Document(s) 10 of 16

The southern African region is awash with light weapons. These are distinguished from major conventional weapons largely by their size, portability and ease of use. The category includes all small arms, such as handguns, pistols, revolvers, rifles, assault rifles and sub-machine-guns. The availability of light weapons exacerbates political violence and social instability, and has direct effects on economic recovery in the region. In Mozambique, northern Namibia and Angola, the proliferation of light weapons, especially anti-personnel land-mines, threatens to subvert social and economic reconstruction. This proliferation is largely the legacy of armed conflicts which contributed to the creation of high levels of poverty and social dislocation, as well as deep-seated ethnic, racial and ideological antagonisms.

Now that there is a pluralist political system and universal franchise in all the states of the region, the consolidation of peace and democracy depends on the transformation of economic and social relations. The proliferation of light weapons is a key issue which threatens the consolidation of regional security on which such transformation depends.

During the anti-apartheid struggle, a popular slogan was ‘no peace without justice’. In the post-apartheid era, peace also requires long-term stability. To achieve peace one needs ‘to focus not just on the sources of violence (such as social and political development issues) but also on the material vehicles of violence (such as weapons and ammunition)’ (Gamba, 1995:xv). Gamba points out that the international community has not effectively straddled these two foci: ‘on the one hand, the international community . . . has put a higher value on peace in the short term than on development and stability in the long term; and, on the other hand, those who do focus on long-term stability have put a higher value on the societal and economic elements of development than on the management of the primary tools of violence, i.e. weapons’ (Gamba, 1995:xiii).

The main source of this proliferation is leakage from the various armed formations involved in past conflicts, including wars of liberation against colonial powers and post-independence civil wars. In Angola, Mozambique and Zimbabwe, the post-independence period witnessed the continuation of armed conflict due to ethnic or ideological differences.

In Angola and Mozambique, and in the wars of liberation in Namibia and Zimbabwe, the apartheid state – specifically the South African Defence Force (SADF), the South African Police (SAP), their surrogate forces and the apartheid arms industry – played a crucial part in opposing these liberation movements. A ‘boomerang effect’ is behind much of the current problem of criminal violence in South Africa, as it is fuelled by light arms flowing back into the country, particularly from Mozambique, but also from Namibia and Angola. Furthermore, not all of the negotiated settlements which marked the end of conflicts in Mozambique and elsewhere involved effective disarmament and thus large quantities of light weapons have been released onto the black market.

This chapter focuses on one category of light weapons: small arms. It is argued that the problem of gun violence has a social dimension; it is connected to social relations, values, beliefs, practices and – most importantly – to different social identities. The demand for guns is socially constructed and embedded in a ‘gun culture’; the supply is socially organised. Much gun violence is about contested identities. The solution includes altering the meanings, allegiances and identities which underlie acts of gun violence. I conclude that a contextualised demilitarisation movement which links ‘peace’ to ‘justice’ is necessary to bring about these social changes.

Gun violence and identity

Gun violence symbolises the crisis in the South African social order in the 1990s. The gun is a symbol of the failure to build a secure society. Firearm killings are the fastest-growing form of violence in South Africa.

• At present, every day 28 people are murdered with a firearm.
• In 1994, 7000 people were murdered with guns and 17700 attempted murders involved guns (this represents an increase of almost 50 per cent since 1991).
• Guns were used in 79 per cent of all robberies in 1995 (an increase of 16 per cent over 1994).
• In 1996, 17 600 licensed guns were reported lost or stolen by private individuals. This figure excludes theft and loss from the SAP and the SANDF (interview with Chris de Kock, SAP Crime Information Management Centre, January 1997).

As Shaw has pointed out, the availability of weapons ‘erodes one of the key requisites of democratic transitions, the state’s ability to monopolise the instruments of coercion’ (The Star, 26 June 1995). It is in this sense that the extent of criminal violence linked to the proliferation of light weapons threatens to subvert the consolidation of democracy in South Africa.

These material vehicles of violence are neglected in much of the literature on the subject. Furthermore, the scanty literature that does address guns focuses on questions of supply, to the neglect of questions of demand. Guns are not value-neutral, ahistorical technologies. We can hope to wean people off firearms only when we understand why people are attached to them. This requires a fresh sociological and historical approach to the problem of gun violence.

Such an approach involves analysing the relation between guns and social identity. A concern with identity is central to the issue but it often involves a paralysing relativism and a retreat from political struggle. As Bondi has written, Lenin’s question, ‘What is to be done?’ is replaced by ‘Who am I?’ (Bondi, 1993:84). I argue that these two questions are inextricably connected.

Identity is neither fixed and essentialist, nor completely fluid and shifting, but rather historically and socially constructed in changing processes of social interaction. Identity depends on a sense of difference which distinguishes ‘us’ from ‘them’. All identities operate through exclusion. The lines of ‘difference’ imply the boundaries of identity. It follows that collective identities are defined negatively that is to say, against others. But as Hobsbawm has written, ‘most collective identities are like shirts rather than skin, namely they are . . . optional, not inescapable’ (Hobsbawm, 1996:41).

Discussions of collective identities and difference need to be linked to an analysis of power relations; of how some social categories have the power to define difference as deficiency or threat. Ignoring difference perpetuates unequal power relations.

A crucial question is how difference and identity are transformed into antagonism. Freud suggested that the smaller the real difference between two peoples, the larger it loomed in their imaginations. He called this effect the narcissism of minor difference. Edward Said connects the process of identity formation in modern society directly to violence. ‘One belongs either to one group or to another; . . . one acts principally in support of a triumphalist identity or to protect an endangered one’ (Said, 1988:54). He concludes that ‘while it would be a mistake to ascribe all . . . violence to . . . identity demands, it would be an even graver mistake to ignore the process altogether’ (Said, 1988:58).

Part of the solution to gun violence involves recasting the relation between guns and social identity. This will be demonstrated in relation to a particular weapon: the AK, the Kalashnikov assault rifle.

March by Pagad (People Against Gangs and Drugs) to protest against drug dealers on the Cape Flats (Photo by Roger Bosch)

The AK-47: contested social meanings and identities

The Kalashnikov assault rifle is not just a gun; it is ‘the most potent symbol of conflict and violence in the closing years of the 20th century’ (Smith, 1996:1). Since its production in 1947, some 70 million AKs have been manufactured. It has been described as the most effective assault weapon in the world and has changed the way wars are fought.

The AK is invested with powerful symbolic force. For many young, black South Africans the AK became a mythical icon during the apartheid era, a ‘marker’ of group identity, a kind of code to assert one’s political allegiance that carried great significance. At this time, the AK was an important ingredient in the state’s portrayal of the ANC as a demonic force. Part of the process of demonisation involved stressing the relationship between the ANC and the Soviet Union, and the AK provided the link. The AK was the bearer, the material evidence of the ‘communist onslaught’; it was constantly described as ‘a Russian-made’ weapon, and there were frequent references to ‘Russian arms and ammunition’ in the state-controlled media and in media displays of captured weapons. This was the evidence to support the regime’s assertion that resistance to apartheid was not indigenous, but inspired by the Soviet Union. Thus the identity of this gun marked the identity of the Russian demon-terrorist.

Ironically, AKs were an important part of the undeclared war of destabilisation directed by the apartheid regime against neighbouring states externally and against the ANC internally. They were included in weapons supplied to Unita, Renamo and Inkatha. For example, almost 40000 AKs were purchased by the apartheid state from Poland, Romania, Bulgaria, Yugoslavia, Hungary and China between 1976 and 1986 specifically to be given to Unita (Cameron Commission, 1995). The Cameron Commission pointed out that ‘it is no small irony that the previous South African government which publicly regarded AK-47s as a symbol of “communist terrorism,” was clandestinely purchasing, selling and distributing thousands of these weapons’ (Cameron Commission, 1995:79).

The AK is attractive for a number of reasons. First, it is relatively cheap: in Namibia, Angola and Mozambique one can be bought for less than $15 or for a blanket, a bag of maize, or even second-hand clothes. In South Africa, the price can be as high as R1 500, so there are substantial profits to be made (Smith, 1996:43). Second, the AK is extremely robust: it has only 16 moving parts, is easy to maintain, durable and rarely breaks down. It is also easy to operate, which makes it particularly suitable for the increasing numbers of child soldiers in the world. For all these reasons, the AK is appealing to criminals and has become a powerful symbol of lawlessness. Criminals and terrorists, or revolutionaries and freedom fighters – these are the contested political identities condensed in the image of the AK-47.

The contestations run deep. South Africa is not a consensual society; there are no shared maps of meaning. Current media accounts of gun violence involve fragments (what Gramsci termed ‘traces’) of knowledge which have acquired the status of ordinary ‘common sense’ and are part of the attempt to impose order and meaning on our experience.

The reaction to gun violence in media accounts reflects a number of distortions concerning the AK. Despite the ‘commonsense’ view, the AK is not the most commonly used weapon in violent crime compared with pistols and revolvers. For instance, in 1995 high-calibre automatic weapons, such as AKs, were used in only 6 per cent of the murders reported that year. Admittedly, this represents an increase from 1992, when less than 3 per cent of all murders were perpetrated with such guns. However, these figures would suggest that the obsessional focus on AKs in the contemporary South African media is an ideological hang-over from the demonisation of uMkhonto weSizwe (MK) guerrillas during the apartheid era.

It should be clear in this discussion of AKs that gun violence is not treated as a uniform, undifferentiated phenomenon but as a relation – the relation between the incidence of gun violence and the social response to it. These social reactions reflect how the transition from authoritarian rule has created a deep well of social anxiety, as familiar social identities and traditional practices have been disrupted and breached. One consequence of this social anxiety is the ‘emergence of a predisposition to the use of “scapegoats” into which all disturbing experiences are condensed’ (Hall et al., 1978:157). In the South African context there are two categories of scapegoats – the ex-combatant and the illegal immigrant. Much press coverage of gun violence reflects a sense of blame and indignation towards these social categories. In the vocabulary of social anxiety, ex-combatants and illegal immigrants are easy symbols of menace, social dislocation and threat.

The policy solution generated by this anxiety – the tightening of border security to block illegal immigration and prevent gun smuggling by Mozambican ex-combatants – is inadequate. Effective policy solutions have to include an understanding of how guns are invested with powerful social meanings and linked to contested social identities. The present romanticisation of the AK and other firearms is, in part, the legacy of colonial conquest and revolutionary struggle in southern Africa.

This legacy includes antagonistic social identities and an ideology of militarism which regards violence as a legitimate solution to conflict and a crucial means of obtaining and defending power. The material legacy of war includes a proliferation of small arms. Together, these elements form a lethal mix: guns provide the power to express social antagonisms in violent ways.

The socially constructed demand for guns

The social categories involved with small arms include not only criminal networks and political groupings with paramilitary formations, but also sportsmen, mercenaries, self-defence and self-protection units, the security forces, citizens and private security firms. There is no homogeneous category of gun owner, but small arms are often the basis of a militarised identity that is lethally connected to gender, ethnicity, race and nationality.

The largest category of people who possess small arms is the ‘security forces’ – the police and the armies of the southern African region. The new Angolan army will have a force level of 70000, the Namibian National Defence Force numbers 7 000, the Mozambican army 12000 (considerably less than the planned 30000), and Zimbabwe is currently reducing its force level from 51 00 to 40 000. The integration process which created the new South African National Defence Force (SANDF) involved hugely inflated, overblown force levels of 100 000 before rationalisation and demobilisation. This structure inherited the weaponry of the SADF, the most powerful army in sub-Saharan Africa. Considerable amounts of weaponry were issued to security force personnel in KwaZulu-Natal, the Transkei and Lebowa before the 1994 elections. For instance, it is estimated that some 3 000 G3 rifles were issued by the KwaZulu police to civilians such as ‘headmen and self-protection units’ at this time (Sunday Times, 20 August 1995). Arms were also issued to commando units of the SADF’s Area Defence System in rural areas. According to Colonel Williams of the SANDF, there was poor weapons control and ‘it is doubtful whether the SANDF can provide an audit of the weapons it has provided the commandos in the past 20 years’ (Williams, 1995:6).

During the 1996 Defence Review process it was disclosed that the SANDF had a total inventory of almost 250 000 R1 rifles, almost 200 000 R4 and R5 rifles, 17 000 pistols and thousands of machine-guns – excluding the ‘war reserves’.

However, many South Africans have no confidence in the capacity of the state to protect them. The outcome has been a ‘privatisation of security’. Increasing numbers of both black and white citizens have come to rely on possessing firearms and arrangements with private security firms. The growth of diverse forms of vigilantism, as demonstrated in the publicised actions of PAGAD (People Against Gangsterism and Drugs) is also an aspect of this process.

The number of private security firms providing armed guards to companies and residences has increased dramatically in recent years. Security is the fastest-growing industry in South Africa after tourism, and the number of private security guards in South Africa now outstrips the number in the police force. There are currently some 3 000 registered security companies and 240 000 registered guards in the country. They have easy access to weaponry, less training than policemen, and there is little regulation of their activities.

In many black communities, ‘privatisation of security’ has involved the formation of community protection groups. Self-Defence Units (SDUs) were established by the ANC as a response to the violence of the apartheid state. Members of MK, the armed wing of the ANC, and SDU members tended to define themselves as soldiers fighting a ‘war’ against the apartheid regime and its supporters. SDUs continue to exist in diverse, fragmented forms and see their key social identities as ‘defenders of communities’. They were established in many areas on the Reef from 1990 to 1994 when political violence peaked in what was widely understood as a ‘war’ (Xeketwane, 1995; Rosenthal, 1994). During this time there was a notable failure on the part of the SAP to protect the private rights of black citizens against violent attack. The SAP was widely viewed as either partisan or ineffective. ‘In many instances weapons are acquired in crime-ravaged areas simply to protect and to provide security for members of households threatened by criminal elements and political opponents. At the level of the [black] community, both self-defence units and hostel residents’ associations demand access to firearms for the same reason, to protect themselves on account of the fact that they have limited or no faith in the security forces’ willingness or ability to protect them’ (ANC, 1992:11). Until recently, because of difficulties in the licensing procedures, black citizens were forced to obtain firearms through the illegal arms market.

A GLOBAL PERSPECTIVE ON LIGHT WEAPONS

Jacklyn Cock

The twentieth century has been marked by a global scattering of violence and the material vehicles of violence – light weapons. Increased intergroup violence is a major characteristic of the contemporary world, with particularly harsh outbreaks in Bosnia, Burundi, Croatia, Kashmir, Nagorno-Karabakh, Georgia, Rwanda, Sri Lanka, Somalia, Sudan, Yemen and the Kurdish areas of Iraq and Turkey. This violence is often explained by essentialist arguments about a primordial ethnic identity.

The dangers of the reliance on a primordialist theory of ethnicity are illustrated in the case of Rwanda. The theory assumes that people are born into particular cultural identities which command their deepest loyalties. The violence in Rwanda between the Hutus and Tutsis – and the violence in former Yugoslavia between Serbs and Croats – is explained in terms of fixed, inherited identities. However, many scholars have emphasised that ethnic antagonism did not cause the 1994 Rwandan genocide, although the form it took involved the systematic manipulation of ethnic identities for political ends.

The 1994 Rwandan genocide also illustrates the impact of the systematic accumulation of modern weaponry. It shows how the growing global diffusion of small arms is contributing to the incidence and intensity of violence in the post-Cold War era. ‘Western journalists have made much of the fact that a good deal of the killing was performed with knives and machetes, giving this episode a distinctive air of primitive savagery. It should be noted, however, that the killings began in a relatively systematic fashion . . . and that the earliest massacres were conducted by the army and organised militias equipped with grenades and assault rifles imported from France, Egypt and South Africa by the Rwandan (Hutu) government’ (Klare, 1995:36). There is evidence that some of the mass murders – for example, of 20 000 people in a stadium – were carried out using machine-guns. This is an item South Africa exported to Rwanda in 1992, but the managing director of Armscor, in a letter to Amnesty International, also used arguments about machetes and ethnicity to minimise South African responsibility.

The Rwandan case is part of a larger pattern; since 1989 there have been global changes in the nature of violence and armed conflict. There has been a shift away from war between states involving major weapon systems, to intra-state war involving light weapons. Instead of wars being fought between states with well-organised armies equipped with major weapons systems confronting each other across defined geographic boundaries, today’s violent conflict is usually an internal struggle with irregular forces using light weapons. This set of changes ‘constitutes a remarkable break with the past’ (Tilly, 1990:180).

The international disarmament agenda has focused on weapons of mass destruction and neglected light weapons. The threat of nuclear war has not disappeared – a threat captured in E.P. Thompson’s warning of ‘exterminism’ – but it is one that has diminished, while light weapons proliferate.

The global diffusion of light weapons is part of recent changes in the global arms trade pattern. Michael Klare has conceptualised these changes as a shift from the ‘proliferation/arms race’ model that focused on the transfer of arms and technology involved in major weapons systems from one state to another, to what he terms the ‘diffusion/global violence’ model. The proliferation model was concerned largely with state-to-state transfers of major weapon systems from six or seven major industrial powers; under the diffusion model, there are more actors involved in the transfer of light weapons.

The changes in the nature of armed conflict in the post-Cold War world have led not only to shifts in the global flow of weapons but to dramatic changes in the social organisation of violence. In intra-state conflicts involving light weapons ‘it is members of the opposing group who are seen as the enemy, not the armed forces of a hostile state’ (Klare, 1995: A1). The boundaries of opposing group identities are socially constructed and defined.

The current predominance of light weapons has reversed the trajectory whereby warfare was becoming increasingly impersonal: a process of killing and maiming by pushing a button or moving a lever. The technology of major weapons systems made its victims invisible ‘as people . . . seen through the sights of firearms could not be’ (Hobsbawm, 1994:50).

Small arms are responsible for most of the deaths in current conflicts, the most common being the automatic assault rifle. The majority of actors in these conflicts are sub-state groupings. All the violent conflict under way in the world involves violence between internal groups, often ethnically defined, rather than states. According to one estimate, ‘only 4 of the 82 armed conflicts recorded in 1989-92 were of a classic inter-state character, while all of the remainder entailed some degree of internal warfare, usually along ethnic and religious lines’ (Klare, 1994:38). Instead of the ‘cats paw’ wars of the Cold War era fought by the superpowers using proxy forces, there are civil wars taking place within failed states, in what have been termed ‘teacup wars’ (Gelb, 1994:36).

States are losing their monopoly of legitimate violence. ‘The key narrative of the new world order is the disintegration of states’ and the ‘key language of that dissolution is ethnic nationalism’ (Ignatieff, 1994:5). This implies the reversal of another historic trend: the disarmament of civilians in the process of Western state formation. ‘Well-armed groups of citizens are forming all over the world’ (Tilly, 1991: 7) and this process threatens to subvert democracy.

SDUs were initially armed with knobkieries, spears and homemade weapons, but later obtained access to a variety of firearms including AK-47s and R4 rifles (Rosenthal, 1994). Rosenthal found that SDUs were not armed by MK, but by gun-runners operating in commercial rather than political interests.

The vast majority of licensed gun owners in South Africa are white. Whites were granted firearm licences more easily than blacks. It was customary during the apartheid era to demand that blacks, coloureds and Indians have training before they were granted firearm licences, although whites were not required to undergo such training (Saturday Star, 12 May 1990). The number of licensed firearm owners has increased dramatically since 1976. At the end of 1996, the total number of licensed firearms in South Africa was 4,1 million.

Licences are easily available and enforcement is minimal. At present, you can get a licence if you are mentally stable (established in a ten-minute conversation, according to one police informant), have never been convicted of a serious (violent) crime in the last 10 years and are at least 16 years old. A number of informants maintained that there is a massive amount of police corruption in processing firearm licences, which extends deep into the Central Firearm Register.

Increasing numbers of licensed firearms are ending up in criminal possession. In contemporary South Africa, widespread poverty and unemployment have contributed to violence as increasing numbers of citizens have come to rely on criminal violence as a means of livelihood.

Analysing gun violence as a social phenomenon involves more than exploring individual biographies, motives and meanings; it involves examining the diverse social organisations, cultural frameworks, social practices, group attachments and institutions built up around guns. Collectively, these constitute a robust ‘gun culture’. Overall, this culture provides a social sanction to the possession of guns and much gun violence follows culturally defined repertoires of behaviour. The values, social practices and institutions which constitute this gun culture include what Raymond Williams calls ‘consumerist militarism’. It involves the normalisation, legitimation and even glorification of war, weaponry, military force and violence through television, films, books, songs, dances, games, sports and toys. None of these are insignificant in the light of Mann’s definition of militarism as ‘a set of attitudes and social practices which regards war and the preparation for war as a normal and desirable social activity’ (Mann, 1987:71).

All of these cultural forms constitute a kind of ‘banal militarism’ which operates near the surface of social life. Banal militarism is embedded in everyday activities; it works through prosaic routines and rituals to make war, weaponry and violence appear natural and inevitable. It is exemplified in war games such as paintball, which has become increasingly popular among white South Africans since 1985. At its core, paintball simulates the sequence of killing. One Johannesburg informant involved in this ‘gun culture’ spent much of his leisure time playing paintball, practising at shooting ranges and cleaning and ‘stroking’ (his word) the 12 guns he owned. This behaviour is chillingly reminiscent of that of the killer responsible for Britain’s worst mass murder, the Dunblane school shooting of 16 children and their teacher in 1996.

This gun culture not only operates to glamorise war and weaponry, but to ‘normalise’ these social arrangements. Part of this ‘normalisation’ is the notion that private gun ownership is legitimate; a right, not a privilege. A key institution which promotes this notion is the South African Gun Owners’ Association (SAGA). This is not as powerful as the National Rifle Association in the United States, but in 1995 SAGA organised petitions to the Constitutional Assembly stating that the Constitution should be amended to recognise the right to own firearms and to place a limitation on government’s power to disarm the civilian population. SAGA maintains that ‘the anti-gun lobby’s emphasis shouldn’t be on guns – it should be on people. Guns are to crime as cameras are to pornography but they don’t ban cameras.’

The notion that private gun ownership is legitimate is linked to the notion that guns are an effective and necessary form of protection. The gun combines two contradictory images: it is a means of order and of violence; paradoxically it is believed to provide protection from violence through the potential threat of violence. A common theme articulated by many informants who had purchased guns for self-protection was a sense of being powerless, of being victims of social forces beyond their control. But the psychodynamic power of the gun as protection is largely illusionary. Legally owned weapons contribute to the problem of violent crime.

The majority of crimes involving firearms are either committed with legally owned weapons used for an illicit purpose, or have been stolen from their legal owners. It follows that the distinction between legal and illegal weapons is a dubious one: guns are long-life commodities and their change of legal status does not affect their lethal power. The legal supply of small arms is generally the seedbed of illegal flows. The fact that over 17 000 licensed firearms probably fell into criminal hands in 1996 dramatises the dangerously self-contradictory potential of guns as a means of protection. There is reliable evidence – ironically, from the United States – that people are safer without guns. Epidemiological research has established that a gun in the home is 43 times more likely to kill a member of the household than to kill an intruder (Kellerman & Reay, 1986:1560).

While the gun culture is extraordinarily resistant to such evidence, there are disturbing parallels. The sociologist James Gibson has identified a highly energised, new paramilitary culture in contemporary America which he relates to a crisis of identity among American men. In South Africa many white male informants articulated a lack confidence in the government and the economy, and seemed uncertain of their future in relation to political change generally and affirmative action policies specifically. Both white and black male informants are also troubled by changing gender relations. There has been a reconfiguration of the discourse on gender since 1990, and women are presenting a challenge to customary male behaviour. Among diverse categories of men there seem to be different versions of a ‘crisis of masculinity’, which reflects a social dislocation and confusion about gender identity.

Men are ‘the primary agents of violence in most societies’ (Beinart, 1992:473). Violence is not an exclusively male practice, but only for men is it bound up with their identity. Guns are part of the dominant masculine code in many cultures. For example, the Afrikaner resistance leader Eugene Terre’Blanche once said, ‘The Boer and his gun are inseparable’ (Cock, 1991).

Guns are often regarded as a marker of status and signal a particular style. For example, to many members of organised crime syndicates in Soweto, ostentatiously displayed firearms indicate the status of being a ‘big man’ (Wardrop, 1996:8). However, the style that guns signal is not restricted to political allegiance or criminal defiance. Guns are also a form of social display which can signal male affluence. As one informant from Lenasia expressed, ‘If you have a BMW, a cell phone and a glamorous woman, you’ve got a lot; if you’ve got a gun as well, you’ve got everything.’

This militarised masculinity evokes an ambiguous response from women. In South Africa, increasing numbers of women are purchasing guns, which could indicate that a male style is being homogenised and spread more widely. This is part of a global trend. The growing power of women within the United States gun lobby was illustrated in 1996 with the election of a woman, Marion Hammer, as the National Rifle Association’s first female president. She has a solution to gun violence: ‘Instead of getting rid of all firearms . . . why not just get rid of all liberals (who moan about gun violence)’ (New York Times, 14 April 1996).

For some informants, gun ownership among women represents an assertion of feminist identity. A South African woman firearm trainer argues that ‘we have come through the sexual revolution to be regarded as equals. We have lost the male protector. Women have to take responsibility for their own protection’ (Saturday Star, 2 November 1996). She advises on how ‘women can carry guns for self-defence and still look feminine, sexy and demure’. This kind of thinking is one response to an increasing trend for women – in both the United States and South Africa – to be the victims of gun violence. In 1996, many of the 36 000 rapes reported, and domestic violence generally, involved firearms.

The gendered nature of gun violence is significant, but gender, class, race and ethnic identities are inseparable: they construct and reinforce each other. Much gun violence relates to deep-seated fears and insecurities grounded in racial and ethnic identities, which are antagonistically defined.

For many South Africans, ethnic identity is the strongest source of social cohesion. The mobilisation of ethnicity by Inkatha to secure economic and political goals has deepened animosities along ethnic lines, and the supply of weaponry enables these animosities to be expressed in particularly lethal ways. Similarly, some right-wing

Afrikaner groupings have mobilised around a politicised ethnicity, and have formed armed, paramilitary organisations. The availability of guns encourages militant political groups to engage in violent rather than democratic opposition. In 1996, for the first time people were being killed by guns more than any other weapon used in political conflict in KwaZulu-Natal.

Some of the worst gun violence in South Africa’s history – such as the random killing of 23 black people by Barend Strydom in 1988 – is explicable partly in terms of the lethal mix of access to weaponry, gender identity and racial antagonism. Barend Strydom maintained that racial difference defined the boundaries of human identity and humane treatment.

The contested notion of ‘non-racialism’ offers an alternative interpretation of difference and identity, and has been linked to an inclusive ANC nationalism. But nationalism also involves identities which legitimise violence and which guns potentially make lethal. Nationalism as an ideology involves two claims: firstly, that while men and women have many different identities, it is the imagined political community of the nation which provides them with a primary, fixed and categorical form of belonging that trumps all other identities; secondly, that violence is justified in defence of one’s nation against enemies. Of course, there is a paradoxical relation here: nationalism is persuasive because it both legitimises violence and offers protection from violence. This relates to nationalism’s two faces: one of group identity, solidarity and inclusion; the other of exclusion.

Until very recently, the nation was the main vehicle of warfare. National identity involved the gender-specific obligation of military service and was the chief justification for lethal combat. Today, however, as Ignatieff (1994) has argued, an ‘ethnic nationalism’ is the main source of contemporary violence; it is what he calls a language of belonging and blood. He distinguishes it from civic nationalism, meaning a shared attachment to certain political institutions and laws. In contemporary South Africa, this ethnic-nationalist identity is being contested in the name of a more inclusive civic national identity which defines a common citizenship.

Less frequently contested is the connection between civic nationalism and militarism. Even the most inclusive statement of a common South African identity – that of Thabo Mbeki marking the adoption of the Constitution in May 1996 – involved invoking the militarist image of his identity as ‘a foot soldier of a titanic African army, the ANC’. This is partly a legacy of the apartheid era when citizenship involved national military service for white males and blacks were denied access to weaponry – a denial which was articulated by Z.K. Matthews to involve a denial of African manhood as well as citizenship (Hellman, 1943:45). Militarised citizenship and militarised masculinism will be very difficult to dislodge. However, a statement from an ex-SDU member points to a crucial aspect of the solution to gun violence in South Africa: the creation of new, demilitarised social identities that are sources of affirmation. Now part of the Daveyton Peace Corps, he commented, ‘I was really disappointed at not getting a gun when I first joined the Peace Corps in 1994.’ He went on to say, ‘After a while I realised that I did not need a gun . . . I now know that the community needs . . . and values us.’

The first step towards creating such new, demilitarised social identities involves confronting the legacy of the past.

Document(s) 10 of 16