Jump directly to the Content
Jump directly to the content
Witchcraft, Violence, and Democracy in South Africa
Witchcraft, Violence, and Democracy in South Africa
Adam Ashforth
University of Chicago Press, 2005
376 pp., 37.00

Buy Now

Robert J. Priest

Witches and the Problem of Evil

Looking behind accusations of witchcraft.

Kenyan Pastor Thomas Muthee sensed that nearby car accidents were caused by a witch whom he publically identified as "Mama Jane," apparently the local pastor of an African independent church.[1] Muthee reports that after another car accident death, his followers drove Mama Jane out of town. The wider world learned of Pastor Muthee when Sarah Palin credited his prayers for her political success, prayers which included the request that God would protect Palin from "all forms of witchcraft."

Why do bad things happen? In the Bible, Job's counselors "knew" his afflictions were caused by his own sin, and called for confession. By contrast, the Aguaruna people of Peru, with whom I did fieldwork, "knew" that afflictions were caused by murderous malevolence in the heart of a neighbor or relative, and called for identifying the "witch" and retaliating. And while anthropologists have found scores of societies which operate out of "moral causal ontologies," where the afflicted are blamed for their own affliction, "interpersonal causal ontologies," where affliction or death triggers the search for a guilty third party, a witch, have historically been more frequent worldwide.[2] (While Wiccans and other Neo-Pagans sometimes apply the term "witch" to themselves, this is altogether different from "witch" conceptualized as a person who is blamed for other people's misfortunes and deaths. It is the latter which this article addresses.)

Because witch discourses were once mainstream to Europe and North America, where today they are largely gone, it has been common to assume that fear of witches dissipates under modernization. However, anthropologists have widely noted that resurgent witch discourses are thriving around the world, even in urban settings. Indeed, research suggests that modernity and capitalism provide fertile soil for the rise of witch anxieties, discourses, and retaliatory practices. An excellent introduction to these dynamics is provided in Australian political scientist Adam Ashforth's engagingly written Witchcraft, Violence, and Democracy in South Africa, winner of the African Studies Association's 2006 Herskovits Award for the best scholarly book on Africa.

Ashforth rented a room in the apartheid township of Soweto in 1990, "a time when the page of world history was turning." His goal? To study South Africa's transition to democracy. Adopted by a host family, he spent 36 months over the course of a decade hanging out in homes and bars, helping friends visit healers, and carrying out interviews. He found himself in a world that was "dynamic and joyful," characterized by familial "norms of sharing, reciprocity, mutual aid, and … love." But he also found himself living in a world where people feared witches, and where affliction and death were attributed to the witch action of relatives and neighbors, who were blamed, gossiped about, denied social support, and occasionally killed. When the government refused to prosecute and even defended alleged witches, many South Africans felt themselves to be in a world without justice, to such an extent that the fledgling democracy faced a crisis of legitimacy. Ashforth concluded that "no one can understand life in Africa without understanding witchcraft." Out of this conviction, he wrote Madumo: A Man Bewitched (Univ. of Chicago Press, 2000), an account of a close friend accused by siblings of killing their mother through witchcraft, but who himself came to feel bewitched. This was followed by the more recent, systematic, and theoretically oriented book reviewed here.

Witch anxiety, Ashforth suggests, involves a fusion of two sources of anxiety. First, there is anxiety over misfortune, affliction, and death. In part because of a history where the police were associated with racial oppression rather than justice, Soweto has ongoing problems of law and order. Alcohol abuse is high. Rates of homicide, rape, property crimes, and fatalities due to traffic accidents are among the highest in the world. The prevalence of HIV/AIDS contributes to the high rate of premature death; Ashforth cites projections of South African life expectancy dropping from fifty-five in 2000 to forty by 2010. Less than half of adults find regular paying jobs. Quite apart from ideas about witches, there are many straightforward reasons why residents of Soweto would experience deep anxiety.

Every year thousands of women and children (and sometimes men) are killed as witches around the world.

Ashforth argues that an underlying assumption about social relationships contributes to a second source of anxiety. The African principle of ubuntu, "a person is a person through other persons," articulates ideals of interpersonal solidarity, love, reciprocity, and communal sharing, which are said to provide the essential underpinnings of the good life. But everyone falls short of such ideals, often feeling resentment or enmity against each other rather than love. The negative corollary of ubuntu, Ashforth suggests, is a pervasive presumption that human malice is what underpins and explains misfortune. In a world where one's well-being depends on others, the awareness that others harbor negative sentiments triggers deep anxieties.

Two sorts of individual are feared. First there are neighbors or relatives who have less than you—in short, those likely to envy and resent you. Repeatedly Ashforth's informants told him that those who are envious or jealous desire the death of the more fortunate. Under ubuntu the "haves" should share with the "have-nots." This means that those who are envied are also those who are asked for help. Those who have resources in the new capitalist order find themselves refusing to share as much as others feel they should. Such refusal is humiliating to persons denied, and triggers guilt on the part of the "haves," who doubly fear those they've spurned. Alternatively, those with resources who refuse to share are felt to be the reason for others' misfortune. They are said to exhibit witch-like selfishness and pride, with witchcraft doubtless explaining their unfair wealth advantage. Those who are suspected of being witches, then, are either less or more fortunate than the ones they are suspected of bewitching.

Hence two sorts of anxiety—anxiety over affliction and anxiety over human malice—coalesce into a single anxiety: anxiety over witches, a single diagnostic system that brings cognitive order to the mystery of evil but that also, once embraced, comes to generate new levels of anxiety.

While witch discourses are not new, recent trends contribute to an intensification of witch anxieties. In Soweto during apartheid, misfortune was relatively evenly shared by blacks, and racial oppression was a compelling explanation for this misfortune. After apartheid, social differentiation for blacks increased and in ways that subverted traditional age and gender hierarchies, with some blacks becoming wealthy and others even poorer. The mystery of new patterns of capitalist wealth and poverty that divided relatives and neighbors as never before required an alternative explanation to that of racial oppression. In apartheid Soweto with a more even distribution of goods and needs, an ethic of reciprocity had served as a safety net to benefit all. Under new patterns of social differentiation with more permanent "haves" and "have-nots," older patterns of reciprocity shifted into one-directional patterns of giving, with resentment for being asked to help increasing among the "haves," and envy/resentment dynamics over the new hierarchies increasing among those with less. And HIV/AIDS introduced escalating levels of mysterious affliction and death. Such changes meant that both the evil of affliction and the evil of human malice were felt to be increasing, and thus witchcraft itself was felt to be increasing.

Ashforth does not deny that some people may engage in thoughts or secret actions which they believe will magically harm others, but stresses that witch accusations seldom depend on actually observing witch action. Since the mechanisms by which witches are said to work, whether through psychic power or through manipulating occult substances (herbs, animal fats, graveyard dirt), are believed to be secret and even invisible, affliction and malice are the primary evidence for witch activity available for examination. Gossip becomes the medium through which "witches" are identified. When someone dies, gossip works to fix blame on a single person as witch. The deceased may have been surrounded by relatives and neighbors who envied or resented her in life, who longed for her death so they could inherit, or who harbored long-term grudges over a perceived insult or slight. In short, the deceased may have been surrounded by people with "witch-like" sentiments toward her. But with her death everyone moves into a mode of gossip designed to deny or conceal their own asocial sentiments and to identify a solitary individual who all can agree is the single repository of evil. That is, people often accuse others of the very witch-like sentiments which they themselves harbor.

Half a million traditional healers (inyangas and sangomas) in South Africa make their living by divining and combating witchcraft—that is, by ratifying the "witch" diagnosis, and hinting at a relative or neighbor who should be blamed for the affliction. Their supposed knowledge is employed in service of affirming an interpersonal causal ontology—affirming that afflictions are best understood as caused by an evil relative or neighbor. The cognitive authority of these "healers" is ratified by widespread rhetoric affirming that "European science" (which gives marvels of technology and Western medicine) is paralleled by "African science" (which these healers are supposed to represent). According to Ashforth, efforts by UNICEF and the World Health Organization to affirm an important role for "traditional" healers help accredit the cognitive authority of healers whose central diagnostic category is that of witch. These healers play their part in constructing a diagnosis which is destructive of interpersonal trust and well-being for society, and which has devastating consequences for those identified as guilty of causing the misfortune of others. (After Ashforth's book was published, the South African government passed the Traditional Health Practitioners Act No. 22, 2007, which requires inyangas and sangomas to register with the government, but which also implicitly accredits their health care practices. Since witches are thought to operate using the same understandings of muthi (medicine/poison) as these healers, Ashforth sees such validations of inyangas and sangomas as contributing to the validation of the belief in the genuine power of witches.)

While some anthropologists argue that witch discourses are functional (discouraging the display of asocial impulses), and others treat witch discourses as symbolic and metaphoric criticisms of global free-market capitalism, Ashforth insists these discourses are literal assertions about the culpability of neighbor or relative for the affliction and death of others. And while he works to respectfully grasp the logic involved, he ultimately concludes that witch discourses construct a world which fosters social mistrust, deep anxieties, and harmful actions. The accused (often elderly widows) are gossiped about, stigmatized, denied social support, and, in the extreme, lynched (hacked to death with machetes or burned with the "necklace" of a car tire doused with petrol).

In Ashforth's account, South African Christianity reflects a quest for "spiritual security in a world of witches." Yet traditional Western missionaries, like African colonial governments, were hesitant to endorse witch beliefs. As a result mainline/mission churches are perceived as failing to understand witch realities, with church members seeking help outside the church. The leaders of Pentecostal and African Independent Churches, by contrast, directly recruit their members by appealing to their own success at divining, healing, and counteracting the work of witches. While Christian faith provides a feeling of security against witches, many Christians are as likely as others to suspect their relative or neighbor of being a witch, and of attributing misfortune to them. Ashforth seems ambivalent on whether such churches produce positive or negative net outcomes.

At this point, I leave Ashforth's book. While Ashforth limits his focus to Soweto, every year thousands of women and children (and sometimes men) are killed as witches around the world—from Java and New Guinea to Peru and Cameroon.[3] For each person killed as a result of such accusations, many more are labeled "witch" and denied all social support. In Tanzania, hundreds of elderly widows are killed each year as "witches," with many more expelled from their homes penniless. In Ghana, thousands of elderly women barely survive in refugee "witch villages" where they have fled for their lives. In Kinshasa, the sprawling capital of the Congo, tens of thousands of homeless children struggle to eke out a living on the streets, a majority of whom, experts report, were expelled from their families as witches.

Missionaries who ignore or express skepticism about witches but without any clearly adequate view of what is going on simply foster patterns of split-level Christianity, where believers feel the church is unable to understand and help them when affliction comes, and who go outside the church to traditional healers for help. Alternatively some missionaries and many pastors adopt approaches which endorse the diagnosis that afflictions are caused by third-party witches, while modifying this in ways stressing that the power is demonic (rather than psychic or magical), and presenting the accused with the option of confessing to bewitching others, being exorcised of the demon of witchcraft, and joining their church, on the one hand, or of having God punish them for being witches while allowing normal mechanisms of social exclusion to be carried out, on the other. Since such pastoral responses fail to address the social dynamics fostering suspicion toward the accused, even the accused who submit to this process seldom escape the suspicion that they are (still) witches. And those who suspect them of being witches treat them accordingly. There is reason to believe that pastors who endorse the witch diagnosis are helping to accredit beliefs with harmful social consequences for vulnerable children, women, and men.

Biblical accounts of disease, infertility, and death never attribute misfortune to human third parties acting by evil supernatural means. And while the word "witch" unfortunately appears in many English Bibles in Exodus 22:18 and 1 Samuel 28, these passages almost certainly were speaking of recognized magico-religious practitioners whom others consulted,[4] not unlike South African sangomas, and provide no charter to treat misfortune as the starting point for suspecting and trying to ferret out some third party who should be blamed for secretly and supernaturally inducing affliction or death. These passages do not appear within discursive formations ratifying interpersonal causal ontologies. That is, English usage of "witch" in these passages not only involves mistranslation but opens the door for others to misread such texts as accrediting ideas alien to the text—horrendously consequential ideas. There are, of course, biblical texts linking affliction with Satanic or demonic agency, but other persons are never blamed for such affliction. Efforts to graft interpersonal causal ontologies onto biblical demonologies introduce an alien logic into the biblical text.

Read one way, witch discourses are discourses on human sinfulness. They feature envy, resentment, animosity, pride, greed, and selfishness. And under interpersonal causal ontologies, each affliction triggers the quest for a guilty third party whose identity as witch is disclosed by evidence of having such sinful sentiments. While everyone may have less than exemplary love for the afflicted or deceased, any feeling other than love must be denied or concealed lest one be named a witch. Those accused of being witches are often those who (not unlike Joseph) wear fancy clothes and tell stories setting themselves above others, or they are those who (not unlike King Saul) bitterly obsess over young upstarts who surpass them in the praises of women, or they may be those who (like Rachel) bitterly resent a more fertile woman. That is, witch discourses, at least some of the time, doubtless are accurately recognizing sinful impulses on the part of those accused. But while the Bible attributes sinful realities to all people, witch discourses deny sin in anyone except the witch. To acknowledge such sins, under an interpersonal causal ontology, is to imply that one is a witch. And yet as we have seen, anthropological research makes clear that witch accusers often exemplify toward the afflicted the very asocial sentiments they attribute to the accused.

Furthermore, anthropologists have documented evidence that accusations typically express sinful and asocial sentiments towards the accused. This may be as direct as the brothers of a deceased AIDS victim blaming his widow for the death. By naming her a witch, they justify themselves in expropriating all marital assets. Or it may be subtle. A destitute widow constantly asks relatives and neighbors for food. In their struggle to feed their immediate families they resent her requests, and provide food only sparingly, if at all, accompanied by humiliating comments designed to discourage her from asking again. Desperate and hungry, she will naturally resent those who spurn and humiliate her. But she is not the only one with hostile feelings. Because she is a constant drain on others, and because each time people fend her off they inescapably feel guilty, she is deeply resented. Unacknowledged guilt toward her gets converted into resentment and anger at her, as people seek a justification for their hostility. The label "witch" declares her to be evil incarnate, for whom normal empathy need not apply. When a neighbor dies, people read the widow's envious resentment as evidence that she is the witch. Everyone claims to have exemplary love for the deceased, breathes a sigh of relief that someone else can be fingered as witch, and enthusiastically endorses an accusation which simultaneously absolves them of sinful feelings for the deceased victim and absolves them of guilt related to the widow's care, while also removing the old woman from the moral community of those for whom care is prescribed.

Sin is everywhere. Communities that are riven with interpersonal animosities (despite discourses of ubuntu or agape) construct themselves as virtuously united against a single evil. Just as Herod and Pilate reconciled with each other in the very act of aggressive violence toward Jesus, so sinners find the scapegoating of another to be a unifying experience. And just as Jesus' supporters fled or denied him when the charges coalesced against him, so all support for the accused vanishes when witch accusations snowball toward action. When "witches" are killed or excommunicated, communities feel cleansed, purified, euphoric. Evil is gone.

Rather than endorsing witch discourses featuring a world of virtuous people being attacked by others who represent evil incarnate, the Bible presents all as sinners, with flawed understandings of evil, and whose discord- and violence-producing accusations against others reflect the attributes of Satan, the great accuser, whose own deceptions lie at the root of discord and murderous violence. The biblical message calls for a recognition that everyone has witch-like sentiments, and requires confession and repentance for these, appropriating forgiveness and strength for personal transformation based on Christ's saving work. Christ, the truly innocent scapegoat victim, willingly bore our sins on the cross, providing the foundation undercutting all other scapegoating. And opposed to Satan, the accuser, we find the Holy Spirit, the paraclete, defender of the accused.

In practice, of course, Christians have struggled in knowing how to address witch discourses. Folk ideas about interpersonal causal ontologies are sometimes linked in unhelpful ways to biblical ideas, creating more toxic patterns than before. In Salem, Massachusetts, scores of individuals were identified as evil witches by equally sinful accusers. The only people with the cognitive authority to question the process were pastors. But they endorsed the witch ontologies, and 20 individuals were killed. The violence stopped only when New England theologian-pastors recognized problems with what they had endorsed.

Throughout Africa, westerners often contribute to conditions fostering witch sentiments and anxieties; and floods of altruistic Western Christians crisscross the continent teaching about HIV/AIDS, most with little understanding of how witch ontologies shape experiences of affliction. But while African pastors are positioned to address with authority realities related to witchcraft, like their pastoral predecessors in New England, they sometimes endorse the very ideas which produce evil—or they struggle with doubt over how to respond. Christian theologians, church historians, and anthropologists, both from Africa and elsewhere, need to prioritize a sustained multidisciplinary conversation aimed at forging understandings which are biblical, culturally informed, and pastorally redemptive. Although not the only ones participating, African pastors and theologians must be central to the conversation. Only they will have the cognitive authority to lead African Christians forward in appropriately addressing these realities.

Robert J. Priest is an anthropologist and serves as professor of Mission and Intercultural Studies and director of the PhD program in Intercultural Studies at Trinity Evangelical Divinity School.

1. According to the investigative work of journalist Zoe Alsop.

2. See Richard Shweder's anthropological review of causal ontologies in Why Do Men Barbecue? (Harvard Univ. Press, 2003).

3. See Jill Schnoebelen, "Witchcraft Allegations, Refugee Protection and Human Rights: A review of the Evidence," Research Paper No. 169, The UN Refugee Agency and Policy Development and Evaluation Service (2009).

4. Although the exact meaning of the Hebrew term in Exodus 22:18 and Deuteronomy 18:10 is unclear, it also appears in Exodus 7:11 and Daniel 2:2, where the term clearly refers to publically recognized magico-religious practitioners of some sort.

Most ReadMost Shared