Worlds of Witchcraft

Witchcraft-based asylum claims raise important questions of how to define witchcraft and how to articulate its terms. These questions are not new; anthropologists studying Africa have grappled with them since the early twentieth century. In his foundational text, Witchcraft, Oracles, and Magic among the Azande, E.E. Evans-Pritchard parsed the supernatural situation of the Zande people in interwar Sudan, defining “witchcraft” as the use of inherited, embodied, supernatural power to do malevolence, “sorcery” as the use of paraphernalia to pursue the same aims, and “magic” as benevolent power used to remedy witchcraft and sorcery (Reference Evans-Pritchard1937:177). He argued that witchcraft spoke to the logics of Zande thought and action, rather than to Zande primitivity. The “concept of witchcraft,” Evans-Pritchard wrote, provided Zande people “with a natural philosophy by which relations between men and unfortunate events are explained, and a ready and stereotyped means of reacting to such events” (64). This work set the stage for analyses in the 1950s and 1960s that aimed to elucidate further the purposes of witchcraft within a given society and assigned witchcraft roles in the management of social relations. Witchcraft provided mechanisms through which tensions created by intimate social conflicts and broad-scale social changes could be managed and social order subsequently restored.

Contemporary work has argued that witchcraft is modern. It has explored how witchcraft interacts with sociopolitical and socioeconomic changes wrought by globalization, neoliberal capitalism, and development. Henrietta Moore and Todd Sanders neatly sum this up: “Far from withering away, witchcraft has purportedly increased in post-colonial Africa. For this reason and others, anthropologists have sought to problematize Africa’s hybrid worlds that contain both tradition and modernity” (Reference Moore, Sanders, Moore and Sanders2001:7). Within these “hybrid worlds,” accumulation has emerged as a key concern of scholars and their subjects alike. Writing about a Cameroonian brand of witchcraft, Peter Geschiere characterizes the socioeconomic landscape of the continent since the 1980s and the place of witchcraft in negotiating it. “Djambe,” he writes, offers “a seductive discourse to address the riddles of modern development; the rapid emergence of shocking new inequalities, the enigmatic enrichment of a happy few, and the ongoing poverty of the many” (Reference Geschiere1997:55). As James Ferguson notes, witchcraft can be conceived of “as the fearsome power that makes it possible for exploiters to exploit with impunity” and it can be the “sanction that checks accumulation…” (2006:71). Whether used to accumulate resources illegitimately or to re-equilibrate inequalities, witchcraft speaks to the intertwined questions: “Why me?” and “Why you, but not me?” These questions frame witchcraft-based asylum narratives.

Diverse sources demonstrate that in recent decades witchcraft has become a much more open subject, a matter of personal discussions and political debates (Roxburgh Reference Roxburgh2019). While a representative from the Office of the United Nations High Commissioner for Human Rights recently stated that there are “no reliable statistics” on how many alleged witches are killed annually, anecdotal evidence generated by states, the press, and NGOs on the continent indicates that violence against alleged witches is frequent and widespread.^{Footnote 1} Ultimately, what matters is that people across Africa are convinced that witchcraft and related harms are on the upsurge, and these convictions hold concrete consequences for everyday life. One of these results is the introduction of witchcraft into the global arena of asylum. At the same time that the (dangerous) experience of “living in a world with witches” (Ashforth Reference Ashforth, Moore and Sanders2001:206) has become much more openly articulated, the expansion of PSG along cultural lines has created space for new categories of asylum seekers to emerge: “perceived witchcraft practitioners” and “victims of witchcraft.”

Researching Witchcraft in Asylum Settings

Publicly available decisions in witchcraft-based asylum cases, rendered by immigration bodies in Canada and Australia, form the core of this article. With Evans-Pritchard’s categories and vernacular (supernatural) lexicons in mind, I used keywords—“witchcraft,” “witch,” “sorcery,” “sorcerer,” “magic,” and even “juju”—as I mined major legal databases (LexisNexis, Quicklaw, and the Global Legal Information Institute) together with national databases. I collected first-tier and appellate decisions in more than forty cases adjudicated since 2000 in which African asylum seekers offered witchcraft-based persecution as grounds for refugee protection. A significant proportion of these asylum seekers, like the bulk of asylum seekers from the continent generally, were Nigerian. Cases where the asylum seeker claimed to have been accused of witchcraft outnumbered those in which they claimed to have been targeted with witchcraft by three to one. The first category of claims was vastly more successful. This discrepancy is unsurprising, given that anti-witch vigilantism produces tangible, legally assimilable evidence in ways that acts of invisible, supernatural malevolence do not. Canadian cases proved more numerous than Australian ones.^{Footnote 2}

This composite digital archive is constantly evolving and increasing as more African asylum seekers pursue protection from witchcraft-motivated violence and as immigration regimes continue to regard such cases as noteworthy enough to publish. While the limits of a purely textual approach are obvious, asylum records nonetheless provide “rich documentary archives tethered to discrete legal contexts” (Lawrance & Ruffer Reference Lawrance and Ruffer2014:3). Close readings of asylum decisions reveal how law is concerned with compressing the complexity of lived experience into consistent, widely applicable rules and norms. They highlight how law’s prescriptive orientation shapes the engagement of immigration authorities—or their failure to engage—with cross-cultural difference, whether the system takes a human rights orientation or “social perception” approach to PSG (Good Reference Good2007:29).

Law and RSD

“Refugee” is a highly circumscribed, legally constituted status. An asylum seeker does not become a refugee until they submit to RSD in the state where they have sought sanctuary; they are subsequently designated as a refugee by immigration authorities in the receiving state, who use a combination of international and state law to adjudicate their claim.^{Footnote 3} The 1951 Convention establishes criteria for determining if an asylum seeker can be legally designated as a refugee. According to that document, a refugee is:

When an asylum seeker is determined to have a “well-founded fear of persecution” based on one or more of the Convention criteria, they are designated a “Convention” refugee. Non-refoulment, or “no return,” is the “primary protection provided by the Convention,” which prohibits states from expelling asylum seekers before immigration authorities have assessed their claims. Yet, “governments disagree over the exact obligations that the principle of non-refoulment imposes on states” (Loescher Reference Loescher2001:353).

Although Canada and Australia both have common law systems, the countries’ national immigration legislation and case law reflect starkly different interpretations of the Convention and RSD orientations. Canada’s 2002 Immigration and Refugee Protection Act (IRPA), reads the Convention widely, evidencing a strong human rights orientation. The 1993 Supreme Court decision in Ward v. Canada, the benchmark for adjudicating PSG claims, reinforces this humanitarian orientation.^{Footnote 4} In contrast, the Convention occupies a fraught space within Australia’s legal landscape. While the 1958 Migration Act provides the “main legislative basis for Australian immigration” (Castles et al. Reference Castles, Vasta, Ozkul, Hollifield, Martin and Orrenius2014:129), recent legislation has codified Australia’s interpretation of who counts as a refugee, in effect significantly narrowing the scope of refugeehood by substituting a revised version of the Convention definition for reference to the relevant chapter of the Convention. The 1997 benchmark High Court decision, Applicant A and Another v. Minister for Immigration and Ethnic Affairs, establishes a sociological approach to PSG. These interpretations reflect Australia’s overall orientation toward deterrence and detention. Over the last twenty years, RSD in Canada and Australia has consistently been revised and remade through new legislation, with the refining of soft law instruments, and as the result of national jurisprudence.

Witchcraft-based asylum cases reflect the shifting terrain of RSD. In turn, close readings of such cases offer important insights into how PSG is assimilated and how “credibility” is assessed in the Canadian and Australian systems. Overall, they show the fraught paths through which immigration authorities and asylum seekers have navigated cultural difference and institutional demands.

Asylum in Canada: Decisions in the First and Avenues of Appeal

Asylum seekers in Canada make their initial claims for protection either at a border point of entry or at an immigration office; those making witchcraft allegations have more typically applied at the border. Using a combination of oral and textual information, the border services officer to whom the claim is made assesses the asylum seeker’s eligibility to have that claim heard by a division of the Immigration and Refugee Board (IRB). If the asylum seeker has arrived without a valid visa, but the officer determines eligibility for a hearing, the claimant is issued a conditional removal order that comes into effect if and when the protection claim is denied. If the asylum seeker has arrived on a valid visa and the officer determines eligibility for a hearing, no removal orders are issued as long as the visa remains in effect. Eligible asylum seekers typically appear before the board within two months of making the initial claim. The officer’s notes taken during the initial assessment can be included as evidence in the hearing.

“The overwhelming majority of decisions that affect the lives of noncitizens in Canada,” Catherine Dauvergne writes, “are made by the IRB,” or the Immigration and Refugee Board, a multi-division, quasi-judicial, independent administrative tribunal established in 1989 and later mandated by IRPA. The board’s Refugee Protection Division (RPD), the branch of the IRB “responsible for ‘first instance refugee decisions,’” carries out asylum adjudication (Dauvergne Reference Dauvergne2012:309). A single member of the RPD is empaneled to hear an asylum seeker’s oral appeal to the division, and nearly every asylum seeker receives an oral hearing before a member. If, during the hearing, the asylum seeker is determined to be a Convention refugee, then the case need move no further through the RSD process. As François Crépeau and Delphine Nakache note, the Canadian system is unusual; with the creation of the IRB, “Canada has chosen to invest its resources in the first-level decision…” (Reference Crépeau and Nakache2008:53–54).

“Persecution” remains undefined by the Convention, and to the degree that Canadian jurisprudence defines “persecution,” it equates it with “serious mistreatment” resulting in the “denial of a core human right.”^{Footnote 5} Case law has broadly construed the meaning of “well-founded fear of persecution,” maintaining that a “well-founded fear” both recognizes the asylum seeker’s subjective fears and is “justified in light of the objective situation” from which they have fled.^{Footnote 6}Ponniah v. Minister of Employment and Immigration (1991) established that “the threshold level of probability of persecution for refugee status was something more than ‘a mere possibility’ and would be equivalent to ‘good grounds,’ a ‘reasonable chance,’ or a ‘serious possibility.’”^{Footnote 7} When PSG is the protection ground, Canada employs the ejusdem generis, or “of the same kind” principle, which holds that because the other Convention categories—race, religion, political opinion, nationality—are broadly defined and based on “immutable” characteristics, PSG should be construed in the same way (Hathaway & Foster Reference Hathaway and Foster2014:226).

Fumni Kogbe and the Immigration and Refugee Board of Canada exemplifies first-level decision making at its most straightforward.^{Footnote 8} In 2000, a thirty-four-year-old Nigerian widow from Edo State named Fumni Kogbe sought asylum in Canada because her late husband’s family believed his “unnatural” death was due to her witchcraft; she stipulated that his death was actually caused by a high fever, which she attributed to malaria. A practicing Christian, Kogbe refused both to enter a levirate marriage and to engage in local funerary rites, and submitted that she faced persecution from her “pagan” affines for these decisions (3). Kogbe reiterated to the single member empaneled to assess her case that “she had been branded a witch and fear[s] that she will be killed by members of her husband’s family in the event of here (sic) return to Nigeria.” She was unable to avail herself of state protection because, as she explained, the police refused to intervene in such “family matters” (4). Although the Nigeria Criminal Code renders a person who “accuses or threatens to accuse any person with being a witch or with having the power of witchcraft” guilty of an imprisonable offense, police nonetheless typically treat witchcraft acts and accusations as quotidian, private affairs.^{Footnote 9} There was no internal flight alternative (IFA), or location where she could relocate and reside safely in Nigeria, available because Kogbe’s brother-in-law was a senior police officer who could track her throughout Nigeria (3).

From her “basic account,” which was found to be credible due to its strong consistency across “port of entry notes, her personal information form (PIF), and her testimony,” and the documentary evidence about the social situation—witchcraft, levirate marriage, etc., in Nigeria—provided by Kogbe’s attorney, the board found that she had a well-founded fear of persecution based on the interrelated criteria of religion and PSG. Kogbe was granted refugee status in an oral decision (4-5). The board recognized the generalized risk of persecution attaching to “accused witches” in Nigeria and how Kogbe’s experiences made her a member of this PSG.

But rarely are cases so easy to affirm. When the RPD denies the claim, the asylum seeker may appeal for a judicial review of the decision. Judicial reviews are undertaken by judges in the Federal Court to examine the decision-making process of the protection division. If the court determines that the member who conducted the original hearing erred along processual or technical lines, it can order a new hearing, conducted by a fresh panel, to redetermine the case. Judicial review offers a critical avenue of recourse for asylum seekers whose allegations about life-threatening danger are assessed by single individuals tasked with hearing an average of two cases per day.

The complex case Gyarchie v. Canada stands in stark contrast to Kogbe and offers a sharp lens with which to examine how Canadian immigration authorities have conceived of the personalized, as opposed to generalized, persecutory risks presented by witchcraft accusations. It provides important insights into the legal reasoning that goes into a grant of judicial review.^{Footnote 10} In 2003, after Mary Efua Gyarchie’s husband took a younger, second wife from a more prestigious and powerful lineage, he began physically abusing Gyarchie, even stabbing and beating her. She appealed fruitlessly to the police for protection from her husband’s abuse; domestic violence legislation did not exist before the 2007 Domestic Violence Act, and the police generally considered violence between intimates “purely as a domestic matter,” as they informed Gyarchie (1).^{Footnote 11}

Gyarchie spent several years working in Jamaica, returning to Ghana to visit her elderly father in 2007. Immediately upon her arrival, Gyarchie’s husband “came to her father’s house, accused her of being a witch and casting a spell that caused his second wife to be infertile, and demanded that she attend a shrine where an oracle would publicly confirm her witchcraft and perform an exorcism,” slapping her twice while leveling the accusations (2–3). She feared for her life and sought relief from the police who “again refused to help and counseled (sic) her to seek help if she was indeed a witch” (3). While the Ghana Criminal Code contains provisions to “protect women accused of witchcraft,” police typically dismiss witchcraft accusations as ordinary family matters rather than treating them as criminal offences.^{Footnote 12}

After this “frightening encounter,” which led Gyarchie to believe her life was in jeopardy, she fled to Jamaica, remaining there until she decided to pursue a midwifery course in Canada two years later, arriving in Canada on a student visa. Distraught by the death of her father and too fearful to return to Ghana for his funeral, she postponed starting her studies and requested a visa extension. When it was denied, she sought the guidance of a lawyer, who advised her to make a refugee claim. Gyarchie was surprised, having believed that refugee protection was “available only for political dissidents or those fleeing civil wars, not those fleeing domestic violence” and witchcraft accusations (2).

The RPD denied Gyarchie’s application, finding it unreasonable that an educated person who had lived abroad would be unaware that she could file a claim for protection from domestic violence-based persecution if this persecution formed a nexus with one of the Convention criteria. The RPD also found it suspect that Gyarchie had not sought asylum in Jamaica if she had indeed feared for her life during her tenure there. Gyarchie next filed for a Pre-Removal Risk Assessment (PRRA) application, a written description that a failed asylum seeker may offer detailing the risks that they would face if returned to their home country. Typically, the PPRA is used to ascertain if there have been any changes in the situation in the home country that would “alter the original protection determination” or if new evidence has arisen since the time of the original hearing that establishes that the asylum seeker qualifies for refugee status.^{Footnote 13}

When Gyarchie’s PRRA was rejected, she sought a judicial review, arguing that she faced a “serious risk of gender-based harm in multiple forms if returned to Ghana,” including violence by her estranged husband, accusations of witchcraft, “cruel, ritualistic abuse to remove her ‘magic,’” and banishment from the community due to her alleged witchcraft. The judge conducting the judicial review concurred with Gyarchie that the board’s analysis had not dealt with the problem of witchcraft-based persecution sufficiently, considering only the individuated risks posed by Gyarchie’s estranged husband’s witchcraft accusations and overlooking “the general dangers in Ghana of being accused of witchcraft” (15). Analyzing the documentary evidence presented by Gyarchie on witchcraft-motivated violence in Ghana, the judge found that the PRRA officer had also neglected to adequately assess the textual evidence about the persecution of alleged witches in Ghana, and by extension, had failed to consider Gyarchie’s case within the context of the generalized risk posed to accused witches in her country of origin (25–26).^{Footnote 14} Quashing the original decision, the judge referred Gyarchie’s case for consideration by a new panel.

Gyarchie illustrates how in a system so conditioned by individual assessments judicial review provides an “invaluable check on the discretion of individual decision-makers in refugee adjudication” (Millbank Reference Millbank2009:24). This is especially crucial in witchcraft-based asylum cases where claimants often encounter adjudicators who not only have little to no familiarity with the bases of their claims, but who also hold a culturally conditioned predisposition to discount witchcraft allegations. Critiques of the judicial review process have focused on the low grant rate (75 percent of unsuccessful claimants request leave for judicial review, but only approximately 10 percent of requests are granted), on the deep reluctance of the courts to “overturn factual findings based on an adjudicator’s firsthand evaluation of the credibility of testimony,” and on how judicial review results only in a new hearing by the RPD, not in a new decision (Macklin 2010:146; Rehaag Reference Rehaag2008:360).

With the goals of offering unsuccessful asylum seekers additional avenues for appeal and of lightening the caseload of the Federal Court, 2001 legislation created a Refugee Appeal Division (RAD) within the IRB. The law explained that “a person or the Minister may appeal, in accordance with the rules of the Board, on a question of law, of fact or of mixed law and fact, to the Refugee Appeal Division against a decision of the Refugee Protection Division to allow or reject a person’s claim for refugee protection.”^{Footnote 15} As Macklin notes, “The mandate of the Refugee Appeal Division was to provide an internal review on the merits…” (2009:145). However, when the legislation was brought into force in 2002, it was done so selectively, and provisions relating to the appeal division were not implemented. The RAD was finally established in late 2012, offering failed claimants a venue through which to appeal a negative decision of the protection division on a “question of law, of fact or of mixed law and fact,” and the opportunity to present new evidence that “was not reasonably available at the time of the Refugee Protection Division process.”^{Footnote 16} It also offers the Minister of Citizenship and Immigration a space in which to appeal a positive decision of the RPD. Both asylum seekers and the Minister retain rights to appeal to the Federal Court. Appealing to the RAD is “paper-based” process, with adjudicators relying on the RPD record and documents presented by the parties to the case; appeals are decided by single-member panels.^{Footnote 17}

Witchcraft-driven asylum cases have already been referred to the RAD. A 2013 case underscores how in this venue, too, witchcraft claims have proven to be challenging for adjudicators whose orientation toward culture-based claims is typically prescriptive and who are largely unfamiliar with and skeptical of the discursive and material “work” that witchcraft performs in African societies. When Emmanuel Okieriete fled Nigeria for Canada, making a refugee claim at the airport, he asserted that his life was at risk from his “Aunty Precious,” whom he described as “being involved in witchcraft.”^{Footnote 18} Okieriete, a Christian, had refused to surrender his son, Samuel, to Precious, who desired to “perform a ritual” as part of the initiation process to make Samuel a “priest-oracle.” Okieriete’s family, including Samuel, had arrived in Canada in 2011 without Okieriete and had made a successful asylum claim. Precious, Okieriete alleged, still blamed him for his son’s disappearance, and continued “to pursue” him together with her “thugs,” putting Okieriete in ongoing fear for his life (1–2).

When the board rejected Okieriete’s claim on the grounds that the Nigerian city of Ibadan offered a viable IFA, Okieriete appealed to the RAD. He argued that Ibadan failed the so-called “two-pronged test”—there was indeed “serious possibility” of his being persecuted there and that circumstances in Ibadan were such that “on a balance of probabilities,” the board “could not be reasonably satisfied that he would not be harmed in Ibadan” (1–2). Okieriete explained that Precious, who owned numerous properties in Ibadan and had many friends among the police and politicians there, would be able to locate Okieriete easily were he to live openly. Precious, Okieriete alleged, had killed her own sister in Ibadan a short time earlier.

The member considered the question of an IFA to be “determinative,” and found that it was “unreasonable” for the RPD to conclude that Okieriete had “failed to provide a reasonable explanation as to why he believes he would be discovered and/or harmed in Ibadan” (3). The member raised important issues of credibility around witchcraft, revealing certain assumptions about what types of people would believe in “witchcraft.” He noted that the conflict over Ibadan’s viability occurred in “the context of a claim of witchcraft,” which he found incredible, given that the “Appellant is a relatively well off, educated Christian who lived in an urban setting and was able to undertake travel to Canada” (5). In his implicit view then, witchcraft is something that (1) Christians reject, (2) educated people do not believe in, (3) urban people do not subscribe to, and (4) wealthy people do not countenance. Concurrently, the member acknowledged that he lacked the evidence to analyze fully how witchcraft figured in the case and focused instead on the problem of Ibadan. Ultimately, he referred the case for redetermination by a newly constituted board. Okieriete underscores how even in a regime with an explicit human rights orientation, witchcraft claims are especially complicated and contentious.

RSD in Canada is largely an exercise in orality underpinned by documentary evidence. Using the discursive resources at their disposal, asylum seekers must convey the “accumulated ideas, images, and associations and so on which make up the wider social and cultural contexts” from which they come (Taylor Reference Taylor2006:94–95). In witchcraft-based asylum cases, this task is particularly complicated, as asylum seekers must draw on all the discursive resources at their disposal to persuade immigration authorities of the real perilousness of being an accused witch or a victim of witchcraft, even if they cannot convince them of the reality of witchcraft per se.

Asylum in Australia—The “Real Chance Test” and Cognizability

In contrast to the Canadian approach, RSD in Australia is a system distinguished by its stringent “culture of disbelief,” its privileging of domestic immigration legislation over international instruments, and its limiting approaches to “cognizability” and “persecution” (Good Reference Good2003:3). Australia’s Humanitarian Programme “allocates a set number of places to refugees and others in need of humanitarian assistance and protection,” privileging asylum seekers who apply for refugee protection and resettlement from outside Australia through UNHCR channels (McKay et al. Reference McKay, Thomas and Kneebone2012:114). Asylum seekers who arrive by air, traveling on a valid visa, may apply for protection as well.^{Footnote 19} Australia also recognizes refugees sur place who make protection claims while their preexisting visas are still valid. Claimants may request that the Department of Home Affairs, the department of the Australian government with ultimate jurisdiction over asylum seekers, recognize that they engage Australia’s protection obligations under the Convention and Australia’s 1958 Migration Act. Asylum seekers making witchcraft-based claims have typically claimed asylum upon arrival or sur place in similar numbers.

Immigration legislation has redefined and narrowed the meaning of “well-founded fear of persecution” to specify that there must be a “real chance” the asylum seeker would be subject to persecution if returned to the country of origin and that “the real chance of persecution relates to all areas of a receiving country.”^{Footnote 20} Amendments to the 1958 Migration Act set out to limit the scope of “persecution,” stipulating that the persecution arising from Convention criteria “must involve serious harm to the person and systematic and discriminatory conduct.”^{Footnote 21}

Finally, in cases where PSG is the protection ground, Australia takes a limiting “social perception approach” to cognizability (Aleinikoff Reference Aleinikoff, Feller, Turk and Nicholson2003:9). The Applicant A decision determined that a “particular social group” is “a collection of persons who share a common characteristic or element which unites them and enables them to be set apart from society at large…not only must such persons exhibit some common element; the element must unite them, making those who share it a cognisable group within their society” (3). This approach yields more limited PSG interpretations than does ejusdem generis.

The cases analyzed below indicate how Australia’s immigration legislation and jurisprudence, which privileges the state’s right to control migration over the asylum seeker’s right to non-refoulement, has narrowed the Convention in regard to fear, persecution, and cognizability. They also illuminate how Australian RSD involves a “tangled network of discursive processes,” both written and oral (Maryns Reference Maryns2005:1). RSD begins with an intensive paper-based exercise, followed by documentary and oral retellings of the asylum seeker’s experience offered across various legal settings. The cases of William Zachariah Alozi and Nnamdi Ojukwu underscore both the profound disparities in the linguistic capacities and discursive resources of asylum seekers and immigration authorities and the critical implications of such inequalities for RSD and for the success of an asylum seeker’s claim.

Presently, RSD begins for asylum seekers who have arrived lawfully when they apply to the Department of Home Affairs for a permanent protection visa, which enables a protected person to live and work in Australia as a permanent resident. They prepare and submit a packet of materials, including a signed “Statement of Australian Values”; claimants may also be interviewed. From autumn of 2013 through 2017, asylum seekers arriving with valid visas applied to a regional office of the Department of Immigration and Border Protection (DIBP). The extensive application included forty-one pages of instructions and forms, concerned largely with the claimant’s personal and family details, to be completed by the asylum seeker and submitted with copies of supporting documentary evidence that had to be certified and translated if they were in a language other than English. After review of the application packet, the asylum seeker was invited to interview with the DIPB decision-maker and possibly to provide additional information; the interview was taped and became part of the case record. The immigration officer queried the asylum seeker about their experience, seeking to ascertain how it aligned—or not—with Convention criteria and the 1958 Migration Act. After a short break, the asylum seeker was invited to make an additional statement. If the asylum seeker engaged a “migration agent” to assist in lodging the claim, the agent was permitted make a “submission,” that is, a legislatively-based document situating the case within Convention criteria and providing additional country information.

The 2015 case WZAL v Minister for Immigration and Anor demonstrates how narrative complexity evolves throughout RSD as the asylum seeker retells their story to multiple immigration authorities across varied legal settings.^{Footnote 22} It shows how authorities gain power over the asylum seeker’s narrative, “decontextualizing and recontextualizing” elements of the story as it moves through the RSD apparatus (Park & Bucholtz Reference Park and Bucholtz2009:486). WZAL illuminates the “level of harm” (and concomitant “degree of suffering”) necessary to amount to legally recognizable persecution and the subjective and objective elements of “harm” necessary to meet the test for determining “well-founded” fear of persecution.^{Footnote 23} It also underscores the broad geographic scope—“all areas of the home country”—that the threat of persecutory harm must encompass for an asylum seeker facing a “real chance” of persecution to be considered a refugee.^{Footnote 24}

When William Zachariah Alozi, the young Nigerian claimant in WZAL, sought asylum in Australia in 2014, he first applied for a protection visa. He asserted that he feared returning to Nigeria because of “my village witch craft” (sic). Specifically, Alozi feared being attacked by “evil doing and demonic people” from his village who had killed his father in retribution for his father’s anti-witchcraft activities. These villagers, he asserted, had attacked him with acid, and furthermore, Abia, a city in Aba State, where he fled to escape the neighbors’ wrath, was wracked with Muslim-Christian religious violence. Nigerian authorities offered no protection, Alozi explained, because “the government is bad, does not care about poor people, only cares about themselves” (4).

Under the RSD procedure in force, a DIPB officer made the first instance decision as to whether the asylum seeker was entitled to protection after assessing the application packet (and information gleaned in the interview if one was conducted). Decisions were granted in writing, and if the application was denied, an asylum seeker could appeal to the Administrative Appeals Tribunal (AAT), which described itself as a “one-stop shop for the independent review of a wide range of decisions made by the Australian government.”^{Footnote 25} The tribunal, composed of one delegate, could consider new evidence, including that not submitted in the original application. When a favorable decision could not be reached on the basis of the packet alone, the tribunal requested a private hearing with the asylum seeker to gather more information about the case.

In WZAL, the DIPB officer rejected Alozi’s application, and Alozi appealed to the tribunal, which invited him to give evidence and lodge his arguments at a hearing. Making his claim orally, Alozi was able to present his narrative of persecution with much more detail and precision than he had been able to muster in his written submission. Alozi reiterated that his father had been “tortured to death in the village in which he lived as a result of his father’s activities against witchcraft practiced by some people in the village” (5). These same villagers had turned on Alozi as well when he began “asking questions about his father’s death which people in the village did not like,” and he was attacked subsequently. He offered to show his acid wound scars (6).

Alozi added another layer to his narrative. The villagers his father had angered through his anti-witchcraft activities had attacked Alozi with witchcraft. Using “magic spells that could make people go crazy,” their “charms” had rendered him “crazy,” insensibly “running naked in the street.” Alozi confirmed that he had fled his village for a town sixty miles away in Abia State, remaining there for several years until sectarian violence escalated. Fearing for his life again, he fled to Malaysia, ultimately arriving in Australia (5). Weighing Alozi’s claims about his experience of witchcraft-driven violence and his fears of future persecution in Nigeria, the tribunal accepted that harmful practices “labelled as ‘witchcraft’ take place in Nigeria.” Here the tribunal stopped short of accepting the reality of witchcraft per se, but instead acknowledged that tangible practices—the delegate had earlier asked Alozi if his alleged bewitchment could not be the result of simple poisoning—could be perceived as “witchcraft” by local populations. The tribunal read witchcraft violence both as harmful (perceived) occult acts perpetrated by “adherents of witchcraft” and acts of ordinary violence carried out against opponents of witchcraft by its adherents (6). Holding that while Alozi had suffered violence motivated by his fellow villagers’ witchcraft beliefs, the tribunal asserted that such harm could be avoided if he returned to the urban center of Aba rather than to his natal village. Drawing on country information, the tribunal noted that rather than being a site of Muslim-Christian violence, Aba, and Abia State more generally, had provided sanctuary to Christian Igbo like Alozi fleeing sectarian violence; such conflict would not preclude Alozi from returning to Abia. The tribunal saw “no substantial grounds for believing the applicant would suffer harm as a necessary and foreseeable consequence of being returned from Australia to Nigeria arising from the circumstances of his father’s death, the adherents of witchcraft in the village of his birth, or from witchcraft” (6). Alozi’s appeal was rejected.

If an appeal to the tribunal failed, the asylum seeker could appeal to the Federal Circuit Court (FCC) for a judicial review of the case on the grounds that the tribunal had made an “error in law,” or “jurisdictional error” in its finding. “Identifying a wrong issue, identifying a wrong question, ignoring relevant material, relying on irrelevant material” all constitute jurisdictional errors as does “an incorrect interpretation and/or application to the facts of the applicable law.”^{Footnote 26} If the court determined that the tribunal made a legal error, it remitted the case back to the tribunal for a fresh decision.

Alozi did not offer a particular ground for jurisdictional review. Rather, he introduced a new claim through a handwritten submission to the court: Were he to be returned to Nigeria, he would again be harmed by the notorious Bakassi Boys gang whom he alleged had attacked and tortured him in 2001 (8–9). He submitted a 2002 report by Human Rights Watch on the Bakassi Boys’ role in the development of vigilantism in Nigeria’s southeastern states. The court read Alozi’s application for protection, the delegate’s decision, and the tribunal decision. It found no mention of the gang and determined that assessing the new evidence about the gang would constitute an “impermissible merits review” (13). Further, the court emphasized that the tribunal had considered the most recent and relevant country information when assessing Alozi’s claims about the persecution of Christians in Nigeria and underscored that “the choice and the assessment of the weight of the country information are matters for the Tribunal.” Overall, the court concluded that whether in reference to new evidence or in regard to country information, Alozi had “not demonstrated that Tribunal had fallen into jurisdictional error” (14). His appeal was dismissed.

Before 2015, the initial phase of RSD was carried out under the auspices of a succession of bodies with subtly different appellations but virtually identical functions and procedures as those engaged by Alozi. From late 2001 through 2006, RSD fell under the portfolio of the Department of Immigration and Multicultural and Indigenous Affairs (DIMIA). DIMIA was succeeded by the Department of Immigration and Multicultural Affairs (DIMA), formed in late January 2006, and a year later, it was followed by the Department of Immigration and Citizenship (DIAC), which existed from January 2007 to September 2013. In 2012, Nnamdi Ojukwu, a Nigerian from Anambra State applied to the DIAC, seeking asylum on the grounds that members of his natal village had accused him of witchcraft and tortured him on two occasions; village elders had subsequently elected to kill him and had hunted him across Nigeria.^{Footnote 27}

Ojukwu’s claims highlight the Australian refugee regime’s distinctive construction of “cognizability” and its stringent “culture of disbelief” (Good Reference Good2003:3). In applying for protection, Ojukwu followed the same procedure and prepared the same sort of paperwork as Alozi, submitting an extensive documentary packet to DIAC. A delegate rendered the first instance decision based upon reviewing the packet. Ojukwu contended that he had been beaten and tortured—including being hanged from the ceiling—and that community members had accused him of witchcraft and threatened his life. Ojukwu submitted photographs of his injuries and explained that he had sustained them during the beatings and torture. His claim was dismissed in the first instance proceedings. The delegate rejected Ojukwu’s contention that he had been accused of witchcraft (4).

Asylum seekers whose applications were rejected at the first instance were entitled to appeal to the Refugee Review Tribunal (RRT), a statutory body that conducted merits reviews of delegates’ decisions. Ojukwu applied in writing for a review hearing and was invited to appear before the tribunal. The RRT decision maker, or “member,” assessed all the documentary materials related to the asylum seeker’s initial application, the recording of the interview with the delegate had one been conducted, and the delegate’s decision. The member also considered any new information that the asylum seeker submitted. The tribunal was empowered to gather additional evidence to use in adjudicating the appeal. The asylum seeker was free to present their case as they thought best, including determining which issues to dispute and which evidence to collect and introduce in the proceedings.

At the hearing, the member queried the asylum seeker about their reasons for seeking asylum in Australia, typically including questions about general conditions in the asylum seeker’s home country—if police protection was available, if IFAs existed, if the asylum seeker had utilized available relocation alternatives—and importantly, what the asylum seeker feared would happen if they were sent back to the home country. The tribunal might also identify “adverse information” which did not support the asylum seeker’s claim. In such instances, the asylum seeker could respond immediately, request to respond in writing at a future date, or respond in a second hearing.

Cognizability figured importantly in Ojukwu’s initial application and in the appeal to the RRT. He submitted a declaration to the RRT before his hearing that recapitulated his initial claims that he had been accused of witchcraft by members of his community and tortured as consequence. In labored English, he wrote, “I was accused of being witchcraft that leads to the communities elders from my to tortured, abused including hanging legs from the ceiling, and the elders and young men started throwing objects including stones on me. I almost die during the beaten I receive from the communities that accused me of witchcraft.” His new declaration added that he had sought protection from the “local police” who, expressing a view now well familiar to us, had informed him that “this is community’s problem to do with witchcraft and the police cannot intervene in such matter and they cannot me and they unable to arrest anybody they suggested that I should the area if I have a chance to do so.” Ojukwu concluded by describing how after moving to Abuja, he was nevertheless attacked at his home “with a cutlass” and then forced back to his village where he was again tortured by members of his community because, he explained, “I’m falsely accused of witchcraft” (4).

At his hearing, which Ojukwu attended with an advocate, the member asked Ojukwu to “describe the nature of the problems he experienced with his community.” Augmenting his earlier narratives, Ojukwu explained that after his cousin died in a car crash, community members consulted an “oracle” who named Ojukwu as a witch responsible for the cousin’s death. Subsequently, Ojukwu was asked to return to his natal village where a crowd had gathered, demanding that he confess to witchcraft (5–6). Up until this incident, Ojukwu had not encountered any difficulties in the community. He was aware, however, that village elders had been “envious” of his father’s landholdings; he read the witchcraft accusation as a means to deprive him of rights to his father’s property. Following this initial witchcraft accusation, Ojukwu was blamed when misfortune befell the village, even after he fled to Abuja.

After nearly a decade away, Ojukwu thought it was safe to return to the village at his uncle’s invitation. Once there he was again attacked, stripped, and stuffed in a container which was left in what he termed “the evil forest”; other villagers found and released him. Ultimately, his persecutors followed him to Abuja; four carloads of village elders armed with machetes arrived at Ojukwu’s urban home and forced him to return to the village with them, where he was “beaten, tied up, and locked inside a room for three days without food or water.” The group’s intent, he noted, was to kill him. He escaped when a young boy “took pity” on him and released him, explaining to Ojukwu that the villagers “were planning to kill him by burning tyres the next day.” Ojukwu sought sanctuary at the police station, explaining his case. The police told Ojukwu his only option was to flee; they termed the attack a “village issue” and emphasized that the community could still kill him. The police, Ojukwu underscored, “also believe in witchcraft” (6–7).

The RRT asked a series of questions to elucidate further how Ojukwu’s case conformed to Convention criteria. Responding to the question as to whether Lagos offered an IFA, Ojukwu explained that it did not; given the ease with which the villagers had located him in Abuja, they could find him in Lagos. The RRT queried Ojukwu as to why the community members wished to harm him—Ojukwu reiterated that they accused him of witchcraft. The member asked about the shape of witchcraft in Nigeria: What is the Igbo term for “witchcraft”? Had Ojukwu “heard of or witnessed others being accused of witchcraft by his community before”? Were men accused of being witches or was it primarily women and children who were accused? What was the purpose of being taken to the pastor?

Ojukwu replied that there was a “history” of people accused of witchcraft, or “amosu,” being buried alive, and that “normally” accused witches were killed or “cast out” from the village. He emphasized that witchcraft was “serious in Nigeria,” adding that “if something bad happens ‘they’ will say someone is a witch.” Since his mother was from “Cross River State where witchcraft is normally practiced,” Ojukwu pointed out, the villagers had even more reason to believe that he was a “witch.” He also submitted country evidence, two articles “about the treatment of those accused of witchcraft in Nigeria and Africa more broadly” (6–7).

Ojukwu’s advocate revisited country evidence submitted earlier, a report about witchcraft in Africa. He added depth and nuance to Ojukwu’s explanations, referring to newspaper articles about Nigerian witchcraft abroad and a High Court case in Australia where a Nigerian was acquitted of witchcraft to illustrate the pervasiveness of witchcraft and the extent to which men are accused of witchcraft. Country evidence, he argued, “highlights how deadly witchcraft accusations can be.” He underscored the absence of state protection, explaining that “even the government in Nigeria and other parts of Africa are unable to crackdown” on witchcraft and that the “police do nothing and fear being attacked too or killed by an evil spirt.” And he unraveled the logic of witchcraft, noting that while a car crash like that which killed Ojukwu’s cousin would be considered a simple accident in Australia, it would not in Nigeria “where they think something is behind it.” Most Nigerians, he added, would “look for something or someone else to blame.”

The member was willing to accept that accused witches constituted a PSG “for the purposes of the Convention” and that “‘witches’ (in other words, those perceived to be witches) constitute a PSG in Nigeria.” Ojukwu, in the RRT’s estimation, was not a member of the PSG “accused witches” (7–8). Citing the “vagueness” of Ojukwu’s claims about the pastor who was tasked to “deliver” him from witchcraft, about the identity of the “oracle” who had named him as a witch, about “who, specifically, beat him in the first attack,” and about the total number of elders in his village, the Tribunal dismissed Ojukwu’s oral and written testimony. Ojukwu’s photographic submissions, the member allowed, indicated that he had been beaten, but did not establish anything about the circumstances of the injuries. The Tribunal did “not accept that the applicant was accused of witchcraft in the past in Nigeria,” nor that he had been “beaten, tortured, detained and starved for three days, humiliated and stripped naked and threatened to be burnt with a tyre, or threatened to be killed—by community members.” As such, if returned to Nigeria, Ojukwu would not face persecution as an accused witch. His refugee claim was dismissed (9–10).

Asylum seekers received the tribunal’s decision in letter form. When an appeal to the RRT was accepted, the primary decision was set aside and refugee protection granted. As we saw above in WZAL, when the tribunal rendered a negative decision, asylum seekers could appeal to the Federal Court for a judicial review. The archival record does not indicate if Ojukwu pursued this avenue. Had he done so, his case would have followed a similar path to that of Alozi.

WZAL and Ojukwu clearly demonstrate the distinctive elements of Australian RSD. They highlight how it centers overwhelmingly on multiple written accounts, supplemented at various junctures by oral testimony, which set out the asylum seeker’s personal history of harm. This “regime of paper documents” builds a series of increasingly detailed, complex narratives of persecution with which oral testimony must align (Hull Reference Hull2012:1). Asylum seekers, who typically engage in RSD with important discursive deficits, face significant challenges in sustaining the consistency of their stories as ever more layered, complex iterations of their experiences are inscribed in the bureaucratic record.