1 Introduction

This chapter is based on extensive ethnographic research in the coastal Kenyan town of Malindi, where concerns about witchcraft and terrorism reflect and inform tensions and conflicts, which sometimes result in violence.¹ In recent years, Kenya’s coast has witnessed terrorist attacks and recruitment efforts by Somalia-based Islamic terrorist organization Al-Shabaab, leading various Western donors to fund programmes to “counter violent extremism” that are implemented by Kenyan civil society organizations. Although Western donors often primarily aim to counter violent extremism that is related to Islam (see Meinema 2021a), a recent “action plan” to “counter violent extremism” by the government of Kilifi County (2017), in which Malindi is situated, also argues that a “strong belief in witchcraft practices” is a “motivating factor” of “violent extremism”. The report subsequently mentions the problem of witchcraft-related killings, which primarily occur among the Mijikenda ethnic group, and the Mombasa Republican Council (MRC), a coastal secessionist movement which the report links to Mijikenda “traditionalism” by mentioning the sacred Kaya forests of the Mijikenda as MRC “hideouts”.

This understanding of “violent extremism” as not only related to Al-Shabaab, but also to “witchcraft practices”, coastal secessionism, and Mijikenda “traditionalism” raises several questions, which I aim to address here. How do discourses about witchcraft and terrorism intersect in coastal Kenya? And how do these discourses shape modes of religious coexistence in Malindi, given that witchcraft and terrorism are often primarily associated with particular religious groups? In particular, the chapter aims to investigate how discourses about witchcraft and terrorism politicize and shape how various religious groups materially manifest themselves in the urban environment of Malindi (cf. Meyer 2015, 597). In these dynamics, particular material objects become ‘things of conflict’ because they are commonly associated with witchcraft or terrorism, which may link widespread fears about witches and terrorists to the actual people who engage with these objects.

The earlier mentioned questions are pertinent, because of the religious diversity that characterizes Malindi and the coastal region of Kenya more generally. While Kenya as a whole has a Christian majority population, Malindi has long been inhabited by Swahili Muslims, whose presence in towns along the East African coast stretches multiple centuries. In rural areas around these towns, other coastal groups such as the Mijikenda often form the majority. In the Malindi environment, the Giriama are the most populous Mijikenda subgroup, who may self-identify as Christians, Muslims, or “Traditionalists”.² Malindi is also inhabited by Christians who come from “upcountry” (bara) parts of Kenya and have varying ethnic backgrounds. While the 2010 Constitution of Kenya states that “there will be no state religion”, Christianity has long played predominant roles in Kenyan law, public life, and politics (Deacon et al. 2017). In this situation, Swahili Muslims and Mijikenda from various religious backgrounds have often complained about political and economic marginalization in relation to people from ‘upcountry’, who are most often Christians. In recent decades, such feelings of marginalization have not only informed recruitment strategies of Al-Shabaab, which has managed to gain support amongst a small segment of Kenyan Muslims, but also a wider popular appeal for the MRC, which was mainly active between 2010 and 2013. This coastal secessionist movement involved people from various ethnic and religious backgrounds and primarily mobilized a regional coastal identity to challenge “upcountry” domination (Willis and Chome 2014). Taking the religiously diverse constellation of Malindi as a starting point, this chapter explores how fears about witchcraft and terrorism set in motion complex dynamics of revelation and concealment, which shape the material ways in which various religions manifest themselves in Malindi.

The chapter is constructed as follows. First, I provide a theoretical reflection on witchcraft and terrorism that describes these phenomena as entangled with social tensions, mistrust, and complex dynamics of revelation and concealment, which can be studied productively through a focus on materiality. Second, I focus on the ways in which people speak about witchcraft and witchcraft-related violence in Malindi and explore how different religious groups position themselves in relation to these phenomena. I pay special attention to the ways in which discourses about witchcraft shape the material religious forms through which people express themselves in the urban environment of Malindi. Third, I investigate how discourses about witchcraft and terrorism intersect within Kenyan understandings of violent extremism and show how attempts of Muslims and Traditionalists to avoid associations with violent extremism shape the material forms by which they organize themselves in Malindi. Finally, demonstrating that Christianity is only rarely associated with witchcraft or terrorism, I analyse how the circulation of discourses about witchcraft and terrorism affects the material expression of Christianity, Islam, and Traditionalism in uneven ways.

2 The Materiality of Mistrust

Anthropological studies of witchcraft have often emphasized that discourses about witchcraft offer insights into the tensions and social bonds around which society is organized (e.g. Evans-Pritchard 1950). Suspicions that particular people secretively practise witchcraft frequently arise around instances of misfortune or social inequalities, focusing on objects, people, or behaviour that are considered extraordinary or anti-social. Such suspicions may crystalize in witchcraft accusations, which link instances of misfortune and the alleged anti-social behaviour of the accused together by treating them as part of an underlying deeper evil, namely witchcraft (Parkin 1985, 228–9). As witchcraft accusations frequently focus on those who fail or refuse to adhere to particular moral and behavioural standards, they tend to reflect tensions within society, as well as providing a vocabulary to express, catalyse, and intervene in such tensions (Stewart and Strathern 2004, 1–28).

While witches are often seen as evil, anthropologist Peter Geschiere (2013, 14) stresses that witchcraft should not be conceptualized as the antithesis of sociality, trust, and kinship but rather as its flipside, since witches are often considered to be dangerous exactly because they disrupt social bonds from within, forming a dangerous threat of betrayal in intimate circles of sociality where sociability and trust should reign. I made similar observations in Malindi, where people often understand jealousy (wivu) among family members or neighbours as a central driver for people to practise witchcraft (uchawi) (Tinga 1998, 183). Yet, witchcraft discourse can also take the nation as the social unit that is threatened by occult forces from within. Noticeably, political relations in Kenya are often described in familial terms, emphasizing the paternal responsibility of a predominantly male and elderly elite to provide guidance and development to women (wamama) and youth (vijana) (Van Stapele 2016, 314). Thus, when witchcraft fears are raised on a national scale, arguably they are powerful because they resonate with similar fears that exist within intimate social bonds on familial scales.

Within the Kenyan context, notions of witchcraft as a threat to the nation have roots in colonial times. British administrators generally conceived fears about witchcraft to be an expression of African superstition, which they expected to disappear under modernization. Paradoxically, however, British officials also used the term ‘witchcraft’ to demonize African movements that challenged colonial authorities. For example, British officials associated the Mau Mau uprising against settler colonialism in the 1950s with “witchcraft and savagery” (Smith 2008, 28). Central to colonial fears was Mau Mau’s appropriation of traditional Kikuyu oaths, which Kenya government commissioner Frank Corfield considered to be “intimately connected with the belief in witchcraft” (Corfield 1960, 163). Mau Mau was also described as a terrorist movement, and subsequently became “the international face of terrorism in the 1950s”, which “embedded itself in the Western mind as the metaphor for deranged violence, primitive savagery, and rejection of ‘civilization’” (Berman 2007, 529). Scholars have argued that these imaginations concealed economic and political motives of Mau Mau, and legitimized its violent repression by colonial authorities (Berman 2007, 542). The condemnation of African ideas and practices as related to ‘witchcraft’ should be understood in relation to broader trends in the history of colonialism, in which concepts such as ‘witchcraft’, ‘idolatry’, and ‘fetishism’ were used to negatively evaluate indigenous African religious traditions along evolutionary lines as backward and primitive (see the introduction to this volume; Meinema 2021b).

Notions of malicious spiritual forces being dangerous threats to peace and public prosperity continued to shape postcolonial political dynamics. In the 1990s, for example, the Mungiki movement rose among marginalized Kikuyu, which took inspiration from Mau Mau and Kikuyu Traditionalism. The movement criticized political elites, involved itself in organised crime, and quickly gained a notorious reputation for violence. Government officials, church leaders, and political commentators subsequently associated Mungiki with devil worship, with concerns again focusing on Mungiki’s use of ritualized oaths (Smith 2011, 58). Given that the state has responded to Mungiki through police crackdowns and extrajudicial violence (Van Stapele 2016), I suggest that such imaginations present Mungiki as a demonic evil that needs to be dealt with in a confrontational and violent manner (Rio et al. 2017, 25).

Furthermore, the association of particular groups such as Mungiki with demonic spiritual powers sets in motion complex dynamics of revelation and concealment. Stating that Mungiki is influenced by demonic powers arguably allows political elites to connect widespread fears about malicious spiritual forces to actual people and the outward appearances with which Mungiki members are commonly recognized. These outward appearances are shaped by particular material objects, such as dreadlocks, snuff tobacco, red eyes which indicate possession by demons, and paraphernalia which are allegedly used in oathing rituals (Knighton 2009, 240; Smith 2011, 74; Phombeah 2003; Daily Nation 2011; Mwangi 2018). These material objects subsequently inform the mental images that people have of those who engage demonic spiritual forces. However, when dreadlocks and snuff tobacco started to serve as indicators for Mungiki membership during violent police crackdowns, many youth – Mungiki members or not – stopped having their hair in dreadlocks to avoid attracting the unwanted attention of security agencies (Rasmussen 2010, 309). These dynamics not only demonstrate that having dreadlocks was insufficient to determine Mungiki membership, but also that state attempts to identify and eradicate Mungiki are constantly complicated as people avoid the material signs that become associated with it. This complexity is further enhanced because Mungiki has never been a uniform movement. The ‘brand name’ Mungiki has been appropriated by different groups, and thrust onto others whom the state wished to suppress, so both the Mungiki movement and the demonic forces to which it is linked continue to resist a total definition (Smith 2008, 39; cf. Meyer 2015, 132).

Similar dynamics can be seen in relation to discourses about terrorism, although terrorism discourses only rarely focus on intimate circles of sociality of neighbours and family members; they focus instead on social relations on a national scale. Taking inspiration from U.S. rhetoric in the Global War on Terror, President Uhuru Kenyatta has described Al-Shabaab attacks as “senseless” or “barbaric” acts of violence, and as the work of “evil-doers” or “the devil” which poses “a threat to the nation” (Standard 2013a; Presidency of Kenya 2015; Ministry of Foreign Affairs 2019). Although Kenyatta stresses that “Islam is a religion of peace and tolerance”, he nevertheless links terrorism to Islam by arguing that radicalization “occurs in the full glare of day, in madrassas, in homes and in Mosques with rogue Imams” (Presidency of Kenya 2015). Such political rhetoric has coincided with increased surveillance of Muslims by Kenyan and Western intelligence services, and attempts of Kenyan security agencies to crack down upon suspected Al-Shabaab members, sometimes through extrajudicial violence. In relation to these developments, young Muslim men in Malindi told me that they avoid wearing particular kinds of Islamic dress (such as having a beard or wearing a kanzu tunic), which could attract unwanted attention from security agents, since they have implicitly come to be associated with radicalization or terrorism.

Several similarities between discourses about witchcraft and terrorism can be noted here. First, they provide a way of speaking about hidden enemies, since both witches and terrorists are thought to covertly plot violence that disrupts social relations from within. While witchcraft suspicions often arise within the intimate circles of sociality, discourses about witchcraft can also take shape on a national scale. It is on this national scale that discourses about terrorism are also often formulated. Second, while the presence of witches and terrorists is frequently feared and suspected, they generally remain invisible, although terrorists often reveal themselves when they execute violent attacks. Third, witches and terrorists are seen as evil figures who deserve extreme punishment, and they are regularly thought to be influenced by external malicious powers (such as the devil). Furthermore, ‘witchcraft’ and ‘terrorism’ are also externally ascribed labels, which conceal the economic and political motives of the accused. Finally, state actors often attempt to expose hidden evils and formulate suspicions that particular groups may involve themselves in witchcraft or terrorism. In response, people aim to evade being linked to these vices, by avoiding particular material religious forms that are commonly associated with witches and terrorists.

These observations demonstrate how discourses about witchcraft and terrorism set in motion complex dynamics of exposure and concealment, in which state actors and others aim to identify and eradicate hidden enemies. In these dynamics, particular material objects or religious forms may function as social stigma which link those who engage with them to witchcraft or terrorism (Goffman 1963, 58). These material objects and religious forms are therefore avoided by those who do not wish to be seen as (potential) witches or terrorists. In this way, discourses about witchcraft and terrorism may be understood to have a formative dimension, because the dynamics of exposure and concealment that they set in motion shape urban environments and the material religious forms that take place in them. Furthermore, these dynamics often impact Christians, Muslims, and Traditionalists in divergent ways, since terrorism is often primarily associated with Islam, and witchcraft with indigenous African religious traditions. In the rest of this chapter, I investigate how these dynamics shape the public expression of Christianity, Islam, and Traditionalism in Malindi in distinctive and uneven ways.

3 Traditional Healers or Witchdoctors?

During my fieldwork, I noticed how Christians, Muslims, and Traditionalists often understand misfortune such as illnesses in reference to an invisible spiritual realm, which can be maliciously influenced by envious neighbours and benevolently by various religious practitioners (see McIntosh 2009; Ciekawy 2001; Tinga 1998). These religious practitioners include waganga, a Swahili term frequently translated to English as ‘traditional healers’ by clients, and as ‘witchdoctors’ by those who condemn them. In the Malindi environment, many waganga are of Giriama ethnic backgrounds, who engage in healing and divinatory practices (uganga) that many consider to be closely associated with Giriama Traditionalism. These waganga can be consulted for diagnosing or treating problems that are explained in relation to spirits (mapepo or majini) and witchcraft (uchawi). Although many Giriama waganga incorporate Islamic elements such as the Qur’an in their ritual practices and may even self-identify as Muslims, Swahili Muslims often condemn traditional healing (uganga) as apostasy (ushirikina, from Arabic, shirk). Such criticism resonates with common Islamic critiques along the Swahili coast, which argue against “mixing” Islam with “traditional religions” (dini za kiasili) (Kresse 2018, 87). According to these critiques, such mixing would lead to unacceptable innovations in Islam (bidaa), such as the acceptance of witchcraft beliefs, traditional healing practices (uganga), spirit possessions, and/or the use of charms (Parkin 1985, 228; McIntosh 2009, 214; Kresse 2018, 87, 195). As unwanted innovations are often thought to originate from indigenous African religious traditions, many Muslims associate witchcraft – and ritual ways of diagnosing and healing it – primarily with Traditionalism. Regardless of their ethnic background, many Christians in Malindi similarly suggested that waganga depend on witchcraft or demonic forces, and that using their services is incompatible with being Christian.

Because of recurrent associations with backwardness and witchcraft, I noticed that many Giriama in Malindi refrained from publicly aligning themselves with Traditionalism, and instead often self-identified as Christians or sometimes as Muslims. Nevertheless, many Giriama (alongside other people from varying ethnic and religious backgrounds) often continued to secretively visit waganga. It must be noted, however, that despite being primarily associated with the Giriama and their traditional practices, fears about witchcraft continued to be shared by Christians and Muslims from various ethnic backgrounds as well. For example, a non-Giriama Christian interlocutor argued that “competition to be at the top of the food chain” in business and politics also finds expression in the spiritual realm, in which people “kill each other” through witchcraft. Similarly, a Swahili Muslim interlocutor admitted that some Muslims keep spirits (majini) with the aim of becoming rich by offering them the blood of relatives, a form of witchcraft that some Giriama also suspect wealthy Swahili Muslims of engaging in (see McIntosh 2009). Furthermore, many Swahili Muslims also visit their own religious specialists who provide treatments for problems that are understood in relation to the spiritual realm.³

I also noticed that besides refraining from self-identifying as a Traditionalist, many Giriama generally avoided public engagement with particular material objects, such as traditional dress, charms, ritual objects, and musical instruments, because occult forces are considered to appear through them. For example, one Giriama interlocutor explained that it had become problematic to take traditional percussive instruments – which are also used in divination practices – to church services. This interlocutor also said that traditional clothes are generally avoided by Giriama in Malindi in favour of Western-style dress, since traditional attire is seen as “devilish” and regarded as a possible indicator that the wearer might be a witch.

Such condemnation may explain why many waganga work secretively and rely on signs placed on trees and lampposts that indicate phone numbers rather than specific whereabouts for advertisement. This demonstrates how waganga have appropriated technologies such as phones and signposts to achieve a degree of public visibility in ways that simultaneously allow them and their locations to remain concealed. Furthermore, when I visited several waganga through the help of interlocutors, I found out that many waganga offer their services in the outskirts of Malindi, in huts or small palm-leaf-fenced spaces that do not display any advertisements to draw potential clients in. This demonstrates how many waganga navigate the politics of public visibility by using informal social networks to be located. I made similar observations when I visited Swahili religious specialists with the help of some of my interlocutors, who often receive clients behind closed doors in houses that do not provide advertisement.

While the location of waganga are usually known by those who live around them, one interlocutor explained that the people who visit waganga usually search for those with good reputations that are located far from where they live so that their visits will not be noticed by their neighbours. This is important not only because uganga is seen as incompatible with Christianity and Islam, but also because those who visit waganga run the risk of being accused of practising witchcraft (uchawi) themselves (also see Ciekawy 2001, 175–6). For example, one Giriama interlocutor explained how using waganga to attract clients to one’s business may be interpreted as witchcraft, as this would imply that one is drawing clients away from others. While engagements with waganga are thus always considered suspicious and potentially dangerous, many Muslims and Christians completely deny that uganga can be used benevolently, and associate all forms of uganga with spiritual danger, apostasy, and an idea of absolute demonic evil, even when it is used for protective or healing purposes (also see Parkin 1985, 228). In relation to such judgements, many Muslims, Christians, and Traditionalists in Malindi publicly distance themselves from waganga and the material forms associated with these practitioners, although many continue to engage with them secretively.

4 Witchcraft-Related Violence

Despite the secrecy that surrounds the topic of witchcraft in Malindi, it sometimes becomes an issue of public contention when suspicions that neighbours or relatives secretively practise witchcraft are settled in a violent manner. In these cases, state actors and Giriama elders both wish to intervene in the social dynamics that inform witchcraft-related violence, with the aim of reducing it. In this section, I analyse how public contestations around witchcraft-related violence further shape material forms of religious expression in Malindi.

During my fieldwork, I heard several stories about Giriama youth who would accuse elderly neighbours or relatives of being witches, after which they would kill them or chase them away so that they inherited their land and other family assets. Such witchcraft-related violence has led the county government of Kilifi to initiate a public campaign to encourage people to abandon “the backward ideologies of believing in witches” (Agoro 2019) through the slogan “old age is not witchcraft” (uzee sio uchawi), which they disseminate via public signboards and occasional public meetings. Through such messages, government officials in Kilifi county stand in a long tradition in which state officials reject belief in witchcraft as backward and superstitious.

Several interlocutors explained that state attempts to halt witchcraft-related violence are complicated by the Witchcraft Act, which was introduced under British colonialism in 1925 and remains largely unaltered today. While the Witchcraft Act criminalizes those “pretending to exercise witchcraft”, convicting suspected witches is often difficult because it requires tangible evidence of a crime that largely remains invisible (Luongo 2008, 41; 2011, 93). The Witchcraft Act also criminalizes accusing others of practising witchcraft and also criminalizes attempts to identify witches via “non-natural means” (i.e. uganga) unless the accusations are presented directly to government officials. Law scholar Katherine Luongo (2011; 2008, 36) stresses that these legal conditions make it virtually impossible to settle witchcraft suspicions through a court. In this situation, people who fear bewitchment sometimes resort to violence to settle suspicions, a problem that has challenged government authorities since colonial times. The observations of Luongo are consistent with my own. The court records show that the Malindi High Court regularly tries people who have allegedly murdered elderly relatives that were accused of practising witchcraft but is generally uninterested in verifying the accusations of witchcraft that inform these murder cases.⁴ Furthermore, while I met several accused elders who were attacked or chased away by younger relatives, I never met youth who publicly made accusations of witchcraft against specific elders, possibly because they feared government interference and criminalization if they openly did so. I also noticed that several Giriama elders in Malindi dyed their grey hair black, because grey hair is sometimes interpreted as a sign of being a witch, and this indicated that they were frightened of being accused of practising witchcraft.

Given the ineffectiveness of government measures in halting witchcraft-related violence, some Giriama elders in Malindi propose that government authorities should work together with ‘traditional’ Giriama practitioners to identify and ritually negate witchcraft, in order to quell persistent fears about witchcraft among Giriama youth. According to several Giriama elders, ‘traditional’ means of dealing with witchcraft accusations only very rarely resulted in suspected witches being killed, since they included ritualized ways of identifying witchcraft through divination (uganga wa kuvoyera) and oath ordeals (kiraho), after which identified witches were ritually cleansed and swore oaths to ensure that they would never practise witchcraft again (see Parkin 1991, 150–1). Yet, persisting associations between Giriama Traditionalism and witchcraft make many government officials reluctant to seek such cooperation. For example, during a conversation I had with a government official in Malindi, this official stated that he rejected cooperation with Giriama practitioners (waganga) to halt witchcraft killings. Instead, he argued that the Kenyan government “puts God first”, implying that the government cannot cooperate with ritual practitioners who engage spiritual forces other than the God of monotheist religions such as Christianity and Islam. Such judgements are perhaps unsurprising, because traditional means to quell witchcraft fears, paradoxically, acknowledge that witchcraft exists, while many state actors wish to distance themselves from such “outdated beliefs”.

The above analysis demonstrates how suspicions about witchcraft may become an issue of public contention when they are settled in a violent manner. In these dynamics, state actors aim to reduce witchcraft-related violence by denying the claims of Giriama youth that their elderly relatives may bewitch them. Instead, they conceptualize accusations of witchcraft as superstitious and resulting from adherence to a “backward ideology” that supposedly ‘real’ Christians would not believe in. Alternatively, Giriama elders propose to defuse suspicions about witchcraft by using traditional means to identify witches and offering ritual solutions. However, many government officials remain reluctant to work together with Giriama ritual specialists, because they associate Giriama ritual practice with “outdated” witchcraft-related beliefs, which in their view contradict the monotheism of Christianity and Islam. Such condemnation of Giriama Traditionalism not only complicates attempts to halt witchcraft-related violence, but also reinforces patterns in which Giriama Traditionalism is denied the same public recognition and material presence as Christianity and Islam because it is associated with witchcraft.

5 Kenyan Discourses on Violent Extremism

In relation to the on-going violence around witchcraft, Giriama elders have at times stated that the government is not doing enough to prevent the killing of elders (Standard 2019). For example, one Giriama elder argued that “if there is a bomb blast and one person dies, the whole government machine comes. But here, six or seven people are injured every week [after being accused of witchcraft]. Does this mean that they are not people?”. Indeed, besides the earlier mentioned public campaign, I have not heard of any systematic efforts by state authorities to prevent witchcraft-related violence.

Contrastingly, the Kenyan government actively cooperates with Western governments to deal with Al-Shabaab. This cooperation not only involves ‘hard power’ approaches by security agencies, aimed at intelligence gathering, arresting terrorism suspects, and military action against Al-Shabaab, but also ‘soft power’ preventive approaches to counter violent extremism. In another article (Meinema 2021a), I argue that while Western donors try to maintain a public stance of neutrality towards different religions, they perpetuate associations between Islam and terrorism by implicitly mobilizing a distinction between ‘moderate’ and ‘radical’ Muslims, which informs their counterterrorism activities in Kenya. Thus, contrary to in the 1950s when Mau Mau was described as a terrorist movement, British policy makers no longer fear that beliefs in witchcraft amongst Africans might inform terrorism and supposedly irrational violence. Instead, they rather see the global spread of ‘radical’ (or Salafi-inspired) Islam as posing a threat to security and their geopolitical interests because it is feared that it may inspire terrorism. Ironically, while Salafi-inspired reformist movements in Kenya often distance themselves from ‘traditional’ African practices in ways similar to those of many Christians and Kenyan state actors, some policy analyses now consider ‘Sufi’ Islam to be more peaceful, exactly because it is thought to embrace cultural elements shared by Muslims and non-Muslims alike, such as “witchcraft beliefs” (Møller 2006, 9; Republic of Kenya 2016, 18).

Despite the preoccupancy of Western donors with Islamic terrorist organizations such as Al-Shabaab, I noticed during my fieldwork that in Kenyan discourses about violent extremism, both witchcraft and terrorism continued to be seen as potential threats to peace and national unity. Whereas Kenyan state actors often envisioned accusations of witchcraft that arise within intimate circles of sociality as resulting from adherence to a “backward ideology”, they continued to recognize witchcraft as dangerous when they suspected that it posed a threat to peace on a national scale. Thus, while the Kenyan state criminalizes people who accuse others of practising witchcraft, state actors paradoxically continue to fear that particular groups engage in forms of witchcraft that may inspire politically subversive activities or violent extremism. In these cases, state actors actively strive to identify and eradicate those who engage in “witchcraft practices”, similar to the ways in which state actors deal with the threat of terrorism. To further explore this overlap between fears about witchcraft and terrorism, the rest of this section investigates how anxieties about witchcraft and terrorism intersect within Kenyan discourses on violent extremism. Additionally, I examine how within Kenyan understandings of violent extremism, witchcraft and terrorism are feared because they are envisioned as dangers to peace and unity on a national scale.

During my fieldwork, I noticed that two Muslim-led civil society organizations (CSOs) that receive funds via the UK Conflict and Stability fund to “build resilience” against “violent extremism” are aware that Western donors primarily focus on Muslims, who are perceived to be particularly susceptible to violent extremism. Consequently, these two Muslim-led CSOs often avoid addressing violent extremism altogether, because they fear that openly addressing this sensitive topic would bring about security risks or the further stigmatization of Islam. Instead, they espouse a view of Islam as a moral and peaceful religion, and thus a potential remedy against radicalization, which in their view rather results from a lack of religiosity amongst young people. In relation to these dynamics, I also noticed how some Muslims aimed to broaden discussions about violent extremism beyond a narrow focus on Islam, by mentioning Mungiki, political violence, the MRC, and criminal gangs as well as Al-Shabaab as examples of violent extremism (Meinema 2021a). Earlier in this chapter, we have seen that Mungiki is often associated with devil worship in public discourse. This wider understanding of violent extremism thus not only focuses on terrorism or Islam, but also resonates with concerns about demonic spiritual forces and witchcraft, which have informed Kenyan political imaginations since colonial times.

Furthermore, during various interfaith dialogues organized in Malindi, I observed how Muslim participants sought to align discussions about violent extremism with broader concerns about violence as related to anti-social attitudes and witchcraft. For example, one Muslim claimed during a dialogue meeting that Giriama youth who accuse elders of being witches should also be considered “part of extremism”. During another interfaith dialogue, a Muslim facilitator reasoned that violent extremism “has nothing to do with religion” but should rather be understood as a selfish, anti-social response to the experience of marginalization and grievances. Since this explanation resonates with conceptions of witches as driven by greed and jealousy rather than by a commitment to communal peace and well-being, it associates terrorism not so much with a ‘radical’ interpretation of Islam but rather with the kind of anti-social attitudes that are also commonly ascribed to witches. Noticeably, such a broader understanding of violent extremism is sometimes also held by Christians. For example, when I asked a Catholic schoolteacher in Malindi about her school’s mission to “reduce radicalization”, she argued that radicalization “is not necessarily Islamic” but should instead be associated with those “who do not identify with any religion” and engage in “devil worshipping” instead.

Furthermore, the previously mentioned action plan to counter violent extremism in Kilifi County similarly associates violent extremism not only with Al-Shabaab but also with “witchcraft practices” among the Mijikenda and the Mombasa Republican Council (MRC). The MRC is a coastal secessionist movement that was mainly active between 2010 and 2013 and sought to address coastal marginalization by mobilizing a regional coastal identity to challenge domination from “upcountry” (bara) (Kresse 2018, 23–5; Mwakimako and Willis 2014, 17–8). Besides having a secessionist agenda, various violent incidents have been ascribed to the MRC in newspaper reports. However, MRC-leaders have generally denied involvement in violence and have emphasized that they choose legal and peaceful means to pursue their political goals (Mwakimako and Willis 2014, 17; Willis and Chome 2014, 4–5).

The previously discussed action plan of Kilifi County similarly associates the MRC with Mijikenda Traditionalism by listing two sacred Kaya forests of the Mijikenda as MRC hideouts. Various newspapers have occasionally reported about MRC-related “oathing” taking place in Kaya forests, while other articles discuss “witchdoctors” being charged for administering MRC oaths, for which they use “witchcraft paraphernalia” (The Star 2019; Standard 2013b; Standard 2013c; Citizen 2015). It remains unclear how oathing in Kaya forests is connected to the MRC, especially since MRC leaders again deny any relation to such practices (Willis and Chome 2014, 4). Yet, concerns about MRC oaths and witchdoctors resonate strongly with anxieties about oathing practices that figured prominently in political imaginations of Mau Mau and Mungiki.⁵ This is another way in which anxieties about witchcraft continue to resonate in Kenyan understandings of violent extremism. The perpetuation of such fears in relation to the MRC arguably undermines the political legitimacy of this movement, even when MRC members choose legal and peaceful means to pursue their political goals and whether members engage in Mijikenda ritual practices or not.

In this section, I have sought to demonstrate how discourses about witchcraft and terrorism intersect in Kenyan understandings of violent extremism. Particularly, I showed how Muslims who participate in CSO programmes often aim to broaden debates on violent extremism by focusing on a much wider range of security issues to evade an exclusive focus on Islam. Such a wider understanding of violent extremism is also shared by some state actors and Christians, who not only mention Al-Shabaab as an example of violent extremism, but also political violence, Mungiki, criminal gangs, and the MRC. In this way, Kenyan understandings of violent extremism often do not exclusively focus on terrorism that is perpetrated by Islamic groups such as Al-Shabaab. These discourses are also informed by anxieties about witchcraft and ritualized oaths which have characterized Kenyan political imaginations since colonial times. Within discourses on violent extremism, however, “witchcraft practices” are feared not so much because they may harm relatives or neighbours within intimate circles of sociality but because – in a way that moves them to the national level – they are imagined to threaten peace and unity on a national scale. In the next section, I will demonstrate how the attempts of Muslims and Traditionalists to avoid being associated with violent extremism shape the material forms through which they organize themselves in Malindi.

6 Avoiding Material Religious Forms

Particular religious groups may not only aim to discursively dissociate their religious traditions from witchcraft or terrorism, but also strive to avoid engagement with particular material objects that inform the mental images that people have of witches or terrorists. In this section, I analyse how both Muslims and Giriama Traditionalists in Malindi attempt to avoid particular ‘things of conflict’ that are commonly associated with witchcraft or terrorism. In this way, I aim to demonstrate how the circulation of fears about witchcraft and terrorism has politicized and shaped the material ways through which different religious groups materially manifest themselves in Malindi.

During my fieldwork, one Muslim interlocutor explained that “Wahhabi-influenced” Muslims used to openly meet in Malindi, but no longer did so, because if you “align with Wahhabi, you are seen as an extremist, you are subjected to a lot of security risks”. Furthermore, during various Friday sermons, semi-public meetings in mosques, and other public meetings, I noticed that Muslim leaders in Malindi generally refrained from publicly airing political dissent, even though criticism of the government has had considerable popular appeal in mosques in coastal Kenya in recent decades, resulting in a “very public politics of overtly Muslim discontent” (Deacon et al. 2017, 12). Given the previously described securitization of Islam, Muslims in Malindi have arguably stopped publicly airing political criticism, since this could easily confirm suspicions that they may have become radicalized.

Some of my observations indicate that Muslims in Malindi have also toned-down discussions and polemical debates with their Christian neighbours. During my stay, I heard several stories about competitive open-air debates (termed mihadhara) that took place in Malindi in the recent past in which Muslim preachers had engaged in polemical theological discussions with Christians while referring to both the Quran and the Bible. Several interlocutors suggested that mihadhara in Malindi were often initiated by Muslims, who invited Christians to these events, just like elsewhere in East Africa (Wandera 2015, 28). Yet, in relation to fears that mihadhara may provoke conflict or heighten tensions between Muslims and Christians because of the polemical tone of these meetings, the Coast Interfaith Council of Clerics called on preachers and state authorities to limit such preaching events, which has led to legal limitations on polemical open-air preaching in the Mombasa region (Petersen and Wandera 2015, 7). Various interlocutors told me that mihadhara were no longer organized in Malindi in relation to similar fears. Indeed, one Christian interlocutor stated that public polemical preaching once sparked tensions in Malindi, which almost led to a clash between Muslim and Christian audience members. I never witnessed any Muslim-initiated public preaching taking place in Malindi, which confirmed the accounts of my interlocutors. These observations suggest not only that many Muslims in Malindi avoid forms of public religious expression which could confirm suspicions that they harbour resentment towards Christians or the Kenyan state but also that state authorities intervene in mihadhara because they are perceived to threaten peaceful religious coexistence.

Broader understandings of violent extremism as related not only to Al-Shabaab but also to the MRC and the witchcraft practices that are associated with the Mijikenda also influence how Giriama Traditionalists materially manifest themselves in the urban environment of Malindi. Suspicions about politically subversive forms of witchcraft sometimes extend to the material forms through which Mijikenda Traditionalists become identifiable. For example, Amason Kingi, the governor of Kilifi County, publicly stated in 2019 that Mijikenda attire has been wrongfully used by security agencies to identify MRC members.⁶ I also noticed that Giriama elders in Malindi are aware that security actors sometimes associate Mijikenda Traditionalism with witchcraft, subversive political practices, and the MRC. In opposition to such associations, Giriama elders often preach peace when they organize cultural events in Malindi, and regularly organize peace walks during which they advocate peace and national unity. Such physical demonstrations of loyalty to the nation are relevant, because several Giriama elders were arrested when they prepared to participate in a peace walk while wearing traditional attire, under suspicion that they were gathering for a MRC-meeting (Kilifi News 2015). This shows how particular material forms that are commonly associated with Mijikenda tradition and ritual practice have implicitly come to inform the mental images that state actors have of those who allegedly engage in politically subversive activities, which some suspect to be informed by witchcraft practices.

Taken together, these observations show how the circulation of fears about witchcraft and terrorism within Kenyan discourses about violent extremism affects the ways in which Muslims and Giriama Traditionalists publicly express themselves in Malindi. Both Muslims and Giriama Traditionalists engage in a (self-)restriction of public religious expression by avoiding particular material religious forms that are commonly associated with witches or terrorists. While Giriama Traditionalists often promote peace when they wear traditional Mijikenda attire during public events, Muslims generally avoid engaging in public competitive preaching, wearing particular kinds of Muslim dress, or publicly expressing political discontent. This is crucial, because these material forms have come to inform the images that policymakers and state actors have of Muslims and Giriama Traditionalists who potentially engage in various forms of violent extremism and require policy or security intervention.

7 Pentecostalism in Malindi

The (self-)restriction of public religious expression and cautious interaction with various material religious forms that I observed among Muslims and Traditionalists have not similarly affected Christians in Malindi. For example, while Muslims stopped organizing competitive public preaching events, Christian evangelization campaigns continue to take place in Malindi. Such ‘crusades’ are organized by Malindi-based Pentecostal churches, and sometimes also by various international ministries. Although crusades often avoid predominantly Muslim neighbourhoods, they regularly involve public stages, preaching through PA-systems, evangelizing in the streets, and playing loud worship music, making it difficult to avoid them completely.

Highlighting a competitive attitude towards Muslims, some Pentecostal preachers in Malindi also associate Islam with Al-Shabaab and demonic spiritual forces. For example, during an interfaith dialogue, a Pentecostal pastor sparked controversy when he explained that he avoids Muslims in the same way that he avoids witches, since Muslims could be “linked to this Al-Shabaab”. During one conversation, another Pentecostal preacher argued that Muslims “use religious demonic forces” and also said that the “Mijikenda religion is more close to witchcraft than religion”. The pastor explained that he hoped to succeed in overcoming the challenge of Muslim-Christian competition in Malindi, which would make Malindi a “peaceful” and “God-fearing town” in which “Islam would lose its grip”. During a worship meeting, he similarly preached about how “Malindi would feel our impact”, mentioning the transport industry as a sector that Christians would come to occupy. Since Muslims currently own much of Malindi’s public transport industry, this preaching is implicitly competitive towards Muslims.

Despite the competitive attitudes of some Pentecostal pastors, Muslims rarely openly complain about the growing public presence of Pentecostal Christianity in Malindi. This signifies how many Muslims have developed “techniques of inattention” towards the sometimes provocative and strongly audible presence of Pentecostalism (Larkin 2014), even while Muslims have toned down their own competitive preaching. Nevertheless, some Muslim interlocutors privately grumbled that Pentecostal worship is too long or too loud, especially when it takes place close to residential areas where may Muslims live. When formulating such complaints, these Muslim interlocutors pointed out that in the past, churches were respectfully built on the edge of town, so that Muslims would not be bothered by them.

How can we explain this situation, in which some Pentecostal churches continue to organize crusades in Malindi while it is unthinkable that Muslims could organize similar public preaching events, especially if they were provocatively labelled ‘jihad’? This difference can arguably be understood in relation to the broader inclination of churches in coastal Kenya to refrain from raising political criticism, and to instead present themselves as peaceful and loyal to the Kenyan nation (cf. Deacon et al. 2017). Consequently, Christianity is generally not associated with witchcraft and terrorism in Kenyan political imaginations, or seen as a potential politically subversive threat. In this situation, polemical public preaching events organized by Muslims are banned because they are perceived to threaten peaceful religious coexistence, while state actors generally do not see polemical Pentecostal crusades as requiring similar state intervention.

On several occasions, I noticed that Pentecostal preaching not only demonized Islam, but also Traditionalism. For example, a Giriama Pentecostal pastor once explained that “the Christian way is the only way” to deal with malicious spiritual forces, and he suggested that both Islam and Traditionalism engage and appease spirits instead of getting rid of them altogether. He illustrated his claim by showing me videos of exorcisms of Islamic spirits (majini). When I later visited his church, I indeed witnessed how majini were exorcised from churchgoers, and the exorcism of both uganga (traditional healing) and uchawi (witchcraft) from an elderly Giriama woman. Since the pastor of this church associated both traditional practices (uganga) and Islam with malicious spiritual powers, he suggested that spiritual well-being and peace can only be achieved when one breaks completely from these religious traditions by invoking the Holy Spirit.

Although Pentecostalism promises deliverance from malicious spiritual forces, this promise of spiritual protection paradoxically confirms and thus perpetuates fears for malicious spiritual forces, such as witchcraft, majini, and ‘devil worship’ (Rio et al. 2017, 12). Ironically, I noticed that despite their increasing presence in Malindi, Pentecostal churches also do not escape being associated with malicious spiritual forces either. In several instances, I noticed that some Giriama and Swahili interlocutors express suspicion that particular Pentecostal churches were involved in devil worship, for example when pastors use loud distorted voices during preaching, which one interlocutor interpreted as an indicator that they were possessed by malicious spirits. This shows how in the diverse religious field of Malindi, people from various ethnic and religious backgrounds share fears of malicious spiritual forces, which are conceptualized through the Pentecostal language of devil worship.

In this section, I have analysed how Christianity in Kenyan political imaginations is not often associated with terrorism or forms of witchcraft that threaten to disrupt peace and unity on a national scale. Consequently, state authorities generally do not see public preaching events organized by Christians as a threat to peaceful religious coexistence in the same way as public preaching that is initiated by Muslims, even though Pentecostal preaching is sometimes polemically aimed against Islam and Giriama Traditionalism. On some occasions, Pentecostal pastors accuse Islam and Giriama Traditionalism of engaging or appeasing malicious spiritual forces. Paradoxically, however, such Pentecostal preaching recognizes that malicious spiritual forces exist, even though it aims at breaking completely with those forces by invoking the Holy Spirit. In this way, Pentecostal preaching perpetuates fears of witchcraft and other malicious spiritual forces that can be found among Christians, Muslims, and Giriama Traditionalists alike. Yet, while Giriama Traditionalism is often perceived to be intrinsically connected with witchcraft in the eyes of Muslims and Christians, Pentecostalism commonly continues to be seen as a powerful solution to witchcraft, including among many Giriama who have become Christians. Since Pentecostal Christianity is generally not associated with terrorism, and is often seen as a solution to witchcraft rather than being associated with it, Pentecostal Christians do not have to engage in the (self-)restriction of public religious expression that I observed among Muslims and Traditionalists. I argue that these dynamics privilege the public expression of Christianity in Malindi.

8 Conclusion

In this chapter, I explored how discourses about witchcraft and terrorism set in motion complex dynamics of revelation and concealment which politicize and shape the material ways in which various religious groups manifest themselves in the urban environment of Malindi. In these dynamics, particular material religious forms become ‘things of conflict’, which may link widespread fears about witches and terrorists to actual people and their outward appearances. Consequently, many Giriama evade public engagement with traditional objects, such as traditional attire, musical instruments, and traditional healing practices (uganga), to avoid being associated with witchcraft, even as many Giriama continue to secretively engage with them. As a result, Giriama Traditionalism generally has a relatively low public profile in Malindi. Since Islam is often associated with terrorism, many Swahili Muslims similarly refrain from wearing particular kinds of Muslim dress or having a beard, competitive public preaching, and the open expression of political discontent. This is relevant, because these forms of religious expression have implicitly come to serve as indicators that one may be susceptible to radicalization, and consequently, they could attract unwanted policy or security interventions. In relation to these dynamics, many Muslims in Malindi also strategically avoid the sensitive topic of terrorism when they participate in Western-funded programmes to counter violent extremism. Yet, while Western donor policies generally focus on countering violent extremism among Muslims, some Muslims aim to broaden discussions on violent extremism by mentioning not only Al-Shabaab, but also the MRC and accusations of witchcraft among the Giriama as examples of violent extremism. In this way, they arguably aim to align debates about violent extremism with fears about the potentially violent and politically subversive threat of witchcraft, which have informed Kenyan political imaginations since colonial times. Within this broader understanding of violent extremism, witchcraft continues to be primarily associated with Traditionalism. This also impacts the ways in which Giriama Traditionalists assume a public presence in Malindi, since they often physically demonstrate their loyalty to the Kenyan nation when they organize public activities to avoid being associated with the MRC, which is suspected of engaging in politically subversive activities that are inspired by “witchcraft practices”.

While Christianity does not remain entirely free from suspicions of engaging demonic spiritual forces, it is generally not associated with the potentially subversive threats of witchcraft and terrorism within Western donor policies or Kenyan political discourse. Since Christians do not have to engage in the (self-)restriction of public religious expression in ways that are similar to those of Muslims and Traditionalists, Christianity has assumed a growing public presence in Malindi at the expense of Islam and Giriama Traditionalism. Pentecostalism in particular is an increasingly appealing religious alternative for many Giriama in Malindi, since it promises deliverance from witchcraft and other malicious spiritual forces, which in Pentecostal discourses are sometimes associated with Islam and traditional healing practices (uganga). Since Christians are only rarely suspected to be witches or terrorists, I thus argue that the circulation of fears about witchcraft and terrorism privileges the public expression of Christianity in Malindi. This perpetuates a colonial pattern in which Christianity is generally seen as a civilized religion that is compatible with modern statecraft. Simultaneously, religious minorities continue to be looked at with suspicion, not only because they may perpetuate “outdated” witchcraft beliefs, but also because it is feared that they may inspire forms of terrorism or witchcraft that threaten peace and unity on a national scale.

Acknowledgements

Foremost, I am thankful to my interlocutors in Malindi, whom I have anonymized in this chapter. I have also benefitted from conversations with Birgit Meyer, Lucien van Liere, Peter Geschiere, and Benjamin Kirby, who read earlier versions of this chapter and provided helpful feedback. I also received useful comments from participants during the seminar ‘The Things of Conflict; A Religious Studies Perspective’ that Lucien van Liere and I organized in 2019, and from the anonymous reviewers of this chapter. The PhD research on which this chapter is based was funded by the Dutch Research Council (NWO, grant number 406-15-072), which is gratefully acknowledged. All mistakes remain my own.