‘But the coast, of course, is quite different’: Academic and Local Ideas about the East African Littoral

Abstract

In recent years, anthropology has paid much attention to the concept of identity. Identity politics is a shifting and complex area, but the trick is to claim the right identity at the right time. This article discusses some of the issues associated with this topic on the coast of East Africa. The quotation in the title is a phrase I often heard when a student studying Swahili in the early 1960s. The East Coast was considered to be different from the rest of East Africa – otherwise known as ‘up-country’ – because it had a long history and impressive material remains as well as a written language with its own literature. What it did not have, unlike the rest of East Africa, were ‘tribes’. In the postcolonial period, ‘tribalism’ has provided a popular and simplistic explanation in the mass media for the conflicts and wars in Africa. Historians, political scientists and anthropologists have argued, however, that modern ‘tribalism’ does not represent indigenous polities but rather the fall-out from the introduction of modern political systems and conflicts over resources. Given all of these factors, why, in the late twentieth century, should there have been calls for the Swahili to be recognised as a ‘tribe’? Seeking answers to this question takes us to an old debate – who are the waSwahili? – sometimes phrased as ‘Is there such an entity as the Swahili?’ In the first section of this article, I consider the arguments of those who have maintained that the Swahili are not a single people, and in the second discuss the contrary case. The third section considers some of the reasons for such differences in approach, including historiography, identity politics, and the relative positions of authors.

In recent years, anthropology has paid much attention to the concept of identity. Friedman notes that: ‘The shift from the mid-1970s to today has been towards an increasing ethnification of public social realms, a generalised increase in identity politics … In such a process there is little room for the hybrid identification discussed and pleaded for by cultural elites’.1 Gillis puts it another way: ‘Today, it seems that everyone claims a right to their own identity … Identity has taken on the status of a sacred object, an ultimate concern’.2 Identity politics is a shifting and complex area, but the trick is to claim the right identity at the right time. This article discusses some of the issues associated with this topic on the coast of East Africa.

The quotation in the title is a phrase I often heard when a student studying for a degree in African Studies (Swahili branch) in the early 1960s. The East Coast was considered to be different from the rest of East Africa – otherwise known as ‘up-country’ – because it had a long history and impressive material remains as well as a written language with its own literature. What it did not have, unlike the rest of East Africa, were ‘tribes’.

Early anthropological studies of Africa took it as axiomatic that tribes existed; this was what anthropologists studied and what colonial governments dealt with.3 John Iliffe, for example, writing of Tanganyika, states: ‘The notion of tribe lay at the heart of indirect rule in Tanganyika. Refining the racial thinking common in German times, administrators believed that every African belonged to a tribe, just as every European belonged to a nation.’4 However, in the latter half of the twentieth century, ‘tribes’ largely fell out of favour as an anthropological concept. Magubane and Faris articulated a critique which had become current in the 1970s and 1980s: ‘When one is referred to as a “tribesman” and actions against oppressive conditions are conceived as due to “tribalism”, the African is being defined as an individual who, for practical purposes (i.e. suppression), is an irrational savage.’5 A decade later, such views had become mainstream. Barnard and Spencer's recent Encyclopaedia of Anthropology notes: ‘The terms tribe and tribal … have a variety of meanings, some of which are taboo in modern anthropology … [T]he use of “tribal” to refer to aspects of culture other than politics is generally discouraged in contemporary anthropology.’6 Barfield's Dictionary of Anthropology takes a similar line: ‘The word “tribe” has a long and ignoble history and remains one of the most variably used terms within and outside of anthropology.’7 It has been recognised that, in the context of Africa, ‘tribes’ were mainly colonial creations, an effort to impose order and supra-local authority on previously largely autonomous communities and that in creating ‘tribes’, what had previously been fluid boundaries became fixed.8 Further, during the struggles for independence, African intellectuals generally rejected the term ‘tribe’, seeking instead to cultivate other concepts such as nationalism, Pan-Africanism, ujamaa, negritude and so on.

In the postcolonial period, ‘tribalism’ has provided a popular and simplistic explanation in the mass media for the conflicts and wars which have plagued the continent. Historians, political scientists and anthropologists have argued, however, that modern ‘tribalism’ does not represent indigenous polities but rather the fall-out from the introduction of modern political systems and conflicts over resources. As Barfield notes, ‘[M]odern “tribalism” does not represent indigenous political systems … [but it has been] skillfully manipulated in post-colonial states by ethnic entrepreneurs seeking control of governments.’9

Given all of these factors, why, in the late twentieth century, should there be calls for the Swahili to be recognised as a ‘tribe’?10 Seeking answers to this question takes us to an old debate – who are the waSwahili? – sometimes phrased as ‘Is there such an entity as the Swahili?’ In the first section of this article, I consider the arguments of those who have maintained that the Swahili are not a single people, and in the second discuss the contrary case. The third section considers some of the reasons for such differences in approach, including historiography, identity politics, and the positionality of authors.

The Swahili Denied: The Colonial and Immediate Postcolonial Periods

In this section, I consider four categories of people who have argued that there is no single Swahili culture and people: the Swahili themselves, the colonial authorities, many scholars writing during the colonial period and its immediate aftermath, and the governments of the modern nation-states of Kenya and Tanzania.

Coastal dwellers have rarely used the term waSwahili to describe themselves, preferring other terms. Some have claimed Arab status, as this placed them high in the coastal social hierarchy, which was a distinct advantage during the colonial period. They have also used the names of their settlements or areas of the coast, for example waAmu, waMombasa and waPate. Some have also used the term waShirazi (or waShirazii), claiming a remote connection with Persia. During this period, the term waSwahili was most often applied to people who did not have a place of origin, and who were thus likely to be of slave descent, the lowest social category. Census figures during the colonial (and indeed postcolonial periods) reveal clearly ‘the fatuity of using statistics to measure anything except people's changing perceptions of their own identity’.11 A much-quoted article by Eastman concludes likewise that ‘Certain people consider themselves Swahili and others not, while some of those so considered [by others] may not have that self concept at all.’12 Her article suggests, however, that the situation by the time of writing was in a state of flux, with even members of the Arab League of Mombasa deliberately emphasising that they were Swahili rather than Arab. This stance is scarcely surprising in a post-independence situation in which people were ‘being encouraged to think of themselves as Africans first, and then as Kenyans or Tanzanians’.13

The British colonial power was ambivalent about the identity of the coastal people. Only those who could satisfy the colonial authorities that they were indeed ‘Arabs’ were entitled to be classified as ‘non-natives’, which gave access to different legal jurisdictions and improved food rations during war-time. Allen has argued that in the colonial hierarchy of European, Asian/Arab, and African, the Swahili were regarded as inferior because they were ‘cross-bred: Arab and African, a ‘half-caste’ or mongrel’ ‘race’:

[A]lmost everything the Swahili represented was anathema to British colonial administrators. The latter liked their Africans to be racially ‘pure’ while the Swahili glorified [sic] in a part Arab ancestry, and they liked their subjects to be rural dwelling and ‘tribally-minded’ … where[as] the Swahili were …'of ‘mixed descent’, detribalised, and living around the towns.14

Allen notes that this view was also reflected in the writings of scholars such as Pearce, Hollingsworth and Ingrams, all of whom also tended to deny the existence of a single Swahili people.15 Similarly, as he points out, Coupland's history of the coast, published in 1938, ‘interpreted the whole of East Africa's coastal history in terms of the colonisation of the region by immigrants from the Middle East, with the indigenous inhabitants relegated to [the status of] wives and slaves’.16

As far as Western scholars are concerned, even in the immediate postcolonial period, much attention continued to be paid to influences on the coast from overseas. For example, the archaeologist Neville Chittick argued that the magnificent ruins at Kilwa Kisiwani should be attributed to the influence of outsiders, rather than to Africans.17 The Dutch anthropologist Prins begins his 1961 survey of the East Coast of Africa as follows: ‘This is a book about Arabs who have settled in Africa, together with Africanized Persians and coastal Africans.’18

As Alamin Mazrui and Ibrahim Shariff note, the postcolonial governments of Tanzania and Kenya have also been reluctant to accord the Swahili a special identity, albeit for different reasons. The government of Kenya has refused to recognise the Swahili as a distinct people and has been especially nervous about political developments on the coast parading as regionalism. Perhaps it fears an attempt to recreate the period during British colonial rule when the Sultan of Zanzibar ruled a ten-mile wide strip of Kenyan coastal territory.19 On the other hand, Mazrui and Shariff suggest, the government of Tanzania, in pursuing policies of nation-building and the promotion of Swahili as a national language, has effectively extended the boundaries of the Swahili to include virtually the whole population of the country: ‘The Swahili are required to give up their ethnic identity so that it can be turned into the public property of the Tanzanian nation.’20 Other commentators have noted that in Zanzibar, which is semi-autonomous within the Union, policies after the 1964 Revolution were to downplay differences and encourage an ‘Afro-Shirazi’ identity, until the recent resurgence of ethnic politics associated with multipartyism.21

Thus, in sum, many of those writing during the colonial period and its immediate aftermath suggested that the term waSwahili could not be applied to coastal dwellers in the same way that terms such as Gikuyu, Luo or Chagga, for example, might be applied to inland people.

The ‘Swahili’ Acknowledged: Recent Debates

More recently, however, both indigenous and foreign scholars have argued strongly not only for the recognition of the Swahili as a people with their own history and culture, but for their continuous existence for over at least a millennium.22 I do not have space here to rehearse all of their arguments, so will note only some of the salient details. Nurse and Spear, on the basis of both historical and linguistic evidence, argue as follows:

Our basic point is that the Swahili are an African people, born of that continent and raised on it. This is not to say that they are the same as other African peoples, however, for in moving to the coast, participating in Indian Ocean trade, and living in town their culture has developed historically in directions different from those of their immediate neighbours … Arabs have been trading along the coast for a long time, and many have remained to settle and to become Swahili … The result has been neither African nor Arab but distinctly Swahili.23

Allen likewise states that ‘[t]o the assertion that “There is no such person as a Swahili” there has always been a very simple retort: “Who in that case, developed the Swahili language and who wrote and read or sang Swahili poetry?”’24 He continues:

‘The Swahili with whom this book is concerned are those who lived in the coastal settlements between 800 and 1600 AD, built places like Kilwa and Gedi, used sewn boats, carved door frames, wrote poetry, and of whom Chinese, Arabs, Portuguese and others have left so many written accounts.’25

The most recent attempt to synthesise work on Swahili history, society and culture is that by Horton and Middleton, respectively an archaeologist and anthropologist. They have argued that it is possible to use the present to understand a past that is visible in the archaeological and historical evidence and which points to the Swahili as being quite different from their neighbours, but constituting a single African civilisation. They maintain that for millennia Swahili society formed one of the major mercantile societies of the Indian Ocean littoral, its members constituting cultural brokers and mediators, as well as commercial middlemen, between the interior of Africa and Asia. According to Horton and Middleton, the features that marked the Swahili off from other societies included a written language and literature, a different religion (Islam) and moral system, a complex internal structure and stratification system, and stone-built towns with a distinctive form of architecture, as well as other aspects of material culture. They sum up Swahili society as follows: ‘It has features of both Africa and Asia and is situated on the geographical and cultural boundary between them, and remains a tertium quid that faces both ways and contains many contradictions within itself.’26

Indigenous scholars have also argued strongly that the Swahili are certainly a people, and have long been so. Indeed, some would go further and argue that the Swahili are actually a tribe because they have a specific ancestry with their own clans and lineages, share a common language and literature, live in the same geographical area and have traditions which can distinguish them from others.27

Mazrui and Shariff maintain that the ‘European fetish’ about Swahili ‘Arabness’ has had its reaction in the Swahili nationalists’ stress on African-ness. They see this as ‘part and parcel of that wider syndrome of intellectual dependency precipitated by the colonial and neo-colonial experiences in Africa rooted in the interests of global capitalism’. For this reason, the Swahili nationalists’ position is ‘strongly affirmative about Africa, but apologetic about Arab links’.28 At the same time, they themselves have attacked many foreign scholars writing not only in the 1960s and 1970s, such as Harries, Knappert and Prins,29 but even more recent ones such as Nurse and Spear for ‘denying the Swahili their Africanity’.30 Indeed, in a recent series of polemical exchanges between Ali Mazrui and Wole Soyinka, this is the issue about which Mazrui appears most incensed: ‘Your most despicable attack involved your questioning my Africanity.’31 Here, then, we see the importance of a recognition not only that the Swahili are a single people, but also that they are an African people, indigenous to the continent.

There is a second recent discourse which appears to lend some support to the Swahili-as-a-people thesis, perhaps somewhat unexpectedly in the light of what has been said in the previous section of this chapter. In their attempts to develop their economies, both Kenya and Tanzania have encouraged the growth of tourism. The two locations stressed in the tourist literature are the game parks for safaris, and the coast for beach holidays. The first coastal area to undergo mass development was in Kenya, especially around Malindi and Watamu. Subsequently, in Zanzibar there has been extensive development of beach resorts and the restoration of the Stone Town.

An examination of recent tourist promotion literature concerning the coast of East Africa,32 much of it government-produced, appears to suggest that, far from dismissing their existence, Swahili history and culture are being actively promoted. For example, a leaflet produced by the Kenya National Tourist Office notes that the coastal belt is ‘truly a distinctive part of the country with a rich and romantic history dating back 1,000 years’. Although the initial emphasis is on the Miji Kenda (‘the nine tribes who inhabit the area’), the leaflet also makes mention of the ‘once thriving Swahili civilization’ of Swahili city-states, and the ‘still thriving Swahili culture’ of the Lamu archipelago.33

Several of the Tanzanian Tourist Office leaflets describe the country as ‘the land of Kilimanjaro and Zanzibar’, and note of the latter that its name ‘evokes a romantic past … ’ and that ‘it has a long and colourful history’.34 Even the governmental ‘Zanzibar Commission for Tourism’ literature utilises terms such as ‘rich heritage’, ‘romance’, ‘magic’, ‘former Sultanate’, ‘thousand and one nights’, and one of its leaflets is termed ‘Zanzibar – island of Swahili culture’, where ‘Arab, Indian, Persian and European influences blended with local African tradition’.35 The palace of the former Sultan, ousted in the 1964 Revolution, is high on the list of not-to-be missed sights.

Thus while the governments of both Kenya and Tanzania might prefer to consider their coastal regions to be just another area of their respective countries, through their encouragement of particular kinds of tourism, they themselves collude to some extent with the notion of the coast as special, different, and redolent of a particular kind of exoticism. In this way, the past and its material remains, and the present, in the form of particular manifestations of culture, are both resources which are being used, some might say appropriated, for economic purposes.

Some of the Swahili themselves have seized an opening here. Recently there has been a sophisticated attempt by a number of individuals living in Zanzibar and Dar es Salaam to promote to an international audience what is described as ‘coastal eco-tourism’ in Tanzania, with the launch in July 1998 of the lavishly illustrated The Swahili Coast magazine.36 Articles, some of which are written by local academics, focus on history and archaeology, sites, local customs such as the Swahili New Year, Swahili cookery with recipes, and material culture such as boats, architecture, carved doors and khanga. Emphasis is laid on the history and culture of the Swahili coast, its beauty and antiquity and its distinctive qualities.

Explaining Differences

Historiography

These arguments must also be understood in the context of wider academic debates about African historiography. Prior to the 1960s, sub-Saharan Africa ‘had’ little history, and most of what it did dated from the onset of the colonial period. Over the last forty years, there has been a move from a history which portrayed sub-Saharan Africa largely in terms of colonial conquest to one which sees Africans as creators of complex cultures or civilizations, a move which began in the 1960s with the Oxford History of Africa, which includes Oliver and Mathew's pioneering History of East Africa.37

A decade later saw the beginning of the massive undertaking of the UNESCO History of Africa, in the ‘Preface’ to each volume of which the former Director General of UNESCO, Amadou-Mahtar M'Bow, notes as follows: ‘For a long time, all kinds of myths and prejudices concealed the true history of Africa from the world at large. African societies were looked upon as societies that could have no history … and [it was] argued that the lack of written sources and documents made it impossible to engage in any scientific study of such societies’.38 He goes on to argue that the consequence of this was a ‘refusal to see Africans as the creators of original cultures which flowered and survived over the centuries in patterns of their own making’;39 for him such a view can be linked to racial stereotypes of black and white.

The distinguished Kenyan historian B. A. Ogot, who was president of the project's International Scientific Committee, also contributes a brief ‘Description of the Project’ to each volume, noting that the project considered Africa as a totality, rather than using the traditional division made by the Sahara. He also explains that two-thirds of the Board's membership was African and that, where suitably qualified people could be found, preference was given to African authors, thus it is essentially a work from the inside, ‘a faithful reflection of the way in which African authors view their civilization’.40

Here then, we can discern two trends. One is the enormous growth in scholarship on and about Africa in the last half of the twentieth century. The other is that much of this scholarship is indigenous. It is also clear from what both M'bow and Ogot have to say that history is not just for scholars: it is of vital importance to Africans living on the continent of Africa, as well as to Africans of the diaspora.

There are two other particularly relevant works to which I should allude briefly. They are, first, Cheikh Anta Diop's Civilization or Barbarism (published in French in 1981 and in English ten years later). The cover blurb sums up the thrust of the argument: ‘Diop offers a critical challenge to orthodox scholarship's interpretation of Egypt as a White civilization, which arose during the nineteenth century to reinforce European racism and imperialism … [H]e demonstrates that Egypt was a Black civilization and that Blacks are the rightful heirs to Egypt's proud legacy.’41 Second, Martin Bernal's Black Athena argues that Europe's classical civilization, that of ancient Greece, actually had deep roots in Afroasiatic cultures;42 it is effectively an attack on the racism endemic in European historiography and a deconstruction of Europe's own myth of charter.

Thus the writing of the history of Africa is inevitably a very political matter, involving not only European views of Africans and their capabilities, but more importantly, both African and European views of themselves, their origins and their relations with each other. In the case of East Africa, it also includes a consideration of its relations with Asia and the Indian Ocean littoral, especially the Middle East.

Identity Politics

An understanding of identity politics on the East African coast necessarily involves a consideration of its political and economic history. As already noted, identity claims have shifted over time, for example the rise in claims to Arab status around the turn of the twentieth century and the more recent stress on African identity in a postcolonial world.43 Furthermore, identity is never absolute: as Horton and Middleton observe, ‘the Swahili have many identities and can select in different situations’ and that ‘as with most social identities, boundaries exist and are meaningful, but they are permeable and changeable in many’.44 Allen's view of Swahili identity is similarly nuanced:

What we have to define, then, it not a tribe or any other group held together, even notionally, by links of blood or marriage, but a highly permeable population whose common factor is cultural in nature … an identity which is neither tribal nor racial but an alternative socio-political structure which is more appropriate for urban living. [Thus] being Swahili is not an absolute state which precludes being anything else.45

Mazrui and Shariff report that Kenyan Swahili complain that they have been marginalised in the current political process, since the government is run by upcountry Africans, while they are discriminated against in the job market. They argue that Islam receives much less time on Kenyan TV than does Christianity, and attempts to form a political party to address these issues have been thwarted by the government. Furthermore, the Swahili themselves have rarely profited from the tourist boom and there are constant complaints about the loss of beach-front land and fishing rights.46

Such issues are also discussed by a number of anthropologists working on the coast who have published on the effects of tourism. In Malindi and other parts of the Kenya coast, Beckerleg and Peake have noted the growth of a ‘beach-boy’ culture and in prostitution and drug addiction which is much lamented by the older generation.47 In Zanzibar, the work of Racine, Larsen and Parkin on the effects of tourism has revealed some interesting paradoxes. Racine has shown how the Swahili New Year ceremony (Nairuz) in Makunduchi has become a ‘cultural event’ which is not only televised, but also attracts a large number of tourists; it now has a political dimension ‘to demonstrate a cultural continuity between the islands and the mainland, given the background of the political problems of the Union’.48 Parkin has considered the effect of tourism upon local art,49 while Larsen has discussed the extent to which tourism impinges upon local concepts of privacy.50 It is thus scarcely surprising that, as Parkin has shown, militant young Islamists in the early 1990s were demonstrating for an end to tourism, arguing that it was destroying their culture.51

Taking this body of literature together, it seems clear that some Swahili are profiting from tourism, while others feel profoundly uncomfortable with its consequences, or else resent the fact that many of its financial benefits appear to be going to entrepreneurs who are not Swahili, but either Africans from inland or foreign investors. They see modernity qua tourism as coming with a very high price tag.

In the light of these situations, it is perhaps not surprising that a new breed of Swahili ‘nationalists’ has called for the recognition, both by governments and scholars, of the reality of a Swahili identity, with its own culture and history, and, indeed, for the Swahili to be recognised as a ‘tribe’ (kabila). It may, of course, be argued that, given the failure of the post-independence Coastal People's Party to achieve its goals of greater autonomy for the coastal region,52 it was inevitable that struggles against perceived complaints – loss of status within the new nation, loss of coastal lands – would be waged by other means.

In Zanzibar, there are other, longer-running grievances, going back to the 1964 Revolution which ousted the Sultan's government and replaced it with one which explicitly favoured ‘Africans’ rather than ‘Arabs’. Such issues of ethnic difference also appear to have been exacerbated recently by the introduction of ‘multi-party democracy’ in Tanzania which has allowed the proliferation of new parties contesting elections. In Zanzibar the elections of 1995 and 2000 have been widely regarded by outside observers as flawed, and the most recent were followed by rioting and police firing on a crowd, which resulted in an exodus of refugees to Kenya.53

Thus the question of Swahili identity, much debated over the years by scholars, has recently become a very live political issue. As Mazrui and Shariff argue, ‘It has become impossible to talk of Swahili identity in terms that are neutral and non-partisan.’54 They themselves go on to support the views of Chiragdin and to argue for the Swahili being regarded as another African ‘tribe’: ‘We once again want to assert that the Swahili are a kabila, clear and simple.’55 It is paradoxical that a people who have spent much of their history seeking to be identified as ‘Arabs’ or ‘Persians’ and stressing the differences between themselves and their upcountry neighbours, should now be emphasising not only their African roots, but even their claims to ‘tribal’ status. In this regard, the Swahili are, of course, following an important trend which has recently been observed in many parts of the world, whereby a group seeks to constitute itself as such, and thereby claim privileges or compensation for victimhood.56 As Cowan et al. point out: ‘Activists from, or working on behalf of, communities making claims are often well aware that they are essentializing something which is, in fact, much more fluid and contradictory, but they do so in order that their claims be heard.’ They cite the well-known example of the account by Clifford of the court-room battle for land by the Mashpee Indians which could only be successful if they could prove that they were now and ‘always had been’ a tribe.57

The Swahili nationalists similarly seek to show not only that they are a people, a group, and have been so for a long time, but, even more importantly, that they are African in origin. The most radical solution to the issue of Arab settlement and cultural influence is espoused by Ali Mazrui, who suggests that the boundaries of Africa be re-drawn to include the Arabian peninsula, citing geology, culture, language and history to argue his case: ‘I personally regard the present boundaries of Africa as not only arbitrary but artificially conceived by European geographers in a former era of European dominance.’58

Positionality

From the brief accounts I have given, it is perhaps clear why increasing numbers of Swahili should, towards the end of the twentieth century, have asserted (or re-asserted) their identity, including their Africanity. It is also not surprising that national governments should have resisted such a move, or at least resisted its expression in terms of the formation of political parties. I would like to suggest, however, that a reading of the recent literature on the Swahili coast should not only be undertaken with an awareness of postcolonial political changes in East Africa but also in terms of the positionality of those engaging in such debates. Where was the research carried out, by whom, and who was consulted in the course of it?

The first significant factor to which attention must be drawn is that much of the material on which recent models of coastal society are based comes from the northern part of the coast, not from areas further south. For example, Middleton discusses the history of his acquaintance with the coast, beginning in 1958 with a study of land tenure on Zanzibar and resuming nearly thirty years later, when he began to carry out research in the Kenyan towns of Lamu, Mombasa and Pate, all situated in the northern past of the coast.59 Allen too, as curator of the Lamu Museum, spent most of his time in that area, while Nurse and Spear also worked primarily on the northern coast. Yet all have produced models which are supposed to apply to the whole of the East Coast, from Somalia to Mozambique. Some aspects of what are claimed to be the structures of all Swahili societies simply do not fit the southern part of the coast, such as the area in which I work, Mafia Island.60 Although there is much less information available for the southern than the northern half of the East Coast,61 it seems clear that it is characterised by thatched villages, rather than the stone towns described by authors such as Middleton and Horton.

Second, while it is not surprising that local scholars who claim Swahili status articulate identity claims in their work, it is noticeable that a number of intellectuals with a coastal background who now live and work abroad, notably in the USA,62 have been particularly vocal in this regard. Perhaps this is not only because they are in a more favourable position to be able to make such arguments but also because some of them operate in an academic milieu in which the ‘recovery’ of Africa for Africans and their diaspora is an important political project.63

However, the varying positions occupied by western scholars are less clear-cut. Mazrui and Shariff argue that the ‘confusion’ over Swahili identity, exemplified by Eastman's 1971 article, came to serve as a ‘lucrative’ research area for Western scholars and that such work raised doubts about the very reality of Swahili identity. Much of the literature they cite in making this charge dates from a period before the 1980s since when, as already noted, most scholars have argued strongly for the existence of the Swahili as a single people and continuous civilisation. So was the change in recognition of the Swahili as a people simply, as Middleton and Horton suggest, because much more evidence is now available? Or is it the case, as has recently been suggested, that Western scholars too are bound by the paradigms of their time,64 and that to deny that the Swahili constitute a people, or that they are of largely African origin, is simply unacceptable? Yet, as Horton and Middleton note, ‘outsiders researching Swahili society must respect these local notions of origins from overseas and not reject them completely, in the modern and politically fuelled project to create a totally indigenous African civilisation to balance earlier views of an Asian creole one’.65

Issues such as the gender and age of scholars also have a bearing, not least because to some extent, they determine levels of access to informants on the East Coast.66 Clearly the kind of study carried out by Fuglesang on the responses of young women to videos could only be done by a young female researcher and the same is true of the work of Peake on beach-boys.67 At the other end of the age spectrum Middleton notes: ‘One point worth making is that in Lamu I have been lucky enough to have been accepted as an elderly scholar and as such have been able to discuss [matters] with both men and women …’68 Swartz, who carried out research in Mombasa, was not so fortunate and notes his regret that he was not able to live with a Swahili family: ‘I could find no-one who would house a solitary, outside man during my solitary, summer stays.’69 While male anthropologists may have difficulties in obtaining access to domestic space, they are of course advantaged in terms of participation in men's meetings (baraza) and mosques,70 spaces in which female participation would be difficult, as Larsen, who worked in Zanzibar City, notes: ‘My position as researcher and woman implied that I could not participate with men in all male contexts. My participation in contexts which privileged men would have been perceived as an intrusion, and thus an unacceptable violation of their social and moral space.’71

Thus women who have worked on the northern coast and Zanzibar, such as Larsen, Strobel, Le Guennec-Coppens, Fuglesang, Beckerleg and Porter, have had different kinds of access than male researchers and, as a result, their work tends to read very differently. Strobel, a historian, has written of the lives of women of different classes in Mombasa in the period between 1890 and 1975;72 le Guennec-Coppens has focused on women's lives in Lamu,73 while Fugelsang has written of the lives of young women in the same town.74

These are by now well-rehearsed issues in anthropology but to them might be added the issue of the background of the informants themselves. Porter's comments on entry into one part of the coast field are worth quoting:

The first people to whom foreign scholars of Swahili culture will be sent in Mombasa, regardless of the focus of their work, are the local male elders, sometimes called ‘sheikhs’. These are senior men recognized in the community for their expert knowledge of local religion and culture. They act as gatekeepers to their communities, and, I believe, see their engagement in the telling of Swahili culture as a way of ensuring its legitimacy and fixity in a world that is increasingly unpredictable … The vision of Swahili culture portrayed to me by these men is hegemonic in community discourse and in much of the published literature … The sheikhs depict Swahili culture as continuous through time and unchanging, highlighting their Islamic and literary tradition. They make almost no distinction between what constitutes Swahili culture and what marks the world of Swahili men, which they depict as tied closely to things Arab and Muslim in terms of religion, knowledge, piety, public oration and education … In the hegemonic narratives that I have described, there is a sense of a reified, even ossified, Swahili culture.75

Porter suggests that there is a distinct risk that, by replicating such messages, some scholars may not only perpetuate a view of the Swahili as a people hankering for past glories but also present a view of Swahili society ‘from above’, in terms of gender, age and social status. Indeed, Horton and Middleton state that their recent book is largely based upon the views of ‘true members of the coastal mercantile society … whom we have used as the main source material for our ethno-historical analysis’.76 This raises problematic issues about informants’ credentialist claims which might be worth further exploration. Conversely, the body of literature which deals with the problems occasioned by modernity, especially tourism, may suggest that there is a danger of ‘degeneracy’ in Swahili society.77

Conclusion

This paper has skimmed a wide body of literature in a necessarily superficial fashion. My aim has been to provide several kinds of context for its reading: the politics of Swahili identity, in which sameness and difference constantly alternate, and the positionality of some of the authors of this literature as well as that of their informants. I started with what had at first appeared to me to be a puzzle: why should some Swahili intellectuals be claiming ‘tribal’ status for their people, when the Swahili have for centuries formulated their identities largely as urban dwellers in opposition to mainland ‘tribal’ peoples? In the first part, I discussed some of those who refuted the existence of the Swahili as a single people, and in the second, the work of those who have argued that the Swahili are a single people, with their own distinctive, long-standing culture. Finally, I considered some of the reasons why people might adopt one side of the argument or the other. One such reason is linked to history and historiography: arguments about the identity of the Swahili are linked to arguments about their Africanity and their indigeneity and, in turn, to wider projects about the reclaiming of Africa and things African by Africans themselves.

Second, identity politics is linked to political and economic processes on the coast of East Africa, thus it is scarcely surprising that Swahili identity is both multiple and shifting. In considering issues of identity and identity claims, it seems important to stress the deployment of skills of adaptation to changing circumstances, long practised by the Swahili, and to consider Swahili society as constantly remaking itself in two ways. The first is by incorporating others ‘in wholly Swahili fashion’.78 The second is by carrying on as the Swahili always have, namely ‘seeking always to be the first to know what “the latest” is and to benefit from mediating between it and non-Swahili Africans’,79 tactics which are well borne out by the work of Beckerleg and Fuglesang.80

Finally, an examination of who the authors are in terms of where they did their research, their age and gender and hence their access to informants, suggests that these issues are salient in ‘reading’ their texts. In short, then, if we return to the question of whether the Swahili are or are not a people, a tribe, African or other, the answer, not surprisingly, depends upon who is speaking or writing, who is being addressed, and the historical moment in which this is occurring.

Acknowledgements

This article was written while I was in receipt of a Leverhulme Fellowship 2001–02. I am grateful to the Foundation for its support, and to the Nuffield Foundation for funding to attend the ASA 2002 conference in Arusha. My thanks also to Lionel Caplan, Janet Bujra and Richard Reid for comments.

Notes

1. Friedman, ‘Global Crises’, 84.

2. Gillis, ‘Memory and Identity’, 2.

3. I have in mind here numerous studies written in the 1940s and 1950s, for example: Fortes and Evans-Pritchard, African Political Systems; Colson and Gluckman, Seven Tribes of British Central Africa and Middleton and Tait, Tribes Without Rulers.

4. Iliffe, Modern History of Tanganyika, 323.

5. Magubane and Faris, ‘On the Political Relevance of Anthropology’, 100.

6. Barnard and Spencer, Encyclopaedia of Social and Cultural Anthropology, 626.

7. Barfield, Dictionary of Anthropology, 475.

8. See for example Vail, The Creation of Tribalism; Southall, ‘The Illusion of Tribe’.

9. Barfield, Dictionary of Anthropology, 476.

10. Mazrui and Shariff, The Swahili.

11. Allen, Swahili Origins, 8.

12. Eastman, ‘Who are the Swahili?’, 228.

13. Eastman, ‘Who are the Swahili?’, 230.

14. Allen, Swahili Origins, 4.

15. Pearce, Zanzibar; Hollingsworth, Short History; Ingrams, Zanzibar.

16. Allen, Swahili Origins, 5.

17. Chittick, Kilwa.

18. Prins, Swahili-Speaking People, ix.

19. Mazrui and Shariff quote from a publication of the Kenya Ministry of Tourism and Wildlife leaflet which states that ‘There is no such thing as a Swahili tribe or group’, in The Swahili, 44. Unfortunately, no date for this publication is given, and a reference to it does not appear in their bibliography.

20. Mazrui and Shariff quote from a publication of the Kenya Ministry of Tourism and Wildlife leaflet which states that ‘There is no such thing as a Swahili tribe or group’, in The Swahili, 44. Unfortunately, no date for this publication is given, and a reference to it does not appear in their bibliography., 45.

21. Cameron, ‘Political Violence’.

22. For example: Nurse and Spear, The Swahili: Reconstructing the History; Middleton, The World of the Swahili; Allen, Swahili Origins; Mazrui and Shariff, The Swahili; Horton and Middleton, The Swahili: The Social Landscape..

23. Nurse and Spear, The Swahili: Reconstructing the History, viii.

24. Allen, Swahili Origins, 11.

25. Allen, Swahili Origins, 12.

26. Horton and Middleton, The Swahili: The Social Landscape, 199.

27. Mazrui and Shariff, The Swahili, 46; Chiragdhin, ‘Kiswahili Na Wenyewe’; Chami, ‘A Review of Swahili Archaeology’.

28. Mazrui and Shariff, The Swahili, 48.

29. Harries, Swahili Poetry; Knappert, Traditional Swahili Poetry; Prins, Swahili-Speaking People.

30. Mazrui and Shariff, The Swahili, 42.

31. This exchange began with a hostile review by Mazuri of Henry Gates’ television series and book, both entitled Wonders of the African World (Gates and Davies, 1999). Most of these polemics were conducted on the web particularly in the journal West Africa Review which published a special edition in 2000 (http://www.westafricareview.com). See references to A. Mazuri 1999, 2000a, 2000b, 2000c, H. Gates, 2000, and W. Soyinka, 2000a and 2000b.

32. The governments of Kenya and Tanzania both maintain Tourist Offices, as does the (semi-autonomous) government of Zanzibar. All of these produce numerous booklets and leaflets, often obtainable in their overseas offices. As will be seen, these make for interesting reading, but unfortunately are rarely dated. I have used the dates of acquisition, which is a good approximation as new ones are continually being produced.

33. Kenya National Tourist Office, n.d. (c.2001): ‘Kenya coastline’ leaflet.

34. Tanzania Tourist Board, n.d. (c.2001): ‘Karibu Tanzania: The Land of Kilimanjaro and Zanzibar’ tourist leaflet.

35. Zanzibar Tourist Commission, n.d. (c.2001): ‘Zanzibar – Island of Swahili Culture’ leaflet

36. In the first issue, the publisher-editor Javed Jafferji states that ‘[O]ur hope is to provide some balance … between the love of one's own heritage which one wants the world to experience [through] – tourism and yet maintain the dignity of a people's culture. Swahili Coast, the magazine, hopefully meanders … through the passages of beauty that is the Swahili Coast while exploring its potential in an enterprising manner … we hope this journey provides you with the richness of our historical roots and culture’: The Swahili Coast: International Travel Magazine to Promote Coastal Eco-Tourism in Tanzania, 1 (1998) p. 7.

37. They go on to discuss the new techniques which were beginning to revolutionise conceptualisations of the history of Africa, indeed, to make it possible to write a history at all. One such was archaeology but there were others: ‘the still scarcely charted ocean of traditional history, tell-tale custom and ceremony, tribal law, family nomenclature and genealogy, place names, and all other substitutes for literary evidence’: Oliver and Mathew, History of East Africa, xii.

38. M'Bow's ‘Preface’.

39. M'Bow's ‘Preface’.

40. M'Bow's ‘Preface’.

41. Diop, Civilisation or Barbarism.

42. Bernal, Black Athena.

43. Horton and Middleton, The Swahili: The Social Landscape, 185, 199.

44. Horton and Middleton, The Swahili: The Social Landscape, 14, 23.

45. Allen, Swahili Origins, 5.

46. Mazrui and Shariff, The Swahili.

47. See Beckerleg, ‘Watamu’; ‘“Brown Sugar” or Friday Prayers?’; ‘Modernity Has Been Swahili-ised’ and Peake, ‘Swahili Stratification’.

48. Racine-Issa, ‘The Mwaka of Makunduchi, Zanzibar’, 173.

49. Parkin, ‘Escaping Cultures’.

50. Larsen, ‘The Other Side of “Nature”’.

51. Parkin, ‘Blank Banners and Islamic Consciousness in Zanzibar’.

52. The Coastal People's Party joined KADU, the Kenya African Democratic Union, made up mainly of smaller ethnic groups which feared domination by the ruling party KANU, which was dominated by the Kikuyu. Just after independence, KADU merged with KANU, leading to a long period of one-party rule. The so-called ‘Ten Mile Strip’ of coastal land had previously been ruled by the Sultan of Zanzibar during the British colonial period.

53. Cameron, ‘Political Violence’.

54. Mazrui and Shariff, The Swahili, 4–5.

55. Mazrui and Shariff, The Swahili, 53.

56. There is a new wave of ‘tribalization’ which is particularly marked among indigenous groups or ‘First Nations’ in the Americas seeking to claim their human rights. See for example Miller, State of the Peoples and Clifford, The Predicament of Culture.

57. Cowan et al., Culture and Rights, 11.

58. Mazrui, The Africans, 38.

59. Middleton, The World of the Swahili.

60. See Caplan, Choice and Constraint; Walley, ‘Making Waves’.

61. Few anthropologists have worked on the coast south of Bagamoyo, although Lockwood, Fertility and Household Labour in Tanzania provides some useful data for the Rufiji. There is, however, much recent archaeological work: see Chami, ‘A Review of Swahili Archaeology’.

62. Most of the Swahili intellectuals cited in this chapter form part of the Swahili diaspora in the USA.

63. Some indication of its significance can be gauged from the fact that the debates already referred to between Mazrui, Soyinka and Gates also spilled over into both American and Kenyan newspapers.

64. Lefkowitz, Not Out of Africa.

65. Horton and Middleton, The Swahili: The Social Landscape.

66. Caplan, ‘Learning Gender’.

67. Peake, ‘Swahili Stratification’.

68. Middleton, The World of the Swahili, xi.

69. Swartz, The Way the World Is, xi.

70. Ibid.; Parkin, ‘Blank Banners’.

71. Larsen, ‘Change, Continuity and Contestation’.

72. For example, Stroebel, Muslim Women in Mombasa.

73. le Guennec-Coppens, Femmes Voilées de Lamu.

74. Fuglesang, Veils and Videos.

75. Porter, ‘Resisting Uniformity’, 622–23.

76. Horton and Middleton, The Swahili: The Social Landscape, 16. Italics are mine.

77. Saleh, ‘“Going With the Times”’.

78. Allen, Swahili Origins, 244.

79. Allen, Swahili Origins, 254.

80. Berkerleg, ‘Modernity Has Been Swahili-ised’; Fuglesang, Veils and Videos.