Ghana's Concentration Camps For Witches

Academics, NGOs and the struggle to explain why Ghana has had witch camps for over a century.

From the years 2005 to 2011, over 3,000 people were put to death in Tanzania for the crime of witchcraft. Before that, around 50,000 to 60,000 more were killed between 1960 and 2000 for the same reasons. The majority of these victims were elderly women, who fit the Tanzanian witch stereotype.

However, accusations have been made on all grounds, such as if a woman has red-tinted eyes. Others were accused of simply living in poverty. Or if the village had a poor harvest. Or blamed for uncontrollable diseases such as HIV.

No matter the justification, these women were shown no mercy in their deaths: beaten, chased, stoned and in more dramatic cases, burned or buried alive. In most instances, the murderers did not face punishment as law enforcement is stretched thin over the region. 

This quote, taken from an LA Times piece about the modern witch hunts in Tanzania, highlights the scale of social violence across many African countries related to witchcraft. Tanzania is not alone of course in carrying this burden - in 2009 the Gambian president Yahya Jammeh had 1,000 people kidnapped on suspicion of witchcraft, and forced them to drink hallucinogenic potions in order to confess; witch-lynchings in Kenya are common enough that officials have turned to sanctioned exorcisms to prevent murder, and sporadic outbreaks of unknown diseases can cause communities to turn on one another, as happened in the Congo in 2001, where around 400 people were killed in a spasm of witchcraft-related violence.

Many of these incidents are the result of intra and inter community tensions, where the authorities attempt to mediate and pacify to prevent rising violence. But sometimes the state itself can look to purge a country of witches, often resorting to magical methods itself to track them down and interrogate suspects:

Through provincial and district conferences, documented in print and radio media, state actors drew on local understanding of witchcraft to construct a standardized witch other' (Ciekawy 1 998) - a deadly figure capable of inflicting magical harm on peasants or state agents in the name of 'feudal lords'. Unlike antisuperstition campaigns launched elsewhere (e.g., in The People's Republic of China), the rhetoric of the Beninese campaigns never construed witch others as the delusions of ignorant peasants (cf. Feuchtwang and Wang 2001); the deadly practices of witch others were regarded as actual challenges to state sovereignty. As such, the goal of liberating the masses did not hinge on ideological reeducation and infrastructural change alone; it depended on a battle between the progressive forces of the state and an army of evil witches, sorcerers, and vodun priests (see d' Almeida 1976).

- Kahn, J., 2011. Policing ‘Evil’: state-sponsored witch-hunting in the People’s Republic of Bénin

Benin is a fascinating example here, a country which underwent a Marxist-Leninist coup between 1972 and 1991, led by the militant revolutionary Mathieu Kérékou. Benin’s modernisation project placed witchcraft front and centre, but as the above quote reveals, the regime took witches seriously on their own terms. In doing so Kérékou and his forces courted and made use of other spiritual and religious figures. This led to the bizarre position that an avowedly atheistic revolution deployed a private army of Fa diviners, Sakpata and Xebyoso priests and azeglovodun anti-witchcraft voodoo cults against non-sanctioned witches.

In 1979, the Beninese government continued to simultaneously uphold and blur the distinction between legitimate state authority and illegitimate witchcraft by awarding three seats to 4 animists in the Assemblee Nationale Revolutionnaire under the condition that each cult- group has to consider itself as a revolutionary cell and that no sorcerers will be accepted in the priesthoods ranks' (Sulikowski 1993, 386). Throughout the 1980s, President Kerekou called on the vodun priest Daagbo Xuno Xuna to conduct state-sponsored rain-inducing ceremonies in service to the nation. By 1989, on the eve of his fall from power (see Heilbrunn 1993, 286), it was widely believed that Kerekou depended on Malian diviners as mystical protectors and political advisors (Sulikowski 1993, 387). In addition, rumors circulated that several of Kerekou's top ministers, including former Minister of the Interior and reported architect of the antiwitchcraft campaign, Martin Azonhiho, were now fully initiated witches (Sulikowski 1993, 387)

Amongst the many strange policies of the Benin project was an attempt to corral and confine suspected witches to special camps or prisons, usually called azexwe - a house of witchcraft. These were largely informal and poorly provisioned places, where the suspects were left to fend for themselves. Strikingly the overwhelming majority of the inmates were older and childless women, a theme we’ll return to with the case of Ghana. Of course the logic of the internment was somewhat undermined by the widespread belief that witches could transform themselves into animals and fly around at night, unimpeded by walls or doors, a contradiction which led to guards and officials seeking magical protection against their detainees.

Witchcraft in Ghana

Ghana is considered to be a modest success story, relative to many of its neighbours. Looking at measurements of democratic institutions, healthcare, poverty and economic growth, Ghana has performed well since its independence in 1957. However, belief in witchcraft is highly prevalent across the country, manifesting in a number of ways. One such is the well documented phenomenon of the ‘spirit child’, called chichuru or kinkiriko. These are believed to be malevolent spirits who inhabit the body of a newborn child, often manifesting in disabilities or deformities. Infanticide in these cases is very common, but difficult to study. One analysis recorded that 15% of all infant deaths under 3 months old was due to chichuru infanticide, typically using a concoction of lethal herbs and/or exposure to the elements. Such is the strength of belief in chichuru that health professionals will fail to report suspected infanticide cases, even in hospitals:

The day after the birth, I went to the hospital to see how mother and infant were progressing. On the ward, the midwives explained that Jampana was fine. However, her daughter had died during the night. The midwives explained further that Jampana's mother, Lamisi, had been with the child during the night. Lamisi reported to the night duty midwife that she had been feeding the infant a herbal preparation when the infant started choking. She had pinched the infant's nostrils to force her to swallow in order to start breathing again, but she had subsequently died. Lamisi wrapped up the dead infant and laid her in a cot, before calling for help. When the midwife examined the infant, she was dead.

In spite of support from a medical officer, Jampana insisted that the incident not be followed up with the staff on duty. Most of the nursing staff were local and were very reluctant to challenge traditional beliefs and practices, some of which they themselves held. Instead it was reported to family and friends that the child was stillborn.

In 1930 the British colonial government outlawed judicial witch-finding, following the previous abolition of the death penalty for sentenced witches. In place of executions came a focus on confession and voluntary examination, particularly by oracles. But even these were far from ideal:

The controversy began in 1926 when J. E. Gresham Williams, a British employee of the Akim Limited mining company, witnessed Nkona priests trying to rid a young mother of witchcraft. Williams reported that he saw the Nkona priest, Kwasi Adjai: "Dance up to the woman and drag her hair with both hands shaking her head from side to side, pulling her cloth down, stripping her naked to waist in the pouring rain. Placing his muddy foot on her head, he would order her to confess...."

Williams, a World War I combat veteran, was painfully sensitive to violence. Over the course of four rainy days, he became increasingly convinced that what he was observing was the "most gruesome torturing of a defenseless native woman.” The woman, Akosua Darebuo, was a new mother and a widow. Darebuo had been living in a nearby town earning her living as petty trader and caring for her baby when her husband, Kwaku Asante, came down with smallpox in the epidemic of 1926. A priest determined through the Nkona oracle that Darebuo was causing her husband's illness through witchcraft

- Gray, N., 2001. Witches, oracles, and colonial law: evolving anti-witchcraft practices in Ghana, 1927-1932

The shift towards social gossip, rumour, exorcism, therapeutic interventions and confession is a legacy of colonial rule, although lethal violence does still occur from time to time. A meningitis outbreak in 1997 killed nearly 550 people across northern Ghana and led to vigilante attacks on older women. Several were lynched and others beaten and stoned to death. Outraged NGOs and journalists began investigating the conditions for women, particularly in northern Ghana, and discovered a shocking reality. Not only were women being killed as witches in higher numbers than expected, but many thousands of women were fleeing to makeshift camps - witch’s camps. Pieces appeared in the international cosmopolitan press, decrying the conditions, particularly at the largest such camp - Gambaga - in the north east region of the country.

Between the 60’s and the 90’s there was a feeling amongst scholars that African witchcraft would simply ‘go away’, under the combined pressures of modernisation and globalisation. But this didn’t happen. Researchers such as Jean and John Comaroff and Peter Geschiere began pointing out that witchcraft and the fears surrounding it intensified and morphed as different African nations began to develop, both politically and economically. Fears of zombies, evil factories, possessed politicians, ambulances roaming at night stealing blood, Satanic murders, ritualistic killings for company profit, the international trade in body parts, digital curses and hexes, penis-snatching, killer mobile phones and more proliferated under what scholars called ‘occult economies’. Success and failure in this new world took on the extra dimension of witchcraft, and whoever might be jealous enough to hold you back could be using magical powers against you.

This ‘African witchcraft in modernity’ paradigm hasn’t really added any value to the scholarship in my opinion. The ghost in the machine for the first wave of researchers was ‘neoliberalism’, which at this point is like the term ‘ritual’ in archaeology, it serves to fill in the blanks when things don’t make sense. Papers trying to explain modern witchcraft as a reaction to the anonymising forces of urbanisation and modernity clash with other studies showing that witchcraft accusations start at home, between intimate family members. Often a young person feeling thwarted in life turns on his mother or grandmother, accusing them secretly of malicious intent and action. Defining Ghanaian witchcraft has been a preoccupation of anthropologists since British rule. The general term bayie is translated as witchcraft, but it is a slippery term, shifting between a person, a spirit, a physical substance and its transformations. Bayie can mean the voracious appetite for human flesh, or the bewitching spells placed on another person. Sometimes witchcraft can be good and helpful, and the position of authority and wisdom that elders possess can oscillate between benevolent and evil.

Bayie, hideous but hidden, became the most trenchant symptom of the contradictory feelings which close relatives may have for one another. Imputations were secretly levelled at people one could not openly criticise, let alone accuse. I agreed with Kennedy (1967: 273): 'Witchcraft is primarily a manifestation of strongly held negative emotions. Any student describing it inevitably finds himself involved with materials which have been the province of psychoanalysis-hatred, fear, anger, jealousy and frustration.' But a psychological interpretation of bayie also falls short of the 'facts'. Several of those who accused someone of bayie denied any ill feeling towards that person. They stressed that the 'witch' was not guilty of his/her deeds since he/she was not even conscious of them. Accusing often went hand in hand with excusing. Bayie remained elusive, typically a concept not to be caught by one-dimensional reasoning, let alone a dictionary

The most remarkable public appearance of bayie, however, is on public transport: lorries, mini-buses and taxis. Again, it was Field (1960: 134-45) who drew attention to the meaning of the popular inscriptions which drivers painted on their cars. In them, she wrote, they express their worries about the future, their anxiety that misfortune may strike. Their security lies in the hands of other people, and it is witches in particular who pose a threat. Some inscriptions beseech God for protection, others are aggressive and refer directly to the potential evildoer: Obi mpc obi yiye, 'Someone does not like someone's success', Sura nea obcn wo, 'Fear the one who is close to you', Abusua d:J funu, 'The abusua loves a corpse', and tan firi fie, 'Hatred come from the house'

- Van der Geest, S., 2002. From wisdom to witchcraft: Ambivalence towards old age in rural Ghana

It seems obvious that swapping out the words ‘success’ and ‘prestige’ for ‘neoliberalism’ hasn’t added anything productive to the conversation. The frustration for modern NGOs and liberal-minded development workers and academics is that witchcraft defies simple explanations, and attempts to intervene can simply make things worse. Poverty relief programs or targeted charity towards accused and banished witches breeds the very resentment which fuels further accusations. NGOs working in the witch camps are seen as benefiting from the arrangement:

The popular image of total banishment and isolation turns out to be more complex in practice. Young children sent to care for the accused also benefit from funds given by the GoHome project to pay for school fees. The local hospital no longer charges for visits. The banished are then ironically resented for charity received. As one informant cynically commented, “The NGOs are happy they (the women) are there” because they are so popular as a source of donor funds

- Crampton, A., 2013. No peace in the house: witchcraft accusations as an" old woman's problem" in Ghana

The Witch Camps of Northern Ghana

As of today Ghana has around six functioning witch camps, although others have opened and closed. Three of them - Gushegu, Nabuli and Kpatinga - are located in the Gushegu district. The infamous Gambaga witch camp is in the East Mamprusi district, and the Gnani and Kukuo camps are located in Yendi and Nanumba South districts. The exact number of residents is unclear, several thousand is a rough estimate. This is in part because these camps are not like a standard prison compound, they are more like an open market/village, demarcated but porous. People come and go and many NGOs and charities are semi-permanently settled. There is no doubt that the majority of the inhabitants are older women, with some younger women and children amongst them. Many anthropological assessments have now been written about the camps, drawing the traditional candidates looking for studies on female autonomy, mutual solidarity and gender subversion. Food is scarcer, life is harder than at home and women typically arrive with little to no possessions. Many combine agricultural work with cottage industries and firewood collection to make a living. Some camps are ruled over by local chieftains, who exchange the women’s labour for protection.

Why are they there? Ultimately the most common reason is that life is better at the camps than at home. Petty accusations, harassment, violence, gossip and local power politics often sees widows targeted for their home and property, at least, that is the view of the aid workers and journalists. The claim that Ghana’s northern ethnic groups are more patriarchal, as compared to the apparently relaxed matrilineal Akan peoples of the south, has become a staple of articles and experts trying to explain the camp phenomena:

Ghanaian sociologist and criminologist Mensah Adinkrah observes that Ghana is very diverse in terms of ethnic composition. In this regard, a woman’s political and social status and recognition depends on the ethnic group to which she belongs. He argues that people in northern Ghana who are mostly Mole‑Dagbani, have patriarchal structures that tend to undermine the social status and power of women as against the matrilineal Akan‑speaking ethnic groups in southern Ghana who are more powerful than their counterparts from the north

- Mutaru, S., 2018. An anthropological study of “witch camps” and human rights in northern Ghana

This idea that the Akan are somehow more immune to violence against women is easily countered by looking at Akan notions of witchcraft, which seem just as dangerous and misogynistic as any northern tribe.

In Akan society, as in many others (Niehaus, 1993), accusations of practicing witchcraft are based on mere suspicion, rumor, or gossip (Bannerman-Richter, 1982). Often, after a witchcraft accusation is leveled, the suspected witch is threatened, drugged, beaten, forced to submit to humiliating ordeals, or is coerced into confessing to imaginary witch activities (Assani, 1996; Parish, 2000; Simmons, 2000). Many accused witches vehemently deny allegations of witchcraft (Drucker-Brown, 1993). Those who “confess” or agree with their accusers often do so simply to avert physical assaults and retributions that may follow denials (Gray, 2000; Niehaus, 1993; Simmons, 2000; Ware, 2001). In some instances, accused witches are banished from their families and communities with threats of violent, retributive reprisal should they return.

- Adinkrah, M., 2004. Witchcraft accusations and female homicide victimization in contemporary Ghana

Neither neoliberal modernity nor simple anti-female sentiment easily explains Ghana’s camps, which are virtually unique in Africa. We must instead look for another motivation or reason as to why they exist. A good place to start is their age. While a number have formed in the last few decades, many go back over a century, before national independence. So whatever this is, it predates the modernity argument.

Ghana’s history is long and complex, too long for this article. But of crucial importance to the history of witchcraft are the tensions between the different northern kingdoms and ethnic groups. Sometime in the 15th century, the Dagomba people moved into northern Ghana and established the Kingdom of Dagbon, displacing the prior Konkomba peoples of the area. The results of this were catastrophic for the Konkomba, who were not only captured and sold as slaves across the Sahara initially, but were then doubly subjugated as the Ashanti conquered Dagbon and demanded slaves were sent south towards the Atlantic coast. This continued from 1744 until 1874, when the British won the Third Anglo-Ashanti War and abolished the slave trade.

The Voltaic peoples, which include the Konkomba, are known for their spiritual attachment to the earth. Not only is the earth a deity, the prime mover of fertility and life, it can also be defiled and polluted by impure activities, which includes the spilling of blood. This is one reason why pre-colonial witch executions often involved strangulation or poisoning, to avoid bloodshed. Amongst the religious structures of the Konkomba is the office of ‘earth-priest’ - ten’daana - whose duties include protecting and conducting rituals at particular ‘earth shrines’. The Dagomba invasion destroyed many of these shrines and resulted in an assimilation of the priestly class into Dagbon society. For the most part these shrines and their priests were allowed to continue operating, but the ten’daana lost their political power to their new overlords. As Ghana became a closer union under the Ashanti, British and then as a new nation-state, there was considerable diffusion of Akan, Voltaic and Dogomba religious belief between the different regions and kingdoms, as well as the presence of Islam and denominations of Christianity.

Over time there was a movement to stop executing witches, as previously discussed, and banishment became more commonplace. Both earth-shrines and chiefly residences became sanctuaries for women and men fleeing persecution. Inhabitants of the camps even today refer to themselves as bagbenye, a ‘witch-slave’, harkening back to when a witch and her children would throw themselves on a chieftain’s mercy.

As far as ethnohistory is concerned, we only have limited sources to draw conclusions from. In 1918 Arthur W. Cardinall, a colonial official cum anthropologist reported about the shrine at Gnani:

Witchcraft is very much feared. There are at Gnani and near Wapuli special villages where the fetish of the place has the power to prevent “child- or man-eating” people – invariably women, in these instances – from continuing to do so. The Gnani village is very large and is divided into three sections, Konkomba, Dagomba and Nanumba. The women seem quite happy and are looked after by their men-folk, who visit them and make farms for them and keep their huts in repair. The fetish is a stone under a big tree. Trees seem generally to have evil spirits, and many baobabs are cut down to drive them away (Cardinall 1918: 61).15

In another source from 1969 Cardinall writes: In Dagomba, Mamprussi and Nanumba and to some extent in Gonja, separate villages are set apart for the use of witches. In Gonja confessed or “convicted” witches become the slaves of the sub-divisional chief (Cardinall, cited in Parker 2006: 353).

- Riedel, F., 2018. The sanctuaries for witch-hunt victims in Northern Ghana.

Complicating the connections between witch camps and shrines is the fallout from the 1994-5 war between the Konkomba and Dogomba-Nanumba-Gonja. Sometimes called the ‘Guinea Fowl War’, this conflict seems to have been forgotten, not even appearing on Wikipedia’s History of Ghana page. Several thousand people died and tens of thousands fled their homes as over 400 villages were destroyed. The root cause was the perceived inequality between the Konkomba and their historical overlords, who still demanded tributes, food and labour from the Konkomba. Many witch camps, in particular Kpatinga, became sanctuaries for women and children fleeing violence, adding another complicated layer to the gender-witch-mistrust mixture.

The camps today seem to be under the aegis of different chiefs and priests, along with NGOs, churches, missionaries and government agencies. The ten’daana still practice different annual rituals, including sacrificing a chicken and giving the women herbal potions to cleanse them of any magical ailments. Attempts to send women home have produced mixed results, with some projects claiming great successes and others watching impotently as the women are chased back to the camps. At least one report cites the underlying ethnic composition as the prime factor in whether reintegration is possible - that where the chief and priest are the same person, as in western Dagbon, there are no Konkombas, and trust is more easily established. But where the chief and priest are different people, and the priest is a Konkomba, such as the Ngani camp, then women who are sent home will most likely be killed.

The Future of the Camps

Mirroring the mood in scholarship regarding witchcraft in Africa more generally, the future of the witch camps seems stuck and unsure. On the one hand the camps are an embarrassment for the urban elite of Ghana and the NGOs who hold the country up as an example of African success, on the other hand closing the camps down seems both impossible and dangerous - vulnerable women and children will lose their protection and livelihoods. No amount of government legislation can force a village to accept an accused witch back in their midst, nobody can abolish the belief in witchcraft overnight, or likely ever.

In 2014 a camp at Bonyasi was shut down by the government, who quickly made it a priority to close all the camps as soon as possible. Whilst lauded by the UN and international media, the complex reality soon hit home. Almost immediately women from the Bonyasi camp relocated to the Gnani camp. The Anti-Witchcraft Allegation Campaign Coalition wrote to the Minister for Gender, Children and Social Protection, making the case that the camps were a pragmatic necessity, and that re-education and poverty initiatives were essential. Similarly, a 2020 opinion piece on GhanaWeb made the case that, until all the underlying belief, economic and social systems which produced witch camps were undermined, the camps should remain open to protect people from harm. Note though that the focus was again on patriarchy and poverty, with no mention of ethnicity or land ownership, two crucial factors in explaining why the camps exist in the first place.

Ultimately the interventions to tackle witchcraft accusations and violence will almost always fall short, as a 2020 paper by Johanneke Kroesbergen-Kamps notes:

Combating witchcraft violence is not high on the agenda of many African governments (Secker 2012). Even if laws are in place, they may not be enforced, and mob justice even involves traditional and community leaders (Eboiyehi 2017: 260). Community groups and civil organizations are called upon to support the fight against witchcraft violence (Federici 2010; Secker 2012). However, this solution proves to be complicated as well, since civil society in many African countries consists mainly of foreign institutions. On a local level, where the witchcraft accusations are made, communities place their trust in traditional and religious leaders, who share the community’s fear of witchcraft, rather than in the more skeptical foreign institutions (Kleibl and Munck 2017). The human rights discourse that civil organizations use has also come under close scrutiny in different African countries, where the feeling is that human rights constitute an imposition of specifically Western values (Secker 2012: 33). Finally, accusations of witchcraft often take place in a context of conflict and contest within families. Social action and programs aimed at teaching people about their rights do not solve the problems within families that lie behind witchcraft accusations (Crampton 2013). Studies that take the human rights approach are not unproblematic. Even in the quantitative scholarship discussed at the beginning of this section, witchcraft is always labeled as a superstition. Despite repeated urges to understand local ways of viewing the world better, there seem to be few attempts to see witchcraft as anything else than an evil, backward illusion

This position is a square peg in a round hole. Western NGOs do not want to be accused of being patronising, colonial or condescending, but neither do they want to give witchcraft any legitimacy. This is probably why Christian denominations which accept witchcraft as real and can offer some protection against it flourish in Africa, particularly the Pentecostal Churches. Approaches which demand a redress in gender relations often seem to miss the mark as well. As missionary Jon Kirby notes in his 2015 essay Toward a Christian Response to Witchcraft in Northern Ghana:

Unlike in the West, in Africa there is nothing more sobering than the threat of witchcraft. The media portray accused women as victims, but few Africans believe this—often not even the accused. As Fr. Joseph points out, in the African mind they are outcasts, and helping an accused witch is itself antisocial witchery. How, then, might one approach a Christian ministry when even the most basic care risks being so grossly misinterpreted?

A new anti-witchcraft bill has been proposed in the Ghanaian Parliament, and the language surrounding the debate sounds very familiar to anyone versed in the concept of ‘The Blob’. If rural villagers accusing one another of transforming into cats or invading their dreams is the reality of witchcraft, the NGO-stakeholder approach looks like another reality entirely:

The discussion forms part of the on-going “Engaging Media and Minorities to Act for Peace building (EMMAP)” project, a two-year intervention that is running from March 2022 to February 2024.

As part of the project, 10 Journalists from Ghana, Sierra Leone and Liberia participating in the online course were selected to undertake the five-day face-to-face training and field visit to some selected Ghanaian communities that host minority groups.

The purpose of the EMMAP programme, which is funded by the European Union (EU), is to raise public awareness of the inter-connections between conflict, migration, and minority exclusion to help build and consolidate sustainable peace in Ghana, Senegal, and Sierra Leone.

The EMMAP is being coordinated by Uganda-based Minority Rights Group International (MRGA) and implemented by the Ghana-based Media Platform on Environment and Climate Change (MPEC) and Media Reform Coordination Group (MRCG) of Sierra Leone, NGOs.

Prof Alhassan reminded the media that, “journalists have the power to frame and therefore let’s use that power to re-frame and shape our narratives.” 

Speaking on the topic, “Conflict, Migration and Minority Rights: Media Perspective on Community Livelihood”, Prof Alhassan insisted he was strongly against advocacy against the disbandment of the Gambaga and other witch camps in the country

In the end, the last words should probably go to the women who actually live in these camps. In a piece published on April 4th 2023, the residents of Gambaga camp had this to say to journalists:

“We’re safe, happy and comfortable at Gambaga Camp”