Nicky Rousseau. Journal of Southern African Studies. Volume 40, Issue 6, December 2014.
Introduction
The Soweto Intelligence Unit (SIU) was a specialised security police unit responsible for recruiting deep-cover agents. Accounts to the Amnesty Committee of the Truth and Reconciliation Commission (TRC) suggest that through the 1980s the SIU operated far beyond its Soweto base, extending its network into Botswana, Swaziland and other liberation movement bases. Membership included several police veterans of the South West Africa1 ‘border wars’, at least two Mozambicans who had previously been deployed to South West Africa, an askari2 abducted from Botswana, and a network of informers and deep-cover agents stretching from Soweto into several of the frontline states. The TRC was given accounts of numerous ‘false-flag’ operations, including bombings and other acts of sabotage ostensibly aimed at providing credibility for informers and undercover agents, as well as several cross-border operations against African National Congress (ANC) targets in the mid 1980s. Several of these were conducted jointly with a covert military unit of the South African Defence Force (SADF) that included ex-Rhodesian soldiers and intelligence personnel. The SIU thus provides a way of exploring the multiple intersections of regional struggles and modes of belonging: its theatre of war extended far beyond its local jurisdiction, crossing multiple physical and institutional borders. Its operatives dragged with them diverse histories and legacies, which in turn both shaped and blurred the boundaries of insurgency and counter-insurgency. These aspects arguably have wider implications for the national, and require us to rethink the frames through which we have come to understand the struggle for and against apartheid to have been waged. They also require us to think more carefully about the evidence, including that of the TRC, on which such frames rest.
On 27 January 1995, an article in the Sowetan newspaper told the story of a young woman, Nokuthula Simelane, who had gone missing in 1982, a week before her graduation from the University of Swaziland. Alongside the article was a photograph: this image of a young Simelane, hand somewhat cheekily on hip, looking out and up at the photographer, starkly contradicted the subject positions she would come to occupy in the public eye following the publication of this article. Simelane’s story traversed the borders of Swaziland and South Africa, and while clearly suggesting that the South African security forces were responsible for her disappearance, it also spoke to the wider historical transmigration of families on either side of the Swazi border, and of South African families sending their children to study in schools and universities of neighbouring countries, thus avoiding the prospect of the hated Bantu education system in South Africa.
The photograph evoked a response: on 6 February 1995, a headline in the Sowetan screamed ‘Cops Trapped and Killed MK Cadre’, accompanied by a photograph of a balaclava-clad figure. ‘Mr X’, as he was named, detailed a story of kidnapping, torture and murder by the security police. Although these events had allegedly taken place more than a decade earlier, a criminal case was opened. Shortly after this, two former security policemen, Willem Coetzee and Anton Pretorius, who had emerged as key suspects in Simelane’s disappearance, were suspended from police duties. As the statements in the case docket accumulated, details of the activities of their former unit, the Soweto Intelligence Unit (SIU), began to emerge.
Alongside the criminal investigation, an entirely different process was unfolding in the national parliament. On 17 May 1995, the enabling legislation of the Truth and Reconciliation Commission (TRC) was signed into law, and on 16 December 1995 the newly appointed commissioners held their first meeting. For the first time, and with considerable controversy both locally and among the international human rights movement, a truth commission was specifically mandated to oversee an amnesty process. As an unintended result, the criminal investigation into several political cases, including the disappearance of Nokuthula Simelane and the wider investigation into the SIU, ground to a halt, as investigators and prosecutors waited to see whether suspects in these cases would apply for amnesty. In the months that followed the establishment of the TRC, just under 270 policemen, overwhelmingly security police, submitted amnesty applications. Among their number were nine SIU operatives who did so for their involvement in the Simelane abduction, as well as for a range of other operations.
These amnesty applications, products of an encounter between a post-Cold War experiment in truth telling and what had now become a homicide investigation, were—at their very moment of constitution—unusually negotiated and crafted documents. While the energies of a guilty perpetrator in a criminal investigation are primarily directed at keeping the facts under wrap, key requirements for amnesty included ‘full disclosure’ for each incident for which the applicant had applied, that the offence should be ‘politically motivated’ and that it should be ‘proportional’ to its intended political objective. Applications and testimony were undoubtedly crafted and shaped to meet such requirements. For security police, most of whom had not bought into the political process of the TRC, the decision to apply for amnesty, what to apply for, and what to disclose, was a complex calculation based on multiple variables—who else had applied and what they may have disclosed, who was likely to implicate whom, what TRC investigators might uncover, and whether the new government would pursue prosecutions should someone not apply. This involved revisiting one’s own career, identifying incidents or moments that could open one to liability or possible leakage, identifying co-conspirators and negotiating with them, or trying to limit damage where co-conspirators fell outside one’s power or influence. Whispered conversations, corridor rumours, official and clandestine meetings often served only to produce uncertainty; the strategies that were devised and the decisions that were made ultimately involved negotiating multiple loyalties (and enmities) that crossed personal, official and political boundaries. Testimony at amnesty hearings was similarly highly mediated: as public hearings they were performative events, addressing multiple audiences (some of whom may not even have been present); amnesty applicants were represented and led by legal representatives, and the transcripts were frequently the product of simultaneous translation.
The establishment and the operations of the SIU to which amnesty applications referred span the period 1981-89. This was a period in which the apartheid state pursued a security strategy named Total Strategy, an approach developed by the South African military in South West Africa. Total Strategy drew on the work of counter-insurgency experts, notably Andre Beaufre and later John J. McCuen, and provided a mechanism through which the South African military inserted itself into the history and trajectory of the Cold War. Its central tenet was that South Africa was in the midst of a total, or—as it was phrased from the mid 1980s—a revolutionary onslaught, the defeat of which was possible only via an equally co-ordinated and managed total response, in which security became the prism through which all levels and policies of government were gathered and singularly directed. Overseen by a re-activated State Security Council (SSC), widely regarded as the de facto government, the massive edifice of the National Security Management System (NSMS) was established, consisting of co-ordinating committees and sub-committees at national, regional, and local level.
The implementation of the early phases of Total Strategy (roughly 1979-84), as articulated by military strategists, saw reform of the political environment as a critical element of counter-insurgency measures aimed at disrupting the fertile ground of political mobilisation. One may note here that the notion of counter-insurgency was something of a misnomer given South Africa’s policy of pre-emptive defence, which wreaked havoc across southern Africa in the early 1980s, the success of which was signalled by security accords struck with various frontline states. In particular, the Nkomati and Swaziland accords significantly disrupted a key ‘pipeline’ for the ANC’s guerrilla army, Umkhonto we Sizwe (MK) and especially its Special Operations unit, through which key personnel in Maputo and the MK machineries in Swaziland connected to underground structures in South Africa via couriers and infiltration routes, and vice versa. According to the TRC report, the apartheid state’s satisfaction with these developments was reflected in a view that ‘it had turned a corner’.
By mid 1985 however, according to the TRC, such confidence was crumbling and its Report referenced a dense cluster of documents. For example, at an SSC meeting on 18 July 1985, President P.W. Botha described the insurgency as a ‘spiralling threat’, declaring that the ‘brain of the revolution’ was inside the country and would need to be destroyed. Despite having adopted 11 principles aimed at ‘countering the revolutionary onslaught’, just over three months later a secret meeting of intelligence heads had concluded—with considerable disagreement from the security police—that negotiations with the ANC were inevitable and that security strategy should direct all its efforts at weakening the ANC (and especially its South African Communist Party [SACP] ally and other radicals) as much as possible before such negotiations took place. The TRC argued that this provided a context for the SSC to develop a strategy that sought to integrate both internal and external threats. Known as Strategy 44, or the National Strategy against the Revolutionary War, and formally adopted by the SSC on 1 December 1986, it stated unequivocally that insurgency and terrorism should be fought at its source wherever in the world that might be. To this end, Strategy 44 suggested, alternative structures in ‘liberated areas’ in the townships, and ‘intimidators’, had to be ‘neutralized’, and ‘moderate blacks’ mobilised ‘to defend themselves’ against revolutionaries. For the TRC, then, Strategy 44 provided evidence of the formal adoption of the policy of counter-revolutionary warfare inside South Africa that had previously been recorded in the scholarly literature: the war waged against the region had ‘come home’.
What happens to forms of warfare that travel across continents and regions? And what happens to them in their recounting? In South Africa, the particular strands of counter-insurgency theory that travelled originated mainly from France and the United States, although stints of duty in Rhodesia and migration of personnel from that war brought to South Africa practices more associated with that of Britain and its empire. Suren Pillay reminds us that theories of counter-insurgency have far longer histories than the Cold War, and that, in this part of the world, we would do well to remind ourselves of ‘the proper degree of terror’ instilled during wars of colonial dispossession, and the concomitant production of the world of the native and the settler. These various strands converged powerfully in southern Africa, are pertinent here to South Africa’s war in South West Africa and southern Angola, and require some re-thinking of the work that the appellation ‘Cold War’ does in the context of the apartheid state.
However, this article pursues a different direction: namely, the way in which wars across the region—in Rhodesia, South West Africa and South Africa—and those who fought them repeatedly converged and crossed each other, calling into question the ways in which these wars are regarded and told as national histories. Through the lens of one security police unit in Soweto, whose membership and operations traversed the region, carrying with them multiple legacies and histories, a picture of ‘mobile soldiers’ fighting ‘(un)national’ wars is made visible. More than this, however, the operations of the SIU blur the boundaries of insurgency and counter-insurgency, calling into question any definitional precision. Indeed, like the excision of earlier imperial and colonial legacies identified by Pillay, the very notion of counter-insurgency—which definitionally places insurgency as being prior to it—functions as a form of discursive counter-insurgency. It may be argued that to position counter-insurgency in this way, and its elevation to the status of warfare theory, is possible only because of a degree of complicity between the two.
Reading the TRC and the Security Archives
While issues of memory and testimony in the TRC have been widely debated, much of this literature has centred on victim testimony. With regard to the amnesty process, the accretion of multiple mediations in a single testimony—requirements for amnesty, the crafting of the hearing both institutionally and by multiple participants, the highly contested nature of the process and the transition itself, to mention but a few—make for a degree of density in these exchanges that should caution those wishing to use them as purely evidentiary sources. Decoding their modes of composition and register may require the skills of the historian of the lie, the secret and the rumour. At the same time, it is precisely these dense mediations that make the TRC proceedings a particularly rich archive. The accounts that emerge through the institution of the TRC provide a rare insight into the multiple forms of negotiation and mediation at a moment when the production of history is marked by both predictability and instability as to the limits of invention and representation.
Alongside the amnesty archive is the archive of state security, traces of which were incorporated into the TRC’s research and investigation work. This archive is fragmented, and reflects gaps not only in TRC research and investigation, and access thereto, but in the security archive itself. This is especially so with regard to the security police archive, a result of both construction and destruction: security police were not encouraged to keep detailed written records of secret or covert practices, and although they amassed a huge intelligence database and generated endless reports for the ever-burgeoning security apparatus, much of these were reportedly destroyed in the period leading up to the 1994 elections. Nevertheless significant and important security holdings survived, most notably here those of the SSC and its associated structures, which were accessed by the TRC. I use the word ‘accessed’ deliberately: this was directed by very particular lines of enquiry determined by the TRC’s specific mandate, and under extraordinary pressures of time. These, however, are as much documents of the TRC as they are of state security, and thus of multiple times, requiring, perhaps, a reading ‘along the grain’, rather than one that seeks unproblematically to unmoor them from the TRC process in the interests of detailing the security context, institutions, and operations in the period of ‘total onslaught’. This documentation is not entirely disconnected from that of the amnesty process, even though most of it arose through the work of the TRC’s main body: somewhat ironically, lawyers of amnesty applicants were quick to seize upon some of this documentation to buttress their own arguments that their clients were not ‘bad apples’, as suggested by the National Party, but that their actions had been authorised at the highest level.
At the time, the TRC read the security archive primarily alongside the scholarly literature rather than the amnesty process, which got fully underway only in mid 1998. Much of the work of tracking Total Strategy and its implementation had been done contemporaneously by a generation of what Pillay refers to as ‘activist scholars’; later studies, including those of the TRC, expanded and refined but largely retained the earlier frames of Total Strategy set down by its strategists and replicated (although written against) by these activist scholars. Given the centrality of the military to policy frameworks of Total Strategy, much of the literature focuses on that arm of the security force. Curiously, with one or two exceptions, scholars have paid little attention to the security police’s role and perspectives in the ‘total onslaught’. Not much has been published aside from the TRC report, a seam of literature on the security police counter-insurgency unit in South West Africa—Koevoet, and the accounts of journalists. Unlike the military, only a very small number of security police have sought to write their accounts into the growing archive of security-force memoirs.
Through the literature on Total Strategy, the South African state is rendered as highly militarised, drawing its resources together, subjugating civilian and political components to a wider war effort. In terms of this, a powerful SSC, through the structures of the NSMS, oversaw state policy, from the pinnacle of government to the individual security policeman or state official in the townships. Although it is this fused archive of amnesty and state security documentation that forms much of the basis for this article, I do not engage further with its constitution beyond these few comments. I have not sought to disentangle or verify particular incidents, institutions or policies; I provide neither a reading of the archive ‘along its grain’, nor do I attempt a close reading of the testimony. Instead, while I have undoubtedly transgressed my own cautions, I have sought to follow a backstory in the testimonies, which brings into view the extent to which protagonists of a single security police unit in Soweto crossed the boundaries of nationality, institution, and war. Membership of the SIU, a covert security police unit apparently designed to further the cause of the apartheid state, does not in its retelling reflect the stereotypical representation of white (largely Afrikaner) men, nor even of allied white and black South African police but rather tracks across southern Africa and beyond, crossing boundaries of race, nation and institution. Nor does it seem to be merely a case of singular migrations to and from South Africa: in several cases, protagonists took circuitous routes through the region and beyond. The accounts of SIU operatives, singly and in concert with other units, suggest that a security police unit based in the massive township of Soweto routinely operated both inside and outside South Africa’s formal borders throughout the 1980s.
At the same time, in traversing the contour of ‘unnational’ liberation, I am mindful of Premesh Lalu’s warning that the question of the ‘liability’ and ‘reliability’ of sources is not one that concerns only their evidentiary value, but rather the apparatus of reading itself. This is a question to which I will return, suggesting that rather than using the TRC archive to add detail to the character of counter-revolutionary war as told by scholars and through the technologies of the TRC, it may require a rethinking of the key frame through which apartheid security policy in this period has predominantly been understood by scholars, including those of the liberation movement—namely, militarisation and ‘Total Strategy’. In the meantime, the reader is advised to follow Lalu’s suggestion to read, not against or along the grain, but with ‘a grain of salt’.
The Soweto Intelligence Unit (SIU)
In April 1981, our first key suspect in Nokuthula Simelane’s disappearance enters the picture. Fresh from a stint in South West Africa as part of the Rhodesian-inspired counter-insurgency unit Koevoet, Willem ‘Timol’ Coetzee was transferred from his base in Oshakati to the Protea security branch in Soweto and appointed commander of the SIU. In that year, according to his personnel records, he also received specialised training in explosives and industrial radiography. A member of the police force since 1968, Coetzee, somewhat unusually, went straight from police training to a much sought-after position in the security police. Between June 1972 and June 1974, he completed five tours of duty in Rhodesia as part of the deployment of the South African Police (SAP) to that country, for which he received two medals for combating terrorism.
Two fellow Koevoet operatives, Mozambicans Adriano Bambo and Manuel Olifant, joined Coetzee at the SIU, having followed a somewhat circuitous route to Soweto, the details of which remain hazy. They were said to be part of a group of five Mozambicans—Bambo, Olifant, Amora da Silva, Abilio Maquoqua and George Sigauke—possibly linked to the Mozambique National Resistance (RENAMO), who, prior to their stint at Oshakati, had been housed at the farm Vlakplaas, just 20 kilometres west of Pretoria. TRC investigation and amnesty records suggest that the farm provided the base for the clandestine work of a security police unit, C1, established in 1979 by two senior security police, J.J. Viktor (Snr) and Jac Buchner, who were also experienced SAP deployees to the Rhodesian war. Ostensibly a rehabilitation centre for ex-guerrillas, Vlakplaas was used to gather intelligence and trace guerrilla suspects, to develop a ‘pseudo’ capacity (a technique much loved in counter-insurgency warfare and widely used in Rhodesia) and to conduct attacks on alleged ANC targets, initially in neighbouring states. In addition to the Mozambicans, three ex-Zimbabwe African People’s Union (ZAPU) guerrillas, employed as ‘farm labourers’, are also said to have lived at Vlakplaas; so, allegedly, did Ntsu Mokhele at the time when he was negotiating arms for the Lesotho Liberation Army (LLA).
At some time in late 1980 or early 1981, all or some of the Mozambicans had been dispatched to Oshakati, where Bambo and Olifant first encountered Willem Coetzee, then second-in-command of the Koevoet unit to which they were attached. Indeed, Coetzee was said to have given Bambo the nickname ‘Strongman’ when, following their first ‘contact’ with SWAPO guerrillas, Bambo jumped from the military-style vehicle, gun blazing. Initially registered as ‘informers’ in the SIU, Bambo and Olifant were later enrolled as constables.
Several others police were transferred to the SIU around this period, including Anton Pretorius and Nimrod Mzimkulu Veyi. In 1981-82, Pretorius (the second key suspect in Simelane’s disappearance), was a sergeant in Benoni with experience in recruiting and handling sources in Tembisa. Fresh from a course on security, he acted as Coetzee’s second-in-command, and over the next few years also received medals for combating terrorism, routinely awarded for tours of duty in South West Africa. Veyi, who had previously worked at the neighbouring Meadowlands police station in Soweto, joined the SIU after having been introduced to Coetzee by an older security policeman. It was Veyi who was to become the whistle-blower in Nokuthula Simelane’s disappearance—the Sowetan‘s balaclava-clad ‘Mr X’.
These were other, more covert, recruits. Young men and women in Soweto intending to join the police were interviewed at a recruitment centre at the Protea police station. They were unaware that ‘spotters’ from the SIU, who sought to identify possible recruits to be trained as deep-cover agents, were observing them. If a recruit were thought to be eligible, the ‘spotter’ would offer the candidate a ‘short cut’ to becoming a policeman. These recruits, having told their families and friends that they had decided against joining the SAP, would then be trained and infiltrated into targeted structures, mainly the ANC’s guerrilla army, MK. This was possibly the route through which Frank Langa came to be agent RS269, and Norman Mkhonza agent RS243 (the RS denoting Republieke Spioenasie—Republic Espionage). Sometime before Langa and Mkhonza’s recruitment, in the early 1980s according to Coetzee’s amnesty testimony, a young woman and Soweto resident known only as ‘Nompumelelo’ was stopped crossing from Swaziland into South Africa at the Oshoek border post. After a hand grenade was discovered in her possession, she was offered the choice of being arrested and charged, or continuing to work in MK’s Transvaal Urban Machinery structures but as a security police informer. She chose the latter option, and was registered as source SWT (i.e. Soweto) 66. To prove her new loyalties, she stole two explosive devices from an MK arms cache and handed them over to her handlers. It was through Nompumelelo that Langa and Mkhonza were introduced into and infiltrated Swaziland MK networks. Sources active within MK in Botswana included two deep-cover agents, RS276 and RS283, and sources R103 and SWT180, who later also operated in Swaziland (‘R’ is explained below).
While the circumstances whereby Nompumelelo came to work for the SIU were seemingly fortuitous, those surrounding Moleke Peter Lengene, an askari, appeared more deliberate. Lengene described himself as the stepson of the first black ‘mayor’ of Soweto, educated at Morris Isaacson High School and a member of the Soweto Student Representative Council, both key organisational nodes of the 1976 Soweto uprising. According to his TRC statement, Lengene went into exile, travelling to Botswana, where he joined the newly formed black consciousness group, the South African Youth Revolutionary Council (SAYRCO). From there he travelled to Nigeria to study, before being sent, seemingly against his wishes, to Libya for guerrilla training. Deployed back to Botswana in early 1981, Lengene encountered one George Khoza, who spoke with a Zimbabwean accent and claimed to have been a Zimbabwe African National Union (ZANU)-trained guerrilla. Khoza began supplying SAYRCO with hand grenades, and later offered Lengene a consignment of AK47s. Lengene’s application narrates being ushered into a dark house in the Gaborone suburb of Broadhurst on the night of 6 February 1982, where the handover of guns was scheduled to take place. Following a burst of brightness as the lights were turned on, he was stormed by ‘a lot of people, both black and white’. Following a scuffle, Lengene was bound and drugged, then taken to the isolated farm owned by the father-in-law of SIU commander Willem Coetzee. There, as the target of a ‘kopdraai aksie‘—literally a ‘turning of the head action’—he was worked over with a mixture of violence and persuasion. On this account, Lengene’s journey from stepson of an apartheid collaborator to Soweto radical once again reversed, from that of an underground SAYRCO guerrilla to becoming a much-hated askari.
The Intelligence Operations of the SIU
According to Simelane suspect and amnesty applicant Anton Pretorius, the SIU was a covert intelligence unit that, although part of the Soweto Security Branch, operated from safe houses and secret locations. Its primary purpose was to collect intelligence both inside and outside South Africa’s borders. Here the notion of liberation movement ‘pipelines’ normalised and justified a parallel security force traffic of information and operations across national borders. Speaking of the difficulty of separating internal from external, a military intelligence operative put it this way:
The Botswana [MK] machinery may be in Gaberone [sic] to day [sic] but tomorrow they are somewhere in the Western Transvaal. So should we stop our operation at the border or should we follow the pipelines through to their courier systems and their safe houses inside the country?
In the early 1980s, Soweto was apparently the only Security Branch division outside headquarters authorised to recruit and run deep-cover agents outside South Africa. Deep-cover agents included policemen such as Langa and Mkhonza who, in intelligence parlance, ‘infiltrated’ targeted organisations, or were members of organisations that had been ‘penetrated’ and who had been recruited as full-time agents (R agents). A further category, and until the second half of the 1980s also the preserve of Security Branch headquarters, was the recruitment and handling of askaris. Although both R agents and askaris were frequently recruited through ‘kopdraai aksies‘ such as that experienced by Lengene (above), R agents were returned to their organisational and social milieu. Askaris were, by contrast, until the late 1980s based at Vlakplaas, and deployed as ‘terrorist tracing teams’ as well as in covert operations, within and across South Africa’s borders.
Operating spy networks took the SIU outside their offices, perhaps recreating in the process something of the wide-open spaces and mobility of the ‘border wars’. Amnesty applicants refer to ‘safe houses’ used to brief and debrief agents and sources, which included a house in Klipspruit where Mozambicans Olifant and Bambo, and possibly askari Lengene, stayed, outbuildings such as a store-room or garage on family-linked farms in Westonaria and locations further afield in the western Transvaal. These were also said to be used in ‘kop-draai aksies‘, which sometimes required ‘camping over’ and eating the ‘rat packs’ usually doled out to soldiers and Koevoet operatives. There was also movement across the borders into Swaziland and Botswana; the relatively short distance from Soweto meant that SIU operatives could travel unofficially into either country and return the same day.
In these accounts, it was not only jurisdictional and national boundaries that were crossed. In any recruitment operation, both sides sought to harness new recruits as well as to test their credentials. Thus, a ‘turned’ agent needed, as in the case of ‘Nompumelelo’ (above), to be quickly compromised so that her or his fear of the consequences, should s/he be unmasked, would prevent hasty confessions to the betrayed side. Askari Peter Lengene was accordingly involved in the abduction of Nokuthula Simelane only months after his own. Conversely, or so the SIU claimed, from MK’s point of view, a new recruit who failed to carry out successive missions would come under suspicion, requiring further scrutiny. Thus Langa and Mkhonza, once they had successfully infiltrated MK’s Transvaal Urban Machinery in Swaziland, were allegedly given sabotage missions by their MK commanders in Swaziland. In order to prevent their unmasking and give them credibility, the SIU, according to their testimony, detonated explosives at the Bryanston and Randburg substations and on the Johannesburg-Durban railway line. Credibility operations carried out in the late 1980s for which amnesty was sought by SIU operatives included establishing arms caches, several grenade attacks on houses or properties, including a migrant hostel, using allegedly weakened explosive or dummy devices, and attacks on government administration board offices in Soweto.
How does one classify an act of insurgency enacted by a counter-insurgent? While broadly regarded as a ‘false flag’ operation in which security forces purport to be guerrillas carrying out the operation, they can be distinguished from a false flag or pseudo operation designed to discredit the opposing side, such as, for example, an attack on civilians that may be regarded as targeting ‘innocents’. Taken over the long view, the aim, and possibly even the effect, of the counter-insurgent’s action in a credibility operation may be consonant with counter-insurgency. But how does one account for its public effect at the time and the mobilising impact it may have on insurgent forces, for which this operation was an apparent success? Such attacks frequently entered daily unrest statistics, and were presumably read by a wider public as attacks by young radicals on collaborators. Indeed, in some cases, MK claimed such operations as its own. Insurgent actions by counter-insurgents may have been seen as the price to be paid for successful infiltration, although in this particular case it is by no means clear that the SIU’s efforts can be regarded as successful; RS agents Mkhonza and Langa seem to have been exposed by 1986.
‘Fighting Fire with Fire’
If ‘credibility operations’ required security police to think and act as insurgents, then counter-revolutionary warfare further strained the boundaries and understanding of insurgency and counter-insurgency. Although Strategy 44, the document officially transacting the shift to a policy of counter-revolutionary warfare, was adopted only in late 1986, it endorsed a strategy and tactics already in practical effect, just as its surfacing in the TRC acted to endorse claims by amnesty applicants that they were operating within an established policy. Thus, for example, a security document recording a joint meeting of senior members of the security establishment at Security Branch headquarters on 2 May 1985 called for ‘the physical gunning down of leaders in riot situations’ as one means whereby ‘ringleaders’ could be ‘selectively eliminated’. A verbal skirmish between amnesty applicants and an implicated General concerned an alleged meeting in February 1986, at which J.J. Viktor (Snr), founder of Vlakplaas, a veteran of the region’s wars and second-in-command of the Counter-Insurgency and Riot unit, had ‘unofficially’ suggested that Pretoria security police fight insurgency with insurgency. Although Viktor strongly denied to the Amnesty Committee that this constituted an instruction, he conceded that he had said that if the security police knew who was responsible for violence they should be ‘haunt[ed] … so that they would find no rest’, and that security police could ‘retaliate against those who were known to have attacked the houses of members of the Police, by attacking their homes in a like manner’, including the use of petrol bombs. The commanding officer of the Pretoria security police, and a fellow veteran of the Rhodesian war, claimed that this meant that ‘there was in reality an active guerrilla war to be waged against the ANC, SACP and PAC [Pan Africanist Congress] activists as well as activists from other liberation organisations’.
According to the applicants, in the months following this meeting, members of their covert security police unit ranged through the townships of Mamelodi and Atteridgeville in a battered car at the dead of night, throwing petrol and more lethal pentolite bombs through the windows of township homes. J.J. Viktor (Jnr), who had been present at the meeting where his father and namesake had advocated attacking the homes of young arsonists ‘in a like manner’, told an amnesty hearing that he had been involved in approximately 40 such bombing incidents between February and May 1986. The commander of the covert unit, one Jacque Hechter, a loner and Koevoet veteran who loved the ‘bush life’, also applied for amnesty for mowing down nine youths in Kwandebele, and a series of abductions and killings in which the bodies were staged variously as victims of ‘necklacing’ or, more commonly, guerrillas who had accidentally blown themselves up.
Amnesty applications for some of these operations drew in not only Pretoria security police but also operatives from the SADF’s Special Forces, some of whom were members of the euphemistically named Civil Cooperation Bureau (CCB). The latter was a covert military structure within Special Forces, with roots in an earlier covert unit called D40, composed of ex-Rhodesians, allegedly responsible for assassinations and other covert operations. A senior general and former head of Special Forces claimed that such operations were the result of a plan he had devised following an instruction given to him by Chief of the Defence Force General ‘Jannie’ Geldenhuys, and later authorised during a social function. At the TRC armed forces hearings, Geldenhuys described the internal security situation in South Africa in the mid-1980s as reflecting the fourth and final phase of ‘revolutionary warfare’ identified by counter-insurgency theorist John J. McCuen. The TRC concluded that this plan, devised just six months after the general’s promotion to head of Special Forces after a decade in South West Africa (the last few of which he was a decorated commander of Sector 10, based in Oshakati), was evidence of the transplanting of personnel and practices from South Africa’s primary external theatre of war to the domestic arena. ‘My Plan’, which aimed to follow McCuen’s doctrine of turning the methods of revolutionary insurgency against the insurgents themselves, ostensibly required a close reading and re-enactment of insurgency’s methods. From the accounts presented to the TRC however, the plan appears to have been implemented in a far more haphazard way.
In Pretoria, joint operations between security police and the Special Forces/CCB unit (for which both security police and military operatives applied for amnesty) included targeted assassinations (one of which used two Angolan ‘hitmen’, João da Pinta and Louis da Silva) and a pre-emptive operation targeting ‘potential’ MK recruits in which ten young activists were killed. A further joint operation involved another assassination by the same unit, this time of controversial Kwandebele Minister of the Interior Piet Ntuli (an ally of the apartheid government, but whose strong-arm tactics were said to have become ‘troublesome’), was actually claimed by MK; in the wake of this event, which further entangled the lines of insurgency/counter-insurgency, violence in the Kwandebele Bantustan escalated.
The amnesty hearings associated with the joint operations between Pretoria security police and Special Forces took place in several clusters over more than three years. Soweto was not mentioned as one of the three ‘hotspots’ identified in the plan devised by the Special Forces general (although the Witwatersrand was, along with Pretoria and the Eastern Cape), nor did the SIU feature in any of these hearings. Nevertheless we may track an itinerary back to this plan from SIU testimony about a car and two key CCB-Special Forces operatives who applied for amnesty for certain of the Pretoria events. These were an ex-Rhodesian (whose records suggest he had ‘completed’ several Selous Scouts courses) and the CCB regional ‘manager’ for Botswana, who was also the alleged point man in liaison with Pretoria security police. In one such joint operation in Soweto, operatives allegedly carried out an explosion at the Ipelegeng Centre, which was used by organisations such as the Soweto Youth Congress for meetings; in another, an explosive device was detonated at the Jabulani Stadium, where Inkatha was due to hold a meeting. While TRC amnesty applicants claimed that they were trying to prevent violence, it can be argued that, read alongside Strategy 44’s explicit support for contra-mobilisation and the desirability of exploiting ethnicity, the opposite was intended, since the action was likely to be blamed on MK or youth activists. The Mozambican Manuel Olifant alleged that he conducted surveillance of the head of the South African Council of Churches, Revd Frank Chikane, following a meeting at which it was said that Chikane was to be ‘eliminated’. One ‘Peter’ and one ‘Mike’, two otherwise unknown military operatives of British origin, were among those who attended the meeting. Olifant further alleged that the vehicle used in one of the Pretoria assassinations was hidden at one of the SIU’s safe properties.
These were not the only joint military-police operations in Soweto: in another SADF-inspired project, Battalion 32 soldiers—composed largely of remnants of Angola’s defeated FNLA—formed part of covert teams who rounded up suspects for clandestine and illegal interrogations during the night. While SIU applicants made no reference to this operation, code-named ‘Xenon’, Battalion 32 soldiers would have relied on lists of suspects provided by Soweto security police.
If the notion of the ‘pipeline’ apparently enabled a Soweto-based unit to conduct intelligence operations outside national borders, counter-revolutionary warfare and the relationship with Special Forces-CCB operatives provided justification for the SIU’s involvement in cross-border raids. This trespassed across national boundaries and the boundary between intelligence-directed operations and offensive operational engagement. According to their testimony, not only did SIU operatives provide intelligence for a major cross-border raid into Gaborone, Botswana on 14 June 1985, Olifant also received additional special training and accompanied one of the attack teams. Olifant was also part of a further (failed) cross-border operation, again involving the CCB’s regional ‘manager’ and the unknown British operatives, ‘Mike’ and ‘Peter’, in an attempt to assassinate two senior ANC members in Gaborone in 1987. SWT180, named by Simelane suspect Anton Pretorius as having provided intelligence for the Gaborone operations, was said to be the source of information about the movements of another senior MK operative, this time in Swaziland. In this case, Coetzee and Pretorius, members of an intelligence unit, applied for amnesty for their role in both the intelligence gathering and the operational team, which, on 13 December 1986, attacked a house in Dalraich, Mbabane, Swaziland, killing the intended target and two others.
A further similar example, this time inside Soweto, involved three young activists whom the SIU claimed had been involved in attacks on policemen. Askari Peter Lengene described infiltrating the group, supplying them with AK47s, hand grenades and a SPM limpet mine, and later introducing them to two Vlakplaas askaris. The askaris, purporting to be MK guerrilla commanders, trained the young men in the use of explosives. Somewhat ironically, one of the askaris‘ own detention documents suggested that he had been tasked by his MK commanders to teach young radicals in the townships to handle explosives and weapons. Following this, the askaris allegedly assisted the youths in identifying three targets, and supplying them with (zero-timed) explosives. The Mozambican Olifant and the askaris accompanied each of the three on their operations, and were instructed to shoot the activists if the devices failed to detonate—two were thus shot (one was left to die of his injuries), while the third, regarded as the ringleader, was killed when the zero-timed limpet mine detonated.
Boundaries of Belonging
The above operation, which took place in 1989, was the last of the SIU operations for which amnesty was sought at the TRC. By then, according to police documents, both key Simelane suspects had moved on: Pretorius had been promoted to head of the unit after Coetzee was transferred to security police headquarters. Deep-cover agent RS243, Mkhonza, claims that after the Simelane operation he acted as an ordinary policeman until 1995, when he left the police; whistle-blower Veyi left the SIU in 1985-86, moving to the detective branch in Cape Town; askari Lengene and Olifant remained in the SIU. It is not clear what had happened to ‘Nompumelelo’, source SWT66. Coetzee and Pretorius testified that, following the 1985 Botswana Raid, three SIU agents (R103, RS276 and RS283) operating within MK had fallen under suspicion and had been recalled to Lusaka, where one was allegedly shot and the remaining two transferred to the MK detention barracks at Quatro. Coetzee was apparently accompanied to Pretoria by his old Mozambican comrade in arms from Koevoet, the ‘strong man’, Adriano Bambo. This was despite the fact that, following some disciplinary infractions, Bambo had previously been arrested and convicted for a criminal offence, spending two to three years in prison. According to applicants, he was later re-arrested for armed robbery, and, while awaiting trial in 1991, Vlakplaas operatives booked him out from police cells. He was shot dead on the side of the road near Nelspruit, allegedly while pointing out an illegal arms cache. The Vlakplaas operatives claimed that they were acting on instructions from the head of the Security Branch, as a result of Coetzee and Pretorius’s fear that Bambo would disclose SIU operations. Both Coetzee and Pretorius denied involvement.
Beyond the personal fates of these operatives, this account of the SIU suggests that wars in southern Africa were fought by a diverse group of men, often crossing boundaries of race, nation and affiliation. From the outset and throughout the 1980s, the SIU appeared to draw in, if only on an ad hoc operational basis, men from diverse backgrounds and multiple locations. To the list provided above we can now add Namibian, Angolan, and British/Rhodesian expatriates. Thus white South African men who had fought for Rhodesia and South West Africa were joined in at least a few cases by black Mozambicans; and black Mozambicans, Namibians, Zimbabweans and Angolans fought for white South Africa alongside various white Rhodesian/British expatriates. Nor was this unique: while the SIU may have occupied a particular niche in the running of deep-cover agents, other operational units, especially C1 (Vlakplaas) and security police units in border zones who ran sources in neighbouring countries, similarly crossed lines of racial and national affiliation. The military unit the CCB, some of whose operatives and structures have featured here, also operated in a number of external regions staffed by both South Africans and a motley crew of mercenaries—ex-Rhodesians, other Africans, British, as well as the odd French or Seychellois.
This raises the issue of the forms of belonging that operated within the SIU. Accounts presented to amnesty hearings provide little sense of a cohesive unit with a clear and singular purpose or commitment to a single cause. Rather, SIU operatives appear to have lived at the borders of different pasts and presents, and the boundaries of belonging seemed brittle. The amnesty application of abductee and askari Lengene, who worked in the SIU for some 18 years, is saturated with anger and bitterness. Labelled a ‘spy, informer, sniffing dog of the police’ by his relatives, he claimed that
[After my abduction] I stayed in the land of apartheid, racism and suffering for the majority … and freedom and riches for the few…. I could feel apartheid and racism everywhere, when one walked with my new masters, we would go to a single building, but once when we had to go the toilets, the famous sign was still there with bold black and white letters, Blacks only and Whites only.
These allegations were strongly disputed by the former SIU commanders, Coetzee and Pretorius, who pointed out that Lengene had subsequently recruited his own brother as a policeman. Murdered shortly before his amnesty hearing, Lengene was unable further to contest such denials.
Of course there were strong reasons for all involved in the amnesty process to recast their experiences in the post-1994 milieu; a new political dispensation evoked numerous anxieties about past and future career paths for both white and black security police, for which new forms of belonging could be required. But even setting this aside, the generally appalling work situation of all black personnel in the police force provided an abundant well of discontent. Used as guards, interpreters, and ‘muscle’ during interrogations, black policeman were poorly paid and permanently trapped in the lowest echelons of the police ranks, with few promotional prospects. Citing whistle-blower Veyi, the Sowetan noted that after ‘almost two decades of serving in the Police Force, he [Veyi] remains a lowly Constable. His white bosses have climbed the ranks to Colonel and General’. Askaris, even when appointed as permanent policemen, seemed to carry an additional burden of residual distrust and suspicion, perhaps arising from the defection of Victor Mnisi, who was captured, ‘turned’ and then escaped, returning to the ANC and subsequently playing a key role in an MK Special Operation’s attack on the South African Air Force headquarters in downtown Pretoria in 1983. Nevertheless, the SIU—and security forces more generally—reportedly relied heavily on such agents: the inside intelligence they provided, especially their nous for everyday life in the camps and abroad in numerous training venues, was invaluable in running a successful intelligence operation, as was their capacity in ‘pseudo operations’ such as the one outlined above involving zero-timed devices. Askaris were also used extensively across the country in numerous interrogations and court cases, not only for their knowledge but also because of their capacity to disarm and soften up detainees.
Coetzee and Pretorius’ own descriptions of their unit are fairly damning. Discussing his general modus operandi in relation to ‘kopdraai aksies‘, Willem Coetzee testified to the importance of instilling fear and terror by means of assault, torture and intimidation, and then of developing a bond via a process of concessions and training. However, any perceived lapses (such as withholding information, or ‘cheekiness’), he indicated, would be met with renewed violence. Physical violence aside, this mind-set appeared to have broader application in his day-to-day working relationships in the SIU. In the words of his old Koevoet associate, Mozambican Olifant, who appeared to hold him in some regard, Coetzee was ‘a rigid commander … you know, he wanted these things to really work on its path’. If violence constitutes particular kinds of community, then this was a community in which suspicion and fear were never too far away.
There is one exception to the rather sour air that pervades descriptions of working relations within the SIU: that of the relationship between Coetzee and the Mozambicans, especially ‘Strongman’ Bambo, arising from their earlier association in Koevoet, from where Coetzee ‘preferred to take … him’ in order to join the SIU. Not that this was a warm and collegial relationship; rather it spoke to mutual ties of obligation and reciprocity. These were ties that—if the testimony of former Vlakplaas operatives cited above is to be believed—were not sufficiently strong to risk exposure, and which therefore had to be erased by Bambo’s murder. It is, however, possible that in conceding a degree of violence Coetzee sought to minimise his own brutality, or, in the face of contrary evidence from some of his former black colleagues, to ‘boost his own credibility’. Similarly, Olifant may have had his own reasons for depicting Coetzee in this manner.
Conclusion
Kevin O’ Brien, drawing significantly on the TRC, describes the apartheid state as having ‘the most effective and structures [sic] programme for the elimination of its opponents, either through counterinsurgency or counter-guerrilla warfare, or through the direct individual targeting of these opponents for selective killing’. Yet in the account of the SIU above, while there were certainly some successful operations, it is none the less hard to escape a sense of marginality, even failure. Citing an apartheid military intelligence officer who said that the Total Onslaught ‘… wasn’t total, and it wasn’t an onslaught’, Stephen Ellis notes that ‘this was not what President Botha wanted to hear. Itself an inveterate maker of myths, in the end the National Party was unable to realise the claims it had always made for itself. Its official view collided with a hard obstacle called reality’. Yet in general, neither scholars nor the TRC have considered the breach between the vision of Total Strategy and the record of its implementation, especially through the period of counter-revolutionary warfare in the mid-late 1980s. To do so is not to recast the apartheid state’s response in less brutal and more benign terms; rather, it is to ask whether the difficulties experienced in relocating the external war internally, inside South Africa’s borders, are a means through which the texts of Total Strategy and counter-revolutionary warfare begin to unravel. At one level this may point to the security state’s ability at a discursive level to talk war, as well as to inflict significant damage through successive highly repressive states of emergency. In contrast, its actual record, while demonstrating intense repression and great brutality, suggests that it was ultimately unable to unleash the forms of violence that it had deployed in the region within South Africa itself, except perhaps in areas of KwaZulu and Natal, in concert with Inkatha. Borders, which had proved highly permeable for the South African security forces in its open war in South West Africa, its covert war in southern Angola and its proxy wars across southern Africa, proved less permeable in reverse motion.
As I have argued elsewhere, this translated into a situation in which there were ‘unofficial/official’ forms of authorisation, and in which operatives were required to devise their own modus operandi in the absence of clear rules of engagement. In doing so, operatives drew on the personal, even the familial, but most significantly on their experience in the so-called ‘bush wars’ of the region. Their attempt to recreate the circumstances of those wars echoed in some of their terrains—safe houses, farms, weaponry and mobility across the region—but never amounted to more than a brutal sideshow. Ultimately the wars of the region (which in any event had already been lost) translated poorly into the urban landscape of South Africa. Writing in a different context, Saskia Sassen suggests that ‘killing civilians in a city is a different type of horror from killing people—far more people—in the jungle and in villages. In that sense, the urbanising of war points to the limits of power’. Counter-revolutionary warfare unfolded in a far more contingent, haphazard and depleted way than is suggested by the grand security strategy enshrined in the moniker, Total Strategy.
In short, Total Strategy and its National Security Management System, as they have been depicted, bequeath a notion of the apartheid state that, ironically, takes as given apartheid’s own imagining of itself. In the words of Ranajit Guha, neither scholars nor the TRC have escaped the binds of the military strategists who inscribed the policy, and thus conform ‘to a counter-code, the code of counter-insurgency’. Reading through the archive of the State Security Council and extant intelligence documents of the 1980s, alongside the archive of the TRC, speaks more to what securocrats dreamed was possible, an attempt to create order out of a disorder that threatened to overwhelm. At the same time, this does not account for the far greater intensity of violence in the early 1990s. Certainly there were continuities between that violence and the period of counter-revolutionary warfare, but here too it may be advisable to take heed of Guha’s caution before tracing the lines of a single master plan.