The Prevalence of Violence in Post-Conflict Societies: A Case Study of Kwazulu-Natal, South Africa

Abstract

After war — so the common definition suggests — comes peace. Recent research, however, shows that post-conflict societies sometimes experience levels of violence comparable to those in times of civil war. What changes are the labels under which violence is recorded and discussed. Political, conflict-related violence between armed groups, the government and civilians becomes ‘ordinary crime’ after conflict is officially resolved. This article argues that the divide between ‘violence’ and ‘crime’ is of conceptual rather than empirical nature. It employs a strictly empirical analysis of forms of violence in KwaZulu-Natal (South Africa) to show how certain manifestations of violence persist long after armed conflict is over. The study concludes that, instead of being treated as an inevitable consequence of apartheid history and enduring socioeconomic ills, violence needs to be an immediate focus of post-conflict development agendas.

Keywords:

• violence
• post-conflict society
• post-conflict violence
• KwaZulu-Natal
• South Africa
• transition to democracy
• civil war and peace
• case study

Introduction

South Africa is often labelled as the ‘world capital of crime’ and is especially known for the brutal violence with which crime is executed (Centre for the Study of Violence and Reconcilisation [CSVR] 2007, 102; Altbeker 2007, 33). Daily newspaper articles about car hijackings, farm killings, people being lynched by angry crowds, or policemen being burned alive in their car come together with alarming statistics of violent crime. Many speak of a ‘culture of violence’ through which people are forced to navigate their day-to-day lives. Publications that tackle the problem frequently mention how South Africa's violent society was shaped by the apartheid system and the ‘political violence’ dominating black residential areas in the late 1980s and early 1990s (see for example Shaw 2002; Simpson 2004; Emmett 2000; CSVR 2009). This term usually refers to physical violence executed by the apartheid government's security apparatus, on the one hand, and the violent resistance movements, on the other. In what is today KwaZulu-Natal (KZN), armed resistance against the apartheid system escalated into a full-scale civil war between the government-affiliated Inkatha and the African National Congress (ANC).1 The ‘magic elections’ in April 1994 mark the turning point at which reports on political violence under apartheid vanished to make space for ever-present narratives of high crime rates in the post-apartheid era.

There is ample reason to view the common distinction between ‘political violence’ and ‘crime’ with scepticism. From a theoretical perspective, political violence and crime are categories referring to motives. Actors; motives are highly difficult to analyse: Is a rampaging mob of ANC-supporting youth exerting political violence? Is a shop robbery to fund the political struggle a crime? The actors themselves are often not of a mere ‘political’ or ‘criminal’ kind. Violence in the area of KZN in the late 1980s and 1990s was characterised by a complex web of power relations between political parties and factions but also warlords, gangs, vigilantes and other highly professional violence entrepreneurs taking opportunity of political antagonisms (Minnaar 1992a; Schuld 2012). These political and, at times, criminal structures did not cease to exist with the 1994 elections, and violence —especially violent crime and protest violence—is still a central topic on KZN's daily agenda. Delineating a clear division between political violence and other forms of violence compatible with our common understanding of a country ‘at peace’ seems to be much more difficult than we commonly assume.

These political and, at times, criminal structures did not cease to exist with the 1994 elections, and violence —especially violent crime and protest violence — is still a central topic on KZN's daily agenda.

These insights are not new to academic discourse. A focus on violence after peace accords or ‘post-conflict violence’ has recently emerged as a new niche of conflict research (Suhrke and Berdal 2012). Authors find that after war there is by no means peace, if these concepts are defined by the presence or absence of violence. Recent studies point out how some post-conflict societies face levels of violence that even exceed death statistics in times of violent conflict (McNeish and Lopez Rivera 2012; Dodge 2012). First steps have also been made to specifically analyse crime in post-conflict societies (Steenkamp 2011; Collier and Hoeffler 2004). The discourse on post-conflict violence is embedded in a wider body of literature discussing the grey-zone of ‘no war, no peace’ configurations after violent conflict (Mac Ginty 2008; Richards 2005).

This article intends to add to these studies and give empirical evidence from the South African province of KZN to support the finding that post-conflict violence is more than the criminal legacy of political violent conflict. It does so by analysing forms of post-apartheid violence in KZN and comparing them with characteristic forms of violence in the so called period of ‘political violence’ in the 1980s and 1990s. The study's argument is twofold: (a) empirical evidence suggests there is no clear distinction between the concepts of ‘crime’ and ‘political violence’; (b) violent practices outlive the official end of violent conflict by many years. While the first point of argument primarily challenges social scientific concepts, the second one provides a lesson for peacebuilding strategists: to take the residual violence itself seriously, not only the conflict that has supposedly produced it.

Post-conflict violence is more than the criminal legacy of political violent conflict.

Theoretical and Analytical Background

The object of analysis here is violence, more precisely the deliberate infliction of physical harm on people. The range of definitions of violence known in conflict research ranges from physical harm to concepts like psychological violence, Pierre Bourdieu's ‘symbolic’ violence (Bourdieu 1990) and Johan Galtung's ‘structural violence’ of social injustice. The focus here is limited to physical violence for two reasons: first, to use a slim concept rather than a wide theory and, second, because the accessible data — mainly collected by KZN's violence-monitoring networks — do not provide sufficient information about other kinds of violence.

In conflict research it became a common practice to think of violence in political conflicts as a distinct type of violence. The moment this kind of violence enters or leaves the stage, a civil war is coded to start or to end. In South Africa, this ‘conflict-related violence’ is commonly labelled ‘political violence’ and associated with the years before the 1994 elections. It is surprising that most works with a focus on political violence fail to give even a vague definition of the term. However, it seems to be silently assumed that it distinguishes itself from other forms of violence through its connection with the political struggle against the apartheid system.2

Violence in post-conflict configurations (as far as it is considered illegitimate) is regarded as a violation of the re-established legal system, and usually referred to as crime, assumed to be driven by personal, emotional or economic motivations. As mentioned in the introduction, the distinction between the two is by no means clear: What causes deaths to be coded as battle-related deaths or murders? Is rape that happens in times of civil war conflict-related, or is it just part of the ‘usual crime rate’ faced by any country? Is a massacre of civilians somehow less criminal only because it has happened in a time of war? Especially on the local level, distinctions between crime, political violence and other theoretical categories of violence become problematic to maintain.

Violence in post-conflict configurations (as far as it is considered illegitimate) is regarded as a violation of the re-established legal system, and usually referred to as crime, assumed to be driven by personal, emotional or economic motivations.

To rethink this broadly unarticulated analytical approach in conflict research, namely to qualitatively divide between conflict and post-conflict violence, is thereby no mere theoretical exercise. Violence classification and recording is the very foundation of conflict research, as it usually defines when civil war begins or ends and what intensity it has (see for example the Correlates of War or Uppsala Conflict Data Programme datasets). How post-conflict violence is discussed furthermore plays an important role in international conflict resolution strategies: the definition of peace and war plays a central role in the steps taken in peacebuilding strategies. The topic is therefore significant for social science discourse as well as peacebuilding practice.

The method used here is a single case study. Opposed to what is usually understood as the ‘case study method’, a single case study usually describes an ‘intensive study of a single unit with an aim to generalize across a larger set of units’ (Gerring 2004, 342). Within-case studies are — although often criticised for their limited scope — effective tools for theory development (George and Bennett 2005, 17f). Instead of generating and testing hypotheses, their goal is located one step earlier in the process of scientific knowledge production: single case studies produce what Barbara Geddes calls ‘descriptive generalisations’ (Geddes 2003) that are yet to be turned into hypotheses and theories. Accordingly, this article merely aims to provide evidence of a continuum of violence flowing from conflict to post-conflict settings and to lay the groundwork for further research into this area.

Violence in KZN is examined empirically and inductively in the following sections, with a focus on forms of violence or violent practices. It concentrates on visible manifestations of variations of violent events, as opposed to political constellations, armed groups and their motives of violent action. The approach furthermore moves away from violence recording in terms of death tolls. This allows us to avoid the disputed labels and categories introduced above. To take violence seriously as an empirical phenomenon and as a social practice of great variety makes it possible to investigate changes and continuities of violence in such (theoretically) different settings as violent conflict and post-conflict configurations. Two phases are examined: what is often referred to as ‘political violence’ during the South African resistance struggle (‘conflict-related violence’) and crime in the post-apartheid era (‘post-conflict violence’).

Violence in KZN in the 1980s and 1990s: Anti-apartheid Resistance

Although commonly portrayed as the miracle of South Africa, the transition from apartheid to democracy has been far from non-violent (Dugard 2003; Du Toit 2001). Taylor (2002) estimates that 20,000 people were killed due to political conflict during the late 1980s and early 1990s. Probably half of them died in the ‘township wars’ in what is today the province of KZN (Benini et al. 1998). There, the conflict between the apartheid regime and the resistance movement translated into ‘black-on-black’ violence. Brutal local wars over territorial control were fought between the ANC and associated groups (many of them part of the United Democratic Front alliance) and the conservative Inkatha governing the black homeland of KwaZulu. Violence erupted in the mid-1980s around school boycotts and — unlike in the rest of South Africa — persisted until the late 1990s (see Figure 1). On the local level, violence in KZN came with a complex net of political and traditional power structures that were exposed to rapid change. Conflicts about taxes and rent increases, land disputes, family feuds, generation/class conflicts and section fighting were accompanied by professional structures of warlord systems, taxi wars, youth gangs and hit-squad activities (Krämer 2007; Minnaar 1992a; Bekker and Louw 1992; Bonnin 1997). Respective antagonists in these local-level conflicts took opposite sides in the Inkatha–ANC struggle for regional dominance, rendering the political dimension to an ‘umbrella’ under which violence was framed and exerted.

Figure 1 Violence Chronology KZN 1984–1998.

KZN also witnessed a diverse range of violent practices. However, these tactics shared some common features. Violence was a young black men's business3 and mainly took place in black residential areas like townships and shack settlements. Violence was initially restricted to urban areas (Bekker and Louw 1992) but spread to rural KZN in the early 1990s (Minnaar 1992a). This picture is reproduced in today's typical settings of violence as presented below and is suggestive of the continuity of violence in the province.

Three central forms of violence typical of the decade before South Africa's transformation into a democracy in 1994 can be distinguished: violent protest action, attacks/assassinations and violent crime including forms of mob justice. These types were identified through a careful study of a rich body of violence recording (a discussion of methodology and sources of violence monitoring in KZN can be found in Schuld [2012]). They will be briefly presented here while the following section collects evidence for the evolution of these types after 1994.

Violent protest action

Strikes, boycotts, stay-aways and similar collective protest action comprised the main tactics of the ANC-led resistance struggle in South Africa. Peaceful incidents outnumbered violent action by far.4 However, violent protests played a crucial role in the early stages of escalation of the political conflict in KZN's largest city, Durban. A prominent example is the students' riot in the townships in 1985. It started as a school boycott. Students, activists and ordinary people marched together in the streets for days, engaging more and more in violent action against institutional targets like policemen's houses and government buildings. As criminal elements took over what started as political protests, looting and rampaging increased and the large crowd dispersed into small groups (Minnaar 1992b). After one week, Inkatha vigilantes managed to take over the scene, with a result of 67 deaths and more than 180 businesses destroyed (Gwala 1985, 10). The 1985 incidents show some typical features of the violence around protest action in KZN in the years of political unrest that followed. Perpetrators of violence acted in crowds composed of political activists, ordinary citizens and rioting youth. Targets shifted from symbols of government to shops and economic opportunities, and casualties mainly consisted of people who happened to be between the frontlines of rioting aggressors.

Attacks/assassinations

The majority of violent incidents were carried out as ‘attacks’. The term is used here to refer to a violent incident where a group of people uses violence against another in a situation of limited possibilities of self-defence. Attacks started as large groups sweeping an area and burning down houses, looting, stealing cattle, threatening, stabbing, beating and occasionally shooting people who did not flee in time. An important example is the so-called seven-day war near Pietermaritzburg. Inkatha vigilantes were bussed to strategic assembly points and attacked the local communities with spears, knobkerries, sjambokks and guns to wipe out insurgent elements from the area ‘infiltrated’ by ANC comrades. Similar coordinated large-group attacks took place in the Durban townships, especially around migrant labour hostels. These violent attacks were an instrument of territorial control (Bonnin 1997). People were systematically threatened to leave their houses (Meer 1989) and forced recruitment was frequently reported alongside such attacks (Minnaar 1992b). Large-group attacks were used not only against residents of a certain area, but also against people assembled at funerals and political rallies (Truth and Reconciliation Commission [TRC] 1998; Meer 1989, 144).

With the increasing separation of districts into ‘no-go areas’ under the control of a certain party and the influx of weapons and trained fighters (TRC 1998; Nebandla 2005), attacks became more covert and professionalised, often taking the form of targeted assassinations. Political assassinations were an instrument used by agents of the apartheid government to eliminate leaders of the resistance struggle during the 1970s and 1980s (Dugard 2003; TRC 1998). In the context of the ANC–IFP (Inkatha Freedom Party) antagonism, assassinations were used for the ‘systematic elimination of the leadership from the two contending forces’ (Gwala 1992).

Assassinations in KZN showed some distinct features. In the few cases where background factors are sufficiently known, killings took place in revenge cycles of ‘killing the killer’ (see for example Taylor 2002). They were carried out at night (mostly at the victim's house) or in daytime and public in so-called ‘drive-by’ shootings. Instead of large groups using traditional weapons, small groups of professionals were said to be behind a large share of violent attacks in the early 1990s (Minnaar 1992b; Dugard 2003, 35). The TRC's findings confirmed professional hit-squad activity in the region, instigated by powerful local warlords like Thomas Shabalala in squatter camp Lindelani and Sifiso Nkabinde in Richmond (Minnaar 1992c). Although these attacks were still labelled as ‘political violence’, it is difficult to determine the rationale behind them, especially the degree of political motivation. What is usually visible to externals is the victim's dead body.

Violent crime

As early as under the apartheid government, crime in black areas was recorded in statistics, displayed for example in the Race Relations Surveys. Information has to be treated with suspicion, since accessibility of data and the South African Police's coding methodology were highly questionable (Turrell 2004, 90; Shaw 2002, 15). According to these statistics, crime rates were high prior to the escalation of violent conflict (Nebandla 2005), and increased significantly during the 1980s and again in the early 1990s (Emmett 2000, 290; Shaw 2002). Shaw (2002, 17) compiles data indicating that ‘conventional’ murder rates were four times higher than fatalities from political violence. Woods (1992) deduces from SAP and Institute of Race Relations data that 70% of all murders in black townships were ‘criminally’ and not ‘politically’ motivated. As argued above, the distinction between the two is questionable, but it is interesting to note that high levels of what is called crime is not a post-apartheid phenomenon.

The reaction against crime produced a very specific form of violence that has then and now been categorised as criminal action although it goes beyond categories of the political or economic, namely mob justice. Incidents where people were killed by the hands of a crowd, hacking, stabbing or beating their victim to death, can be observed in the whole period of violent conflict in KZN. Mob executions often took place in a context of popular justice of so-called ‘people's’ (or pejorative: ‘kangaroo’) ‘courts’. The method of ‘necklacing’, in which a tyre around the victim's neck is set alight to burn a person alive, has become a style as well as a symbol for mob justice in South Africa.

Violence in the Post-apartheid Era: A Look Behind the Frame of Recorded Crime

After levels of what has been recorded as political violence significantly dropped after 1994, violence has predominantly been discussed and recorded as ‘violent crime’. Although stories of violent crime fill the newspapers every day, it is difficult to gain detailed information on its frequency and social background in KZN. This is even true for murder (Turrell 2004, 83), the most reliable of all crime statistics (CSVR 2009; Shaw 2002). There is, especially, ‘limited awareness of the nature of crime in poorer areas’ (CSVR 2007, 31). It is therefore useful to start by consulting nationwide statistics.

After levels of what has been recorded as political violence significantly dropped after 1994, violence has predominantly been discussed and recorded as ‘violent crime’.

Violent crime is certainly not a mere post-conflict phenomenon. In fact, the yearly number of murders in South Africa in the early 1990s was higher than the total of fatalities in the conflict (Du Toit 2001). Different types of violent crime show very different trends. For example, murder declined between 1994 and 2000 while rape and assault rates increased (Shaw 2002; see also CSVR [2007] with South African Police Service (SAPS) data 1994–2006). Organised crime shifted from small-scale actions to nationwide syndicates in the 1990s (Shaw 2002, 65). Illicit possession of guns is a prevalent phenomenon in many areas, accompanied by high rates of gun violence (Keegan 2005). These numbers indicate that South Africa, although far away from being perceived as a war-torn country, still faces high levels of violence.

Altbeker (2008) remarks that race, sex and geography matter; victims of murder are predominantly black men in townships or city centres of Johannesburg, Cape Town and Durban. Males furthermore account for 87% of violent deaths and 98% of the prison population (South African Institute of Race Relations 2009; 2011). In this regard things seem to be the same as in the 1980s and 1990s.

The few systematic studies on violent crime by the CSVR and the National Injury Mortality Surveillance System were conducted to learn more about why post-apartheid South Africa has had such a high rate of violent crime. The 2008 CSVR study compares 1,161 police dockets dealing with murders committed in eight areas with exceptionally high murder rates. In only 41% of the dockets are circumstances known.5 A finding consistent in all areas was that most murders were classified as related to an argument, still a big share were classified as caused by an economic opportunity and only very few could be traced back to fights between organised groups. This seems to contradict the argument made here, since it traces post-conflict violence back to economic or personal motives. Has violence in KZN since 1994 been merely the criminal legacy of political violence in the 1980s and early 1990s?

A different picture appears, however, if one moves away from numbers and statistics to the broader context.

… we can detect threads of continuity between current and past trends in violence, fuelled by the legacy of apartheid repression, ‘revolutionary violence’ and a general aversion to peaceful conflict resolution expressed in the use of weapons as instruments of social, economic and political opportunity and power. (Marais 2003, 185; see also Du Toit 2001, 49)

Today's violence thus seems to show direct links to the years of political struggle. As we see in the following, these links are more than the roots of today's violence in the past or the ‘legacy of apartheid’. The continuity includes the actual manifestations of violence. To trace the continuity of violence in KZN, a look beyond the label of crime is needed. There we find merely all types of violence that were so typical of the armed conflict before 1994.

To trace the continuity of violence in KZN, a look beyond the label of crime is needed.

Mob justice in KZN today

First of all, the phenomenon of mob justice briefly mentioned above under the category of crime is found in the KZN region up to today. The picture of a crowd chasing after an alleged criminal and beating him or her up is reportedly a familiar scene at the busy Durban market area, where informal traders sell their goods to people on their way to work. Recent cases that have made it into the newspaper read like these two: In January 2012, a sangoma (traditional healer and/or fortune teller) was killed by a mob while a crowd of 500 people was watching, because he gave wrong information on the whereabouts of an allegedly abducted child. In the same month, two accused cellphone thieves were stoned to death in Georgedale, KZN. The Human Rights Commission even deemed it necessary to have a workshop on mob justice in Durban on 22 March 2012. The numbers of people killed or ill-treated through People's Courts remains unknown, probably because township residents are reluctant to report cases to the police in order to protect the only way of fighting the devastatingly high crime level in their area. But even though quantification is missing, this violent practice seems to have survived the transition from apartheid.

Violent protests as a means of political contestation

Violent protests are still an important feature of political contestation in South Africa in general and KZN in particular. Protests almost uninterruptedly transcended the political transition, and while protests before 1994 were usually directed against the apartheid regime, they now challenge the ANC government's performance (Mottiar and Bond 2009). Especially, service delivery protests became a tool to demand the citizen's right of basic services (Booysen 2011). In recent years, protest action manifests more and more as non-orchestrated, localised community protests.

Available data on the share of violent protests is ambiguous and (without access to unpublished SAPS material) only aggregated in nationwide statistics. Statistics mentioned by Karamoko (2011), Alexander (2012) and the SAPS records on public violence all testify to an upward trend in total numbers of violent protests and unrest in South Africa. In KZN violent mass action seems to be a common feature of political contestation between the people and the local government (see Bond 2011, 13–16).

Violent protests in KZN often take the following form: a crowd of young men (according to newspapers often consisting of members of political groupings like trade unions and party factions as well as unemployed youth) gathers spontaneously or around a registered protest march and turns violent at some stage, resulting in clashes with the police, low-level forms of violence against people (such as those who undermine protest action), looting, burning tyres and destroying property for many hours. A recent example is the violence in Umlazi T-Section in late February 2012, where a violent mob wandered about the area, apparently starting with a protest march of 600 hostel dwellers and leaving two people (who seemed to be bystanders) dead. Another example is the Durban taxi drivers' protests on 18 May 2012 against ‘harassment’ by strict and corrupt police controls, which ended up with the city centre being a blocked-up ‘war zone’ for hours, shoot-outs with the police, and passengers being beaten and violently dragged out of taxis.

What is surprising is the similarity between these two incidents and the violent anti-apartheid protests in the 1980s as exemplified by the 1985 riots. Especially the composition of the crowd, the escalation dynamics, targets of violence and the role of the police share a common pattern. Further research into the shape of protest action pre- and post-1994 would be an invaluable way to gain systematic evidence regarding the continuity of the shape of protest action in KZN.

Assassinations and the xenophobic disguise of large-group attacks

Political assassinations seem to be another unbroken tradition from the 1990s in KZN. Local news reports frequently mention the selected killing of party leaders. The targets shifted from the ANC–IFP antagonism to the respective breakaway parties Congress of the People (COPE) and National Freedom Party. Systematic studies of political assassinations since the end of apartheid are missing so far and evidence is mostly based on newspaper articles and researchers' personal contacts. Some academics vaguely mention the existence of political assassinations (Piper 2004; Taylor 2002). The Violence Monitor, a provincial violence-recording publication from the 1990s that has been continued by violence expert Mary de Haas, mentions some cases of targeted killings of political leaders in 2008 (De Haas 2008). The South Africa Survey speaks of ‘numerous killings of high-profile politicians belonging both to the African National Congress (ANC) and the Inkatha Freedom Party (IFP)’ (South African Institute for Race Relations 2005).

Patrick Bond (2011, 15–16) gives a detailed account of assassinations of the political elite in Durban in recent years: In 2011, for example, Durban's leading ruling party official Sbu Sibiya, the ANC councillor Wiseman Mshibe and the leader of Chatsworth township Gundu Makhanya, as well as several of his fellow members of the Inkatha breakaway, party National Freedom Party (NFP) were killed. But the years before were not less violent and the list can be continued with names like the traditional leader Inkosi Mbongeleni Zondi, the local leader of the South African National Civic Organisation Jimmy Mtolo, the South Durban Community Environmental Alliance's leader Ahmed Osman and many others.

The incidents that led to their deaths resemble the old spectrum of being gunned down at home or in the streets as well as firebombings of the victim's house. Assassinations are also on the local agenda of civil society again: The ‘alarming increase in politically motivated violence’ over the past two years even caused the Durban-based NGO Democracy Development Programme to launch an open forum on ‘Political Killings in KZN’ in November 2012. The invitation further speaks of 40 deaths that have been ‘attributed to fighting within and between political parties’.

It is of course difficult to determine whether a killing of a prominent political actor is politically motivated. What might be concluded from this collection of clues is that the problem of the killing of political leaders, a major feature of the early 1990s violent political conflict, still occurs today in KZN.

Apart from assassinations of political leaders, there are no other obvious continuities of politically motivated large-scale attacks in the region. Attacks in the late 1980s predominantly took the form of raids of large groups of men armed with traditional weapons and determined to punish elements supporting the opposite party. While these do not manifest in the exact same way today, we can find a very similar kind of violence in what has become known worldwide as the xenophobic eruptions of 2008.

Violence against foreign nationals broke out in the Alexandria township of Johannesburg in 2008 and spread all over the country. In KZN, especially urban areas around Durban were affected (Amisi et al. 2010; Monson and Arian 2011). Xenophobic violence shows characteristics from many types of violence encountered in the years before South Africa's official transformation into a democracy. It ‘shares some elements with protest violence’ but ‘seems to draw from repertoires of more personal violence by vigilantes against criminals, who are not infrequently beaten or burnt to death, as well as from anti-apartheid repertoires such as the necklacing of izimpimpi’ (Von Holdt et al. 2011, 28–29). Some say that service delivery/community protests are behind what is called xenophobic violence (Monson 2011). But large-scale attacks are also found in the repertoire of violence re-applied in 2008. In Bottlebrosh or in Cato Manor shack settlement in Durban, groups of aggressive young men armed with traditional weapons swept through the area in order to drive foreign nationals out:

In the evening of that Tuesday, around 7pm, a group of unknown people entered the Mazithanqaze area … These people were uttering negative words about ‘foreigners’, saying ‘amashangane out! amakwerekwere out! People who harbour amakwerekwere as tenants must let them out or they will be in trouble. We don't want them here’. They were carrying sticks, bush knives, and sjamboks, and even stones … Residents tend to think that xenophobic attacks in the Cato Manor/Crest areas were forms of criminal opportunism. (Amisi et al. 2010, 71–73; see also Monson [2011] for Bottlebrosh)

Although these events only occur sporadically and on a smaller scale than in the late 1980s, they clearly resemble earlier styles of violence, contrary to the contestable media image of the violence as new phenomena. It seems as if all manner of violent practices are mixed together in the xenophobic attacks: protests, vigilante violence and mob justice, large-group attacks on areas and criminal or semi-criminal activities. This is a remarkable observation, as it shows how the same styles of violence recur in various conflict settings, directed against different targets and under different motives. It furthermore underlines the power of an approach that ignores political constellations, causes and actors but focuses on the phenomenon of violence itself. To make sense of violence by attributing it to a certain conflict narrative (like ‘anti-apartheid struggle’ or ‘xenophobic eruptions’) can mislead us in mono-causal directions. The blind approach of leaving out causes and politics can be a way to detect the continuities of violence as traced here.

Although these events only occur sporadically and on a smaller scale than in the late 1980s, they clearly resemble earlier styles of violence, contrary to the contestable media image of the violence as new phenomena.

In conclusion, many forms of violence characteristic of KZN in the years before South Africa's first democratic elections in 1994 are still found today in one way or another. Information remains difficult to access and evidence collected here can only be taken as a first attempt at documenting the similarities, but the escalation dynamics of violent protests, high crime levels and specific types of crime like mob justice as well as political killings are mirrored in the ‘consolidated democracy’ South Africa is called nearly twenty years after the political conflict was declared over.

Conclusion

This article has argued that specific types and patterns of violence applied in the 1980–1990s political conflict in the KZN region did not die with the end of apartheid; what have changed are rather the labels under which violence is discussed. Violent protests, now directed against the ANC government instead of the apartheid regime, seem to have transcended the 1994 elections; assassinations of party officials and leaders of political groupings were part of political contestation in the past as they are (in apparently limited quantities) today; large-scale group attacks aim to drive people of foreign origin out of an area instead of supporters of a certain party. Crime as a central phenomenon of the apartheid legacy — although not qualitatively analysed here — has been equally prevalent before and after 1994. The weapons used, characteristics of perpetrators and victims as well as geographical indicators are still of striking similarity to those during KZN's violent transition to democracy. Most notably the 2008 ‘xenophobic eruptions’ appear to have been a culmination of several types of violence in a shifted conflict setting. This study does not deliver mature enough evidence for a straight conclusion, but it should make us wonder whether violence in KZN could be more than a legacy of the civil war period; it shows a direct continuation — although apparently with overall decreased quantities — from the period of violent conflict.

This purpose of this article was to demonstrate how violence transcends political conflict, thereby questioning central concepts of the line between war and peace in mainstream conflict research. As a phenomenological rather than a comprehensive analysis, it does not deal with the question of why violence survived with such striking continuity. The last paragraph shall therefore suggest a few implications of this overview for peace and development in South Africa.

Today's public and scholarly discourse on violence is influenced by two main arguments. First, violence is embedded in and caused by its socio-economic setting. It is not controversial that in South Africa widespread poverty, inequality and repressive police interventions survived the end of apartheid. These socioeconomic conditions still foster violence as people turn to desperate and criminal measures to make a living. Second, the roots of today's violence in the apartheid system are often mentioned. Violence is thus seen as a long-term reaction to the former government's imposition of structural violence on the majority of the population. Violence as a means of political liberation is furthermore widely regarded as a legitimate expression of social disobedience.

Both arguments quickly lead to the conclusion that violence is a result or sign rather than a direct part of the flaws and setbacks in South Africa's post-conflict development. The development agenda (as written down in the nationwide 1994 Reconstruction and Development Programme as well as each municipality's Integrated Development Plan) consequently mainly focuses on economic improvement and service delivery. Another emphasis of post-conflict reconstruction, especially in the early years after 1994, was reconciliation and the development of a non-racial society. Education in nonviolent conflict resolution and comprehensive measurements to tackle violence as an overarching theme in the South African society were not focus topics of the post-apartheid political agenda.

However, the variety and intensity of violence in post-conflict South Africa explored here lead to the conclusion that violence should be a crucial point on today's political agenda. Xenophobia, political assassinations, mob violence and protest violence should be addressed not only though structural approaches (like reducing poverty and strengthening the legal system), but through direct measurements. Community meetings and workshops on violence and violence prevention can be used to create awareness and to gather information on how violence affects local communities. Hotspots like universities, the taxi industry and the police are other important sites of education and investigation.

On the academic side, violence can be taken as a serious indicator to analyse where exactly development policies fail. More research on socioeconomic and political consequences of violence is required. What impact do political assassinations have on the democratic system? In what way is xenophobia employed to structure the informal economy? How do the legal system and mob justice interact? For this end, it is essential to revitalise violence monitoring in KZN by nongovernmental institutions.

The aim of this article was to be another little step in the new-emerging research on post-conflict violence. A lot still needs to be done to understand the relationship of peace, development and violence in countries departing from violent conflict.

Notes

*This article has been written in my position as a visiting scholar for the Centre for Civil Society of the University of KwaZulu-Natal. My thanks go to Patrick Bond, Shauna Mottiar and all other colleagues at the Centre who supported my research.

¹ The Correlates of War threshold of ‘civil war’ is 1,000 battle-related deaths per year. This threshold was met by the violent conflict in KZN for at least the time frame 1990–1994.

² This is most visible in violence-monitoring reports and statistics that usually distinguish between deaths from political violence and others without further reflection. Examples are police statistics, local media (Daily News, Natal Witness, The Mercury), NGO reports assembled in the Network of Independent Monitors as well as the TRC's reports.

³ Although violent actors were dominantly male, women were the most vulnerable to the violent environment, as shown by Debbie Bonnin's extensive research (1997). Gender-related violence plays an important role in KZN's society and is not considered in this article.

⁴ An extensive documentation of protest action is found in Indicator South Africa (see for example Jarvis [1990]) and in the University of KwaZulu-Natal Alan Paton Centre's armed struggle archive in Pietermaritzburg.

⁵ A striking result is that in the two areas located in KZN the proportion of murder cases coded as ‘unknown circumstances’ was significantly higher than in the six other areas.