Centre for the Study of Violence and Reconciliation

From the SADF to the SANDF:
Safegaurding South Africa for a better life for all?

by
Noel Stott

Violence and Transition Series, Vol. 7, 2002.

Noel Stott is a former Researcher at the Centre for the Study of Violence and Reconciliation.

Contents:

Acknowledgements

Thank you to all the participants who shared their time, stories and information.

Deep felt thanks go to all who participated in this research project, including those interviewed and those who offered advice and encouragement. Thank you to Wardie Leppan, Project Officer, International Development Research Centre (IDRC), for his ongoing support of, and insight into, the Violence and Transition Project. Special mention of Piers Pigou must be made not only for his editorial input but more importantly for his enduring tolerance. I am also indebted to Bronwyn Harris for her support, editorial and otherwise.

Introduction

South Africa's military became synonymous with repression during the height of the anti-apartheid struggle. It was powerfully influential within the state's security structures, and these in turn carried much authority within political decision-making circles at the time. Eight years after South Africa's democratic transition, however, the South African military is held to be one of the most, if not the most, successful state sector to deal with transformation.

This paper focuses on violence in relation to the South African Defence Force (SADF) in the 1980s and early 1990s, and the South African National Defence Force (SANDF) in the post-1994 dispensation. During this latter period, the military has undergone significant change in the context of South Africa's broader transition from authoritarian to democratic rule.

While the key military protagonists of South Africa's apartheid conflict are no longer at war, the legacy of the recent past will not be overcome simply by the laying aside of arms. Neither the old SADF, nor the liberation armies, were created or structured to serve in a democracy and none had an unblemished history of respecting human rights (Nathan, 1994a). For these reasons, it is of the utmost importance that potential problems and pitfalls in the drive to consolidate democratic gains, and fundamentally transform the armed services, are adequately addressed.

This paper explores the role of the apartheid state's military force in South(ern) African society and the consequent rise in militarism in the 1980s. It also raises the question of whether, and if so how, this past function contributes to the high levels of conflict and violence that, to a large extent, continue to plague post-apartheid South Africa.1 This requires an examination of past manifestations of violence perpetrated by the SADF, and "new" forms of violence within the SANDF, including assaults; gender, domestic and racial incidents; suicides; and violent property crimes, such as thefts and armed robberies. This, in turn, requires an examination of how processes of transition, integration, rationalisation and demobilisation have impacted on, or resulted in, new or different forms of violence.

Internationally-recognised independent monitoring groups such as Amnesty International (AI) and Human Rights Watch (HRW) continue to report on alleged abuses committed by the South African military. For example, AI asserted in their 2001 survey that "there were reports of torture, ill-treatment and unjustified use of lethal force by the security forces, including military units based in KwaZulu-Natal province, primarily in the context of crime investigation and the search for illegal weapons" (Amnesty International, 2001, p. 220).

Potential sources of conflict that threaten South Africa's democratic transition are manifold, and many are related to the legacies of apartheid, especially in terms of socio-economic factors and massive structural inequalities. However, that threats to security and stability have also emanated from within the armed formations of both statutory and non-statutory forces underscores the importance of securing institutional control across the board. Both the process of integrating the various armed formations into a single defence force, and the resultant end product had the potential to cause conflict. Not only were sworn enemies expected to work together, but also affirmative action, down-sizing and demobilisation had the potential to stimulate resentment and further conflict.

While a number of specific violent incidents have indeed taken place and integration and demobilisation have caused considerable tensions and conflicts, this paper strongly suggests that systematic violations no longer occur within the military and that the South African military is no longer considered to be a perpetrator of gross human rights abuses against South Africans more generally. The violent culture that permeated the SADF is apparently not endemic in the SANDF.2

Methodology

In addition to an extensive review of the available literature, the author conducted in-depth discussions and interviews with a number of leading military and sociological analysts and with former members of the SADF and the so-called non-statutory forces, MK, APLA, as well as with serving members of the SANDF.

Limitations

Understanding the past and present requires access to information. Many former and present members of the SADF/SANDF and non-statutory forces were reluctant to be formally interviewed and quoted. Much of this paper is, therefore, based on private, off-the-record discussions and secondary source material. Official documents were difficult to obtain, mainly because of the systematic destruction of material that took place prior to, and during, the transition period - a process described in some detail below. Potentially crucial material that is known to have been destroyed includes records of the National Security Management System (NSMS), and of bodies associated with the South African Defence Force (SADF) such as the South West African Territory Force (SWATF) and other special directorates and units such as the Directorate of Special Tasks (DST) and the Civil Co-operation Bureau (CCB).3 Nevertheless, the problem of access to information for empirical research is not a new phenomenon being an obstacle encountered throughout the 1980s and before, especially when attempting to analyse security-related events (Sarakinsky, 1989, p. 71).

In stark contrast to the secrecy of the past, the advent of democracy in South Africa has introduced an unprecedented level of openness about defence matters. Nonetheless, there is still considerable uncertainty within the SANDF itself as to what information should be readily available to the public. "The question of public access to defence information in a democracy in a time of peace is inevitably characterised by a tension between the imperatives of transparency and accountability on the one hand, and the legitimate need to preserve the confidentiality and secrecy of certain information on the other" (De Jager, 1999). Paradoxically then, the national interest in defending democracy becomes the justification for limiting individuals' and society's rights to information. In this way, the process of transition itself has contributed to a lack of clarity regarding the violence of the past.

Another important factor inhibiting access to information was that prior to 1999, record keeping was decentralised within each arm of the military. While this has now changed and a process is underway to centralise information, cumulative statistical data covering the years before, during, and shortly after the transition are not readily available.

Our insight into how the military operated was further constrained by the limited availability of detail on the previous state's security structures and the evolution of their relationship to the police and broader political structures.

Despite ongoing official secrecy about the SADF's involvement in past violence, a vast amount of information has emerged. Rather than a comprehensive analysis of such violence, this report offers an overview of violence surrounding the SADF and the SANDF (primarily through the lens of the TRC). When read in conjunction with other VTP reports (especially those dealing with ex-combatants and the police), this "big picture" overview complements the analytical framework of the CSVR's VTP project and offers a starting point for future analytical research.4

A Brief Overview of the South African Military5

The SADF's predecessor, the Union Defence Force was created in 1912 following the establishment of the Union of South Africa in 1910. Prior to the 1960s, expenditure on defence was not a priority for the statem,6 by the mid-60s, the arms budget had increased significantly in response to a United Nations (UN) arms embargo, and an Armaments Development Corporation (ARMSCOR) was established. By the mid-1970s, South Africa "had a powerful and significant arms industry", and by 1979 over 14% of state expenditure was allocated to defence. In 1982-83 expenditure had risen to almost R3 billion, and by the mid-1980s almost 20% of the government's total annual budget was devoted to defence (Saunders, 1983, p. 111; Frankel, 2000, pp. 149-50). Despite a prolonged period of militarisation in other countries in the Southern African region, the SADF's military machine was still regarded as more powerful than the combined forces of all its neighbouring states (Heitman, 1985, p. 8).

By the early 1990s, the SADF consisted of 75 479 full-time members, 21 695 civilians and a part-time component of 526 702 citizen-force personnel, as well as 76 476 commandos (primarily the white reserve force) organised into 200 units.7 In addition, the four "independent" homelands of Transkei, Venda, Bophuthatswana and Ciskei (the TVBC states) also had small defence forces, nominally independent, yet effectively under SADF control.

Historically, the South African military has a range of combat experience and has participated in a number of conventional wars, including the First and Second World Wars. South Africa was one of 14 countries to offer military assistance in the Korean War and, "as a gesture of solidarity against communism" in 1950, sent its 2nd Air Force Squadron to Korea, where it served until 1953.8 From 1966 to 1974, South Africa also provided "policing services" and military support to neighbouring white colonial governments in Rhodesia (now Zimbabwe), Mozambique and Angola and, as will be noted later, engaged in a number of "low intensity" military deployments, incursions and cross-border raids against these and other "Front Line States" on a regular basis from 1975 to 1989. With respect to Rhodesia, the South African House of Assembly passed a defence amendment bill in March 1973 making it possible for South Africans to be seconded to the Rhodesian forces without loss of seniority or pay.9 Moreover, South Africa was also covertly involved in a number of attempted coups, either directly or by supporting so-called "liberation" armies or dissident military personnel and supplying weaponry. These include the 1981 Seychelles coup attempt, and other initiatives in the late 1980s and early 1990s, in the Transkei and Ciskei as well as the 1986 economic blockade of Lesotho which facilitated a Major General Lekhanya deposing Chief Leabua Jonathan's government and who immediately deported approximately 100 ANC members (Stiff, 2001).

The defence force also has experience of operational deployment inside the country. As early as 1914 and 1922, troops were used to suppress strikes by white mine workers. During the 1980s, the South African military also took on increasing responsibility for internal security matters, providing the South African Police (SAP) with considerable support in their attempt to crush anti-apartheid resistance.

On the eve of the 1994 first democratic general election, the SANDF replaced the SADF, as a result of the decision to integrate non-statutory forces of MK and APLA with SADF and TBVC forces. Since then, and while undergoing complex processes of transformation, restructuring and downsizing, the military has been involved in a range of security issues. In both the 1994 and 1999 elections, for example, the SANDF assisted the Independent Electoral Commission (IEC) in the registration and voting processes. It was also deployed in high-risk areas around the country to reduce tensions between members of the various political parties and to provide the necessary security on voting days.

Currently, the SANDF plays a major role in border patrols, with 23 companies engaged at any one time along South Africa's extensive frontiers. The military also performs a central role in assisting the "new" South African Police Service (SAPS) in crime-combating operations (both internal and cross-border), and is also deeply involved in the rural protection plan (utilising part-time forces, Commandos and Rear Area Protection Units) for farming communities that have increasingly come under criminal attack since the early 1990s. As early as 1996, the SANDF was involved in over 20 000 operations across the country. In 1997, 54 army companies (between 6000 to 8000 troops) were maintained on a routine basis in providing support to the police. These operations reflect the contemporary nature of crime and violence and the prioritisation of particular crime-fighting objectives. According to the Department of Defence, these actions have resulted in the recovery of large numbers of narcotics, vehicles and livestock, confiscation of weapons and apprehension of undocumented migrants.10

The effectiveness of these operations in the ongoing war against what are, in some respects, the results of past SADF operations to destabilise the region: arms smuggling, drug and wild-life (flora and fauna) trafficking and the like is, however, a matter of interpretation. The continuing high crime rate has resulted in calls for military involvement to be expanded, but military leaders have thus far resisted this move (Frankel, 2000, pp. 159-160). "The SANDF is not an institution for fighting crime … but, we do deploy members of the SANDF in the battle against crime because the Constitution, and Parliament have declared that it is our secondary duty to assist and stabilise our society. Generally speaking we must not see combating crime as the primary duty of the SANDF."11

Externally, the new South African military has been, or is presently, involved in a number of operations. The SANDF is cautiously engaged in peace-keeping missions in Burundi and the Democratic Republic of the Congo (DRC). The former deployment, which is officially called the South African Protection Support Detachment, is South Africa's "biggest and most expensive and riskiest military mission since 1994".12 There have already been problems. One SANDF member has been found murdered in circumstances that remain a mystery and one black soldier has been shot dead after shooting and wounding a white officer - the particulars of which have also not yet been released.

In September 1998, under the auspices of the Southern Africa Development Community (SADC), launched Operation Boleas to quell an army rebellion in the tiny mountain kingdom of Lesotho. Because nine South African soldiers, over 50 Lesotho Defence Force (LDF) soldiers and 40 civilians were killed, the operation was widely presented as a debacle and a failure. The large number of deaths in what should have been a relatively small operation was also interpreted as evidence of the SANDF's lack of fighting capacity. For some, the perceived lack of fighting capacity was attributed to the absorption of former fighters from the liberation movement, "which has created confusion and division" (International Institute for Strategic Studies, 2000). "The Lesotho debacle (also) appears to fulfil some of the worst predictions of brutality, ill-discipline and poor leadership" (Frankel, 2000, p. 200).

The South African government and the military itself have conceded that tactical shortcomings and poor intelligence resulted in a force too weak to handle the level of resistance encountered.13

Despite these real or perceived shortcomings, there is substantial progress on many of the internal policy goals of the SANDF. Its civilian control mechanism is well established. Its emergency relief operations have been widely acclaimed - especially those pertaining to the floods in Mozambique in 1999-2000. Its overall staff reduction target has been realised (the SANDF now has a full-time component of 78 823 members with further downsizing in the pipeline). Its affirmative action policies are well on the way to being fully implemented (Blacks making up 60.6%; Asians 1.2%; Coloureds 12.1% and Whites 26.1% as of 1 April 2001).14 Despite a number of problems, some of which are explored in more detail below, "transformation is clearly in motion, but one suspects it is far from concluded" (Frankel, 2000, p. 216).

"Total Strategy" and the South African Military

In the context of cold war politics and ongoing decolonisation, apartheid South Africa was faced with an increasingly precarious situation in the early 1970s. As the then Minister of Defence saw it:

Like the rest of the Free World, the RSA is a target for international communism and its cohorts - leftist activists, exaggerated humanism, permissiveness, materialism, and related ideologies. In addition, the RSA has been singled out as a special target for the by-product of their ideologies, such as black radicalism, exaggerated individual freedom, one-man-one vote, and a host of other slogans employed against us on the basis of double standards. (Defence Minister P.W. Botha, 1973)15

In the mid-70s, the international and continental political landscape was substantially altered by the 1974 "Revolution of the Carnations" in Portugal and the subsequent independence of Angola and Mozambique. When Zimbabwe gained its independence in 1980, South Africa had effectively lost what it considered a protective buffer of white-controlled states. In its place were what it saw as hostile, black-ruled regimes. All were avowedly Marxist and publicly opposed to white minority rule.

After the détente of the Nixon presidency and the end of the Vietnam War, the international politics of the Cold War became increasingly polarised. Liberation movements in Africa and elsewhere increasingly looked to the Communist East bloc for support while the South African government defended its own political evolution within the context of fending off the international communist threat.

The threat to the RSA within the ambit of the communist international battle for world domination is also related to the increase and establishment of communist influence and presence in South Africa" (South Africa. White Paper on Defence and Armaments Production, 1975).

This "communist battle" was depicted as a "Total Onslaught", to which the South African government responded by adopting a "Total Strategy", ostensibly to fight off the godless communists, but primarily designed to replace the lost cordon sanitaire of formerly white minority ruled allies by whatever means necessary. At the same time, the state embarked on a process of internal "reform" designed to give the appearance of movement towards a more inclusive and representative political system whilst effectively retaining the status quo of minority political and economic control.

Under the leadership of P.W. Botha (Minister of Defence from 1966 to 1980; Prime Minister from 1979 to 1984,16 and State President from 1984 to 1989), South Africa's military establishment achieved unprecedented political influence.17 As Minister of Defence, Botha had championed the "Total Strategy" policy, which had been first mooted in the 1977 Defence White Paper and was later adopted as official government policy. Throughout Botha's subsequent premiership and presidency, the government developed and maintained this anti-communist stance, aligning itself with the West, buying both support and time to try to control the situation at home.

The Total Strategy addressed both the domestic, and the regional security situation. In 1979, South Africa's security agencies convened in the Western Cape to decide who would take responsibility for the security of specific geographic areas. "Specific tasks were allocated, areas of responsibility defined and a broad outline given of the tactics to be employed" at a conference of security chiefs in what has become known as the Simonstad Raad or Simonstown Council.18 A task force identified targets involved in the armed struggle (Seegers, 1996, p. 163). Despite being termed a "defence" force, the SADF was to become the primary vehicle for the destabilisation of most neighbouring countries. The security situation inside the country during the 1980s, however, continued to deteriorate, and although the SAP remained the main agency for internal security (and repression), the SADF's role inside the country steadily increased. Both the SAP and the SADF would "be in the forefront of the counter-revolutionary efforts, with the National Intelligence Service and Department of Foreign Affairs providing backup" (Seegers, 1996, p. 163).

Political support from powerful western nations, including the United States, Britain, France and West Germany, effectively allowed South Africa to develop and use its formidable military machine to defend the apartheid regime from any internal or external threat to national security. Consequently, the state was able to develop both a nuclear defence programme, (developing six nuclear bombs, all of which were officially dismantled in the early 1990s) and "one of the most aggressive chemical and biological warfare programmes to have been implemented anywhere in the world since the Second World War" (Ellis, 2000, p. 62).19

The State Security Council (SSC) and the National Security Management System (NSMS)

The Total Strategy also involved the development of an intricate politico-security infrastructure that was to stretch from the highest echelons of government down to remote rural communities. At its apex was the State Security Council (SSC), which many observers claim became the de facto Cabinet. This process has been described as a "bloodless coup", and indicates just how much coercion and state-enforced control was needed to preserve white control over both politics and economics in South Africa. (Ellis, 2000, p.62).

However, while it is true that security issues came to pervade virtually all major aspects of governance under Botha's leadership and that the SSC took on a central policy-making role - sometimes surpassing the Cabinet as the most important decision-making body in the country - "political society and the dual state remained largely intact". What did emerge was a highly centralised bureaucratic-cum-military structure with the power and ability to coordinate the implementation of both security and political policies, and which took direct responsibility for the formulation of strategies "with the common objective of winning the war" (Swilling & Phillips, 1989, p.143).

This bureaucracy was known as the National Security Management System (NSMS) with a network of Joint Management Centres (JMCs) as its operational arm. The SSC's Secretariat was dominated by members of the National Intelligence Service (NIS), with SADF and Defence representation making up 10 per cent of the total staff (Seegers A, 1996: 166). The Chief of the Defence Staff acted as the nodal point of the SADF for co-ordination and liaison with the Secretariat of the State Security Council.20 Below the national command centre was a hierarchy of JMCs that generally corresponded to the SADF's regional command areas but were later to correspond to economic regions. Military or police Brigadiers, each directing about fifty officials, chaired the JMCs. These JMCs consisted of members of the security forces from both the SAP and SADF (16% of its membership consisted of SADF personnel), local government officials, community councillors and other organisations such as churches and civil defence units. JMC officials also directed the activities of sub-JMCs, or sub-centres composed of officials from municipalities, police, and the military in the area. At the bottom of this security apparatus pyramid was a host of mini-JMCs, directed by municipal officials, fire chiefs, local defence officials, and other community leaders. The SADF therefore participated in the in the overall strategic and operational management of the state.21

The state made extensive use of Low-Intensity Conflict (LIC) theory and developed a dual strategy of repression and reform, assisting and utilising surrogate opposition forces in neighbouring states such as the Resistência Nacional Moçambicana (RENAMO) in Mozambique and Uniao Nacional para a Independência Total de Angola (UNITA) in Angola for destabilisation and repression, to complement the direct involvement of its own forces, including those of SWATF. Inside South Africa the security forces provided direct and indirect support for vigilante groups22 and others, including Inkatha (subsequently renamed the Inkatha Freedom Party), that opposed the swelling ranks of the United Democratic Front (UDF), widely perceived as a proxy of the ANC. The ensuing violence enabled the government to feed the myth that the violence was simply "black on black" (Hanlon, 1986).

LIC also involved at least the appearance of reform. Even P.W. Botha realised that "you can't control violence only with violence and force - you also need socio-economic and other measures to stop the onslaught against South Africa".23 Consequently the state attempted to "win the hearts and minds" of black South Africans by increasing economic aid to local authorities and by trying to improve the image of the SADF in the townships.

As envisaged by the 1977 Defence White Paper, by the early 1980s control and concession and divide and rule had become the state's modus operandi.

The resolution of the conflict in the times we live in demands interdependent and co-ordinated action in all fields - military, psychological, economic, political, sociological, technological, diplomatic, ideological, cultural etc … (South Africa. Defence White Paper, 1997)

The NSMS, therefore, provided the infrastructure for this new strategy, encompassing "hard war" methods aimed at "eliminating revolutionaries" and "maintaining law and order", and a "soft war programme" designed to win black support through social and economic upliftment schemes (Swilling and Phillips, 1989, p. 145).

The NSMS or, as it became known later, the "National Co-ordinating Mechanism" and the Inter-departmental Committee on Security, provided the SSC with the necessary information to ensure a constant national security profile and to take decisions at both national, and local levels.24 These decisions could then be implemented either by formal law enforcement agencies or by other structures acting covertly (Coleman, 1998b, p. 106). "Total Strategy" thus laid the basis for the SADF's increasing influence on the formulation of state policy, while the security-focused SSC ensured the military's place at this focal point of all national decision-making and government power (Frankel, 1984).

The SSC thus functioned as the NSMS's national command centre, evaluating current intelligence, formulating policy, and directing a nation-wide organisational network dedicated to implementing the Total Strategy in terms of security measures, and at the same time "winning the hearts and minds" of the masses by token township upgrading (Coleman, 1998b, p. 21).

In response to these and other political developments and government "reform" initiatives, the United Democratic Front (UDF) was formed in 1983, as a coalition eventually comprising over 600 domestic anti-apartheid organisations. The UDF was soon regarded by the state as part of the revolutionary problem. By 1984, SADF training courses described the Front as one of

five revolutionary groups which are conducting a political or military war against the RSA … and which … is largely responsible for the total onslaught against the RSA.25

It was thus not surprising when in February 1988, the state announced the restriction of 17 extra-parliamentary organisations, including the UDF and the judge in the "Delmas Treason Trial" ruled that it was a revolutionary organisation conceived by the ANC (Sarakinsky, 1989, p. 75). These restrictions and comments were integral to the WHAM approach, which sought to legitimise "existing institutions and institutions which will be established in the context of forthcoming reforms" (Sarakinsky, 1989, p. 80).

The SADF's Regional Destabilisation Strategy

During this period, one of the chief tasks of the SADF, and in particular of its Directorate of Military Intelligence (DMI), was to implement a destabilisation strategy in the Southern African region. Military (and economic) destabilisation of the neighbouring states included invasions and "hot pursuit" operations into Angola, pre-emptive strikes against the South West African People's Organisation (SWAPO) in Namibia and Angola, actions against the African National Congress (ANC) in exile in countries such as Mozambique, Swaziland and Lesotho, and military support for rebel anti-Marxist groups such as RENAMO and UNITA.

From 1975, the SADF invaded southern Angola regularly - Operation Smokeshell (1980), Operation Protea (1981) and Operation Askari (1986) being prime examples of this. In response to ANC refugees in Lesotho, the SADF attacked "targets" in the capital, Maseru, in 1982 and 1985, killing 36 ANC members and 15 Lesotho citizens.26 In addition, the SADF's Directorate of Special Tasks (DST) also clandestinely supported the Lesotho Liberation Army (LLA) under Operation Latse. These actions eventually culminated in the Lesotho government's formal denial of refuge to ANC members and supporters. In Mozambique, South Africa's support for RENAMO and its direct military actions against ANC bases in Maputo, led the ruling party, the Frente de Liberta‡ao de Mo‡ambique (FRELIMO) to curb the ANC's presence in its territory officially, as part of the infamous 1984 Inkomati Accord. The Kingdom of Swaziland was also not immune. Considerable tension had developed with Pretoria over the presence of ANC and PAC activists seeking refuge there and using it as a rear base from which to launch guerrilla attacks on the Republic. A secret non-aggression pact signed in 1982 resulted in an unknown degree of co-operation between Swazi and South African security forces.

Between 1960 and 1990, the South African government's involvement in the region expanded from occasional cross-border interventions in the 1960s to a situation in the 1980s where the SADF was involved in various levels of warfare in six Southern African states, while covert units conducted attacks particularly in the BLS states of Botswana, Lesotho and Swaziland (TRC, 1998, Volume 2, Chapter 2, Paragraph 4).

While it is impossible to pinpoint exactly "how many deaths in those countries were the result of South Africa's interference … what can be said is that hundreds of thousands of people died throughout Southern Africa in conflicts manipulated by South Africa for reasons of its own, and to some extent those people could be considered victims of the struggle for South Africa itself" (Ellis, 2000, p. 65). These wars then had catastrophic human and economic consequences - thousands of deaths, damage to billions of Rands' worth of property and infrastructure, and almost unquantifiable costs in lost economic growth, tourism and trade agreements (Hanlon, 1986, p. 265). Unfortunately, the TRC only briefly examined these operations and in some senses, "diverted attention from some of the greatest abuses of human rights inflicted by the South African state" (Ellis, 2000, p. 65).

De Klerk's eventual announcement in November 1989 that the NSMS was to be abolished and replaced by a co-ordinating system under civilian control, known as the National Co-ordinating Mechanism, was a deliberate move to remove security force influence and consolidate decision-making powers, returning them to the Cabinet. De Klerk, thereby re-established Cabinet as the "highest policy making co-ordinating authority".27 Few, if any, analysts of the South African transition would however describe this process as a "military transition", that is, the transfer of state power from a military government to a constitutional government headed by a civilian with a popular mandate.

The Militarisation of South African Society

Complementing South Africa's destabilisation programme and internal security strategy was a broader militarisation of South African society, in which the SADF played a central role. The expansion of military power and influence as a social institution is evident at a number of levels, including the political, economic and ideological (Tomaselli, 1984, p. 215).

At an economic level, militarisation was epitomised by the strong links between the SADF and the private sector, particularly with the phenomenal growth of the arms industry. The TRC found "business was central to the economy that sustained the South African state during the apartheid years … businesses benefited from co-operating with the security structures …" (TRC, 1998, Volume 4, Chapter 2, Paragraph 161). By 1984, over 2000 private sector firms were involved in the arms industry, and employed over 120,000 people. Moving from a position of almost complete dependency on arms imports during the 1950s and early 1960s, "by the end of the 1980s, the arms industry had reached a relatively high level of self-sufficiency and could meet most of the equipment requirements of the SADF" (Batchelor, 1998, p.100).

At the political level, the SADF influenced foreign relations, particularly in neighbouring states, and was increasingly involved in domestic security arrangements, in terms of "traditional" law and order responsibilities, and also in other arenas as diverse as education, labour and health.28 In 1985 alone, for example, some 35 500 troops were used in the townships to evict rent defaulters, occupy classrooms, identify the injured seeking treatment in health clinics, and break strikes. Murder by "death squads" and the disappearance of prominent anti-apartheid activists were also common during this time.

The degree to which the SADF and the SAP were linked in this period, in suppressing black resistance, was an important indicator both of the level of violent conflict and the role of the SADF within that conflict" (Cock, 1990, p. 88).

Direct evidence of the military's involvement in covert actions, such as assassinations, is available but limited. This appears to reflect limitations with regard to disclosures before the TRC, rather than a limited number of incidents that military elements were actually involved in (Pigou, 2001).

Militarisation at the ideological level manifested itself across a range of areas; school cadet systems, youth preparedness programmes, the promotion of war toys and games, the compulsory registration of 16-year-old white boys for conscription and the progressive extension of compulsory military service for white male youths to two years plus annual camps.29

A militarised and, therefore, violent response to prevailing conditions, came to be seen as the acceptable solution to apartheid South Africa's problems. Violence was either sanctioned by law or at least, unrestrained by the law (Cock, 1990, p. 103). Furthermore, the NSMS ensured that state employees believed in national security concerns and did as ordered (Seegers, 2000). Government policies were also supported by the bulk of the white electorate who continued to return the National Party (NP) to power with large majorities. Elections in 1977, one year after the Soweto uprising, gave the National Party its strongest-ever mandate: 67% of the white vote. Moreover, during the 1980s, surveys conducted by the Human Science Research Council (HSRC) and the South African Institute for International Affairs (SAIIA) show that the vast majority of white South Africans favoured the exclusion of black South Africans from the political system, white schools, residential areas and public amenities, and approved of the policy of cross-border raids (Theissen & Hamber, 1998, pp. 8 - 12).

These endorsements of the government, legitimised an increasing military involvement in all aspects of political, social and economic decision-making. McCuen's counter-insurgency theory,30 and in particular his "winning hearts and minds" (WHAM) strategy was embraced by the South African state (Cock, 1990). The situation was complex, as it became increasingly necessary for the state to create the impression that it was "reasonable", wanting to upgrade and reconstruct townships, whilst at the same time retaining control over the security situation. State violence became more covert and selective; with the aim of intimidating without antagonising and with the (somewhat forlorn) hope that large-scale upliftment programmes in the townships would alter political attitudes and erode support for radical elements. This subsequently evolved into a cruder programme of divide and rule in the mid-1980s, as vigilantes and municipal constables became the primary agents of violence and repression. Concerted efforts were also made to forge relations with more "traditional" and conservative elements within the black community.

As white South Africa became increasingly militarised, so too did South Africa's black communities, either in collaboration with the state and its security forces, or in direct opposition to it. But the militarisation of members of South Africa's white community effectively breathed life into the apartheid leviathan, as they seemed impervious to, and largely insulated from, the security force abuses committed in their name.

In May 1996, shortly after the first public hearings of the TRC, the Centre for the Study of Violence and Reconciliation (CSVR) conducted a random nation-wide telephone survey amongst 124 white South Africans, to gauge whether white South Africans would sooner deny their complicity in the apartheid past than take any responsibility for what had happened (Theissen & Hamber, 1997). At the time CSVR concluded that not only did white South Africans turn a blind eye to ongoing human rights violations, but also, that many of them supported the way in which the security forces dealt with black opposition. The survey found that less than a quarter (23%) of respondents claimed to have heard about the atrocities only through the TRC. Over half (55%) said that they knew there were abuses, but were unaware of the extent. A further 22% said they had been more or less fully aware of the scale of the atrocities (Theissen & Hamber, 1997).

A consequence of this pervasive militarisation continues today: South African society has become "a battlefield littered with the bodies of damaged people" (Cock, 1988, p. 23). Almost twenty years of border wars and township violence has left South Africa with hundreds of thousands of war casualties of a psychological and physical nature. Indeed "what the TRC uncovered [was] not a few bad apples … but a system, a culture, a way of life that was organised around contempt and violence for other human beings" (Ignatieff, 2001, p. 21).

The Truth and Reconciliation Commission and the Military

One of the problems that confronted the negotiating parties during South Africa's transition to democracy was how to deal with gross human rights violations under apartheid (Nattrass, 1999, p. 373). In order to resolve this, the South African Truth and Reconciliation Commission was established in terms of The Promotion of National Unity and Reconciliation Act (No. 34 of 1995) which was, according to the then Minister of Justice, "a necessary exercise to enable South Africans to come to terms with their past on a morally accepted basis and to advance the cause of reconciliation";31 in other words, "to balance peace and justice, forgetting and forgiveness, healing and punishment, truth and reconciliation" (Ignatieff, 2001, p. 15).

However, the TRC had to make difficult choices and operate within a number of constraints. Firstly, it attempted to establish a middle path between the Latin American experience of not prosecuting perpetrators of gross human rights violations because of their ability to destabilise an emerging democracy, and the German (post-World War II) case, where the existence of an outright victor made punishment possible. Secondly, the TRC arose out of a negotiated settlement between the ANC and the National Party; without one side being able to dominate the other. This demanded compromise and sufficient consensus. Thirdly, and importantly, the nature of South Africa's transition meant that many of the individuals who (allegedly) committed acts of human rights abuse in the past conflict would remain in, or become part of, the state (Newham, 1995, p. 10).

Although the principle of amnesty had been agreed upon before South Africa's transition to democratic governance, the National Party government and the security forces continued to argue for a blanket amnesty. Amidst an array of other priorities, the issue was not finalised before the elections in 1994, which subsequently gave room for the development of a unique conditional amnesty process. This examined individual violations and required, amongst other things, the establishment of a political motivation and provision of full disclosure. It was "not a general, blanket amnesty for say, all generals and all top politicians, as had been the case in Chile and Argentina … it was to be a specific amnesty, on a case by case basis, decided by a panel of amnesty commissioners sitting in public hearings" (Ignatieff, 2001, p. 16).

In its submissions to the TRC, the National Party did not provide details of SADF operations, but instead chose to support those who had served under its successive administrations. Security force actions occurred within the context of defensive counter-revolutionary actions and, as such, were regarded as legitimate.32 Although there was an admission that some units of the SAP and SADF had acted outside of the law, this it was claimed, had not been authorised. As such, the National Party would not accept responsibility for these abuses. De Klerk and the National Party led the TRC to believe that separate submissions by the SANDF would deal with the details of the operational dimensions of the past conflict. No attempts were made to explore the consequences of, and responsibility for, the militarisation of South African society, and no admissions were made that the military had in fact achieved unprecedented influence in general policy making. The general tenet of the National Party's input was defensive and negative towards a process that it felt was biased against it.33

Unfortunately, the armed forces remained largely distant from the deliberations of the TRC, and the lack of co-operation on the part of the National Party and the obscurity of the SADF meant that many perpetrators could not be identified (Cherry, 2000, p. 15). Very few members of the security forces came forward to request amnesty, and the refusal of former president PW Botha to participate in the Commission probably suggests that many perpetrators of apartheid refused to take the Commission seriously (Campbell, 2000, p. 2).

Of the 7 115 amnesty applications received, 1 154 were granted amnesty and 150 were granted partial amnesty. In total only 267 applicants were security force members, and only a handful came from the military.34 These did, however, include the former head of the SADF's elite Special Forces, Brigadier Joep Joubert, for his involvement in the assassination of activists in joint internal operations with the SAP in 1986. They also included some members of the CCB, DMI and other units. It should, however, be noted that many of those who did apply for amnesty, did so only on hearing of their impending prosecution.

It is increasingly evident that the military, by and large, simply boycotted the process, as they saw no need to participate. With very few exceptions, members of the military did not face potential prosecution, in contrast to many former security police officers who were implicated as a result of investigations conducted by the Office of the Transvaal Attorney General (Pigou, 2001, p. 210).

It now seems as though there had never been any intention to prosecute. MK/SADF negotiations paralleled the multi-party negotiations, and during 1993 MK chief and the first post-apartheid Minister of Defence, Joe Modise, told SADF Generals that "Nuremburg-style trials and investigations of military personnel of any persuasion did not (and would not for the indefinite future) form part of the ANC perceptions of a new South Africa". By November 1993, both sides had agreed on the "absolute necessity of a general amnesty" after elections, in the interests of reconciliation (Frankel, 2000, pp. 23 - 25). This helps to explain the generally unaccommodating attitude of the military during the process.

Unlike the ANC, which provided a number of submissions - however imperfect - and (even) admitted that its security department had overstepped the boundaries with respect to events in its detention camp in Angola (camp 31 or Quatro), the TRC experienced a severe lack of co-operation when investigating allegations against the former SADF. One example of this was the SADF raid on an alleged APLA base in Transkei in October 1993, which resulted in the killing of five children. The nodel point claimed that the children were members of APLA. Later however, General Meiring was "forced to admit that he had not clearly verified intelligence on the target, or the existence of a large cache of weapons".35

The SANDF did, however, make a written submission to the TRC on behalf of the SADF during 1996.36 This submission was, however, condemned by the then Deputy Defence Minister, Ronnie Kasrils, as "arrogant and disappointing". According to the TRC's Deputy Chair, the submission was "breathtakingly one-sided and almost arrogant in that no responsibility at all was taken for any events. The defence force emerges as the knight on the white horse".37 In both the SANDF submission and in the subsequent armed forces hearings, at which the former Chief of the SADF General Viljoen took the stand, the military presented themselves as mere functionaries of National Party policy and as such, "politically neutral". No admissions of gross human rights violations were made, and exception was taken to the depiction of the military as "apartheid security forces".38

In February 1998, the SA Defence Force Contact Bureau (consisting of the panel of the four former Chiefs of the SA Defence Force, Generals Malan, Viljoen, Geldenhuys and Liebenberg, and the convenors, Maj. Gen. Marais and WO1 Holliday) conducted an assessment of the probable results of the TRC's activities. While expressing their unequivocal support for the process of reconciliation taking place in South Africa, they also expressed their concern at the management and conduct of the TRC concerning matters of a military nature in the first two years of its existence. The generals suggested amongst other things that the TRC should:

The TRC Report

The TRC's report was released to the public on 29 October 1998.39 Volume Two of the report focuses on government-inspired violence and Volume Three, on state-sponsored political violence as experienced by the victims. Volume Five includes the findings and recommendations made by the TRC.

It is clear that the TRC was particularly frustrated by the SADF.

A number of party leaders, some prominent past politicians, and representatives of … the SADF made submissions to the Commission. The usefulness of these submissions varied greatly, but were generally disappointing and did little to further the work of the Commission (TRC, 1998, Volume 5, Chapter 6, Paragraph 7).

The first submission by the SADF was so insubstantial that the Commission asked for a second, more comprehensive one. For the TRC, this

reflected the enormous - perhaps unbridgeable - chasm between the perspectives of those who wielded power in the apartheid era and those who suffered at their hands (TRC, 1998, Volume 5, Chapter 6, Paragraph 14).

According to the TRC, "nowhere was this more clearly illustrated than in the opening remarks of General Viljoen's submission on behalf of the SADF at the Commission's armed forces hearing":

The former SADF was politically neutral whilst your Commission is highly politicised … the governing party of the former government did not demonstrate interest in the former SADF. You really erred in your assumption, and the expectations you created in public, that the SADF was guilty of gross violation of human rights on a substantial scale (TRC, 1998, Volume 5, Chapter 6, Paragraph 14).

For the TRC, the above statement epitomised the overarching sense of denial that seemed to have enveloped so many of those who were the leaders and beneficiaries of the former state. The TRC felt that it was "unfortunate that a 'whistle blower' did not emerge from the ranks of the SADF" (TRC, 1998, Volume 5, Chapter 6, Paragraph 33). Nevertheless, some of the information provided by former members of Military Intelligence (MI) and the Special Forces helped the Commission to obtain some insight into the role played by the SADF in cross-border target identification and other operations, and provided a broader insight into its role in the formulation of security policy. Important, but limited, information about a handful of internal operations was also received.

The TRC's access to security force files and personnel was through nodal points established at the beginning of the Commission's life. Although the SANDF nodal point was designed to facilitate access, the TRC concluded that it acted as gatekeeper to the SADF's activities (TRC, 1998, Volume 5, Chapter 6, Paragraph 41). Access to SADF archival material was also limited and given only towards the end of the time available for sustained research. Of even more concern to the Commission was the fact that the nodal point appeared to have played a similar screening function when channelling amnesty requests from its former members. In at least one case, a former member of the Civil Co-operation Bureau (CCB) was advised not to apply for amnesty. Disturbingly, appeals to the (post-apartheid) Minister and Deputy Minister of Defence for assistance bore little or no fruit. This led the Commission to believe that the role of the nodal point was decided at the highest (present and past) officer level (TRC, 1998, Volume 5, Chapter 6, Paragraph 42). These concerns were to surface again when evidence of a vast military archive emerged in late 2001.

The agreement between the ANC and "old-guard securocrats", that all information relating to the military would flow through the nodal point was, in the words of the TRC's Head of Investigations, "perhaps the biggest mistake" of many made by the TRC.40

The TRC however did find that the National Party government had been involved in extra-judicial killings and that the SSC was responsible for a number of human rights violations, including the creation of a political climate that facilitated the perpetration of gross violations on a wide scale. It also held the SADF (together with other state forces) accountable for the destruction of state documentation, which would have provided information on the inner workings of the state's security apparatus. Within this context, the Commission found that:

The state - in the form of the South African government, the civil service and its security forces - was, in the period 1960-94 the primary perpetrator of gross violations of human rights in South Africa, and from 1974, in Southern Africa (TRC, 1998, Volume 5, Chapter 6, Paragraphs 78 - 101).
The TRC's Findings on the "Third Force"

The early 1990s saw unprecedented levels of violence in South Africa. A "hidden hand" or "third force" was alleged to be involved in orchestrating and fomenting such violence in order to derail the negotiation process or, at least, to weaken the ANC. "This 'third force' was seen to involve covert units of the security forces acting in concert with other individuals or groupings, such as the Inkatha Freedom Party (IFP) and various right-wing paramilitary structures" (TRC, 1998, Volume 5, Chapter 6, Paragraph 126). This interpretation of the violence fits with the state's WHAM strategy and is illustrative of lessons learnt by the South African security forces when it mounted a successful covert campaign to prevent SWAPO from gaining a two-thirds majority in Namibia's independence election of 1989 (Ellis, 1998).

Unlike in the 1980s, the TRC had limited success in uncovering the cause of violations in this period, although "the nature and pattern of political conflict in this later period changed considerably, particularly in its apparent anonymity" (TRC, 1998, Volume 2, Chapter 7, Paragraph 1).

Nevertheless, the TRC found evidence of:

Furthermore, the Commission found some evidence regarding the involvement of security force operatives and IFP members in train violence and in right-wing agendas and structures; the existence of alliances between certain security force operatives, the right-wing and sectors of the IFP, who clearly believed they were arming and training people for a full-scale war.

While little evidence exists of a centrally directed, coherent and formally constituted "third force" on the basis of the above, the TRC found that a network of security and ex-security force operatives, often acting in conjunction with right-wing elements and/or sectors of the IFP, fomented, initiated, facilitated and engaged in violence that resulted in gross violations of human rights, including random and targeted killings. Such networks had established "partnerships" during the 1980s with pro-government individuals or groups at a local level, which then acted in concert to perpetrate such violations (TRC, 1998, Volume 5, Chapter 6, Paragraph 128).

Importantly, the Commission found that the sanction of illegal activities by security force operatives during the 1980s provided the basis for their continuation in the 1990s, that these networks functioned with the active collusion and/or knowledge of senior security force personnel, and that the former government, either deliberately or by omission, failed to take sufficient steps to end to such practices.

The Commission also found that the success of "third force" attempts to generate violence was at least in part, a consequence of extremely high levels of political intolerance, for which both the liberation movements and other structures such as the IFP were held to be morally and politically accountable (TRC, 1998, Volume 5, Chapter 6, Paragraph 129)

Although the military were deployed in trouble spots in KwaZulu-Natal and the Gauteng regions, little detail of how and where the military were specifically responsible for violations was uncovered. The relationship between the military and the communities they worked in was not confirmed and varied over time and from location to location. In some areas, such as Katlehong on the East Rand, the military was increasingly regarded by embattled hostel residents as "anti-Inkatha" and "pro-ANC". In the neighbouring community, the military was seen as a "lesser evil" than the local SAP Internal Stability Unit. In general, all security force agencies lacked legitimacy.

In response to these and other findings made by the TRC with respect to the military, the South African Defence Force Contact Bureau provided a 60-page analysis of the report in May 1999. The generals again accused the TRC of deliberately manipulating the facts as set out in the SADF's submissions (SA Defence Contact Bureau, 1999).

Impact of the TRC

In the context of limited insight and disclosures, the TRC could make only a series of broad recommendations regarding the military (TRC, 1998, Volume 5, Chapter 8, Paragraphs 63-67). These covered a range of issues from civilian oversight to civil education within the military (as recommended in the Defence White Paper). The TRC made a series of specific recommendations relating to its investigations into the SADF's Chemical and Biological Warfare programme (also known as Project Coast), including one that a Judicial Commission of Inquiry should be established to investigate:

The TRC also recommended:

Despite concerns raised regarding the quality and nature of civilian oversight, the Department of Defence (DOD) claims that "civil control" is well established. It also points out that the civic education programme continues and that "a significant percentage" of DOD members "have been exposed to it".42 It appears that little if any attention has been paid to the other TRC recommendations.

Despite the potential importance of the TRC for the military's achieving its transformation and the broader goals of re-orienting civil-military relations, the military, to all intents and purposes, ignored the TRC. Consequently, the impact of the TRC on the military has been minimal. Its recommendations were regarded as too weak and lacking specificity and clarity as to what steps should be taken. Moreover, the military distrusted the TRC process, which many former SADF members regarded as a witch-hunt. SANDF members with a background in the non-statutory forces were more likely to co-operate with the TRC, which, in turn, bred a certain resentment against SADF colleagues who simply refused to divulge details of their complicity in gross human rights violations. It appears that much more could have been done in terms of institutional transformation, if a refined set of recommendations had been developed, so that the TRC's recommendations could have been synchronised with an ongoing transformational dialogue seeking to address concerns of the rank and file members and the leadership cadre (Williams, 2000).

South Africa's Transition: From Authoritarianism to Democracy

South Africa's political settlement entailed the inheritance of state institutions and their bureaucracies. This resulted in a complex set of relationships between former allies and enemies, including former members of the SADF, Bantustan armies and the military formations of the liberation movements.

While some define the period of the South African "transition" as between 1986 and 1994, that is, from the very beginnings of "negotiations" (talks about talks) until the first non-racial and democratic elections in April 1994 (Seegers, 1996, p. 283), others use this term to refer to the new social and legal order achieved in South Africa since that election, or at least the transformation of the political and economic fabric of society that began with the election.

Whatever the time period, it is clear that South Africa's "transition" has been characterised by an ongoing "process of liberalising an authoritarian political regime as well as the abandonment of the policy of racial separation". As such, this process set out to "achieve a more acceptable economic and social system as well as a more democratic system of governance" (Tucker & Scott, 1992, p.12). One could argue that this process was already underway in the early 1980s, when former President P.W. Botha introduced his own constitutional "reform" package.

Consequently, in South Africa, "transition" is generally defined as something more than the interval between one political regime and another (mere "regime shift") or a shift from authoritarian to democratic rule. It includes the notion of "transformation" (Cock, 2000). South Africa, including its military and other public and private institutions, remains (some eight years after the first democratic elections) in a process of transformation. As one political commentator points out "… if there is a reasonable expectation among a significant section of society that the basic rules governing society could change, then that society is still in a period of transition" (Friedman, 2000).

Basic evidence of this is clear from the fact that the integration process with respect to the SANDF was due to be completed (only) towards the end of 2001 after more than a seven-year process. A final intake, comprising 190 former MK and 145 former APLA members, to follow previous intakes in July 1998 and November 1999, was planned for September 2001.43

In addition to the fact that the transition did not follow the collapse of the State or the economy, an important difference between the South African transition and the transitions in Latin America and Eastern Europe, with which South Africa is frequently compared is that, for white people at least, South Africa before 1990 was democratic albeit with many features of authoritarianism under National Party rule (Guelke, 1999, p. 18). The semblance of constitutionalism, the separation of powers and the "rule of law" were not alien, even though the security forces were used for illegal activities, which were not intended to be subject to judicial scrutiny.

The Transition and the Military44

Political transition can be divided into the following three broad categories: overthrow, reform and compromise (O' Donnell, Scmitter & Whitehead, 1986). When the apartheid regime and the opposition (liberation) forces realised that further conflict would not result in a victory for any party, and that an inclusive negotiated settlement was possible, a compromise similar to those reached in two other Southern African states, Namibia and Zimbabwe, was achieved. In these situations of compromise, the achievement of "lasting peace" and the viability of formally agreed peace accords are directly tied to the future role of the armed forces and the reform of the "security sector" (Berdal, 1997, p. 813).

Consequently, the future of the military was seen as central to South Africa's transition. For political reasons, the past image and role of the military forces had to be excised and trust in the army restored (Edmonds, 1994). An important vehicle for achieving this was the integration of the various armed forces into a new, single unit. For the new government, the integration process and the Defence Forces' subsequent rationalisation and reduction to a manageable and affordable size were primary tasks (Motumi & Hudson, 1995). The military, as a state institution, needed be transformed and its defence policy re-conceptualised. The challenge was to integrate or demobilise [currently] distinct and conflicting elements, thus creating a force structure which is both efficient and politically acceptable (Mills, 1990).

In 1990, this view was not shared by the National Party government or the SADF, both of which believed that it was not necessary to change the racial composition or political orientation of the SADF, as it "was the only organisation in South Africa that spans (sic) the whole political spectrum, [and] had succeeded in welding a diverse (sic) of cultures, population groups and also political beliefs into a formidable fighting machine that had a unifying effect on the country" (Meiring, 1990, pp. 13 - 17). Similarly, in his briefing to the parliamentary defence debate in May of that same year, the then Minister of Defence, General Malan, emphatically rejected the possibility of integrating the SADF and MK (Nathan, 1991, p. 3).

For the ANC, by contrast, "it was inconceivable that the SADF would continue to serve as the South African defence force after apartheid" and what was required was "an entirely new defence force, comprising elements of MK, the SADF, the homeland armies and the armed wings of the PAC and Azapo", in part, at least, to quell any possibility of SADF intervention in the political process. This new defence force could prevent the SADF from acting as a disruptive force through the use of its network of front companies and rural commanders and its access to funds from the Special Defence Account or, worse still, staging a coup d'‚tat. Some political commentators (but not all) saw this latter scenario as a distinct possibility (Nathan, 1991, p. 4).45 That Wouter Basson, the SADF's Chemical and Biological Warfare expert, who was tried and controversially acquitted for a range of crimes including murder and fraud, was taken into the SANDF illustrates fears regarding the capacity of the SADF's top brass to spearhead a coup, and the need to incorporate it into the new order - in a way that would leave them acquiescent (Cock, 2000). According to one commentator from this period, SADF negotiators reminded MK throughout the negotiation process of the potential for de-railing the process (Frankel, 2000, p. 24).

The PAC was as emphatic as the ANC about the need for the SADF to be "rehabilitated and remoulded". "For the successful formation of the new army it must be accepted as a fact that the African people and their armies must form the basis of the future army" (Romero, 1994, p.50).

There was, thus, initial resistance from the National Party government, to the idea of integrating the various armed forces. After a series of clandestine meetings between MK and the SADF in 1991, there was, by 1992, however, a "remarkable degree of consensus between the South African government and the ANC on defence matters", resulting in the start of formal negotiations between the SADF and MK in April 1993 (Gutteridge, 1994, p. 7).

These negotiations focussed on, inter alia, the need to integrate all the armed forces. South Africa was unique in that each player had its own armed forces (SADF, homeland forces, MK, APLA). The process, however, was dominated by the monopolistic inclinations of the SADF and MK, to exclude the TBVC militaries and minor liberation movement armed formations (Frankel, 2000, p. 2).

Negotiations went ahead in the context of the ongoing political talks, which had also increasingly taken on a bi-lateral flavour in the wake of the breakdown in talks following the Boipatong massacre in June 1992. A breakthrough meeting in April 1993 at Simonstown led to the initiation of three processes that were to dominate the military agenda: the Transitional Executive Council's (TEC's) Sub-Council on Defence; Assembly Areas; and the National Peace-Keeping Force (NPKF).

Under the patronage of the Transitional Executive Council's Sub-Council on Defence and the Joint Military Co-ordinating Council (JMCC), the various military formations that were to integrate into the new SANDF embarked on a process of planning and policy formulation. The process eventually brought on board the SADF, MK and the four homeland defence forces, and later included APLA.

Despite a range of frustrations and problems, a remarkable degree of consensus and progress was achieved between the former protagonists. Some commentators argue that it was the violence of the early 1990s that drew the key parties together, and "the common understanding" that "the power of the SADF became the guarantor (of) the political system that the ANC sought to inherit" (Frankel, 2000, p. 21). Ironically, it now seemed that the overall success of political negotiations ultimately depended on the acquiescence of the military formations on all sides and on the urgent need to provide some sort of institutional umbrella under which some level of control over all armed protagonists could be engineered.

This situation fed perceptions that MK's negotiators were selling its members short. Nevertheless, with no apparent external threat, and despite competing political and economic priorities, both sides agreed that they should create a smaller highly professional army. The process, however, continued to be frustrated by an unresolved debate over "retaining standards", versus "affirmative action". When SADF held out an olive branch of a range of senior MK appointments and a promise of bridging training for all MK members, "the deal was virtually struck" (Frankel, 2000, p. 25). This has appeared to many, in retrospect, a recipe for MK's absorption into the "old order" rather than a process of balanced integration (Frankel, 2000, p. 28).

The first attempts to instil institutional control over the disparate military forces came in early 1994, with the establishment of a 10 000 strong National Peace-Keeping Force (NPKF) drawn from several, but not all, rival armies. In a context of ongoing political violence, the need to establish a legitimate security force presence became essential. Drawn primarily from MK and TBVC forces, the NPKF was deployed in the strife-ridden Katorus townships on the East Rand during April 1994, only weeks before the elections. Some analysts had warned of the possibility that such a move could be ineffectual or counter-productive (Nathan, 1994). The NPKF was intended to replace the Internal Stability Division (ISD) of the South African Police (SAP) and SADF units, both of which had been deployed in the area. The ISD had earned a notorious reputation among township residents, who had actively campaigned for its withdrawal. SADF units were accused by IFP-supporting hostel residents of targeting and victimising their members. The IFP called for their withdrawal, and in contradiction to the surrounding residents, the retention of the ISD. In this very volatile, antagonistic context, the subsequent NPKF deployment was almost inevitably disastrous, "because of disorganisation, politicisation and little, logistical support" (Davenport, 1998, p. 30). The NPKF was disbanded shortly thereafter.

On the eve of the 1994 first democratic general election, the SANDF replaced the South African Defence Force (SADF). At this time is was already recognised that the success of demobilisation and integration of the armed forces would hinge on how the new government handled the questions of race and ethnicity, which had been the cornerstones of the policies of the outgoing government (Holomisa, 1994).

The Integration Process

Although the transition from the SADF to the SANDF was one of those processes that has looked very smooth and easy, it has been tremendously complicated and marred by tension and dissatisfaction. Amalgamation involved unifying seven armed formations into one force and, at the same time, converting non-statutory force members into a conventional army.

"Integration" normally refers to a process in which armed forces and military traditions are merged into one defence force after the end of a war. However, the short-term to long-term process of replacing the SADF with a truly National Defence Force (SANDF), consisting of the so-called "statutory" and "non-statutory" forces that had been at war for 37 years, was more complex. "Logic dictates" that demobilisation and rationalisation should have preceded integration, but "due to the political circumstances surrounding civil-military relations in South Africa, the converse has occurred" (Frankel, 2000, p. 197). The South African example was therefore "largely an experiment in uncharted waters" (Frankel, 2000, p. 69). As such, it provides valuable lessons and insights into the impact of political transition on contemporary manifestations of conflict, violence and reconciliation within the SANDF structures, and on a society built on militarism. Integration "met a political and symbolic need of bringing two sides together in a manifestation of national unity and the practical expedience of coping with thousands of armed ex-guerrillas" (Edmonds, 1994). "The exercise of merging and transformation was undertaken as a political imperative, independent of the processes of defence policy formulation and the exercise of SANDF force structure and operational review" (Edmonds, 1994). In many ways the creation of this new national defence force was the symbolic culmination of the negotiations process in South Africa.46

The integration process was numerically unequal, with the former SADF dominating. This imbalance created the sense that the other forces were assimilated and absorbed into the existing SADF rather than that a new structure was being formed. It was virtually "pre-ordained" that the former SADF would provide the framework for integration, given its size, organisational differentiation and infrastructure (Frankel, 2000, p. 48).

Integration started badly, with high levels of discontent amongst the largely non-SADF members, who felt excluded. In September 1994, 2 500 combatants walked out of Wallmansthal, a temporary transit post in the integration process and 265 did not return. This was followed by a violent protest in Durban at the beginning of 1995, by 200 MK soldiers. MK cadres also marched on Parliament and the ANC offices in Cape Town, to protest against their exclusion from the integration process. Although the situation improved later, the process continued to be plagued by a host of contentious issues, including security of tenure, standards, status, promotions and salaries. Poor communication skills, organisational and bureaucratic difficulties, ill discipline and unrealised expectations, compounded this. Non-statutory members with less than seven-years' service (including recruits from the late 1980s and early 1990s) were offered only two-year short-term (renewable) contracts.

All sides felt vulnerable. Bridging training, which was obligatory for all non-statutory force members, was regarded as inadequate. Concerns arose that the rationalisation process would favour former SADF and TBVC members. At the same time, many former SADF members felt threatened by deteriorating prospects for advancement in the face of affirmative action.

"It was at this level - in the crucible of training - perhaps more than anywhere else in the integration process that the contrasting cultures of the SADF and MK-APLA came up hard against each other - with inevitable friction on both sides". Between January and August 1995, for example, almost 500 instructors (mainly former SADF officers and NCOs) alone, resigned from the army (Frankel, 2000, p. 72).

These integration-related problems also contributed directly to the frustrations of those who were subsequently demobilised. Studies conducted in 2001 on the re-entry of demobilised military personnel into South Africa's economic life have clearly shown that the majority remain unemployed and struggle to support themselves and their families.47 This supports earlier research suggesting that demobilisation did not "provide for the effective reintegration of former combatants in society".48 The problem still exists. In November 2001, about 2 500 former and serving soldiers attempted to "invade" Botswana, in order to re-occupy their former MK-bases and to re-enlist with the SANDF.49

The demobilisation policies particularly have been very weak. The process has been apparently smooth but it has not been a process of integration - it has been a process of absorption of the liberation armies into the SADF. So even though theoretically there is the new defence orientation, its still very much white controlled and white dominated and even though one now has black leadership - they talk the talk of the old militarists. It's been absorption rather than integration and has been full of difficulties.50
Defence Budgets

The ending of South Africa's occupation of Namibia and the unbanning of the liberation movements prompted a dramatic cut in defence expenditure, by 10% in the 1990-1991 budget and a further 10% in the 1991-1992 budget. Allocations to defence (at least initially) saw a steady decline in expenditure, in terms of the percentage of GDP, as Table 1 suggests:


 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
In constant Dollars 3,565 3,138 2,858 3,023 2,691 2,337 2,151 1,921 1,833 2,127
In local currency (m. Rands) 10,699 10,724 10,713 12,352 11,942 11,143 11,124 10,622 10,656 13,007
% of GDP 3.2 2.9 2.5 2.6 2.2 1.8 1.6 1.4 1.3 -
Table: 1. Sourced from SIPRI Yearbook 2001

In budget year 1998/99, the defence department received R9 958bn, R720m less than the previous financial year. The 1990s, however, saw a rise in overall defence spending; partly because of the expensive nature of the integration and demobilisation processes, and partly because of the controversial and costly decision, popularly known as "The Arms Deal", to purchase defence equipment to upgrade the SANDF's capabilities. These increases have been criticised by a number of civil society groups who claim that money is being redirected from women and children for submarines and weapons, and by some members of the ANC - one of whom controversially declined to vote for the defence budget.51 R 18.4 billion is allocated to the 2002/3 defence budget, an increase to 1.7% of GDP. This represents the first real increase in defence expenditure within the last ten years.52

Force Numbers

The integration of the various armed formations initially led to an increase in SANDF force numbers, which peaked at 102 600 in 1995/96. This was below projected expectations of 138 000, mainly because thousands of non-statutory force members whose names were on the Certified Personnel Register (CPR) failed to turn up at the Assembly Areas (Institute for Security Studies, 2001a, p. 9).53 Thousands began to take voluntary demobilisation. By April 2000, force numbers stood at 82 258,54 and by March 2001 these had decreased further to 78 823 members.55

Most of the reductions have been the result of natural attrition, demobilisation and voluntary severance or the voluntary non-renewal of contracts and the newly created Employer Initiated Retrenchment (EIR) package.56 However, the "voluntary" nature of the demobilisation has been questioned: "they may call it natural attrition - but how natural is it really. Resignations, for example, follow the perception that one is not able to stay - that one does not have a future in the SANDF".57

In general, problems associated with the integration and transformation process have also been linked to questions of morale. As the SANDF expanded (with the combining of the various armed formations) there was a very public announcement that force numbers would need to shrink. This made the transformation much more difficult because it created uncertainty as to who would be competing against whom. The "other" factor became the threat. For "straight people" it became "gays" and lesbians; for whites it was blacks; for men it was women.58 The extent to which the rationalisation process was discussed with or amongst, combatants is moot. Its contribution to a climate of uncertainty is self-evident (Frankel, 2000, pp. 199 - 200).

It should also be noted that in June 2001, Defence Minister Mosiuoa Lekota recommended to Parliament the introduction of a service system that would entail the recruitment of 10 000 young people per year on the basis of a two-year voluntary military service. "This would enable the SANDF to retain the necessary average age and best physical capacity among the bulk of [its] members".59 Such a process would facilitate the goal of achieving a pared-down professional force. The SANDF thus seems at present to be moving towards an American model, where a career in a defence force is not seen as a life-time commitment. "One comes in for a ten-year period or so, thus creating a situation where personnel shouldn't be older than 50 or so and the lower ranks between 18 and 28 years of age. So demobilisation has implied a skills-training programme. The problem is that it is easier for a 55-year-old colonel to find a job in management consulting or whatever and it is very difficult for the large number of non-commissioned officers to find work in the private sector".60

The scale of the proposed rationalisation is unprecedented. "The pervasive sense of being at risk is concretised by the fact that all NDF members (the CSANDF excepted) are now temporary staff as the vanguard of the rationalisation process moves into full acceleration." Disillusion and fear pervades the heterogeneous armed forces. Whites continue to see themselves as "pawns in a political chess match" where their futures have been negotiated away and "combat capability is sacrificed on the altar of representation". Thousands have left the military, many gravitating towards the security industry and "allied work in defence industries". Junior non-statutory force members are probably the most vulnerable, as they have fewer convertible skills, and generally have access to a more limited social security net. The situation is particularly bad for those former self-defence and self-protection unit members who were not able to access the integration process. "This group (with generally low skills and a history of disrupting the integration process) is a prime but explosive target for forced retrenchments. It is also the least absorbable into civil society" (Frankel, 2000, pp. 199 - 202).

The rationalisation process will continue throughout 2002 and 2003, with reductions to as few as 65 000 force members "to ensure that the envisaged force design will be aligned with both future national defence requirements and defence budget realities" (Institute for Security Studies, 2001a, p. 9). Despite these reductions South Africa will probably still have the largest armed force in Southern Africa, constituting 57% of the total armed forces and personnel in the region (Satgar, 2000, p. 12).

Attrition Figures for the SANDF

The figures for personnel who have left the military for various reasons, including death, retirement and voluntary severance packages, retrenchment and misconduct from April 1994 to March 2000, per former force, were as follows:

Former Force Total
APLA 1,243
BOP 1,550
CISKEI 1,207
KZSPF 253
MK 3,885
SANDF 3,093
SADF 40,901
TRANSKEI 2,071
VENDA 126
TOTAL 54,326

Interestingly, the SADF total includes 13 969 resignations, 2 498 discharges and 14 316 personnel members who took voluntary severance packages. The MK figure includes 669 deaths.

As the rationalisation and demobilisation processes continue, the SANDF continues to shed personnel monthly. The figures for September 2001 alone are as follows:

Former Force Total
APLA 21
BOP 1
CISKEI -
KZSPF 1
MK 27
SANDF 47
SADF 187
TRANSKEI 1
VENDA 1
TOTAL 288

Between April and September of the financial year 2000-2001, a total of 389 SANDF members died and a further 725 resigned. Of those who died during the financial year, 335 were below the age of 50 and of those who resigned, 700 were younger than 50.

South Africa's new Defence Policy

After the formal ending of apartheid in 1994, the defence policy of the South African military changed dramatically in concert with the formation of the Government of National Unity (GNU) and the adoption of a new interim constitution. All sides negotiating the military's future had agreed that the SANDF must be a functional component of the new democracy and would therefore "have to be reengineered on different social foundations". This, in turn, meant addressing issues such as representivity and appropriate civilian controls, and facilitating the military's visible participation in social reconstruction (Frankel, 2000, p. 101).

The Interim Constitution provided for the future military's political neutrality and parliamentary oversight. It also made provision for a "balanced, modern and technologically advanced military".61 The final constitution, adopted in 1996, defined the role and character of the military in more classical military terms, stating that:

In May 1996 a new White Paper on Defence63 was ratified by Parliament, with the aim of transforming both broad defence policy, and the SANDF. This draft legislation represented "a single fundamental break from the aggressive strategies of the National Party government" and sought to "bring defence policy in line with the new democratic dispensation" (Nathan, 1998, p.41). It was followed in 1997 by an extensive Defence Review that effectively translated the White Paper into practical terms aimed at making the proposals a reality.

Transformation, however, remains contested in many areas and "there is little consensus as to what this means or implies" (Frankel, 2000, p. 142). Specific aspects of the White Paper on Defence relate particularly to inculcating respect for and protection of human rights. These include the constitutional provision of the right of soldiers to disobey orders that contravene international law, and importantly, stipulations that would ensure civil supremacy over the armed forces.64 Both provisions aim to prevent a repetition of the use of the military in violent illegitimate activities "beyond the reach of the courts and outside the law" (Cock, 1990, p. 86).

The Defence Review process of interpreting the principles of the White Paper and developing clear policy frameworks began in February 1996. Final approval was given by Parliament in August 1997. The Review aimed to provide details of force design, human resource needs, defence planning, and finances, thereby underscoring the primary function of the SANDF to defend the country against external aggression. While the deployment of the SANDF in internal policing was considered undesirable, allowance was made for emergency situations and possible constitutional threat in South Africa.

The White Paper and Defence Review were thus essentially political processes designed "to emasculate and gain control over the defence complex".65 The ANC had come to power deeply suspicious of the SADF, fearing possibilities of a coup or some form of destabilisation. In 1995, the ANC had called for the disbanding of the commando system, which it saw as constituting the apartheid government's "self-defence units", and a repository for right-wing racism. This process, in turn, resulted in the declining ability of the military to support the police generally and, more specifically, to contribute to safety in rural areas. However, the issue of rural "safety" remains controversial,66 in light of allegations that certain commandos have taken the law into their own hands.67

The Defence Review did establish the parameters for future civil-military relations. Given the complexities of the overall process, it is not surprising that the South African experience has been touted as "perhaps [being] one of the most comprehensive and successful transformation processes … South Africa has managed successfully to implement [an] integration process during which seven different armies, each with their own traditions, culture and military histories were integrated into a national defence force admitting to a common culture and identity and united in a common allegiance to the country's new Constitution" (Chuter, 2000, p. 2). The image of success, however, has to some extent made it difficult to publicly address some of the residual and emerging problems that plague the ongoing process of transformation. "The NDF is, to an extent, hoist on its own petard since it continues to present itself as one of the most (if not the most) successful state sector in dealing with transformation" (Frankel, 2000, p. 204).

There is no doubt that the constitutional changes in South Africa have had a profound influence on the military establishment and civil-military relations (Van Wyk, 1996, p. 71). In addition to the White Paper on Defence and the Defence Review, the SANDF has also introduced a Code of Conduct and, for the first time, allowed for the establishment of two military trade unions, the SA National Union of Soldiers (SANUS) and the SA Security Forces Union (SASFU). Policy developments continue as a new Defence Act is being formulated to replace existing legislation that has been on the statute books since 1957. Furthermore, efforts to formulate a "Military Strategy", as a key component of South Africa's "National Security Policy", remain in progress.68 With regard to the Defence Bill that was submitted to Parliament during August 2001, key principles contained in the White Paper have been retained, including those maintaining:

Code of Conduct

On 15 February 2000, the Minister of Defence, Mr Mosiuoa Lekota, launched a code of conduct that has to be signed by all uniformed members of the SANDF. Significantly, the Code, which was formulated in consultation with civil-society interest groups and members of the Department of Defence, commits each individual member of the SANDF to, inter alia:

Transformation however remains contested in many areas and "there is little consensus as to what this means or implies" (Frankel, 2000, p.142). Controversy continues to rage over the debate around "guns versus butter", and remains focused on the controversial "Arms Deal", and its ethical, social and economic implications. There are also some concerns regarding the extent to which policy has been translated into practice in a number of areas, "especially in relation to transparency; affirmative action for women and black soldiers; the return of SANDF-controlled land to communities that were dispossessed under apartheid; and arms exports". Concerns also extend to the limitation of civilian-military relations and the dominance of former SADF personnel within the SANDF hierarchy and Defence Secretariat (Nathan, 1998, pp. 53 - 57). Despite these evolving matters, important quantitative indicators, in terms of personnel and expenditure, for example, show that the military has downsized significantly and re-orientated its policies.

Linking Periods of Transition to Violence and Crime

Since the Sharpeville Massacre in 1960, South Africa has been portrayed as a state ruled by violence and even as the quintessential example of a society in conflict with itself (Davidson and Strand, 1993, p. 81). As such, South Africa was, and still is, seen to be prone to various forms of violence.

There were widespread expectations that the collapse of "communism" in Eastern Europe and the Soviet Union would bring about an era of global peace. The decision of former South African State President, F.W. de Klerk, to release Nelson Mandela and other political prisoners, and to enter into negotiations with political organisations (some of which had been proscribed for almost three decades) such as the ANC, PAC and the South African Communist Party (SACP), which would inevitably bring about the "demise of the last legally sanctioned, racially segregated society in the world" evoked similar hopes (Habib, Pillay and Desai, 1998, p. 95). However, in South Africa, unparalleled levels of violence continually frustrated the negotiation process. Why was this so?

Experiences from other societies undergoing transition have, over the preceding decade or so, resulted in a growing recognition that understanding contemporary dynamics requires a detailed understanding of the past - and in particular, a violent political past. In order to understand the present situation in South Africa and deal with future challenges, it is therefore necessary to grasp how South Africa got to where it is presently (Oden & Ohlson, 1993, p. 12).

Periods of political transition are, in general, the most volatile for any society (Louw, 1992). Transitional societies such as those in Latin America and the former communist countries of Eastern and Central Europe have shown similar patterns (Shaw, 2000, p. 2). Political transition is often accompanied by dramatic changes in social and economic circumstances. Volatility frequently manifests itself in terms of high levels of crime, although why this is so "is a complex phenomenon difficult to analyse" (Shaw, 2000, p. 3).

Understanding violent crime in South Africa is complicated by the fact that apartheid-related violence (both in terms of state repression, and liberation movement resistance) led to a blurring of the boundaries of political and criminal activity, because the state itself was often a significant source (although not defined as such at the time) of criminal activity (Shaw, 2000, p. 3). In the politically charged atmosphere of the pre-1994 election period, "the distinction between political and criminal motivations [was] somewhat arbitrary" (Louw, 1994, p. 16).

Whilst crime in general remains pervasive, violent crime in the post-1994 dispensation is particularly endemic. According to police statistics, 15 999 South Africans were murdered in 1994 and a further 20 046 people were victims of attempted murder.71 Between January and September 2001, 15 054 cases of murder and 21 207 cases of attempted murder were reported to the police.72 South Africans are now more than twice as likely to be murdered than to die in road accidents.73 "Crime in South Africa has ceased to be a phenomenon on the fringes of political and economic society … and is not only, and perhaps not even primarily, the result of poverty. It is a social and even a political artefact" (Ellis, 1999, p. 49 - 51).

The last 30 years of South African history shows that politics, violence and crime are inter-connected (Ellis, 1998, p. 296). One can therefore be easily tempted to deduce that many participants in local "criminal" violence are those who were regarded as political actors when apartheid was still in place. This is, as the VTP project suggests, a complex phenomenon that is difficult to reduce to a simple continuity between past and present perpetrators, victims and forms of violence. Care must be taken to avoid labelling specific role-players, especially those with a military background, as the current "criminals". This is particularly the case given the context of South Africa's militarised social order. Nevertheless, and because of popular perceptions regarding their alleged involvement, it is important to explore the role of former soldiers in violence and crime.74

The Role of Former Soldiers in Violence and Crime

In his address to the 50th National Conference of the ANC in 1997, president Mandela, warned against believing that "the obligation to defend, advance and deepen democracy has disappeared and that anti-democratic forces of counter-revolution no longer exist in our society". He accused "various elements of the former ruling group of establishing a network to intensify a campaign of destablisation". He claimed that the use of crime to render the country ungovernable was an important feature of this campaign, which aimed to "limit the possibilities of the democratic order to such an extent that it [the ANC] would not be able to create a society of equality …"

Consistent with the objectives we have just mentioned, it [the counter-revolutionary network] has engaged in practical activities which include: the encouragement and commission of crime; the weakening and incapacitation of the state machinery, including the theft of public assets, arms and ammunition being among these; the hiding of sensitive and important information from legal organs of state; and the building of alternative structures, including intelligence machineries as well as armed formations.75

In February 1998, former SANDF chief, General George Meiring, personally handed Mandela a report alleging that there was a left-wing plot to topple the government. The report, produced by Military Intelligence, was based on dubious sources, and claimed that senior politicians, including Winnie Madikizela-Mandela, Bantu Holomisa and current SANDF chief, Siphiwe Nyanda amongst others, had conspired against the government and were linked to a range of criminal activities, including gun-running. The "Meiring Report" was subsequently dismissed by government as a disinformation attempt "linked to the destabilisation campaign", and Meiring was effectively forced into early retirement. The debacle also prompted the formation of a Commission of Inquiry to examine the function and use of the intelligence community in South Africa.76

The potential for disgruntled former combatants to pose a threat to South Africa's security and stability was another matter for serious concern. Ex-combatants, it was feared, might have easy access to weapons or, at least, form extensive networks to guarantee the continual supply of weapons. The alleged establishment of paramilitary groupings, such as the South African Total Liberation Force and the Committee of Four exacerbated this fear (Motumi, 1995, p. 15). Limitations and constraints on demobilisation and vocational skills training further reinforced concerns that former combatants may turn to a life of crime. The involvement of renegade ANC self-defence units and Inkatha-aligned self-protection units in the violence of the early 1990s was alleged by the Human Rights Committee.77 Fears regarding ex-combatants were exploited by unscrupulous elements within the security forces. A controversial report released by former SAPS Commissioner, George Fivaz, for example, attempted to link disillusioned elements within MK and APLA to a series of arms thefts from military bases. The report claimed that a group calling themselves Mkapla planned to disrupt the 1999 elections. The key source for the report subsequently denied making the allegations.78

While foreigners and illegal immigrants have increasingly become scapegoats for problems of crime and unemployment, ex-guerrillas have also become "easy symbols of menace, social dislocation and threat".79 Although some former guerrillas have been involved in robberies, including cash-in-transit heists, the extent of their participation in crime remains unclear and should not be exaggerated. They constitute only one factor in a complex situation (Laurence, 1998). Few robberies are carried out using AK-47s, a weapon commonly associated with MK combatants. Criminals rarely use the Kalashnikov assault rifle. Criminals use AK47s in less than 3% of murders (and, probably, in a minority of robberies).80

It can be speculated that certain former MK-APLA combatants have turned to crime as a result of the failure to integrate a sizeable number of MK-APLA fighters into either the SANDF or the wider economic community. Perhaps similar problems propel ex-combatants from disbanded battalions of the old SADF towards crime. "Demobilised but neglected ex-combatants who faced one another in battle during the old days gravitate towards one another in post-apartheid South Africa and join hands to rob the society that abandoned them" (Cawthra, quoted in Laurence, 1998).

The failure to integrate MK and APLA fully into the new SANDF is implicit in the following figures: of the 32 000 MK and 6 000 APLA fighters on the certified personnel register drawn up to facilitate integration of the different armies in 1994, 6000 TBVC force members and 2000 KZSPF, only 15 000 were accepted for service within the SANDF (Institute for Security Studies, 2000a, p. 9). Many of those not integrated for various reasons, such as disqualification or voluntary departure, were members of the self-defence units that were hastily recruited in the last phases of the armed struggle and were therefore less well trained and disciplined than MK veterans who were trained abroad.

In the same way that former MK combatants are blamed for the cash-in-transit heists, former APLA fighters are often accused of responsibility for attacks on farmers.81 According to Agri South Africa, more than 1000 people have died in 5 594 attacks on farms since 1991. In the first six months of 2001, 67 people died in 461 incidents.82 Many farmers suspect a political motive, and believe that the attacks are intended to drive them from the land. Because the attacks are often carried out with "military precision", APLA has become the main suspect. In several cases, APLA slogans have also been daubed on the walls of properties that have come under attack.

The South African Agricultural Union (SAAU) requested that a judicial inquiry into farm killings be implemented, but this call was rejected by (former) President Mandela, who subsequently asked the National Intelligence Agency (NIA) to investigate. The consequent report, which was not publicly released, indicated that the majority of attacks were criminally (and not politically) motivated. These findings were greeted with scepticism by (former) Freedom Front leader Constand Viljoen, because the attacks were "well-planned". The PAC's Free State leader, Thomas Likotsi, is on record as attributing the murders to disenchanted APLA and MK members. He believes that, frustrated by their experience in the new Defence Force and having spent their R15 000-exit payment, they now use their guerrilla skills to survive.83

The ideologies of the PAC and APLA reinforce the belief of farmers that APLA combatants are responsible for many of the attacks on them. The PAC and its armed combatants have always been associated with the view that whites are settlers and the land needs to be returned to its true owners. Newspaper cuttings, however, record only one case where an ex-APLA fighter has been convicted for murder. Recent research into the profiles of convicted farm attackers found that not one of the 60 incidents reviewed was politically motivated.84 These facts underline the risk of pinning a disproportionately high share of the blame on the APLA factor and are a reminder that violence and violent crime is multi-faceted.

In April 2001, the national SAPS Commissioner, Jackie Selebi, appointed a Committee of Inquiry into Farm Attacks. Its mandate is to inquire into the ongoing spate of attacks on farms, including violent criminal acts such as murder, robbery and rape, to determine the motives and factors behind these attacks, and to make recommendations on its findings.85 The Committee has yet to release its findings.

While former liberation force members are commonly linked with violent crime (cash-in-transit heists and farm attacks) within popular perceptions, less attention is given to the potential involvement of former SADF members in violence, for example, through private security companies. The scope for a "racialisation" of violence and the scape-goating of particular groups in this regard requires further study.

Criminality and the Military

The Department of Defence has recognised the importance of addressing issues of criminality within the SANDF, and is currently implementing a Strategy to Combat Criminality (Department of Defence, Annual Report, 2000/2001).

Recorded violent incidents within the SANDF during 2000-2001 show that there were 226 cases of assault and 149 shooting incidents. Between July 1999 and February 2000, over R7 million worth of equipment was stolen from various bases and units of the SANDF. These included: R1 and R4 rifles, 9mm pistols, an FT5 Missile, mortars, and a large number of motor vehicles, computers, "camo" shirts and jackets and personal belongings.86 Furthermore, unlicensed firearms have been confiscated by police from army bases, a number of SANDF reservists are currently facing charges of "abusing their uniforms to carry out hate crimes", and a colonel and a retired colonel are facing fraud charges relating to a demobilisation fund.87 In August 2001, a SANDF member was caught selling stolen firearms in the Eastern Cape town of Grahamstown. Twenty-six automatic rifles and thirteen R1 rifles, all of which had allegedly been stolen from the military base in the town, were recovered.88

In July 2000, over 100 alleged victims of police and army brutality converged on the South African National Defence Force (SANDF) base in Ermelo to call for the disbanding of local commando units.89 Although the commando units fall under SANDF control, units operating in the greater Wakkerstroom and Piet Retief areas have allegedly been subverted to serve the interests of local white farmers.90 The commandos in these areas have a history of allegedly abusing local farmworkers.91 This is not unique to these areas and reflects a more widespread problem involving certain elements within the commando structures. According to a Human Rights Watch report released in August 2001, black farm workers in rural South Africa are widely subjected to physical abuse, including rape and sexual harassment. In its report HRW states that various complainants have filed criminal charges against the commandos.92

Members of the SANDF have also been accused of "terrorising and unleashing a reign of terror" in urban townships. For example, on the East Rand, the members of the army base in Thokoza, which includes members of the former liberation armies, Umkhonto we Sizwe and the Azanian People's Liberation Army, have been involved in arrests, kangaroo courts and the destruction of property.93

The Violence of the Transition Period

Unprecedented levels of violence marked South Africa's transition. In contrast to that of the mid-70s and mid-80s, however, the violence of the early 1990s bore a different hallmark. Widespread and indiscriminate on the one hand, focused and targeted on the other, the violence was characterised by a series of massacres at places where people gathered socially (bars, restaurants, and night vigils); by attacks on commuter trains, taxis and workplaces; and by shootings and assassinations. Affected communities were divided along political, ethnic and geographical lines, and in the Gauteng province, frequently along residential (hostel and neighbouring community) lines.

Exactly how many people were killed is unclear. What is clear is that over 14 000 South Africans lost their lives in violence between 1990 and 1994; more than at any other period of the war to overthrow apartheid (Ellis, 1998, p. 263). Estimates suggest, for example, that between September 1984 and the end of 1989, 3 500 people died in political violence throughout South Africa; the number estimated to have been killed in about 90 000 incidents between 1984 and the end of 1993 is 11 000 (Van Vuuren, 1995, p. 11). Official figures show that in 1990 alone, 17 088 violent incidents resulted in large numbers of deaths, compared with 17 957 recorded incidents in the three years immediately preceding 1990. The fatality figure for the first four years of the transition period (1990 - 1993) is estimated to be in the region of 11 186 (Van Vuuren, 1995, p.11). This translates into an annual rate of more than four times that of the period prior to 1990. From 1993 to June 1995 over 7 000 people were killed, often with the AK47s and R4s.94 Disturbingly, July 1993, when the Multi-Party Negotiating Forum set 27 April 1994 as the deadline for holding elections, was the second-most violent month of the transition period (South African Institute of Race Relations, 1997, p. 600). The extent of the violence can also be gauged from the estimated annual number of deaths in politically related unrest in Natal alone: 1989 (559); 1990 (1685); 1991 (1324); 1992 (1481); 1993 (1975)95 and 1994 (1603) 1995 (905) 1996 (536) 1997 (334).96

The period between 1990 and 1994 may thus be considered as the time when the war "for South Africa, previously fought most ferociously outside the county's borders now enveloped South Africa itself" (Ellis, 1998, p. 263). The deterioration of the situation raises a number of important questions, such as who was responsible, who benefited, and what role did violence play in the negotiations themselves (Guelke, 1999, p. 45; Guelke, 2000, p. 242).

There is some suggestion that organised violence during this period was disconnected from the narrative of negotiations. It has been suggested that many of the people who perpetrated this violence continued to play an important role in South Africa, and in some cases were believed to be persisting in the use of violence in pursuit of their aims (Ellis, 1998, p.263). Others have argued that the predominant conflict in the pre-1990 era involved those who were in control of the process and that in the 1990s the locus of the conflict moved to those excluded from political power (Olivier, 1992).

Oden & Ohlson (1993) identified three main types of violence in this period: political-ideological, criminal, and ethnic. All of these had their roots in the combined political, socio-economic and mental impact of racial oppression, segregation and apartheid in South Africa. Between 1990 and 1993 South African politics was marked by two competing trends: the coming together of political adversaries to negotiate solutions and a struggle for ascendancy. For the National Party government, this translated into a double agenda of portraying itself as a responsible political participant preparing, through dialogue, to create a democratic South Africa, while at the same time undermining, through different forms of covert action, the African National Congress as a credible political force (Oden & Ohlson, 1993, p. 1).

Each of the main political parties had a very different concept of what constituted legitimate political activity. All publicly called for an end to the violence, and key protagonists regarded the violent actions of their supporters as defensive and, therefore, legitimate actions. Although the Government had lifted the ban on political organisations, removing many apartheid laws and freeing political prisoners, in the violent early 1990s there were widespread allegations that the security