Ball, J. (1994). The Ritual of the Necklace. Research report written for the Centre for the Study of Violence and Reconciliation, March.


Joanna Ball

Research report written for the Centre for the Study of Violence and Reconciliation, March 1994.

Joanna Ball is a former economics and social anthropology student at St Andrews University in the United Kingdom.


In this paper an attempt is made to develop a clearer understanding of the 'necklace' form of burning which has evolved in South Africa. The notion that it is a political punishment, used by the youth since the mid-eighties, is questioned and three areas of exploration are suggested. Firstly, traditional Bantu ideas of such things as punishment, witchcraft and fire are discussed. Consequently a link is suggested between the victims of necklacings – collaborators, witches, murderers and rapists – in that they are all traitors to the social solidarity of the community. In the second area of research, possible explanations are examined for the resurgence of the practice of burning which had virtually disappeared this century. When placed in their social context, these burnings of social deviants are seen to be related to unsettling social, economic and political change. They may be a means of asserting some control over social life. This leads onto the third area where the necklacing act, which includes the preparations, is analysed as a ritual. It is seen to give meaning and to define the beliefs of those involved by the use of two symbols: that of the necklace and that of fire. It is suggested that the necklace symbol is highly potent in its verbal usage, whereas in the actual enactment of the burning the fire takes over as the dominant symbol. These symbols combine to form a very powerful performance which perhaps gives people something to grasp onto in this time of great change in South Africa.


In the upheavals of the last two decades South Africa has witnessed the establishment of burning as a means of extra-legal punishment. A widely renowned form of this burning has been developed, the 'necklace', which involves placing a tyre soaked in petrol around the victim's neck and setting it alight. This behaviour tends to be seen by outsiders as savage, as something which should not occur in this rational day and age, and is gladly shrugged off as an inappropriate subject to ponder upon. Consequently the 'why' behind the action has been largely unexplored, the cliches left unquestioned. Burnings are on the increase once again and, as South Africa negotiates the hurdles towards its democratic ideals, a deeper understanding of these burnings would appear to be vitally important.

The focus of this paper is on 'necklacing', but as my research progressed it became clear that this practice could not be analysed in total isolation from the wider phenomenon of burning. This is for a number of reasons. One is that all forms of burning have increased considerably, it is not just the necklace which has gained notoriety. Another reason is that while the distinction between a necklacing and other forms of burning is clear in the physical sense, it becomes blurred in the cultural arena. The term necklace is applied more broadly than the word implies, for example the media will describe a burning where the tyres are placed on top of the victim's body as a 'necklacing'. Presumably the broader use of this term is related to its symbolic potency and the graphic horror of burning that it captures. When the victim is actually burnt various forms of burning appear to be interchangeable. To contend with this hazy border I shall use the term 'necklacing' for the act in its strict sense and also where the symbolic value of the term seems important. Elsewhere I shall refer to burnings.

The first widely reported necklacing occurred in the Eastern Cape in March of 1985. The victim, Mr Tamsanqa Kinikini, was a member of the town council in the township of KwaNobuhle and was believed by some people to have been involved in corruption and violence. A few weeks earlier he had refused to follow the lead of his colleagues and hand in his resignation (Kane-Berman, 1993). His punishment was to be the necklace, and thereafter the use of this method seemed to escalate. This may have been aided by the media who gave much publicity to the event, while Dutch television cameras captured the image of a chanting mob dancing around the burning body. The next burning to receive much attention was on 20 July 1985 in Duduza, a township outside Johannesburg. A 25-year-old woman was accused of being an 'informer'. She was beaten and stoned, stripped, soaked in petrol and burnt to death. This appears to have been the final straw for President P W Botha with regard to his fears of uncontrollable unrest within South Africa. That evening he declared a state of emergency.

This new fashion provoked utterances of horror and incomprehension. The apartheid critic, Anglican Bishop Desmond Tutu warned: 'If you do this kind of thing, I will find it difficult to speak for the cause of liberation. If the violence continues, I will pack my bags, collect my family and leave this beautiful country that I love so passionately and so deeply … I say to you that I condemn in the strongest possible terms what happened in Duduza.' (Time, 5 August 1985). Foreign Minister 'Pik' Botha stated that he would not allow the future of South Africa to be decided by 'perpetrators of violence who burn people alive'. (Time, 5 August 1985). The Sunday Times described these burnings as 'an offering to the demon that has entered the body of black township politics' (16 February 1986). And The Sowetan commented that 'the stoning and burning of people is becoming almost fashionable among blacks … this is the beginning of a terrible phase that spells doom to the black nation … . This kind of savagery is completely out of step with the sophisticated thinking and behaviour of the twentieth century. The children should know from their history books that this type of thing was common in the Dark Ages and cannot be countenanced in 1985.' (The Star, 18 April 1985).

On the other side there are the performers who are saying such things as 'For how long will you tell us to show mercy to these "system"1 people when they kill us once we are in their hands? Why don't you allow us to deal with these dogs in the same way they treat us?' (The Star, 11 July 1985); 'It works, after this you won't find too many people spying for the police.' (Newsweek, 16 September), 'It's not possible to necklace a victim if you're not angry enough.' (Mail & Guardian, 28 August-3 September, 1992); 'Tshis'inyama' – Braai2 the meat' (The Star, 25 May 1993). When two rapists were tried by a kangaroo court3 and found guilty, they were necklaced in front of the school staff 'for everyone to see how comrades dealt with criminals.' (City Press, 4 July 1993). And Winnie Mandela created an outcry when she said at a funeral on 13 April 1986: 'With our boxes of matches and our necklaces we shall liberate this country.' (Uniform, August 1992).

The statistics on necklacings and burnings are bound to be inaccurate given the situation in South Africa, where there was much manipulation of information accessible to the public under apartheid, where events went unreported or undiscovered. Nevertheless, they may offer some indication of the general trends. According to the South African Institute of Race Relations (SAIRR), 672 people were burnt (half of whom were necklaced) between the beginning of 1984 and June 1987. The Herald newspaper in July 1990 stated that there had been 428 necklacings since 1985, and that between October 1989 and February 1990 there were 29 necklacings and 476 murders by other methods of burning (The Herald, 2 July 1990). According to the SAIRR, between July 1987 and June 1989, 15 people were burnt out of 661 political deaths. Between July 1988 and June 1989, 5 people were burnt out of 237 political deaths. Thus necklacings and other burnings appear to have died down as the nationwide uprising against apartheid died down in 1987, and to have increased at the end of 1989. They appear to be escalating once again.

Necklacing seems to be perceived by the general public as something which was suddenly invented by the black youth in the mid-eighties as a political tool, and was used to discipline 'collaborators' and political opponents. Some initial clarifications should be made here. Firstly, burnings began to be evident at least 7 years prior to the first widely reported necklacing in 1985. In Lebowa, during December 1976 and January 1977, 14 suspected sorcerers were burnt to death. Two of the victims were thrown onto a fire, the rest were locked in huts and burnt alive (Daily News, 5-15 January). Anderson (1990) writes that these burnings gained a gradual acceptance, and that by 1983/1984 this had become the standard means of disposing of witches. Various burning procedures were deployed. Some victims had petrol poured over them and lit, some were burnt in the open on a pile of wood, and others were locked inside their huts and burnt alive (1990:34). Thus the initial victims of burning would seem to have been witches, with the later inclusion of traitors and collaborators.

Secondly, following from the above, the perception that the first necklacing was in 1985 needs to be questioned. Anderson discovered a report suggesting that tyres had been used as early as 1981 (1990:34), and a newspaper report in 1984 describes how a witchdoctor and two women were burnt alive on a truck with a large tractor tyre placed over them (Rand Daily Mail, 25 February 1984). One will probably never know when the true necklacing (as defined) came into being, and when it was christened as such. Thus labelling Mr Kinikini as victim number one is perhaps more fiction than fact.

Thirdly, necklacings and burnings in general have been retribution for a vast array of misdeeds. The phenomenon does not simply belong in the political arena, it is a punishment used also against criminals, rapists, murderers, shebeen4 owners, other ethnic groups, witches and wizards. Indeed, in my research political burnings were certainly not the most prominent category which is what one would have expected given the general perception. Out of 65 random instances of burnings between 1984 and 1993, 39 victims appear to have been burnt following witchcraft accusations, 15 burnings were for political reasons (against collaborators, policemen or opposing political groups), 5 for medicine murder,5 4 for murder and 2 for criminal conduct. Male and female, young and old, have all been attacked. However, it should be noted that these burnings appear to be largely confined to the black community in South Africa; only one report was discovered where the victim was white.

Fourthly, the youth are not the only participants in these acts. They are often the more prominent actors, but there are many reports of elders as passive supporters as well as being directly involved. A final point of clarification: this is not purely an urban phenomenon. The initial burnings that I mentioned above occurred in rural areas, and from 1985 until the present day burnings and necklacings are found in both rural and urban areas.

Necklacing cannot be put into a neatly labelled box with its clear time, place and reason. The further one delves, the more complex the situation becomes. What follows is an attempt, no more, to develop a clearer understanding of necklacing. Given my limitations in research and space this paper will be restricted to three areas:

  • I will look at whether the present actions bear some relation to traditional beliefs. Do they make more sense when the cultural background is taken into consideration? I consider this justified because traditional Bantu notions in some form appear to still play an important role in daily life, even in the urban areas. The recent increase in witchcraft accusations and medicine murders, in both rural and urban areas, is evidence of this.

  • Possible reasons for the development of burning as a punishment will be examined. When and why do 'witch-hunts' develop? What relation do they bear to economic or social conditions? I will look at anthropological work done in South Africa, at other relevant theory and at other times in history. Here my focus will be on all methods of burning, and not specifically the 'necklace'. As I mentioned, necklacing is part of this wider phenomenon, and it is this which is important in the above context.

  • Finally, I will explore the attraction of these necklacings. I will argue that there are two important symbols which need to be distinguished, one is the necklace and the other is fire. I suggest that the necklace is a symbol which gains its potency more from the connotations that the word implies, whereas fire is a potent symbol in the visual sense. I will look at why these symbols are perpetuated, what they may 'mean', what role they may play in social life. I found anthropological writing on ritual helpful in this respect, and will use it as a guideline.

The research for this paper was conducted in Johannesburg. For a month I was based at the Centre for the Study of Violence and Reconciliation. I gathered information from libraries, from research organisations and from speaking to people. The anthropology department at the University of the Witwatersrand was helpful, and my thanks goes especially to Isak Niehaus.

The Relation to the Traditional

Here traditional ideas about justice, punishment, witches, burning and fire will be discussed, and used to suggest some possible meanings associated with the present day burnings. By traditional ideas I mean those beliefs which have been in the past, and may still be, common among the various Bantu-speaking people of Southern Africa. The arrival of western culture had a deep impact on these people, and it is this more recent moulding of traditional beliefs which would seem of greater importance when exploring the relation to present practices. The reason for including a discussion of witch beliefs is because the recent epidemic of burning appears to have begun with the burning of witches. From this discussion it becomes evident that those burnt – be they murderers, witches or collaborators – share a common crime.


Myburgh (1974) divides traditional Bantu ideas about justice into two areas – public law and private law. He writes that in the past those branded as public enemies were those who committed treason, contempt of court, assault, verbal abuse of elders, incest, sodomy, bestiality, rape, self-exposure, murder and sorcery. The punishment for these criminal offences was death, banishment, bodily correction, confiscation of property belonging to the accused's family, or a fine in livestock. However death, banishment and confiscation were confined to treason, murder, sorcery and incest (1974: 295-297). 'Death was brought about by stabbing, clubbing, impalement, decapitation, strangulation, stoning, burning, or exposure to ants or hot stones'. (1974: 299). He also writes that criminal conduct could be justified on the grounds of self-defence, defence of others or of property, and protection of public safety (1974: 296). It was the intent behind the act which was punished.

Western contact altered the political system. Chiefs gained greater protection from the government but lost much of their legislative and military power. Their judicial authority also weakened – they are unable to order the death penalty, severe corporal punishment or banishment (Myburgh, 1974: 291-2). Thus there would appear to have been a decrease in the harshness of the traditional justice system. Schapera writes of the Tswana:

The whole criminal jurisdiction of Tswana courts has been considerably curtailed since the introduction of European Administration; and for all practical purposes the punishments nowadays inflicted are restricted to fining and thrashing, with banishment as the penalty for extreme offences. (1970:50)

This decline in the severity of punishment is evident in the treatment of witches. Hammond-Tooke (1974) writes that the reaction to a discovered witch was violent in the past, especially among the Nguni, Tsonga and Sotho where death was the usual penalty. Various methods were employed, however he makes no mention of burning. He does say that among the Nguni the homestead of a witch was razed by fire (1974:357). Banishment was resorted to when the crime was seen to be not so serious. Similarly, no mention of burning was found in Junod's writings on the Tsonga (1962). In Hammond-Tooke's book on the Kgaga people he writes that in the old days balôyi (workers of evil) would be taken to the capital, and if found guilty they would be publicly executed by being battered with a club or by being pegged over an antheap with their heads shaved and smeared in fat.

The meting out of this violent punishment to alleged witches was curtailed as colonial ideas about the inappropriateness of witch beliefs became concrete in the form of laws. The Witchcraft Suppression Act of 1957 forbade the practice of witchcraft, accusations of witchcraft and consultation of witchdoctors. As Hammond-Tooke notes of the Kgaga practice: 'These times are no more…"Today many people are being killed by the balôyi but there is nothing that you can do about it except take him to the magistrate. The trouble is that they often employ a clever lawyer and get off. The White people have destroyed this custom because of money"' (1981:101). Public accusations of witchcraft were seldom made after the imposition of the Witchcraft Suppression Act, as was also the case among the Lovedu. Monnig, who did fieldwork among the Pedi in the early 1960s, writes that in former times witches were either clubbed to death, or were tied down at the entrance to a kraal where they would be trampled to death by cattle, or they were banished from the community. At the time of his fieldwork anti- witchcraft action was confined to consulting a diviner and taking protective medicine, fines and occasional banishments (1967:94-95).

From the 1950s and up until the end of 1976 the killing of witches appears to have been a rare event. Then, as mentioned earlier, 14 sorcerers were burnt to death in Lebowa during December 1976 and January 1977. I shall be considering possible reasons for this resurgence of witch killings from the late 1970s onwards in the next chapter.

What is a Witch?

As witches appear to have been the first victims in this latest epidemic of burning, and continue to be a high proportion of those burnt, it would seem useful to look at what a witch represents and how this relates to the idea of a traitor. Anthropologists, following Evans-Pritchard, tend to distinguish between witches and sorcerers. A witch is born into his/her position and harms by means of psychic powers, a sorcerer is not born into his/her position and uses material substances such as medicines and charms. However, Hammond-Tooke writes that this distinction is not always expressed terminologically, for example abathakathi among the Zulu and the Xhosa refers to both witches and sorcerers (1974:337). Given this, and the fact that there is much confusion in the use of the terms, 'witch' is used here in the broader sense to refer to witches, wizards and sorcerers. Hammond-Tooke writes: 'None of the South African Bantu have the concept of chance in their world-view. Apart from death from extreme age and minor illnesses such as chills and stomach upsets, all deaths and occasions of misfortune are believed to be caused ("sent") by some external agent' (1974:336). This agent may be an ancestor or a human being using supernatural means (a witch). This second type of misfortune is through no fault of one's own, one is a victim of another person's malevolence.

Witches represent behaviour that deviates from the accepted norms of a society, they are evil and create disharmony in social relationships. 'To call someone a witch is to say that they are a traitor, that the person stands in an antagonistic relationship to the rest of society' (Ritchken, 1989b:18). To the believer:

It represents the dark, malevolent urgings that lie at the heart of man – the product of hate, envy, revenge and malice. It threatens the very basis of social life and it must be detected and combated with every means at one's disposal. It represents the African version of medieval demonology with which it has, of course, close parallels. (Hammond-Tooke, 1974:336)

And further, 'it is this idea of treason against the kin group itself, an attack on the very basis of the social structure, which makes witch activity such a heinous offence. It is the quintessence of immorality' (Hammond-Tooke, 1974:337).

I suggest that if one considers witches at this level, instead of as toothless hags flying around on broomsticks, the connection between these and other victims of necklacings (or burnings in general) become apparent. Victims such as 'collaborators' are also seen to be traitors, they break the social solidarity of the group, they 'sell out' to the enemy and assist in the continued suppression of the black community. They deviate from the expected norms of society. Ritchken has hinted at this when he writes:

In as much as a community councillor is necklaced as a symbol of a system of antagonistic relationships (apartheid) and in as much as the public nature of this punishment serves as an example for all to see and learn from, so the burning of witches served the same function for the people living in Mapulaneng. (Ritchken, 1989b:18)

The fact that at the start of these burnings witches appear to have been the dominant victims, and that only later did collaborators begin to receive the same punishment is of secondary importance. Given the situation in South Africa during the mid-eighties with much political unrest and frustration, the transferral would seem almost inevitable, and from the mid-eighties 'apartheid's spies and puppets and "witches" often became interchangeable' (Ritchen & Koch:1991). The issue here is why the need was felt to expunge the society of evil traitors by such a dramatic method, not what the specificity of each traitor's crime was. Thus, through looking at what traditional notions of a witch symbolise, one has been able to gain a clearer understanding of the similarity between a witch and a collaborator, and of the reasons for their punishments.6


The symbolism behind burning needs to be discussed in relation to traditional beliefs. Minnaar et al write about notions which they claim are still held in Venda in the present day: 'The belief is that a witch's body must be completely destroyed together with all the magic paraphernalia that the alleged witch might be storing in the hut and the best way to do this is by burning' (1992:23). It has also been suggested by Minnaar that when a witch is burnt it is believed that his/her soul is destroyed. This is said to be the worst possible punishment as it breaks the link with the ancestors. Ancestor worship is important among the Bantu, and for the dead person's children the burning of the soul would imply that the link with the ancestors has been broken (conversation with Minnaar). This is supported by a quote which Bergland gives from a Zulu informant:

When snakes have been killed they must be burnt with fire on the garbage heap outside the homestead because fire destroys them totally. They are rubbish. Snakes and abathakathi (evildoers) are the same. Both are just rubbish. So that is why abathakathi ought to be treated with fire (ie. burnt). … Yes, there was a reason for the burning of abathakathi like rubbish. It was the shade. If a person was burnt with fire, like the fire from the sky (ie.lightning), there could never be a shade of that person. So if a person was umthakathi that person must be destroyed totally (ukuqedwa), nothing remaining (kungasali nokunci). His medicines and familiars together with all his things must be destroyed. (1976:305)

However as imputation of witchcraft has been a criminal offence under South African law since the 1950s, this burning of evildoers did not appear to happen. This informant had witnessed only the burning of the homestead of a person who was accused of ubuthakathi. Whether this idea is common to all Bantu people is debatable. I could find no other reference in the ethnographies examined and Niehaus commented that he had never come this idea in his fieldwork. Rather, among the people of Green Valley in Lebowa, the soul (or the shadow as it is sometimes known) is believed to leave the body of the person as soon as he/she is dead, however he/she had died.


Further beliefs about fire might be helpful in understanding burnings. Krige & Krige (1947) write that among the Lovedu fire is seen in three ways. It is seen to be valuable, it is seen to give out a harmful 'heat' which must be 'cooled', and thirdly fire belongs to a pattern 'which links it up with crises or great change, ritualises its value, and uses it to extinguish an evil or to purify a state of defilement' (1947:168). According to Hammond-Tooke (1981), a sacred fire must be kept burning during the initiation period of the Kgaga. The initiates are seen to be in an impure state, and spend this time in a secluded lodge. At the end of the initiation period the hut is burnt as the initiates walk away, they must not look back. This burning represents a symbolic destruction of their impure state, and of all childhood things (Hammond- Tooke, 1981). This use of the fire during initiations is also found among the Lovedu (Krige & Krige, 1947:160). Junod (1962) writes that if taboo children such as twins have not been buried correctly they will cause great misfortune to the whole land. These children must then be exhumed and reburied near the river. Among the Bavenda of the Northern Transvaal the burial party goes through a purification ritual to cleanse themselves from the dangers of contact with the corpse. This involves washing in the river, and on their return drawing their hands and feet through a fire (Stayt, 1931:164). From the above examples it would appear that there may be some idea among those who participate in burnings that fire can destroy evil, that it can cleanse the society.

The use of the all-consuming fire as a punishment can be found throughout history. It has been employed elsewhere in Africa. For example, among the Nyoro, sorcerers were wrapped in banana leaves and burnt, and among the Ndembu a person divined as a troublemaker, sorcerer or a witch was put to death by burning (Mair, 1973:11). Witches were burnt during the European witch-craze which grew after the Renaissance. De Coning and Fick quote John Mountcastle as saying: "The Old Testament is full of references to instances in which fire was credited with destroying the houses of wicked sinners, serving as a vehicle for sacred pronouncements, or consuming unbelievers. The inherent fear of flame was incorporated into the Judeo-Christian religious tradition and capitalised upon by religious leaders" (1986:29). Allister Sparks writes, as he watched a victim burn at the Mxenge funeral: 'it put me in mind of the burning of the great whore in Revelations who corrupted the earth with her fornication' (1991:263). And Ebersohn gives further examples: 'The Slavs believed that the holy fires of the holy night of Kupula possessed a purifying virtue … Assyrian legend speaks of the greatest possible punishment being the all-consuming fire. The Judaic scriptures tell of burnt offerings to God and the New Testament threatens a punishment for the unrepentant that is described as an everlasting fire'(1987:41). It seems that this notion of the purifying flame can be found in many diverse cultures.

In this chapter there has been a brief discussion of ideas which would seem helpful for a deeper understanding of the practice of burning witches and others. The Bantu system of justice in the past sanctioned harsh punishments (including burning) for those guilty of witchcraft, incest, murder and treason. But with the imposition of western precepts there was a marked lessening in the severity of retribution, especially during the latter half of this century as the power of the chiefs was eroded and anti-witchcraft laws were introduced. Possible reasons for the recent resurgence in brutal punishment will be looked at in the next chapter. Further, a witch was seen to represent behaviour that is anti-social and evil and is seen to betray the society. Thus a link can be established between the victims of necklacings: collaborators, witches and murderers. They all threaten the cohesion of the community in that they break the social norms in a devastating way. And finally, there would appear to be some notion that fire can destroy evil, that it can purify.

Burnings in their Social Context

In the previous chapter a witch was seen to be anti-social and evil, as someone who threatens the social cohesion of the society. Here I will look at how these beliefs function when placed in their social and political context. I will discuss some anthropological theory on the occurrence of witchcraft accusations and of 'witch-hunts' in the broader sense, especially in times of social and economic change. I will then look to the situation in South Africa, and compare it to other places and other eras in history. I have included this section because I felt it was important to look at why the initial burnings began. Given that the first victims appear to have been witches, I have focused on witchcraft but much of what will be said could be applied to the broader category of social deviants. As has been mentioned, these first burnings were in rural areas and the details in this chapter reflect this. However the broader idea that economic and social change is related to 'witch' (traitor) burnings can be applied to both the rural and urban way of life (see Chapter 4).

Out of Evans-Pritchard's Witchcraft, Oracles and Magic, comes the idea of witchcraft as a 'homeostatic control system' (Douglas:1970), where a belief in witchcraft is seen to be a controlled outlet for anxiety and hatred, helping to maintain the social system. Following from this, anthropologists such as Gluckman have associated a rise in witchcraft accusations with the moral breakdown of that society. Witchcraft was seen to increase in Africa as contact with missionaries increased and as people began to live and work in towns. Other anthropologists, such as Marwick, Middleton and Turner have taken the contrary view. Here witchcraft accusations are seen to reflect underlying social tensions, and function as a means of breaking off social relations. Through the accusation he/she is able to break off an unwanted social relationship with, for example, a close kin. According to this theory we should see a decline in witchcraft with urbanisation, as there is no longer this close-knit kin group and the accompanying social obligations.

For Douglas (1970) the functional model of witchcraft (which both of the above are) is equivalent to a scientific paradigm in the sense suggested by Kuhn in The Structure of Scientific Revolutions. Kuhn argues that the anomalies within a paradigm will gradually accumulate to the point where it must be rejected and replaced by a new paradigm. Douglas feels that the functional paradigm for witchcraft has too many inconsistencies, and that a new paradigm needs to be found to understand witch beliefs. She suggests: 'The symbols of what we recognise across the globe as witchcraft build on the theme of vulnerable internal goodness attacked by external power' (1970: xxvi). She classifies witch beliefs into two main patterns. The first is where the witch is an outsider and the accusation serves to reaffirm group boundaries and solidarity. The second pattern is where the witch is an insider; here 'The symbols of the attack by witchcraft tend to make the body of the victim into an image of the betrayed community' (1970:xxvii). She further writes: 'So it would seem that the way the witch works, his sources of power, the nature of his attack on his victim, all these can be related to an image of the community and the kind of attack to which community values are subject' (1970:xxviii). The new paradigm which Douglas proposes is that witchcraft beliefs will be found where there is intense social interaction which is ill-defined. In contrast, where social relations are diffuse, or fully ascribed, they will not be found. The former situation is what we find in both the rural and the urban areas in the contemporary South African scene, and this theory may thus be useful for an understanding of the increase in witch burnings.

Steadman (1985) argues that the purpose of witchcraft accusations is to intimidate the opposition. From his fieldwork among the Hewa of Papua New Guinea he proposes that, given the fact that witch accusations are untrue and unverifiable, they are purposefully directed at the weaker members of threatening groups. This serves as a warning to the threatening group, an attempt is being made to intimidate the opposition through the killing of their 'witches'. In his explanation, Steadman seems to disregard the often very real belief in witchcraft; as Niehaus writes, 'witchcraft presents a personalised conception of evil which enables people to explain, diagnose and compensate for unmerited misfortune' (1991:3). Thus while this idea of intimidation may be useful in understanding the South African situation it must be considered as part of an explanation and not the entire reason for the witch burnings.

Witch-hunts are the organised enactment of witch accusations. One can consider these 'witch-hunts' on a more general level than that above, where the 'witch' is any internal enemy that society hunts out. As was shown in the previous chapter, a witch deviates from the expected norms in the same way as political subversives and criminals do. Any of these deviants, as defined by the society, can be subject to 'witch-hunts'. Bergesen suggests that these hunts serve 'as ritual mechanisms for the periodic rejuvenation of collective sentiments' and that it 'is a mechanism for renewing common moral sentiments and redefining the contours of the social reality' (1977:220). The crimes may be imaginary but this becomes unimportant here; what is important is that people believe that they are cleansing their society and through this feel their collective morality to be reaffirmed.

Let us now turn to some South African material. As mentioned earlier, 14 sorcerers were burnt to death during December 1976 and January 1977 in Lebowa; prior to this burnings seem to have been an unusual event. As Anderson writes, 'There can be little doubt that the events of January 1977 represented some kind of rupture, both in the way matters of witchcraft were perceived by the press and the public at large, and in the more fundamental significance that the incidents concerned seemed to herald for patterns of witch murder in rural Transvaal' (1990:24).

There appears to be a gap in research over this period. However the work done by Ritchken and Niehaus in connection with burning epidemics in the 1980s in Lebowa suggests possible explanations which may also be applicable to the 1977 burnings. Niehaus looks at the increase in witchcraft accusations and witch-hunting in the village of Green Valley, Lebowa. He describes the gradual erosion of the power of the chiefs, and how this political vacuum was filled in the mid-eighties by the 'comrades' of the Brooklyn Youth Organisation. This youth organisation, in a culture where age is associated with authority, 'faced a severe legitimation crisis' (1991:15) and one way in which they sought to gain acceptance was by pledging themselves to the elimination of evil. During 1988 and 1989 witch-hunts were conducted, and over the December holidays of 1990 34 people were accused as witches. Niehaus suggests that the witch beliefs should be seen as real, and that these witch-hunts were seen to be necessary by the community. Rather than seeing these witch-hunts 'as a covert attempt to mystify subordinates or intimidate political opponents' (1991:3) they should be seen as 'an attempt to eliminate misfortune, perform a valuable social service, and thereby attain political legitimacy' (1991:24). 'Engagement by political actors in the management of misfortune presents a potent source of political legitimacy' (1991:3).

Ritchken has written about the Mapulaneng district of Lebowa where, in April and May of 1986, 150 people were labelled as witches and about 30 of them were necklaced. He gives a very detailed description of life in this area in the time leading up to the killings. Migrant labour provided the main source of income to the homelands prior to the recession of the early 1980s. With the men away, the women became the effective household heads. This recession led to a very high unemployment, especially among the male youth. Returning jobless to the homeland, they found themselves without a role to play in the society. They were forced to live off their grandparents' pensions, which resulted in much friction. Further there was an increase in neighbourly conflict as resources became scarce, and people could no longer afford to be charitable. These new family forms and the recession had a deep impact on the unemployed youth. Ritchken writes:

These experiences within a world-view that sanctions pure harmony have the status of anomalies. And it is these anomalies that are explained with reference to herbs and witchcraft … . The witch attacks are a warning of a potential punishment for all those people involved in relationships that the youth define as anti-social … . It was an implicit threat and it produced fear and anxiety. It was the imposition of a new discipline in each and every household, a demand for co-operation, in the context of a society competing for scarce resources and riddled with potential points of conflict. (1989a:14)

In a later article Ritchken & Koch write: 'Witches provide an explanation for hardships as well as an active way to remove the agents of oppression' (1990).

Thus in the above studies of South Africa burnings are related to upheaval in the political, economic and social arenas. The resettlement of people with the destruction of old kinship patterns of residence, a decline in agricultural subsistence, migrant labour, high unemployment, fierce competition for resources, loss of authority of the chiefs and the social ostracism of the youth are all seen to contribute to the smouldering coals of social tension. This argument can be supported by events in history such as the European witch-craze. MacFarlane relates witch accusations in Essex to economic and social upheaval. He writes that as social relationships changed, witchcraft came to 'provide many acceptable mechanisms for overcoming uncertainty and anxiety … a whole detailed set of counter-witchcraft rituals, and myths about the evil doings of witches, gave ample scope for activities and outlet for repressed fears and worries' (1970:95). Ardener (1970) looks at the Bakweri of Cameroon and, in a similar vein, relates their witchcraft beliefs to their economic situation. The Bakweri faced difficult times between 1894 and 1954, and Ardener notes that during this time witchcraft beliefs took on a morbid form. However, between 1954 and 1961, when the Bakweri earned a large amount of money from selling bananas the witchcraft threat came to be seen as less ominous. Some of this new money was used to rid the community of the witches. Thus it would seem that changes in the social, political and economic arenas can result in the fierce hunting out of traitors or witches or scapegoats, as a means of overcoming uncertainty and restoring order.

In summary, while these burnings appear to be effective political tools with which to assert power and to intimidate those in opposition, they cannot be solely explained by this (as Steadman would be content to argue). An unprecedented misfortune, such as a car accident or the death of a person by lightning, seems often to ignite a burning episode. An anonymous schoolteacher in Lebowa commented: 'Something must be done … to root out this firm belief in witchcraft … . We welcome the rains for our crops, but live in mortal fear of what the lightning may bring' (Drum, March 1984:6). They also serve to explain the world on what would appear to be two levels. Firstly, the 'why' behind an event is addressed in contrast to our notion of 'chance', and witchcraft would seem to be a very real explanation for events which disturb social harmony and it should not be lightly dismissed as an explanation for burnings. Niehaus respects the validity of witch beliefs, which leads him to conclude that it is through the useful task of cleansing society that the youth gain political legitimacy. On a second level, I would argue that through this expunging of the internal enemy (as Douglas argues) burning fulfills a deeper need. It gives meaning to social life, which is clearly needed to mask the uncertainty left by the destruction of former beliefs and practices. As Bergesen suggests, it redefines the moral norms and collective sentiments of a society. People need meaning to exist as social beings and I would argue that this becomes an important part of why the burnings occur, through these communal acts sense can be created where none seems to exist. This will be further explored in the next chapter in relation to ritual. The argument is supported by Ritchken, who sees the burnings as a means by which the youth can give some meaning to their anomalous position by this implicit threat to those who oppose them. It gives them a feeling of power and a sense of control over their lives. The argument also fits in with Douglas' proposal that witchcraft accusations will occur where social relations are ill-defined and yet intense.

The Attraction of the Ritual

In my reading of reports on necklacings and burnings, I was struck by the similar pattern which many of them follow. My interest here is in what would appear to be the ritualisation of burning. I will be using the term 'necklace' because symbolically it seems to play an important role in this ritual. It should be remembered that the victim may go through this ritual and be burnt by some other means, such as having petrol poured on him/her which is then lit. With the aid of theoretical writing on mainly secular ritual I hope that a deeper analysis of the attraction and meaning of the necklacing act will be achieved.

There has been much written on ritual. Durkheim forged the way in The Elementary Forms of Religious Life where he analyses the ritual of the Australian Aborigines. He argued that collective rites symbolised the social cohesion of the group, while at the same time ensuring their social solidarity (Moore & Myerhoff, 1977:5). Until fairly recently ritual in the religious domain has been the focus of analysis, while the ritualisation of aspects of secular life was hardly explored. Ritual had also been largely seen to uphold the status quo and its power as a political tool in times of change went unnoticed. I have found the writings of Moore & Myerhoff, Turner and Kertzer especially helpful in that they address this use of ritual as a political tool and explore the meaning of secular ritual.

Turner claims to see ritual in a post-modern light, where the emphasis has shifted from structure to process. Ritual is a 'complex sequence of symbolic acts', a 'transformative performance revealing major classifications, categories, and contradictions of cultural processes' (Ronald Grimes, quoted in Turner, 1992:75). Turner focuses on 'social dramas', as units of 'disharmonic social process, arising in conflict situations' (1992:74). This sequence of acts has four main public phases, with ritual action (as Turner sees it) possibly occurring in the third stage. These are:

  • A breach of the regular, norm-governed behaviour;
  • the crisis;
  • the redressive action within the frame provided by institutional procedures which may be law-like or in the form of ritual;
  • the reintegration of the disturbed social group or the recognition of irreparable schism (1992:74-76).

He writes of the third stage:

Cognition reigns primarily in judicial and legal redressive action. Where such action fails, however, to command sufficient assent, will and emotion reassert themselves … there may be an attempt to transcend an order based on rational principles by appealing to that order which rests on a tradition of coexistence among the predecessors of the current community … . This kind of ordering is better regarded as the crystallisation of joint experience, handed down in striking or potent cultural forms and symbols and bears rather the character of orexis (feeling and willing) than rational planning … . Legal action itself, of course, is heavily ritualised. But in these more fully ritualised procedures what is being introduced into situations of crisis is the non-rational, metaphorically 'organic' order of society itself, felt rather than conceived as the axiomatic source of human bonding. It is the 'social will'. (1992:91)

All of the above seems applicable to the burnings. However I would question his argument that the ritual only comes in where legal redress fails. In necklacing, the legal and the ritual aspects would appear to go hand in hand, there is not one or the other. The legal aspect, whether a kangaroo court is held or a witchdoctor is consulted, is undertaken with the knowledge that it will be followed by a burning. Whether the necklacing occurs is not dependent on the legal procedure commanding sufficient assent; for example, in the kangaroo courts the evidence is usually based 'on the word of faceless accusers, and the "accused" is almost always guilty even before the hearing' (Sunday Star Review, 7 July 1991), there is no defence and no appeal. The necklace is the punishment that goes with the verdict. It would thus seem that one has to either say that the necklace is part of the legal procedure and as such is not the type of ritual that Turner writes about, or else that the whole of this redressive stage is a ritual with the cognitive part reinforcing the emotional part and vice versa. The second seems more plausible here; the distinction between the legal and the ritual does not seem valid.

As Parkin (1992) notes, there are many problems with defining a 'ritual'. Kertzer sees it as 'symbolic behaviour that is socially standardised and repetitive' (1988:9), 'an analytical category that helps us deal with the chaos of human experience and put it into a coherent framework' (1988:8). Moore & Myerhoff define ritual as 'a set of formal acts which deal with or refer to postulated matters about society or ideology' (1977:22). With reference to necklacing, these definitions enable one to consider both the events surrounding the burning and the actual burning itself. I feel that this is necessary, because the necklace would appear to have more symbolic value when spoken about than it does when it comes to the actual burning. I mentioned the apparent interchangeability of different forms of burning. Whether physically the necklace is used is probably largely dependent on the availability of tyres, and where the tyres are used it may be simply that they are an effective aid in burning and that they debilitate the victim. During the actual act, I would suggest that it is the burning rather than the necklace which gives 'meaning' to people.

Moore & Myerhoff see collective ritual as 'an especially dramatic attempt to bring some particular part of life firmly and definitely into orderly control' (1977:3). Some features of these collective ceremonies are repetition, acting (not a completely spontaneous event), the use of symbols and actions which are out of the ordinary, an evocative presentation style, order in the events and a collective message. Ritual 'veils the ultimate disorder, the non-order, which is the unconceptualised, unformed chaos underlying culture' (Moore & Myerhoff 1977:17). Through its structure, the connections it makes and the explanations it provides, it portrays ordered 'reality' as postulates that are difficult to question, it endows particular interpretations with legitimacy. Any aspect of behaviour or ideology can be ritualised, and 'once used in a collective ceremony, whether performed for the first time or the thousandth, the circumstance of having been put in the ritual form and mode, has a tradition-like effect' (Myerhoff 1977:8). Ritual can be used to make and mark change, often with the use of modified traditional symbols. Through the use of symbols, ritual 'shows' ideas or values that are invisible in daily reality and they are objectified and reified. These symbols are presentational rather than discursive. Parkin writes that 'it is precisely because ritual is fundamentally made up of physical action, with words often only optional or arbitrarily replaceable, that it can be regarded as having a distinctive potential for performative imagination that is not reducible to verbal assertions' (Parkin 1992:11-12). The message is conveyed dramatically thus enabling the connection to be made between the tangible and the invisible ideals. Emotion, imagination and belief are called into play and 'participants are able to conceive the invisible referents of symbols used in rituals. Sometimes, especially in religious rites, they have the sense of envisaging the essential pattern of human life, its relation to the natural and cosmic orders, and achieve thereby an immediate triumphant sense of knowledge and belief' (Moore & Myerhoff 1977:13-14).

Kertzer looks at ritual and its importance in politics, both for those in power and those in revolt. He sees rituals and their symbols as crucial to politics, 'Political allegiances flow not from culture-free judgements but from symbolically nourished conceptions of the order of the universe' (1988:182), 'It helps provide legitimacy at the same time as it mystifies actual power relations, it facilitates popular solidarity even where consensus is conspicuously absent, and it leads people to conceive of their political universe in certain ways' (1988:153). He further notes that it is one of the most potent weapons that the powerless have. Ritual is not only to commune through, but it defines relations with those outside the group. Said one Ukrainian, 'It is impossible to break people, to make slaves out of them, until you steal their holy days from them, and until you trample upon their temples. There could be no Mau Mau movement without Mau Mau rituals, nor any popular anti-Vietnam war movements without its public rites' (1988:181).

What follows is an attempt to analyse necklacing procedure as a ritual, using the above as a guideline. It should be remembered that my analysis is limited, given the difficulties in understanding the different levels of meaning and the extent to which the ritual relates to a larger reality. As Moore & Myerhoff write: 'the messages carried by rituals are extremely complex with implicit and unconscious elements not easily interpreted, and consequences difficult to measure' (1977:15).

Often the victims of a necklacing are killed after a generally structured process. An accusation is made, either of suspected witchcraft, murder, rape or collaborating. If it receives wide support a witchdoctor is consulted to verify the accusation or to name the accused, or a people's court7 is held to 'try' the accused. Having 'confirmed' the guilt of the person, he/she is then burnt. This may not be done immediately, often the victim is told to report the next day for the necklace. While only a few people are directly involved in the trial and burning, they are often witnessed and passively supported by large crowds (up to 500 people).

The actual burning is also largely ritualised. The victims are often stabbed, beaten, hacked with axes and stoned. They may be dead or alive when the actual burning takes place. They may have had their hands chopped off or their hands and feet bound with thin wire. In some cases, once they have been doused in petrol and a tyre has been placed around their neck, they are asked to give R1 (20p) for 'prayer money'. In return they receive a marijuana cigarette to 'ease the pain' of burning. They are told to light it and as soon as the match is struck the petrol ignites (Sunday Times, 27 October 1985). Numerous reports were read of the perpetrators dancing around the burning body, chanting such things as 'Siyitholile Izinja. Siyazitshisa' ('We found the dogs and we are burning them') (unknown source of newspaper article, 1993). The bodies are left to rot in the open, sometimes the heads are decapitated and paraded around on broomsticks (City Press, 30 August 1992). The necklace is also referred to as the 'guillotine', '"young revolutionaries" … compare their actions to the time when the French used the guillotine to eliminate "collaborators" during the French Revolution' (City Press, 30 August 1992). The victims become what is called "Kentucky fried chicken" (Newsweek, 16 September 1985).

Moore & Myerhoff suggest five ways in which the outcome of secular ritual may be looked at. I found this useful and have used it as a framework.

  • The explicitly stated purpose

This is to condemn and punish those who are said to have betrayed the society. Collaborators, informers and policemen are punished for 'selling out' the people, for working in the system of apartheid. Witches are accused of killing people to use them as zombies,8 of killing people by lightning strikes and falling stones, of poisoning and bewitching people. Others have been necklaced for rape, for ritual murders and for being a political opponent. Moore & Myerhoff comment that this meaning 'is often likely to be the most superficial' (1977:16).

  • Explicit symbols and messages

As Kertzer notes, 'a flag is not simply a decorated cloth, but the embodiment of a nation; indeed that nation is defined as much by the flag as the flag is defined by the nation' (1988:7). Two symbols need to be looked at in this context, the notion of the necklace and the consuming of a person by fire. The necklace image is of a person with a tyre pinning his/her arms to the body, he/she is on display as an abnormal member of society by being artifically rendered as physically abnormal. One is reminded of a person being put in the stocks. Given that those receiving this punishment are those who have deviated from the social norms of the group, the symbol of the necklace is perhaps associated with ideas such as that a social community exists which is important, that it is immoral to act anti-socially and that it will not be tolerated, that communal interests should be placed above individual greed, that social harmony is important. For example, two youngsters who had necklaced a former mayor were reported to be 'convinced that the gruesome method sends a clear message to would-be collaborators that "our struggle is serious"' (Mail & Guardian, 28 August-3 September 1992). It also conveys a message of what the relationship is to those outside the group. For example, where the youth gain political power through the threat of the necklace, it intimidates those who would oppose them. The necklace defines what behaviour is expected of the community, and defines the power relationship. On a broader scale it can define the relationship of the community to the system of apartheid and white supremacy, that it is unjust and that it must not be supported.

Other messages that the necklace could convey need to be considered. Franklin (1963) writes of the Cewa that the neck and the stomach area are seen as links between the three main contrasting divisions of the body. This horror of the separation of the head from the body is perhaps fairly universal, and the burning necklace could be seen as a decapitator in a sense. It is perhaps of some relevance that the 'guillotine' (another term for the necklace) portrays this same idea. The term necklace could also have an ironic association. It usually carries positive emotional associations. Necklaces are given as gifts, they are decorative and are normally seen to enhance one's appearance. Also protective medicine is sometimes worn around the neck. For example, Sosibo (1992) writes that the 'comrades' in Pietermaritzburg would be given protective medicine from traditional healers, and that this could be tied in a small bottle around their necks. With the necklace becoming a burning tyre, there is a direct symbolic reversal of the term which is the more potent because of the strong positive connotations on the one side and the very strong negative connotations on the other. This potency would seem to explain why the term has become so renowned and is used outside of its strict sense, as was mentioned in the introduction.

Turning to fire, Kertzer suggests that rituals are invented out of pre-existing symbols, and that they become established because of the social circumstances of the participants, and not because of the inventor. It was seen that burning as a punishment was used in the past; thus one could say that it is a traditional symbol, which still holds some common association that has been used in a new ritual. Burning was seen to possibly symbolise a number of things. It may be associated with the destruction of the soul of the person, thus breaking the link with the ancestors, it may signify the destruction of evil or the purification of the society. Obviously, in the physical sense, the victim is unrecognisable after he/she has been burnt, there is a visual destruction of the person. It should be noted that as the victim is often dead when he/she is burnt, fire must be seen to fulfill some purpose other than as a means to kill.

This message of the fire may be different for each participant and observer but this is not a source of conflict. This visual act enables solidarity to be built without the necessity of shared belief. As Fernandez (1965) writes, the 'cultural consensus' is sacrificed for the 'social consensus'. Related to this idea is Parkin's suggestion (see above) that messages are transmitted much more powerfully by drama than by words. I would suggest that when the victim is actually burnt it is the fire that takes over from the necklace as the potent symbol. The burning of a person is a highly visible and dramatic act, and it would seem a powerful stimulant of the senses. Emotion rather than rational argument makes sense of the act, the messages behind the necklace are now rather felt than heard. This is evident in the accompanying chanting and dancing and euphoria which accompanies burnings. Kertzer suggests that this increased emotion makes one less able to distinguish categories of others: 'Carried to an extreme, the emotionally-charged individual may operate with an overriding cognitive division of people into just two categories: "with me" and "against me"' (1988:82). The whole ritual performance involves a powerful combination of two potent symbols with the fire taking over from the necklace and making the final culmination focused and intense.

  • Implicit statements about less conscious social and material aspects

Moore & Myerhoff write that at this level the ritual 'may express deep contradictions in the social or cultural system – all kinds of troubles, uncertainties, conflicts and paradoxes' (1977:16). It was shown in Chapter 3 how the witch burnings were largely seen to be a result of social changes. These were caused by such things as forced resettlement, the gradual erosion of the power of the chiefs, the breakdown of agricultural subsistence and the reliance on migrant wages. Kinship patterns of residence were disrupted, resources became scarce and unemployment increased. The new social relations became uncertain and ill-defined.

The first official necklacing occurred in an urban township, and this great social uncertainty can also be found there. In the mid-eighties the townships became more politicised. The economic situation worsened, there was a recession with high unemployment. At the end of 1984 school pupils began a campaign against poor education and many schools were boycotted. The administrative structures collapsed, and by 1985 only 5 of the 38 black local authorities could operate effectively (SAIRR, 1985:89). Morris & Hindson write that the government reforms in the 1980s led to the development of shanty towns. There was increasing class differentiation as rapid urbanisation led to a shortage of space and resources. 'The struggle for new economic and social positions by the poorer and more marginalised sectors of black society, within shifting and uncertain social and economic contexts, is one of the fundamental roots of the violence' (1992:46-47). It is evident that at the time of the 'first' necklacing there was much uncertainty in the townships. Social and political tensions were high. Also, one should not forget that apartheid was maintained by the use of force. People were often killed cruelly and unjustly. The necklacing of Mr Kinikini occurred a few days after 20 people had been shot in the same area by the police while they were on their way to a funeral (15 were shot in the back). At the time it was very difficult for black people to oppose apartheid and so their frustration would appear to have turned inward, to the excising of the internal 'enemy' in an attempt to grasp some form of control over their lives. It should be remembered that burnings are powerful visually, this dramatic and ritualised punishment creates a strong impression of order and control.

  • The effect on social relationships

Turner writes that the restoration of peace in a social drama involves a change in social relations, 'New power may have been channelled into new authority and old authority lost its legitimacy. Closeness may have become distance and vice versa' (1992:92). How could social relations change through these ritual burnings? They may place the perpetrators in a new position of power and authority, either through what is seen to be socially useful action (as Niehaus suggested), or through intimidation. The often very large crowds evident at these events would suggest that a new sense of closeness and unity is generated, also De Coning & Fick write that those who do not actually partake in the burning may throw stones at the corpse as they depart (1986:35). If the victim is seen to be the internal enemy, there must be some sense of the reasserted strength of the community.

  • A cultural statement about cultural order as against a cultural void

Moore & Myerhoff write that at this level ritual 'is a declaration of form against indeterminacy' (1977:17), 'it celebrates man-made meaning, the culturally determinate, the regulated, the named, and the explained. It banishes from consideration the basic questions raised by the made-upness of culture, its malleability and alterability' (1977:16-17). A way in which cultural order is given meaning is by linking the past to the present, and the present to the future (Kertzer, 1988:10). This is supported by a comment of Chief Ledwaba, in reference to some recent witch burnings in Lebowa: '"White people do not understand black magic. We deal with witches according to our custom, the way our ancestors taught us"' (Drum, March 1984:6). The act is made to appear timeless, it is sanctioned by tradition which makes it seem fixed and unquestionable. The indeterminacy of life is hidden by defining reality. The extent to which this is true is certainly debatable. It should be noted that it is still a minority of the black population that is associated with this ritual, and that there is strong condemnation of the act from all political parties and many independent observers. It may give meaning but it also raises questions.


In this paper I set out to explore a new form of punishment, the 'necklace', which has become well-established in South Africa as a means of extra-legal redress. Burning had been a means of punishment in the past, but had become a rare occurrence this century. However, from the late 1970s burnings began to escalate. In 1985 the necklace gained notoriety, and until this day necklacings and other forms of burnings continue to claim a significant number of lives.

Traditional Bantu notions were explored. The initial victims of the burnings were witches and they continue to be a large proportion of those necklaced or burnt, thus the notion of a witch was discussed. It was suggested that there is a strong connection between the victims of burnings, such as witches and collaborators, in that they are seen to 'sell out' to the community and break its social solidarity. The fact that the initial burnings were to punish witches and only later came to include collaborators then becomes understandable, as it was only in the mid-eighties that the urban areas became increasingly political and where 'collaborating' became another means by which communal values could be betrayed.

As to why burnings began, it was suggested that the initial rural witch burnings developed in a climate of great social and economic change. There was intense social interaction which had become ill-defined, due to such things as high unemployment, migrant labour and the demise of chiefly authority. This great uncertainty was also to be found in the urban townships in the mid-eighties, which was when the necklace gained notoriety. It was suggested that these 'witch hunts', of traitors or murderers or witches, are a means of reasserting communal values. The internal enemy is expunged in an attempt to grasp some form of control over the new uncertainties.

To try and understand why the 'necklace' gained such notoriety and continues to exist, the necklacing act was analysed in the framework of a ritual. Possible different levels of meaning that people might gain from this behaviour were suggested. The wider context of burning was seen to be indispensable to the discussion of necklacing; it has even been argued that the necklace has more value when it is spoken about and that whether a person is actually necklaced probably depends largely on the availability of those effective aids, the tyres. I have suggested that the 'necklace' is an extremely potent term because of the strong contrast with its usual positive connotations, and because of the image of severing one's head from one's body. During the actual burning of the victim it would seem that, rather than the necklace, it is the all- consuming fire which makes the act so powerful and attractive. This would account for the fact that many forms of burning are found. The victim is defaced beyond recognition, and perhaps through this destruction comes the feeling of purification from evil and control over one's life. I do not deny that the necklace has symbolic value in the actual burning; when an actual necklacing occurs the combination of the two symbols is probably even more powerful. I simply suggest that the necklace is the more powerful symbol when spoken about, and that the fire comes to dominate in the actual burning.

My argument developed out of what information there was available to me in this research. The restrictions of time and the difficulties in researching a topic such as this have determined the boundaries of this project. If I have stimulated debate, or further research into this area then I feel that it has been a fruitful endeavour.


1 Those who uphold the system of apartheid.

2 A South African term for a barbecue/to barbecue.

3 These courts, illegal under apartheid's laws, developed in the townships of South Africa during the mid-eighties as part of the informal justice system.

4 An illegal drinking venue.

5 A person is killed for bodily parts such as the genitals, nails, skin and tongue. The medicine made from these is believed to help attain wealth and in some cases bring good luck in warfare. It should be noted that these 'muti murders' appear to have also increased considerably. Balic (1990) writes that there had been at least 30 'muti murders' since 1987 and quotes Hammond-Tooke as saying: 'The incidence of medicine murders rises dramatically at times of political and social tension, insecurity and competition for economic resources' (1990:32-33). Thus the turmoil in South Africa has seen the escalation of both witch accusations and ritual (muti) murders.

6 De Coning and Fick (1986) focus on political burnings in their article, and suggest that others explore burnings done for religious, criminal or ethnic reasons. I would disagree with this. If it is the reasons for burnings that are being explored then these categories must be taken as a whole. A song, sung by the Brooklyn Youth Organisation on patrol at night serves to illustrate the point:

Informers, we will destroy you. Haai! Haai!
Witches, we will burn you. Haai! Haai!
Those who commit abortions, you will be destroyed. Haai! Haai!…
Kill our people's enemies. Haai! Haai!
Let our people know comrades. Haai! Haai!
Our motto is: an injury to one is an injury to all.
(Ritchken 1989a:2)

The youth will discipline any person who deviates from the expected norms and threatens the social framework, be that person a witch, an informer or a criminal. It would seem to me that it is the fact that they are all traitors in the broader sense that is more important than the specificities of their crimes. All could be burnt or otherwise killed.

7 A synonymous term for a kangaroo court.

8 These are either hypnotised humans or the resurrected dead. They are used to plough the witches' fields at night and to steal money for them (Niehaus, 1991:5).


Anderson, R L. 1990. Keeping the Myth Alive: Justice, Witches and the Law in the 1986 Sekhukune Killings. BA Honours Dissertation. Department of History. University of the Witwatersrand, Johannesburg.

Balic, S. 1990. Changes Bring Muti Murders. New African. May (272): 32-33.

Bergland, A. 1976. Zulu Thought Patterns and Symbolism. London: C Hurst & Co.

Bergesen, A. 1977. Political Witch Hunts: The Sacred and the Subversive in Cross- National Perspective. American Sociological Review. 42 (April): 220-233.

Brain, J L. 1982. Witchcraft and Development. Journal of the Royal African Society. 81: 371-284.

De Coning, C and Fick, J. 1986. Menslike Verbrandings: 'n Sosio-Politieke Ondersoek na die Aard en Impak van Kataklismiese Geweld in Suid Afrika. Politikon. 31(1): 22- 53.

Douglas, M. (ed.). 1970. Witchcraft, Confessions and Accusations. London: Tavistock.

Ebersohn, W. 1987. Ring of Fire. Leadership SA. 6(1): 39-42.

Fernandez, J. 1965. Symbolic Consensus in a Fang Reformative Cult. American Anthropologist. 67: 903-926.

Franklin, K. 1963. Kewa Ethnolinguistic Concepts of Body Parts. Southwestern Journal of Anthropology. 19(1): 54-63.

Hammond-Tooke, W D. (ed.). 1974. The Bantu-speaking peoples of Southern Africa. London: Routledge & Kegan Paul Ltd.

Hammond-Tooke, W D. 1981. Boundaries and Belief: The Structure of a Sotho Worldview. Johannesburg: Witwatersrand University Press.

Junod, H. 1962. The Life of a South African Tribe. New York: University Books, Inc.

Kane-Berman, J. 1993. Political Violence in South Africa. Johannesburg: South African Institute of Race Relations.

Kertzer, D. 1988. Ritual, Politics and Power. New Haven and London: Yale University Press.

Mair, L. 1973. Witchcraft. London: BAS Printers Ltd.

Marwick M (ed.). 1970. Witchcraft and Sorcery. Great Britain: Penguin.

Minnaar, A de V, Offringa, D & Payze, C. 1992. To Live in Fear, Witchburning and Medicine Murder in Venda. Pretoria: Human Sciences Research Council.

Moore, S & Myerhoff B. 1977. Secular Ritual. Assen: Van Gorcum.

Monnig, H O. 1967. The Pedi. Pretoria: Van Schaik.

Morris, M & Hindson, D. 1992. South Africa: Political Violence, Reform and Reconstruction. In Review of African Political Economy. 43-59.

Motshega, M. 1988. The Ideology behind Witchcraft and the Principle of Fault in Criminal Law. In Law and Justice in South Africa (ed.) J Hund. Institute for Public Interest and Research.

Myburgh, A C. 1974. Law and Justice. In The Bantu-Speaking Peoples of Southern Africa (ed.) W D Hammond-Tooke. London: Routledge & Kegan Paul Ltd.

Niehaus, I. 1991. Witch-Hunting and Political Legitimacy: Continuity and Change in Green Valley, Lebowa (1930-1991). Unpublished. Department of Social Anthropology, University of the Witwatersrand.

Parkin, D. 1992. Ritual as Spatial Direction and Bodily Division. In Understanding Rituals (ed.) D De Coppet. London: Routledge.

Ritchken, E. 1989a. The Meaning of Rural Political Violence. Seminar Paper, Centre for the Study of Violence and Reconciliation.

Ritchken, E. 1989b. Burning the Herbs. Work in Progress. July (49): 17-22.

Ritchken, E & Koch E. 1990. The Political Economy of Witchcraft. Mail & Guardian. 23 March.

Ritchken, E & Koch, E. 1991. A Chief is Always a Witch. Cross Times. Feb/March.

Sosibo, M B. 1992. Healing, Ritual and Violence at Edendale, Pietermaritzburg. Paper presented at the Annual Conference of the Association for Anthropology in Southern Africa. 16-19 September, University of Durban-Westville.

South African Institute of Race Relations, SAIRR. 1985, 1986, 1987, 1988, 1989, 1990, 1991. Survey of Race Relations in South Africa. Johannesburg.

Stadler, J. Age, Tradition and Household. MA Dissertation, Department of Social Anthropology, University of the Witwatersrand.

Steadman, L. 1985. The Killing of Witches. Oceania. 56(1): 106-23.

Thornton, R J. 1990. The Shooting at Uitenhage, South Africa, 1985: the Context and Interpretation of Violence. American Ethnologist. 17(2): 217-236.

Turner, V. 1992. The Anthropology of Performance. Baltimore: The Johns Hopkins University Press.

Copyright © 2019 CSVR. All Rights Reserved.