Most of the violence in the pre-election period of 1990-1994 was black-on-black violence between the Inkatha Freedom Party (Zulu) and the African National Congress (Xhosa).

Whenever negotiations reached a critical or delicate point there were sudden outbreaks of township violence.

This gave rise to questions about the "Third Force". There is speculation that the National Party (apartheid government) was deliberately initiating the violence and supplying Inkatha members with weapons and military training. Thereby increasing black-on-black violence in an attempt to break trust in the ANC through the Inkatha Freedom Party.

There seems to have been a close relationship between the National Party and the Inkatha Freedom Party. In fact, many white South Africans voted for the Inkatha Freedom Party after being led to believe that although the National Party would be voted out, the Inkatha Freedom Party would pose a threat to the ANC.

The following is part of a report of investigation findings regarding the "Third Force".

1. In its Final Report to President Mandela in 1998, the Truth and Reconciliation Commission (the Commission) made reference to the ‘Third Force’ in its discussions on the subject of ‘Political Violence in the era of Negotiations and Transition,
1990-1994 In addition, the Commission made a set of specific findings on
the‘Third Force’ during this period.

2. The early 1990s witnessed unprecedented levels of political violence, with over14 000 people killed and many more thousands injured. Although violence permeated the country, most violations occurred in the homeland of KwaZulu and the neighbouring Natal province, and in the PWV (Pretoria – Witwatersrand- Vereeniging) region of the Transvaal. In the latter region, the Human Rights Commission (an independent non-governmental organisation) estimated that some 4756 people were killed between July 1990 and June 1993 alone. While many of these killings can be attributed to the internecine conflicts that developed in many communities, primarily between supporters of the IFP and ANC, there
were frequent allegations about the role and complicity of elements within the security forces .

3. The ‘third force’ label was first used by ANC leadership figures in the wake of a wave of seemingly random attacks on the Witwatersrand and Vaal areas in August and September 1990. As the attacks continued, allegations were made that a ‘hidden hand’, or ‘Third Force,’ was involved in orchestrating and fomenting violence – to derail the negotiation process and/or to undermine the ANC’s efforts to consolidate its political presence. These attacks were believed to involve covert units of the security forces acting in concert with individuals or groupings, such as the IFP and certain right-wing paramilitary organisations.

4. Although the Commission wishes to restrict the understanding of this
phenomenon to the post-1990 period, its origins and genesis can be found in the philosophy of the ‘Total Strategy’ and the practices of covert counter- insurgency that developed throughout the period from the 1960s to the 1980s.

5. The apartheid state’s counter- insurgency efforts intensified during the 1980s, and especially after 1986. As testified to by a number of security force amnesty applicants, the methods employed by specialist covert units included murder, torture, kidnapping and various other covert illegal actions. They also involved the use of proxy and surrogate forces, including freelance criminal elements.

6. The development of intelligence-gathering units with an offensive capacity had proved effective in the Rhodesian situation and was subsequently adapted to the South African context by both the police and military. The devolution of decision-making powers resulted in police units such as Vlakplaas and the Namibian-based hunter-killer unit Koevoet operating with virtual impunity, making it extremely difficult to establish lines of command and accountability.

7. Unlike the police, the military made no disclosures to the Commission about its role in violations, with the exception of admissions about two sets of assassinations executed by South African Defence Force (SADF) Special Forces in 1986. In these cases, the head of Special Forces, Brigadier Joep Joubert, claimed that the chief of the defence force gave him approval. Allegations of complicity in ‘third force’-type activities in the 1990s were denied, including those relating to an array of charges generated by General Pierre Steyn’s preliminary investigations into covert military operations in late 1992.

8. Such denials and the limited evidence available make it difficult for the
Commission to make specific findings, especially on the role of the military. This does not mean, however, that such activities did not take place. Indeed, the security forces were repeatedly involved in a long line of cover-ups of illegal or unlawful activity. This is evident, for example, from evidence about torture and killings that emerged in inquests and trials, which again, in cases such as that of Stanza Bopape, reached the highest echelons of the police.

9. T h e re is no evidence to suggest that this practice was halted during the 1990s. The Harms Commission is a significant example of this: not only were witnesses instructed by their seniors to lie, but the Harms Commission failed to deter them from embarking on further operations.

10. The March 1994 Goldstone Commission report on the criminal activities of the South African Police (SAP), KwaZulu Police (KZP) and Inkatha Freedom Party (IFP) provides further compelling evidence of the fact that senior police officers attempted to subvert a government-appointed commission of inquiry. According to the report, senior members of the SAP repeatedly approached police officers
associated with the Goldstone Commission in the course of the investigation – in ways that could only be construed as obstructive. Further, once the police became aware of Goldstone’s interest in false passports, those in possession of such passports were requested to bring them in for destruction. Similarly, Goldstone investigators learnt that Major General Engelbrecht, the last head of the SAP Counter- I n s u rgency Unit (C-section), had ord e red the destruction of all
documentation relating to the SAP’s involvement with Inkatha.

11. The fact that such cover-ups involved senior officers and continued well into the 1990s reflects the extent to which such groups felt they had the authority to act with total impunity. In such a context, the impression must have been conveyed to the more junior members of such structures that, despite negotiations, they were still at war and could make use of whatever means they had at their disposal, if not to rout, then at least to weaken ‘the enemy’. The continued practice of referring to the ANC as ‘the enemy’ in SADF operational commands clearly underscores this.

12. While allegations of ‘third force activities’ in no way account for all or even the bulk of violent incidents during this period, these attacks were particularly significant as they appeared to be largely indiscriminate, and consequently spread terror amongst hundreds of thousands of township residents. The types of attacks included drive-by shootings, attacks on trains and taxis, and massacres at social gatherings such as night vigils and shebeens. Regular allegations of collusion between elements of the security forces and the IFP were refuted as propaganda. Although a number of these attacks could be placed within a matrix of revenge violence, many could not. Indeed, they gave the impression of being deliberately designed to provoke further violence.

13. In late 1991, a Johannesburg-based non-governmental organisation, the Community Agency for Social Enquiry (CASE), published a research report analysed the first twelve months of Reef violence and highlighted the major actors, victims and alleged patterns of control of the violence that erupted during the period from 22 July 1990 to 31 July 1991. The report was based on thousands of reports from a range of newspapers and figures published by human rights monitoring organisations, including the Independent Board of Inquiry into Informal Repression (IBIIR), the Human Rights Commission (HRC), the Centre for Applied Legal Studies (CALS) and Lawyers for Human Rights.

14. The report stated that IFP supporters and the police were reportedly responsible for the vast bulk of the classifiable acts of violence. Furthermore, it was revealed that the targets or victims of the violence were mainly ordinary citizens. Of the 2271 people killed during the period, 87 per cent were recorded as ‘general members of the community’. There were reports of direct collusion between members of the SAP in 257 confirmed incidents. The report also showed that IFP-supporting hostels provided the base for massive attacks on squatter camps,
and that at least 915 of the total number of 2271 killed during the period were the result of attacks from these hostels.

15. The report suggested that the violence could not simply reflect a violent power struggle between the ANC and IFP, and that the monthly breakdown of the deaths reported during the period made it difficult to believe that the sharp monthly variations were random.

16. The sudden escalation of violence in 1990 coincided with the establishment ofInkatha as a national political party in July, and its attempts to develop a politicalbase in the Transvaal.

17.Inkatha’s relationship with apartheid security force agencies had a long history. In April 1986, the State Security Council approved guidelines for a strategy for a counter- revolutionary war, which, amongst other things, emphasised the fact that the forces of revolution should not be combated by the security forces alone, but also by ‘anti-revolutionary groups such as Inkatha … or sd well as the ethnic factor in South African society’.

18. In 1986, the State Security Council also authorised the provision of military training for 200 Inkatha members by the SADF. The special project to support Inkatha was called Operation Marion and was the responsibility of the Directorate of Special Tasks, a section within the office of the Chief of Staff (Intelligence), which was also responsible for supporting insurgency initiatives in neighbouring front-line states.

19. Support for Inkatha continued in the early 1990s. It is now known that President de Klerk approved a Strategic Communication (Stratcom) propaganda project in 1990, which included financial support by the SAP for Inkatha. In July 1991, the existence of a secret police project to fund Inkatha was revealed in the media. In response to these revelations – which became known as ‘Inkathagate’ – President de Klerk reshuffled his two leading security ministers, Malan and Vlok, and ordered a new review of secret projects.

20. It is also now evident from evidence presented to the Commission that elements in both the police and the IFP continued to collude with one another throughout the negotiation period, and that the police, mainly through Vlakplaas operatives, supplied considerable amounts of weaponry to the IFP during the 1990s. This was also covered by the March 1994 Goldstone report, which implicated members of the SAP, KZP and IFP in the supply of weapons to the IFP. This included a
massive arms cache unearthed in KwaZulu-Natal during 1999, which was provided to the IFP by Vlakplaas, ostensibly for the purpose of training self-protection units. The Commission received a number of amnesty applications corroborating this evidence from both Vlakplaas and IFP operatives. Eugene de Kock, for example, claimed that his unit provided and sold weapons directly to hostels on the East Rand and elsewhere. When Vlakplaas was officially closed down in 1991, unit members were redeployed to work on the recovery of illegal firearms . This provided a perfect cover for the further distribution of weapons and other
fraudulent activities. Chapter One of the Investigation Unit’s Gun RunningReport deals with the ‘receipt of weapons by the IFP’ and describes the background and systematic distribution of weapons in the PWV re g i o n .

21. In addition, the Commission received detailed testimony from Vlakplaas operatives about the specific nature of relations with senior IFP officials operating in both the Transvaal and Natal. Security police resources were used, and a core group of IFP members was allegedly placed on the Vlakplaas payroll for a short period of time. According to De Kock, the relationship was known about, approved and even encouraged by senior police officials.

22. When the IFP’s Transvaal Youth Brigade leader, Themba Khoza, was trapped in the grounds of the Sebokeng hostel after the massacre of nineteen hostel inmates on the night of 3/4 September 1990, the local police fabricated evidence to ensure that Khoza and the 137 IFP supporters arrested with him could not be linked to the firearms found in the boot of Khoza’s vehicle and apparently used in the massacre. According to Vlakplaas operatives, the weapons found matched
those they had provided to Khoza the previous day, while Khoza’s car was provided by the Security Police. Vlakplaas also allegedly put up Khoza’s bail money.

23. Amnesty was granted to the head of the Vaal Triangle Security Police, Jacobus Francios Conradie (AM4123/96), who admitted to ‘defeating the ends of justice’. The officer investigating the massacre, the head of the Vaal Triangle Murder and Robbery Unit, Jacobus Jacobs (AM 4373/96), and an officer at the scene of the crime, Arthur John van der Gryp (AM 4146/96) were also granted amnesty. Conradie denied that his actions to assist Khoza were approved or authorised, but claimed that he had acted unilaterally when he found out how important Khoza was to the police.

24. While the three amnesty applicants’ versions of events largely corroborate one another, other important issues that are not thoroughly covered in their applications saw the light of day at the Section 29 in camera hearings. Regrettably, these amnesty applications were heard in chambers, preventing any further opportunity to explore the case and its broader implications in terms of collusion between the security forces and the IFP.

25. Although no admissions have been made by the IFP regarding these allegations, several investigations undertaken by the National Department of Public Prosecutions are believed to have reached an advanced stage, indicating that therere is prima facie evidence against certain individuals.

26. Disclosures made re g a rding the Sebokeng incident support the assertion that ‘ ‘third force’ elements were at play. Not only did one of the security forces’ most ‘successful’ counter- insurgency units supply weaponry to the Inkatha attackers, but the police were also successful in protecting one of the most prominent Inkatha leaders in the region in the legal process following the massacre.

27. Consistent allegations that Themba Khoza and other IFP leaders in the region were involved in the distribution of weapons and had regular meetings with security forces re p resentatives such as Eugene de Kock further supports the findings in the Commission’s Final Report.

28. The Commission thus finds that, while little evidence exists of a centrally directed, coherent or formally constituted ‘Third Force’, a network of security and ex-security force operatives, frequently acting in conjunction with right-wing elements and/or sectors of the IFP, was involved in actions that could be construed as fomenting violence and which resulted in gross human rights violations, including random and target killings.

29. Allegations of ‘third force’ activity reached a crescendo in the wake of the Boipatong massacre in June 1992. The Commission did not undertake detailed investigation into all allegations of security force complicity. Instead it relied on a number of reports submitted to it by monitoring groups who went into Boipatong immediately after the massacre and compiled reports based on the testimony received. The Commission made a series of detailed findings in which it alleged that there had been direct collusion between the security forces and the IFP, and that the security forces’ direct participation in the massacre was alleged. Conversely, the Amnesty Committee accepted the version of most amnesty applicants, who denied any security force involvement in the massacre , and rejected the evidence of Andries Matanzima Nosenga, the one applicant who supported victims’ assertions that the security forces were complicit. The Amnesty Committee did, however, leave open the possibility of security force complicity: it acknowledged the victims’ allegations, while accepting that there was no evidence to connect the amnesty applicants with them. The Commission does not accept that allegations about this complicity were fabricated.

30. An analysis by the Commission’s Investigation Unit into available material relating to the massacre and subsequent investigations raises a number of serious shortcomings or issues that raise doubts about the credibility of the police investigation into allegations of police involvement. Within two weeks of the massacre, for example, the SAP officer tasked to investigate the allegations reported that they had been proven to be untrue. The Commission’s enquiries established that approximately fifty witnesses testified about the direct or indirect involvement of the security forces. It appears that several of the witnesses were either ignored or deliberately not presented. A number of other shortcomings were presented in the Investigation Unit report.

31. October 1992 seems to have been a turning point for ‘third force’ activities. The Goldstone Commission’s discovery that a Military Intelligence operation against the ANC was still operational led to the appointment of General Steyn by President de Klerk to investigate the allegations of armed forces involvement in the violence. At the same time, Colonel de Kock was approached by his superiors and asked
to resign from the police force.

32. Although General Steyn’s and subsequent investigations were largely
inconclusive in terms of proving SADF involvement in a wide range of illegal and/or unauthorised activities – including ‘third force’-style attacks – the pall of suspicion and incriminating evidence has not been lifted. Analysis of violence statistics indicates that typical hit-squad attacks declined after mid-1992 while the level of political violence rose steadily until 1994.

33. By the 1990s, patterns of security force conduct that crossed the boundaries of legality emerged. This conduct was condoned and in some instances encouraged. A network of security force operatives – bound by oaths of blood and secrecy – had been developed, with informal channels of communication and in possession of, or with access to, material resources and weaponry. While the new De Klerk government significantly dismantled many formal securocrat structures, little obvious attempt was made to dismantle these networks or to change the mindset of operatives’ intentions to continue an all-out war on the ANC and its allies. Indeed, where efforts were directed at uncovering such networks – as with the establishment of the Harms Commission – security force personnel were instructed by their seniors to lie, sending a clear signal that these were simply public relations initiatives and that they were not intended to change the status quo. The fact that Vlakplaas personnel continued with unlawful activities at the very time that the Harms Commission was sitting is clear testimony to this. Given this background, it is unsurprising that evidence emerged of security
force involvement in the violence and destabilisation of the 1990s.

34. Various explanations for the violence in the early 1990s have been proffered that do not necessarily take into account the role of the security forces. These have included assertions that the violence was essentially symptomatic of the rapid political change that was taking place, fuelled by ethnic divisions and socio-economic pre s s u res. Certainly, existing tensions between Inkatha and the ANC were exacerbated, fuelled by the mobilisation of ethnic and political tensions. There is also evidence that individuals and groups in some areas were targeted on the basis of their ethnic background. But such ethnic division was far from being monolithic or monocausal in its manifestations.

35. While such explanations for the violence addressed critical aspects of why it was happening, they were largely unsatisfactory in that they failed to engage either with the specific dynamics of violence in particular communities or with emerrging patterns and trends. Violence often appeared to take on a life of its own – cycles of revenge often triggered by indiscriminate and unprovoked attacks. Assumptions of guilt and responsibility were manifest. Attempts to instil reason and constraint were drowned by calls for action. The need to defend frequently
and rapidly against attacks metamorphosed into offensive action.

36. There were also understandable concerns that the picture being drawn by many commentators and analysts was manipulated and skewed – unconsciously or even deliberately (as much of the internecine conflict in certain communities in the 1980s had been) – as part of a broader depiction of ‘black on black’ violence. This was interpreted in some quarters as part of a deliberate strategy to undermine the ANC alliance and the broader objectives of democratic transformation by creating the perception and imagery that political opportunities for the black majority would result inevitably in conflict and violence. As such, violence was interpreted as a political tool in the power play for the negotiated settlement.

37. We are there forerepresented with a spectrum of views about the violence and ‘third force’ allegations. These range from benign interpretations of government and security force action at one level to allegations of a specific agenda to destabilise political opponents at the other.

38. Within this spectrum of views, a host of important questions and issues has been raised, many of which the Commission was not able to address adequately, due to a lack of re s o u rces and time constraints. Other relevant structures have not really addressed the outstanding issues either. Unresolved issues include an analysis of exactly what the government did to address the violence, and whether its response can be classified as adequate or reasonable; the extent to which the government had lost control of its security forces; a detailed analysis of how the security forces reacted to and addressed violence in various locations,
and the role of leadership and the rank and file, their attitudes and the limitations and obstacles to their work.

39. During the late 1980s, the security forces intensified their counter-insurgency efforts at the same time as the government was developing its preliminary contacts with the ANC. Repression and violence were evidently an integral component of a broader political strategy at this stage. However, the extent to which this was carried over into the negotiations period of the early 1990s is less clear.

40. Understanding the characteristics of violence in the 1990s, therefore, requires a more nuanced appreciation of security force practices and policy within the context of political change and the limits of institutional transformation that accompanied this.

41. While the involvement of security force individuals and structures in ‘third force’ violence was to some degree corroborated, the quality and quantity of available evidence, whilst significant, is generally thin. No detailed or focused investigations were initiated; few amnesty applications were received, and lines of command and accountability were not established.

42. The Commission was also unable to establish the extent of covert networks and how they evolved and mutated during the conflict period. Consequently, it is not clear whether the senior security force personnel involved represented their own, state or right-wing agendas. In a rapidly changing political situation with shifting alliances, the Commission accepts that it is probable that there were several agendas involved, at different levels within political and security force hierarchies.

43. In this context, explanations for and allegations relating to the dynamics of and rationale behind ‘third force’ activities remain vociferously contested. The Commission believes that more light must be shed on the role and activities of the military and the police, and especially on the role of covert and other specialist units during the violence in both the 1980s and 1990s. Ongoing researchrch suggests that there is considerably more to be uncovered in this respect. The Commission there f o re believes that further enquiries and investigations regarding ‘third force’ allegations are an essential part of a broader process in terms of developing our understanding of past conflict and those responsible for it.