“She appeared to be in some kind of trance”

This work is licensed under the Creative Commons | © Joost Fontein. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau4.1.003

“She appeared to be in some kind of trance”

Anthropology and the question of unknowability in a criminal trial

Joost FONTEIN, University of Edinburgh

This is a personal account of a recent criminal trial in the United Kingdom that the author was involved in as an expert witness, involving a young Zimbabwean woman who attacked her mother with a knife, when she (as she, her mother, and relatives claimed) was possessed by an evil spirit as the result of another family member’s witchcraft. Evidence for her abnormal state of consciousness was corroborated by police evidence that described her as “in a trance” on the night in question, and despite a wide range of medical and psychiatric assessments, no clear neurological, medical, psychiatric, or sleep-disorder causes for her “possession” were ever established. The article describes the difficulties encountered in producing anthropological evidence for the criminal court that sought to go beyond the limitations of conventional forms of “cultural defense” to argue for the limits of knowledge and the “possibility of other possibilities.” With a nod to Harry West’s notion of “ethnographic sorcery,” this unusual court case illustrates how anthropological expert evidence can be constrained by courts constructing their own kinds of certainty, and yet still have efficacy in unintended ways.

Keywords: witchcraft, possession, courts, cultural defense, uncertainty, anthropology and the law

It was only moments after I entered the courtroom on day two of the trial that she said it. Despite the knife attack she had suffered almost two years earlier, and her long separation from her only daughter, not to mention the uncomfortable formalities of the crown court trial in which she and other relatives were prosecution witnesses, she emanated a remarkably calm and dignified presence in the witness box. The defense solicitor later described her as an excellent witness. But now under cross-examination by the defense barrister about the state her daughter (the defendant) had been in when she was attacking her mother, this dignified middle-aged Zimbabwean woman’s frustration seemed evident. The judge was seeking clarification. “When you say, ‘she didn’t look like Rachel’ what do you mean? Are you talking about her expression, mannerisms, or her movements? What was it, that wasn’t like Rachel?” insisted the judge. “It is hard to describe …” she responded, unable to find words that might explain for the judge, the jury, and the cross-examining barrister, how her daughter appeared that night: recognizable but different, known but unknown, present yet absent. Then turning to face the whole courtroom, the tremor of frustration reverberating in her voice, she demanded: “Is there anyone here in the court who has experienced spirit possession, who might understand this?” Sitting in the back row of the public gallery, I resisted the strongest urge to put up my hand. Later, as the trial unfolded over the next ten days, and even after the trial had concluded, part of me wished I had.

The barrister paused; “OK,” he said gently. Then he continued step-by-step through the minutia of the events of that evening two years before, “not to be rude, but I’m trying to establish a history.” The barrister drew on her previous witness statements and led her through what happened after the witness woke to find herself being attacked by her daughter, dressed in dark clothing, gloves, and an improvised mask, “engulfed by this thing” as she put it. “She felt stiff in my arms” followed by “moments of being very scared,” the woman explained. “Rachel said, ‘they are going to kill us both…. I want to go to my dad,’” and then “she tried to hide in the small space behind the wardrobe … like she was trying to get away from something.” There was some lengthy discussion between the barrister, the witness, and judge about the exact size of the space Rachel had been trying to hide in. About a foot, the judge suggested. “Would she fit in that space?” the barrister asked. “No, she wouldn’t,” the woman replied.

The detailed examination continued. Photographs of the house and the woman’s stab wounds were discussed. A cut on her cheek needed four stitches and she had three or four stitches for each of the four cuts on her arm. The woman kept her composure until the barrister asked about her relationship with her daughter. “You described these events as totally out of character, that Rachel was never aggressive … despite being a teenager.” Then when the barrister continued, “So when the police arrived Rachel was in your arms …” the woman started to cry and the judge suggested a short break. Later the woman described the deep distress she experienced being separated from her daughter since the event, “other than in [remand] prison, I have had no time alone with her, from then until now.”

A short while later the defense barrister asked the woman if she had any view of what was behind the knife attack she suffered from her daughter. “I do have a view,” she replied, but the judge immediately intervened, querying whether this was a fair question. The prosecution barrister confirmed that he was “concerned where we are going, given the opening statements.” As the two barristers spoke to each other in hushed whispers, the judge explained to the jury: “we have rather precise rules about what a witness can be asked about in court. A witness, except for expert witnesses, cannot be asked their opinion about things.” The conferring barristers agreed that the defense would ask another, second, much more indirect question.

Slowly, even laboriously, the defense barrister asked Rachel’s mother about her Christian beliefs, about prayers being said in her church for Rachel, and about the services of a Botswana muprofiti (a prophet)—he articulated the word very carefully—engaged for “Rachel’s spiritual welfare.” Already chastised by the judge, and on unfamiliar ground (or “in Indian territory,” as he put it), his questions were imprecise and vague. Rachel’s mother answered the awkward questions as best she could. “The muprofiti has been there … to pray for her two times. I was concerned because her actions were so out of character, so I asked the muprofiti to pray for her.” “For what purpose?” the barrister asked. “A muprofiti cleanses evil spirits,” she replied. “So it was to cleanse a spirit?” the barrister pressed on. “Yes, because in my view she was …” but she was interrupted again. “I cannot ask you about this, your view is not relevant to the court.” He changed tack, asking about the stature and authority of the muprofiti from Botswana. “Did you discuss what had happened with this senior religious figure?” he asked. “I told him I had a problem, but as a muprofiti you don’t really need to tell him anything,” she replied. Seeking clarity the judge interceded again: “You told him, ‘I had a problem with my daughter.’ Did you tell him your daughter had attacked you?” “Yes,” was the reply, repeating, “but you don’t really need to tell a muprofiti many things.” Still seeking clarity, the barrister continued, “So you told him your daughter attacked you but not any of the details?” “I asked him to pray for Rachel, but it was not necessary to go into the details,” she responded. Trying to make sense of this the barrister suggested that the muprofiti “was dealing with spiritual matters, but not what materially happened …” But before he could finish, and almost as if the muprofiti was listening, the fire alarm went off, and with orderly haste people left the courtroom, as the building emptied onto the street outside.

The story I tell here is about what happens when different ways of knowing, of ordering the world, of making sense of the unusual and the uncertain, confront each other within the disciplined, highly structured world of UK courts. But it is a story less about cultural difference than about the social (cf. Mosse 2006) and legal means—practical, performative, and discursive—through which certainty is constructed, and uncertainty, indeterminacy, and the possibility of other possibilities are encountered and vanquished through the courts. In his account of sorcery on the Mueda plateau in northern Mozambique, Harry West (2005) suggests that both sorcerers and countersorcerers (and a host of others: villagers, local leaders, but also anthropologists, missionaries, colonial authorities, FRELIMO, and most recently neoliberal reformists) are drawn into a never-ending game of kupilikula, through which they temporarily transcend the material, everyday world, in order to envisage, articulate, and remold it “in accordance with their own visions of a world reordered” (West 2007: x, 2005).1 West argues that anthropologists too make “transcendent manoeuvres” that “reorder the world,” and hence ethnography is a kind of sorcery, albeit (one hopes) without the malevolence normally associated with witchcraft.2 Here in this story, which can be read as a decidedly “postcolonial,” “multicultural,” and “diasporic” encounter3 between different epistemologies, and perhaps even ontologies (Henare, Holbraad, and Wastell 2007)—between commonplace but diverse and inevitably contested Zimbabwean understandings and practices of spirit possession, and Western knowledge/epistemology filtered through the determining space of legal practice—courts of law too can appear like arenas for sorcery in West’s sense, constructing a transcendental position in order to enforce certainty upon an otherwise “inescapably inchoate world” (West 2007: 69).

But more than simply an account of how a court asserts its kind of certainty upon events lying at the edges of its epistemological reach, this is also a story about the difficulties anthropologists can face in making their evidence commensurable and useful for courts of law. For those providing evidence on spirit possession and witchcraft, the consequential nature of the circumstances throws into sharp relief our normal handwringing about the political imbrications of anthropological representations of the “others” we study. If such self-conscious handwringing normally turns on problematics inherent to the textual distancing and objectification of social worlds encountered/generated by fieldwork—in the movement from field to desk and toward what David Mosse calls “anti-social anthropology,” akin to West’s ethnographic sorcery, if you will—then as Mosse discovered (2006), such anxieties can be deeply amplified when our own “sorcery” is in turn constrained and transcended in socio-political contexts in which we feel ourselves more pawns than players. Yet three years in the wake of Rachel’s attempted murder trial, I am less convinced now about the extent to which anthropological evidence was marginalized and ineffectual in the courtroom than I was in the immediate wake of the trial. The story I tell here therefore raises complex questions both about the difficulties anthropologists can face as expert witnesses in criminal trials, and about the ways in which anthropological “evidence” can have unexpected efficacy in courtroom contexts.

To protect those involved, in what follows, the names, places, and contexts have been deliberately (and quite extensively) anonymized and obscured. The events described are filtered through my (necessarily imperfect) notes and recollections, an interview I conducted with Rachel as an expert witness for the defense, as well as various witness statements and other court documents. So to some extent this account does amount, in West’s sense, to a final sorcery of my own. Much of what I describe here, about the way particular regimes of expertise and knowledge are reproduced in legal settings, reflects well-rehearsed arguments made in legal anthropology (cf. Conley and O’Barr 1990, 1993, and 1998; Good 2006, 2008). There are also important resonances with ongoing debates about cultural and legal pluralism, political liberalism, and the law (cf. Shweder 2010a; Webber 2006; Inksater 2010; Merry 1988), provoked particularly by the work of Clifford Geertz (e.g., 1983); as well as with earlier anthropological accounts in African contexts (such as by mid-twentieth century scholars working from the Rhodes-Livingston Institute—see Gluckman 1963, 1965, 1967), examining the social and cultural constitution of law, and how in legal contexts “reasonable behavior” is usually defined with reference to social, cultural, and political factors not contained by the law.4 Unfortunately, lack of space means I cannot do justice to these other works, considered and relevant though they are, and therefore I restrain myself to a reflexive account of Rachel’s trial, and my involvement with it, as one of two anthropologists called as expert witnesses by the defense.5

Background: “In some kind of trance”

Let me provide a basic account of the case, and the “agreed facts.” In late spring 2009, Rachel’s mother (who we shall call Mrs. Moyo), a practicing nurse living with her daughter in the Midlands, awoke around midnight one evening to find herself being attacked in her bedroom by what she initially thought was an intruder armed with a knife. In wrestling the knife off her attacker, she realized it was her teenage daughter Rachel. But Rachel did not appear herself: “It didn’t look like Rachel but it was her,” she told police in a witness statement. “Rachel’s voice was very strange,” the “expression on her face was so weird—terrified, so scared.” Rachel was “saying that she was hearing voices,” “telling her they were going to come to kill us,” and she “kept screaming ‘I want to go to my daddy.’” Bleeding from her stab wounds, and restraining her daughter, Mrs. Moyo managed to call her brother, whom we shall call Toby Ndlovu. He called the police and went straight around to the house. Hearing him knocking on the front door, it took her fifteen minutes to struggle herself and her daughter downstairs to open the door. Toby found his sister “drenched in blood,” with an improvised bandage around her upper arm. Rachel “was just sobbing” and “petrified, I could not describe it, she was not herself…. I saw fear in her eyes.” She “was still shouting” when the police arrived. His wounded sister was calm and holding onto her daughter like she was “cradling someone.” Toby explained to the court how Mrs. Moyo described to him what had happened as she was treated at the scene by paramedics; how Rachel had said to her mother that “voices had told her to kill her mum as someone else was coming to kill both of them anyway…. She thought it would be better that she kills her mum and herself … so that they could go and be with her dad.” “It was like Rachel was possessed,” Mrs. Moyo had told her brother, adding that “during the struggle for the knife Rachel had been so strong … it had taken all her strength to get the knife off her.” “Everybody who knows her is just so stunned and shocked,” Toby told the police in his witness statement, “this is completely out of character. Rachel is never aggressive.”

Mrs. Moyo’s and her brother’s descriptions of Rachel’s “altered state of consciousness” on the night of the event were corroborated by police statements. Various officers described her as “extremely distressed,” “hyperventilating,” and “sobbing uncontrollably … rocking backwards and forward,” despite their efforts “to calm this female down,” which were “having little effect.” The most vivid description came from the arresting officer, who arrived on the scene only moments after Toby, and described how Rachel was shaking, crying, and “kept saying ‘my dad, my dad…. My Nan told me my mum had killed my dad, she wants me to do it, she will kill me.’” “I asked her where her Nan was,” the officer told the court, “she said ‘she’s here to get me,’” describing how Rachel had been looking over her shoulder towards the front door. “I was expecting that her Nan lived with her or visited the address,” the police officer explained under cross-examination. She had not known that Rachel’s grandmother passed away almost a decade before in Bulawayo, about a year after her father had died in April 2000. Once arrested on suspicion of attempted murder and cautioned without any response, it was only en route to the police station, that Rachel “quieted down,” and “upon being booked she appeared calm and composed” according to another officer. In the arresting officer’s words, “she was as if she was another person, calm and collected, but when I first arrived she appeared to be in some kind of trance.” “In the van, she seemed to have changed into a different person,” the officer told the court. When told she had been arrested for attempting to murder her mother, it was with “astonishment” that Rachel had then asked, “Why would I do that?”

Before and throughout her trial, Rachel Moyo was consistent in her claims that she did not remember the events of that evening after she had gone to bed, nor changing into other clothes, gloves, and an improvised balaclava, going downstairs to collect a kitchen knife, or attacking her mother with it in her bedroom. She did remember an unusually vivid dream, “this weird dream that seemed real” (police interview), in which she saw her dead grandmother, her “gogo,” and her paternal aunt, talking to her in Sindebele “telling me that my mum is responsible for my father’s death and that I should do the honorable thing by my father and … kill her just like she did my dad.” “It’s as though she was guiding me,” she added during her police interview, “and then the only thing I remember after that is when my mum was actually fighting back.” “Okay, so the obvious question is,” the police interviewer asked the day after the events, “are you awake?” “That’s the thing I can’t explain because to me it seemed like a dream … where I was going to wake up. Its like, in the dream, I could see myself actually stabbing my mum, but … in reality when I think about it, I just don’t remember my actions except when she was fighting back,” Rachel replied. “My first conscious recollection where I was knowing where I was and what I was doing was when the police were actually here.” “So when you’re attacking your mum and she’s fighting back, you’ve no conscious recollection of that? That’s what you’re saying?” the police interviewer pressed. “Yeah to me it was just like a dream,” she replied. “So you can see it happening and you can see you doing it … but you are not consciously aware of you doing it. Does that make sense … ? I think I know what I’m saying,” the policeman hesitated. “Yeah,” Rachel replied, “that makes perfect sense.” “So,” the policeman continued, “you come to your senses when the police arrive, and do you remember any conversations with the police officers?” “The woman taking care of me, she just kept trying to tell me to calm down, but I couldn’t calm down because it’s like at that point I was still sort of seeing my Nan and at the time she didn’t look very happy …”

Noting how other police statements “tie up with what you’re trying to say,” the police interviewer suggested that “at that point, we accept that you are, in reality, back with the rest of (pause) reality really”; illustrating something of the difficulty everyone at the interview, and later in the courtroom (where Rachel’s police interview was rehearsed in fine detail), had understanding the events of that evening. In fact, the arresting officer’s evidence and that of Toby Ndlovu, as well as Rachel’s claim in court that she only recognized the officer from when she had been booked into custody at the police station, suggests that she was only “back with the rest of reality really” much later that night. Before and throughout the trial Rachel remained adamant she had no reason to attack her mother and did not remember doing it. As she tearfully told the court from the witness box, at the climatic end of the fourth day of the trial: “They were my hands, my body, that did that to my mum, but not me. I love my mother; she is the only person I have. I don’t even talk back to my mother. Why would I do that to my mother?”

No ordinary “cultural defense”

Over the past decade, there has been a proliferation of cases in the United Kingdom and elsewhere that involve beliefs and practices to do with possession and witchcraft.6 Where such cases have gone to court, they have sometimes resulted in verdicts of not guilty by reason of insanity, but where there is a lack of medical or psychiatric evidence or where other beliefs are perceived to be involved, a cultural defense has occasionally been mounted. Something like a cultural defense was involved in Rachel’s case, but this was by no means straightforward.

In cases involving Africans living in the United Kingdom and Europe, so-called child witch exorcisms and “muti murders” have gained the highest media profile, and have been the subject of a number of investigations and television documentaries.7 Coverage of such cases has sometimes left scholars profoundly disconcerted, provoking acrimonious debate about the reification of sensationalized, exoticized, and primitivizing images of an aggregated “African occult.”8 Not withstanding that such cases are without doubt profoundly disturbing, it has scarcely been acknowledged beyond academic circles that trade and use of body parts for ritual purposes has a long and complex history across Africa and beyond (cf. Fontein and Harries 2013; Bernault 2013; Espirito Santo, Keretetzi, and Panagiotopoulos 2013), and that discourses and practices of spirit possession are as common as they are culturally diverse, contested, contemporary, and historically emergent; and therefore decidedly not erroneous (or necessarily dangerous) “hang-overs” from a “primitive” past. Indeed, extended scholarly debates about the “modernity of witchcraft” (Geschiere 1997) illustrate how, more often than not, discourses and practices to do with spirit possession (in all their complexity and diversity) can not only perform “positive” social “healing” functions but also reflect the messy contemporary politics of postcolonial states in an often alienating, globalizing world. In this respect, the alarmist sensationalism with which such (relatively rare) cases of child witch exorcisms and muti murders have been met in Western media, speaks volumes more about the dilemmas of inclusion and exclusion facing contemporary European societies—what Geschiere calls “the perils of belonging” (2009)—than about the salience of such beliefs and practices in particular African contexts. Scholars (e.g., Ranger 2007) have therefore often responded by emphasizing the immense diversity of African religious beliefs and practices, and the need for detailed empirical historical and ethnographic research situated in particular historical, cultural, social, and political contexts.

These debates are part of the broader context within which Rachel Moyo’s case was inevitably entwined; yet her situation was somewhat different. In the absence of any clear medical, psychiatric, or sleep-disorder related cause for Rachel’s actions, despite a cacophony of tests by experts, “guilty by reason of insanity” was never an option the defense or the prosecution was likely to entertain, even if the judge held it up as a possibility for the jury to consider. At the same time, however, no ordinary cultural defense could be mounted because Rachel was, by all accounts, not conscious of what she was doing. If by “cultural defense” it is meant that someone tries to justify their conscious and deliberate actions in terms of responding rationally to a deeply held cultural belief—in Rachel’s case perhaps that her dead grandmother or aunt wanted her to avenge her father’s death—then it simply did not apply here because the accused, her mother, and other family members all believed it was not Rachel who intentionally did this, but an evil spirit possessing her. Having interviewed Rachel herself a year before the trial, my conclusion was that she and her relatives had come to the belief that it was another, maternal aunt (not the paternal aunt in her dream) who used witchcraft to attack Rachel’s mother using her body, for an entirely different malicious purpose unrelated to her father’s untimely death in 2000. Yet because an ordinary “witness cannot be asked their opinion about things,” but also because even the oral evidence of expert witnesses is heavily constrained in court (especially by the cross-examination of barristers), Rachel and her mother’s (the “victim”) views were never, in the end, discussed in court in any detail. For most of the trial—but perhaps not the sentencing (to which I return below)—their beliefs in the existence of malicious spirits, witchcraft, and the possibility of possession were only briefly acknowledged and then set aside as irrelevant to the business of the court.

Of course the way cultural defense (Renteln 2005) has been deployed in legal contexts has often made anthropologists feel distinctly uncomfortable.9 As Good (2008: 51, 53–54) explains, one of the problems of cultural defense in court cases similar to this one, particularly in the United States, is that it is often based on rather outmoded and simplistic notions of “cultures” as separate entities with distinct and separate belief systems that determine people’s thought and actions. Few anthropologists today adhere to such deterministic and bounded notions of culture because of the overwhelming evidence that cultural boundaries are in no way static or fixed, nor does culture in any straightforward way determine people’s thoughts, actions, and motivations. Rather, “culture” and its “boundaries” are fluid, processual, and emergent, and importantly, always subject to people’s creative, scrutinizing, conscious, and continuing construction and indeed contestation. As Good (2008:51) describes, “most contemporary anthropologists would I imagine agree with Zygmunt Bauman that culture ‘does not cause behaviour, but summarises and abstracts from it, and is thus neither normative nor predictive’” (1996: 11). Therefore, belief in the possibility and prevalence of witchcraft and spirit possession may in part determine or structure how a strange event can subsequently be interpreted, understood, or deciphered, but could not simply, or in any deterministic or mechanical way, cause someone to do or experience something unintentionally or unconsciously.

Due to these reservations, as well as Rachel’s insistence that she could not remember her attack upon her mother, the defense case to which I envisaged contributing as an expert witness (in my consultations with the barrister before and during the trial, and in my formal written report), was rather different. Indeed my report tried to do several slightly different things, some of which might have fitted a more conventional cultural defense even as I attempted to reach beyond that. These turned out harder to reconcile than I initially imagined. Let me discuss these briefly, to show how I approached my difficult task (and why), before returning to the trial itself to explore what traction (if any) this anthropological “evidence” actually had.

The first aspect of the report drew specifically on my country expertise, and did reflect the kind of material that might be drawn upon for a conventional cultural defense. Outlining the different witchcraft and spirit possession explanations that might arise in Zimbabwean contexts, I tried to show the cultural consistency and coherence of Rachel and her family’s beliefs about what had happened.10 I drew heavily on my own experience and the wider literature, but also my interview with Rachel in Edinburgh in early 2010, which had focused on how she (and her family) understood the events, and how these explanations were arrived at, in relation to other possible explanations.11 For example, aware that witchcraft accusations frequently occur between extended, affinal kin around the death of a senior family member, we discussed whether Rachel’s paternal aunt (who appeared in her “dream”) might have been responsible for the attack her body carried out on her mother. This aunt had in the past indirectly accused her mother of being responsible for Rachel’s father’s death, an accusation that itself implied earlier witchcraft by her mother. Rachel had discounted this idea following communications with her family, and in consultation with the muprofiti, concluding that her maternal aunt, umama omdala (mother’s older sister), was responsible for the attack, motivated by jealousy and a desire to acquire Rachel’s mother’s property in Bulawayo. In my report I linked this to a common belief among some Zimbabweans that the threat of witchcraft can be most severe from much closer (i.e., “blood”) relatives. Emphasizing the uncertainties often provoking, and provoked by (Niehaus 1997), witchcraft events/accusations,12 I discussed how Rachel’s deference to others’ explanations (her relatives and the muprofiti) reflected structures of authority related to age, kinship, and status that continue to have salience among Zimbabwe’s UK diaspora, particularly in family matters; and that in some ways mirrors the “hierarchicalization” of knowledge in UK courtrooms.13 My purpose of presenting this kind of material was to show how Rachel and her family’s explanations made sense in a Zimbabwean context, which of course would have been vital for any ordinary cultural defense.

The second part of the report drew more on my anthropological rather than my specific “country” expertise, and was primarily targeted at ensuring the court could (if it wanted) engage in an informed discussion of the efficacy of witchcraft and possession beliefs and practices. This material too could fall within the requirements of building a conventional cultural defense, however my intention was particularly to try to reduce distance between Rachel and her family’s understanding of what had happened, and what the court might find reasonable and commensurable. I therefore discussed how the anthropology of witchcraft had built upon three important points made by E. E. Evans-Pritchard (1937): First, that beliefs about “witchcraft” and magic are not evidence of illogical, prelogical, or “primitive” thought, and indicate not a lack of but rather “an excess of rationality,”14 even if Evans-Pritchard himself stated the Azande were wrong to believe in oracles. Second, witchcraft accusations closely reflect social tensions and can therefore be seen as a “social strain gauge” (Marwick 1965). And third, that the distinction between “witchcraft” and “sorcery” Evans-Pritchard observed, remains of huge analytical significance because it reflects how questions of individual agency, responsibility, and intentionality are often key for how any “witchcraft event” is judged, and for any measures taken in response.15

This last point, I argued, suggested that questions of intent and consciousness that are often of central concern in legal settings such as Rachel’s trial, might be mirrored in the inevitable concerns that arise from the uncertainties of agency, consciousness, and intention provoked by witchcraft situations, in Zimbabwe or elsewhere. In this particular sense, I suggested, there might be little difference between how a case like Rachel’s is handled in a UK courtroom and how it is handled in Zimbabwe, even if the 2006 repeal of colonial-era legislation criminalizing witchcraft allegations means that the possibility of witchcraft and spirit possession are now given due hearing in that country.16 To illustrate this, I discussed a widely publicized case (the “winnowing basket” case) that appeared before Harare magistrates in May/June 2009, to show how questions about responsibility, consciousness, and intention can lie at the very core of responses to alleged witchcraft, as indeed one expects they might in any serious criminal trial. This involved a woman found naked in Harare’s Highfield suburb, who claimed she was flown there on a winnowing basket by her father-in-law and aunt, intent on using her to kill another family member. After hearing “expert evidence” from Zimbabwe’s National Association of Traditional Healers (ZINATHA), the Harare court recognized “the accused is also a victim who was being used by a relative,” gave her a suspended sentence and ordered her “to seek help to recover from the spell apparently cast on her by her father-in-law.”17

This example related to the third, most ambitious (and controversial) aspect of the report, where I considered the importance of anthropology’s long engagement with the difficult epistemological and ontological problems—or “uncomfortable questions” (Niehaus 2013: 210)—raised by experiences of witchcraft and possession, to do with ways of knowing, their relationship to the nature of the world, and ultimately about the limits of knowledge. Here I sought to suggest that (and why) a contemporary anthropological stance to such events might emphasize profound uncertainty; what I framed as the “possibility of other possibilities.” This, I felt, could be key to an unconventional kind of cultural defense for Rachel’s case because it implied that the possibility that Rachel and her family’s understanding of what had happened might be correct—to put it crudely, that spirit possession/witchcraft might actually “work”—be left deliberately open, as such things are not empirically provable either way without recourse to very particular and deep-founded cultural, philosophical, or ontological assumptions about how the world really operates. This argument built on a notion (which some might question) that an anthropological approach involves seeking to uncover how the world might work according to the often profoundly different cultural, philosophical, or ontological assumptions of differing peoples. This open-ended or even agnostic stance suggests, I argued, that we leave open and indeterminate the possibility that there are possibilities that lie outside the limits of our knowledge.

In making this assertion, I suggested anthropology has moved on from Evans-Pritchard’s notion that Azande beliefs about oracles and sorcery were “rational but wrong,” and that it is now well-recognized that all “truths,” including “scientific knowledge” are necessarily socially, culturally, historically, and politically constructed and therefore contingent (Bloor 1976; Shapin 1996). With hindsight, and in response this article’s various reviewers, perhaps this assertion about anthropology’s common cause were exaggerated.18 Certainly many anthropologists remain wary of the kind of cultural or even “cognitive” relativism long ago propounded by scholars like Peter Winch (1958, 1964), which posited that “rationality” itself was culturally constructed, and if taken to extreme, denies the possibility of “cross-cultural comparison,” undermining the very usefulness or even possibility of a cross-cultural discipline like anthropology itself.19 Yet in the context of anthropology’s so-called post-modern (Willis 1996, 1999) and more recently “ontological” turns (e.g., Henare, Holbraad, and Wastell 2007), some anthropologists have argued that “magical” phenomena are or can be “objectively real,” even if they are necessarily inexplicable in terms of “western scientific rationality” (cf. Stoller and Olkes 1987). While more extreme versions of such approaches have insisted upon the existence of “separate realities” or “other worlds,” based on very different fundamental (ontological/metaphysical) assumptions about the nature of reality, others continue to be concerned about notions of “radical difference” such analyses imply (cf. Fontein 2011; Carrithers et al. 2010). Reservations have also been expressed about how such approaches exacerbate the “long history of exceedingly ‘generous’ or ‘charitable’ anthropological treatment of witchcraft and of divination” (Niehaus 2013: 211), and the serious consequences that can result. Beyond fears about a proliferation of “absurdity” (cf. Lett 1991: 318), some argue that anthropologists can “ill afford to adopt an uncritical, purely interpretive, stance towards witchcraft and divination” because too often such beliefs can enhance very real “fears, anxieties and spiritual insecurities” with sometimes devastating consequences; particularly, for example (but by no means exclusively), in the context of South Africa’s AIDS pandemic (Niehaus 2013: 212; Ashforth 2005: 69).

In writing about these debates and trying to remain relevant for the court, I therefore argued that a more moderate approach, based on a compromise between these positions, probably represents the moral and analytical position adopted by many (but clearly not all) anthropologists today. This involves leaving open the possibility that magic, witchcraft, and similar occult forces might operate in any particular context, in order to avoid making judgments on the basis of necessarily contingent assumptions about the nature of reality, while at the same time not essentializing cultural difference, or championing simplistic notions of cultural determinism that deny agency, change, and historicity. Importantly, such an approach is based less on “radical cultural relativism,”20 and more on a recognition of the limits of knowledge and the prevalence of uncertainty. Similarly, in contrast to some “interpretative” approaches (cf. Jackson 1989) criticized for treating questions about the reality of witchcraft, possession, and divination as irrelevant as well as unknowable (Niehaus 2013: 211; Lett 1991: 313), the stance I was advocating might treat such questions as potentially significant exactly because of their profound “unknowability.”

It was my contention that in Rachel’s case a similar stance could be adopted by the court. That is, leaving open as undecided and unanswerable the possibility of people doing things outside of their own consciousness and intentionality, without clear medical or psychiatric cause and for whatever reason—including the explanation Rachel and her family had come to—yet without relying on a vague cultural determinism positing causal links between beliefs in the existence of witchcraft or spirit possession and the difficult-to-explain events of that day. Emphasizing the diversity of cases of people in altered states of consciousness anthropologists have encountered, I suggested focusing attention on Rachel’s intentionality might be the only way to allow the unanswerable and ultimately unknowable questions (to do with the possibility of “demonic possession,” malevolent “witchcraft” by a “witch,” “sorcerer,” relative, or a unhappy spirit, or of “other cultural explanations” as the defense psychiatrist later glossed it) to be set aside and left open-ended. The case might then ultimately depend upon how the court judged the credibility of the key witnesses and the accused, and again this would make it much less “strange,” “exotic,” or “different” than any other “normal” criminal trial.21 In this way the kind of cultural defense I envisaged contributing to involved what anthropologists are often said to do—“making the familiar strange, and the strange familiar”—unlike more conventional, glib kinds of cultural defense, which tend to reinforce outmoded, simplistic, bounded, and overdeterministic notions of “cultural difference.” In other words, I argued more for the recognition of profound uncertainties for everyone concerned rather than for the existence of profound alterity.

Dissociative states, other possibilities, and “noninsane automatism”

But as it happened, none or few of these arguments, what I envisaged as my anthropological contribution to the trial, was made much use of. The sorcery of the court, as I later came to think of it, meant that neither the detailed account of the complex nuances of Zimbabwean beliefs and practices, nor of anthropology’s long critically engagement with witchcraft and spirit possession, let alone the more ambitious (but perhaps naïve) attempt to encourage the court to recognize the limits of knowledge, made much of an impact, at least directly. The real courtroom drama occurred in the exchanges among barristers, witnesses, and the judge, and most of all, between the two psychiatrists called as expert witnesses by the prosecution and the defense.

The prosecution’s case was simple: Rachel Moyo was lying. She had deliberately tried to kill her mother, and then tried to cover up her actions by pretending to be unaware of what she had done. They built their case by seeking to discredit the accused, focusing on a series of lies and inconsistencies to do with events in the months before the night she attacked her mother, when her attendance at school had become patchy, and some months afterward when she broke her bail conditions and lied to cover it up. Central to the prosecution’s case was the argument that the lack of any known medical, psychiatric, or sleep-related disorder, and the complexity of the actions she undertook before the attack—getting dressed, preparing a mask, collecting a knife, and so on—discredited her account of not being aware of what she had been doing. Therefore the only possible explanation was “criminogenic”; Rachel was simply guilty.

The problem the prosecution had was establishing a motive, and the barrister’s attempt to suggest Rachel harbored resentment against her mother to do with her father’s death almost ten years before was not very convincing. Both Rachel and her mother were clearly as distraught from their long forced separation as from the events that had led to it. The defense barrister was able to effectively cast doubt on the prosecution’s attempts to portray Rachel as a compulsive liar, by thoroughly and convincingly explaining the context of each inconsistency and untruth. For her part, Rachel explained her “terrible lies” after she had broken her bail conditions, by referring to the lack of efficacy her earlier truths had had. As she put it, when asked in the witness box why she had lied to police after breaking her curfew, “I had given the truth [before], and that truth had not helped anything, so why tell the truth?” “It was so unfair,” she explained to the court, “in May I was just a normal person, [then] my whole life turned upside down for something I can’t remember … now I can’t see my mum or my friends, that’s what I mean by unfair.”

The argument the defense barrister was aiming at was that Rachel was both “not guilty” and “not insane,” and that she had been in a once off (thereby unlikely to be repeated) “dissociative” state, probably triggered by her “dream” or “nightmare,” and therefore her actions amounted to a form of “noninsane automatism.”22 This was a very thin target to aim for, yet I felt that in this possibility of “noninsane automatism” there might be room for the open-ended stance and the recognition of “other possibilities” and profound uncertainty my report had advocated. However it became clear early on that the barristers and the judge had problems making sense or use of the anthropological evidence in the report. My report was too complex, too detailed, and probably too ambitious; perhaps it spoke the wrong language. In fact by day four of the trial, when the defense barrister requested the normal sequence of witnesses be changed to allow the anthropologists to give evidence immediately after the psychiatrists, it became clear the judge had not yet read the full report but only its brief summary.23 In our discussions the defense barrister seemed open to the suggestion to “suspend his disbelief,” and tried to build his defense around the possibility of “other possibilities,” but had difficulties expressing this in a way that the prosecution barrister and judge would not object to.

The anthropological evidence seemed to have more impact on the two opposing psychiatrists called by the prosecution and defense. This became apparent to me both by talking to them informally, but also when the oral evidence each presented to the court differed in varying degrees from the reports they submitted in advance. In the courtroom, cultural “factors,” “aspects,” “context,” “influence,” and even “contamination” took on far more significance than in their written reports. There was real drama in the psychiatrists’ exchanges as they battled in the front of the jury on the Wednesday and Thursday of the first week of the trial.24 The two psychiatrists cut very different figures in the witness box. The prosecution psychiatrist was softly spoken, hesitant, and often inaudible and unclear. In contrast the defense psychiatrist filled the entire witness box with his presence, his outstretched arms leaning on its sides with a confidence that substantiated his professorial authority, his voice dominating the whole courtroom. His eyes never leaving the jury, his was an impressive performance even as that of his “opponent” appeared weak and fey. Both the defense barrister’s cross-examination and the defense psychiatrist’s performance amounted to such a robust rebuttal of the prosecution psychiatrist’s oral evidence, that the prosecution recalled him to give further evidence on the Thursday morning in an effort to respond.

Yet both psychiatrists agreed that the cacophony of medical, psychiatric, and sleep-disorder tests carried out on Rachel indicated no obvious psychological, neurological, medical, drug, stress, or sleep-related issues that could have caused her to go into an abnormal state of consciousness. The prosecution psychiatrist argued this probably meant the accused was simply lying, and indeed their entire case built on this. His conflation between the lack of clear medical or psychiatric causes and his statements about the reliability of the defendant was effectively exploited by the defense barrister. But his strongest argument was that Rachel could not have been afflicted by so-called trance/possession disorder because the complexity of her acts before her mother was attacked were not “ordinary or everyday learned behavior.” It was hard, he said, “to see that as anything other than consciously driven.” If some kind of “dissociative state” was involved, he conceded, it could only have occurred after she was interrupted by her mother waking up during the attack. He did, albeit briefly, acknowledge that psychiatric assessments were usually based to some degree on “subjective interpretation,” and later under cross-examination, that there were “other possibilities” to do with “culture,” which lay outside his expertise.

The defense psychiatrist argued that despite the lack of obvious psychiatric or medical causes, all the evidence nevertheless pointed strongly to the accused having been in a “dissociative state.” He mentioned known cases of people doing very complex actions in dissociative states, and argued that Rachel’s dissociative state probably had something to do with the transient period between sleep and wakefulness, the “twilight state,” he suggested she was in at the time. Importantly, he argued that in the absence of clear medical or psychiatric explanation, cultural factors were probably highly significant to the case, and could come into play in two ways. First, known psychiatric disorders might manifest in “peculiar” or unexpected ways according to cultural context (or even “contamination”); a point I suspect few anthropologists would (in principle) disagree with. As he put it, Rachel’s description indicated a “dissociative state that is consistent with her culture.” Second, more interestingly, he suggested that there may be “states of abnormal consciousness,” which are not explainable in conventional medical or psychiatric terms but do find explanation in other “cultural narratives.” This was a point of agreement that he and I had established during a brief recess just before he went into the witness box. This, I sensed at the time, might have given me a window to express the argument in my report about allowing for the “possibility of other possibilities”; that Rachel and her mother’s explanation for her dissociative state might be true, or at least provided explanation where other known explanations fell short.

Under cross-examination, the defense psychiatrist explained that a dissociative state would be considered a “disease of the mind” if caused by something “inherent” or “internal”; unless “triggered from the outside” by an “external” factor. In the absence of an “external blow” or “trigger,” any automatism must be legally an “insane automatism.” But, he added, if there is “a cultural explanation” for the dissociative state, it would be “for the judge to decide” whether such “other,” “cultural” factors amounted to an “internal” or “external” cause, and therefore whether it was “still a disease of the mind.” This became the issue upon which the whole case turned. If cultural factors were to be understood as external, this would mean the court accepted the possibility that Rachel’s actions were the result of “noninsane automatism,” and therefore that she might be found not guilty. If, however, such cultural factors were deemed an internal cause for a dissociative state then her actions would legally be defined as “insane automatism,” and she could only be found not guilty by reason of insanity. As a point of law it could only be up to the judge to rule on this, the defense psychiatrist stated, warning at the same time that “of course culture is not a disease.”

Any anthropological audience would likely have been aghast at the crude use being made of its still hallowed (and much debated—see Carrithers et al. 2010) notion of “culture,” but the fact these questions were raised at all seemed to indicate that the anthropological evidence in my written report had had some indirect effect on the psychiatrists’ expert evidence in court. Of course, the question of whether cultural factors are external or internal not only manifests the kind of Cartesian distinctions anthropologists have become hypersensitive to, but also echoed the concerns of anthropologists about the way cultural defense has often been invoked in courts of law. But listening from the back of the court, these were not concerns preoccupying me at that moment, when I was preparing myself to take the stand. Rather, I was thinking about how to construct a point of convergence with the defense psychiatrist’s evidence. In particular, how to suggest to the court that the cultural narratives through which Rachel, her mother, and the muprofiti explained the events of two years earlier—that she had been the unwitting tool (and victim) of a third party’s witchcraft causing her to be possessed by an evil spirit and attack her mother—did exactly the opposite to what psychiatric or medical evidence was expected to offer the court, and had manifestly failed to do here (i.e., offer an internal cause), by providing an external explanation for Rachel’s abnormal state of consciousness.

The defense barrister, however, tried to exploit the psychiatrist’s evocation of other cultural possibilities in a different way. Perhaps recognizing that the “possibility of other possibilities” would never gain traction with the jury or the judge, he focused upon the first of the defense psychiatrist’s explanations for how cultural factors could be relevant for the case; that the way known “abnormal states of consciousness” manifest could be culturally dependent. Reexamining the defense psychiatrist he asked whether it was possible Rachel’s dissociative state had been caused by her “random nightmare.” “I can’t rule it out,” the psychiatrist acknowledged, adding, however, that in his opinion Rachel had not experienced a nightmare but a “dream-like event” common to “twilight states between sleep and wakefulness.” “It’s possible, but a bit unlikely,” he concluded. “So it is possible,” the barrister pushed on, “that a random nightmare with particular cultural sources or context could have caused Rachel to go into a dissociative state?” “The cultural origin of such a dream does seem evident,” the psychiatrist hesitated. “And it remains possible that a nightmare could have caused her dissociative state?” the barrister pressed. “Yes,” the psychiatrist conceded uncomfortably, “this was a possibility.” The defense barrister had maneuvered him into a difficult position, and after the trial closed for the day, he warned the barrister that a nightmare could only be interpreted as an internal event.

This idea that a nightmare triggered a dissociative state did not really tally with anyone else’s ideas in the courtroom—not Rachel and her family, not the two psychiatrists, not the anthropologists, and ultimately not the judge.25 By the end of the following Monday, after a day of legal argument, it became clear that, as the defense psychiatrist had warned, it was impossible for the judge to consider a dream or nightmare an external cause of any dissociative state, however consistent with Rachel’s “culture.” This would have profound implications for the whole case. It seems very doubtful, however, that the vague gloss of “other cultural narratives” offering explanation for Rachel’s abnormal state of consciousness where medicine and psychiatry had failed, would have fared any better, or been more easily accepted as offering an external cause for a dissociative state by the judge.

As it was, though, the opportunity never arose. In contrast to the multiple appearances of psychiatrists, it was much harder for the barristers and judge to agree upon the admissibility of anthropological evidence at all, let alone the “possibility of other possibilities” my report had advocated. After the two psychiatrists had given their evidence, and the prosecution closed its case, the defense requested to call the two anthropological “experts” (myself and my colleague) ahead of the accused who was the main defense witness. In the face of the prosecution’s objection to any anthropological expert being required at all, this provoked a very long discussion about the nature of anthropological evidence and what might be admissible. This culminated in the judge requesting the defense barrister and I refine the oral evidence I would give from my written report into a short document to be circulated by email to the judge and prosecution that evening, and decided upon the next day. In the meantime the defense called the accused to the witness box, and Rachel began her examination in the witness stand on the Thursday afternoon.

I had been in court for three days waiting to give evidence, but that night I was up through the early hours with the defense barrister and his assistant rewriting questions for him to ask me, which might allow me to suggest that Rachel and her mother’s witchcraft beliefs offered an explanation for her nonconscious actions in a way that medicine/psychiatry could not, and that this should be taken seriously. But I never gave oral evidence in the witness box. The next day, after Rachel’s grueling cross-examination, the prosecution and defense barristers continued to argue with each other and the judge about what admissible evidence an anthropologist could offer that the prosecution had not already acknowledged. The prosecution barrister stated that he did not understand what an anthropologist could contribute beyond confirming what he was already happy to agree with, that beliefs about witchcraft and spirit possession are “deeply held” across Sub-Saharan Africa. In this view anthropology was limited to a kind of ethnology—the study of “other” peoples’ “cultural” beliefs, and had nothing to contribute to a discussion of dissociative states or the possibility of noninsane automatism. The defense barrister argued there was much more an anthropologist could offer, but could not articulate this in way the prosecution did not object to. Perhaps the prosecution was employing a delaying tactic, knowing I had to leave that afternoon having spent the week in court. Eventually I ran for my train while the court went for lunch, but my anthropological colleague did give evidence later that day, holding her own under cross-examination.26 But beyond stating that “the beliefs held by Rachel and her mother seemed characteristic of people in Zimbabwe,” she was not asked any questions “that needed detailed ethnographic information” from my report or elsewhere (pers. comm.). Like all witnesses, the oral evidence of experts is limited by the questions they are asked in court. Therefore it is very unlikely that had I taken the stand an opportunity would have arisen to offer much detailed ethnographic and historical context to Rachel’s beliefs about witchcraft and spirit possession, or contribute to the debate about noninsane automatism, let alone to argue for the possibility of other possibilities.27

The verdict: “A fudge”

Nevertheless my colleague must have done well because that evening the defense barrister phoned to say the judge was considering allowing the jury to weigh up the possibility of noninsane automatism and therefore a not guilty verdict. This was a positive moment to end the first week on. But by the end of the following Monday, the judge decided that “a nightmare would in law, be an internal cause of a dissociative state” (Defense solicitor, pers. comm.). This “had a major consequence on how best to proceed with the defense” as it “removed the option of a not guilty verdict (which would be ‘non-insane automatism’) meaning an external cause of any dissociative state could not be a decision open” to the jury (Defense solicitor, pers. comm.). In effect, the jury’s options were now limited to a guilty or a not guilty by reason of insanity verdict, neither of which the defense had advocated for. “The rug was pulled from under my feet,” the defense barrister later told me, and this required a change of tactic. The barristers and judge agreed that two lesser counts of “wounding with intent to cause serious harm” and “malicious wounding (intent to cause some harm)” be added to the indictment, to which Rachel pleaded not guilty. The judge “felt this course was fair to both sides” (Defense solicitor, pers. comm.).

With all the evidence already submitted, the barristers delivered their closing speeches on the Tuesday of week two, and the judge began his summing up, finishing on the Wednesday when the jury retired around midday to consider their verdict. I was not in court for the second week of the trial, but the defense barrister told me later that because his defense of noninsane automatism had been denied, he spent his two-and-half-hour closing speech trying to persuade the jury to disagree and produce a hung jury. By any reckoning this was an unusual thing for a defense barrister to advocate, but this was an unusual trial. That the jury spent nearly three days considering their verdict, and then it was a 10-2 majority decision, and not unanimous, suggests he may have come close to succeeding. On Friday, the tenth day of the trial, the jury reached a verdict of not guilty of attempted murder, not guilty of wounding with intent to cause serious harm, but guilty of malicious wounding (intent to cause some harm). Rachel was bailed pending further “presentencing reports,” and allowed to have contact with her mother in the presence of other family members. Rachel was very upset. Her reaction, the defense barrister later told me, was “to collapse into her mother’s arms repeating for twenty minutes that she couldn’t remember.” It was first time they had direct contact for almost two years.

The defense barrister later described the verdict as a “total fudge.” Rachel was convicted on the least serious charge the jury had to choose from. It was a charge neither the defense nor prosecution had argued for. Yet the final result of malicious wounding was perhaps the best possible result. This became clearer with the sentencing. Four months after the verdict, following further psychiatric tests and various risk assessments, Rachel Moyo was sentenced to twelve months imprisonment in a Young Offenders Institution, suspended for eighteen months, with an eighteen-month supervision order and 120 hours of community work. She therefore avoided a custodial sentence, and was able to return to live with her mother and family after almost two years. Despite his direction to them that disallowed a defense of noninsane automatism, in passing sentence, the judge stated that the jury had rejected the argument that Rachel had been in a dissociative state by convicting her of malicious wounding.28 Yet it seems the anthropological evidence so marginalized during the trial itself, did have some efficacy in the judge’s deliberations. Echoing the more conventional kind of cultural defense I have discussed, which was never actually made by the defense team during the trial, the judge stated that the probability was that Rachel had “attacked her mother because she believed that was what the spirits were telling her to do” or because her beliefs made her “think she was possessed.” For the judge then, Rachel had acted consciously but in the belief that she was possessed, or that her grandmother’s spirit wanted her to do it. In ordering Rachel to attend supervisory sessions, he further pressed her to try to “understand her beliefs so she could fight any urges that tell her to commit any crimes in the future.” Therefore, in the final count, the judge revealed himself as sympathetic both to Rachel and her mother’s plight—hoping they “can come to terms with what has happened” and wishing them well for the future—and to the cultural context of the case, acknowledging that “in large parts of Africa, including Zimbabwe” beliefs about spirit possession “are taken for granted” and “regarded as no more irrational than the belief in the existence of god.” Yet he had also discounted the possibility of other possibilities and the limits of knowledge, and his comments ultimately rehearsed Evans-Pritchard’s (1937) assertion that Azande beliefs about sorcery are “rational but wrong.”


It was my first experience in a criminal trial, and I was surprised how much the proceedings matched my expectations. There was a real sense of drama, of barristers performing to the court, especially the jury, within constraints of its hierarchies and rules, closely disciplined by an attentive judge. The judge remonstrated with the barristers but also witnesses, asking them for clarifications, or to speak up, even chastising one police officer for interrupting him. There were stronger and weaker performers, and as the barristers questioned witnesses, they established personas not only for the witnesses but also for themselves. The prosecuting barrister was demure, serious, with an organized air about him. The defense barrister by contrast was charming, witty, even maverick, playing to the emotions of the witnesses and the jury. There was plenty of legal argument, the barristers and judge playing a complex, invisible game of chess with each other by rules opaque to the rest of the court, despite the judge’s frequent legal explanations to the jury as they were escorted in and out of the court. Each day the court finished fifteen minutes early to enable one member of the jury to attend afternoon prayers. Timing was important for the barristers too, who might spend hours going through tedious minutia of details with witnesses, to ensure the right words and emotions filled the court at the right moment, such as when Rachel burst into tears at the very end of the fourth day, subduing the room with an affective scene that everyone took home with them.

There were moments of misapprehension too, when questions did not provoke the right answers and emotional responses, such as Rachel’s mother’s efforts to describe how Rachel appeared that night.29 Or when Rachel was asked about the daily relevance of her cultural beliefs about spirits and witchcraft, to which she had answered “they don’t impact on my life,” confounding the defense barrister’s line of questioning and giving opportunity for the prosecution’s grueling cross-examination. Some witnesses withstand the verbal manipulations of barristers, others less so. Rachel was confident and affective, her mother dignified, but her uncle emanated a less sure figure. Expert witnesses too can feel the pressure of inquisition; some command the court with authoritative presence (such as the defense psychiatrist), others less so (like the prosecution psychiatrist). Even the most accomplished can be maneuvered into saying something not anticipated, such as when the defense psychiatrist conceded a nightmare may have triggered a dissociative state. With this insight I was later grateful I did not take the stand, doubting I would have been very good under the prosecution’s cross-examination.

All of this points to the social, cultural, and performative aspects of how courts construct particular kinds of certainty, which have frequently been discussed in the wider (and older) anthropological literature on juridical processes and legal contests, which I have not had room to engage with here (see, for example, Benda-Beckmann 1981; Canter 1978; Colson 1974; Comaroff and Roberts 1981; Moore 1973; Von Benda-Beckmann and Von Benda-Beckmann 2009; Werbner 1977, 1982). Yet during the trial, and immediately afterwards, I was most struck by how both the defendant and her mothers’ explanations for what had happened, and the evidence of the anthropologists appeared so marginal to the proceedings. Not only the prosecution (who were always likely to object) but also the judge and the sympathetic defense barrister struggled to know what to do with the anthropological evidence in my report, just I had struggled in writing it. While I worried about the exoticization of the so-called African occult and the problematic assumptions embedded in conventional forms of cultural defense, and felt compelled to engage with anthropology’s long discussion of the epistemological/ontological problems raised by witchcraft and spirit possession, the trial really turned on the debates among the psychiatrists, barristers, and judge about internal and external “triggers” for dissociative states and the legal possibility of noninsane automatism. Rachel’s mother’s plea to the court on the second day of the trial for someone “who has experienced spirit possession, who might understand this” was similarly irrelevant to the proceedings. Rachel and her mother’s views never seemed to be taken very seriously. The jury never heard a very full account of their explanations, or the complex process by which these were arrived at, nor of what anthropology might make of their accounts. They were acknowledged and then sidelined; relegated to the irrelevant. This marginalization of their “cultural narratives” (as the defense psychiatrist put it) seemed to resonate sharply with anthropology’s struggle to be taken seriously in the court.30

This case therefore illustrates the difficulties anthropologists can face as expert witnesses, particularly (and surprisingly so) in cases revolving around the kind of phenomena—like witchcraft, magic, and spirit possession—that our discipline has long debated and obsessed about. It is a problem of making anthropological knowledge production commensurable and useful to the functions and expectations of the court, without falling back on analytical or theoretical assumptions that no longer hold disciplinary sway. We could say, following West (2007:69), that the “sorcery” of the court means that in the effort to construct a “transcendental position” in order to enforce certainty upon an otherwise “inescapably inchoate world,” certain forms of expertise and knowledge, like medicine and psychiatry, are given a credence that anthropology is not. And of course when anthropologists find their own “transcendental maneuvers” to “reorder the world” (our own kind of sorcery) constrained and transcended in socio-political contexts in which our contribution is at best ambiguous, this can be very disconcerting.

Yet however much our discipline’s ethical and analytical stance might sometimes appear an anathema to how UK courts function, the dynamics I have been describing were also a result of the particular peculiarities of this case. It is interesting to speculate, for example, whether if a more conventional cultural defence had been presented—perhaps with the defendent admitting conscious intentionality in attacking her mother but appealing to some cultural imperative that motivated her—then Rachel and her mothers’ explanations, and what anthropologists had to say about them, would have got a better hearing during the trial. Rachel Moyo’s case was different and unusual exactly because there was such considerable evidence that she was not in control of her own actions during the events in question. It was the lack of any known medical, psychiatric, or sleep-disorder related causes for her abnormal state of consciousness, that raised the very possibility of noninsane automatism, or in my framing, the “other possibilities” that might be involved. Although some anthropologists are likely to disagree, I maintain that recognizing the limits of knowledge, the prevalence of a profound uncertainty, and acknowledging the possibility of other possibilities does not have to equate with the reification of a radical alterity or relativism that might threaten to undermine the very function and legitimacy of any court or other mechanism for establishing “truth” or “justice,” however contingent. I think we can allow for the “ontological self-determination of the world’s peoples” (Viveiros de Castro 2003:18), and still have legitimate mechanisms for establishing individual responsibility and accountability (as indeed Geertz might have argued, see Shweder 2010a, 2010b). The difficulty lies in how to make such arguments commensurable, useful, and relevant for a criminal court. How can (if indeed they should) anthropology’s theoretical debates, particularly those to do with its much contested “ontological turn” (Killick 2014; Pedersen 2012; Laidlaw 2012; Fontein 2011; Course 2010; Carrithers et al. 2010), be usefully applied in work we do as expert witnesses in legal contexts?

Three years on I am less convinced about the extent to which anthropology was ineffectual in the courtroom than I was in the immediate wake of the trial. This is partly due to the verdict and particularly the sentencing, which were, as Terence Ranger commented soon afterward—mindful no doubt of the dangers of reified, aggregated images of the so-called African occult—“as merciful as could have been expected and better really than a verdict which implied that if you were possessed by a witch spirit you could kill!” (pers. comm.). Maybe the fudge of the verdict was the best of all alternatives. Certainly the sentencing was very sympathetic to Rachel and her mother. She was allowed home, her custodial sentence was suspended, and she could begin to rebuild her life, her relationship with her family, and most of all with her mother. My arguments about profound uncertainty, indeterminacy, and the possibility of other possibilities may have gained little traction, and often courtroom discussions did fall back on essentialized notions of culture that would make any anthropologist squirm. Yet the psychiatrists did engage, to some degree, with the anthropological evidence in their oral testimonies, which did seem to influence their courtroom discussions, if only in a small way. It also had an effect on the (ultimately flawed) defense case. In the end the anthropological evidence presented to the court may also have had efficacy in the judge’s deliberations, even if his statements did fudge something akin to the kind of cultural defense anthropologists are often at pains to avoid. In acknowledging that “in large parts of Africa” beliefs about spirit possession “are regarded as no more irrational than the belief in the existence of god,” and in suggesting that Rachel attacked her mother because she thought she was possessed, he inadvertently invoked a legacy that continues to loom large in anthropology, even if more recent debates have moved decidedly on. Evans-Pritchard might have been pleased, and perhaps, in the end, so should we.


Ashforth, Adam. 2005. Witchcraft, violence and democracy in South Africa. Chicago: University of Chicago Press.

Bauman, Zygmunt. 1996. “From pilgrim to tourist: Or a short history of identity.” In Questions of cultural identity, edited by Stuart Hall and Paul du Gay, 18–36. London: Sage.

Bernault, Florence. 2013. “Carnal technologies and the double life of the body in Gabon.” Critical African Studies 5 (3): 175–94.

Bloor, David. 1976. Knowledge and social imagery. London: Routledge.

Canter, Richard S. 1978. “Dispute settlement and dispute processing in Zambia: Individual choice versus societal constraints.” In The disputing process: Laws in ten societies, edited by Laura Nader and Harry F. Todd, 247–80. New York: Colombia University Press.

Carrithers, Michael, Mattia Candea, Karen Sykes, Martin Holbraad, and Soumhya Venkatesan. 2010. “Ontology is just another word for culture.” Critique of Anthropology 30 (2): 152–200.

Colson, Elisabeth. 1974. Tradition and contract: The problem of order. New York: Aldine.

Comaroff, John L., and Simon Roberts. 1981. Rules and processes: The cultural logic of dispute in an African context. Chicago: University of Chicago Press.

Conley, John M., and William M. O’Barr. 1990. Rules versus relationships: The ethnography of legal discourse. Chicago: University of Chicago Press.

———. 1993. “Legal anthropology comes home: A brief history of the ethnographic study of law.” Loyola Law Review 27: 41–64.

———. 1998. Just words: Law, language and power. Chicago: University of Chicago Press.

Course, Magnus. 2010. “Of words and fog: Linguistic relativity and Amerindian Ontology.” Anthropological Theory 10: 247–63.

Department for Education (DfE). 2007. “Safeguarding children from abuse linked to a belief in spirit possession.” Accessed January 28, 2014. http://webarchive.nationalarchives. http://webarchive.nationalarchives.gov.uk/20130401151715/https://www.education.gov.uk/publications/eOrderingDownload/DFES-00465-2007.pdf

———. 2012. “National action plan to tackle child abuse linked to faith or belief.” Accessed January 28, 2014. https://www.gov.uk/government/publications/national-action-plan-to-tackle-child-abuse-linked-to-faith-or-belief.

Espirito Santo, Diana, Katerina Kerestetzi, and Anastasios Panagiotopoulos. 2013. “Human substances and ontological transformations in the African-inspired ritual complex of Palo Monte in Cuba.” Critical African Studies 5 (3): 195–219.

Evans-Pritchard, Edward. 1937. Witchcraft, oracles and magic among the Azande. Oxford: Oxford University Press.

Fontein, Joost. 2011. “Graves, ruins and belonging: Towards an anthropology of proximity.” Journal of the Royal Anthropological Institute 17: 706–27.

Fontein, Joost, and John Harries. 2013. “The vitality and efficacy of human substances.” Critical African Studies 5(3): 115–26.

Geertz, Clifford. 1983. “Local knowledge: Fact and law in comparative perspective.” In Local knowledge: Further essays in interpretive anthropology, edited by Clifford Geertz, 167–234. New York: Basic Books.

Geschiere, Peter. 1997. The modernity of witchcraft. Charlottesville: University of Virginia Press.

———. 2009. The perils of belonging. Chicago: University of Chicago Press.

Gluckman, Max.1963. “The reasonable man in Barotse Law.” In Order and rebellion in tribal Africa, 178–206. London: Cohen & West.

———. 1965. The ideas in Barotse jurisprudence. New Haven, CT: Yale University Press.

———. 1967. The judicial process among the Barotse of Northern Rhodesia. Manchester: Manchester University Press.

Good, Anthony. 2006. Anthropology and expertise in the asylum courts. London: Routledge-Cavendish.

———. 2008. “Cultural evidence in courts of law.” Journal of the Royal Anthropological Institute 14, Issue supplement s1: s47–s60.

Henare, Amiria, Martin Holbraad, and Sari Wastell. 2007. “Introduction: Thinking through things.” In Thinking through things: Theorising artefacts ethnographically, edited by Amiria Henare, Martin Holbraad, and Sari Wastell, 1–31. London: Routledge.

Hoskins, Richard. 2012. The boy in the river. London: Pan Macmillan.

Inksater, Kimberly. 2010. “Transformative juricultural pluralism: Indigenous justice systems in Latin America and international human rights.” Journal of Legal Pluralism 60: 105–43.

Jackson, Michael. 1989. Paths towards a clearing: Radical empiricism and ethnographic enquiry. Bloomington: Indiana University Press.

Killick, Evan. 2014. “Whose truth is it anyway?” Anthropology of this Century 9.

Kuper, Adam. 2005. “Clueless.” London Review of Books 27: 8.

La Fontaine, Jean S. 1994. “The extent and nature of organised ritual abuse: Research Findings.” Report commissioned by Department of Health, published by HMSO.

———.1998. Speak of the devil. Cambridge: Cambridge University Press.

———. 2009. The devil’s children. London: Ashgate.

Laidlaw, James. 2012. “Ontologically challenged.” Anthropology of this Century 4.

Lett, James. 1991. “Interpretive anthropology, metaphysics, and the paranormal.” Journal of Anthropological Research 47: 305–29.

Marwick, Max G. 1965. Sorcery in its social setting. Manchester: Manchester University Press.

Merry, Sally E. 1988. “Legal pluralism.” Law and Society Review 22 (5): 869–96.

Meyer, Birgit. 2009. “Response to Ter Haar and Ellis.” Africa 79: 3.

Moore, Sally F. 1973. “Law and Social Change: The semi-autonomous field as an appropriate object of study.” Law and Society Review 7 (4): 719–46.

Mosse, David. 2006. “Anti-social anthropology? Objectivity, objection, and the ethnography of public policy and professional communities.” Journal of the Royal Anthropological Institute 12: 935–56.

Niehaus, Isak. 1997. “‘A witch has no horn’: The subjective reality of witchcraft in the South African Lowveld.” African Studies 56: 2.

———. 2013. Witchcraft and a life in the new South Africa. Cambridge: Cambridge University Press.

Pedersen, Morten. 2012. “Common nonsense: A review of certain recent reviews of ‘the’ ontological turn.” Anthropology of this Century 5.

Ranger, Terence. 2007. “Scotland Yard in the Bush: Medicine murders, child witches, and the construction of the occult: A literature review.” Africa 77: 2.

Renteln, Alison D. 2005. The cultural defense. Oxford: Oxford University Press.

Reynolds, Pamela. 1992. “Dreams and the constitution of self among the Zezuru.” In Dreaming, religion, and society in Africa, edited by Marian C. Jedrej and Rosaland Shaw, 21–35. Leiden: Brill.

Rutherford, Blair. 1999. “To find an African witch.” Critique of Anthropology 19 (1): 89–109.

Shapin, Steven. 1996. The scientific revolution. Chicago: University of Chicago Press.

Stobart, Eleanor. 2006. “Child abuse linked to accusations of ‘possession’ and ‘Witchcraft.’” Research report No. 750, Department for Education and Skills.

Stoller, Paul, and Cheryl Olkes. 1987. In sorcery’s shadow: A memoir of apprenticeship among the Songhay of Niger. Chicago: University of Chicago Press.

Shweder, Richard A. 2010a. “Geertz’s challenge: Is it possible to be a robust cultural pluralist and a dedicated political liberal at the same time?” In Law without nations, edited by Austin Sarat, 185–231. Stanford, CA: Stanford University Press.

———. 2010b. “Clifford Geertz: Obituary.” Proceedings of the American Philosophical Society 154 (1).

Ter Haar, Gerrie, and Stephen Ellis. 2009. “The occult does not exist: A response to Terence Ranger.” Africa 79 (3): 399–412.

Thomas, Robert. 2011. Administrative justice and asylum appeals. Oxford: Hart Publishing.

Viveiros de Castro, Eduardo. 2003. “And.” Manchester Papers in Social Anthropology 7.

Von Benda-Beckmann, Keebet. 1981. “Forum shopping and shopping forums, dispute processing in a Minangkabau village in West Sumatra.” Journal of Legal Pluralism 19: 117–62.

Von Benda-Beckmann, Franz, and Keebet Von Benda-Beckmann, eds. 2009. Spatializing law. London: Ashgate.

Webber, Jeremy. 2006. “Legal pluralism and human agency.” Osgoode Hall Law Journal 44 (1): 167–98.

Werbner, Richard. 1977. “Small man politics and the rule of law: Centre-periphery relations in east-central Botswana.” Journal of African Law 21: 24–39.

———. 1982. “The quasi-judicial and the absurd: Remaking land law in north-eastern Botswana.” In Land reform in the making: Tradition, public policy, and ideology in Botswana, edited by Richard Werbner, 131–50. London: Collings.

West, Harry. 2005. Kupilikula: Governance and the invisible realm in Mozambique. Chicago: Chicago University Press.

———. 2007. Ethnographic sorcery. Chicago: Chicago University Press.

Willis, Roy. 1996. “Magic” and “Witchcraft and sorcery.” In Encyclopedia of social and cultural anthropology, edited by Alan Barnard, and Jonathan Spencer. London: Routledge.

———. 1999. Some spirits heal, others only dance: A journey into human selfhood in an African village. Oxford: Berg.

Winch, Peter. 1958. The idea of a social science and its relation to philosophy. New York: Humanities Press.

———. 1964. “Understanding a primitive society.” In Rationality, edited by Bryan Wilson, 78–111. Oxford: Blackwell.

“Elle semblait être dans une espèce de transe”: l’anthropologie et l’inconnaissable dans un procès criminel

Résumé : Ceci est un récit personnel d’un récent procès criminel au Royaume-Uni auquel a participé l’auteur en tant qu’expert. Le procès impliquait une jeune femme du Zimbabwe qui avait agressé sa mère au couteau pendant qu’elle était (de l’aveu de sa mère, de sa famille et de la femme elle même) possédée par un esprit maléfique, après avoir été ensorcelée par un autre membre de la famille. Une preuve supplémentaire de son état d’aliénation était apportée par le rapport de la police, qui précisait qu’elle était « en état de transe » lors de son arrestation. Bien qu’elle fût soumise à une série d’examens médicaux et psychiatriques, on ne trouva aucune explication neurologique, médicale, psychiatrique ou reliée à des troubles de sommeil pour l’épisode de possession. L’article détaille les difficultés qu’a rencontrées l’auteur lors de cette expertise anthropologique, qui visait à dépasser les formes habituelles d’une « défense par la culture » pour mettre en avant les limites de la connaissance et la possibilité d’autres possibilités. L’article s’appuie sur l’idée de « sorcellerie ethnographique » proposée par Harry West pour montrer comment l’expertise anthropologique, tout en étant sujette aux contraintes de tribunaux qui cherchent à construire leur propre certitude, garde des formes d’efficacité non intentionnelle.

Joost FONTEIN teaches social anthropology at Edinburgh. In August 2014 he is taking up an appointment as Director of the British Institute in Eastern Africa. He has done field work in Zimbabwe since the late 1990s, and is editor of the Journal of Southern African Studies, cofounder of Critical African Studies, and a founding member of the “Bones Collective” research group.

Joost Fontein
Social Anthropology
University of Edinburgh
Cmb, 15a George Square
Edinburgh EH8 9LD


1. According to West, Kupilikula literally means “to invert,” “to overturn,” “to negate,” “to annul,” and “to undo.” He describes how Muedan sorcerors transcend the everyday by “rendering themselves invisible,” “producing and inhabiting an invisible realm from which they gain powerful perspective on the visible” enabling them to reorder it in such way as to “feed their insatiable appetites.” But they “do not go unchallenged,” and different figures of authority, as “sorcerors of construction,” “transcend not only the world inhabited by ordinary Muedans but also that of “sorcerors of ruin,” fixing the latter in their gaze … unmaking sorcerors’” acts, and remaking the world in accordance with their own visions of a world re-ordered” (2007: x).

2. As one reviewer noted, for some anthropologists “West’s insistence on seeing ethnography as sorcery is unwise” because while anthropologists do “re-order the world,” they “do not cause misfortune” or “kill as sorcerers are said to do.”

3. This case was discussed in detail on Zimbabwean websites, where a range of opinions were expressed, some hostile and some sympathetic to either or both Rachel and her mother. I cannot cite them because of the need to ensure anonymity. Apart from illustrating the stigma often associated with witchcraft afflictions, one resounding theme of many comments was that this case involved aspects of “African culture,” chivanhu (loosely “tradition” in Shona), ngemadlozi/masvikiro (ancestors/spirit mediums in Sindebele/Shona) or mhepo (“wind” or spirits in Shona) that “white people,” British courts, and psychiatrists cannot understand.

4. See also Canter 1978; Colson 1974; Comaroff and Roberts 1981; Moore 1973; Von Benda-Beckmann 1981; Von Benda-Beckmann and Von Benda-Beckmann 2009; Werbner 1977, 1982. I am grateful to the anonymous reviewers for pointing me to this useful wider literature, and regret I do not have more space to do justice to it.

5. I was first approached by the defense team during the summer of 2009 on the basis of my country expertise and anthropological knowledge of Zimbabwe. This was the first time I have acted in this capacity for a criminal trial, having previously been a country expert for Zimbabwean asylum cases. In early 2010 I held an informal interview with Rachel Moyo in Edinburgh. I then drafted my report, which I revised later that year and submitted to the court as part of the “bundle of papers” making up the defense case. Both barristers, their expert witnesses, and the judge had access to it before and throughout the trial. During the long period before the trial, a more senior barrister was drawn into the defense team, who subsequently sought, a few weeks before the trial in early 2011, an additonal, second anthropological expert witness, who I cannot name because that would jeopardize the anonymity of the case. This senior colleague was drawn into the case because of her public standing, seniority, and previous experience in similar trials, but she, although an Africanist, has no particular knowledge of Zimbabwe, and was therefore requested to provide evidence in a short report to supplement my much more detailed fifty-page report. Her report was based on a brief meeting with Rachel some weeks before the trial, and we communicated about each other’s reports before, during, and after the trial. She also commented closely on early drafts of this paper and I remain grateful for the useful and critical comments she provided.

6. Such cases are hard to quantify, and yet it is clear new public concerns have been raised in recent years, that differ from earlier “satanic abuse” cases (La Fontaine 1994, 1998). Current concerns are focused on the risks of child abuse linked to accusations of “possession” and “witchcraft” (La Fontaine 2009). After some high profile cases, these concerns have reached such proportions that government reports have been commissioned (Stobart 2006; DfE 2007), a National Action Plan developed (DfE 2012), and various police task forces (London Metropolitan Police’s Project Violet, see http://content.met.police.uk/%20Article/Abuse-linked-with-spirit-possession/1400010000897/%201400010000897, accessed January 28, 2014) and nongovernment organizations established to focus on the problem (African Unite Against Child Abuse [AFRUCA—see http://www.afruca.org, accessed January 28, 2014] and the Churches Child Protection Advisory Service [CCPAS—see http://www.ccpas.co.uk, accessed January 28, 2014]). Partly as a result of what I discuss in this article, in April 2013 I cohosted a closed workshop at the LSE entitled “Witchcraft, spirit possession, and anthropological expertise in legal contexts” bringing togther anthropologists, lawyers, social workers, and police to begin a conversation to promote greater mutual understanding and better use of anthropological evidence in such cases.

7. The “Torso in the Thames” case is perhaps the most well known UK example of a suspected “muti murder,” although the investigation never proceeded to trial (Kuper 2005; Ranger 2007). So-called child witch exorcisms have recived more coverage recently, particularly after the horrific killing of Kristy Bamu (“Witchcraft murder: Couple guilty of Kristy Bamu killing” BBC News March 1, 2012).

8. Of particular concern has been the emergence of dubious “expert witnesses” who police have sometimes relied upon. Richard Hoskins is one example who has appeared in documentaries (e.g., Witch Child, a 60-minute documentary by October Films, broadcast on BBC 2 on April 6, 2006, http://www.octoberfilms.co.uk/recentproductions.php?production=184, accessed January 21, 2014) and recently wrote a book on the “Torso in the Thames” case (Hoskins 2012). Some academics too have stood accused of reinforcing such simplistic and alarmist impulses by implying that there is something peculiarly “occult” and dangerous about “African societies” (see Ranger 2007; Meyer 2009; Ter Haar & Ellis 2009).

9. I experienced this when I initially encountered the questions I was asked to address in the report requested by the defense team. These included: “Is there any empirical or anecdotal evidence that belief in witchcraft or black magic has caused people to awaken having carried out an act that they do not remember?”

10. So I outlined how beliefs about witchcraft and possession are not only widespread across Zimbabwe, but are also diverse, contested, and can vary according to regional, cultural, and religious factors and affiliations. The precise content of a person’s, a family’s, or a group’s beliefs may vary locally and regionally, and according to religious affiliation, church membership, education, and even class, and so on. Often such beliefs appear under a generic term such as “muti,” as Rachel herself named it, although this word itself has many connotations ranging from medicine to poison, and from “magic” to witchcraft “substance.”

11. During the interview we talked about a range of possible cultural explanations for the events, including posession by a benevolent ancestral spirit, or by a malevolent one, or by animal or other nonhuman spirit, a “demon,” or as the result of a deliberate act of witchcraft by someone.

12. Which mean that determining the presence of witchcraft, and identifying those responsible can involve a multiplicity of factors and contested processes, in which anthropologists too can be implicated (Rutherford 1999).

13. I am grateful to the comments of two anonymous reviewers that helped me to articulate this point, even as its rendering here reflects my interpretation.

14. I have borrowed this apt expression from Anthony Good.

15. So that if a person is identified as being guilty of being a witch or doing witchcraft, but it is recognized that this is an involuntary/unintentional act (through being possessed or bewitched), then this usually provokes a very different response than where someone has deliberately sought to acquire/use witchcraft for their own (usually malicious) purposes.

16. “Witchcraft recognised, laws tightened” April 22, 2006 Andnetwork.com, cited from http://www.zimbabwesituation.com/apr23_2006.html#Z4.

17. This shows how there was a recognition by the court that at the root of the issue were family disputes/tensions, which threatened to worsen dramatically. This led to the conclusion that the matter should be dealt with by those local (traditional) authorities who normally deal with such family disputes. So the accused was given a suspended sentence by the Harare magistrate and ordered to get herself “cleansed” and her family disputes resolved. See “Woman ‘flew’ in basket on witchcraft mission,” Newzimbabwe. com May 28, 2009; “Naked basket flight woman a witch—expert,” Newzimbabwe.com, May 29, 2009; “Witchcraft case takes new twist,” The Herald May 29, 2009; “Witchcraft case—Woman gets suspended sentence,” The Herald, June 5, 2009.

18. Although I was careful to circulate my draft report to colleagues for informal review before I submitted it, mainly because I was concerned about overstating this point, and wanted to ensure that my account of anthropology’s long critical engagement with “magic,” “spirit possession,” and “witchcraft” was representative.

19. As Jean La Fontaine indeed commented (pers. comm.).

20. As one reviewer commented; but I would argue ultimately turns on a problematic distinction between “subjective” and “objective” truth that needs to be avoided.

21. Or indeed any immigration tribunal that, as Thomas (2011) has argued, often depend on the perceived credibility of the claimant.

22. There is a legal concept of “noninsane automatism,” which has sometimes gained traction in court cases where, for example, sleepwalking is involved. See R v. Burgess 93 Cr.App.R.41, CAM cited on the Crown Prosecution Service website: www.cps.gov.uk/legal/d_to_g/defences_-_sleepwalking_as_a_defence_in_sexual_offence_cases/

23. At that stage he was not even aware I had interviewed Rachel.

24. This was heightened by the agreement between the barristers and judge that their two experts’ evidence be heard “back to back,” at the end of the prosecution case, rather than according to the normal sequence whereby defense witnesses are heard after the prosecution has closed its case. There was no such straightforward agreement about when and what anthropological evidence would be admissible.

25. A “cultural explanation” that might have had traction among some Zimbabweans would be that the dream itself was a manifestation of witchcraft, as it is commonly held that spirits and witches do communicate and act through dreams (Reynolds 1992). In our interview Rachel suggested the appearance of her grandmother and her paternal aunt in her dream was a deliberate effort by her maternal aunt to disguise the true source of the sorcery. But Rachel and her mother’s cultural explanations were never discussed in such great detail in the courtroom.

26. See note 3 above for a discussion of our roles as expert witnesses in this case. Although my involvement was deeper and longer than hers, with hindsight I soon felt that it was probably better only she ended up taking the stand, with her far greater experience of giving oral evidence and especially cross-examination. She later told me that in cross-examination, the prosecuting barrister tried to suggest that because she herself had spent many of her early years in Africa, she perhaps was partly “African” (which she denied); a line of questioning that she felt suggested the prosecution might have intended to argue by analogy that because Rachel had spent nearly a decade in the United Kingdom, she was acculturated to British society, and thereby to imply that any cultural defence could not apply. But this argument was never actually made by the prosecution barrister.

27. As Anthony Good commented (pers. comm.), this marginalization of anthropological evidence may have had something to do with the fact that this was a jury trial. As he discusses in relation to the case of R v. Turner (Good 2006: 144–45), when judges are the decision makers, rather than juries, anthropological evidence is often given a greater hearing.

28. After the sentencing the defense barrister told me an appeal was being considered, based on the judge’s decision to disallow noninsane automatism and therefore a not guilty verdict, but this was never pursued. Later he told me he was worried “the Court of Appeal may have clipped the anthropologist’s wings even further” and “an anticipated ruling that an ancestral spirit is no more than an internal demon at work, i.e., a behavioral deficiency arising from within the ‘perpetrator’ would hardly help the Rachels of this world” (pers. comm., April 2013).

29. Or the confusion that surrounded the muprofiti, who of course should not need to be told the reason for their consultation, they should already know it, as Rachel’s mother was trying explain before the fire alarm went off.

30. Perhaps the two were connected because, as one reviewer suggested, the “marginality of anthropological testimony” may reflect an assumption that anthropologists “deal with ‘primitives’”; or alternatively, that “we deal with the relativity of practices, whereas legal bureaucracies deal with absolutes.” It may also, however, reflect that this was a jury trial, because, as Good suggests (see footnote 27), in trials where judges make verdicts anthropologists are sometimes taken more seriously.