institutions, which was nevertheless ?not wholly arbitrary?, although ?often indistinct and ambiguous? (Wilson, 2001, p. 81, own emphasis). In identifying the symptoms of the exception in the ambiguity of these distinctions between the political and the criminal, what again remained implicit was the fundamental biopolitical fracture between political life and bare life, which this distinction attempted to mediate, and the relation of ban such a distinction articulates, which places the one who has been banned, friedlos, ?the criminal?, on the boundaries of the juridico-political order. 198 These difficulties of adjudication of the political and non-political actor arose in a number of arenas, exposing the aporia of an instrumental arbitration of these complex boundaries, where for example ?racism? was invoked as a political motive, which was not the stated public policy of any organisation but which had systematically informed the biopolitical classifications of the apartheid system, and was interpreted in the context of the TRC variably as either ?personal malice? or in other cases as the legitimate basis of ?politically motivated? violence. ?Witchcraft?, which was invoked as political motivation, presented similar problems of adjudication around the distinction between ?personal malice? and public political objective that ?might have seemed to fall outside the requirement of having a political motive?.yet on closer investigation?frequently masked profoundly political issues? (TRC, 1998b, p. 12). The actions of members of informal defensive structures or politicised gangs, or people involved in various forms of internalised conflict whose subject positions were so starkly enacted in the fluidity of praxis, presented similar problems of adjudication in terms of the TRC?s framing of the singularity of an a priori subjecthood as ?victim? or ?perpetrator?, which meant that the Commission ?had great difficulty with these cases? (TRC, 1998b, p. 77). The adjudication of the Moleleki execution itself within the context of the Truth and Reconciliation Commission also produced ?great difficulties? and problems of decidability as the institution attempted to adjudicate the ?political? nature of these actions in purely juridical terms, as the product of a prior ?political? intention directly related to the policies of a political party. However, these were actions which had in fact taken place in non-relation to law, in a context of exception at the boundaries of the juridico-political order, a place of juridical void. Ultimately, therefore the nature of these actions, whether legislative or transgressive, human or divine, was in fact undecidable in law. 199 CHAPTER SIX: THE MOLELEKI EXECUTION AND THE TRUTH AND RECONCILIATION COMMISSION Introduction It is in terms of the principle of the biopolitical fracture that it is possible to begin to understand the complexities of juridical inscription encountered by the post-apartheid juridico-political order in its attempts to adjudicate the actions of the Moleleki protagonists within the contexts of the TRC and the subsequent courtroom processes. However, the significance of these early adjudications, which would establish the biopolitical principle on which the sovereignty of the post-apartheid state would be founded, remained implicit. Thus, the grounding of sovereignty in the ?people? would require a continual mediation of the biopolitical fracture that sought to trace a threshold between ?inside? and ?outside? the political domain through a distinction between ?authentic life and life lacking every political value? (Agamben, 1998, p. 132). The protagonists of the Moleleki execution were some of the earliest subjects of these ongoing processes of arbitration. Critically, the mediation of this biopolitical fracture, which defines inclusion and exclusion in the political community, is substantiated against the boundary of homo sacer, one who has been judged on account of a crime. It was in these terms that the struggles of juridico- political ordering took place in Moleleki prior to the execution, and it was in these terms that the adjudication of the actions of the Moleleki protagonists implicitly took place, first within the context of the institution of the TRC (an adjudication of the ?political?) and then within the context of the post-apartheid courtroom (an adjudication of the ?criminal? at the boundaries of the ?political?). 200 However, the biopolitical nature of the fracture that both institutions sought to mediate in relation to each other remained completely implicit and were instead assumed to concern a process of juridical classification and inscription that could take place through an appraisal of the subjectivity of the Moleleki protagonists, which would putatively establish an instrumental relation with a prior intention, either ?political? in the assertion of a political motive or ?criminal? in the admission of mens rea or criminal intent. These processes of adjudication were critically premised on an instrumental conception of agency as the direct product of a predetermined consciousness and identity. As Lwandile Sisilana writes, what the law understands as ?human action? involves two elements, ?intention and conduct? (Sisilana, 1999, p. 303). Thus human action in law is constituted through the demonstration of the existence of these two bounded, separable elements and the establishment of an instrumental relation between them. While it was assumed that the two process within the TRC and the courtroom were significantly different, both institutions, in their attempts to inscribe the execution within the juridico-political order and their mutual efforts to give legal form and consequence to the actions which had led to the massacre, took the law ?as their model and their code? (Agamben, 1998, p. 5). Within the context of the TRC, both SDU and ANCYL members contested this ?legal? conception of the ?political? as they unsuccessfully sought to articulate the biopolitical nature of the fracture which the conflict in Moleleki had attempted to negotiate. In this process they evoked the ?criminal? rather than the party political opponent as the ?originary figure? of the political relation. Such assertions, however, raised in a context of juridical instrumentality, evoked multiple problems of comprehension. 201 The problems of comprehension and decidability around the Moleleki execution were significantly exacerbated by the fact that the conduct which the TRC and the courtroom sought to adjudicate concerned actions that had taken place in a context of exception, a context of suspension of the juridico-political order, of both the law and the norm of law. These actions had occurred at the boundaries of the law, in a context of sovereign exception, and thus presented the post-apartheid institutions which attempted to adjudicate them with a limit concept as actions which stood outside the law. What law would apply to these acts committed during such a juridical vacuum? How could their nature be ascertained? Moreover, the execution in Moleleki concerned the killing of homo sacer, a figure who is exempted from both human and divine law. How could this figure now be retrieved for law? State of Necessity Critically, the arbitrations of the actions of the Moleleki protagonists, in the context of both the TRC and the courtroom, concerned a contestation around the existence and nature of the state of necessity ?on which the exception is founded? (Agamben, 2005, p. 1) and which grounds the ?right? to self-defence ?in those urgent situations in which the protection of the community fails?when the community is in danger and the magisterial function breaks down? (Mommsen cited in Agamben, 2005, p. 43). Crucially, the state of necessity which establishes the right to self-defence exists in an ambiguous zone ?at the limit between politics and law? (Agamben, 2005, p. 1), an ?uncertain borderline, fringe, at the intersection of the legal and the political? (Fontana cited in Agamben, 2005, p. 1). It was exactly this limit which the processes of adjudication around the Moleleki execution sought to mediate. Located within these mediations was the principle 202 of sovereign decision on the exception, which is necessitated by the ?impossibility? of ascertaining with complete clarity when a situation of necessity exists, nor can one spell out, with regard to content, what may take place in such a case when it is truly a matter of an extreme situation of necessity and of how it is to be eliminated. (Schmitt cited in Agamben, 2005, p. 55) What emerges in this context when juridical authority breaks down or is deliberately suspended is a form of power which is not juridical and is not vested in formal office. Instead it is a biopolitical power that exceeds formal office and in a context of exception is vested in every citizen regardless of rank. It therefore cannot subsequently be adjudicated in terms of the ?proper? exercise of juridical authority mandated by legislative power. Agamben uses the paradoxical institution of Roman law, the iustitium, to ?observe the state of exception in its paradigmatic form? (Agamben, 2005, p. 41). The iustitium suspended the law and created a juridical void, specifically in response to a political context of civil war or tumultus, through a senatus consultum ultimum (final decree of the Senate) which ?called upon the consuls?in extreme cases, all citizens to take whatever measures they considered necessary for the salvation of the state? (Agamben, 2005, p. 41). Importantly, the iustitium put aside the restrictions of the law on the actions of those in formal office, in particular the prohibition against magistrates putting a Roman citizen to death without orders from the people (Agamben, 2005, p. 45). In this context of a suspension of the law, of a putting aside of legal restrictions on the actions of magistrates, consuls and governors, what emerged was not a form of dictatorship, a new form of office, but instead a power that was outside the law and inside every citizen, regardless of formal rank, which consequently ?resists definition within the terms of the normal order? (Agamben, 2005, p. 43). 203 However, the efforts of the both the TRC and the courtroom to exercise this sovereign decision on the existence of the condition of necessity, ran aground in multiple aporiae, created by the assumption that the state of necessity and the ?right? of self-defence which it generated, was an objective situation that could be legally adjudicated. SDU members claimed this situation of necessity as their defence in the proceedings of both the TRC and the courtroom, a situation which therefore ?demanded violence? (O. Motlokwa, personal interview, 2001) and which all protagonists in the conflict in Moleleki in fact understood to be the circumstance at the time. As ANCYL and civic leader Lethusang Rikaba concurred, ?According to the situation, anybody who attacked was regarded as an enemy? (TRC, 1998a, p. 682, own emphasis). However, the state of necessity in fact concerns an ambiguous zone between fact and law, an originary context that ?cannot be regulated by previously established norms? but which at the same time ?makes law?, and ?itself constitutes a true and proper source of law? (Romano cited in Agamben, 2005, p. 27), as in the case of revolution, which is a state of fact that ?cannot be regulated by those state powers that it tends to subvert and destroy? and is in this sense ?anti-juridical even when it is just? (Romano cited in Agamben, 2005, p. 28). While revolution may be anti-juridical in relation to the positive law of the state it intends to subvert, it is nevertheless concerned with the generation of new norms of law (Agamben, 2005, p. 29). This is a context therefore ?where fact and law appear to become undecidable? (Agamben, 2005, p. 29). It was these ambiguities that the TRC and the courtroom struggled so much to negotiate as they attempted to adjudicate the actions of the protagonists of the Moleleki execution and inscribe this violence unleashed from law within the law again. However, if the power which 204 emerges in a context of exception, and which underlies the prerogative of ?defence? in such circumstances, does not concern a juridical power exercised in terms of the mandate of formal office, but is an embodied power, how can such biopolitical power be adjudicated subsequently within the terms of the juridico-political order and what is the nature of this human praxis ?that is wholly delivered over to a juridical void?? (Agamben, 2005, p. 49) As Agamben argues, these actions are in fact ?radically removed from any juridical determination? (Agamben, 2005, p. 50) and cannot be given the conventional forms of law, whether ?legislative, executive, or transgressive? (Agamben, 2005, p. 50). The TRC and the court proceedings which followed them thus grappled with this ambiguity of human action performed in a context of exception, which evoked the ?impossibility?of clearly defining the legal consequences of those acts committed during the iustitium with the aim of saving the res publica? (Agamben, 2005, p. 49), critically, in terms of the Moleleki execution ?whether the killing of an uncondemned Roman citizen was punishable or not? (Agamben, 2005, p. 49). TRC Adjudication of the Moleleki Execution In 1998 13 members of the SDU in Moleleki Extension 2 applied to the TRC for amnesty for their involvement in what the Commission termed the ?Katlehong massacre? in the hope that this amnesty application would release them from the legal consequences of their actions. Interestingly, however, the SDU members who applied to the TRC for amnesty included not only the seven SDU members who had been charged by the state at the end of 1997 but also a number of other SDU members who came forward, as a result of the efforts of the TRC?s investigative unit. The applicants before the Amnesty Committee were listed as: 205 Michael Armoed 1st Appl icant (Am 5596/97) Michael Langa Nkoma 2nd Appl icant (Am 7265/97) Mohale Oscar Motloka 3rd Appl icant (Am 3135/97) Petrus Vusumuzi Mthembu 4th Appl icant (Am 7350/97) Zola Nceba Michael Sont i 5th Appl icant (Am 7177/97) Paulos Matlalepule Shongwe 6th Appl icant (Am 7183/97) Tsekimyana Rodrick Singo 7th Appl icant (Am 7179/97) Norman Sol ly Mashinini 8th Appl icant (Am 7081/97) Lefu Patr ick Mokoena 9th Appl icant (Am 7097/97) Thobi le Phi l l imon Luphindo 10th Appl icant (Am 7753/97) Joseph Moshe August 11th Appl icant (Am 7701/97) Bennet Mnyaka Ndaba 12th Appl icant (Am 7051/97) Themba Chris topher Mtshal i 13th Appl icant (Am 7349/97) (TRC, 1998c, p. 1) According to Section 18 of the Act establishing the TRC, an amnesty application ?could be made in respect of any act or omission that amounted to a delict67 or offence, provided that it had to have been associated with a political objective and was committed in the prescribed period? (TRC, 2003, p. 5), which fell ?between 1 March 1960 and?10 May 1994? (TRC, 2003, p. 8). The granting of amnesty was also contingent on the ?full disclosure of all relevant facts relating to acts associated with a political objective? (Promotion of National Unity and Reconciliation Act, 1995). In adjudicating whether an offence was politically motivated, the TRC Act provided that, ?unless the context otherwise indicates?, the act or 67Delict is a concept of civil law in which a willful wrong or an act of negligence gives rise to a legal obligation between parties even though there has been no contract between the parties. (Delict. Retrieved from Wikipedia: http://en.wikipedia.org/wiki/Delict). 206 delict for which amnesty was sought, in addition to being associated with a political objective, must also be shown to have been ?planned, directed, commanded, ordered or committed? by: (d) any employee or member of a publicly known political organisation or liberation movement in the course and scope of his or her duties and within the scope of his or her express or implied authority directed against the State or any former state or any publicly known political organisation or liberation movement engaged in a political struggle ?which was committed bona fide in furtherance of the said struggle (s 20(2), Promotion of National Unity and Reconciliation Act). Whether an act was deemed to be an act ?associated with a political objective? was to be adjudicated with regard to the following criteria: (a) The motive of the person who committed the act, omission or offence; (b) the context in which the act, omission or offence took place, and in particular whether the act, omission or offence was committed in the course of or as part of a political uprising, disturbance or event, or in reaction thereto; (c) the legal and factual nature of the act, omission or offence, including the gravity of the act, omission or offence; (d) the object or objective of the act, omission or offence, and in particular whether the act, omission or offence was primarily directed at a political opponent or State property or personnel or against private property or individuals; (e) whether the act, omission or offence was committed in the execution of an order of, or on behalf of, or with the approval of, the organisation, institution, liberation movement or body of which the person who committed the act was a member, an agent or a supporter and the relationship between the act, omission or offence and the political objective pursued, and in particular the directness and proximity of the relationship and the proportionality of the act, omission or offence to the objective pursued. The provisions explicitly excluded acts committed for ?personal gain?, or ?out of personal malice, ill-will or spite.? (s 20(3), Promotion of National Unity and Reconciliation Act) The actions of members of SDUs, operating on the boundaries of the juridico-political order in conditions of necessity, immediately provoked multiple problems of decidability for the Commission, which assumed, as did the ANC which initiated their formation, that the ?right? to self-defence concerned a juridical right that could be legally adjudicated. 207 A Problem of Decidability, Adjudicating the ?Right? to Self-defence The problems of decidability encountered by the Commission evoked those encountered by the legal theorist Mommsen who, while acknowledging the essentially extra-juridical nature of the ?right? to self-defence, i.e. that ?in a certain sense it stands outside of the law? and that it is a ?right? evoked by an extra-juridical context, nevertheless insisted that ?it is necessary to make the essence and application of this right of self-defence [juridically] intelligible? (Mommsen cited in Agamben, 2005, p. 43). However, it was exactly this ?intelligibility? which eluded the post-apartheid institutional contexts that attempted to adjudicate the Moleleki execution, as they assumed that ?self- defence? denoted an instrumental power exercised in causal relation to an immediate and objectively identifiable threat. Instead the nature of the power that SDU members evoked when they spoke of their role in community defence was the type of power that emerges in the context of suspension of law, namely a ?floating imperium?, which ?resists definition in terms of the normal order? (Agamben, 2005, p. 43). Thus the Amnesty Committee found that ?applications by former members of SDUs presented the Committee with formidable problems? (TRC, 2003, p. 41). The Amnesty Committee was only obliged to hold public hearings for amnesty applications that related to acts that amounted to ?gross human rights violations?. While the activities of SDU members, like those involved in witchcraft activities, and the acts of violence of some members of the Azanian People?s Liberation Army (APLA), appeared, ?at first glance?to be common crimes? (TRC, 2003, p. 39), the Commission reported that, ?it became evident that these matters could only be properly decided at public hearings where all the relevant circumstances could be fully canvassed? (TRC, 2003, p. 39). 208 Therefore the Amnesty Committee decided, ?after intense discussions prior to the finalisation of SDU applications?, that it would hold public hearings for these matters, even though, ?some?did not, strictly speaking, require a hearing? (TRC, 2003, p. 39). At these hearings, ?the context of the conflict and the activities of the SDUs could be fully ventilated? (p. 43). Although, as we shall see, in terms of the adjudication of the Moleleki execution, the public hearings on this matter did create the opportunity for the ?ventilation? of the context in which the execution took place, the Commissioners, faced with actions that occurred essentially in a juridical vacuum, in a context of necessity and biopolitical power, struggled to interpret the nature of the ?context? that they were confronted with. Thus although ?The hearings helped clarify the political background and context within which these offences occurred? (TRC, 2003, p. 43), the Committee admitted that the various submissions from witnesses, non- governmental organisations and political parties, while ?generally helpful?did not always enable the Committee to reach an informed decision on every individual case? (TRC, 2003, p. 43). The ?legality? of self-defence The Amnesty Committee faced several points of undecidibility in its adjudication of the acts of members of the SDU in Moleleki, as elsewhere. The first concerned the question of the ?legality? of the right to self-defense. Thus while the TRC acknowledged the context of necessity that establishes the right to self-defence, it saw this ?right? as a juridical claim, that could be legally adjudicated in order to ascertain the ?legitimacy? of such actions, a ?legitimacy? that once established would release the perpetrator from the legal consequences of their actions. Therefore, ?a legitimate killing in self-defence still amounts to the deprivation of life and a violation of the right to life, but the law does not hold the perpetrator liable for the consequences of this conduct? (TRC, 1998b, p. 72). 209 Crucially, however, as the TRC acknowledged, the legitimacy of self-defence is often difficult to establish...Amongst the most difficult issues the Commission faced in this regard were cases involving SDUs and SPUs68?where it was usually not clear who was ?innocent? (defending) and who was ?guilty? (attacking). (TRC, 1998b, p. 88) Therefore, while the amnesty provisions that an act had to constitute a legal offence or delict, sought to affirm the law and legality as a norm, meaning that ?there would be no Nuremberg- type trials for the many human rights violations legally committed in the course of implementing apartheid? (TRC, 1998b, p. 52), it raised a critical ambiguity in terms of the amnesty applications of members of SDUs, which concerned the legal consequences of acts committed during the iustitium, with ?the aim of saving the res publica? (Agamben, 2005, p. 49). Paradoxically, those who could prove they had acted in ?legitimate? self-defence had not acted illegally and therefore could not be said to have committed an act which amounted to an offence or delict, and were therefore not eligible to apply for amnesty in terms of TRC legislation. As the Amnesty Committee explained: not unnaturally, SDU members stated in their applications that they had acted in self- defence. On a strict legal interpretation, such conduct is not unlawful and does not, therefore, amount to an offence. As one of the statutory requirements for amnesty is that the applicant?s conduct must constitute an offence associated with a political objective, SDU applicants did not qualify for amnesty. (TRC, 2003, p. 42) The Moleleki applicants? legal representative argued that they carried out the execution because they feared for their lives after the killing of an SDU commander. In response, one of the TRC commissioners asked: if indeed?this was the perception that they were all going to be rounded up and killed, should that not be construed, at best for them, as a self-defensive ? an act of 68 SPUs refer to Self Protection Units, which were informal defensive structures similar to SDUs established by the Inkatha Freedom Party, primarily in response to the formation of SDUs by the ANC. 210 self-defence, and therefore falling outside of the ambit of the Act, that it?s not politically motivated, but motivated by the protection of their own lives? (TRC, 1998a, p. 764) While ?legitimate? self-defence was not an offence in South African law and therefore did not fall within the ambit of acts for which amnesty could be granted, SDU members in Moleleki as elsewhere articulated a more complex notion of ?community defence? as the ?motive? for their actions. This notion of community defence was articulated by SDU members in numerous hearings. The Moleleki protagonists unanimously articulated their motivation for their actions in terms of a duty of ?protection? of a ?defenseless community? (Dwane-Alpman, 1998). However, the TRC was unable to recognise the biopolitical nature of the power which this notion of community defence signified. Community defence and political objectives This obligation of community defence did not concern the individual defence of personal safety, which could be instrumentally and juridically adjudicated to ascertain the ?legitimacy? of these actions as the TRC envisaged, but denoted a much broader mandate that concerned the dispensation that all citizens carry in a context of necessity when the juridico-political order has broken down or is suspended. As SDU member Oscar Motlokwa explained of the activities of SDU members, ?the situation forced us? (TRC, 1998a, p. 171, own emphasis). ANCYL and civic leader Lethusang Rikaba made a similar argument in relation to the carrying of weapons by Youth League members: ?They were eligible to use those guns according to the situation that existed in Katlehong? (TRC, 1998a, p. 662, own emphasis). This necessity of defence could not be given the instrumental forms of law as actions in pursuit of a strategic ?political objective?. SDU members facing questioning from TRC commissioners regarding the ?political objective? of their actions were not able to articulate 211 such instrumental intent. SDU member Michael Sonti applying for amnesty for the Moleleki execution explained to the TRC as follows: ADV MADASA: Now, in relation to the offences which you have admitted. What was your political objective when you captured those people? MR SONTI: There was nothing that I was going to gain in as far as that was concerned. It?s just that all these incidents took place during the time of violence. In other words there wasn?t any objective that I was after or I will have?. we were protecting the community during the time of violence. (TRC, 1998a, p. 306, own emphasis)69 SDU member Singo expressed similar incomprehension when asked to articulate a political objective for his actions: ADV MADASA: What was your political objective in killing them, in participating in their killing? MR SINGO: Sir, can you repeat your question, I don?t understand what you mean by the political objective? ADV MADASA: What did you want to achieve by complying with Dondola?s order [to kill the ANCYL members], what did you want to achieve? MR SINGO: I expected nothing, sir, when I carried out Dondola?s order. (TRC, 1998a, p. 426) Advocate Madasa, who represented the Moleleki protagonists, spuriously attempted to argue that this duty of defence did indeed constitute a juridical ?political motive?, asserting that the Moleleki applicants? ?conduct was definitely associated with a political objective, namely to defend the community? (TRC, 1998a, p. 763) and that this political objective to ?defend the community? was part of ?the mandate from the ANC?, which was ?not only to kill the identifiable, the known enemy, which is perhaps the State and its agent and IVP [IFP], [but] was to eliminate?individuals within the community who were perceived to be harassing the community? (TRC, 1998a, p. 765). However, he was unable to demonstrate an instrumental 69 An SDU member from Katlehong who was applying for amnesty for the killing of an IFP member similarly attests to the dispensation that necessity carries with it, which does not concern an instrumental intent, ?the death of Mr Nqobo is not my intention, but because of the situation prevailing at the time in the Katlehong community? (TRC, 1999, p. 599, own emphasis). 212 link between the actions of SDU members and the political objectives of a ?publicly known political organisation?. He instead evoked the ambiguity of the biopolitical fracture that articulates the margins of the political community against the boundary of bare life, namely those ?perceived to be harassing? the community who were consequently seen by all protagonists in the conflict in Moleleki as the licit object of elimination. However, Madasa?s paradoxical attempt to articulate this biopolitical fracture in juridical terms, made little sense in the context of the TRC proceedings, as stated by chairperson of the committee hearing the application: ?The only issue is whether the conduct of the applicants, who are admittedly members of the SDU and members of the ANC, had any political objective?whether the killing of the deceased had the objective envisaged in the Act? (TRC, 1998a, p. 762). Thus, the difficulties which the Commission faced in its adjudication of self-defence within the context of the conflict of the 1990s, was not centrally about ascertaining the legitimacy of the various actions carried out during the course of this conflict, an adjudication of an instrumental relationship between act and author, that would allow the TRC to determine who was ?innocent? and who ?guilty? but instead concerned the ambiguity of the floating imperium, which emerges in the context of exception. As the report of the Amnesty Committee of the TRC stated in relation to the conflict of the 1990s: While it was possible to draw a sharp distinction between those involved in the clandestine military operations of MK70 and those engaged in other forms of protest in the pre-1990 era, such distinctions become far less clear in the early 1990s. During this period, the borders began to blur as MK operatives became involved in community SDU structures and activities and civilians were increasingly drawn into paramilitary activities. (TRC, 2003, p. 296) 70 Umkhonto we Sizwe, the armed wing of the ANC prior to its de-prohibition in 1990. 213 However, the TRC, operating within a paradigm of juridical instrumentality, was not able to recognise the nature of this authority, which emerges in a context of exception, and instead attempted to adjudicate the actions of SDU members within the amnesty process, by determining whether ?orders? had been given to commit various acts of violence. Nevertheless, as the TRC acknowledged at the conclusion of its amnesty processes, It was clear? that it had not always been possible for SDU members to receive a specific order before launching an attack or operation. The areas in question were, moreover, gripped by large-scale, ongoing and indiscriminate violence where the maintenance of law and order had all but collapsed. Testimonies at the hearings depicted a grim picture of day-to-day survival as communities came under attack by clandestine forces, often operating with the tacit approval and even support of the security forces. The East Rand in the early 1990s offered a clear example of this, with young people testifying about their involvement in violent operations in defence of their communities. (TRC, 2003, p. 43) SDUs and the ANC Thus in attempting to adjudicate the actions of the participants in the Moleleki execution, the nature of the authority for their actions raised a number of significant ambiguities. While all SDU members in Moleleki were members of a ?publicly known political organisation?, namely the ANC, the extent to which the acts of violence they committed could be construed to have been carried out ?within the scope of his or her express or implied authority? (s 20(2), Promotion of National Unity and Reconciliation Act) as a member of the ANC, raised one of many complexities in the adjudication of their actions. In its attempts to adjudicate the Moleleki execution, the TRC operated in terms of a juridical conception of the ?political? premised on a separation between an atemporal origin of power and its violent effects, which allows for a process of legitimation and rationalization by attempting to trace an instrumental link between formal political rationalities (the origin of power) and its violent effects (the offences or delicts carried out by those who applied to the Commission for amnesty). The TRC ran aground, however, in multiple problems of 214 decidability when confronted with the ?manifestation? of violence of the Moleleki execution. This violence, as Van Zyl argued in relation to much of the violence which took place during the 1990s in South Africa, was ?political in origin but not politically motivated? (Van Zyl, 1990, pp. 1-2, own emphasis). Moreover, the ambiguous relation of SDUs to the ANC created significant difficulties in terms of determining an instrumental relation between the origin of power (formal political rationalities) and its violent effects in terms of an adjudication of the ?political motive? of members of SDUs, in particular whether their actions could be linked to the formal political rationalities of the ANC. While SDU structures had initially been established under the auspices of the ANC, once established they were ostensibly accountable to the ?local community?, purportedly represented by the national civic organisation, SANCO. Thus, as the ANC stated in its submission to the TRC, ?the SDUs were established in communities under attack as a joint project between the ANC and the communities concerned? (African National Congress Statement to the Truth and Reconciliation Commission, Further Submissions and Responses, 1997). However, it also acknowledged that while ?the units should have been controlled by the communities in which they operated?many communities were entirely destabilised by low intensity violence, and organised structures at grassroots levels were almost non-existent? (African National Congress Statement to the Truth and Reconciliation Commission, 1997). In this context the location of the political authority to which SDU members should refer in the exercise of their ?duties? remained extremely ambiguous throughout the period of their operation. These ambiguities were deepened by the context of necessity in which SDU structures had been established, which concerned a ?war situation? in which it was difficult if 215 not impossible to secure clear and ambiguous ?orders? for actions committed in the context of ongoing conflict and violence. As SDU member Oscar Motlokwa explained, ?if [the] commander was not there and the enemy was attacking, we had to take the initiative to defend the community and account later? (O. Motlokwa, personal interview, 2001, own emphasis). Thus as Thomson argues in relation to the law in such situations, ?if there is?a sudden danger, regarding which there is no time for recourse to a higher authority, the very necessity carries a dispensation with it, for necessity is not subject to the law? (Thomson cited in Agamben, 2005, p. 25). In preparation for the hearing of the amnesty application of the Moleleki protagonists, the investigative unit of the TRC had sought formal confirmation from the ANC that the SDU members who had applied for amnesty were ANC party members. The response of the ANC essentially confirmed the dispensation that necessity carries with it in a ?war situation?. However, the organisation also attempted to distance itself from juridical responsibility for these actions, which it argued would have had to have been exercised through a direct command: ?No one has instructed them to do the killings because it was a war situation according to the information we got from the PWV office then? (P. Molekane, letter, January 19, 1998, own emphasis). Thus, while the SDU members in Moleleki articulated their defence at the TRC in terms of an extra-juridical power, there remained a critical ambiguity in this articulation, which also invoked a juridical claim to authority, namely that SDUs were fundamentally an initiative of the ANC and therefore were a critical part of the struggle for ?rights? on the East Rand. As SDU member Michael Sonti explained, ?I responded to the call of our political organisation, our political leadership, the ANC? (TRC, 1998a, p. 323). On the other hand, the practical nature of the link between the ANC and SDUs operating in areas where there was conflict was unclear. ?In most cases, SDUs had some form of contact with ANC structures, albeit in 216 an ad hoc and unstructured way. Some existed in areas where there were no strong ANC branches that could provide political leadership? (TRC, 2003, p. 305, own emphasis). Even where there were strong political links between the ANC and SDUs, as in Thokoza township on the East Rand, this tended to occur at a local level and significantly, the existence of such ?political control? did not ?lessen the ferocity of the conflicts or the offensive character of the attacks carried out by the SDUs. Thus, despite political control the Tokoza SDUs engaged in extreme forms of violence? (TRC, 2003, p. 305, own emphasis). In Moleleki, despite the fact that both sets of protagonists in the conflict were ANC members, neither the SDU nor the ANCYL approached the ANC for assistance to help address the division between the organisations prior to the execution. Instead both parties turned to the civic organisation in the township as the appropriate authority to intervene in the conflict, despite the fact that its secretary general in Moleleki, Lethusang Rikaba, was actually partly the source of contention, making this a highly unlikely avenue for the successful resolution of the conflict. Oscar Motlokwa explained to the Commission: MR MAPOMA: When you had problems with the ANCYL before the incident itself happened, did you ever raise the matter with the ANC as an organisation of which you were a member? MR MOTLOKWA: Sir, I told you already that the civic solved the problems because within the SDU, the SDU had members from different political organisations, PAC and ANC. (TRC, 1998a, p. 171) It was only after the execution had taken place that ANC headquarters sent a representative to defuse the situation. However, prior to the execution, ?there were no [ANC] meetings, we were busy at all time because Katlehong was fighting? (TRC, 1998a, p. 171). Asked about ?political guidance? from the ANC prior to the execution, Michael Sonti explained, ?all the decisions that are made in ANC meetings we would obey, even though we did not get enough time or did not get clarity on the decisions made or the code of conduct what we are supposed 217 to do as the members of the ANC? (TRC, 1998a, p. 317). It was this lack of a code of conduct that Oscar argued had led to the emergence of ?Kangaroo Courts?71 in Moleleki, ?because of certain reasons we did not have a code of conduct the situation forced us? That is why most of the time the Kangaroo Court was in operation? It was not the ANC policy to have Kangaroo Courts? (TRC, 1998a, p. 172). Instead as Oscar explained, the Moleleki SDU?s primary link to the ANC was the informal weapons training provided to SDU ?operators?72 such as himself by a former member of Umkhonto we Sizwe (MK), Julius Gadebe (TRC, 1998a, p. 127). Although an MK member may have trained Oscar in the use of weapons, this did not extend to the provision of weapons by the ANC: ?the ANC did not give us guns? although the ?sectional committee of the ANC? ?was aware? that the SDU had guns (TRC, 1998a, p. 129). The TRC amnesty report on the ANC and ?allied organisations? therefore states that ?SDUs?can claim some level of practical and moral authorisation from the ANC? (TRC, 2003, p. 108). However, the nature and extent of this ?practical and moral authorisation? remained unclear both at the time that SDUs actually operated and during the subsequent adjudication of the actions of their members during the processes of the TRC. The TRC amnesty report noted that during the period in which SDUs operated, although ?the ANC kept its distance from the command and control of most of the SDUs, it was forced to intervene in several instances when SDU structures drifted into criminality or internecine 71 (idiomatic) A judicial or quasi-judicial proceeding, or a group which conducts such proceedings, which is without proper authority, abusive, or otherwise unjust. (Kangaroo Court from Wikitionary: http://en.wiktionary.org/wiki/kangaroo_court ). The term ?kangaroo court? may have been popularized during the California Gold Rush of 1849. The first recorded usage is from 1853 in a Texas context. It comes from the notion of justice proceeding ?by leaps? (Kangaroo Court. Retrieved from Wikipedia: http://en.wikipedia.org/wiki/Kangaroo_court). 72 Members of SDUs who were trained and authorised to use guns. 218 conflicts (TRC, 2003, p. 305), as occurred in Moleleki. In the context of the amnesty process of the TRC, the ANC in its ?Statement of Responsibility? to the TRC73, in which senior ANC leaders collectively took responsibility for the armed actions of members of the organisation, appeared to indicate that the organisation was taking direct juridical responsibility for SDUs within the context of the amnesty process: Due to the circumstances which prevailed in the townships, in the early 1990s as a result of third force activities, the leadership of the ANC established and, in some instances encouraged the establishment of self-defence units (SDUs), which played a critical role in the defence of defenceless communities. In the event, and to the extent that any of the activities of any of the abovementioned institutions and structures including the SDUs could in any manner whatsoever be regarded as the kinds of acts or omissions or offences envisaged in the Promotion of National Unity and Reconciliation Act, we collectively take full responsibility therefore applying for amnesty in respect thereof? [AM5780/97] (TRC, 2003, p. 272) In practice the extent to which the ANC as a political organisation did take juridical responsibility for the actions of SDU members was less consistent. The TRC Amnesty Committee therefore found that an instrumental relationship between the ANC and SDUs could only be concretely demonstrated by the limited supply of weapons to some SDU structures by the ANC: ?It is probably in the supply of weaponry by MHQ74 that the strongest case for a link between the ANC and SDUs can be made? (TRC, 2003, p. 272). Therefore, although in its establishment of SDUs the ANC had argued that they should be accountable to the national civic organisation, SANCO, in the context of the amnesty processes of the TRC the actions of SDU members were adjudicated in terms of the extent to which they were an instrumental reflection of the political objectives of the ANC and were in 73 This collective amnesty application was at first granted by the Amnesty Committee but was later overturned by the Supreme Court because individual ANC leaders had not specified the particular violations for which they were taking responsibility. 74 Military Headquarters of the ANC?s armed formation, Umkhonto we Sizwe. 219 fact exercised under the direct command of that organisation. Thus, as stated by the TRC evidence leader in the Moleleki hearings, the [amnesty] committee needs to look into the actions, were those actions geared in furtherance of the political struggle waged by the ANC?...my emphasis is on the command that was given. I?m saying, Chairperson, that the command that was given is not a command which came from a publicly known political organisation, that is the ANC. (TRC, 2003, p. 303, own emphasis) These ambiguities reflected the difficulties faced by the ANC in its early attempts to juridically inscribe this ?right? to self-defence at the time of the internal strife of the 1990s after it had agreed to suspend the armed struggle in terms of the Pretoria Minute signed with the National Party government in August 1990. This occurred a month after what subsequently came to be known as the Sebokeng massacre, which launched what was to be a bloody four-year civil conflict. Faced with a massive escalation of violence across the country and increasingly, in response, a spontaneously organised groundswell of defensive activities around the country, the ANC sought to contain this violence, which was ?loosed from law?. As SACP leader Jeremy Cronin argued at the time, ?communities will spontaneously resort to arms to defend themselves and?it is better for political organisations to exert some kind of control over these activities ? rather than press for disarmament? (quoted in Koch, 1992). ANC spokesperson Siphiwe Nyanda argued that ?the ANC?s involvement is a bid to direct the process so that these things don?t degenerate into something ugly? (quoted in Sole, 1991). Critically the way in which the ANC attempted to inscribe this floating imperium was in terms of a juridical notion of self-defence. The ANC?s key policy document on SDUs For the Sake of our Lives! (1991), authored by South African Communist Party leader Jeremy Cronin, was the most systematic articulation of this juridical right. The document asserts, 220 Our people have the moral right to state: We do not intend to attack anybody but we demand the right to protect our lives, our families, our homes and communities. We are forced to create defence units for the sake of our lives. (Cronin, 1991, p. 5) Therefore in the light of the suspension of the armed struggle, the ANC sought to reassure its constituency of its continued support for their ?right? of self-defence. Although this ?right? could no longer be exercised through the formal structure of the ANC?s armed formation, Umkhonto we Sizwe, it would now be exercised through the formation of SDUs. For the Sake of our Lives! explains, ?At present, in the light of the Groote Schuur and Pretoria Minutes, Umkhonto we Sizwe (MK) alone cannot undertake the task of our people?s defence, although this is a right we need to forcefully demand and struggle for? (Cronin, 1991, p. 4). This juridical ?right? was allegedly a ?non-partisan? right that belonged to all protagonists in the conflict, a right which had already reportedly been acknowledged on a partisan basis by the apartheid government, which had lifted a ban on ?traditional weapons? as IFP members mobilised to attack townships and also refused to act against the Afrikaner Weerstandsbeweeging (AWB75) ?self-protection? commandos. Consequently, the SDUs established by the ANC would ostensibly not ?be affiliated to any political party or movement but [would] be a protective force which serves the community as a whole? (Cronin, 1991). Nor would they in their initial conceptualisation be directly armed by the ANC, although the ANC did eventually start to supply SDUs with small amounts of arms as the conflict intensified and SDU members used illegal means to acquire arms. However, at the start of the conflict the ANC emphasised the importance of the principle of the right to self-defence. ANC spokesperson Gill Marcus argued therefore: ?We are not talking about 75 A rightwing organisation established in 1973 under the leadership of Eugene Terreblanche to agitate for a ?homeland? for the ?Boer volk? (people). (Retrieved from Afrikaner Weerstands Beweeging, AWB information: http://www.awb.co.za/inligting_e.htm) The AWB were involved in a considerable degree of violence during the period of negotiations including several bomb attacks in the days before the first popular elections took place in 1994. 221 arming our people but about their right to self-defence?? (quoted by Own Correspondent, Sowetan, November 1990). The ?community? to which SDUs were purportedly accountable was a ?community? represented through SANCO, the civic organisation operating in townships, ostensibly open to members of all political parties but which clearly could only represent the interests of a small section of ?the community? in a context of ongoing conflict, which had already balkanised areas of conflict into geographical zones of homogenous political affinity. As the ANC later acknowledged in its testimony to the TRC, There was a basic assumption, which again may also have been wrong?which was that there would be in these communities local political structures, local structures of civil society strong enough to be able to constitute these committees that would then take charge of the self-protection units?it?s clear that we should perhaps have reviewed the matter of that control but we continued to proceed as though you could as ANC armed (sic) the units and surrender them to these local civil and political structures to control?.there was a carry through of a particular concept of self- protection units which was perhaps then not founded on reality with regard to the control and so on within those communities. (African National Congress Political Party Recall, 1997, Day 2) Homo Sacer, Mediating the Boundaries of the Community In evoking the ?community? as the key referent for the authority of SDUs, the ANC was in fact evoking the ambiguity of the exception which articulates the boundaries of the juridico- political order. In a context where the juridico-political order had essentially been suspended, the ?community? to which the ANC referred did not reflect a community that could be juridically or electorally represented; instead it could only denote the fundamental ambiguity of the biopolitical fracture, mediated in terms of the relation of exception. The mediation of this biopolitical fracture, which defines inclusion and exclusion in the political community, is substantiated against the boundary of homo sacer, leading the protagonists in the Moleleki 222 execution to evoke the ?criminal? rather than the party political opponent as the ?originary figure? of the political relation. For the TRC in its adjudication of the political motive of the Moleleki applicants, the fact that their actions, by their own account, had not been directed at a ?political opponent? as defined in the TRC legislation but against people whom they perceived to be ?criminal?, played a defining role in the decision that the actions of the SDU members in Moleleki had not been politically motivated. The chairperson of the Amnesty Committee hearing the Moleleki application questioned SDU applicant Oscar Motlokwa in this regard, evoking in his line of questioning the enlightenment notion of a ?transcendental? political subject who exists prior to action (Feldman, 1991, p. 4). This was a notion on which the conception of intention, which informed the TRC?s adjudication of political motive, was heavily reliant, i.e. the extent to which the actions of political protagonists could be said to have been carried out in terms of their political intention as particular types of political subjects, adjudicated in terms of their membership of political organisations. However, the protagonists in Moleleki, both ANCYL members and SDU members, articulated a different type of subjecthood, one that is embodied in praxis, rather than as a product of a pre-existing political consciousness. As ANCYL and civic leader Lethusang Rikaba argued, ?anybody who attacked was regarded as an enemy? (TRC, 1998a, p. 682, own emphasis). SDU member Oscar Motlokwa similarly articulated the ?political objective? of the SDU as the protection of ?the community against an enemy?, defined not in terms of organisational affiliation but in terms of praxis that subordinated juridical identity, MR MADASA: So the Youth League, you perceived them as an enemy within you? MR MOTLOKWA: The way they acted, that is so. (TRC, 1998a, p. 167, own emphasis). 223 It was the actions of ANCYL members that made their political party membership irrelevant and rendered them as homines sacri, ?criminals?: CHAIRPERSON: At the time when these persons were killed you believed that they were members of the ANCYL? MR MOTLOKWA: I believed that they were criminals Sir. CHAIRPERSON: You did not believe that they were members of the ANCYL? MR MOTLOKWA: We did not believe that they [were] members of the ANCYL according to what they were doing. CHAIRPERSON: But did these persons who were killed hold themselves out as members of the ANCYL? MR MOTLOKWA: That is correct Sir. CHAIRPERSON: But insofar as you are concerned you were killing criminals? MR MOTLOKWA: According to their actions yes, they were criminals. (TRC, 1998a, pp. 194-195, own emphasis) These assertions of criminality by the SDU would lead the Amnesty Committee to conclude in its decision on the Moleleki execution: All the applicants testified that they regarded the members of the ANCYL as criminals who were harassing the community in Moleleki section. They also testified that the deceased were killed because they were a gang of criminals? On the applicants? version, therefore, the deceased were killed, not because of their membership in a political organisation, but because they were a gang of criminals who had killed Blanko in particular. (TRC, 1998c, pp. 8-9, own emphasis) In their evocation of the ?criminal? rather than the party political opponent as the figure of the original political relation, all protagonists in the conflict in Moleleki were evoking the ambiguity of the biopolitical fracture that mediates the boundaries of ?community? in terms of the figure of homo sacer. This ?community? was evoked by all protagonists in Moleleki as the source of authority for a variety of decisions to commit acts of violence rather than an ?order? or command from a specific leader. It was this ?community? that leaders also drew their authority from when they gave ?instructions? to carry out various acts of violence. This ?community? was not organised as a representative juridical authority but was simply embodied by fluctuating groups of ?community? members who gathered in various meetings in the township of Moleleki as tensions escalated in the months prior to the execution and 224 who congregated even more fluidly at various sites of death during the course of the night prior to the execution. A number of writers have noted the centrality of ?community? to African conceptions and practices of justice: punishment and the resolution of disputes will lay emphasis upon law as expressing the will and traditions of the community. There is no distinction between legal and moral issues?The person at the bar of judgment is there, in principle, as a whole man, bringing with him his status, occupation, and the entire history of all his social relations. Justice is substantive and is directed to a particular case in a particular social context and not to the creation of a general rule or precedent. Punishment, as Foucault has stressed, is a social drama, symbolising the awesome power of the group over the individual ? there is a sharp dichotomy of reconciliation and outlawry? (Hund and Kotu-Rammopo, 1983, p. 201, original emphasis). Crais similarly notes: Typically infractions involved as much the person directly involved as the community of which he or she was a member. Thus people could be found guilty of offences they themselves did not commit?the central issue revolved around the restoration of balance within the community, less personal culpability or the restoration to the individual of the harm done to them. (Crais, 2002, p. 86) Sch?rf and Ngcokoto also write that populist notions of democratic justice tend to be characterised by a frank acceptance that the decisions in collective tribunals, courts, street committees?or whatever they are called, should express the moral/political will of the most powerful collectivity within the particular constituency at a particular time. (Sh?rf & Ngcokoto, 1990, p. 343). However, the evocation of this ?community? provoked significant problems of comprehension for the TRC in its adjudication of the Moleleki execution. Who was this ?community? that was invoked so readily by all protagonists, where could it be found, how did it make and communicate its decisions to authorise violence? 225 As the TRC could only recognise a juridical form of authority, where decisions would be taken and communicated in terms of the mandate of this representative office, the nature of the authority claimed by protagonists in the conflict raised considerable confusion in the interaction between the officials of the TRC and the protagonists from Moleleki who gave evidence to the Amnesty Committee as this authority evoked an auctoritas that is realised in the person of the actor rather than through the mandate of representative office. For example, in his evidence to the Amnesty Committee, SDU member Oscar Motlokwa evoked the ?community? as the source of authority for SDU actions, such as the 9pm curfew that the organisation had imposed on the township. He claimed: ?All members of the community took that decision?the community took a decision that as from 9 o?clock people should not roam in the street? (TRC, 1998a, p. 130, own emphasis). In response TRC Commissioners attempted to make material and hence intelligible this concept of community: MR MALAN: Could I again just ask a question here just to clarify for me. The community that were taking all these decisions, were they meeting regularly? How did you interact with what you?re referring to all through your evidence as the community?? MR MOTLOKWA: As I said before, we would meet with the community if we encounter problems. We would tell the sectional committee and the sectional committee would call the community to explain our problems. MR MOTATA: But I suppose Mr Motlokwa, before you could tell the community what your problems are or problems which you encounter, there should have been a meeting...what did the community which you are referring to do so that your formation could come into existence? MR MOTLOKWA: May you please repeat your question Sir? (TRC, 1998a, p. 130, own emphasis) In the context of the unravelling of juridical authority that occurred during the night after the death of civic activist Bulelwa, the fluidity of the location of this ?community?, claimed by all protagonists in the conflict as the source of the authority for their actions, caused considerable confusion for the Commission, who sought the instantiation of a juridically represented ?community?, fixed in time and space. 226 The nature of the authority for the decisions that were taken to authorise the key incidents of violence that took place during the night following the killing of young civic activist Bulelwa Zwane, therefore evoked a radical problem of understanding for the Commission, firstly in terms of the killing of SDU leader Blanko and then, in response to this killing, the decision to round up and later to kill ANCYL members allegedly involved in the killing of Blanko. Thus Lethusang Rikaba, who acknowledged his role in Blanko?s killing at both the TRC hearings and the courtroom proceedings that followed, evoked the authority of a ?community? of approximately 50 to 70 residents who had gathered in Block C of the township after the killing of Bulelwa, as the source of the sanction for the killing of Blanko. However, TRC commissioners and advocates who interrogated Lethusang regarding the nature of this authority, sought in vain to find evidence of a juridical mandate for this killing and to understand the nature of the ?meeting? itself, which was not addressed by a community leader but had simply gathered in response to the ?screams? of grief of Bulelwa?s friend and fellow civic member, Alfred Buthelezi. It was in the context of this spontaneous gathering that Lethusang drew the authority to ?hunt down? members of the SDU allegedly involved in the killing of Bulelwa and in other incidents of violence in the township. However, this authority did not flow from the ?deliberations? of a juridically representative leadership or authority, nor was it expressed in terms of a juridical mandate, but was instead embodied and realised in the very persons of the members of the ?community? who had gathered in response to Bulelwa?s death. Lethusang could no more understand the juridical nature of the authority that the Commission required him to articulate than the Commission could understand the biopolitical authority he invoked: MR MOTATA: Whilst gathered at this tuck shop or Spaza shop you deliberated with the community what next to do, isn?t it so? 227 MR LEKABE (sic): The community pronounced its dissatisfaction ? they said enough is enough. I did not address them. MR MOTATA: I want to understand, Mr Lekabe, that at that stage nobody knew who killed Bulelwa. How did you arrive at the decision that you should now round up the members of the SDU and start with the top structure, probably Blanko. MR LEKABE : Chairperson I explained already that the community was dissatisfied of the activities of the SDU in Mololeke, they took a decision that enough was enough, they were not going to wait for yet another incident. (TRC, 1998a, p. 681, own emphasis) MR MOTATA: What I want to find out from you? how did you come to a conclusion or decision by the community that you should go for the SDU and Blanko in this instance? MR LEKABE: My lord, I explained that the community said enough was enough and the allegations that members of the community, the leaders of the community, were supposed to be killed was already flying. I think I explained that my lord. MR MOTATA: Yes, my brother, I heard that one but you are not answering my question. MR LEKABE: Maybe I do not understand your question. (TRC, 1998a, p. 682, own emphasis) The context of ?community? evoked in relation to the later abduction and ?presentation? of ANCYL members to this ?community? before the decision to kill them, posed similar problems of comprehension for another advocate involved in the juridical interrogation of the Moleleki protagonists, MR MADASA: When you say you took them [ANCYL members] to the community, what do you mean by that? Where was the community where you took them to?? MR MOTLOKWA: At about 7 o?clock in the morning there was a community meeting to discuss the incident of the previous day, then the community lost control and they said these children must be killed. (TRC, 1998a, p. 154, own emphasis) The ?decision? to kill ANCYL members who had been rounded up and captured in response to the killing of Blanko evoked similar ambiguities. TRC commissioners attempted to identify and fix the ?location? of the decision to kill the ANCYL members in time and space in order to ascertain the extent to which this decision reflected the mandate of a juridically represented community. Thus Advocate Madasa sought to clarify from SDU member Oscar Motlokwa, ?Where was the killing ordered, was it a meeting or what was the situation when these three ordered the killing? What was going on at that time?? (TRC, 1998a, p. 156, own 228 emphasis). Advocate Mapoma sought to identify the time at which the order had taken place in order to ascertain its relation to a juridical mandate: ?Now this decision to kill them [ANCYL members], when exactly was it communicated to you...? (TRC, 1998a, p. 182, own emphasis) The ?orders? for the killing were in fact given by SDU commander Njebe Ndondolo in a shadowy zone on the boundaries of a ?meeting? in the ?yard? of SDU leader Blanko. This meeting was actually a spontaneous gathering of people, who had congregated throughout the night in response to Blanko?s killing. The meeting became more formalised as it was eventually addressed by civic leader Machinini at approximately 7am. The killings were ?ordered? on the boundaries of this meeting in the early hours of the morning. According to Oscar Motlokwa ?after the meeting of the community Ntjebe called the members of the SDU aside and he said Manyala commanded him that these people [ANCYL members] must be killed? (TRC, 1998a, p. 156). SDU leader Njebe Ndondolo explicitly drew authority from the ?meeting? in Blanko?s yard and the anger of the crowd gathered there, but as Oscar Motlokwa explained, did not act in terms of a juridical mandate openly articulated in the context of the public meeting: CHAIRPERSON: Who came with the suggestion that these nine people that were caught had to be killed? MR MOTLOKWA: I already mentioned that Ntjebe Ndondolo issued out an order which was given to him by Manyala. CHAIRPERSON: Did he announce that at the meeting? MR MOTLOKWA: No, he did not announce that in the meeting, he told us as members of the SDU. CHAIRPERSON: What happened at the meeting, what was the meeting about? MR MOTLOKWA: I said the vice chairperson addressed the meeting, the vice chairperson of the ? the zonal vice chairperson. The issue on the table was the situation around Moleleki and the death of the two people [Bulelwa and Blanko]. He was addressing the meeting as the leader, he did not take out such an order at the meeting. Because you know, if people are angry you can?t control their emotions, they just decided that these people must be killed and he called us aside as members of the SDU. It is there where he took out an order. 229 MR MOTATA: What interests me Mr Motlokwa, you were present at this meeting where the community was outraged by the happenings of the previous night, that is the deaths of Bulelwa and Blanko. Were you present in that meeting? MR MOTLOKWA: I was present Sir. MR MOTATA: This meeting as we know by now was addressed by Mr Machinini, would I still be correct? MR MOTLOKWA: That is correct. MR MOTATA: Now, whilst the community is there and Machinini is addressing the community, you being present with some members of the SDU?s, did the community itself in your presence take a decision about the fate of the nine? MR MOTLOKWA: The community was talking that they should be killed. No decision was taken by the community or the chairperson in agreement with the community that they should be killed, people were just saying it out of their own feelings. (TRC, 1998a, pp. 156-157, own emphasis) While Machinini did not explicitly order or sanction the killing of the ANCYL members, the ambiguity of the boundary between what was formally authorised in terms of an overt juridical mandate, and what actions drew implicit authority from the angry crowd gathered at Blanko?s home, was made clear by the fact that civic leader Machinini arrived to address the meeting at Blanko?s home in the company of Katlehong SDU commander Mzaapindile Ntsingola (known as ?Manyala?) and Moleleki SDU commander Njebe Ndondolo, who would later issue the order that the youths should be killed. Therefore, the extent to which this ?community? had or had not been part of the decision to kill the youths, in terms of a juridically expressed mandate, could not be interpreted in terms of the juridical authority the Commission attempted to extract from the testimony of SDU applicants. Instead they invoked the biopolitical authority of the ?community?, on the basis of whose ?feelings? and ?views? SDU leaders Njebe Ndondolo and ?Manyala? purportedly issued the ?order? to kill the captured ANCYL members: MR MAPOMA: In that meeting which was chaired by Machinini, was there any resolution taken that the youth be killed? MR MOTLOKWA: I said that Machinini did not pronounce the death of these people, he was actually addressing the issue of stability in Moleleki, that crime must 230 be stopped. And the people, the community, raised their views, they said they must be killed but the Chairperson of that meeting did not agree to that decision. MR MAPOMA: So out of those views which were given by some people in the meeting, then an instruction from the SDU leadership came to you that they be killed, is that what I understand? MR MOTLOKWA: That is correct. (TRC, 1998a, p. 182, own emphasis) In a juridical conception of the ?community?, it was only the juridical representative of this ?community?, namely civic leader Machinini, who, in the eyes of the TRC, had the legitimate mandate to authorise actions on behalf of this ?community?. MR MAPOMA: Now Machinini, let me get Machinini?s role, was he chairing this meeting or was he addressing as a speaker in the meeting? MR MOTLOKWA: He was chairing the meeting Sir. MR MAPOMA: Now in the process of chairing the meeting there are some decisions which are taken and the chairperson conveys those decisions to the meeting, is that not correct? MR MOTLOKWA: In general that is the procedure. MR MAPOMA: Yes. Now in this particular instance, is there any stage where Machinini announced that it is resolved by the community that these youths who were captured must be killed? MR MOTLOKWA: No, Sir, he never said that, he never announced that. MR MAPOMA: Will I be understanding it correctly if I say that it was never a decision of the community that the youths be killed? MR MOTLOKWA: A decision was not taken by the chairman of the meeting that those youths be killed, an order was issued out by the commander of the SDU that these youths be killed. MR MAPOMA: Thank you, through you Mr Chairperson. Now how, in your opinion, did the meeting resolve the issue of the captured youths, what should be done about them? MR MOTLOKWA: Some members of the community wanted these youths to be killed. I have already said that Machinini addressed the issue of stopping crime in Moleleki, the order came from the commander. (TRC, 1998a, p. 183) Thus, the fact that SDU member Oscar Motlokwa explicitly did not invoke a juridical authority in attempting to explain the mandate for the killings but instead invoked a biopolitical authority embodied by the angry residents gathered at the site of the Blanko?s killing, led the TRC to reject the legitimacy of the claim that the actions taken by the SDU were ?authorised? by the ?community?, and therefore could have been ?politically motivated?, as this motivation could not be deduced from a juridical authorisation. 231 MR MAPOMA: ... how do you say that it was politically motivated to kill those people? MR MOTLOKWA: The task of the SDU was to defend the community. MR MAPOMA: Yes. But the community had an opportunity to decide what it is that has to be done about these youths and it did not resolve that these youths must be killed. How do you explain this one? MR MOTLOKWA: Chairperson, if we speak of the SDU and the civic, these are similar organisations but an order is not issued out in front of everybody, it must be directed to members of the SDU. Even members of the community who wanted to be part of the killing, Ntjebe did not allow them to take part in the killing because he only wanted members of the SDU. MR MAPOMA: So the killing of the youths was an instruction which came from neither the community nor the ANC? MR MOTLOKWA: I said the order was from Ndondolo. MR MAPOMA: Thank you Mr Chairperson, those are my questions. (TRC, 1998a, p. 184) Therefore, in a context of floating imperium, commissioners struggled to understand the location of this ?community? claimed as the source of authority for violent actions, which denoted not a stable juridical entity, but an embodied polis. Clifton Crais has noted in this regard that ?Western forms of cognition?had an important spatial component that, at a minimum, required the existence of stable political boundaries or jurisdictional units? (Crais, 2002, p. 71). On the other hand, The common pre-colonial tradition of political process had entailed the existence of multiple and overlapping political domains and the creation of unadministered areas on the borders of chiefdoms, a complex map of power in which boundaries more or less remained in permanent flux. This tradition rested on the basis of principles that fundamentally, indeed, radically differed from those upon which colonial rule rested. (Crais, 2002, p. 76) In fact, this Western cognition, which premised the boundaries of the ?imagined nation? (Anderson, 1991) on a geographic outline, constructed a fiction of neutrality, of essentiality, and in this fiction, was able to conceal the essential politicality of this ?nation?, the essential ambiguity of this jurisdictional unit. This was a politicality which the fluidity of ?community? articulated by the protagonists of the Moleleki execution made explicit. It therefore represented not a ?pre-modern? traditionalism, but in fact, shared an essential 232 semantic and conceptual ambiguity with the Western juridico-political order that sought its adjudication. What was evoked therefore by these processes of inscription of the Moleleki execution within the context of the TRC was the ambiguity of human action in a context of exception, which led to a mutual problem of comprehension, a radical problem of understanding between SDU members and the TRC. While the TRC sought a juridical ?explanation? from the protagonists of the execution, an explicitly articulated political motive, what they elicited from the amnesty applicants was an invocation of ?community?, a biopolitical ?explanation?, which evoked the criminal as the boundary of the ?political?. However, the TRC, operating in terms of a juridical understanding of the political, could not interpret the biopolitics of the Moleleki execution as political action. In the classification of the actions of the Moleleki protagonists as criminal actions, the TRC sought to re-establish the political as the juridico-political, a political freed from the stain of violence. Paradoxically, however, implicit in the classification of the actions of the SDU members involved in the execution as criminal was a fundamental biopolitical principle that seeks to define the boundaries of the political in terms of a distinction between ?authentic life and life lacking every political value? (Agamben, 1998, p. 132). 233 CHAPTER SEVEN: FIXING THE LIMITS OF THE LAW, THE MOLELEKI EXECUTION AND THE COURT TRIAL Introduction In having established that the Moleleki killings had not been ?instructed? by a juridical representative of a ?known political party?, namely the ANC, the Amnesty Committee refused the Moleleki protagonists amnesty for their actions. The criminal proceedings, which had been suspended for the duration of the amnesty process, were reinstated, leading to the trial of seven members of the SDU in Moleleki, four of whom, Zola Sonti, Michael Armoed, Oscar Motlokwa and Michael Nkomo, had testified at the Truth and Reconciliation Hearings. Whereas during the amnesty proceedings, most applicants, excluding Michael Sonti, had admitted culpability for the killings, even providing detailed descriptions of the killings, in the context of the court proceedings, all seven accused denied responsibility for the execution. A newspaper report at the conclusion of the court case noted: A six-year long wait for justice by four Katlehong women came to an end this week when the Johannesburg High Court finally convicted five men of the deaths of their sons. The wait was especially painful as the accused had confessed to the crimes before the TRC, but denied all when the matter came to trial. (Mbhele, 1999) According to the court judgement Siviwe Malcoms Ngama (accused 2), Michael Armoed (accused 3) and Bethwell Tabiso Ntoma (accused 4), relied on ?alibi defences?76, while Zola Michael Nceba Sonti (accused 1), Petros Mtembu (accused 5), Langa Michael Nkomo (accused 6) and Oscar Noqola Motlokwa (accused 7), ?admitted being present at the scene of Bulelwa?s body and adjacent areas, but denie[d] any material involvement or participation in 76 Bethwell Ntoma argued that he was in the former ?homeland? of the Transkei at the time of the execution. He was acquitted due to lack of evidence and on the basis of the credibility of his alibi. Ngama claimed that on the night of 6 December 1993 he worked night-shift at the South African Breweries and therefore that he was not present at the time the ANCYL members were rounded up and killed. The court rejected his alibi and found him guilty of ?associating himself? with the common plan or purpose to round up and kill the nine ANCYL members who were executed, as well as with the death of ANCYL leader Isaac Motloung, who was killed separately after being abducted by SDU leader ?Manyala?. 234 the offences? (The State v. Michael Sonti, 1999, p. 1773). Zola Sonti and Michael Nkomo admitted to having ?partaken in the rounding up of the [ANCYL] youths but deny any involvement in their killing? (The State v. Michael Sonti, 1999, p. 1773). The adjudication of the actions of the Moleleki protagonists in the context of the courtroom, as in the processes of the TRC, raised a number of ambiguities which concerned the difficulties of attempting to determine the legal consequences and indeed the nature of actions, which had taken place in a juridical void, in a context of necessity, at the boundary between fact and law, the legal and the political. The assumption that the arbitration of these actions concerned merely the juridical, therefore, caused further complications in terms of the attempt to adjudicate the actions of the Moleleki protagonists as ?criminal?. While criminal law in general requires the demonstration of a direct instrumental relation between the criminal intent of the actor (mens rea) and the actions of this actor, the collective nature of the actions which had led to the Moleleki execution meant that such an instrumental adjudication simply could not be made. Despite this, in the effort to weld life to law, and inscribe this anomie within the juridico-political order, the judge in the Moleleki case used the doctrine of common purpose to attempt to establish the criminality of the actions of the Moleleki protagonists. While ?under South African law murder requires the accused to have done something which he foresaw might lead to death and which in fact did [which is] a causal requirement? (Parker, 1996, p. 79), the common purpose rule significantly expanded the basis of liability for murder in contexts where more than one individual was involved. Critically, it did away with the requirement that there had to be a causal connection between the acts of each individual and the eventual consequence of murder. This was a development, which Parker describes as a reinterpretation of the law, 235 ?in order to facilitate convictions in crowd murders? (Parker, 1996, p. 78) in a context of rising political tensions. The Common Purpose Rule However, the common purpose rule had in fact been on the South African statute books since the late 19th century and was originally imported from English law via the Native Territories? Penal Code to incorporate crimes in which more than one party was involved77. The basic principle was that ?where two or more people associate in a joint unlawful enterprise, each will be responsible for acts of his fellows which fall within their common design or object? (Snyman, 1989 cited in Davis, 1990, p. 135). In its earliest interpretations, the common purpose rule was narrowly applied in terms of the application of the principle of a specific mandate78 for unlawful actions, in terms of which criminal liability could be established. In a judgement in 1917 where a case had been brought against a Boer rebel officer by a farmer because some of the rebel officer?s troops had allegedly stolen cattle and damaged the fences of the farmer, the judge ruled that ?a common purpose to rebel did not encompass the actions of rebel troops who had acted without orders and whose deeds could not be said to form part of the mandate that might be read into a rebellion? (Parker, 1996, p. 82). 77 Section 78 of this Code provided that: if several persons form a common intention to prosecute any unlawful purpose, and to assist each other therein, each of them is party to every offence committed by any one of them in the prosecution of such common purpose, the commission of which offence was, or ought to have been known to be a probably consequence of the prosecution of such purpose (Rabie cited in Davis, 1990, p. 139). 78 In terms of the principle of common purpose by mandate, ?where A agrees with B to commit an offence, she will be liable therefore even if it cannot be shown that she acted upon such agreement and therefore causally contributed to the commission of the offence, if B having committed the offence, is liable for it?(Unterhalter, 1988, p. 674). However, in the case of common purpose by mandate, the critical principle that limits liability is the principle of agreement i.e., ?a person?s subjective consent to be bound by the acts of another?(Unterhalter, 1988, p. 674). Unterhalter argues in terms of the notion of the sovereign self, ?the law shall only attribute the acts of B to A only if A had voluntarily assumed responsibility for B?s actions...[this]is a necessary entailment of the criminal law?s profound commitment to the separateness of persons. Blame attaches to individuals in virtue of their actions because each person is sovereign over his actions and thus responsible for them? (Unterhalter, 1988, p. 674). 236 However, the common purpose rule was used intermittently over the next seventy years to secure convictions for murder in cases where crowds had been involved. Over this period, the basis for criminal liability in terms of the common purpose rule was gradually, if subtly, expanded, particularly in response to crowd murders, where ?evidence of crowd crimes is normally unsatisfactory: witnesses see the result of what has happened but rarely the details of who does what? (Parker, 1996, p. 85). A significant decision by the Appellate Division as early as the 1940s held that it would no longer be necessary to prove a ?previous conspiracy? to develop a common purpose or plan to murder; the courts would now recognise that such a plan could ?spontaneously? crystallise, ?it was enough for people to have acted with common intent on the spur of the moment?, even if ?this co-operation has commenced on an impulse without any prior consultation or arrangement? (R v Mkize cited in Parker, 1996, p. 86). The expanded liability implied by the adoption of the notion of a ?spontaneously? generated common purpose meant that subsequently it would make it possible to impute to various members of a crowd an association with a ?common purpose? to kill, whether or not they had with prior knowledge voluntarily and consciously consented to be party to such a killing and notwithstanding the fact that they may have been participating in the activities of a crowd for a variety of other reasons. This evidentiary change thus ?permitted evidence of this spontaneous common purpose to be read into the multitude of disconnected actions that accompany crowd behaviour? (Parker, 1996, p. 86) It was during the 1980s, against the backdrop of rising civil unrest in South Africa, that the basis of criminal liability in terms of the principle of common purpose would be dramatically expanded to completely do away with the causal link which would normally be required to prove liability for murder. In 1982 Judges Hoexter and Botha argued that it was a ?fallacy? to suggest that proof of causation was necessary. Instead, ?the criminal act of an accused in a 237 common purpose murder consisted, not in an act which is causally linked with the death of the deceased, but solely in an act by which he associates himself with the common purpose to kill? (S v. Khosa, 1982 cited in Parker, 1996, p. 94, own emphasis). Thus what Botha did was to replace causation with active association, therefore cutting ?through the evidential problems of crowd violence? (Parker, 1996, p. 95). However, as Parker points out, because an accused performs an act of association, it does not...follow that he thereby associated himself with the crime, because the law insists that the criminal act and the fault or mental element are distinct. Both must be proved?whereas an act of association must be proved on an objective basis?to prove the mental element it must be shown that the individual accused actually did or must have subjectively known what was happening and thereby associated himself with the crime. (Parker, 1996, p. 95) The application of these principles in the context of the case that came to be known as that of the ?Sharpeville Six? (S v Safatsa 1988 (1) SA 868 (A)), led to a guilty verdict and the imposition of the death sentence on six members of a crowd who had participated in a march to the home of the deputy mayor of the town council of Lekoa during protests against rent increases in the Vaal Triangle79. This case caused enormous local and international controversy, particularly in terms of the application of a sentence of capital punishment for a guilty verdict obtained on highly contestable grounds. However, it was a derivation of the principles of this judgement, which abandoned the requirement of causation, that the court in the Moleleki case invoked nearly a decade later in securing the conviction of the Moleleki accused. 79 The formerly ?white? towns of Vereeniging, Vanderbijlpark and Sasolburg in the Vaal Triangle were bordered by six black townships, which from 1984 fell under the auspices of the Lekoa municipal authority. The Lekoa council was established in accordance with the provisions of the Black Local Authorities Act of 1982. The municipal area of Lekoa was divided into 39 wards and the council was elected in a 14,7% poll in 1983. Economic rentals (i.e. rentals which were expected to fund the township?s administration and services) were introduced in the Vaal Triangle area in 1984 by the Lekoa Council, much against the wishes of residents. The new local authority raised tariffs for rents and services in September 1984, which prompted violent protests. (Sharpeville. Retrieved from South African Townships: http://www.saweb.co.za/townships/township/gauteng/sharpvil.html) 238 The judgement in the case of the Sharpeville Six, drawing on the earlier judgement of Hoexter and Botha, which rejected the necessity to prove a causal link between the act and the actor in common purpose murder cases, went further by appearing to ignore the principle whereby it was necessary to provide evidence of the mens rea or intent of each of the accused to ?associate? with a common purpose to kill, in favour of a reference to the ?intent? of the ?mob? of which the accused formed part and from whom their intent could purportedly be derived. Although the six individuals prosecuted for involvement in a common purpose to kill the deputy mayor could all be proved to have committed a variety of acts of violence during the course of the march, Judge Botha acknowledged that there was ?no proven causal link between the actions of the accused and the death of the deceased? (Unterhalter, 1988, p. 671). Nevertheless Judge Botha argued that proof of an active association with the common purpose of the ?mob? was enough to convict the accused: The individual acts of each of the six accused convicted of murder manifested an active association with the acts of the mob which caused the death of the deceased. These accused shared a common purpose with the crowd to kill the deceased and each of them had the requisite dolus80 in respect of his death. Consequently the acts of the mob which caused the deceased?s death must be imputed to each of these accused. (Botha cited in Unterhalter, 1988, p. 672) Critically, in his ruling, Judge Botha did away with the obligation to prove mens rea or a subjective intent on the part of each of the accused to have killed the deputy mayor. This was simply ?imputed? from the common purpose of the ?mob?. Thus David Unterhalter argued that the judgment could not be legally upheld as a result of its attribution of blame to a ?mob? or group mind: 80 Criminal intent 239 Blameworthiness is a concept attaching to individuals, not abstractions. We judge individual actions. These actions have their origin in the beliefs of and intentions of individuals and cannot be the product of a collective mind or entity. So too we understand responsibility as the respect owed to the sovereignty of the individual?s will and not as undifferentiated collective guilt.? (Unterhalter, 1988, p. 677, own emphasis) If blame attaches to a ?mob?, then, with whom can a person enter into a common purpose? In Safatsa81 reference is made to the ?accused sharing a common purpose to kill the deceased with the mob as a whole??such description presupposes an identity of interest between each member of the ?mob? and the ?mob? itself. But what is this collective persona, ?the mob?? From the point of view of the criminal law, the mob is an irrelevant abstraction? (Unterhalter, 1988, p. 676) While in the Moleleki case blame was not attributed to a ?mob?, nevertheless in the light of the fact that there was so little direct causal evidence, all the five accused who were found guilty, were not found guilty of directly causing the deaths of the youths who were executed, but of making common cause with those who did. The identity of the perpetrators of the Moleleki execution was never revealed by the court proceedings although the testimony of the Moleleki protagonists at the TRC had made explicit their direct involvement in the execution. Nevertheless in the effort to yoke this violence to law, the court was compelled to stretch and make malleable this law, which continually sought, if no longer a direct causal relation then at least a relation of some instrumentality between the actions of the accused and the final result, namely the killing of the ANCYL members. However, the evidence to hand, in a context of exception and ongoing violence, did not lend itself easily to the forms and conventional precepts of law. Ten months later, while the Sharpeville case was still on appeal, another case of appeal for a conviction of murder on the basis of common purpose, S v Mgedzi, came before the courts 81 Safatsa refers to the case of the ?Sharpeville Six? referred to above (S v Safatsa 1988). 240 and in giving judgement, Justice Botha, drawing on principles already identified by a Professor Whiting, sought to restrict the ambit of liability in common purpose murders by elucidating ?certain additional requirements? which had to be proved by the prosecution, in order for a participant in a common purpose leading to the death of another to be guilty of murder of that person: (a) presence at the scene of the violence; namely for every crime there is an actus reus, or the physical act that constitutes the crime, and the mens rea, or the mental element of varying standards that is held by the perpetrator. (b) awareness of the assault. (c) intention to make common cause with those who were actually perpetrating the assault. (d) manifestation of a sharing of a common purpose with the perpetrators of the assault by performing ?some act of association with the conduct of others? and the possession of the requisite mens rea. (Burchell, 1997, p. 132) While, as Unterhalter argues, this account was useful in explicating what South African courts had understood as ?active association?, ?these requirements do not provide a compelling normative basis for attributing the acts of one person to another? (Unterhalter, 1988, p. 675). Applying the Common Purpose Rule to the Moleleki Execution Nevertheless, it was this judgment which was to guide Justice Boruchowitz in his application of the common purpose doctrine to the actions of the Moleleki accused. In contrast to the findings of the TRC that the Moleleki protagonists had acted in spontaneous revenge to the killing of civic activist Bulelwa and SDU leader Blanko and had no foresight in terms of their actions, i.e. a prior political intention in terms of which they carried out their actions, the judgment of the court, in seeking to adduce a criminal intent, found the opposite, namely that SDU members had acted together with significant prior planning and criminal intent. The court found therefore that ?the rounding up and killing of the youths was executed in furtherance of a plan or common purpose by SDU members? (The State v. Michael Sonti, 1999, p. 1768). Critically, the ?reason the state relies heavily on the doctrine of common purpose as enunciated in S v Mgedzi and others (1989 SA 687A at 705F-706B)? was that while ?there is some evidence that certain of the accused contributed causally to the deaths of 241 the deceased? that evidence is insufficient to render all of the accused criminally responsible? (The State v. Michael Sonti, 1999, p. 1772). The reason for the lack of causal evidence was the fact that only two ANCYL members, Vuyani Tshabalala and Albert Mangane, survived the execution and were able to directly testify as to who was involved in the killings. However, the court found that, little weight can be attached to Mangane?s evidence. It was apparent to the court that Mangane suffers from a mental disability. He spoke in a slow stilted fashion and was unable to recall dates, places and times. No psychiatric evidence was led to indicate whether his disability arises from the fact that he was attacked in the incident. (The State v. Michael Sonti, 1999, p. 1862) When Albert was asked questions in relation to his account of the execution, he appeared to be unable to answer any of the questions regarding who had captured him, or whom he had been captured with: ?MR FERREIRA: ?did you recognize any of those persons that had been arrested [i.e. ANCYL members]?? (SILENCE)? (The State v. Michael Sonti, 1999, p. 471), and of the people who captured him: ?Are any of those persons here sir? (SILENCE)? (The State v. Michael Sonti, 1999, p. 471). This led the court to comment, ?I would like it noted for the record, that the witness is silent?we have had long silences concerning these questions that you have put? (The State v. Michael Sonti, 1999, p. 469), and later: ?I would again record that there was, when you asked your question, do you know these people, there was a silence? (The State v. Michael Sonti, 1999, p. 469). The Court also made note of Albert?s physical demeanor during his testimony, as Albert complained of a headache: COURT: Is he in pain your client, Mr Sawyer?...Is he in a position to proceed?Well if he has a headache or can?t concentrate we can perhaps assist him with, does he want to stand down a few moments? Because I see he has got his head back? (The State v. Michael Sonti, 1999, p. 473) Albert himself explained, ?This injury or injuries affected my head or brain?my mind or brain is not functioning well, I have also been chased away from school because my mind is 242 not functioning well? (The State v. Michael Sonti, 1999, p. 475). While he could not remember ?these other things?82 such as the names of people, ?the things which happened to me do not go away from my thoughts? 83 (The State v. Michael Sonti, 1999, p. 477). In the absence of a coherent account from Albert, what the court has is the single evidence of Tshabalala as to who was present in the veld at the time of the shooting. At the time of the shooting matters were chaotic and Tshabalala must have been subject to extreme stress. It stands to reason that he would not have had the best opportunity to make a proper observation. (The State v. Michael Sonti, 1999, p. 1807) Merely recalling the execution caused Vuyani to break down in court: VUYANI TSHABALA: They fired my lord and I also heard the sound of big firearms and there was a big noise. My lord may we stop at this moment my lord? COURT: Yes, I noticed that the witness is extremely upset. COURT: Yes, we will adjourn at this point so that the witness can compose himself?(The State v. Michael Sonti, 1999, p. 220) The court consequently found, in relation to Tshabalala?s evidence, Tshabalala?s evidence is not entirely without blemish. In evaluating his evidence the court is mindful of the fact that Tshabalala was deeply traumatised by the incident, he was seriously injured and was lucky to escape with his life. (The State v. Michael Sonti, 1999, p. 1809) Most importantly, Tshabalala?s evidence with regard to who was involved in taking the group from the shack where the ANCYL members were incarcerated to the veld where they were executed differed from that of another ANCYL member, who was released just before the execution and hid in a toilet where he watched the execution take place. Vuyani also made other apparent errors in identifying exactly who was involved in which events during the lead-up to the execution and mistakenly identified Michael Sonti as carrying a gun, while all other witnesses stated that he had carried a large spear. 82 Hierdie ander dinge kan hy nie onthou nie. 83Die dinge wat aan my gebeur het gaan nie weg van my gedagte af. 243 In the light of this lack of direct causal evidence, the judge relied on the testimony of a number of witnesses whose evidence it used to establish, in terms of the principles outlined by the S v Mgedzi judgment, the presence of the accused at various sites of violence prior to the execution, actions which ?manifested? an active association with a common purpose to kill the youths, awareness that the youths would actually be killed and an intention to make common cause with those who intended to kill the youths. In order to establish these different elements, the court used the evidence of various witnesses to establish the presence of the accused at various sites during the process of the abduction of the youths. Accused 1, Michael Sonti, was implicated by seven witnesses in several of the events preceding the execution, including the stabbing of two people, Alfred Zulu and ANCYL member Jabulani Nxumalo. He was also implicated in the rounding-up of various youths, their ?presentation? at the body of Bulelwa, their incarceration at the first shack where the youths were held and their eventual removal to a second shack from where they were taken into the veld to be killed. However, the court could not establish Sonti?s actual presence at the site of the execution. What it did find is that the ?accused was instrumental in the preparations for the removal of the youths from the first to the second shack? (The State v. Michael Sonti, 1999, p. 1813, own emphasis). However, it was not possible to implicate Sonti directly in the actual removal of the youths from the second shack to the veld. Despite this, the court found that ?accused 1 was seen in the company of the youths shortly prior to their being killed. This suggests that he was present at the killing or handed the youth over to their killers? (The State v. Michael Sonti, 1999, p. 1816). On this basis and on the basis of his identification by witnesses at various sites of violence prior to the execution, the court was able to conclude that ?accused 1 actively and knowingly participated in the plan to kidnap and kill the ANC Youth League members who had been rounded up? (The State v. Michael 244 Sonti, 1999, p. 1815, own emphasis). Mens rea was deduced from the following: ?accused 1 at no stage sought to distance himself from what was happening. He clearly had no concern for the physical welfare of the youths? (The State v. Michael Sonti, 1999, p. 1816). In relation to accused 2, Malcoms Ngama, the court found him guilty of an active association with a common purpose to kill the ANCYL members on the basis of two conversations attested to by witnesses, one outside the first shack where the youths were held, where he was overheard discussing the need to move the youths to conceal their presence from the Internal Stability Unit and a second interaction with Jabulani Nxumalo outside the second shack where he told Jabulani that ?they had been looking for him for a long time and they were going to follow, they referring to the youths who had been apprehended, were going to follow the other people that had already died? (The State v. Michael Sonti, 1999, p.1828). According to the court, evidence that accused 2 had been carrying a particular kind of weapon called a tshomentshos had, ?a ring of truth about it? (The State v. Michael Sonti, 1999, p. 1835) and provided further evidence, for the court, of an association with a common purpose to kill the youths. With regard to accused 3, Michael Armoed, the court found that although the evidence implicating accused 3 in the killing of the deceased is inconclusive the court is satisfied that he actively participated in the plan to round up and kill the youths, he knowingly facilitated the killing of the youths and acted as an accomplice in respect of their murders. (The State v. Michael Sonti, 1999, p. 1855) In relation to accused 6, Michael Nkomo, the court found him guilty of an association with a common purpose to kill the ANCYL members on the basis of evidence that the weapon he carried could have been used in the killing. Although the court can make no finding that accused 6 participated directly in the killing of the youths it is clear from the type of injuries sustained by the various deceased that the type of axe or weapon carried by accused 6 could well have been used. (The State v. Michael Sonti, 1999, p. 1873) With regards to accused 7, the court made a similarly ambiguous finding: 245 Although the state has not proved that accused 7 personally participated in the killing of the youths?the court is satisfied that he knowingly assisted in the rounding up of the youths and facilitating their killing. At minimum he permitted the youths to be handed over to persons who he knew and foresaw would kill the youths. (The State v. Michael Sonti, 1999, p. 1899) Despite the evidence of the ways in which the various accused had allegedly ?manifested? an active association with a common purpose to kill the youths, the court ?could make no finding as to precisely when the plan or common purpose was originated or formed? (The State v. Michael Sonti, 1999, p. 1772), nor was there any ?direct evidence of an agreement between the accused to kidnap and kill the youths? (The State v. Michael Sonti, 1999, p. 1773). Thus this common purpose was deduced from the actions of the accused during the course of events on the night of the killings, rather than direct evidence that such a plan had actually been at any point consciously formulated. In fact there had been ?agreement? between the accused in terms of a direct mandate to firstly abduct the ANCYL members and then later a separate mandate to kill the youths, which was revealed by the SDU members? testimony at the TRC. This notion of a direct mandate was the basis on which the earliest interpretations of the common purpose rule relied. According to evidence led at the TRC, the first direct mandate to abduct the youths responsible for the killing of SDU leader Blanko, came from SDU leader Sugar Ramabele directly after Blanko?s killing. Ramabele, according to testimony at both the TRC and the court proceedings, instructed SDU members that those who were involved in this killing should be ?fetched? (The State v. Michael Sonti, 1999, p. 1819). This order to ?fetch? the youths was separate from a later mandate to kill the youths in terms of an ?order? given by SDU commander Njebe Ndondolo in the context of a large meeting of township residents at Blanko?s home in the early hours of the morning. While the judge was not in possession of information about the second order from Ndondolo, as the accused concealed this information in an attempt to 246 protect themselves from conviction, he rejected the assertion by some of the accused that they had ?rounded up? the youths who were later executed in response to an instruction from Sugar Ramabele, in favour of his hypothesis that a prior plan existed to kill the youths. As in the TRC process, the nature of the authority for the abductions (none of the accused admitted involvement in the killing of the youths) was at issue. Dismissing the assertion by certain of the accused that the abduction of the youths was a spontaneous response to the death of Blanko, Justice Boruchowitz instead found that the ?death of Blanko, in the court?s view, appears to have been a pretext for the rounding up of the youths, and to put the plan [to abduct and kill the youths] into operation? (The State v. Michael Sonti, 1999, p. 1822). He also rejected the assertion by certain of the accused that the ?rounding up? of the ANCYL members had occurred as a result of an order given by SDU commander Sugar Ramabele shortly after Blanko?s death. These accused suggest that the decision to round up the youths was a spontaneous one which arose after, and as a consequence of Blanko?s death. The facts and probabilities indicate otherwise and that the plan or common purpose in fact originated, or was in existence prior to the attack on Blanko. (The State v. Michael Sonti, 1999, p. 1772, own emphasis) The ?facts and probabilities? which ?indicated otherwise? primarily concerned the fact that SDU members could, according to the court, not have known who to ?round up? if they truly sought to find the perpetrators of Blanko?s murder, as they already knew from SDU member Oscar Motlokwa that ANCYL and civic leader Lethusang Rikaba and ANCYL leader Isaac Motloung were directly involved in Blanko?s killing. Of the large crowd who had been present at the time in addition to the two identified perpetrators, which individuals could be legitimately identified and held accountable for his killing? 247 Advocate Ferreira, in cross-examining Michael Sonti, sought to clarify the process through which those who had been captured had been identified as the perpetrators of the killing, but encountered an irresolvable ambiguity in seeking to find an instrumental relation between those who were captured and the killing of Blanko. Michael Sonti instead evoked an amorphous ?community? as referent in both identifying who the perpetrators of Blanko?s killing were as well as the source of authority for the order to capture the youths who were eventually executed. MR FERREIRA: But sir what I don?t understand is, apart from Lethusang Rikaba and Wips, how did the SDUs including yourself, know who to catch? MICHAEL SONTI: Nobody knew who the people were to be arrested. MR FERREIRA: So how could Sugar?s instructions be complied with? MICHAEL SONTI: Oscar had already explained at that stage that there were two people but besides the two people, there were many other people who had attacked Blanko. MR FERRERIA: Am I correct in stating that Oscar stated that it was members of the ANC Youth League? MICHAEL SONTI: I would agree and also not agree that it is so. MR FERRIERA: What do you mean by that sir? MICHAEL SONTI: The reason that makes me to say that, it is because he had explained that there were also adults or elderly people. MR FERREIRA: So am I correct in saying that Oscar stated that it was members of the ANC Youth League as well as older people? MICHAEL SONTI: I stated that it was said even elderly people were there but it was not exactly the word that was used. The word that was used was, the community? (The State v. Michael Sonti, 1999, p. 1356, own emphasis) And later, MR FERREIRA: Sir what I don?t understand is, you must help me is, you have all these different groups going in different directions?And the only information that you have at your disposal is that there was, two of the perpetrators were Rikaba and Wips, is that correct? MICHAEL SONTI: Yes it is so. MR FERREIRA: So would I be correct in stating that then you would have had to apprehend the ANC Youth League members because you only had two names? MICHAEL SONTI: I do not know my lord where we would have apprehended those members of the ANC Youth League because they were being sought after. MR FERREIRA: So I am going to review your answer and you must please try and explain your answer to me because it makes absolutely no sense. MICHAEL SONTI: I do not know where we would have apprehended those members of the Youth League because they were sought after. MR FERREIRA: That is your answer sir. I don?t understand it. 248 MICHAEL SONTI: Let me explain it this way my lord. I did explain my lord that the message came from Sugar Ramabela, he said we should look for the perpetrators of this incident or crime, even though two names of people were given. What I am explaining is that it was not only two people who had attacked Blanko? MR FERREIRA: Yes, yes, before we continue there, let us clear this up first. Because you knew that, because it was known that Wips was a member of the ANC Youth League, and Rikaba as well, it was then decided to apprehend the ANC Youth League members? MICHAEL SONTI: Could you please repeat the question, I have not quite understood it? (The State v. Michael Sonti, 1999, pp. 1358-1359, own emphasis) The state advocate encountered similar ambiguities when he questioned Petros Mtembu about who was to be captured on the basis of Sugar?s order at Blanko?s home: PETROS MTEMBU: It was at that stage when Ramabela said the culprits should be sought after and found. MR FERREIRA: What did you understand by this? PETROS MTEMBU: That they should be sought after and found? MR FERREIRA: When you say ?they? who are you referring to? PETROS MTEMBU: The people who committed the offence. MR FERREIRA: In other words members of the ANC Youth League? PETROS MTEMBU: Even though you say it that way it was not only members of the ANC. MR FERREIRA: Who else was it then? PETROS MTEMBU: It was not only the youth, there were adults as well? MR FERREIRA: Sir I want an answer as to how you knew as which people to go and catch. PETROS MTEMBU: The ones I knew were the ones whose names were mentioned [by Oscar Motlokwa] (The State v. Michael Sonti, 1999, pp. 1534-1535) As the advocate in the court proceedings attempted to interpret the mandate for the ?rounding up? process in juridical terms as directed against members of the ANCYL, Michael Sonti instead evoked the biopolitical fracture, that is substantiated against homo sacer, a subject who is constituted in praxis rather than in terms of a pre-existing juridical identity. MICHAEL SONTI: Here is another thing my lord that I would like to put straight, I would say my lord we were not looking for people but we were looking for criminals, thugs or those who had committed crime. MR FERREIRA: So you don?t consider those persons that committed the crime, the thugs, as people? MICHAEL SONTI: They are people my lord but the whole thing is that what they have done. (The State v. Michael Sonti, 1999, p. 1362, own emphasis) 249 As ambiguous as who should be captured was whom the instruction to go and capture was directed at, a direct juridical mandate to SDU members or, as Michael Sonti attempted to explain, a generalised mandate to members of the ?community? to exercise their authority in terms of a floating imperium in a context of exception: MR FERREIRA: But Sugar was the commander of the self-defence unit, is that correct? MICHAEL SONTI: He was a deputy. MR FERREIRA: But you would agree with me that he was in a position of leadership? MICHAEL SONTI: Yes it is so. MR FERREIRA: And he was one of the leaders of the self-defence units? MICHAEL SONTI: Yes it is so. MR FERREIRA: So is it not correct that the instructions that he was giving, was to the self-defence units? MICHAEL SONTI: I would explain it this way that even though Sugar Ramabela was someone with whom we used to go about with, but at that stage he was explaining to each and everybody who was present at time, at the scene of the incident. MR FERREIRA: Let me ask you this first, what is the function of the self-defence units? MICHAEL SONTI: At that time the duty of the self-defence unit was to protect the community that was not protected. MR FERREIRA: Will I be correct in stating that they were policemen, in inverted commas, of the community? MICHAEL SONTI: Yes I do agree with you. MR FERREIRA: So is it correct then if I state that the instructions were principally or mainly aimed at members of the self-defence unit? MICHAEL SONTI: Let me explain it this way before I answer the question my lord. The members of the self-defence unit were not people who were chosen. Let me put it this way, the whole community in that place or area, was protecting themselves. In other words my lord in short, I am explaining it this way that any person who was a man, was a member of the self-defence unit. (The State v. Michael Sonti, 1999, p. 1356-1357, own emphasis) SDU member Langa Michael Nkomo similarly argued: LANGA NKOMO: Then Sugar Ramabela took out an order that people should go out, to go and look for Lethusang Rikaba and Wips and the group who was with them when they attacked Blanko. MR FERREIRA: Surely this order was given to the SDUs by Sugar Ramabela who was the deputy of the SDUs? LANGA NKOMO: It could be so but he was giving that order to each and every person who was there, even to the community itself. (The State v. Michael Sonti, 1999, p. 1632, own emphasis) 250 The court faced similar difficulties in attempting to distinguish between SDU members and the ?community? in relation to the testimony of Bethwell Ntoma. Speaking of the gathering at Bulelwa?s body, Bethwell?s evocation of the duty of defence as an obligation of all male members of the ?community? rather than a formal mandate as a member of a juridically recognisable organisational formation, provoked incomprehension on the part of the advocate questioning him: MR FERREIRA: So now Tsepo and Njebe have left and you stayed there behind with the community and other SDU members, is that correct? BETHWELL NTOMA: What I explained was, I was there with the community. MR FERREIRA: So you are saying that there were no SDUs there? BETHWELL NTOMO: The SDU member who was present, that I know was myself. MR FERREIRA: Now let us get this straight sir, you say that there were, that you were there with the community? BETHWELL NTOMO: Yes. MR FERREIRA: So apart from yourself, there were no other SDUs? BETHWELL NTOMO: Maybe they were there, let me say they were there. MR FERREIRA: Sir were they there or were they not there, what is it? You have got to choose. BETHWELL NTOMO: Let me please you, I will say they were there. MR FERREIRA: How do you know they were there sir? BETHWELL NTOMO: In short I am going to say any male person who was a resident of Moluleke (sic), was automatically an SDU member. MR FERREIRA: Sir if what you are saying is correct, then why is there the need to distinguish between SDU members and members of the community? BETHWELL NTOMO: There is nowhere where I distinguished. (The State v. Michael Sonti, 1999, pp. 1522-1523) While the court itself in the application of the common purpose doctrine had abandoned the principles of direct causality and instrumentality, it insisted on these in terms of the conduct of SDU members. Thus the assertion by some of the accused that there had been a general instruction by Sugar Ramabele that the persons responsible for Blanko?s death should be ?rounded up? was rejected as ?improbable? on the basis that two individuals who were responsible for Blanko?s death had already been identified by Oscar Motlokwa. How then, the court asked, could the accused know who to apprehend, unless this ?rounding up? process 251 was in fact the result of a prior plan to target ANCYL members specifically? As the court stated in relation to the evidence of accused 1, Michael Sonti: When pressed to explain how it was possible to properly carry out Ramabela?s instruction [to find out who the perpetrators of the killing were] accused 1 conceded that it would not have been possible to have rounded up unidentified persons and that in fact Sugar Ramabela had stated that the ANCYL members had to be rounded up. (The State v. Michael Sonti, 1999, p. 1798) The claim by several of the accused that the youths had been ?rounded up? in order to ?present? them to the ?community? for a decision as to their fate was met with the same incomprehension and dismissal as at the amnesty hearings of the TRC, as neither institution could recognise the biopolitical nature of the authority invoked by this reference to ?the community?. Michael Sonti thus attempted to articulate the authority for his ?fetching? of ANCYL leader Lethusang Rikaba for presentation to the ?community? gathered at the site of Bulelwa?s body, in terms of a community embodied by Sugar Ramabela. The court, however, could not interpret the nature of this auctoritas or biopolitical authority: MICHAEL SONTI: I had already explained the reasons why he [Lethusang Rikaba], why I wanted him or what the reasons were that he was wanted by the community. MR FERREIRA: ?How can you say he was wanted by the community? You have now said it twice, how can you say that? MICHAEL SONTI: I have already explained my lord that everything that happened at the place where we were staying, we were protecting the community. MR FERREIRA: But surely the instruction came from Sugar Ramabela, not the community? MICHAEL SONTI: Yes my lord it is so, but Sugar Ramabela is still part of the community. (The State v. Michael Sonti, 1999, pp. 1364-1365) Nor could it interpret the biopolitical ?presentation? of people to the ?community?, which concerned an embodiment in the presence of this community, rather than a juridical presence that would have involved logos or language: 252 MICHAEL SONTI: ?I showed him [Lethusang Rikaba], or referred him to Sugar Ramabela. After we had taken Mr Rikaba with we met certain other members of the self-defence unit at a corner of block E and D my lord. Those members were going with another boy. (The State v. Michael Sonti, 1999, p. 1369) MR FERREIRA: Sir you say that the SDU were part of the community, you made a definite distinction between the boy and the group of SDU. Now my question is on what basis did you make that distinction? MICHAEL SONTI: I made that distinction between the SDU members and this boy because this boy and Lethusang Rikaba were shown to the community. The self- defence unit are the people who showed them up to the community. MR FERREIRA: Sir can you just repeat that please. MICHAEL SONTI: I say my lord, I say as this man were walking with this boy and me walking with Rikaba we arrived simultaneously at the body of Bulelwa where there were people who were standing guard over the body of Bulelwa. Sugar Ramabela was also present there. I was there when they were shown to the community and Sugar Ramabela? MR FERREIRA: Sir now this just doesn?t make sense if one looks at your evidence?You said you showed, your words were along the line that you showed?Rikaba and this boy were shown to the community, what do you mean by that? (The State v. Michael Sonti, 1999, pp. 1370-1371, own emphasis) Another ambiguity arose in relation to the nature of the instruction that Sugar Ramabela issued to the ?community? gathered at the site of Blanko?s killing, namely, whether it constituted a mandate to kill those who were to be captured. Michael Sonti again evoked the authority of the community in terms of any further decision regarding what was to be done with those who were captured and was again met with disbelief: MR FERREIRA: Did Sugar Ramabela say what was to happen with the perpetrators of those heinous crime, to Bulelwa and Blanko? MICHAEL SONTI: That was left in the hands of the self-defence unit leaders as well as the leaders of the community? (The State v. Michael Sonti, 1999, p. 1415) MR FERRERIA: Sir, what were you supposed to do, once you had apprehended these persons or these killers, what were you supposed to do with them? MICHAEL SONTI: I have already explained that they were brought to the community at the body of Bulelwa. MR FERREIRA: Why did you take them to the body of Bulelwa? MICHAEL SONTI: ?I just want to put it straight my lord, the reason was also that the leaders of the community as well as the leaders of the self-defence unit, should get an opportunity to ask what was happening? (The State v. Michael Sonti, 1999, p. 1416) 253 The fact that the abducted youth were never handed over to the police or any community structure that the court recognised, and were instead deliberately concealed from the police when they came to remove the body of Bulelwa, provided further evidence for the court that this reference to ?community? was a spurious attempt by the accused to avoid taking responsibility for a prior decision to kill the youths. Michael Sonti gave a familiar explanation regarding lack of trust in the police as the reason the youth were not handed over to them, MR LUVUNU: Why were the boys not given to the Internal Stability Unit? MICHAEL SONTI: My lord I would say according to my knowledge at that stage in the whole of Kalerus (sic) the whole of the stability unit was not trusted (The State v. Michael Sonti, 1999, p. 1434). The court found that, while some of the accused contend that the youths were unlawfully apprehended because they were responsible for the deaths of Blanko and Bulelwa, none of the youths were handed over to the police to community structures to enable the law to take its course. The fact that the youths were not handed over to the authorities, and that their deaths ensued shortly after their apprehension gives rise to the inference that the intention of those who had apprehended them was to kill them. (The State v. Michael Sonti, 1999, pp. 1771-1772, own emphasis) However, in the absence of any direct evidence of an agreement between the accused to kill the youths, the court inferred this plan or common purpose from an eclectic mix of factors, which spoke to the difficulties of deducing such a plan from an array of actions and activities committed during the chaotic context of events which unfolded during the night following Bulelwa?s murder. Parker warns of the ?danger? in such circumstances of ?muddying? the distinction between the ?hindsight enjoyed by the judge and the foresight available to the accused? (Parker, 1996, p. 86), which provides the requisite mens rea. Thus ?those in a large group rarely know what others are doing or want to happen. In retrospect it is easy for a court 254 so minded to declare that everyone knew that what did happen was going to happen? (Parker, 1996, p. 86). In the Moleleki case, the judge ?inferred? the existence of a common purpose to kill from several factors. One was that all those killed were ANCYL members. Another was that ?most of the youths were treated in the same way is indicative of a plan or common purpose? (The State v. Michael Sonti, 1999, p. 1770). The judge also argued that in the light of evidence from both sets of protagonists in the conflict in Moleleki that each had been informed of the possibility of an imminent attack on them by the other party, immediately prior to abduction and killing of ANCYL members, that the SDU ?would not have acted supinely but would have embarked on some action against ANCYL members to thwart an attack? (The State v. Michael Sonti, 1999, p. 1822). In fact it was ANCYL members who first launched an attack on the SDU by killing SDU leader Blanko in response to the death of civic activist Bulelwa. Civic and ANCYL leader Lethusang Rikaba acknowledged his personal involvement in this killing to the TRC but later invoked his privilege not to disclose the details of his involvement in the events surrounding Blanko?s death in court, where he appeared as a state witness. As the court explained, According to the witness, Lethusang Rikaba, a meeting of the members of the community was held after the death of Bulelwa when a decision was taken to, as he termed it, ?hunt down? the members of the SDU which operated in Block E. Between the hours of 02:00 to 03:00 on 7 December 1993 an attack was launched on Malusi Jackson Keyana, known as Blanko. Blanko was a leader of the SDUs. He was shot at his house in Block E and his house burnt and razed to the ground. The attack on Blanko was followed by the abduction of ANC Youth League members [by SDU members]. (The State v. Michael Sonti, 1999, pp. 1766-1767) Subsequently, During cross-examination, Rikaba was asked questions concerning the killing of Blanko but declined to answer same on the ground that his answers were privileged 255 and might incriminate him. After argument I upheld the privilege (The State v. Michael Sonti, 1999, p. 1779) In Western jurisprudence it would usually be required that the basis of liability of the accused would be decided individually and hence such evidence of planned attacks on the ANCYL and the SDU as corporate entities would not be relevant. However the common purpose rule and the way in which it had been developed in the South African context, allowed the judge to pursue a corporate basis of responsibility and ironically in doing this he implicitly recognised the biopolitical nature of the power which emerges in a context of exception. The other factors from which the court deduced the existence of a ?common purpose? primarily involved the ?means? through which the crimes were allegedly committed, namely evidence of a ?list? allegedly identifying who was to be eliminated, the fact that the ?rounding up process was completed speedily and efficiently? and ?all of the abductees were taken to a predetermined place? and the possession of ?heavy duty? rope that was later found at the site of the execution and ?could not have been obtained in the early hours of 7 December 1993? (The State v. Michael Sonti, 1999, pp. 1769-1770). In addition there was evidence that ?those who participated in the rounding up process carried an array of weapons? and ?the undisputed medical evidence? showed that the wounds sustained by the deceased were ?consistent with the use of such weapons? (The State v. Michael Sonti, 1999, p. 1770). However, evidence of these various means, whether a rope, a list, or the possession of a weapon, is not adequate in and of itself to prove the existence of a common purpose to use these various means to commit a killing, nor does it provide evidence that every person in possession of a rope, a list or even a weapon, was party to or aware of a common purpose to 256 kill the abducted youths. Thus, SDU members, who were usually armed, and would have been even more likely to be armed in the light of the events which unfolded after Bulelwa?s death, could not automatically be deduced to be carrying these weapons with a specific intent to kill the youths. In fact the evidence led at the TRC, that SDU members carried out the killings in response to a second order from SDU commander Ndondolo, indicates that their original intent, based on Sugar Ramabele?s initial orders, was indeed to ?round up? and capture the youths they believed might be involved in Blanko?s death. However, it is highly likely that this initial ?intent? changed during the course of the night in response to the escalating conflict, finally culminating in the large emotionally charged gathering in Blanko?s yard during the early hours of the meeting, which was the context in which the ?order? for the killing of the captured youth was given by Ndondolo. In the court?s endeavour to deduce a single, uninterrupted, common intent among several people over period of more than 24 hours, what was exposed was the paradox of attempting to impose a critical precept of Western jurisprudence, namely mens rea as a reified construct, a manifestation of subjectivity fixed prior to action and unchanged over time, in order to create the ?myth? of sustained intentionalities that are so important to maintain law?s relation to life, through the construction of a fictive relation between a prior origin of power (intention) and its violent effects. The weakness of the hold of law on life was evidenced by the judge?s reliance as evidence of intention on a series of threatening remarks made by some SDU members to some ANCYL members, which ostensibly demonstrated that all SDU members foresaw that the youth would die and therefore possessed the requisite mens rea in terms of a common purpose to kill the youths. These included statements by different SDU members that the youths should 257 ?smoke their last cigarette?, that ?their mothers should buy coffins? and that they were ?going to follow the other people that died? (The State v. Michael Sonti, 1999, p. 1771). Other factors which the court proposed as indicating that there was a plan or common purpose to kill the youths were a variety of incidents of violence that occurred during the ?rounding up? process, such as the stabbing and shooting of Alfred Zulu, who had attempted to escape after being abducted, the drowning of ANCYL leader Isaac Motloung and the stabbing of ANCYL member Jabulani Nxumalo. However, the killing of Alfred Zulu, while certainly providing evidence of a direct intent to kill on the part of SDU member Oscar Motlokwa, who shot dead Alfred at point blank range, did not provide evidence that Oscar intended to kill the other youths who were abducted in the same way, or that there was a common purpose among other members of the SDU who had abducted youths after Blanko?s death, to kill them in a similar fashion. No evidence was led with regard to a similar intent to kill ANCYL leader Isaac Motloung. Isaac, nicknamed ?Wips?, was abducted separately by SDU leader ?Manyala? and later found drowned in a stream. Despite the lack of any direct evidence implicating any of the accused in Isaac Motloung?s death, the court found that his death formed part of the ?common purpose? to abduct and kill the deceased youth by the SDU members. The court found that all the SDU members, barring those whom it acquitted, were guilty of being part of a common purpose to kill Isaac Motloung. With regard to accused 1, Michael Sonti, the court found that while Isaac had been apprehended separately, there are sufficient indications on the evidence that Wips? death fell within the ambit of the plan or common purpose to round up and kill the youths. Although accused 1 has not been found to have directly participated in his killing, or to be directly causally connected thereto, he nonetheless, on his own evidence, intended to apprehend Wips, who was the leader of the ANCYL and associated himself actively 258 in the plan to round him up. (The State v. Michael Sonti, 1999, p. 1824, own emphasis) Wips and Lethusang Rikaba were both identified by Oscar Motlokwa, Blanko?s neighbour and fellow SDU member, as being directly involved in the killing of Blanko, and he shared this information with other SDU members at the site of Blanko?s killing, so there is no doubt that when Sugar Ramabele issued an ?order? that ?the people? responsible for Blanko?s death should be ?fetched?, it would indeed be the ?intent? of SDU members to ?apprehend? Wips. However, in the absence of any causal evidence whatsoever, linking the accused to Wips? death, either in terms of actions, manifest intent, or even identificatory evidence placing them at or near the scene of his death, the already enormously flexible basis for liability implicit in the common purpose rule was stretched beyond credulity thus exposing the fallacy of attempting to incorporate within the law the anomic violence of the exception, which spoke not to a ?common purpose?, a juridical intent that could located in time and space, but a manifestation of biopolitical power that blurs praxis and logos. Thus what was produced by the adjudication of the Moleleki execution within the context of the courtroom was another problem of comprehension, a radical problem of understanding, as the court attempted to incorporate within law the actions of the Moleleki protagonists in terms of a notion of a prior collective intent. The SDU members instead invoked the ?spontaneity? of their actions in the context of the anger of the ?community?. As during the TRC process, the court struggled to understand the nature of this ?community? and the ways in which it authorised action. These explanations, which evoked the community as the source of authority for actions, were incomprehensible to the court and were rejected as cynical falsifications, intended to obfuscate the criminal intent of the protagonists. While the conviction of some of the SDU members involved in the execution was intended to establish 259 a new norm of law, in fact what it exposed was the irrevocable ambiguity of human action in a context of exception, whose nature, is fundamentally irresolvable. 260 CHAPTER EIGHT: CONCLUSION, THE EXCEPTION, A RADICAL PROBLEM OF UNDERSTANDING This thesis began by presenting the Moleleki execution as a radical problem of understanding. How was it possible to make sense of this atrocious act of violence? However, the Moleleki execution was not an isolated incident. During the period leading up to South Africa?s first popular elections, approximately 16 000 people died in ongoing conflict, many in just as horrific circumstances. How, then is it possible to understand the nature of the power that emerged during this period, in particular its ?excess? of violence? This thesis has sought to explore the radical problem of understanding posed by the Moleleki execution, within two historical junctures, the period in which the execution took place during South Africa?s negotiated transition and the post-apartheid period where the execution became the subject of significant efforts to inscribe its violence within the law. What has emerged through these investigations is that the radical problem of understanding and the mutability of the ?truth? that was produced around the execution critically concerned the ambiguity of the exception in which the execution took place and the misrecognition of this condition of exception, both in the pre- and post-apartheid contexts, as a juridical problematic, in which violence was conceived of as a implacable ?exception? to the proper functioning of juridical power, a consequence of a lacuna in law, a breakdown or distortion of a normatively conceptualised sovereignty. In fact the exception concerned something entirely different, namely the boundaries of the juridico-political order, an ambiguous site of articulation between juridico-political and biopolitical power, law and life ? the ?hidden nucleus? of sovereign power (Agamben, 1998, 261 p. 6). The sovereign decision on the exception is the original political structure in terms of which what is inside and what is outside the juridico-political order acquire their meaning. Therefore what is at issue in the sovereign exception is the very condition of possibility of juridical order, its initial conditions of constitution. Thus while it was assumed that what was being witnessed during the period of negotiations was a breakdown or distortion of state sovereignty in a context of transition, in fact what was revealed was an exposure of sovereignty at the boundaries of the juridico-political order as the form of the state crumbled and the juridical order was essentially suspended during the negotiation process. Therefore the violence of the negotiation period was not a ?paradox?, an extraordinary exception to the proper functioning of juridical order but instead concerned the initial conditions of constitution of this order in the transition from authoritarian to democratic governance. As Schmitt argues, ?the exception appears in its absolute form when it is a question of creating a situation in which juridical rules can be valid?there is no rule applicable to chaos. Order must be established for juridical order to make sense? (Schmitt cited in Agamben, 2005, p. 16). It was in the course of exactly such an originary struggle for sovereign exception in a context of national exception that the Moleleki execution took place. Invoked in this struggle was the principle of the exception that opens up the space in which juridical order can be applied, a principle that would create the conditions of possibility for the post-apartheid juridical order. However, the boundary of sovereign exception which would open up the space for juridical order proved elusive, permeable, and irrevocably ambiguous. The originary struggle for juridical order between the SDU and ANCYL in Moleleki as the blocks of the new settlement 262 were being laid out took place in a context of anomic violence, unleashed on the East Rand by the national context of exception. In this context, the struggle between the SDU and the ANCYL sought a violent tying of juridical order (Ordnung) to territory to expiate the ambiguity of the exception, and irrevocably define the boundaries of juridical order in the township. This attempt to localise what was essentially unlocalisable, namely the exception, and grant it a visible and permanent expression, led to the atrocity of the execution in the violent localisation of exception in the bodies of the Moleleki dead. The radical problem of understanding faced by analysts and the juridico-political institutions of the post-apartheid state in attempting to interpret the nature of the power that emerged in this context of exception, thus concerned not only a juridical conception of power but the nature of the exception itself, which ?defies general codification? (Schmitt cited in Agamben, 2005, p. 15), and consists essentially in a lacuna, a juridical void created by the suspension of the juridical order. How then could this empty space be interpreted and the actions of the protagonists of the Moleleki execution, which took place in this context of exception, be adjudicated? The ambiguity of the exception was manifested in the multiplication of explanations of the conflict during the negotiation period, where analysts working within a juridical conception of power assumed that a causal relation could be traced between an ?original? site of power and its violent effects. However, the multiplication of explanations of the conflict which this line of enquiry produced, spoke to the aporia of this line of investigation, which was merely able to describe the fleeting articulations of power in terms of a particular manifestation of ethnicity, generational cleavage or economic division. It could not elucidate the why of power, the articulation of juridico-political and biopolitical power at the boundaries of the 263 juridicio-political order. Thus analysts working within a long philisophico-political tradition, in which violence was conceived as a transparent instrument that could be put to the service of political ends, assumed that the basis of the sovereign state?s authority was its ability to monopolise the means of violence. They could not separate out sovereignty from the state form and its putative ability to monopolise the means of violence. The eruption of sovereign power at the boundaries of the juridico-political order, the functions and purposes of which could not easily be ascertained or linked to an ?intelligible public cause? (A. du Toit, 1993, p. 6), therefore defied comprehension in juridico-political terms, constituting a radical problem of understanding for analysts at the time. Critically the violence that emerges in this context does not concern an instrumental relation between a juridical origin of power and its violent effects, a separation of the origin and effects of power, which is in fact, as Nietzsche argued, a mythological structure, but concerns what Benjamin has called a ?manifestation of violence? (Benjamin cited in Agamben, 2005, p. 62), violence which is not a means to an end but stands in relation to its own mediality. The mythic nature of the relation between the origin and effect of violence is made particularly explicit in a context of exception, where the norm of law is suspended and what is exposed is the ?force of law?, which floats as an indeterminate element that can be claimed by all protagonists, either the state or its opposition. Importantly, this force of law exists in non-relation to law, it is not a means, whether legitimate or illegitimate, to an end, whether just or unjust, it is instead a ?pure? anomie (Benjamin cited in Agamben, 2005, p. 61). It is the struggle to capture this anomie, to make it the ?thing? of politics, that is, as Agamben argues, the ultimate stakes for Western politics (Agamben, 2005, pp. 59-60). 264 These struggles to capture the ?pure violence? of the exception in the efforts to inscribe the Moleleki execution within the juridico-institutional contexts of the TRC and the courtroom, exposed the complexity of this capturing of pure violence, which denoted, not a straightforward process of juridical classification and adjudication as these institutions assumed, but the uncertainty of the exception which articulates a relation between law and life. Thus both the TRC and courtroom grappled with the ambiguity of human action in a context of exception, whose nature is unclear and which evokes the impossibility of defining the legal consequences of acts committed during the exception. The nature of the authority which emerges in this context does not concern the authority invested through representative office, as the post-apartheid institutions which adjudicated the execution assumed, but a floating imperium, which is invested in all citizens regardless of rank, a biopolitical power embodied in the person of the actor, which resists ?definition within the terms of the normal order? (Agamben, 2005, p. 43). It therefore cannot subsequently be adjudicated in terms of the ?proper? exercise of juridical authority mandated by legislative power. Nevertheless the post-apartheid juridical institutions sought to understand the nature of the actions relating to the Moleleki execution in terms of a notion of juridical intent, a prior origin of power, which could purportedly be arbitrated in instrumental relation to the violent effect, through an adjudication of the subjectivity of the Moleleki protagonists, either a political intent in the context of the TRC or a criminal intent within the courtroom processes. In the multiple problems of decidability which both these institutions encountered during their adjudication of the execution, the ambiguity of the exception and the attempt to inscribe it within the juridical political order was exposed, reproducing, in the post-apartheid context, the radical problem of understanding produced by a purely juridical conception of power. 265 These processes of re-inscription therefore traversed both the period of juridico-political exception in which the execution took place and the period subsequent to the inauguration of the post-apartheid state and shared, paradoxically, an essential contiguity in their efforts to recapture this ?pure? violence leashed from law, within the law again. However, while the pre- and post-apartheid processes of inscription around the execution shared an essential contiguity in their efforts to capture this violence as the object of the juridico-political, the process of inscription in the post-apartheid period was critically concerned with the establishment of the sovereignty of the new state in terms of a new norm of law, the norm of ?human rights?, which sought define the boundaries of the juridico- political present, in terms of the ?exception? of the violence of the past. Paradoxically, however, despite the explicit intention to constitute the present ?political? as the juridico-political in opposition to the biopolitics of the past, the norm of human rights in fact introduces a critical biopolitical principle into the functioning of modern nation states. The Declaration of the Rights of Man was the instrument of transformation from the dynastic system in which sovereignty was vested in the prince to the modern nation state in which sovereignty is vested in the ?people? who are protected by a set of inalienable rights. However, while purportedly making birth the only criteria for the protection of the regime of rights, what remained implicit was that political rights or the rights of ?active citizenship? would only be allowed to certain people, specifically ?excluding (among others), women, children, foreigners, the insane, minors and those condemned to punishment? (Agamben, 1998, p. 130). 266 These people would have no influence on public matters and would not be considered citizens. Therefore the ?Rights of Man? emerge as a biopolitical status that is acquired within the context of the political community. Thus the shifting of sovereignty from the prince to the people creates an essential ambiguity, how to determine the limits of the political community? This has led to a constant need to define the boundaries of the political community in terms of a distinction between ?authentic life and life lacking every political value? (Agamben, 1998, p. 132). Therefore, while it was assumed that the transfer of sovereignty to the people through the country?s first democratic elections was a purely juridical phenomenon, which would establish a juridical political in opposition to the biopolitics of the past, what remained implicit was the biopolitical nature of the relation of exception that would constitute this present political, through ?the biopolitical fracture in terms of which, the human species [is constituted] into a body politic?, through a fundamental split, ?which defines the boundary of the political community in terms of a distinction between ?bare life? (people) and political life (People)? (Agamben, 2000, p. 30.1-2, original emphasis). Critically, however, this split is profoundly ambiguous, and has to constantly redefine itself in terms of a series of exclusions which demarcate its boundaries. The TRC was established to respond to this ?highest political task?(Agamben, 1998, p. 130), namely to articulate the biopolitical fracture and delineate the People in whom sovereignty would be vested in order to establish a post-apartheid ?nationhood?. However, the TRC misrecognised the nature of the exception it was established to articulate, assuming that it concerned the instantiation and arbitration of an instrumental relation with the biopolitical past. This would occur through a legislative process that would adjudicate this relation by 267 granting amnesty for those who had committed acts in pursuit of ?political objectives? on all sides of the political conflict as well as administratively deciding who had been the victims of such gross human rights violations. In this adjudication, the relationship between the juridical origin of power and its violent effects could putatively be sundered, making an exception of the biopolitics of the past. However, the ambiguities of this endeavour, which could not recognise the biopolitical nature of the relation that the Commission was implicitly established to articulate, created multiple problems of decidability in attempting to determine an instrumental relation between a prior source of power and its violent effects in order to render a political present free from the stain of violence. This finally led the Commission to acknowledge the difficulties of such a juridical classification of the political when it stated in its report that ?the political nature of specific acts was hard to define? (TRC, 1998b, p. 82). The limits of this juridical conception of the political were particularly exposed by actions like the Moleleki execution, which had taken place in a context of exception. Thus, the protagonists of the Moleleki execution implicitly articulated the nature of the conflict in Moleleki in terms of the biopolitical fracture, in which the boundaries of the political community are substantiated against the originary figure of the political relation, homo sacer, one who has been judged on account of a crime and who stands in a relation of exception to the juridico-political order. However this invocation by the Moleleki applicants of the ?criminal? rather than the juridico-political opponent, and the ?community? rather than a juridical mandate as the source of the authority for their actions, led to the TRC to conclude that the execution had not been politically motivated and therefore was not political in nature. 268 In contrast to the TRC, which had found a lack of prior political intent in a coherently articulated political motive, the adjudication of the Moleleki execution in the context of the courtroom found significant prior intention in a common plan or purpose to commit the execution among the Moleleki protagonists, which spoke to the ambiguity of the adjudication of actions committed during the exception and the mutability of the ?truth? that such adjudications produce, when it concerns actions that occurred in a juridical void at the boundaries of the juridico-political order. The complete abandonment of the principle of causality in the court proceedings where the judge used the doctrine of common purpose to adjudicate the actions of the Moleleki protagonists, therefore spoke to the ?impossibility? of adjudicating actions which have occurred in a context of exception in juridical terms, and of attaching legal consequences to actions which had occurred outside law, as a manifestation rather than an effect of a prior source of power. Thus while five of the Moleleki accused were sentenced to life terms of imprisonment as a result of the court trial, in using the doctrine of common purpose the judge implicitly acknowledged the difficulty of welding life to law, of incorporating the political in law. However, bound by the aporia of a juridical conception of power, which could not conceive of a realm outside law, the Moleleki execution was left to remain, essentially, a radical problem of understanding. Thus, what this thesis has to conclude as a result of its interrogation of the Moleleki execution as a moment of sovereign decision that evoked the struggle to create the conditions of possibility of post-apartheid juridical order, is that contrary to popular wisdom, the sovereignty of the post-apartheid state was not established merely through the exercise of juridical ?choice? at the ballot box, but was in fact was a much more fundamental process, which concerned the sovereign decision that articulates the boundaries of the juridico- 269 political order, in terms of a set of biopolitical inclusions and exclusions. However, as the processes of inscription around the Moleleki execution made clear, what this thesis also has to record is that, despite the exposure of the ?hidden nucleus? of sovereign power during the founding moments of South Africa?s democracy, this exposure of sovereignty would remain fundamentally misrecognised, dismissed as an ?archaic aberration? which, it was assumed, would not recur in a post-apartheid juridico-political order, putatively freed from the stain of violence. These obfuscations of the source of power of the post-apartheid state and the basis of its sovereignty in a domain of the ?political? that is not usurped by the realm of law have reproduced a dangerous blindness to these initial conditions of constitution of the current juridical order. In this conception, the violence of sovereign decision and the realm of exception - Hobbes? ?state of nature?, in which man is wolf to men - slips below visibility under the form of the post-apartheid state, to lurk ominously unnoticed, waiting to be exposed when the forms of this state decay. 270 REFERENCES African National Congress Statement to the Truth and Reconciliation Commission. Further Submissions and Responses by the African National Congress to Questions Raised by the Commission for Truth and Reconciliation. (1997). Retrieved from http://www.doj.gov.za/trc/submit/anc2.htm African National Congress Political Party Recall. (1997). Cape Town, 12-13 May, Hearing Transcript, Retrieved from www.doj.gov.za/trc/special/party2/anc2.htm, Truth and Reconciliation Commission of South Africa. Agamben, G. (1998). Homo Sacer: Sovereign power and bare life (D. Heller-Roazen, Trans.). California: Stanford University Press. Agamben, G. (2000). Means without end: Notes on politics (V. Binetti & C. Casarino, Trans.). Minneapolis: University of Minnesota Press. Agamben, G. (2005). State of exception. Chicago: The University of Chicago Press. Aitchison, J. (1988). Numbering the dead: Patterns in the Midlands violence. Pietermaritzburg: Centre for Adult Education, University of Natal. Aitchison, J. (2003). KwaZulu-Natal: The pre-election wars of the 1990. In R. Greenstein (Ed.), The role of political violence in South Africa?s democratisation. Johannesburg: Community Agency for Social Enquiry. Anderson, B. (1991). Imagined communities: Reflections on the origin and spread of nationalism (rev. ed.). London: Verso. Apter, D. E. (Ed.). (1997). The legimization of violence. New York: New York University Press. Asmal, K., Asmal, L. & Suresh, R. (1996). Reconciliation through truth: A reckoning of apartheid?s criminal governance. Cape Town: David Philip. Ball, J. (1994). The ritual of the necklace. Johannesburg: Centre for the Study of Violence and Reconciliation. Bantu Homelands Citizenship Act (National States Citizenship Act) No 26 (1970). Retrieved from South African History Online. http://www.sahistory .org.za/pages/chronology/special-chrono/governance/apartheid-legislation.html Bassiouni, M. C. (1974). International extradition and world public order. New York: Oceana Publications. Beinert, W. (1992). Introduction: Political and Collective Violence in Southern African Historiography. Journal of Southern African Studies, 18(3), 445-477. 271 Bhargava, A. (2002). Defining Political Crimes: A Case Study of the South African TRC. Columbia Law Review, 102(5), 1304-1339. Black Communities Development Act 4 of 1984. Retrieved from Sabinet Law. http://0-search .sabinet.co.za.innopac.wits.ac.za/netlawpdf/netlaw/BLACK%20COMMUNITIES%2 0DEVELOPMENT%20ACT,%201984.htm Bonner, P. & Ndima, V. (1999). The roots of violence on the East Rand, 1980-1990. Seminar Paper. Johannesburg: University of the Witwatersrand Institute for Advanced Social Research. Bonner, P. & Nieftagodien, N. (2001). Kathorus: A history. Cape Town: Maskew Miller Longman (Pty) Ltd. Bonner, P. & Nieftagodien, N. (2002). The TRC and the pursuit of ?social truth?: The case of Kathorus. In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC (pp. 173-203). Johannesburg: Witwatersrand University Press. Boraine, A. (2000). A country unmasked: Inside South Africa?s TRC. Oxford: Oxford University Press. Bozzoli, B. (1998). Public Ritual and Private Transition: The Truth Commission in Alexandra Township South Africa 1996. African Studies, 57 (2), 167-196. Brewer, J. D. (1994). Introduction: South Africa in transition. In J. D. Brewer (Ed.), Restructuring South Africa. Belfast: St Martin?s Press. Bruce, D. & Komane, J. (1999). Taxis, Cops and Vigilantes: Police attitudes towards street justice. Crime and Conflict, 17(Spring 1999), 39-44. Burchell, J. (1997). Joint Enterprise and Common Purpose: Perspectives in English and South African Criminal Law. South African Journal of Criminal Justice, 10(2), 125-140. Burman, B. & Sch?rf, W. (1990). Creating People?s Justice: Street Committees and People?s Courts in a South African City. Law and Society Review, 24(3), 693-744. Buur, L. (2002). Monumental historical memory: Managing truth in the everyday work of the South African TRC. In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC. (pp. 63-93). Johannesburg: Witwatersrand University Press. Buur, L. (2005). The sovereign outsourced: Local justice and violence in Port Elizabeth. In T. Blom Hansen & F. Stepputat (Eds.), Sovereign bodies: Citizens, migrants and states in the postcolonial world. Princeton and Oxfordshire: Princeton University Press. Buur, L. (2006). Democracy and its discontent: Vigilantism, sovereignty and human rights in South Africa. Paper presented at the Wits Institute of Social and Economic Research, April 24 2006. 272 Carton, B. (2000). Blood from your children: The colonial origins of generational conflict in Southern Africa. Charlottesville and London: University of Virginia Press. Chaskalson, M. & Seekings, J. (1988). The Challenge. From Protest to People?s Power. In Indicator SA (Issue Focus, 40). Chipkin, I. (2007). Do South Africans exist? Johannesburg: Wits University Press. Coleman, M. (Ed.). (1998). A crime against humanity: Analysing the repression of the apartheid state. Cape Town: New Africa Books, David Philip. Comaroff, J. L. & Comaroff, J. (1997). Of revelation and revolution. Volume two: The dialectics of modernity on a South African frontier. Chicago & London: The University of Chicago Press. Comaroff, J. & Comaroff, J. (2004). Criminal Obsessions, after Foucault: Postcoloniality, Policing and the Metaphysics of Disorder. Critical Inquiry, 30(4), 800-824. Crais, C. (2002). The politics of evil magic, state power and the political imagination in South Africa. Cambridge: Cambridge University Press. Cronin, J. (1991). For the sake of our lives: Guidelines for the creation of people?s self- defence units. Johannesburg: Centre for the Study of Violence and Reconciliation. Davis, D. (1990). Capital punishment and the politics of common purpose. In D. Hansson & D. van Zyl Smit (Eds.), Towards justice? Crime and state control in South Africa. Cape Town: Oxford University Press. Du Toit. A. (1990). Discoures on political violence. In Chabani Manganyi, N and Du Toit, A (Eds). Political violence and the struggle in South Africa. Basingstoke: Macmillan. Du Toit, A. (1993). Understanding South African political violence: A new problematic? Geneva Switzerland: United Nations Research Institute for Social Development. Du Toit, P. (2001). South Africa?s brittle peace: the problem of post-settlement violence. New York: Palgrave. Dugard, J. (2003a). Low-intensity conflict. In R. Greenstein (Ed.), The role of political violence in South Africa?s democratisation (pp. 1-21). Johannesburg: Community Agency for Social Enquiry. Dugard, J. (2003b). South Africa?s internal low-intensity conflict. In R. Greenstein (Ed.), The role of political violence in South Africa?s democratisation (pp. 22-46). Johannesburg: Community Agency for Social Enquiry. Dwane-Alpman, P. (1998). Moleleki Massacre Investigative Report. Investigation Unit. Johannesburg: Truth and Reconciliation Commission of South Africa. Emmet, T & Butchart, A., (2000). Behind the mask. Getting to grips with crime and violence in South Africa. Pretoria: HSRC Press. 273 Everatt, D. (2003). Analysing political violence on the Reef, 1990-1994. In R. Greenstein (Ed.), The role of political violence in South Africa?s democratisation (pp. 95-142). Johannesburg: Community Agency for Social Enquiry. Everatt, D. & Sadek, S. (1992). The Reef violence: tribal war or total strategy? (No.1). Johannesburg: Community Agency for Social Enquiry. Feldman, A. (1991). Formations of violence: the narrative of the body and political terror in Northern Ireland. Chicago & London: University of Chicago Press. Foster, D., Haupt, P. & De Beer, M. (Eds.). (2005). The theatre of violence: Narratives of protagonists in the South African conflict. Cape Town: Institute of Justice and Reconciliation. Foucault, M. (1971 original, published in Rabinow 1991). Nietzsche, Genealogy, History. In P. Rabinow (Ed.), The Foucault reader: An introduction to Foucault?s thought. Penguin Social Sciences. New York & Canada: Penguin Books. Foucault, M. (1976). ?Society Must Be Defended? Lectures at the College De France, 1975- 1976 (D. Macey, Trans. First Edition: January 2003 ed.). New York: Picador. Foucault, M. (1976 original, published in Gordon 1980). Two Lectures. In C. Gordon (Ed.), Power/Knowledge: Selected interviews and other writings 1972-1977. New York: Pantheon, pp. 78-108. Foucault, M. (1977). Discipline and punish: The birth of the prison (A. Sheridan, Trans.). London: Allen Lane. Foucault, M. (1977a). Truth and power. In P. Rabinow (Ed.), The Foucault reader: An introduction to Foucault?s thought. Penguin Social Sciences. New York & Canada: Penguin Books. Friedman, S. (1993). Slow walk to freedom. In S. Friedman (Ed.), The long journey: South Africa?s quest for a negotiated settlement (pp. 171-193). Johannesburg: Ravan Press. Geschiere, P. (1997). The modernity of witchcraft: Politics and the occult in postcolonial Africa. (P. Geschiere & J. Roitman, Trans.) Charlottesville: University of Virginia Press. Giliomee, H., & Schlemmer, L., (Eds.). (1994).The bold experiment: South Africa?s new democracy. Halfway House: Southern Publishers. Gordon, C. (Ed.). (1980). Power/Knowledge: Selected interviews and other writings 1972- 1977. New York: Pantheon. Gotz, G. (1995). Violence in the Vaal. Unpublished Draft, January 1995 (pp. 1-26). Graybill, L. (2002). Truth and Reconciliation in South Africa: Miracle or model? Boulder Co: Lynne Rienner Publishers. 274 Guelke, A. (1999). South Africa in Transition: The misunderstood miracle. London & New York: I.B. Tauris. Hansen, T. B. & Stepputat, F. (Eds.). (2005). Sovereign bodies: Citizens, migrants, and states in the postcolonial world. Princeton and Oxfordshire: Princeton University Press. Hanssen, B. (2000). Critique of violence: Between poststructuralism and critical theory. London & New York: Routledge. Hayner, P. (2001). Unspeakable truths: Confronting state terror and atrocity. New York: Routledge. Heribert, A. & Moodley, K. (1992). Political Violence, "Tribalism", and Inkatha. The Journal of Modern African Studies, 30(3), 485-510. Hornberger, J. (2004). ?My Police -Your Police?: The Informal Privatisation of the Police in the Inner City of Johannesburg. African Studies, 63(2), 213-230. Horowitz, D. L. (1991). A democratic South Africa? Constitutional engineering in a divided society. Cape Town: Oxford University Press. Hund, J. & Koto-Rammopo, M. (1983). Justice in a South African Township: The Sociology of Makogtla. Comparative and International Law Journal of Southern Africa, 6, 179- 208. Independent Board of Inquiry & Peace Action. (1993). Before we were friends. Johannesburg: Independent Board of Inquiry. Institute of Criminology. (1991). Violence in South Africa during the first four months of 1991: A Comment from the Institute of Criminology. Cape Town: University of Cape Town. Jeffrey, J. A. (1992). Spotlight on disinformation about violence in South Africa. Johannesburg: South African Institute of Race Relations. Jensen, S. (2003). Security and violence on the frontier of the state (pp. 1-19): Wits Institute for Social and Economic Research. Johnston, A. (1997). Politics and violence in KwaZulu-Natal. In W. Gutteridge & J. E. Spence (Eds.), Violence in Southern Africa. London & Portland: Frank Cass. Kane-Berman, J. (1993). Political violence in South Africa. Johannesburg: South African Institute of Race Relations. Kentridge, M. (1990). An unofficial war: Inside the conflict in Pietermaritzburg. Cape Town & Johannesburg: David Philip. Khosa, M. M. (1991). Routes, Ranks and Rebels: Feuding in the Taxi Revolution. Journal of Southern African Studies, 18(1), 232-251. 275 Killian. (1998). Investigative unit memorandum on the Moleleki Execution. Investigation Unit. Johannesburg: Truth and Reconciliation Commission of South Africa. Kistner, U. (2003). Commissioning and contesting post-apartheid human rights: HIV/AIDS- Racism-Truth and Reconciliation. M?nster, Germany: Lit Verlag. Koch, E. (1992, 11 June 1992). A time when myths can kill. Weekly Mail. Kynoch, G. (2005). Crime, Conflict and Politics in Transition-Era South Africa. African Affairs, 104(416), 493-514. Kynoch, G. (2006). Urban violence in colonial Africa: A case for South African exceptionalism. Seminar Paper, 15 May 2006. Johannesburg: Wits Institute for Social and Economic Research. Levin, N., Ngubeni, K., & Simpson, G. (1994). Meeting the challenge of change? Notes on policing and transition in South Africa (Occasional Paper). Johannesburg: Centre for the Study of Violence and Reconciliation. Mabokela, H. (1993) Police Statement. Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong. Mahlangu, D. (1999, 3 October). Tears still flow for mother of slain and beheaded son. City Press. Mamdani, M. (1998). 'A diminished truth', Siyaya, 3 (Spring). Marinovich, G. & Silva, J. (2001). The Bang Bang Club: Snapshots from a hidden war. United Kingdom: Arrow Books. Markell, P. (2003). Bound by recognition. Princeton: Princeton University Press. Marks, M. (1993). Organisation, identity and violence amongst activist Diepkloof youth, 1984-1993. Unpublished Masters of the Arts Dissertation, University of the Witwatersrand, Johannesburg. Marks, M. (1995). We are fighting for the Liberation of our People: Justifications of Violence by Activist Youth in Diepkloof Soweto. Temps Modernes, 585. Mashabela, H. (1988). Townships of the PWV. Johannesburg: South African Institute of Race Relations. Mbhele, F. (1999, 13 June 1999). Moms? long wait for justice pays off. Sunday World. Mbembe, A. (2003). Necropolitics. Public Culture 15(1), 11- 40. Minnaar, A. (1994). Self Defence Units or "Comtotsi" criminal gangs. Centre for Socio- Political Analysis. Pretoria: Human Sciences Research Council. Minnaar, A., Keith, T & Pretorius, S. (1994). An analysis of massacres in Transvaal and 276 Natal: 1990-1992. Pretoria: Human Science Research Council. Mokoena, M. C. (1995). Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong. Moloi, J. (1995). Police statement, Katlehong MR 144/12/93. Unrest and Violent Crime Unit. Katlehong. Moodley, D.L. (1993) Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong. Mootsi, T. (1997). Statement concerning Gross Violations of Human Rights. Human Rights Violations Committee. Johannesburg: Truth and Reconciliation Committee of South Africa. Morris, M., & Hindson, D. (1992). The Disintegration of Apartheid: From Violence to Reconstruction. South African Review, 6, 152-170. Moser, C.O. & Clark, F.C. (Eds). (2001). Victims, perpetrators or actors? Gender, armed conflict and political violence. London and New York: Zed Books. Motloung, S. (1998). Statement concerning Gross Violations of Human Rights. JB/03300/01ER/KAT. Human Rights Violations Committee. Johannesburg: Truth and Reconciliation Commission of South Africa. Moya, F.N. (1998, 16 Septermber). Katlehong massacre recalled. The Star. Mthembu, M. (1998). Views of Families of Victims of Moleleki Massacre. Investigation Unit. Human Rights Violations Committee. Johannesburg: Truth and Reconciliation Committee of South Africa. Mugadi, E. (1995). Police statement MR 144/12/93. Unrest and Violent Crime Unit. Katlehong. Nathan, L., & Phillips, M. (1992). Cross-Currents: Security Developments under FW de Klerk. In Moss, G. & Obery, I. (Eds.), South African Review 6: From Red Friday to Codesa. Johannesburg: Ravan Press. Ndlozi, G. (1997). Report on Moleleki Investigation. Investigation Unit. Johannesburg: Truth and Reconciliation Commission of South Africa. Ngwenya, J. (1993). Police Statement Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong Nxumalo, J. (1995). Police statement MR 144/12/93. Unrest and Violent Crime Unit. Katlehong. Olivier, J. (1994). Political violence. In J. D. Brewer (Ed.), Restructuring South Africa. Belfast: St Martin?s Press. 277 Own Correspondent. (1990, 1 November). The right to self-defence. Sowetan. Parker, P. (1996). South Africa and the Common Purpose Rule in Crowd Murders. Journal of African Law, 40(1), 78-102. Peace Action. (1993). November and December 1993 Monthly Report. Johannesburg: Peace Action. Perry, G. C. (1988). The Four Major Western Approaches to the Political Offence Exception to Extradition: From Inception to Modern Terrorism. Mercer Law Review, 40(2), 709- 735. Pieterse, J. L. Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong Posel, D. (2002). The TRC report: what kind of history? What kind of truth? In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC. Johannesburg: Witwatersrand University Press. Posel, D. & Simpson, G. (Eds.). (2002). Commissioning the past: Understanding South Africa?s TRC. Johannesburg: Witwatersrand University Press. Posel, D. & Simpson, G. (2002a). The Power of Truth: South Africa?s TRC in Context. In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC. Johannesburg: Witwatersrand University Press. Potgieter, D. W. (1992, 24 May). Rebels on rampage: ANC cadres turn defence unit into murderous gang. Sunday Times. Promotion of National Unity and Reconciliation Act. No 34 of 1995. Retrieved from http://www.doj.gov.za/trc/legal/act9534.htm Pugliese, J. (2002). "Super Visum Corporis": Visuality, Race, Narrativity and the Body of Forensic Pathology. Law and Literature, 14(2), 367-396. Rabinow, P. (Ed.). (1991). The Foucault reader: An introduction to Foucault?s thought. Penguin Social Sciences. New York & Canada: Penguin Books. Rauch, J. & Storey, D. (1998). The policing of public gatherings and demonstrations in South Africa 1960-1994. Johannesburg: Centre for the Study of Violence and Reconciliation. Retrieved from the Centre for the Study of Violence and Reconciliation website at http://www.csvr.org.za/ index.php?option=com_content&task=view&id=689&Itemid=200 Ross, F. C. (2003). Bearing witness: Women and the TRC in South Africa. London, Sterling and Virginia: Pluto Press. Sapire, H. (1992). Politics and Protest in Shack Settlements of the Pretoria-Witwatersrand- Vereeniging Region, South Africa, 1980-1990. Journal of Southern African Studies, 18(3), 670-697. 278 Sarkin, J. (2004). Carrots and sticks: The TRC and the South African amnesty process. Antwerp and Oxford: Intersentia. Sch?rf, W. & Ngcokoto, B. (1990). Images of punishment in the people?s courts of Cape Town 1985-7: From prefigurative justice to populist violence. In N.C. Manganyi. & A. du Toit (Eds.), Political violence and the struggle in South Africa. Basingstoke: Macmillan. Schyff, Baylis, Gericke & Druce. (1988). Town and regional planners. Township application of Moleleki Ext 2. Situated on portion 141 (A portion of the remainder of portion 79). Farm Rietfontein 153 IR. Rivonia. Johannesburg. Schyff, V. d., Baylis, Gericke & Druce. (1991). Town and regional planners. Application in terms of regulation19 (5) of the Township Establishment and Land Use regulations, 1986 for the cancellation and amendment of the General Plan for Moleleki Extension 2. Schlemmer, L. (1991). Dimensions of turmoil: Position paper on current violence in South Africa. Johannesburg: Centre for Policy Studies. Schutte, C. (1998). How many third forces, South Africa? In C. Shutte, I. Liebenberg & A. Minnaar (Eds.), The hidden hand covert operations in South Africa: Revised edition (Second ed.). Pretoria: HSRC Press. Sebolao, P. (1998, 4 March). Truth body told of victims? abduction. Business Day. Segal, L. (1991). The human face of violence: Hostel dwellers speak. Johannesburg: Centre for the Study of Violence and Reconciliation. Shaw, M. (1997). South Africa?s other war: understanding and resolving political violence in KwaZulu Natal and the PWV. Unpublished Master of Arts Dissertation, University of the Witwatersrand, Johannesburg. Shutte, C., Liebenberg, I. & Minnaar, A. (Eds.). (1998). The hidden hand covert operations in South Africa: Second Edition. Pretoria: HSRC Press. Simpson, G. (1993). Explaining Endemic Violence in South Africa (?Gewalt in Sudafrica?). Weltfriedensdienst Quesrbrief, 3. Simpson, G., Mokwena, S., & Segal, L. (1992). Political Violence: 1990. In M. Robertson & A. Rycroft (Eds.), Human Rights and Labour Law Yearbook 1991, 2, pp. 193-219. Cape Town: Oxford University Press. Simpson, G., & Rauch, J. (1993). Political Violence: 1991. In N. Boister & K. Ferguson- Brown (Eds.), Human Rights Yearbook 1992 (First ed.). Cape Town: Oxford University Press. 279 Simpson, G. (2002). "Tell no Lies, claim no easy victories": A brief evaluation of South Africa?s TRC. In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC. Johannesburg: Witwatersrand University Press. Simpson, G. & Van Zyl, P. (1995). South Africa?s TRC Temps Moderns (585), pp. 394-407. Sindane, P. (1999, 10 October). Killers get life sentence, mother pleads for son?s head. Sunday World. Sisilana, L. (1999). What?s wrong with Common Purpose? South African Journal of Criminal Justice, 12(3), 287-316. Sisk, D. T. (1995). Democratization in South Africa: The Elusive Social Contract. New Jersey: Princeton University Press. Sitas, A. (1992). The Making of the "Comrades" movement in Natal 1985-91. Journal of Southern African Studies, 18(3). Sitas, A. (1996). The New Tribalism: Hostels and Violence. Journal of Southern African Studies, 22(2), 235-248. Slabbert, F. (1992). The quest for democracy: South Africa in transition. London: Penguin Books. Sole, S. (1991, 21 April 1991). Street Wars. Sunday Tribune. Tau, F. (1993) Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong. Taylor, R. & Shaw, M. (1998). The dying days of apartheid. In D. R. Howarth & A. J. Norval (Eds.), South Africa in transition (pp. 13-30). London: Macmillan. The State v. Michael Sonti. (1996). MAS 13/12/95. Senior State Prosecutor. Office of the Attorney-General (Witwatersrand Local Division). Alberton. The State v. Michael Sonti. (1997). Katlehong MAS: 144/12/93. Senior State Prosecutor. Office of the Attorney-General (Witwatersrand Local Division). Alberton. The State v. Michael Sonti. (1998). Attorney General. High Court of South Africa (Witwatersrand Division). The State v. Michael Sonti. (1999). Case No. 378/97. High Court of South Africa (Witwatersrand Local Division). Thornton, R. (1995). The peculiar temporality of violence: A source of perplexity about social power. Johannesburg: Centre for the Study of Violence and Reconciliation. TRC (Truth and Reconciliation Commission of South Africa) (1998a). Amnesty Hearing. Katlehong Massacre. Amnesty Committee. Boksburg:TRC. 280 TRC (Truth and Reconciliation Commission of South Africa). (1998b). Truth and Reconciliation Commission of South Africa Report. Cape Town: CTP Book Printers (Pty) Ltd. TRC (Truth and Reconciliation Commission of South Africa). (1998c). Decision on Katlehong Massacre Amnesty Application, AC/98/0013. Amnesty Committee. TRC. TRC (Truth and Reconciliation Commission of South Africa). (1999). Amnesty Hearing. Application No: AM 7011/97. Mr T A V Nhlapo. Amnesty Committee. Johannesburg: TRC. TRC (Truth and Reconciliation Commission of South Africa). (2003). Volume 6. Truth and Reconciliation Commission of South Africa Report. Cape Town: Juta and Co. Ltd. Tshabalala, V. (1995). Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong. Tutu, D. (1999). No future without forgiveness. New York: Doubleday. Unterhalter, D. (1988). The Doctrine of Common Purpose: What Makes One Person Liable For The Acts Of Another? South African Law Journal, 105(4), 671-678. Van den Wijngaert, C. (1980). The political offence exception to extradition: The delicate problem of balancing the rights of the individual and the international public order. Deventer: Kluwer. Van der Merwe, H. (1999). The TRC and community reconciliation: An analysis of competing strategies and conceptualizations. Johannesburg: Centre for the Study of Violence and Reconciliation. Van der Merwe, H. (2002). National narrative versus local truths: The TRC?s engagement with Duduza. In D. Posel & G. Simpson (Eds.), Commissioning the past: Understanding South Africa?s TRC. Johannesburg: Witwatersrand University Press. Van Vreden, A. H (1993). Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong Van Zyl, S. (1990). Explaining violence in South Africa: Some psychoanalytic considerations. Johannesburg: Centre for the Study of Violence and Reconciliation. Viljoen, A. (1993). Police statement, Katlehong MR144/12/93. Unrest and Violent Crime Unit. Katlehong Wilson, R. (2001). The politics of truth and reconciliation in South Africa: Legitimizing the post-Apartheid state. Cambridge: Cambridge University Press. Woodbridge, J.E. (Ed). (1958). Hobbes Selections. New York: Charles Scribner?s Sons. Xaba, T. (1995). Repressive and Restitutory Sanctions in the Townships: The Social Origins of the Different Forms of Punishment in the People?s Courts. Transformation (28). 281 Young, R. (Ed.). (1981). Untying the Text: A Post-structural Anthology. Boston, MA: Routledge and Kegan Paul. Zwane, W. (1992, 18 June). Phola Park unit hijacked-police. Business Day.