Over the last two decades, bitter confrontations have taken place among Muslims in East Africa. The main dividing line runs between defenders of a well-established repertoire of ritual practices (such as the maulidi, a recitation of poetry to celebrate the Prophet's birthday), on one hand, and reformers who dismiss such practices as un-Islamic ‘innovation’ on the other (Ludwig Reference Ludwig and Westerlund1996; Lacunza Balda Reference Lacunza Balda, Rosander and Westerlund1997; Becker 2006). The defenders of established ritual practice are often, though not always, associated with Sufism. Their practices are the outcome of a long process of exchange with the wider Indian Ocean, and as they evoke historical precedent to justify their practices, they can be spoken of as traditionalists.

The antecedents of these debates reach back into the colonial period, when a number of influential shehe among the old Muslim congregations of East Africa's coastal towns began to reject as bida’, ‘innovations’, those ritual practices that were not traceable directly to the Qur'an or hadith (sayings of the Prophet). Kai Kresse has recently traced the transmission of critiques along these lines between shehe in Lamu, Mombasa and Zanzibar back to the 1940s. At that time, though, these differences by and large remained limited to erudite discussion among highly educated urban notables (Bang Reference Bang2003; Kresse Reference Kresse2007).

In a new departure, similar criticisms have become the subject of popular debate over the last three decades, involving many people with no connection to the old maritime scholarly networks. Both sides in the confrontation are divided into multiple organizations and ideological currents, but on the whole the reformists tend to be younger. They are inspired by local clerics who, since the 1970s, have attended training courses in Saudi Arabia sponsored by the Saudi government (Becker Reference Becker2008: 241–75). Many of the people now participating in the debate have little education, and their stances are not so much reflectively explicated as practised and performed. Moreover, as criticism of the region's post-colonial governments has become linked to reformist discourse, the differences have acquired political overtones.

Both the wider Tanzanian public and Western observers tend to focus on characteristics broadly shared by these reformist currents: their emphasis on austerity and piety, condemnation of innovations and insistence on sticking to the letter of the Qur'an (Rosander and Westerlund Reference Qadiria1997; de Waal 2004). Moreover, the reformists’ emphasis on the need for Muslims autonomously to maintain an appropriate religious way of life connects with the discontent of many, including traditionalist, Tanzanian Muslims over their relative marginality in higher education and national politics. These issues have been the subject of confrontations that sometimes turned violent. Book-length statements by Muslim and Christian intellectuals indicate that both sides ascribe to them the potential for major political conflict (Njozi Reference Njozi2000; Mbogoni Reference Mbogoni2004). The Saudi connection, meanwhile, highlights the indebtedness of Tanzania's Muslim reformists to inspiration from abroad, which also includes the Iranian revolution, Sudan and the Pakistani Tabligh (Pouwels Reference Pouwels1981; Lacunza Balda Reference Lacunza Balda, Rosander and Westerlund1997; Becker Reference Becker2008).

The present study examines popular debates among Muslims in Tanzania – as reflected in media such as pamphlets and audiotapes, as well as interviews – to focus instead on a topic very salient in these debates, but rarely addressed in the existing literature: conflicting views of funerary ritual. It aims to show that, in this regard, reformists’ criticism constitutes an intervention in, and is to some extent shaped by, a secular, ongoing process of ritual change. Moreover, this change does not concern simply the ‘purification’ of Muslim practice, as both Muslim reformists and Western academics would tend to put it. Rather, I argue that the debate concerns the place of the dead in social relations, and hence ultimately the shape, reach and cohesion of society. This also means that Islamization or Islamic reform is but one of a number of factors that Tanzanians concerned about funerary ritual perceive to be at work.

These place-specific doctrinal debates counteract the interpretation of Tanzanian events as an instantiation of a unified, homogenizing wave of global Islamism. Bernard Lewis and Samuel Huntington, among others, have interpreted the many current expressions of Islamic radicalism as signs of a generalized reaction of ‘Islamic civilization’ against ‘the West’ and ‘Western’ modernity (Huntington Reference Huntington1996; Lewis Reference Lewis2004). Olivier Roy, in a similarly global argument, explains the current wave of Islamism as a mediatized intellectual fad, piggy-backing on the new-media revolution (Roy Reference Roy2004). In fact, overseas influences that might be described as global are undoubtedly important in the present case, but so are factors that need to be traced to the local and regional levels.

To highlight these place-specific factors does not mean to deny the importance of trans-regional influences. Rather, it helps appreciate that the success of Islamist ideas does not rest on their instant identical reproduction in different contexts, but instead on their ability to adapt so as to speak to different contexts. It enables the observer to place events in their African as well as their Muslim context. Moreover, ‘place-specific’ is not the same as ‘local’: it includes influences that travelled between different parts of the African continent and the Western Indian Ocean. The aim is to keep in focus a point that has been noted by a number of observers of the current wave of Islamic reform: the polyvalence of the only seemingly cut-and-dried Islamist doctrines; the fact that they work at different levels simultaneously, and in different ways in different places (Eickelman and Piscatori Reference Eickelman and Piscatori1996; Qasim Zaman Reference Qasim Zaman2007).

The importance of funerary ritual in religious debate in itself connects Tanzanian Muslims to a long strand of Islamic history. Leor Halevi has recently argued for the importance of funerary ritual in shaping Muslim identities at the beginning of Islamic history, and it has long formed part of diverse place-specific praxes (Halevi Reference Halevi2007; see also Smith and Haddad Reference Smith and Haddad1981). Moreover, debates over funerary ritual have also emerged in the course of dissent between traditionalist Muslims and reformers in other parts of the world, including Indonesia and Central Asia (Abashin Reference Abashin2006; Federspiel Reference Federspiel1970).

Yet specific concerns over funerary ritual in Tanzania also highlight the connections of the current debate to changes in longstanding indigenous ritual practice, with its wider social and cultural context (Parkin Reference Parkin1972; Metcalf and Huntington Reference King, Fiedler and White1991). The reformists, despite their undeniable indebtedness to doctrines learned abroad and their frequent condemnation of ‘innovation’ in Islamic practice, also participate in these continual, creative and place-specific processes. The significance of the trans-regional doctrinal exchanges examined, then, arises from their interaction with a very dense and specific societal context: Tanzanians’ negotiation of changing forms of personal interdependence in the colonial and post-colonial world (Haykel Reference Haykel2003).

ISLAM, SOCIAL CHANGE AND FUNERARY PRACTICE DURING THE NINETEENTH AND TWENTIETH CENTURIES

Since at least the late nineteenth century, Islamization in Tanzania has been implicated in a broader process of change in funerary ritual. Through this process, the responsibility for the dead has passed from descent groups to a less narrowly defined Muslim congregation. Moreover, the nature of the relationship – in particular, whether the living need the dead or vice versa – has come into play. These large-scale changes, though, arose slowly from the gradual adoption of Islamic notions and practices. The adoption of Muslim ritual did not, in this process, imply the active disowning of indigenous precedents. Rather, ritual innovation was treated as cumulative.

The evidence for this process adduced here derives from travel accounts of the hinterland of the southern Tanzanian and Mozambican coast, and missionary sources as well as interviews from south-eastern Tanzania. It does not provide a great deal of detail, and different emphases would be found in other parts of the East African interior, but the outlines of the process are clear. The following is a description of a clearly non-Muslim grave in the hinterland of the southern Tanzanian coast, dating from 1866:

This funeral site illustrates why present-day respondents tend to claim that their non-Muslim forebears ‘threw their dead into the bush’: the inhabitants had abandoned the settlement to the wilderness because of the death (Weule Reference Weule1909: 194–201). Such small-scale migration in response to a death was widespread among the fairly thin and mobile populations of eastern and south-eastern Africa.

It is not clear when exactly this form of migration ceased, though it was obsolete by the end of the First World War.Footnote ¹ It is clear, though, that its disappearance was connected both to the secular social and political changes of the late nineteenth and early twentieth centuries, and to the growth of Muslim and Christian congregations. During the nineteenth century, a rapidly expanding long-distance trade in ivory and slaves created new security needs as well as new forms of dependency and political control. Villagers sought safety from slave hunters in larger settlements and local leaders exploited trade connections to set themselves up as ‘big men’ who protected their own followers and persecuted those of their competitors (Alpers Reference Alpers1975; Wright Reference Wright1993; Becker Reference Becker2004).

In this context, evidence for changing burial practices goes back as far as the 1870s. The British traveller J. F. Elton, visiting a large, tightly governed village of the kind that had come into being under slave-trading Yao ‘sultans’ in Malawi and northern Mozambique, remarked that its inhabitants had retained the habit of abandoning homes in which somebody had died (Elton Reference Elton1879: 287–91). But rather than moving away after burying the deceased at their former home, the villagers now tore down their old homes over the graves, planted crops on the remains, and built new houses right next door. Elton compared the resulting townscape to choppy water, but this apparent spatial disorder reflected the greater permanence of these settlements and their inhabitants’ dependence on the strongmen who controlled them.

A generation later, in 1906, the German anthropologist Karl Weule, travelling upcountry from Lindi on the southern Tanzanian coast, deplored a lack of genuine ‘pagan’ burial sites.

In fact, the markings on this grave constitute a mixture of indigenous and Islamic practice. The frame and headstones recalled centuries-old Muslim graves on the island of Kilwa Kisiwani, where Islam had long been indigenous. The hut above the grave of local rulers, meanwhile, is attested to in older evidence by Portuguese missionaries among ‘pagan’ people in northern Mozambique, whence Makua and Yao-speakers had come to present-day Tanzania (Alpers Reference Alpers, Ranger and Kimambo1971).

Weule suggested that the ‘fashion’ for Islamicate graves prevailed particularly among notables (former ‘big men’ recently disempowered by German colonialism). This corresponds to the fact that at this time access to Islamic ritual knowledge was one of the privileges of members of the network of ‘big men’ who lived off the long-distance trade routes. But the graves – sometimes Islamicate in appearance – of some of the notables Weule interacted with (and of others he did not meet) are still being maintained and visited today.Footnote ² This practice is not considered in any way Islamic: rather, Muslims and Christians alike discuss it in terms of indigenous ‘tradition’. The Muslim influence on the architecture of these graves has been recuperated into a notion of ‘African’ practice.

During the early decades of the twentieth century, Muslim allegiance lost its elitist connotations and became instead a sign of equality among villagers (Becker Reference Becker2008). This shift was part of a far-reaching recalibration of social relations under Indirect Rule. In this context, Islamic ritual provided a new focus for social interaction. Thus its provisions for funerary ritual recur in informants’ narratives as one of the fundamental advantages of the new religion. Of some of the earliest proselytizers in the rural south-east, a Somali family that had settled in the town of Liwale around 1900, it is said that they told their neighbours and customers, ‘Don't just bury each other like chickens, you just throw each other [your dead's corpses] away, religion does not allow this.’Footnote ³ Present-day informants emphasize the ordered, dignified character of Muslim funerary ritual and its origins in divine revelation.

Respondents who opposed Muslim funerary practice to ‘throwing the dead into the bush’ referred to the fact that Muslim graves were integrated into human settlements rather than segregated in the bush. In rural areas, graveyards continue to be sited on the fringes of the village, but they constitute a shared facility and graves are tended carefully. In today's towns, moreover, Muslim graveyards are visible, tended and interspersed with the homes of the living, quite unlike the isolated graves noted by early travellers. Whereas the earlier habit of moving away from a residence whose inhabitant had died left gravesites in a liminal area between wilderness and settled space, Islamic graveyards and Muslim burial allowed for the recovery of graves from the wilderness.

Some Muslim burial practices were adopted even among non-Muslims. In 1942, a German missionary gave the following account of burial among ‘heathen’ Makua:

While the roof and the cloth on the grave recall the description by Kirk, the use of white cloth for a shroud, known as sanda in Swahili, is typical of Muslim funerals, as is the niche within the grave (Knappert Reference Knappert2001: 34–7). The missionary even notes that the corpse is often placed in the niche in such a way that it faces north, in other words, in the direction of Mecca, without making the connection with Islam.

It is nevertheless unlikely, given missionaries’ concerns over Islamic competition, that he would not have noticed if the people performing this burial had actually professed Islam. His account forms evidence of the selective adaptation of practices of coastal Muslim origin into the funerary practice of rural ritual congregations. The use of the niche is also described as required among Muslims on other parts of the East African coast (Knappert Reference Knappert2001). Later on, Christians in the region too were reported to use these niches within graves: it occurs in Muslim, Christian and local funerary practice, and followers of all three religious traditions treated it as part of their own ritual.

Notwithstanding this passage of specific ritual acts across the boundaries of different religious praxes, followers of the new religious congregations also began to evoke burial as a mark of distinction by the 1930s. This applied not only to the dismissal of funerals according to indigenous rites as ‘like throwing rats into the bush’. In the 1930s, an Anglican missionary among the by now predominantly Muslim population of Tunduru District, far inland from Lindi town on the Tanganyikan–Mozambican border, remarked on his experiences:

Hence mistrustful Yao Muslims attended the first Christian funerals at Tunduru in great numbers. They came away reassured by their decorousness, yet:

Funerary practice was both a site of exchange between religious praxes, and of the assertion of separate religious allegiances.

Followers of every religious tradition, then, spliced together funerary practices of different religious derivations, while asserting the integrity of these practices as the sign of a specific religious allegiance. Notwithstanding the kind of mistrust described by the Anglican Lamburn for Tunduru's Muslims, there is no evidence that funerary ritual at this time (during the mid-twentieth century) became the object of conflict, whether among co-religionists or between followers of different faiths.

The importance that funerals acquired as a token of social cohesion among Muslims is evident also in the way Tanzanian Muslims reinterpreted the Sufi practice of ziara. Literally ‘visit’, this ritual typically, in the Middle East, involves the visitation of the grave of the founder or important leader of a Sufi brotherhood. Among Muslims in rural Tanzania, though, ziara came to denote the visitation of the graves of believers by living Muslims, led by a sheikh.

During the 1940s, the ziara were threatened by divisions among urban Muslims over another ritual issue, the use of drums in worship. The dispute led to separate ziara being held by the supporters and opponents of their use in Lindi town. Notables in the town were sufficiently concerned about this split to appeal to the Provincial Commissioner for mediation, and even to Dar es Salaam when the Commissioner proved unhelpful. With the settlement of the dispute over drums, the ziara were reunified.⁵

Through this process of rituals crossing boundaries and boundaries being asserted, burial grounds have become ritual sites integral to settlements, and elements of practice have gone back and forth between Muslim and indigenous religious realms. The ability to hold funerals together is considered an essential characteristic of a functional settlement. One of the most common ways for people of different religious persuasions or ethnic origins to emphasize the fact that they get along is to say that tunazikana; ‘we bury each other; we attend each others’ funerals’. To show respect for the dead, it appears, is the absolute minimum of regard to be shown for each other.

Moreover, the relationship between the living and the dead is not lived out only at grave sites. Here as elsewhere in Africa, the living continue to have an active relationship with their ancestors. In the rural region adjacent to the southern coast, the practice of making donations of flour and beer is documented by missionaries and administrators for the earlier twentieth century, and still well remembered today. Donations of this kind, left typically at trees, can be traced through Portuguese sources as far back as the seventeenth century among closely related populations in northern Mozambique (Alpers Reference Alpers, Ranger and Kimambo1971).

When German officers arrived in Mikindani, the next coastal town to the south of Lindi, around 1900, they found Muslim notables in the town making symbolic offerings of strips of cloth at a baobab tree near the town (Berg Reference Berg1901). Present-day interlocutors from old, thoroughly Muslim families in Lindi describe a way to share one's troubles with the ancestors by telling them on the doorstep of one's home, a liminal place where the deceased can hear the living.Footnote ⁶ Further up country, the term once used for the trees at which sacrifices were offered is used by present-day Muslims to describe a shared meal accompanied by prayers and Qur'an readings, intended to invoke the goodwill of the ancestors for troubled or ailing family members.⁷

Thus change in funerary ritual and in the way the living address the dead is not only a recent phenomenon. It has been connected over the last 130 years to changing forms of settlement and social control as well as to religious innovation. While propitiation of the dead has become a more domestic concern, funerary practice has been placed more squarely in the public realm, with the greater visibility of graves and the involvement of Muslim congregations rather than lineages. These changes have been slow and cumulative. Members of the new Muslim congregations did not conceive of the adoption of Muslim funerals as a radical departure from indigenous religious practice, but rather as an improvement on it. This inclusive approach to religious change, where specific practices can be exchanged and added, contrasts with present-day reformists’ insistence on a comprehensively and consciously Muslim way of life.

THE DEVELOPMENT OF CONFRONTATIONS OVER FUNERARY RITUAL BETWEEN REFORMISTS AND TRADITIONALISTS

The movement of Islamic reformists that has developed in Tanzania since the late 1970s is far from unified and combines disparate political and religious foci. Nevertheless, there are clear patterns in the way they have challenged established Muslim congregations. Both in Dar es Salaam and in provincial towns, confrontations between Muslim factions have centred typically on control over mosques. They started with reformists objecting to the way the daily and Friday prayers were performed: in particular, to the posture used by traditionalist worshippers and the additional prayers on Friday. They quickly widened into recriminations over each others’ level of education, use of funds and general conduct. Sometimes reformists sought to take over the committees through which mosques are run, or challenged traditionalists’ ownership in court.

Disagreement on funerary ritual forms part of this more general clash over ritual practice. In some provincial towns, particularly Lindi and Mtwara on the southern coast, it quickly became the main focus. Here, the reformists have become known colloquially as Chimumuna: a word taken from the indigenous Makonde language, and commonly translated as ‘quietly, hastily, furtively’. It is intended as invective against the reformists’ practice of walking quickly and silently behind the bier during funerals, and departing quickly from the grave once it is filled in.

Such ‘silent’ funerals caused a sensation in Lindi in the early 1990s. Their initial supporters were mostly young men with connections to more metropolitan towns on the northern coast, whose lifestyle as travelling petty traders was considered less than respectable by established Muslim notables. Over the next fifteen years, a confrontation on the typical pattern ran its course. The reformers found support from a hitherto relatively marginal member of the older generation of sheikhs. Their challenge to the politically well-connected leading sheikh of the town took on political overtones. A lengthy confrontation over control of a mosque in the town was resolved in favour of the traditionalists, while the reformists have started building separate mosques.

Although all groups still use the same graveyards, the facilities for preparing the corpse for burial have effectively been segregated in Lindi: the relatives of a traditionalist Muslim would not want to use the outbuildings of a mosque perceived as Chimumuna-dominated. There are fairly regular deaths among the young adults where the Chimumuna stance is most common, due to both AIDS and the dangers of the road journeys on which many of them depend for a living. Therefore Chimumuna funeral processions can regularly be seen moving through town at a sort of canter, the bier – as with traditionalist Muslims, an upturned bed – followed exclusively by young men. Yet the mutual avoidance at funerals has not become a general principle. There is still ample opportunity for visiting and condolences during the three-day-long gathering surrounding the actual burial. Most Chimumuna have non-Chimumuna relatives, and there is no way to exclude them from these encounters.

Reformists do not necessarily see the focus on funerals as programmatic for their project. In fact, one of the more politically minded leaders in Dar es Salaam has dismissed the debates on funerals in the south as a distraction. Nevertheless, in places like Lindi uneasiness about the ritual divide runs deep. Traditionalists take the reformists’ criticism as a rejection of their entire tradition of learning, including the foundational Sufi sheikhs of the twentieth century in whom traditionalists take great pride. They describe the rejection of their funerary rituals as akin to a return to paganism, to the ‘throwing away’ of the dead. Reformists, in their turn, can be very blunt in their rejection of traditionalist ceremony. During an informal, but quite public conversation, from behind the counter of his corner shop, a reformist sympathizer asserted pithily that the corpse of a human was, ‘no more than a slaughtered goat’; something to be disposed of hygienically, yet without further meaning.Footnote ⁸ It is worth taking a closer look at the doctrines from which this assertion was derived.

RUPTURE VS SOLIDARITY: THE FOCUS OF DOCTRINAL DEBATES BETWEEN TRADITIONALISTS AND REFORMISTS

The acrimonious tone of debate notwithstanding, traditionalists and reformists share some starting points. People on both sides accept the notion of the ‘torture of the grave’, according to which deceased sinners are punished (and good Muslims rewarded) already before judgement day. Again, this proposition has well-established antecedents among old Muslim congregations in the Middle East, where it is documented since the early centuries of Islam (Smith and Haddad Reference Smith and Haddad1981; Halevi Reference Halevi2007).Footnote ⁹ As Charles Hirschkind has shown, it is also a characteristic concern of popular reformist Muslim discourse in contemporary Egypt; Tanzania thus appears in tune with regions considered part of the heartlands of Islam (Hirschkind Reference Hirschkind2006).

Given the peril the dead are perceived to be in, the obligation of the living to mitigate their fate is also undisputed between the two sides. But the way this is to be done, and the extent to which the living are able to do so, is the subject of debate. As one would expect, the reformists put much narrower limits than the traditionalists on the ability of the living to help. In the contrasts between the two sides’ views on this matter, the differences between their visions of what it means to live as a Muslim become particularly clear. It is also evident, though, that contrast is not all there is: that the reformists do not represent a complete departure from what has gone before.

We can trace the differences by taking a closer look at two publications which fall (albeit not neatly) on different sides of the divide. Again, it should be remembered that there are no ‘official’ doctrinal platforms and that the reach of publications that circulate among Muslims is hard to assess. Nevertheless, voices that made it into print and into the offerings of Dar es Salaam's struggling Muslim booksellers represent a contribution to the debate that is accorded a minimum of legitimacy.

On the reformist side, Safari ya roho akhera, ‘the journey of the soul into the afterlife’, was published in 1999 by Said Amour Al-Habsy in Oman. I bought it from an open-air bookstall outside the Ngazija (‘Comorian’) mosque in Dar es Salaam, one of the first mosques in the city centre to become embroiled in reformist controversies (Al-Habsy 1999). The name Al-Habsy suggests a Zanzibari with Omani connections, and the author is clearly bilingual in Swahili and Arabic. He terms his pamphlet a short introduction to Islamic beliefs about death, burial, the afterlife, the last judgement and heaven and hell. Al-Habsy does not spend much of his writing on condemning others’ views: maintaining a didactic tone, he focuses on what is appropriate. Among reformists, he is moderate. Nevertheless, he finds time to reject bida’, innovations at large, and matanga gatherings in particular (Al-Habsy 1999: 158).

Al-Habsy's exposition of funerary ritual and of supplication on behalf of the deceased is built around a succession of Qur'an quotes, supplemented with prophetic traditions. In particular, he upholds the Qur'anic injunction that those accompanying a bier should pray to God with modesty and in a low voice, without making noise: the same one invoked by the patron of the reformists among Lindi's established sheikhs against traditionalists’ reciting of the shahada (Al-Habsy 1999: 144–5).Footnote ¹⁰ While he goes into great detail regarding the handling of the corpse, the prayers integrated into the process are few and short. He condemns the reading of the Qur'an on behalf of the deceased after burial as an abuse of the divine word, all the more strongly if the person reciting is paid for doing so (Al-Habsy 1999: 152–3).

Regarding the ability of the living to mitigate the suffering of the deceased, Al-Habsy focuses on the belief that every Muslim will be treated according to his own actions. The most important charity the living can show the deceased is at the time of their dying, rather than afterwards: it is then that they should remind them of their faith in God, so that they can die declaring their faith. After death, only the continuing effects of alms given or education provided by the deceased, or the supplication of a good child can ease their fate. Yet Al-Habsy offers some qualifications: God will listen if survivors (related or not) remember the good deeds of the deceased, and supplicative prayer performed during the preparations for burial ‘reaches’ the deceased (Al-Habsy 1999: 157–8).

Compared to the vociferous condemnation by some reformist preachers of any ritual act not vouchsafed by the Qur'an, Al-Habsy is moderate. Nevertheless, he takes care to derive his prescriptions closely from the Qur'an and a limited number of hadith, emphasizes the virtue of austerity and restraint, and puts clearly defined limits on the ability of the living to support the dead. Where he uses contemporary scholars as reference points, they are from the Middle East; despite using the Swahili language, he makes no explicit reference to the region he is addressing.

A spirited defence of the traditionalist view on funerals, meanwhile, was published in 2002 or 2004 (the cover and front page give conflicting dates) by members of the Qadiriya Sufi order in Dar es Salaam. It is worth quoting at some length, as it offers vivid illustration of what, to the mind of traditionalists, is at stake in funerary ritual. The title translates as ‘Debate [Swahili: Mjadala] between the society of the Answaru Sunna [that is, an outspoken reformist group] and the Ahal-Sunna wal-Jamaa, the latter represented by the youth of the Qadiriya. Held on 11 October 1997 until 27 October 1997, equivalent to the 1418th year of the Hijiriya, in Tandika – Dar es Salaam’ (Qadiria: 2002/4). The book, then, is the written account of a 16-day debate between the reformists (Answaru Sunna) and the Qadiri Sufis, as recorded by the Sufis. According to the introduction, the debate was arranged by elders in a poor southern suburb of Dar es Salaam who were concerned about the doctrines taught by some younger teachers at a mosque in the neighbourhood.

Of a total of fifteen subjects of debate recorded, six deal with funeral practice, making it the topic most extensively discussed. The disagreements addressed revolve around the supererogatory rituals practised by the tarika: recitation of the shahada (confession of the faith) and the Qur'an, collective prayer in the mosque and individual supplication on behalf of the deceased, the reading of a set of instructions (known as talkini, after the Arabic talqin, ‘prompting’) for the encounter with the angels of death at the just-filled-in grave. Points of doctrine are also addressed: whether the dead can hear in the grave, and the likelihood of Muslims entering hell.

The Qadiris’ assertions here reflect the central role of the shahada in traditionalist Muslim practice generally, and in funerary ritual in particular. It may be evoked in ordinary conversation, and it accompanies a traditionalist Muslim's corpse all the way from the home to the grave. Colloquially, some traditionalist Muslims refer to themselves as ‘lailah Muslims’ in counter-distinction to the Answaru. But the text also emphasizes the deceased Muslim's entitlement to divine grace, which the lailah Muslims see as under attack from the Answaru. It is not only ritual propriety that is at stake, but also the way believers can secure themselves, and their pre-deceased co-religionists, from the dangers of the afterworld.

The underlying concern about the well-being of the dead becomes most palpable in the arguments over the talkini ritual. This is how (according to the traditionalists’ account) the reformists dismiss it:

To counter this claim, the Qadiris cited events around the funeral of Ibrahim, Muhammad's son, who predeceased his father. It is an account of the first Muslims demanding reassurance from the Prophet for the Day of Judgement:

This pronouncement of the Archangel is taken from the Qur'an sura bearing Ibrahim's name. The entire story is an elaboration of the context in which this sura was revealed. As such, the Qadiris used it without giving a separate reference for the origin of this hadith, describing it instead as a ‘tradition’, mapokezi. They imply that the sanctity of the text substantiates the circumstantial story they tell about it. Above all, the story itself emphasizes the believers’ dire dependence on divine grace for safety in the afterworld, and the dependence of the dead on the living in seeking this grace.

The overriding concern of the Mjadala as regards funerals is to assert that the living are able to care for the dead: to prepare them for the tests awaiting them in the grave, to atone for their sins, to obtain God's blessings for them. The Answaru stand accused not only of hubris, but also of cruelty: they threaten to deny the dead the care of the living. As the Qadiri put it in defending their use of the shahada: ‘It is quite impossible that God in his mercy and compassion should deprive a deceased person of the benefits and thawabu [divine reward] provided by their fellows’ (Qadiria Reference Qadiria2002/4: 70) The Answaru, to the Qadiris’ mind, are denying God's generosity and grace towards the deceased.

In the Tanzanian context, then, the most striking thing about a reformist statement such as Safari ya roho is how much is not part of funerary ritual according to this book. The use of the shahada, the talkini, and the many sources of intercession that the Mjadala speaks of are not even mentioned. To appreciate the appearance of this pared-down ritual to traditionalist mourners, it is helpful to remember the longer story outlined above. In their first hundred years of growth, Muslim congregations cultivated careful funerary rituals that were considered an improvement over earlier generations’ way of ‘throwing out’ the dead. In a sense, the dead were allowed a more tangible presence in society than before, with the establishment of permanent burial grounds. At the same time, the notions of judgement day and hell (and, as far as it was known, the torture of the grave) implied a new need to look after the dead. Against this background, the invective of reformists against these rites in the name of Islam presents to traditionalists an almost perverse twist in the interpretation of what it means to be Muslim.

CONCLUSION

Thus in the debates on funerals, Tanzanian Muslims, reformists and traditionalists alike, are working through the implications of the monotheist convictions they have acquired. The understanding of mortality is in flux; the right way to die as a creature of God, and to bury people with the anticipation of a day of judgement, of eternal bliss and damnation. Against this background, the concern of present-day Muslims for the dead appears familiar, yet it also shows a perceptible shift in emphasis. Donations to the ‘elders from the past’, as ancestors are often familiarly described, used to be made to conciliate them, to demonstrate loyalty and secure their goodwill. Now, it is the task of the living to secure God's goodwill for the dead. It could be said that where once the living needed the goodwill of the deceased, now the dead need the care of their living descendants. Yet, although the most clearly discernible direction of change is towards a loss of power of the dead in the lives of the living, the relationship between living and dead clearly remains mutual. The dead are still deemed to hear the living.

Set in context, this discussion of funerary practice reveals an ongoing concern about the relationship between the living and the dead, shared by reformists and traditionalist Muslims. This concern, in turn, is informed by long-term, place-specific changes not only in religious affiliation, but also in the participants’ wider life world, from settlement patterns to forms of social control. It is telling that the debate on funerals is at its liveliest not in the capital, but in provincial towns and villages, within a region that is economically and politically marginal within Tanzania (Seppaelae and Koda 1998). The villagers in question are sharply aware of this marginality, and concerns about the ability of local society to both hold together and hold its ground in interaction with both national power centres and international market forces are strong. Their disagreements on funerary practice resonate with a general sense that village society, tied in by these national and transnational concerns, is fraying.

In effect, the reformists propose a rupture in the process of accumulation, reinterpretation and experimentation in ritual practice through which Tanzanians had earlier learned to live as Muslims. Nevertheless, the significance of this proposed rupture arises from the way the protagonists interpret it with reference to their own history: from place-specific conditions. It is, moreover, never clean-cut, whatever the rhetoric: the reformists cannot actually set out with a clean slate. Thus their insistence on personal, individual responsibility for the fate of one's soul, for instance, cannot be expected necessarily to translate into a sense of personal responsibility of the sort associated since Weber with the Calvinist reformation in Christianity. It remains to be seen how it will shape actions and dispositions among Tanzanian Muslims.

The debate also offers insight into the way East African Muslims conceive of their efforts to define correct practice. The process whereby African Muslims integrate Islam into their lives has been described as the ‘Africanization’ of Islam, complementary to the Islamization of parts of Africa (Robinson Reference Robinson2004). While this description is certainly adequate for observers interested in cultural change, the actors themselves do not construe their debates as part of a process of cultural interaction, because they do not think of Islam as the bearer of a specific culture. Rather, they seek to distinguish between essential and inessential or correct and incorrect Islamic rules. This ‘casuist’ perspective is in keeping with doctrinal debates among Muslims in many places and times, but also harkens back to the cumulative and experimental approach to ritual practice so evident in the earlier development of funerary ritual.

The present case, then, shows that the dynamics of this debate among Muslims cannot be fully understood with reference to a standardized account of Islamic doctrine and scholarly debate. More generally, the culture of a group whose members are Muslim is not necessarily most succinctly described as a ‘Muslim’ culture (any more than it would be conclusive to describe culture in south-east Tanzania as ‘African’). Even though the inspiration for the arguments over funerals clearly came from abroad, they need to be traced also in place-specific terms. Ultimately, this insight points towards the dissolution of the venerable notion of the ‘Muslim world’ as an analytical unit, and towards the recognition of the Islamic presence in Africa (and elsewhere) as no less diverse than the Christian one has long been accepted to be.