About this site

This resource is hosted by the Nelson Mandela Centre of Memory, but was compiled and authored by Padraig O’Malley. It is the product of almost two decades of research and includes analyses, chronologies, historical documents, and interviews from the apartheid and post-apartheid eras.

Violence and the State

National Party Policies

The original cause of the violence was the National Party (NP) government. When black people demonstrated against the pass laws at Sharpeville in 1960, it mowed them down with sten guns. When schoolchildren demonstrated against Bantu Education in Soweto in 1976, it shot them down too. It banned and house arrested people after stigmatising many of them as communists, and several dozen of them it beat to death or drove to suicide in detention. Not content with banning people it closed down newspapers and banned numerous political organisations. The more it drove black opposition underground and into acts of desperation and violence, the bigger its majorities got in parliament. During the worst of these years-the 1960s-the whites who objected to all of this were so few in number that they managed to scrape together enough votes to send precisely one MP to Cape Town-Mrs Helen Suzman. When she opposed some of the worst of the detention laws, she did so alone, for the then official opposition, the United Party (UP), was more concerned with fighting elections on slogans like 'Want TV? Vote UP'. For most of the time all these outrages against black people had the backing, implicit or explicit, of some of the business community, the Afrikaans press, and, of course, very important in a supposedly God-fearing country, the Dutch Reformed Churches. Those few clerics who objected, like Ds Beyers Naudé and Professor Albert Geyser, were branded as heretics. Among the organisations stigmatised was the South African Institute of Race Relations, whose 60 years of annual Surveys and other publications are the Encyclopaedia Britannica of the apartheid system. They constitute a meticulous, factual, and damning indictment, unrivalled anywhere.

The darkest days of apartheid are gone, but the legacy and the wounds remain. They cannot be eliminated by a single speech, no matter how decisive a break State President F W de Klerk tried to make with the past when he opened parliament on 2nd February 1990. The NP's track record helps to explain why so many people believe the allegations that the government is orchestrating a 'third force' consisting of the South African Police (SAP), the South African Defence Force (SADF), the Inkatha Freedom Party (IFP), vigilantes, and miscellaneous 'hit squads' to destroy the African National Congress (ANC). The theory has such wide currency both in South Africa and abroad that it merits examination in some detail.

The 'Third Force' Theory

The president of the ANC, Mr Nelson Mandela, has been among those making the accusations. For example, he said in October 1991 that Mr de Klerk had 'let loose his hounds against the people'. In an address on 15th July 1992 to the United Nations Security Council in the wake of the Boipatong massacre, in which more than 40 people were killed, allegedly by IFP-supporting hostel residents, a month previously, Mr Mandela spoke of a strategy of 'state terrorism' in which the IFP had 'permitted itself to become an extension of the Pretoria regime, its instrument and surrogate'. In December 1992, Mr Mandela referred to 'the fact that the state security services, using certain black organisations, have been responsible for the death of no less than 15 000 people since 1984'.

The Commission of Inquiry Regarding the Prevention of Public Violence and Intimidation (the Goldstone Commission), which was established in 1991, has said no evidence has been submitted to it to substantiate accusations that the government is orchestrating violence, but they persist. Speaking in September 1992 Mr Chris Hani, former chief of staff of Umkhonto we Sizwe, the military wing of the ANC, and then general secretary of the South African Communist Party (SACP), said that the government was pursuing a twin-track strategy which involved negotiation with the ANC and 'at the same time deliberate [and] violent destabilisation' of its forces. Mr Hani said that whereas the ANC-led alliance had not deployed violence strategically since 1990, the government was deploying it systematically as 'a central component of a political strategy' making use of proxy forces including 'Contras, Unitas, Renamos, Askaris, Lesotho Liberation Armies, the Ciskei Defence Force etc backed up by small, special forces and seconded officers from the main army'. Mr Hani went on to say that violence-in train massacres, for example-was being 'switched on and off', and that 80% of the victims were ANC members or supporters. The government's objective was political rather than military, he said. Its aim was to 'impose a political settlement on a war-weary people'.

Such allegations should not lightly be dismissed. The conviction in 1992 of a white police officer, Captain Brian Mitchell, and four black non-commissioned officers for the massacre of 11 people in December 1988 in the village of Trust Feed near Greytown in the Natal Midlands was not the first time black and white policemen had been found guilty of murder. The murders were, moreover, part of a strategy to promote the IFP at the expense of the then-banned ANC's de facto internal ally, the United Democratic Front (UDF) (although the killers murdered IFP supporters by mistake). The judge also found that the initial police investigation into the Trust Feed massacre was so incompetent as to suggest an attempt at concealment, possibly in top police ranks. He called for an independent inquiry into the investigation and the possible involvement of police officers in 'counter-revolutionary strategies'.

In another case, seven IFP members were acquitted for lack of evidence on 38 counts of murder at a massacre at an ANC funeral vigil in Sebokeng in 1991. The judge said the legal process had failed to bring the murderers to justice, although he pointed out that the police had said the ANC had prevented them from speaking to some of the witnesses. The Association of Law Societies found it necessary to criticise the fact that a member of the KwaZulu police was released from prison in February 1992 after serving less than a year of a 27-year sentence for murder despite having been described by the judge as a 'beast in policeman's clothing' who seemed to enjoy his reputation as a 'hit man' and the fear this caused in the KwaMashu township outside Durban. Other members of the IFP, including a KwaZulu deputy minister, have been convicted of murder. In April 1993 four IFP members were convicted of smuggling arms into South Africa from Mozambique.

Yet another cause for suspicion is the publication of a document implicating the State Security Council and top military officers in the 'permanent removal'-and alleged murder-of political activists, among them Mr Matthew Goniwe, in the Eastern Cape in 1985. At the time of writing, an inquest into the murders was in progress. The government's insistence in October 1992 on legislation that could be used to indemnify state employees (and others) who may have committed acts of violence has done little to diminish suspicions, especially since details of the crimes in question may be kept secret.

After seizing a number of files from a covert military operations unit on 11th November 1992, the Goldstone Commission said that 'highly placed members of military intelligence' had been party to the employment of Mr Ferdi Barnard, 'a notorious former member of the Civil Co-operation Bureau' (CCB), to obtain information about Umkhonto we Sizwe, and in particular its involvement in crime. The chairman of the commission, Mr Justice Richard Goldstone, said that Mr Barnard had two previous convictions for murder, one for attempted murder, and three for theft, and that the SADF had denied employing him when this was 'quite untrue'.

Mr Barnard was alleged in the press to have been involved in the murder of a UDF activist and university lecturer, Dr David Webster, in May 1989. In January 1993, however, Mr Justice Michael Stegmann came to the rather unsatisfactory conclusion that responsibility for the crime could not be pinned on any person in particular. Although Dr Webster's death had been brought about by acts constituting a conspiracy to murder, Judge Stegmann said, it could not be conclusively proved that Dr Webster had been an 'officially sanctioned CCB project' or that the organisation or any of its members had killed him. This was despite the fact that evidence before him had indicated that the CCB included as part of its programme the proposed elimination of certain 'targeted' people. Evidence incriminating Mr Barnard was circumstantial, the judge said, and he was 'at this stage entitled to the benefit of the doubt'. Judge Stegmann added, 'For the present the truth has not been brought to light.'

The CCB, which was formally constituted under that name in 1986 in the defence force, was widely suspected of murdering ANC supporters and other anti-apartheid activists. It was formally disbanded in 1990, various attempts to penetrate the secrecy surrounding its activities having failed, partly as a result of the disappearance of some of its records. At the time of writing it seemed likely that many of the mysterious deaths of anti-apartheid activists down the years would never be finally cleared up. After the Webster finding, an ANC official said the organisation reserved the right to re-open all inquests into political assassinations when it came to power.

Mr de Klerk has admitted the involvement of security force members in violence. While denying the existence of any 'third force', he said in November 1992 that 'obviously there might be individuals who have their own agendas'. A month later, announcing that 23 senior officers of the defence force were being retired or sent on leave, following investigations by the Goldstone Commission and by the SADF's chief of staff, Mr de Klerk said SADF personnel had been or were still involved in illegal activities, including, apparently, activities that 'led to the deaths of people'. The ANC interpreted this statement as confirming its contention that there was a 'third force' operating within the security forces, but Mr de Klerk repeated his denial of such a force and stated that only a limited number of persons and a few units were involved. At the time of writing is was not clear whether subsequent investigations would result in prosecutions.

The Sunday Star reported soon after the departure of the 23 officers that more than 60 former Rhodesians had been dismissed in the 'latest government purge' of military intelligence. One was quoted as saying, 'They used us over all those years. They were happy to get us to carry out their dirty work. Who do you think they used to do eliminations? And now they want to dump us.' The newspaper said that the ex-Rhodesians were planning to bring lawsuits against the government and the SADF. A former managing director of the CCB, Colonel Joe Verster, said nobody in the organisation had done anything without authority.

More is known about shootings by the police than about what military personnel may have done. The police have killed more than 560 black people in political demonstrations since the shootings at Sharpeville in 1960, and possibly as many as 660. Most of these were shot dead, some in the back, in the riots in Soweto and elsewhere which began on 16th June 1976.

As far as is known, no top-ranking police officer or politician has paid the price for the Sharpeville or Soweto or any other police shootings. Instead of firing the general or the minister, National Party governments have put indemnity legislation through parliament. Almost inevitably, a culture developed in which some members of the police force thought themselves above the law.

Nor is it surprising that several dozen people have died in detention under security laws.

The first such death on record is that of Mr Solwandle Looksmart Ngudle, a trade unionist and ANC member who was said to have committed suicide by hanging himself in Pretoria in September 1963. Six years ago the Institute published a study of deaths in detention, listing across several columns the name, date of death, place, age, and so on. There are 63 names on this list between 1963 and 1984. One column is headed 'Criminal Proceedings'. The word 'no' appears no fewer than 59 times in this column. If governments enact laws empowering policemen to detain people indefinitely beyond the reach of lawyers, courts, private doctors, families and friends, and those people keep dying and very few prosecutions are brought against the police, then the message that the government is in effect sending to the security police is that it is acceptable to torture people. How different South Africa's history might have been if Mr John Vorster, minister of justice in 1963, had been forced to resign when Mr Ngudle died. Mr Ngudle might have been the only South African to die in this way.

Even though the powers of the police to detain people without trial have been severely curtailed (from an indefinite period to a maximum of ten days), people continue to die at a high rate in police custody-153 in 1991 and 210 in 1992, according to the government's own figures, many of them people arrested on suspicion of offences unrelated to politics.

South Africa has a problem of lawless and proven criminal behaviour on the part of policemen, 1 871 of whom were convicted of criminal offences in 1991 (while 5 314 were disciplined for departmental offences).

The 'third force' theory is plausible from a political point of view. Grand apartheid sought to divide and rule the black population, and for the National Party to seek an alliance with the more conservative black parties against the radicals and then to exploit ethnicity to undermine other parties would have a kind of logic about it. Suspicions are reinforced by the secret funding of the IFP and an IFP-linked trade union, the fact that the police have been less than zealous in preventing IFP members from brandishing dangerous weapons in public, and the fact that the Goldstone Commission found it necessary to take the government to task for inadequately implementing its recommendations to curb this practice.

Scrutinising the 'Third Force' Theory

It is necessary to probe a little more deeply into some of the theories or reports which, explicitly or implicitly, attribute violence in South Africa almost exclusively to the state or even see violence as the result of a government strategy to destroy its opponents.

The Community Agency for Social Enquiry (Case) and a group known as the Human Rights Commission (HRC) have claimed that the government switches violence on and off at 'key moments'. They say it turns violence up when it would weaken the ANC and its allies. The HRC and Case thus claim that violence is turned up during ANC conferences and campaigns. Their reports fail to mention several conferences and campaigns when violence dropped. The HRC and Case further argue that the government switches violence off when Mr de Klerk travels abroad. Their reports fail to mention overseas trips by the state president that coincide with increases in violence. Case says that the 'government-sponsored peace conference in June [1991], boycotted by the liberation movements but attended by Inkatha, saw a major diminution of violence'. June 1991 did see a major drop in violence. But the conference was in May 1991-a month which, according to Case's own figures, had the third highest incidence of fatalities during the period under consideration. A forensic specialist at the medical school of the University of Natal involved with autopsies of victims of violence in Natal said he had done a week-by-week analysis of incidents of violence and also of political developments from 1989 to 1992 but had found no correlation between the two.

The HRC has also contended that 'vigilante-related causes' accounted for 83% of all politically-related deaths in one 12-month period. Vigilantes were said to be acting 'at the government's command'. But the HRC used the term 'vigilante-related action' in such a way as to blame on 'vigilantes' not only attacks which they may themselves have carried out but also attacks in which 'vigilantes' were killed. This definition was so wide as to tar both victims and perpetrators of violence with the same brush. In one report the HRC even blamed 'vigilantes' for murders that by its own admission were carried out by 'unidentified' and 'unknown' attackers. Without citing evidence it blamed vigilantes for no fewer than 49 killings in a single month, among them necklace executions, when this method of killing is often employed by militant 'comrades' rather than by 'vigilantes'. Subsequent to criticism by the Institute of the attempts to blame nearly all violence on 'vigilantes', the HRC stopped using this categorisation. Whereas the HRC said 71,3% of all politically related fatalities in 1992 were 'vigilante-related', and that 'private armies' of vigilantes 'of late have considerably widened the scope of their attacks to destabilise township communities at large', its report for February 1993 made no mention of vigilantes. Most fatalities (76%) were now said to have occurred in 'general incidents'.

In July 1992 the HRC issued a report on massacres since 1990, among them 11 massacres in Natal since July that year. One of the massacres occurred in June 1991 in Richmond in the Natal Midlands. According to the commission, 16 ANC supporters were killed by the IFP on this occasion. However, there was another massacre in Richmond three months before that, at the beginning of April. On this occasion not 16 but double that number of people were killed, most of them apparently IFP supporters. The HRC makes no mention of this massacre, even though it says it looked into its records of more than 5 000 politically-related incidents and even though it defines a massacre as an incident in which ten or more people died.

Not only is the report selective in which massacres it mentions, but it in effect pronounces certain parties guilty of genocide on the basis of what it describes as media reports of 'allegations or deductions sometimes corroborated by legal evidence'. No court would use such methodology to convict even a minor offender.

A book on 'apartheid death squads', published by a journalist, Mr Jacques Pauw, in 1991 contains a list of 225 'political assassinations of anti-apartheid activists' between February 1971 and February 1991. Mr Mandela has relied on this list to back his accusations that Mr de Klerk is an accomplice to murder. It is almost certain that 'hit squads' exist. Captain Mitchell and his four subordinates could be described as one, although there is so far no evidence that they were acting on higher orders. Dr Webster may also have been a victim. Given the gravity of the allegations sometimes made about 'hit squads', however, it is striking that some of the assumptions on which the allegations rest are dubious and some of the information incorrect.

For example, the last victim listed in the book on 'apartheid death squads' is Chief Mhlabunzima Maphumulo and his alleged killer is listed as one Sipho Madlala. Mr Madlala claimed to have been attached to the military police since 1986, that the murder had been plotted from the Alexandra Road police station in Pietermaritzburg, and that members of the hit squad of which he was part had each been given R5000 for their part in the murder. The ANC said Mr Madlala had shown that government-controlled hit squads continued to assassinate people to 'stoke the flames of conflict and lay the blame at the door of Inkatha'. In the event, Mr Justice Page found that unknown people were responsible for the killing and that no reliance could be placed on Mr Madlala's story, which had been contradicted by every other witness and a 'vast body of documentary evidence'. The evidence of a second witness who claimed to be a 'hit squad' member was found to be 'appalling' and 'even worse' than that of Mr Madlala.

Mr Pauw's list describes as 'unknown' the killer of Mr Saul Mkhize, who was shot dead in April 1983 at Driefontein near Piet Retief in the south-eastern Transvaal. However, the man who shot him was a 21-year-old police constable called Johannes Andries Nienaber whose photograph appeared in the press and who was acquitted a year later on a murder charge when the circuit court in Volksrust accepted his plea of self-defence-a verdict greeted with outrage. Mr Mkhize was shot dead in broad daylight in front of a large number of witnesses by an easily identifiable policeman wearing a uniform-not the way a 'hit squad' would operate.

The HRC attributed the murder of Mr Andre de Villiers in Addo in the eastern Cape in August 1992 to 'hit squads', which it defined as 'faceless professional hitmen whose objective is to assassinate identified political figures and/or cripple anti-apartheid organisations'. In April 1993 one member of Umkhonto we Sizwe and another man were convicted of the murder.

The HRC and Case reports and the 225-name list mentioned above are not the only documents that are open to question. In June 1992 Amnesty International produced a report which correctly quoted the Goldstone Commission as saying it had received evidence of police bias towards the IFP at the time of a massacre in Bruntville near Mooi River in Natal in December 1991, but failed to mention the Commission's finding that there was no unlawful conduct by the police.

Similarly, a press release issued after a visit to South Africa in 1992 by the International Commission of Jurists (ICJ) carries a section headed 'dangerous weapons' but this section makes no mention of any weapons other than 'lethal traditional weapons' carried by members of the IFP. Indeed, nowhere in the ICJ's press release is there any reference to any other particular kind of weapon, a surprising omission in view of the fact that most incidents of violence involve firearms, petrol bombs, and hand grenades. The ICJ calls for hostel residents to be disarmed but makes no recommendation that anyone else should be. The ICJ describes the KwaZulu police as a private army. No reference is made to any other private army. Like Amnesty, the ICJ quotes selectively from the Goldstone Commission and fails to mention its finding that the police were not to blame for the Bruntville violence.

Amnesty, the ICJ, the HRC, and Case all have in common the view that violence is to be blamed mainly, if not almost entirely, on the government and the IFP. The involvement of both of these in violence is beyond question. However, one of the other things that Amnesty, the ICJ, the HRC, and Case have in common is that they fail to deal with violence directed against agents of the government and the IFP. Indeed, the problem with their reports, and with the 'third force' theories, is not what they do explain but what they do not explain. They do not explain why, according to the IFP, more that 1000 of its members have been killed in violent attacks, why 275 of its office-bearers and leaders have been assassinated since 1985, or why the police have made so few arrests in connection with these killings, which the IFP describes as 'serial murders'. Nor do these theories explain the violence directed against the police, nearly 700 of whom have been killed on duty in the last six years (although not necessarily always in political situations).

The 'third force' theory has been doggedly promoted by The Weekly Mail. In an issue in July 1992 it ran an editorial attacking first as 'meaningless' and then as 'political' Mr Justice Goldstone's statement that 'no evidence' had been submitted to his commission 'which in any way justifies allegations of any direct complicity in or planning of current violence by the state president, any member of the cabinet, or any highly-placed officer in the South African Police or Defence Force'. The editorial stated: 'This newspaper for one has spent a great deal of time, energy and money assisting Justice Goldstone in his investigations.' The Weekly Mail was nevertheless unable to provide proof. One of the co-editors of the paper admitted in August 1992 that its 'third force' allegations were based on 'patchy evidence' which is 'not always reliable'.

Two other journalists who have written about the 'third force', Messrs Jacques Pauw and Shaun Johnson, said in August 1992 that the only real evidence forthcoming was the Trust Feed case-which was finally broken by a determined police officer. The definition of 'third force' then began to be changed-'third force activities' would now be 'any illegal activities which have the effect of stoking violence and instability'. This definition of the 'third force' is radically different from some of the earlier versions, which depicted the 'third force' as acting upon the command of the government itself. On the new definition Umkhonto we Sizwe, the Congress of South African Trade Unions (Cosatu), the ANC and the SACP join the IFP, the SAP, and the SADF in the 'third force'. The term becomes so wide as to lose all meaning-rather like Mr Vorster's definition of 'terrorist'. Mr Vorster also introduced laws to presume guilt until innocence was proved. On this too the left is beginning to resemble the right. The Weekly Mail's rebuke to Mr Justice Goldstone was, in effect, saying to him, 'We have already pronounced these people guilty. Why do you bring up such a "meaningless" factor as our inability to produce evidence?'

It should nevertheless be borne in mind that governments involved in covert or criminal activity can use further such activity to cover their tracks. Some of the covert operations of NP governments have been exposed over the years, but there may be others as yet uncovered. It is conceivable that the 'third force' contains agents provocateurs who are killing IFP members and policemen by the hundred in order to provoke the IFP and the police into acts of violent retaliation against communities of ANC supporters. This would set up a vicious circle of attack and counter-attack which becomes self-perpetuating. It has been argued that this violence would in turn serve the interests of the government by derailing negotiations leading to majority rule and so leave the NP in power. So Machiavellian a strategy would be compatible with the known dishonesty, corruption, and brutality of apartheid rule, so it is as well to keep an open mind.

It is consistent with the NP's divide-and-rule policy that it might have sought, as a matter of deliberate strategy, to foment conflict between the ANC/UDF/Cosatu alliance on the one hand and the IFP on the other. An alliance between the ANC and the IFP would have presented a formidable challenge to the NP in constitutional negotiations. Instead, the ANC and the IFP have been at loggerheads. However, the theory that the conflict between the ANC and the IFP has been fomented by the government in order to weaken the ANC at the negotiating table overlooks the fact that the conflict goes back to the 1970s-long before the NP ever contemplated negotiations with the ANC. At that time NP policy was to hive off all ten of the homelands into constitutional independence of the rest of South Africa, not negotiate with black organisations for black people to be brought into parliament.

Without dismissing the possibility that a state-orchestrated 'third force' strategy does exist, this paper will put forward another explanation of the violence. Since it relies in part on material which could not lawfully be published in South African newspapers prior to 2nd February 1990, it will require more elaboration than the 'third force' theory, which has had such wide currency in the local and foreign media.

This explanation is put forward not as an academic exercise but because a full understanding of the violence and the dynamic that continues to drive it is a pre-condition of bringing it to an end. The explanation will widen the picture, by adding to it what the 'third force' theories usually omit-the strategies adopted in the struggle against apartheid, and the moral climate in which they were adopted.

This resource is hosted by the Nelson Mandela Centre of Memory, but was compiled and authored by Padraig O’Malley. Return to the Nelson Mandela Centre of Memory site.