Islam and Ethnicity in Northern Kenya and Southern Ethiopia (Eastern Africa Series, 13) 1847010466, 9781847010469

Citation preview

EAS_Schlee Islam_PPC_15mm:B+B

06.01.2012

16:00

Page 1

Günther Schlee is director of the Department of ‘Integration and Conflict’, Max Planck Institute for Social Anthropology, Halle, Germany; Abdullahi A. Shongolo is an independent scholar based in Kenya. Photograph: Somali traders and their Rendille customers. Ngoronit, Kenya © Günther Schlee, 2001

an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF (GB) and 668 Mt Hope Ave, Rochester NY 14620-2731 (US) www.boydellandbrewer.com www.jamescurrey.com

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

A companion volume to Pastoralism and Politics in Northern Kenya and Southern Ethiopia, this book is based on over thirty-four years of field research and synthesizes findings from history and political anthropology.

SCHLEE with SHONGOLO

The recent ethnic violence in Kenya has been preceded by a process of territorialization and politicization of ethnicity. This study examines a marginalized part of Kenya, the semi-arid north inhabited by pastoralists of three language groups – speakers of Oromo, Somali, and Rendille. It spans different periods of time, from early processes of ethnic differentiation between groups, through the colonial period when differences were reflected in administrative policies, to recent times, when global discourses, particularly those related to indigeneity and Islam, are tapped by local political agents and ethnic entrepreneurs.

Günther Schlee with Abdullahi A Shongolo

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Eastern Africa Series ISLAM & ETHNICITY IN NORTHERN KENYA & SOUTHERN ETHIOPIA

Eastern Africa Series Women’s Land Rights & Privatization in Eastern Africa BIRGIT ENGLERT & ELIZABETH DALEY (EDS)

War & the Politics of Identity in Ethiopia KJETIL TRONVOLL

Moving People in Ethiopia ALULA PANKHURST & FRANÇOIS PIGUET (EDS)

Living Terraces in Ethiopia ELIZABETH E. WATSON

Eritrea GAIM KIBREAB

Borders & Borderlands as Resources in the Horn of Africa DEREJE FEYISSA & MARKUS VIRGIL HOEHNE (EDS)

After the Comprehensive Peace Agreement in Sudan ELKE GRAWERT (ED.)

Land, Governance, Conflict & the Nuba of Sudan GUMA KUNDA KOMEY

Ethiopia JOHN MARKAKIS

Resurrecting Cannibals HEIKE BEHREND

Pastoralism & Politics in Northern Kenya & Southern Ethiopia GŰNTHER SCHLEE & ABDULLAHI A. SHONGOLO

Islam & Ethnicity in Northern Kenya & Southern Ethiopia GŰNTHER SCHLEE with ABDULLAHI A. SHONGOLO

Foundations of an African Civilisation* DAVID W. PHILLIPSON * forthcoming

Islam & Ethnicity in Northern Kenya & Southern Ethiopia GÜNTHER SCHLEE with

ABDULLAHI A. SHONGOLO

James Currey is an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF, GB www.jamescurrey.com and of Boydell & Brewer Inc. 668 Mt Hope Avenue, Rochester, NY 14620-2731, US www.boydellandbrewer.com © Günther Schlee & Abdullahi A. Shongolo 2012 First published 2012 1 2 3 4 5 15 14 13 12 All Rights Reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner The right of Günther Schlee & Abdullahi A. Shongolo to be identified as the authors of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988 The publisher has no responsibility for the continued existence or accuracy of URLs for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library AISBN 978-1-84701-046-9 (James Currey Cloth)

Papers used by Boydell & Brewer are natural, recyclable products made from wood grown in sustainable forests

Typeset in 8.5/9.5pt Cordale by CPI Typesetting Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY

Contents

List of Maps, Figures and Tables List of Abbreviations

Introduction

vii ix

1

1

Pax Borana

19

2

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

39

3

Modern Trends

71

4

Ecology and Politics

111

5

The Impact of War on Ethnic and Religious Identification in Southern Ethiopa in the Early 1990s (with ABDULLAHI A. SHONGOLO)

159

References Index

171 179

v

List of Maps, Figures and Tables

Maps Map 0.1

Boundaries in colonial Kenya

8

Map 4.1 Movement of a camel fora based on Kiliwe Hirre (KH) March 1989–March 1990

150

Map 4.2 Cattle camp movement from Kiliwe Hirre (KH) March 1989–March 1990

152

Map 4.3

155

Stock routes

Figures Figure 0.1

Colonial movement permit

6

Figures 1.1–1.5 Moieties and major divisions of ethnic groups mentioned in the text

20

Figure 2.1

Heptagram illustrating the days of the Gods

45

Figure 3.1

Dabel from a nearby elevation

74

Figure 3.2

Sakuye elder in a nomadic hamlet

74

Figure 3.3

Somali hut under construction

75

Figure 3.4

Somali hut frame covered with mats

75

Figure 3.5

Migrating nomadic hamlet in Mandera District

77

Figure 3.6

At the well

77

Figure 3.7

Hussein, a Sakuye Chief, intent on tracking a lion

78

Figure 3.8

Sakuye Chief ensures all is going smoothly at the well

79

Figure 3.9

The anthropologist’s camp in the village of Dabel

80

Figure 3.10

The anthropologist’s tent

80

Figure 3.11

Chief’s third wife

82

Figure 3.12

Freie Deutsche Jugend shirt worn by a Sakuye

82

Figure 3.13

Ali Jarso and the ethnographer

83

Figure 3.14

Constructing an instrument in lieu of a scalpel

85

List of Maps, Figures and Tables

Figure 3.15

Woman experienced in blood-letting

85

Figure 3.16

Blood-letting the author

86

Figure 3.17

After blood-letting the author

86

Figure 3.18

Somali traders and their Rendille customers, Ngoronit, Kenya 106

Figure 4.1

Water is drawn from a natural cistern at Lensayu

124

Figure 4.2

The water is poured onto the rock

124

Figure 4.3

The water flows down to where it is collected

125

Figure 4.4

The water is loaded on camelback in wooden containers

125

Figure 4.5

Reer Abdi Adan, smallstock hamlet

128

Figure 4.6 Agnatic relationships between the household heads of Reer Abdi Adan, smallstock hamlet 128 Figure 4.7 Agnatic relationships between husbands and wives (No. 1, No. 5)

129

Figure 4.8

Household 10, relationship between the spouses

129

Figure 4.9

Genealogical relationships between spouses

130

Figure 4.10

Reer Abdi Adan, camel hamlet

130

Figure 4.11 Agnatic relationship between the married men in the camel hamlet

131

Figure 4.12

Relationships between spouses (1)

131

Figure 4.13

Relationships between spouses (2)

132

Figure 4.14

Relationships between spouses (3)

132

Table 2.1 Seven celestial bodies and their periodic movements in relation to the Earth

45

Table 2.2 Gabra forms of the names of days of the week and the respective names for boys and girls born on each day

47

Table 2.3 Gabra set of names of days of the week and their associations

48

Table 2.4

Comparison of words in Maa and Rendille

57

Table 2.5

Examples of Arabic loanwords in Rendille

61

Table 2.6

Comparative chart of names of months in six languages 64

Tables

List of Abbreviations

Amh.

Amharic (Language)

Ar.

Arabic (Language)

B

Boran (Language)

CDF

Constituency Development Fund

DC

District Commissioner

EPRDF

Ethiopian People’s Revolutionary Democratic Front

FFZDDD

Father’s father’s sister’s daughter’s daughter’s daughter

KANU

Kenya African National Union

KNA

Kenya National Archives

Ksh

Kenya Shilling

MoBr, MB

Mother’s brother

NFD

Northern Frontier District

OALF

Oromo Abo Liberation Front

OLF

Oromo Liberation Front

OPDO

Oromo People’s Democratic Organization

PC

Provincial Commissioner

PRS

Proto-Rendille-Somali

R

Rendille (Language)

S

Somali (Language)

SALF

Somali Abo Liberation Front

Sw.

Swahili (Language)

TPLF

Tigrayan People’s Liberation Front

UNHCR

United Nations High Commissioner for Refugees

ix

Introduction Between two East African Highlands, the Ethiopian plateau to the north and Mt. Kenya to the south, there is a vast arid and semi-arid lowland, which, to the east, stretches into Somalia and down to the coast of the Indian Ocean. It comprises the larger part of Kenya and a substantial part of Ethiopia along its southern fringes. Most of this land is used by nomadic pastoralists, and only pockets of it are suitable for cattle. The rest is pasture for camels, sheep and goats. Without wishing to revise political maps (there are many good reasons against such revisions) one can say that in many ways these lowlands belong to the Horn of Africa or North-East Africa rather than East Africa. Linguistically, Cushitic languages dominate the entire area from the eastern shore of Lake Turkana (formerly Lake Rudolph) down to the Ocean. The largest of them is the Oromo language, which is also the language with the widest distribution in neighbouring Ethiopia. Here it is represented mainly by the Boran dialect. To the same Lowland East Cushitic sub-branch of that language family belong various Somali dialects spoken in the region, as well as Rendille, a language which might be very similar to what the Somali language once was before the Somali became Muslims and their language incorporated a vast number of Arabic loanwords. Economically, the lowlands of northern Kenya also form one region with the adjacent areas of Ethiopia and Somalia. There are the same groups of nomads and agro-pastoralists on both sides of the international boundaries, and animals are driven across them depending on where the highly erratic rainfalls have fallen. There is a fateful correlation between the amount of rainfall and its distribution in time and space: the lower the amount, the less predictable whether it will rain in a given year and where – and the average rainfall in northern Kenya is rather low. Much of the territory is enclosed by the 200 mm isohyet. There may be torrential rains on one day in one spot and years of drought in the immediate vicinity. Evaporation is that of an equatorial lowland.1 Apart from rainfalls, it is fluctuations in price levels which determine the movements of livestock. Before the collapse of statehood in Somalia, sheep were exported from Kenya into Somalia for Islamic festivals like ‘Id ul Fitr and ‘Id ul Hajj; especially for the latter vast numbers of sheep were taken from Somalia to Mecca, where millions of pilgrims from all over the world are obliged to sacrifice a sheep (Schlee 1992a). Now, as the roads to the Somali ports are obstructed by greedy warlords, animals find their way into Kenya to be exported via Mombasa (Little 1996). The catchment of the Nairobi market  For detailed maps about rainfall patterns and their implications see those which come with the Range Management Handbook of Kenya (Shaabani, Walsh, Herlocker and Walther 1992a,b,c,d). They include maps on meaningful measures of production risks like ‘reliability of rainfall’ and ‘maximum numbers of consecutive months without relevant rainfall during the last 30 years’. 1

1

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

for cattle often extends far into Ethiopia. At times, there are also flows northwards. In a quite recent development (2008), Rendille camels from 300  km south of the Ethiopian border found their way to the border town Moyale and on to Saudi Arabia for slaughter. Most of this cross-border trade is neither taxed nor does it find its way into any statistics. Although this book mainly deals with Islam and ethnicity, two modes of collective identification, it contains a chapter which has a focus on ecology and even two district-level perspectives on livestock management and livestock trade (Chapter 4). The identity politics in which people engage, the social links which they maintain or cut, cannot be divorced from the types of mobility and exchange their livelihood demands. Precarious ecological or economic conditions in given areas and/or opportunities elsewhere may motivate people to maintain or create cross-cutting ties to other peoples or areas. There are also, of course, forces acting in the opposite direction. If territorialized ethnicities and exclusive identities are rewarded by politics, they tend to gain importance at the expense of wider connections and crosscutting ties. Apart from linguistics and economics, a third field in which northern Kenya can be said to belong to the Horn of Africa is that of tribal politics. There have been movements of ethnic expansion across the boundaries of the modern nation states long before they came into existence. The Oromo expansion of the sixteenth century is the most important of these.2 There have also been Somali immigrations which added new elements to the Somali groups present in Kenya since time immemorial.3 The following chapter, on pax borana, will cover a part of this history. Further down, in Chapter 5 and in our companion volume (Schlee & Shongolo 2012), we shall see that ethnic politics in northern Kenya is still interwoven with conflicts going on in Ethiopia and Somalia. The ethnicization of politics in the core areas of Kenya – the Rift Valley and Central Province – is however not less important as a factor shaping the fates of people in these northern lowlands. A HISTORICAL SUMMARY Northern Kenya was incorporated by the British colonial power into Kenya as a buffer zone against rival expansionists, the Ethiopia of Menelik II and the Italians who managed to colonize Eritrea and Southern Somalia and were continuously pushing and pulling for a larger and contiguous East African empire. Finally the Italians managed to subject Ethiopia for a short while, from 1936 until 1941, when the British and their South African allies ended their rule and reinstalled Haile Selassie. Northern Kenya kept rival powers at a distance from those parts of Kenya which the British regarded as an asset: the White Highlands – the high, temperate, fertile lands reserved for white settlement. There was a fresh influx of aspirant white farmers in the course of demobilization after World War I, so new areas needed to be vacated by the Africans, including those who had fought alongside the British against the Germans in Tanganyika. Another asset which required protection was the Uganda Railway. The north itself was a liability rather than an asset.  Hassen (1994); Merid Wolde Aregay (1971); Schlee (1989a).

2

 That there may have been people speaking Somali dialects in Kenya for a very long time does not imply that they referred to themselves as Somali or that they were part of a Somali nation in the modern sense. In fact, as we shall see (Chapter 1) the ‘Somali’ identification has in some cases evolved, and not always in a straight line, only since the early twentieth century. 3

2

Introduction

The British aimed at minimizing the cost for keeping it under their administration: administration was kept at a lowest possible level of personnel. One wonders how the British with one or two of them per district and a couple of Somali or Indian clerks managed to maintain an administration which by those old enough to remember is often favourably compared to that of independent Kenya. Their secret lies in moving around a lot. A District Commissioner was expected to be travelling most of his time, and travel they did and, in the early decades of British rule, mostly on foot. They spent much time listening to people in remote localities and gave them a feeling of participation, even though what they materially did for them was very little. Formal education and health care were practically absent. In Moyale, for example, there was just one primary school which served both Moyale and Wajir districts (now there are 25, plus three secondary schools, in Moyale District alone) and there was a Hospital to which patients came from as far as did the primary school pupils. These practices of minimal interference were in harmony with some romantic ideas about the preservation of contemporary cultures and their capacity for self-regulation – ideas derived from the structural-functionalist persuasions of those days that cultures are equilibrated systems whose balance should not be disturbed beyond the capacity of the system to compensate influences and to re-establish its balance where necessary. The administration struggled to make the districts pay for the cost of their own administration. That was practically all. But there were obstacles on the way to even this modest aim. If poll tax was introduced or raised too high in a border area, entire segments of the population might move across into Ethiopia, like the Garre did repeatedly.4 But to some people at some times in the early colonial period the taxation burden appeared higher in Ethiopia: a group of Boran who came to be known as Hofte moved the other way, into Kenya.5 Tax had to be paid from some monetary income. People were expected to sell animals. But there were also the white settlers who now owned ranches on former Maasai pasturelands and who made sure that their livestock economy was protected against competition from the nomads in the north by a quarantine belt, a special marketing channel for Africans, and a grading system for beef. Some of these structures have survived independence and now protect the African political elite whose members have taken over the ranches, against competition by the grandsons of the nomads of those days.6 In all these aspects the northern lowlands were tied to Kenya, for better or worse. After introducing spatial classifications by discussing in which way these lowlands fit into North-East Africa and in which way they belong to East Africa, let us now try to introduce some classifications of time by dividing the history of the area into periods. Historical periods are always fuzzy because they overlap, and they are always arbitrary because one might as well apply alternative criteria which lead to different periodization. Their only use is rough guidance, a quick and preliminary orientation. But as an introduction is expected to provide just that, this might be the appropriate place to define such periods. Colonialism is an important phase in African chronologies. It is usual to divide history in the pre-colonial, the colonial and the post-colonial phase. In analogy to the pax imperica, the (internal) peace of the Roman Empire, the state of affairs after colonial pacification by the British has often been called  For the first time this seems to have happened in 1923, when the attempt of the British to collect ‘tribute’ from the Garre led to eight Garre casualties and an exodus to Ethiopia. 4

 Interviews by Abdullahi Shongolo with Hofte elders, Daudi Dadacha and others.

5

 Raikes (1981); Schlee (1984, 1990b, 1991a, 1998, 1999); Walz (1992: 35–41).

6

3

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

pax britannica. The British themselves have not failed to notice that before them there was another hegemonial power in the area which succeeded in maintaining stability for relatively long periods: the Boran. In colonial records we find the phrase pax borana. For the period before the coming of the Boran we have no written records. But we can conclude from oral history and cultural comparisons that people who spoke Somali-like languages not unlike modern Rendille, and who had a culture or cluster of related cultures which had camel husbandry at its core, were widespread in the area. I have called this stratum Proto-Rendille-Somali or PRS (Schlee 1989a). We can, on the ground of these distinctions, define the following four periods: The Proto-Rendille-Somali era

before 1550

pax borana

c. 1550–1920

pax britannica

c. 1920–1963

Independence

since 1963

These periods become shorter as we approach the present. This may simply have to do with the fact that we know more about the recent past than about the remote past and therefore apply more distinctions to the former. The following summary of the history, broken down into these periods, is shorter than that in Identities on the Move (Schlee 1989a) and only serves the purpose of making the present volume self-contained: to round it off by putting the different parts into one historical framework for those readers who do not have the earlier work at hand. It is difficult to mark the beginning of the Proto-Rendille-Somali era. The Somali dialects in the south differ much more from each other than those in the north. As languages tend to have undergone more differentiation in their old areas of distribution than in new ones, it is plausible to conclude that Somali origins need to be looked for in and around the southern parts of the present distribution of peoples speaking Somali dialects or Somali-like (Somaloid) languages. As the presence of the Somaloid Rendille language shows, what is now northern Kenya was part of this early, probably pre-Islamic or marginally Islamized Somaloid dispersal. Also the existence of the related Baiso language in Ethiopia makes us think of the present Kenyan-Ethiopian borderland as a candidate for an early nucleus of Somali-like peoples.7 Apart from some form of early Somali-like speech the other criterion for classifying people as Proto-Rendille-Somali is the cultural emphasis on camel husbandry and the associated calendar and social organization. This cultural complex may have developed in the area with the arrival of the camel or may have been brought along with it at some unknown time, presumably in the Christian era (A.D.). The clan histories of nearly all Rendille and most Gabra, Sakuye and Garre seem to go back to groups which were once characterized by the combination of features characteristic of the PRS. To the east, in the complexities of Southern Somalia local cultures and dialects, it is difficult to say which groups derived from the PRS. Some may have preserved PRS specific traits, others may have lost them and yet others, maybe groups with a stronger orientation to cattle and farming rather than camels and smallstock, may never have been typical representatives of the PRS complex. As the 4

 Fleming (1964); Lewis (1966); Lamberti (1983); Schlee (1987).

7

Introduction

focus of the present work is on northern Kenya and southern Ethiopia, we can reserve the question of cultural strata in Somalia for future research. The next phase, that of Boran hegemony, is already shorter. The expansion of Oromo from their southern Ethiopian cradle lands started around the middle of the sixteenth century. The first Oromo group to move south were the Warr Daya whose remainders can now be found among the Tana Orma (Schlee 1992b). We assume that the Oromo expansion took place more or less simultaneously on all sides, from the sixteenth century to the eighteenth century, to the north and west as far as Wollo and Wollega on the present border between Ethiopia and the Sudan, to the south and east as far as the Indian Ocean. The most important group to represent the Oromo in northern Kenya were the Boran. They played a hegemonial role in the area which was taken over by the British around 1920, when they succeeded to establish colonial rule more or less effectively on the local level. The Oromo expansion had important effects on the ethnogenesis of non-Oromo groups. Many PRS groups tried to evade the periodic raids by the Boran whose freshly initiated raaba age-grades had to undertake ritual warfare and to secure loot in the form of livestock and the cut-off genitals of slain male enemies. These initiations took place once in eight years. The Gabra Miigo were separated from the Garre in this period, because the Gabra Miigo submitted to the Boran while the Garre first fled and only submitted after a long circular migration which took them into Jubaland in Southern Somalia. Also the ancestral Gabra were separated from the Rendille in this period. The fact, that many former PRS groups now speak Oromo while others, the Rendille and some Garre, have maintained their Somaloid languages, goes back to this period. The Rendille are the only modern ethnic group among those which mainly descend from bearers of the PRS culture who have neither become linguistically assimilated by the Boran Oromo, nor undergone Islamization and inclusion in the wider category ‘Somali’.8 The Boran managed to incorporate many non-Boran groups in a network of ritual dependence on their two High Priests, the qallu9 of the Sabbo moiety and the qallu of the Gona moiety, and to unite them into an internally peaceful military alliance, the Worr Libin, called after an area in southern Ethiopia. The groups they managed to incorporate into this alliance in what now is northern Kenya are all of Somaloid origin. They include the Gabra (Malbe and Miigo), the Sakuye who appear to have emerged as an ethnic group only after the Boran expansion and, later and less permanently the Garre, parts of the Ajuran and indirectly – under the Ajuran umbrella – groups of Degodia-Somali. No Maa-speakers like Laikipiak Maasai or Samburu became part of it. They always remained outsiders and enemies to it. It is not clear by which point of time the Boran had become established to this degree, but after it, it is surely justified to refer to this period as pax borana. In the 1920s and 1930s, after set-backs during the First World War, the  There have been some recent politically motivated polemics accusing me of anti-Oromo and pro-Somali leaning. I was said to ascribe ‘Somali’ origins to the Gabra and others. It therefore needs to be stressed once more that the PRS were not Somali in the modern sense (‘modern Islamic Somali’ as I used to say) but the bearers of that cultural stratum from which many modern Somali and non-Somali groups (Rendille, Gabra, etc.) derived to a large extent. To put it very bluntly: they were among the ancestors of some (not all) Somali and many non-Somali. Some of these polemics have been summarized in the introduction to Baxter, Hultin and Triulzi (1996). More of it can be found in non-reviewed internet sources. A somewhat more scholarly discussion of the controversy about these issues can be found in Kassam (2006) and Schlee (2008b). 8

 Variously spelled qallu, qaalu, qaallu, qaalluu etc. We prefer the first two. The variation appears to be due to compensatory lengthening: if the geminate consonant is simplified, the vocal is lengthened. 9

5

Figure 0.1  Colonial movement permit, in possession of Abdullahi Shongolo, the son of the holder

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

6

Introduction

British established and slowly rigidified their rule. For some decades before, there had been an influx of Somali groups from the east, by people who claimed the status of pseudo-relatives or genealogical dependents (sheegat)10 of the Somali groups, mainly Ajuran, who were part of the Worr Libin. These immigrant Somali had started to overcrowd the Boran and their allies at the wells of Wajir and elsewhere.11 The British responded to this by delineating tribal grazing lands. With motorized travel setting in, the tracks connecting the towns and trading centres of northern Kenya often acquired an additional function as grazing reserve boundaries. A system very similar to that of apartheid, later formalized in South Africa, developed in colonial Kenya. Not only was there a division between White and Black settlement areas (the ‘reserves’, risaaf), not only was the white economy given privileges and shielded against competition by Africans, nay, within the African sector itself numerous boundaries were also drawn. Like the Asians (Indians) and migrants from other parts of Kenya, the northern Somali veterans who had fought for the British in the First World War and who, often after serving additional years as Dub ’as, [‘Red Fez’, later: Administration Police, AP] in Kenya, had opted for a lump sum rather than a pension, were not allowed to live as traders among the Rendille, Gabra, or other local groups. Alien traders were restricted to the towns. Many of them decided to marry into local societies, typically at the upper end, to be allowed to live among their customers. Pass laws, like in apartheid South Africa, also existed in Kenya. To travel from one district to another one needed a permit. As can be seen from the pictures of one such permit (Figure 0.1), this system endured for years after Independence. The ethnic macro-categories which the British recognized in northern Kenya were Galla (a term which later came to be replaced by ‘Oromo’) and Somali. The most important of the new boundaries drawn was therefore the Galla-Somali line, which, after a couple of shifts, developed into the present boundary between the Eastern and North-Eastern Provinces. The term ‘Galla’ referred to Oromo speakers like the Boran, Gabra, and Sakuye. ‘Somali’ by the Somali themselves and by the British, who became keen genealogists, was defined by patrilineal descent from Somali ancestors. Inconsistencies predictably developed wherever people claimed Somali genealogies but were found to speak Oromo.12 Thus not all minor units fitted very neatly into the major ones of ‘Galla’ and ‘Somali’. Much of the subsequent chapter, Pax Borana, will be about the Ajuran re-affiliating themselves from the former to the latter, and further down in chapter 5 (and in our companion volume, Schlee & Shongolo 2012), we shall learn how and why at least some of them now wish to revert to the prior situation. Being Somali and being Muslim is, in this part of the world, closely intertwined. Islamization = Somalization was the other factor which differentiated the former PRS. As some of them acquired Boran as their new language and became part of the Worr Libin, others – and not only others, because the process was also at work among the Worr Libin – became Muslims and came to be regarded as a kind of Somali, irrespective of their language, be it Boran,  From the verb sheeganaya – ‘I name’ [somebody else’s ancestor as my own] which in turn is an autobenefactive derivate from sheegaya – ‘I tell, narrate’. It is used for unrelated people who have rendered themselves under the protection of a strong clan. 10

 Kenya National Archives, summarized in Schlee (1989a).

11

 About the linguistic complexities associated with being ‘Garre’ cf. Schlee (2001) or Schlee (2008a: 100, 101). 12

7

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Map 0.1 Boundaries in colonial Kenya (based on Schlee 1999: 222)

8

Introduction

Somali or a Somaloid language. These identifications are still relevant in the most recent ethnic clashes. We shall see below that Islam, a universal religion with a great potential to bring peace and understanding, has been used mostly in the opposite way in the regional context: to define others as enemies and to draw lines of distinction between people who have been Islamized some time ago and people who have become Muslims more recently, to denounce the latter as kufaar, unbelievers, and to declare war against them to be jihaad. In many ways the British colonial system can be seen as the precursor of the modern ethnicization of politics. It not only sets the pattern by allotting chunks of the territory to certain categories of people (often defined by rather arbitrary criteria) which may or not have to do with applying the model of the modern European nation state on a smaller scale in Africa. Many of the boundaries drawn then now serve a new generation of strategists of exclusion as legitimization of murder or the expulsion of people who originate from other parts of the country. Shortly before independence, in 1962, the British held a referendum, asking the people of the Northern Frontier District (NFD)13 whether they wanted to belong to the newly emerging independent Somalia or to Kenya, which was on its way to independence. Uhuru na Somalia (‘Independence with Somalia’) and Uhuru na Kenya (‘Independence with Kenya’) were the competing slogans. There was a majority for Somalia. But as the counting was over, negotiations with the delegation led by Kenyatta in London were too far on their way. The process towards independence of Kenya in the colonial boundaries which comprised the NFD could no longer be reversed. The British ignored the results of their own referendum. There was no better way of telling the Government of Kenya, which became independent in 1963, that the north, with the exception of certain Boran inhabited areas, was a hostile territory which would have to be taken by force. Violence erupted immediately. A Somalia sponsored separatist guerrilla movement, known as the shifta from the Amharic word for ‘bandit’, operated in northern Kenya throughout the 1960s and Government retaliation was drastic, especially against unarmed civilians, mainly the pastoralists, who were preferred as targets because they could not fire back. Isaaq and other northern Somali traders (the above mentioned veterans and their descendants) were restricted to Isiolo, the Sakuye were concentrated in camps, ‘keeps’, where those parts of their herds which had survived machine gunning by Government troops died of hunger, because the interdiction of nomadic movements led to immediate overgrazing of the surroundings of these ‘keeps’.14 Given the close links the Sakuye had had as members of the Worr Libin alliance to the Boran, it can be debated whether it was their rather recent conversion to Islam, their PRS-derived camel culture or this treatment by the Government which made many of them join the Somali shifta. After the dissolution of the ‘keeps’, the surviving Sakuye were marched by Boran guards from Eastern Province to North-Eastern Province, a march which turned out to be a

 The Northern Frontier District is a name which remained in use for a long time after the administrative unit it referred to had been abolished. This new entity was the amalgamation of British Somaliland and the Italian UN trust territory Somalia which lasted from 1963 to 1991. 1991 is the year of the expulsion of Siad Barre from Mogadishu which was followed by the era of the warlords. De facto, Somali unity had ceased years before, at the latest with the aerial bombings of Hargeisa by Siad Barre in 1988. On the other hand some people claim that Somali unity exists even now, although that must be an existence on paper. 13

 Oral accounts.

14

9

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

trail of blood. Babies were torn from their mother’s backs and smashed against trees.15 The last decades of British rule had deserved the name of pax britannica. Violence was at an all-time low. After independence, raiding started to resume. In the Rendille age-sets Ilmeuri, who were warriors from their circumcision in 1937 to that of the subsequent age-set in 1951, and Ilkimaniki16 (1951–1965), I do not know a single man who has the honoured killer status. In the following age-set, Ilkichili (1965–1979) there are about 300. The killer beads worn around the neck do not show whether the bearer has killed one male enemy or more then one. The number of killings is larger than the number of killers. The killing of a woman does not entitle one to the killer status, but that does not mean that women are not killed. In former times they were often taken as captives and subsequently married as second wives by members of the groups of their captors. Government control, ineffective as it may be in other ways, now prevents this. Therefore, in modern forms of raiding, women are killed. If we double the number 300 to estimate the number of victims of this age-set, and assume that the Gabra and Turkana inflicted similar losses on the Rendille, we get a considerable death toll of violence, especially taking into account that in 1962 there were only about 8,000 Rendille.17 The Rendille themselves do not attribute this sudden increase of killings to the end of the pax britannica. They are aware that a major proportion of their victims were Sakuye and Somali who had come to raid the neighbouring Samburu because they had been deprived of their means of subsistence in the troubles of the shifta emergency. They got into an ambush at the Larabasi pond on their way through the Rendille area. The Rendille also see the connection between the absence of effective Government control and the increase of hostilities between themselves and the Gabra. But the main factor to which they attribute the uneven distribution of killers to the different age-sets is daji, the return of everything after two generations or six age-set periods (Schlee 1979: 89–96). As their Dismaala grandfathers who were initiated 84 years before them, Ilkichili are an ‘age-set of blood’, strong but troublesome. They told me while Ilkichili were still warriors, that with the circumcision of the following age-set the killings would be over. Although the security situation in general has not improved, so far the Rendille theory has proved correct. There are very few killers in the age-sets Ilkororo (circumcised 1979) and Ilmole (circumcised 1993). To a small part this decrease may be due to the proliferation of both legal and illegal arms among them. If you shoot somebody at a distance, it might not be wise to walk up to him to see if he is properly dead. He might be lying in his blood between the lava boulders with his finger at the trigger, waiting for somebody who does just that. Also the local chiefs and other government representatives who formerly looked the other way when the rather elaborate ceremonies for those who have killed an enemy were carried out nowadays might ask questions inspired by the penal code. For these reasons more men might be killed than genitals cut and killer insignia acquired. But even allowing for that, the number of people killed by the subsequent age-sets is nowhere near that of those killed by Ilkichili. This recent decrease, however, is a local  Numerous oral accounts.

15

 These are the Samburu names which have a wide circulation. The proper Rendille names of these age-sets are Libaale and Irbandif. 16

 The 1962 census gives the figure as 13,724. From this the estimated number of Ariaal needs to be discounted to arrive at the population of the Rendille in the narrow sense (Spencer 1973: 4). 17

10

Introduction

peculiarity. In northern Kenya in general violence is now at a level unknown during both the pax borana and pax britannica periods. The colonial forerunner of the Administration Police had units mounted on camels which could actually patrol the area and pursue raiders and thieves. The range of action of the post-independence motorized police stops at the base of each lava field, as no four-wheel drive vehicle can move across areas strewn with basalt rocks half a metre in diameter. The local nomads can walk across them; heavy booted, beer-bellied policemen from Central Kenya cannot. Personally, I mostly left the task of walking through these arduous stretches to the camel I was riding. In 1974 the emperor Haile Selassie was deposed and the Ethiopian government was taken over by a council of military officers, the Derg. Through a process of elimination, over the years Mengistu Haile Mariam emerged as the ruler of the country. The oppressive nature of that regime in combination with the war it waged in Eritrea for which young men were forcefully recruited from all over the country, led to growing numbers of Ethiopian refugees in Kenya. Marsabit and Nairobi had increasing Ethiopian populations which enriched the culinary scene with Ethiopian restaurants and became an important part of business life. Many Ethiopians also moved on to Europe and America. Later refugees who were intercepted were restricted to a huge camp run by the UNHCR18 at Woldá near Turbi, in the shrub land just south of the Ethiopian border. By 1991 the Eritreans had gained the upper hand and their comrades in arms, the TPLF (Tigrayan People’s Liberation Front) marched into Addis Ababa. Mengistu went into a comfortable exile in Zimbabwe and left the remainders of his followers to fend for themselves. His disintegrating troops made it into northern Kenya, sold their weapons for a meal and flooded the country with Kalashnikovs to an unprecedented level. Then they mostly ended up in the same camp as those who had escaped from them earlier. Earlier the same year President Siad Barre had had to flee from Somalia. That is the conventional date given as the beginning of the period of turmoil and the warlords in that country. It had been Siad Barre who had created rifts between the clans of Somalia which were deeper than ever before19 and who had been at war with his own people20 for many years. He was at the peak of his power in the Ogadeen war (1977–78), when he managed to invade and occupy for a brief period the part of Ethiopia which is inhabited by Somalis. With the pan-Somali cause lost and clanism becoming more and more divisive within Somalia itself, he had the loyalty of a rapidly shrinking part of the population. As I have summarized that history elsewhere (Schlee 2002), suffice it to say here that Somalia was in pieces long before Siad Barre finally left. The state of affairs in that neighbouring country did nothing to improve the security situation in Kenya. There was a huge influx of refugees, trade in arms, and at times the Somali war spilled over when Somali militias pursued each other onto Kenyan soil.

 (Office of) the United Nations High Commissioner for Refugees.

18

 He made it a legal offence to talk about clans while basing his own power on an increasingly exclusive clan alliance. Cf. Schlee (2002 and 2004) on the role of clanship and other factors in Somali politics. 19

 ‘Somalia: A Government at War with its Own People’ is the result of an Africa Watch Report (1990) which reflects these events. 20

11

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

ECOLOGICAL FACTORS After the chronological one, an overview of the ecological conditions may be in place here. About half of northern Kenya east of Lake Turkana falls into Pratt’s and Gwynne’s21 eco-climatic zone VI, ‘very arid’, with a discontinuous vegetation of dwarf shrubs and some mostly annual grasses. This half is mostly the western half, the Rendille and Gabra country. Gabraland is the harsher of the two, not only because of its climate but also because of its less even distribution of water spots and its surface formation. Much of it is strewn with large lava boulders which make it not only difficult to live on something in this country, but even to move through. The other, the eastern half, the Somali area around Wajir for example, falls into zone V, ‘arid’. It is more densely wooded with thorn shrubs and trees and has a higher proportion of perennial grasses. This large arid and semi-arid area is interrupted by some inselbergs: Nyiro, Kulal, Marsabit, the Hurri Hills, and the promontories and outliers of the Ethiopian highlands along the international boundary. On these mountains we find all climatic zones up to the ‘humid to sub-humid’ tropical climate with thick forests. Most of the hot dry lowlands lie around an altitude of 500 m while most forests can be found above 1,200 m on the eastern side, the weather side of the mountains. The peaks of Marsabit and Kulal are between 1,650 and 1,900 m high. The forest often ends abruptly on the peak and gives way to bare western slopes of grassland and shrubland. Mt. Nyiro in the Samburu District is 2,700 m high and in the southern part of that district the Leroghi Plateau is not just an isolated mountain but a vast temperate area of above 1,500 m. All nomads have to keep large numbers of smallstock for domestic slaughter. Cattle and camels are too large for a family to consume before the meat is rotten and wide distribution in the hope of later reciprocity puts an unrealistically high strain on human morality. So we find sheep and goats in all climatic zones because they are of the right size for a meal. Large stock, which are mainly kept for milk, however, have characteristic spatial distributions. We find mainly camels in the vast, low, dry, hot areas and cattle in the limited, high, wet, more temperate areas. This is not because the average year in the lowlands is too dry for cattle. It is because a dry year in the lowlands is too dry for cattle. Because of the erratic nature of rainfall in the low areas there are few average years and enough dry years to cut the cattle population down from time to time. Traders gamble on these disasters: one can buy a herd of cattle cheaply in a long drought and either lose them all when the drought continues or if they are too weak to survive the chill of the first rain, or make a huge profit if they do survive. Camels, on the other hand, are not objects of speculation but comparatively safe investments. The use of biological metaphors is dangerous in social sciences. One is easily accused of ‘biologism’ and put into the conservative corner. Let me therefore clarify: in the world of experience we all share, there are two distinct systems of information. One is genetics and everything which is guided by the genes, the other one is culture in its widest sense: the learned components of our behaviour, the sum of our knowledge, our norms and values, all acquired skills, and all systems of signs we have developed like traffic signs, arts, music, natural and computer language. Both systems are interlocked in many ways. Any task, even a simple one like drinking from a cup, requires the coordination of complex organic structures, instinctive behaviours and a great deal of experience, as any parent knows. But there is no bridge between the two systems. 12

 Pratt and Gwynne (1977: 42, map I).

21

Introduction

One cannot feed information from one of them into the other.22 But parallels there are many. It is no coincidence that one applies a linguistic term to biology by talking about genetic codes.23 The inverse case is more frequent. As biology has been ahead of social science in its development for over a hundred years, social and cultural sciences have borrowed many theoretical concepts from biology. One of them is the ecological niche. Two species of animals which eat the same food in the same type of habitat at the same time of the day cannot coexist. If there are two such species, they must be divided by a geographical barrier. If the geographical barrier breaks down as was the case with American squirrels which were taken to Britain, one of the species will have some slight advantage over the other one and replace it in the long run completely. Now there are almost only American grey squirrels in Britain instead of the British red ones. Human beings can change their strategy more easily than animal species which often die out before their genetic outfit has adjusted by variation and selection to new circumstances. When European ship builders were obliged to notice that the Japanese could produce large tankers as good as and much cheaper than they could, they did not react by perishing but by building different and more sophisticated types of ships. Some even stopped building ships and used their yards to weld together various types of land-borne industrial equipment. That is how European shipyards survived for a while – until the Koreans entered the market. Whether we talk of ecology or economics, the principle remains the same: we cannot coexist with those who do exactly the same things as we do. One of the competitor groups, either us or them, must have a slight advantage over the other and replace them in the long run. This replacement in the case of human beings does not necessarily take the form of one dying and the other taking over the resources, although frequently it is just this that happens. Identity politics may enable people to change sides and to become the other. Unlike the red squirrels which had no way of redefining themselves as grey and did not mate with the grey ones, people who give up their separate social identities may continue to reproduce biologically, be it in isolation (if, for example, the newcomers are not fully accepted and practice involuntary endogamy) or in the form of mixed populations. In fact, at different periods, many Rendille have become Gabra, others Ariaal or even Samburu. Some of these dynamics of sloughing off people who are then accepted by more accommodating neighbours have been discussed elsewhere (Schlee 1989a: 49–50). The Rendille and the Gabra Malbe are both camel herders along the fringes of the Chalbi saltpan. Many Rendille are named after places of birth well to the north-east of the Chalbi, where now Gabra live. Rendille and Gabra are competing for the same pastures. What one of them breeds, is a valuable prey to the other. In the age-set Ilkichili, warriors from 1965 to 1979, one could see many killers wearing necklaces of white beads. The Elemo sub-clan (Schlee 1989a: 197, where the underlined names in the genealogical chart stand for killers) is by no means untypical. Nearly a third of the men of this age-set wear killer beads and many of them have killed more than one man. With interruptions by peaceful periods, Gabra/Rendille hostilities continued in the 1990s. Because of the clan links between the two ethnic groups, some of the consequences of these hostilities were rather paradoxical (Schlee 1997). The northern neighbours of the Gabra are the Boran, cattle herders who seasonally disperse into the lowlands from their central areas on higher ground  This was written before I became aware of a field of knowledge called epigenetics.

22

 Also, whether capitalism inspired Darwinism or Darwinism capitalism can be debated.

23

13

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

and along the southern promontories of the Ethiopian highlands. The Boran on Marsabit Mountain also predominantly occupy the higher ground. Gabra and Boran clearly do not share the same ecological niche. Their long-standing, mostly peaceful relationship is part of what we describe below under the heading pax borana. The friendly relationships between the Gabra and the cattle keeping Boran are mirrored among their southern enemies by the close alliance between the Rendille and the cattle keeping Samburu which is the theme of Spencer’s book (1973). The recent loss of peace, like the deterioration of the Gabra/Boran relations, can often be explained by a loss of specialization and an increase of competition for the same resources. Several examples of such processes can be found below. Attitudes towards strangers, exclusion or assimilation (Schlee 1989a: 49–51), are not fully explained by ecology. But, at least with respect to warfare in northern Kenya, ecology explains a lot. Those who are most likely to fight are those who strive for the same food resources, whatever ideological, historical or political reasons for fighting may be given in addition to this. Diversification and combined strategies do not blur the boundaries of ecological niches but define new and distinct ones. As in the animal kingdom, where there are some species with specialist and others with generalist food habits, we can find specialist and generalist types of people, strategies, organizations and enterprises in all walks of economic, social and political life. Such a mixed strategy among northern Kenyan pastoralists is represented by the Ariaal (Spencer 1973) who linguistically, culturally, spatially and socially straddle the Rendille/Samburu boundary and have a mixed cattle-cum-camel economic strategy. Despite separate camps for camels and cattle their habitat and the grazing cycles made possible by it might neither be optimal for camels nor for cattle but reasonably good for both, and their double strategy might enable them to better balance risk and a fuller exploitation of their resources and thus give them – at least in one year or another – an edge over their specialist neighbours. The ‘Hofte Boran’, now sedentarized on Mt. Marsabit, seem to have straddled the Gabra/Boran boundary in a similar way. Ethnic re-affiliation might be sought according to herd-mix. Somebody who has a number of cattle but does not want to give up his camels or entrust them to others might be tempted to join the Ariaal or Hofte (depending on whether he is a Rendille or Gabra) or, if he wants to try his luck with cattle only, the Samburu or Boran. (But see Schlee 1989a: 49ff for other factors influencing such decisions apart from ecology.) An Odoola Rendille elder24 explained the reason why many members of his clans had joined the Ariaal like this: lolyo lolyo raahta – ‘cattle follow cattle’. Another Rendille saying points into the same direction: gaal lolyo magarti – ‘the camels do not know the cattle’; they have different grazing and watering cycles so it is often useless to ask in a camel camp about the whereabouts of a cattle camp: the herdsmen might not know. The differences in the grazing and browsing preferences of livestock species correlate with the spatial and social distance between their herders. Mixed strategies do not only concern herd mix. Gudrun Dahl (1979) has explained how the modern sector (the market, irrigation agriculture, wage labour by one member of a family, trade labour by one member of a family, trade, etc.) is incorporated into the risk-balancing and diversification strategies of the Waso Boran. The pastoralist who intelligently opts for one or the other strategy has to make a complex cost-benefit analysis. Sending one family member to town for wage labour might reduce the labour available for 14

 Ballo, interview Hammaleyte, June 1979.

24

Introduction

control and care of livestock, while staying closer to the market might expose one’s stock to epidemics, and dispersal of the family in different occupational niches might reduce their coherence, etc. Every advantage must be paid for by a disadvantage somewhere else. One can therefore not say a priori that generalist strategies are better than specialist ones or vice versa. All one can say is that they are different in perfect analogy to ecological niches of biological species which differ along the same lines. One key factor in the ethno-genetic processes which led to the ethnic map of northern Kenya as we know it today was the Boran hegemony. Let us once more consider how it came about and why it lasted so long. In the chapter on pax borana (below), and more fully under the heading ‘The Long Trek’ in Identities on the Move (1989a: 92ff), we have explained that the establishment of tiriso relationships, i.e. adoptive clan relationships, between the Boran and PRS peoples was a way by which the latter sought protection from the ritual raids and homicides of the former. The Boran hegemony and the spread of the Boran language were results of the pacification of large numbers of others. It is difficult to determine to what extent this development was accompanied by an expansion of the human and livestock population of the Boran. The removal of the Boran from Wajir and their resettlement in Isiolo District was, as we have seen, the result of a colonial policy and it is difficult to guess how the ethnic distribution map would now look if the local forces had been allowed a free interplay. But it is clear from the colonial records that even before their expulsion the Boran only formed a fraction of the Wajir population. In the Marsabit District, where no expulsion of Boran ever took place, the Boran only inhabit the comparatively favourable cattle areas on Marsabit Mountain and along the Ethiopian border in what is, since 1997, once more the Moyale District. The pax borana seems to owe some of the stability it has had over long periods of time to the fact that the Boran restricted themselves to their niche. They left the lowlands at least for the dry season to their camel-keeping Worr Dasse dependents, the ‘people of the mat [tents]’. We should, considering the principle of the ecological niche, rather expect conflict between different groups of the Worr Dasse than between Worr Dasse and Boran.25 Also the causes of conflict which involve the Boran as such, and not just as allies of a rival Worr Dasse, might have to do with the principle of the ecological niche, or rather with the breakdown of the borders of these niches. Some Boran have recently acquired large herds of camels, and Garre and Boran are competing for the same types of business in town. So much, in form of a rough sketch, about the regional and chronological setting of the subsequent chapters. ON THE HISTORY OF THIS BOOK One substantial chapter of this book has been co-authored with Abdullahi Shongolo. Abdullahi is a Boran from Moyale and a local scholar with a profound interest in the history of his group. By profession he was a primary school teacher, later a headmaster and still later part of the administration of education at district level. One of his main concerns is the preservation of traditional Boran wisdom and the production of written texts in the Boran language. In 2004, his efforts were rewarded by the President of the Republic of Kenya, by awarding him the Head of State’s Commendation (HSC) – Civilian  More on this in our book Pastoralism and Politics in Northern Kenya and Southern Ethiopia (Schlee & Shongolo, 2012). 25

15

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Division. Apart from teaching materials for schools, he has collected Boran proverbs and has written stories to illustrate their use (Shongolo and Schlee 2007). In the present volume, chapter 5 goes back to a text we have published together earlier. For a portrait of Abdullahi Shongolo and more bibliographical information the reader can consult Max Planck Institute for Social Anthropology Report (2004–2005: 86) (www.eth.mpg.de), and for more information about our cooperation the introduction to Shongolo and Schlee (2007). For other parts of this book I have the sole responsibility. They have been written at different times for different purposes and betray their original contexts to some degree, although I have also tried to make clear why they are relevant in the present context of ethnic and religious identification. The oldest texts included here go back to my Habilitation thesis, submitted to Bayreuth University in 1985. That book was found to be too voluminous for a marketable book publication. So I decided to focus on just one aspect it covered, namely the clan links between the peoples of PRS derivation, and published a much more slender volume as Identities on the Move in 1989. This has given rise to the criticism that my description is one-sided. It was said to overstress the PRS elements or to have a Rendille bias. This criticism was misguided: in the book I say quite clearly: I cannot present a complete or even well-balanced picture of all the different cultures studied. Other aspects of these cultures have to be studied elsewhere. One entire dimension of cultural history has had to be left out almost completely: the early and recent Islamic influences (Schlee 1988). I also have to reserve for the future a fuller treatment of the influences of the African cultures of the interior on the PRS-derived cultures. Some political institutions of the groups in question (hayyu, jallab, etc.) seem to have been fashioned after Boran models, and linguistic and cultural borrowings from the Boran can even be shown in the case of the Rendille, the group most remote from Boran control. Much would need to be said also about Samburu and possibly other Maa influences on elaborations of Rendille age-set rituals, on Rendille youth culture and on Rendille kinship terminology. All these cultural strata deserve the same attention as . . . the PRS culture . . . (Schlee 1989a: 90f)

16

Many of the ‘omissions’ for which I have been criticized, like not giving enough emphasis to the role of the Boran Oromo, refer to things which were there in the original text. The following chapter on pax borana comes straight out of my 1985 thesis. In the published version of that thesis, Identities on the Move, I have devoted a chapter to the reconstruction of a complex of cultural features common to the modern Rendille, Gabra, Sakuye and some groups which are now counted among the Somali. This complex of features, the Proto-RendilleSomali complex, has already been repeatedly mentioned in my attempts to give a rough chronological and terminological order to events and cultural influences in this introduction. Traits which can be attributed to the PRS complex can be found in an area which stretches from the Rendille in the west to Somali-speaking groups in the east and seem to form a layer historically underlying the modern cultures of this area. This complex comprises a particular rather complex calendar which takes both lunar and solar cycles into account without trying to make the two match in any way by intercalary units, it further comprises a set of rules concerning camels which is linked to the calendar because the rules often state what to do or not to do with camels on a particular day or in a particular month, and last but not least it provides a grid of old clan identities which predates the emergence of the modern ethnic

Introduction

groups and is responsible for the interethnic clan identities which form the main theme of that book. All that has been covered extensively in Identities on the Move, and in order to have enough space for that, I cut out much of what I had found out about non-PRS elements in northern Kenyan pastoralist culture. Within the broader scope of the present volume, I can now come back (in chapter 2) to these non-PRS-derived elements of local culture of the Cushitic lowland pastoralists as well, like those brought by the Nilotic Samburu or the Boran Oromo. Chapter 3 also contains texts which were part of the Habilitation thesis (1985). Due to the focus on the PRS stratum and what can be derived from it, much of what I had written about modern world religions, Islam and Christianity, and their role in the area had been left out of the published version. All that has now become part of chapter 3, ‘Modern Trends’. Of course, as a quarter of a century has elapsed since, I have tried to update these texts for the present volume. The section on ‘Acceptance and Rejection of Christianity and Islam by the Rendille’ goes back to a paper published in a rather hidden place in German in 1982 (Schlee 1982b). I left that chapter as it was and restricted updates and reflections from 2007 and 2008 to the footnotes, so that the chapter now reflects two clearly distinguishable levels of time. Chapter 4 is based on two reports I wrote for the Range Management Handbook of Kenya (Shaabani, Walsh, Herlocker and Walther eds) which were first published in 1992. They have a clear development focus and stress the necessity to keep the range open, the borders permeable and to help the pastoralists to maintain high levels of mobility. A critical reader who is very well versed in the debates about pastoralism has reminded me that this was the emphasis of the 1980s. To most students of pastoralism this is no longer new. I can only add that it was not new to them even then, and even now it has not been generally accepted by those who take decisions which affect the pastoralists. For an effective use of the scattered natural resources a mobile form of livestock economy is required, and even now, after many more years in which obstacles to pastoral mobility have continued to grow and the cause of maintaining mobility has been lost in many cases, it is justified to reiterate this point to administrators and politicians to whom it still does not appear to be evident and whose lack of comprehension has a much stronger impact on the lives of the pastoralists than the comprehension or otherwise of scholars. In rewriting the chapters in response to this criticism, I have tried to make clearer what mobility implies for the present context of ethnic and religious identification. I did so without cutting out the original emphasis on development issues, which gives these chapters much of their internal coherence. Although the order of the chapters in this book was guided by other considerations, it happens to correspond to the order in which they were written. The first chapters contain lots of older text, the later ones larger proportions of newer text. If the reader thinks that something sounds a bit dated, he or she should just read on and might find recent developments which let the older parts appear in a new light in later chapters. I have become part of local history as well. Rather than re-writing my texts from the 1970s, 1980s or 1990s in the latest jargon, I leave them as they are as a reflection of what I perceived then. Local identity discourses have taken up my writings and made all sorts of legitimate and illegitimate use of them. In combination with my writings from the 2000s and the writings of others, the reader will be able to make his or her own synthesis. 17

1 Pax Borana

There is a progression of tales widespread in the lowlands of northern Kenya and the Horn, variously known by its Somali name as kedi guur, the time of migrations or, by the Gabra, gaaf Iris, the time when people went to Iris. These tales describe how, in what must have been the sixteenth century, people fled from the Boran while others remained behind, or how treks of migrating nomads got cut into two by emerging or re-emerging bodies of water (the Moses motif). These stories refer to the ethnogenesis of the modern ethnic groups out of the earlier Proto-Rendille-Somali (PRS) stratum. They describe how the Gabra Miigo were separated from the Garre, the Hawiyya Somali from the ancestors of the Rendille and Sakuye, the Gabra Malbe from the Rendille. Because of their importance for the origin of the interethnic clan relationships these tales have been discussed at some length in Identities on the Move (Schlee 1989a) and are indeed the main theme of that book. With the exception of the Rendille who remained hostile to the Boran for centuries and might have been saved the choice between annihilation and final submission by the arrival of the British in the early twentieth century, all other Cushitic groups who now live in northern Kenya later submitted to the hegemony of the Boran. They include groups like the Garre who have traditions about escaping from the Boran to the east and later coming back. The return of the refugees from their exiles and their ritual affiliation to the Boran marked the beginning of a multiethnic political system that was the major unifying factor in the region before the introduction of modern statehood. In what is now Kenya this system comprised the Gabra (Malbe and Miigo), Sakuye, Garre, Ajuran and Warra Daya, and other peoples belonged to it on the Ethiopian side. These peoples were also collectively known as Worr Libin or, in Somali, Reer Libin, i.e. the people of Liban after a landscape in southern Ethiopia. Liban is not only named as the homeland of the Warra Daya and Boran in the oral traditions but has also, on the grounds of the analysis of written sources, been described as belonging to the central regions of Oromo expansion in general (Braukämper 1980c). The Boran of Dirre and Liban even today are regarded by other Oromo as the occupants of the former common homeland, the guardians of the holy sites and the keepers of traditions. Their political and ritual influence, however, extends far beyond the ethnic boundaries of the Oromo.

19

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

20

Figures 1.1–1.5  Moieties and major divisions of ethnic groups mentioned in the text. For more detailed clan lists and genealogies see Schlee (1989a)

Pax Borana

The earliest military expansion oral traditions mention is that of one Oromo group at the expense of another. Thirty Boran warriors are said to have expelled the Warra Daya from Liban. The warriors had eight tresses (gutu) each instead of only one as usual. Boran with eight tresses can allegedly be found in Liban even today.1 The Boran are reported to have encountered the Garre next: Gaaf [Boraní] Liban gad d’ufe, Garriit itti ga bae. Duub chuf gos Borána dibbi hin qábdu: Karrayyu himati. Qodaalle hin qábdu. When the Boran came from Liban, they encountered the Garre. They all now have no other Boran clan: they call themselves Karrayyu [i.e. all of Garre became affiliated to Karrayyu]. And there are no subdivisions. Ajuranaaf Baallád: tokkoca, iti-gabae, Buna, Wajer. Wajeriileen durru lafum Borána. Akkás. Waaso Boraní kees hin jíru. Guyya keesa d’ufe. Waason Korre. Rendillaanin, Korre. Ajuran is the same as Balad,2 they met them at Buna and Wajir. Also Wajir belonged to the Boran country. That is how it is. There were no Boran at Waso3 [i.e. during the early colonial period, the Boran were moved to there when they were evacuated by the British from Wajir]. At Waso there were the Samburu. Not Rendille, Samburu. [Question: Gaaf Garriifi Boraní iti ga bae, Garrin ‘nu bojjia’ jed’e moo irreen qában? When the Garre and the Boran met, did the Garre say ‘take us as captives’4 or were they caught by force?] Isiit iti gale. Hín gibirt. Borána finni jed’an. Wan inni d’ufeen: roco gaala, sanga looni, korbes ree, qallutiin d’ufti. They entered by themselves. They became tributaries. Among the Boran it is called finn [observance for the ritual benefit of the performer]. What they brought: a loading camel, an ox of cattle, a billy goat, they brought it to the qallu. Ajurán, ini Gona, qallu Odítuutti d’ufan. Sakuye, abbaan isii qallu, isiin garbici qalluu, garbic hin jéd’anín, duub hín finniti. The Ajuran belong to Gona [a Boran moiety], they bring it to the qallu of Odítu. The Sakuye, their father5 is the qallu [of Karrayyu, belonging to the other moiety, Sabbo], they are vassals (garbic) of the qallu, one does not say vassal, but they perform finn.

 Interview with Waako D’iriba, Marsabit, April 1980.

1

 Further down in this chapter we quote British colonial archival sources which divide the Ajuran in Balada and non-Balada but also refer to this division as ‘entirely artificial’ and to be abolished. 2

 Along the Ewaso Ngiro River, in what now is Isiolo District.

3

 The formula of surrender which, if accepted, can either lead to adoption in an inferior position or to full adoption. 4

 Or ‘owner’.

5

21

Islam & Ethnicity in Northern Kenya & Southern Ethiopia Waani Sakuyeen gurbo gaala hoofti, korbes ree, eleemo, ini qool fuute deemt. Waat, tulleen gosa, damm kinisa, gorjo guute, balgúd guccii, – balgúd adi garte? – kan fuuté, qallu geesit, finn. Aki akán. The Sakuye drive a young camel bull, a billy goat, a ram,6 they take their firesticks and go. The Waat [an endogamous hunter and gatherer caste], they also belong to the tribe, they take bee honey, fill a leather container, and ostrich feathers – have you seen the white feathers? – and take it to the qallu, as finn. That is how it is. [Question: What do they get in return?] Hori qalleefi, qumbi kennaaf. Hawasoolle akas, Hawaso ka Sidam agarte? Hímmuudi. He slaughters livestock for them, gives them incense. Also the Awash [the Oromo-ized population of the Awash valley, east and north-east of Addis Ababa] do that, have you seen the Awash of Ethiopia? They come on pilgrimage.7 Loon fuute d’ufti. Qumbi kennaaf. Hawaso muudi diide. Qalluulle irrá tae. Loon isii yo elman aanan him báán, dííg baan. They took cattle and came. He gave them incense. [One day] the Awash did not want to go on the ritual journey. The qallu did not complain [lit: stayed away]. When they [the Awash] milked their cattle, they did not produce milk but blood. Nadd’een isii hin d’altu. Ijolleen d’alattu ballá d’alatti. Jarsi isii wolti bae. ‘Maant akan nu olce’ jed’. Aada dabne’, jed’, ‘nutt aadá lakkisse’. Their women did not give birth. If children were born, they were born blind. Their old men came together. ‘What has affected us like this?’, they said. ‘We omitted the custom’, they said, ‘We left out the custom’. ‘Kaa, ímmuunna’, jette Hawason ya uf sodaate, muudi deemt. Yo fuute d’ufte qallu ‘ya sii muunne, nutti muud kanke lakkisse nu ho!’ jed’. ‘Let us go, let us make the pilgrimage’, they said. The Awash had become afraid of themselves [i.e. remorsefully feared further negative consequences of their omission] and went on the ritual journey. When they came with their presents, they said to the qallu: ‘We have come to pay reverence to you, we had omitted the journey to you, accept us [or: ‘have us, here we are’, as if giving themselves as presents]. Qumbi taan iti kennu diide, haamp’e badanna akkanum fuute isiilleen hím beeti, ‘qumbin aabon kiyy nu kenne chufumá qumbi’, jette, fuudatte. He [the qallu] did not want to give them the incense [as customary acknowledgement and reciprocation] and gave them gum of badanna8 and they noticed it, [but] said ‘whatever incense my father gives us is just incense’, and took it. Waan isii kaan ka durra hammate hím mid’aadee, loon aanan baani ijolle ilan d’alatte, nadd’een akka aada d’alte. ‘Qumbin aabo kiyy nu kenne chufumá qumbi’, gaaf sun baat. Aki an d’agae sun akkan.  Boru Galgallo (Odoola, Gabra) add jico (plaited milk containers) and bute (water containers) to this list. 6

 Muuda, here translated a ‘pilgrimage’ [ritual journey, payment of tribute, reverence, literally: anointing] or ‘ritual journey’ literally means ‘anointment’. In the course of it the visitors anoint the qallu (cf. Goto 1972: 37f). 7

22

 R: kullum (balanites sp. cf. B. orbicularis Sprague), another thorntree, instead.

8

Pax Borana Their things which before had become bad now improved, the cows gave milk, the children were born with eyesight and the women gave birth properly. [The proverb] ‘All the incense my father gives us is just incense’ sprang from that.9 That is how I heard it.10

I quote this lengthy episode to illustrate three points. First, it describes the relationship between the Boran and their ‘vassals’, as they are called in most of the literature, as reciprocal. The qallu of the Boran gives important ritual services to his clients, he makes things prosper and without his mediation barrenness and anomaly spoil the course of life. Another point of which we are reminded is that the limited regional perspective of this book, i.e. northern Kenya and adjacent areas of southern Ethiopia, should not make us blind to the fact that the ritual influence the Boran had, and in some ways probably still have, in Ethiopia reaches as far or further in the opposite direction and involves even more people. The Awash described here as payers of tribute populate a region that drains north (the Awash dries up in Djibouti) and are separated by longer stretches of rougher territory from the centre of Boran ritual life than the peoples of northern Kenya are. To complete the picture on the Ethiopian side, we would also have to consider the numerous Oromo groups, numbering millions of people, who keep some memory of coming from the Boran area and often regard the Boran as their ritual seniors and the most conservative keepers of custom. Further we would have to examine the Oromo-ized populations partly integrated into these, partly called by separate ethnonyms, some of whom earlier spoke various Cushitic, others Semitic, languages before their linguistic assimilation and many of whom were Muslims.11 We do not know how many of these groups undertook similar journeys to the qallu of the Boran, but we would expect that some of them did. Lastly, we can extract from the above text a list of adoptive affiliations by northern Kenyan peoples to specific Boran clans. Such relationships are called tiriso by the Boran. We can supplement this list from other sources. The Ajuran are counted with the Gona moiety and pay reverence to the qallu of this moiety who belongs to the clan Odítu. This fits in with the saying that the Ajuran are ‘Jilítu without gut’, i.e. without the tress on the back part of the head characteristic of the Boran proper.12 Jilítu, i.e. Worr Jidda, is one of the clans of Gona moiety. The Sakuye are tiriso of the qallu clan Karrayyu of the other moiety, Sabbo, and so are the Garre13 who more specifically belong to the Worr Boku [Bokkic] part of the Karrayyu.14 The Warra (i.e. Worr, Warr) Daya, like some Gabra lineages (Schlee 1989a: 172, 200), had no tiriso because they fitted into the Boran moiety structure by their origin, so that adoptive relationships were superfluous. We have discussed Warra Daya clan stucture elsewhere in connection with the Tana Orma who are remainders of them.15  Lit.: ‘came out at that time’.

9

10

 Waako D’iriba, Marsabit, April 1980.

 Braukämper (1980c).

11

12

 Waako Huqa (Boran).  ibid.

13 14

 ibid.

15

 Schlee (1992b).

23

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

The relationship between the qallu – our information refers to the one of Karrayyu – and his ritual protegés from northern Kenya acquires another element of mutual dependence because of one rule that by its consequences binds the qallu to maintain these interethnic links. The qallu may not drink water which has been transported on a donkey. To quench his thirst, the Gabra phratry Alganna, segments of which originate from Karrayyu, send him eight young camel bulls in every Friday-year16 in which they hold an age-set promotion. One such time, a man of the Gaar phratry insisted on adding one young camel of his own. This generous action proved to be very propitious for him, since henceforth his camels multiplied.17 Also the Ajuran, Garre18 and Sakuye19 are reported to have sent young camel bulls to their respective qallus. The close political and ritual Boran connection of the Ajuran and Garre seemed to have been compatible with a certain degree of adherence to Islam. Islam however, contrary to its present status, must then have been a ‘minority culture’ inside a non-Islamic major system and might have been somewhat watered down. Worri sun chuf gaafas aada isa kara takká jirt. Islaaniille hín jirt. Ammo Ajuraní gaarayú aki nadd’een Boran wol baafate, Garri nadd’een Borana wol baafate, nadd’éénille wolti buusani. Worri sun chuf sheria Borana. The customs of those people [the Worr Libin (Liban), the peoples under the Boran umbrella] were the same [lit.: on the same road]. There were also Muslims. But some Ajuran intermarried with Boran, Garre and Boran intermarried, they mixed the women. All these people followed the Boran law. Aadan isa lam. Ta Borana gul deeman sun, bisaanille ka Borana d’ugan, lafaalle ta Borana gallan. Lubbum tan ufiin bulcan. Their customs were twofold [Islam plus Boran-ness]. They followed the Boran customs because they drank even the water of the Boran and settled in Boran country. They just maintained that livelihood.20

The Boran as the lords of the land kept the others, it seems, in a position similar to guests. In the course of time, however, too many guests invited themselves onto Boran pastures and it may have been lack of foresight of the Boran to allow them to stay. While many generations of Somali were given access to Boran pastures, granted peace and a loose client status, now that most of the area is in Somali hands, a non-Muslim Boran who ventures there without disguise and recognizable by his tress would not even be safe in a bus or at the gate of a police camp.

THE ALLIANCE BREAKS UP All northern Kenyan branches of the former Proto-Rendille-Somali after the return from their migrations found themselves under the umbrella of Boran hegemony, with the exception of the Rendille. The Gabra Miigo were  Robinson (1985) has collected the names of years used by the Gabra. See also in Chapter 2 under ‘The origin of the cycles of seven’. 16

17

 Interview Waato Katelo, Gabra, Alganna, May 1980 near Dukana.  Interview Mu’allim Mukhtar, Garre, June 1980 in Merti.

18

19

24

 Waako D’iriba, Marsabit, April 1980.  Halake Guyyo, Burot, Galbo, Gabra, August 1980.

20

Pax Borana

prevented by the Boran from making the long trek, the Garre and (some) Ajuran have traditions of a migration to the Jubaland and back, the Rendille and the Gabra went to Iris, a legendary place thought to be west of Lake Turkana, and returned from there, and the Sakuye share Garre traditions of a migration to the east and claim, on the other hand, descent from the Rendille. Whatever might be the historical truth behind these tales, we can safely say that there was a period of perturbations which ended with the establishment of pax borana and the acceptance of Boran supremacy and Boran language by Somali and Somaloid peoples. We now address the question how this pax borana crumbled, and examine what is left of it today. If people idealize the past, they also give a turning point from which time onwards things became as bad as they are now. Often the arrival of the colonial rule is given as such a turning point, but not so in our case. Here the trouble started earlier and the British only added to it. The ‘spear’, i.e. war, is said to have been brought by the Kipyai,21 the Laikipiak Maasai, who are well known from the accounts of the Maasai civil war by early European travellers. These Laikipia scattered after their defeat by the other Maasai22, and a large body of their warriors moved north from what is now the Laikipia District, took the whole of Rendille – people and livestock – as their spoil and divided them up among themselves, leading a brief and happy life as parasites. When, however, they wanted to move the whole Rendille society and make them follow themselves elsewhere, the Rendille warriors who had been hiding in the bush chased them away in a bloody battle.23 The Laikipia then moved north, raiding cattle and driving large herds along, until they were beaten by the Worr Libin cavalry near Buna.24 A British compiler25 of the accounts of ‘some old men’ gives 1876 as the probable date of the battle and Korondile as its place.26 Whatever the exact location might have been, it is clear that these Maasai had ventured far into the Boran heartland. The father of our informant Waako D’iriba27 took part in this battle and also Ido Robleh, the Ajuran leader, is reported to have borne ‘the mark of an arrow got near Buna in one fight with them’.28 Apart from the Boran and the Ajuran also the other member peoples of the Worr Libin, the Garre and Gabra29 were involved in these fights and the subsequent pursuit of the intruders and temporary solidarity was extended even to the Rendille who stood outside this alliance.30 The only people who did not join the Boran in these fights were Warra Daya, who had had a bloody conflict with the other Worr Libin in 21

 Boru Galgallo, Odoola, Gabra, September 1979 in Bubisa.  Thomson (1968 [1885]).

22

23

 Common Rendille tradition.

24

 Waako D’iriba, Boran, Marsabit, April 1980.

 Kenya National Archives (KNA) (1917), Moyale Station Report, microfilm Reel No. 43.

25

 Robinson (1985) has collected the names of years used by the Gabra. Here we find Arba Kibiye ‘the Wednesday year of the Laikipiak’for 1879 and Kamis Kibiyi baasan the Thursday year for 1880 when the Laikipiak were chased. See also in Chapter 2 under ‘The origin of the cycles of seven’. 26

 Waako D’iriba, Boran, Marsabit, April 1980.

27

 ‘Notes on the traditional history of Wajir tribes,’ Wajir Political Record Book Vol.II, KNA, compiled 1939 from accounts by Ido Robleh, Dima ‘Abdi, other elders and from other sources. Here, ‘Korei’ (Korre i.e. Maasai and related peoples) are not specified as Laikipia. 28

 No information on Sakuye participation.

29

 Boru Galgallo, Odoola, Gabra.

30

25

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

the preceding decades and had withdrawn to the south31 where the Daarood Somali were to continue their decimation. As far as the Somaloid associate peoples of the Worr Libin are concerned, we can say that in the Laikipiak war they were strongly engaged on the side of the Boran, as they had earlier been in the Boran/Warra Daya conflict. The wars of the nineteenth century thus show that the alliance then was strong and functioning.32 A central figure in strengthening this alliance and later in weakening it was Ido Robleh who must have been an adult by 1876 and died on 12  December 1941.33 Before pursuing the chronology of events, we should pause to reconstruct as much as we can about his personality which must have been remarkable. Although Ido Robleh was the head of the Gelbaris Ajuran (Walemügge), he was not a descendant of that or any other Ajuran clan. His patriline went back to the Haskul clan of the Hawiyye clan family.34 According to eye-witnesses he was a dark-skinned, tall and strong man (Gurracc deera, furda,35 ini jabba36) whose physical strength was paralleled by other characteristics like cunningness (nami garo) and a potent curse (harrabaalle jabbaalle – also his tongue was strong, waan afaan isa abaare, hin káu – what his mouth cursed could not get up). He is also said to have been a sheikh and in possession of a book (Qur’aan): iniille qallu, kitaab qaba.37 In times of unrest he filled the role of war leader (abba duula) himself, and he occasionally also led peaceful expeditions like the ones to pay tribute to the Boran. By 1913, using the British for his aims as they used him, he had become the highest paid Government chief of the Moyale Station, receiving 40 Rupees per month. Yattani Kuno [Jattani Kuni] of the Boran and Qalla (‘Galla’) Rasa of the Gabra got 30 and others 10 or 20 Rupees.38 The explosive load that accumulated within this alliance and at its borders, and finally cracked it, were the ‘Somali’. From early district annual reports it is clear that the term ‘Somali’ without further qualifications in the first decades of the twentieth century did not include the Garre and Kenya Ajuran, although it was known to the authorities that both peoples claimed Somali origin. Politically, however, they were something other than Somali. We find this terminological distinction throughout the early official reports, e.g. ‘But for the Somali the Gurreh [Garre] would live peacefully here, I believe’,39 ‘. . . from complaints made by the Ajuran, Boran, Saquieh [Sakuye] and Gabra living at Wajheir40 it appeared there would be trouble at the place unless the Somali who were robbing them of their wells were ordered to evacuate and turned out  KNA (1939), Wajir Political Record Book Vol.II.

31

 As this is the only point I wanted to illustrate here, I refrain from unfolding the rich and vivid traditions about the Laikipia war which have been collected from all over Northern Kenya. 32

33

 KNA (n.d.): PC/NFD/1/6/4, Moyale District Annual Report.  KNA (1938), Wajir Political Record Book Vol II. (Boran) gives ‘Abdi Ibrahim as ancestral names.

34

 Waako Huqa, Boran.

35 36 37

 Waako D’iriba, Boran, Marsabit, April 1980.

 Waako D’iriba, Boran.

 KNA (1913), Moyale Station Reports. The Indian Rupee was the currency in the then British East Africa until 1920. 38

39

 KNA (1914), Gurreh District Annual Report.

 Later officially but wrongly spelled Wajir, pronounced ‘wajeer’ (read consonants the English way and vowels the Italian way, as in most current transcriptions of non-European languages). 40

26

Pax Borana

of place.’41 When talking about ethnic origins, the reports use the word ‘Somali’ in a wider sense, including Ajuran and Garre: Along the Abyssinian border the inhabitants are mostly Boran who are of Galla race and language. Intermingled with them are to be found many Ajuran, Somalis of the Hawujah [Hawiyya] division, and Gurreh [Garre], another large division of the Somali race. The latter together with some of the Gabra have their headquarters at El Wak. These five tribes, Boran, Sakuyu, Gabra, Ajuran and Gurreh are bound by treaties of friendship and to some extent intermarry and share their villages, even though the Gallas still remain pagans while their Somali neighbours are Muhammadans.42

From this terminological comparison it is clear that ‘Somali’ without further qualification refers to a new wave of arrivals, while the Ajuran and Garre, although Somali as well, are perceived as separate entities. This distinction was not a misinterpretation by the British alone but it is shared by many northern Kenyans today. If a Garre informant full-heartedly laments about the ‘Somali’, it is obvious that he does not include himself in this category; the ‘Somali’ are said to have spoiled the Garre customs and since the arrival of the ‘Somali’ the numbers of soor camels, the Garre strain with white faces, have dwindled. And proudly this informant43 describes the wounds his father received in the fights against these ‘Somali’. While the Oromo population of the Jubaland had suffered the onslaught of these Somali already in the nineteenth century, they made their pressure felt in Wajir and Mandera only in the first decades of the twentieth century. These Somali, in the narrow sense of the word, comprised many branches of the Daarood tribal family and the Degodia who stood in a similar relationship to the Hawiyya tribal family as did the Ajuran, deriving from it by a female link which, the dominant reckoning of descent being patrilineal, would have been forgotten or never claimed unless stressed for political reasons. The hypothesis therefore is justified that both may originally have claimed other identities: separate identities or affiliation to other clusters. This would fit in with the impression that ‘Somali’ for many central and southern groups is a fairly recent label. Both of these migrant groups arrived from the north, the Daarood from the peak of the Horn including what is now the eastern part of Somaliland and the Degodia from the Shabelle River in the relative vicinity of Northern Kenya. These recent north-south migrations in combination with pious traditions of descent from Arab Muslim ancestors have misled generations of scholars to believe that the cradleland of the Somali is on the shores of the Gulf of Aden instead of in southern Ethiopia and its borderlands. While the Daarood, represented mainly by the Muhammad Zubeir Ogadeen who were followed by sheegat of diverse origins, expanded mainly in the southern Juba Tana area, and the Marrehan [Marexan], also Daarood, eventually were confined to the northern Jubaland of what now is Somalia,

41

 KNA (1916), Gurreh District Annual Report.

 KNA (1916/1917), ‘The country between the Juba River and Lake Rudolf’, in the Gurreh District Annual Report. 42

 The reader will understand that I cannot identify informants by their names in cases where this might cause them inconvenience. 43

27

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

the Degodia44 succeeded in establishing themselves in the northern Wajir District and in Mandera District. The Degodia expansion which took a violent form in the early decades of this century seems to have had peaceful pre-history. Various Degodia segments joined the Ajuran as sheegat (in Boran we might say tiriso) and thus became clients to a people who in their turn were clients of the Boran. These Degodia later became known as the earlier representatives of the Reer Liban Degodia, i.e. the northern migrational wave as distinct from the Reer Dakach Degodia who penetrated into Kenya by the southern route, following the Muhammad Zubeir Ogadeen who now live in Garissa District. This distinction is based entirely on the migrational history and has nothing to do with genealogy. We thus distinguish between the Reer Muhammad Liban and the Reer Muhammad Dakach who belong to the same tribal segment. The Gelible is one group which today forms an important part of the Degodia, but is not counted among them at all in the early sources however, but treated as a separate entity or a part of the Ajuran.45 We cannot decide the question of whether they originally were a Degodia group who tried earlier and succeeded better than other Degodia to incorporate themselves in the Ajuran society or whether they were something different and decided only later to become Degodia, at a time when such decisions had become influenced by the British practice of allocating grazing rights on the grounds of the British perception of ‘tribes’. All informants claim that the Gelible always were Degodia, which may have to do with the British experience and the researcher being white. A separate origin, however, appears plausible from the distanced attitude of other Degodia who even abuse them as pagans (kufaar, gaalo). In the district records, the Gelible were in 1913 recognized as a part of the Ajuran. In 1920 they were characterized as a ‘small tribe which is fairly obedient but is always quarreling with the Ajuran’. And in the 1924 report we read: The Gelibleh have always been closely connected with the Ajuran. The Gelibleh are not allowed in the Boran grounds outside the Wajir district, and so call themselves Ajuran when they wish to enter Moyale district.The Abdi

Majid section of the Gelible was recognized as being actually Ajuran.46

In following the records about the Gelible we have jumped the course of events and now return to the beginning of the twentieth century, the wells of Wajir being yet in Boran hands. The turn of the century is described as a time of ‘migrations between British Somaliland, Somalia and Southern Ethiopia’47. The Reer Matan (Degodia) took part in these migrations ‘and attached themselves to the Ajuran who were at this time living in Southern Ethiopia and in part of what is now the N.F.D. under the protection of Anna Buru [Boru], the Kalu [qallu] of the Gonna [Gona] Boran’. These Reer Matan had a sheegat group called Reer Gedid, of Hawadle origin, i.e. not Degodia at all but members of an entirely different branch of the Somali family, at least by the genealogies now accepted among the Somali.  The Daarood are one of the major Somali clan families. The Degodia belong to a completely different branch of the Somali genealogy (cf. Schlee 1989a: 28). 44

 Degodia and Ajuran are also patrilineally unrelated. Their only genealogical connection is that both claim an uterine link to the Hawiyya. 45

 KNA (1924), Wajir Political Record Book, Vol II.

46

28

 KNA (1950), Wajir Political Record Book, ref. Rer Gedid, compiled.

47

Pax Borana On their arrival here the Rer Gedid were, therefore, shegat to the [Degodia] Rer Liban who were shegat to the Ajuran who were virtually shegat to the Boran. This is an excellent example of how Somali migrations are carried out.48

We do not need to enumerate all other Somali groups that succeeded in getting admission into the tiriso network of the Boran and thereby got watering and grazing rights. The position of Ido Robleh in this network seems to have been especially favourable, since some Degodia re-arranged their alliances to render themselves under his personal protection. Ido Robleh ‘had secured permission for his people to use all the Gonna wells in Wajir and Buna’.49 At first the Degodia only watered their stock at the Boran wells by night, so as to cause as little friction as possible. Worri marr gul ya’a, adum ya’um ya’u naman bae. Guyyá eel hin d’úddu, halkan yo Booran irra gálle, wodaamo buufate, halkanum kan d’uddte. Gaaf sun nuu fa Ajuran woliin jir. Wolumaan qubanné. Deemun deemte, laf barraté, guddate d’uub laf chuf qabate. The tribe [Degodia] followed the pasture and for as much as they moved and moved they appeared to the people [Boran]. At daytime they did not drink from the well[s], at night when the Boran had left they watered their stock, they drank only at night. At that time we [the Boran] and the Ajuran were together. We settled among each other. [The Degodia] went on and on, came to know the country, became many and then took the whole country.50

Their link to the Degodia also became a strain to the relationship between the Ajuran and the Boran, especially as the numbers of non-Ajuran sheegat increased and these became more and more self-conscious, so that the Boran eventually had to wait for admission to their own wells and might also not obtain it. The ‘help’ the British awarded the Boran also did a good deal to weaken their position. By 1909 the Somali pressure at Wajir on the Boran, Adjuran and Sakuye, all of whom were regarded as having a traditional right to the grazing supported by the wells, was so great that the Administration was forced to the conclusion that there was little hope of being able to expel the Somalis and to drive them back to the east and the north; nor were the forces available sufficient to protect the Boran and their associates in their villages or to ensure for them the undisturbed use of wells. Zaphiro, the frontier agent,51 accordingly moved all the Boran and Sakuye, and many of the Adjuran, from Wajir, and settled them between Buna and the Dirre country below the Ethiopian escarpment. The scheme was wholly unsuccessful; on the wells, the Ogadeen spread the report that the Boran and their sheegats had left for good; and in the north the Boran found themselves in as much danger from the Ethiopians as they had been from the Ogadeen. Finding the dangers and the uncertainties of the frontier even greater than those of Wajir, the Boran, in 1910, returned to the south – only to find themselves crowded out by the very much augmented colony of Somalis which was then in occupation of the wells. The Adjuran, in spite of having offered their friendship and protection to the Degodia, fared no

 Ibid.

48

 Ibid.

49

 Waako Huka, Boran.

50

 A Greek and the highest ranking representative of the British side in the Anglo-Ethiopian boundary commission. 51

29

Islam & Ethnicity in Northern Kenya & Southern Ethiopia better; and even Ido Robleh complained of being excluded from his traditional grazing areas by these Somali immigrants.52

Even further north, at Buna, to where the British had unsuccessfully tried to evacuate the Wajir Boran, a similar conflict evolved. By 1913 the Buna Boran under Yattani Kuno [Jattani Kuni] were complaining bitterly that the Adjuran had brought the Degodia with them and were driving them from the Buna waters.53

Further east, the Garre were engaged in a constant guerilla war of mutual raiding with Degodia infiltrating from the north-east, and with Marrehan. The 1914 District Report blames chiefly the Degodia and speaks of [. . .] the bitterness which the Gurreh felt towards the Degodia in [on] account of past events & constantly recurring instances of theft at the present time.54

It is also reported that the Garre and Degodia had no common language and that reminds us not to extrapolate the present day linguistic situation (in which Standard (i.e. northern) Somali has found a wide influence by radio) into the past. Earlier in this century the Garre only spoke Boran or Garre Kofar, a language similar to Rahanweyn, or both, but none of these languages is immediately intelligible to speakers of the Degodia dialect which in its turn – at least today – with some good will is mutually intelligible with Standard Somali. On the other hand the newly arriving Degodia obviously did not know any Boran, which now, after decades of contact with mainly Boran speaking peoples like the Ajuran, many of their descendants master quite well. It is no wonder that in this linguistic situation a common Somali national feeling did not exist. In 1916, Serenli, a government post in the Jubaland, which later was to belong to Somalia, was sacked by Aulihan Somali, the garrison killed and the British maxim guns captured. The British though it wise to evacuate also Wajir for a couple of months. [. . .] the sack of Serenli gave the local Somalis an immense feeling of confidence. The Degodia did what they could do to chase the Boran away from Wajir and Buna and the Rer Mohamud (Rer Liban) and the Gelibleh, who at that time had virtually lost their identity as Adjuran shegats, broke away from their hosts and asserted themselves as separate entities.55

In an attempt to re-establish the old order after this and to guarantee access to the wells for the Boran and their allies, Ido Robleh escorted the District Commissioner. The fighting force at this time was estimated as 1,000 men on the side of the Ajuran, Garre and Boran of Wajir, and 1,500 ‘Somali’ opposing them, the largest fraction being the Degodia with 700.56 No lasting changes were effected by these efforts. The position of Ido Robleh on the side of the Boran was becoming more and more difficult and growing numbers of Ajuran sided with the newcomers.  Turnbull (1955: 12f).

52

 KNA (1950), Wajir Political Record Book, ref. Rer Gedid, compiled.

53

 KNA (1914), Gurreh Annual Report.

54 55

30

 KNA (1950), Wajir Political Record Book, ref. Rer Gedid, compiled.  Ibid.

56

Pax Borana

In the end also Ido Robleh broke his loyalty to the Boran. This development was accelerated by new arrivals, this time not sheegat, but Ajuran who came from Somalia, who had gone through a quite different history, who had never had any association to the Boran, and were Muslims in a fuller sense than the Kenyan Ajuran who were often regarded as pagans by Somali immigrants. While Ido Robleh’s people were Walemügge, most of them Gelbaris, the new arrivals were Waghle, another tribal section, but both were Ajuran and regarded each other as tribesmen. Thus there were two kinds of Ajuran. Those who had been tiriso of the Boran for a long time are described by a Degodia like this: Ajuran-ta Boran-ta kujiro, sida gaalada lodilijirey. Usu [i.e. ayaga] iyou Boran malakala-saco. Sain geri aiyey tida’nayeen oo inta ey kuxidhanayeen, luqun-ta. Borantu na wa saas oo malakalakaso. Samin-kas iyeguna gaal ey ahayeen, Gelbaris. Gelbaris wa raidoo Ajurane, markas gaal ey ahayeen. These Ajuran who lived among the Boran were killed like unbelievers. They were not distinguished from the Boran. They plaited tail hair of giraffes and tied it here, around the neck. The Boran did the same, they could not be distinguished. At that time they were pagans, Gelbaris. Gelbaris are one section of Ajuran, they were pagans then. Markas Soomaali imadey, wa dad-ki ladhashay (as imadey) wei islameen. Mugi hore iyeguna dad-ka sooguurey baaba’ aiyey ahayeen. Then the Somali came, they were related people [lit.: born together], they [the Gelbaris] became Muslims. Before they [the Gelbaris] had been people who had migrated to here, they had been lost. Ajuran badh-kis wolibba meeshi ai jiran, Shabelle ai jiran. Badh-kis kol dhow Afgoi ka-imadeen, aniga nin weyn an ahay. One part of the Ajuran are still in their [original] location, at the [river] Shabelle. One part came a short time ago from Afgoi [on the Shabelle, N of Jowhar] when even I was grown-up. Hadda raidoo way, mid-da gaaloudey. Baaba’ saas oo kaley ayeguna gaal soogaleen, oo sheegteen. Sidi Rendille kukufredey ayey ayeguna kukufreyeen. Mar-k’ey walala­dod yimadeen, wayislameen. (Now) they were only one section who became pagans. They were just lost, they joined pagans, they became sheegat [of them]. Like the Rendille became pagans,57 they too became pagans. When their brothers came, they became Muslims [again]. Dad-kii ladhashay ayya imaadeey muslim-ka kusoonoqdeen. Saasa sheekadu na soogarteey. Markii hadda ani dhalanaaye, dad-kii Gelible ayey layeen, ayeko gaala’a, ay dameiyen. Bacadi niman-kii oo kelia ragi lagadameyeey, ‘waxa rabnaa’, yiradeen, ‘[…], diiga marabné. Dhal-ka inno logugadan an degna a-rabna’. Sidas dhal-ka kudegeen. Digooday kudegeen, dhal-ka. People who were their brothers came to them and they became Muslims again. In this form the tradition has reached us. After I was born [i.e. after 1909], they raided a group of Gelible, they [the Ajuran] being unbelievers, massacred them. Afterwards the people [Gelible] of whom men were killed, said, ‘What we want, [. . .] we do not want this blood[-money]. We want to  The extent to which the Rendille have actually ever been Muslims is doubtful. The Reer Diid traditions that Muslim Somali have about the Rendille, according to which the Rendille were original Muslims who were forced to give up their religion, possibly just serve the purpose to deny that the Somali are closely related to a non-Muslim group and by implication might not always have been Muslims themselves. These traditions are discussed at some length in Schlee (1989a). 57

31

Islam & Ethnicity in Northern Kenya & Southern Ethiopia buy the land to settle there’. That is how they settled in the land. They settled in this land with their blood.58

Reading the last episode we have to keep in mind that it was not ‘the’ Ajuran who raided ‘the’ Gelible, but one group of Ajuran, presumably Walemügge, who raided one settlement of Gelible, presumably a group of new-comers of unclear status, because other Gelible had old and stable arrangements with the Boranized Kenya Ajuran which would have precluded such incidents. Later these Walemügge by the influence of other Ajuran, most of them Waghle, who came from the east and were Muslims in name and in practice, were brought ‘back’ into the Muslim Somali fold. We have to write ‘back’ in inverted commas because, apart from their own and their allies’ claims, we have no hard evidence that any or all of these Walemügge had been Muslims in an earlier phase. They might have belonged to the same marginally Islamized PRS culture as the ancestors of the Garre, Gabra, Rendille and Sakuye with whom they share so many traditions. We have seen from archival materials that at the time of the sack of Serenli, 1916, the Worr Libin alliance (which temporarily also included the British) was still intact. Oral traditions59 agree on one other incident which must have happened around the same time and which also foretells the role the British were going to play. A group of Degodia had raided a settlement of the Boran clan Nonítu. Thereupon the Boran and the Ajuran mounted a joint expedition against the Degodia. The fighting force consisted mainly of cavalry with spears. One of our informants, Waako D’iriba, took part in this expediation along with Ido Robleh, and Diido Doyo, a Boran and the father of Galma Diido, later chief of the Waso Boran.60 Among the Ajuran, the names of ‘Uthmaan Ache and Diima ‘Abdi Qotara, the later Government chief, are remembered. The expedition was stopped by the British and persuaded to accept compensation from the Degodia. Waako D’iriba (Boran) thinks that the Boran and their allies would still hold the wells of Wajir if they had not repeatedly followed the peace call of the British. This may be true or not. The more time passed the weaker the Boran position became because of the constant new arrivals of Somali. As the matter was not fought out early, the Boran later had no choice but to follow the British order to evacuate the Wajir area and to move the present day Isiolo district. Although the Boran still had, in 1925, a traditional claim to the use of the Wajir waters, their strength had been so undermined by open Somali attacks, and by the fifth column of their Ajuran shegats, that they scarcely dared approach the wells. And by 1933, victims not only of direct Somali pressure but of the treachery of the very people that had sought protection from them, they no longer pretended to exercise any of their own rights; so that the Administration, recognizing the inevitability of the eclipse of the Boran, had no alternative but to move the remnants of the tribe to the Uaso [i.e. Waso, Ewaso Ngiro, i.e. to the Isiolo District]. (Turnbull 1955: 15f)

 Interview Sheikh Isma’il ‘Abdille, Degodia, Arba Jahan, 1978, language: dialectal Somali.

58

 (Boran) and Waako Huka (Boran).

59

 Diido Doyo was later killed by Somali around Wajir. His people then were evacuated to the Waso, Galma being a young boy. In 1963 Galma, by then a Haji, was killed by Somali shifta together with Dabasso Wabera, a Gabra who was the first post-independence District Commissioner of Isiolo. 60

32

Pax Borana

One factor that contributed to the ‘eclipse’ of the Boran is ignored by Turnbull: the British. If these had fulfilled the security functions they had assumed or if they had admitted that they were unable to do so and given a free hand to the Boran and their allies to fight the matter out themselves at an earlier stage, history might have taken a different course. By 1933, after so many decades of unopposed, ineffectively opposed or just verbally opposed Somali expansion, the Boran decline may indeed have appeared as an ‘eclipse’, i.e. a natural event. Another factor that contributed to this decline – or rather resulted from it and accelerated it – is described correctly by Turnbull: the ‘treachery’ of the Ajuran, Ido Robleh’s switch of alliance. It is difficult to give a point in time for this switch. We have seen that as early as 1913 the Boran had reason to complain about the Ajuran. On the other hand we shall see that there are ritual ties between the Boran and Ajuran even now. The relationship between the two peoples has never formally ended, and in a way can be said to be still existing. Elements of it have even undergone a recent revival (Schlee & Shongolo 2012). Instead of conspiracy and a sudden breach by treason we have to imagine the breaking up of the alliance as a long, slow and painful process of alienation, differentiation of interests and re-orientation of loyalties again and again mended by temporary arrangements. A Boran informant gives the cause of the alienation in the following paragraphs. He seems to distinguish between good and bad Ajuran along the same lines as our Degodia informant but with inverted value judgement. Waghlen d’ufte yennani gos wol tolfate wolti d’ufté, qayyaanen d’ufte duub. [. . .] Ado Boorani gosuma herregatu, Waghlen duub ‘gos tiyy’ jed’e; eel irra jirru isi bisaan obaasét, Boran d’owwe. The Waghle came, so they formed one tribe, came together, so they brought evil gossip. [. . .] While they [the Walemügge] first regarded the Boran as their tribe, the Waghle then said ‘they are my tribe’; they [the Walemügge] let them [the Waghle] use the wells with the higher water tables and refused the Boran [access]. Yennán duuba deebi duub. Durri sheri [seer] hin-jíru; guyya kan Ingresi sher jiddu base nuf worraalle. [Mpakan amm Korondille baate suni?] Hmm, sun. [Gashéni ammá-lle wolumanjirtani, hiyu?] Wolumánjir. So, then they moved apart. Before there was no border; that day the English established the border between us and those people. [Question: the border which now passes at Korondille?61] Yes, that one. [Question: you are staying with Gashe62 even now, aren’t you?] We are together. Gar keen gad hin gódaantu, sher isiin kanum Ajurána suni, isiin ammo olki nam buutuf hin qábtu. Ajurana gosum takka wolin Walmügga. Maqa akasi himacu malle ammó isiin worraan kees hin jírtu. They do not move to our side, their border is that same one, of those Ajuran, but they do not bring quarrel (to people). They [and Gelbaris, the people of Ido Robleh] form the same tribe within Ajuran: Walemügge. Although they call themselves by the same name, they do not bring trouble [lit.: spears are not among them].63

 Today, this boundary separates the Moyale and Marsabit Districts of Eastern Province from the Wajir District of North-eastern Province. 61

 One Ajuran section.

62

 Waako Huka (Boran).

63

33

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

In the Archives64 we find the following fragments of information which might illustrate the development of the Ajuran-Boran relationship and the changes in the composition and wider affiliation of the Kenya Ajuran underlying it. 1924 At heart (however) the Ajuran remain Somali. For many years they had been gaining strength by admitting Degodia to their ranks and consolidating their position with a view to ousting the Boran. They thought the chance [had] come with the migration of [one wave of] Degodia from Abyssinia in 1924. The Ajuran received them and then together they turned on the Boran at Buna and began murdering them. The Abyssinia Boran massed over the border but were prevented from taking action. A baraza [meeting] was held at Buna, when Ido Robleh was reminded of his original agreement with the Boran and debarred from the Buna wells. 1933 The Ajuran [. . .] though of Galla origin [. . .] are now entirely Moslem and are rapidly adopting the Somali language. They can now be considered to have been absorbed by the Degodia.

This quotation does not contain a grain of truth but compensates us with a number of highly interesting misunderstandings. Of course the Ajuran are neither of ‘Galla’ (i.e. ‘Oromo’) origin nor have they ever shown signs of becoming absorbed by the Degodia. But the fact that a British administrator could get this impression shows that the changes undergone by the Ajuran were so dramatic that to superficial observers and the compilers of their views they appeared like a trans-ethnic process. We can conclude that formerly the Ajuran must have been culturally very much like the Boran and by 1933 had become very much like the Degodia. 1934 Ajuran moved into the areas vacated by the Boran who were moved to Isiolo District in 1932. For a number of years the Ajuran have been drifting further from the Boran and closer to the Degodia, and early in the year it was decided to move them over entirely to the Somali side of the Somali-Galla line. This necessitated a slight alternation in the boundary. 1936 The chief problem has been the trespassing of the Ajuran and Gelibleh into Moyale district [i.e. into Boran territory].

We also find an Ajuran census which informs us about the administrative view of the organizational segmentation of the Ajuran. Sections Gelbaris (Headman Dima Abdi) Gashi Gelbaris (Headman Ido Robleh) Garen Waghleh Dulhadda Yibid Ali Total

Number of adult males 189 144 164 122 222 40 26 897 [907]63

 KNA, Wajir District Annual Reports and Wajir Political Record Book Vol.II.

64

 The source (cf. above) gives 897, the correct addition results 907. A copying mistake cannot be excluded. 65

34

Pax Borana 1937 The Ajuran request to be allowed to water at Gubatto was refused, though the Boran had no objection to the Gashe sub-section watering there. The discrimination did not please the Ajuran headman.

We have seen above that such distinctions are made by Boran even today. 1938 [.  .  .] the Ajuran have been and apparently still are, the spearhead of the Somali invasion. 1939 The outstanding problem with regard to the Administration of the Boran & Sakuye is still the ‘Shegat’ system whereby worthless Ajuran & Gurreh [i.e. Garre] and ‘maskin’ [from Arabic: poor] Aulihan etc. are taken into the tribe; they are seldom properly absorbed and tend to form cysts in the body politic which still further advance the rapid disintegration of the Boran. Lists were made of the most flagrant cases but few evictions were effected owing to the regrettable lack of time available for safari. This custom of adopting Somali and semi Somali ‘Shegats’ is to some extent responsible for the deplorable lack of internal authority amongst the Boran. An appreciable part of the tribe is already Islamised and the young converts and the Mohammedan ‘Shegats’ demand settlement of disputes by the half understood ‘Shara’a’ rather than by Boran customary law; there is as a result an undercurrent of discord between the pagan fathers and their sons who have accepted the dogmas but neglected the discipline of Islam.

This quotation from the Isiolo District Annual Reports refers to the Waso Boran. The pastoral Boran of Marsabit District have not embraced Islam to this day. Apart from undisguised value judgements which we do not need to discuss here, this text illustrates one important factor of ethnic politics: by mentioning ‘flagrant cases’ and ‘evictions’ it shows how little the British hesitated to make it subject to their administrative decisions who belonged to which tribe. 1939–1941

In the north, facilitated by the Second World War, the control of the area falling into Italian hands regained by the British, intertribal frictions ‘culminated in what amounted to a Pagan-Islamic war’.66 As the Garre, however, seemed to have a better understanding with the Boran and Sakuye than with the Degodia and Ajuran, there were also Muslims fighting Muslims. 1941

Ido Robleh died on 12.12. and was succeeded by his son Hussein. 1957

In connection with Chief Maalim Aden, grandson of Ido Robleh, the Moyale District Annual Report discusses the problem of unifying the political authority of the Ajuran. ‘The family is of Hawiya, Haskul, descent which is the only reason that the non-Balada sections accept a Gelbaris as Chief.’ 1959 The [new grazing] Order carried a profounder significance, not lost on the Ajuran, in that it abolished the entirely artificial division of the tribe into Balada and non-Balada, grouping them instead as they group themselves, viz: Walemega [Walemügge] and Waghleh.  KNA (1941), Moyale District Annual Report.

66

35

Islam & Ethnicity in Northern Kenya & Southern Ethiopia It is not unduly subtle to say Order LXXXVI brought into the Boranised Gelbaris and Gashe fold, which is least oriented on Somalia, the important, partly Boranised, Garen, to whom the concession was gratifying, coming as it did on top of the appointment, also in 1959, of their principal Sagale Osman Golicha as a paid sub-Chief.

Osman Golicha [Golija] was one of our informants (Schlee 1989a: 94) and Sagale is a gada title, i.e. Boran age-set title! (In the later part of his life the late Osman Golija would have flatly denied that the Boran ever played a hegemonial role in the area.) Apart from this, the discussion of internal divisions of the Ajuran in the 1957 and 1959 reports needs a few more explanations. Balad, if used as a tribal name, is largely used as synonymous with Ajuran, but only by older informants and only in northern Kenya.67 In traditional genealogies Balad (Bal’ad, Balada) figures as an ancestor of Ajuran, the eponymous ancestor of the Ajuran, which would imply that every (descendant of) Ajuran is also a (descendant of) Balad but not every Balad is an Ajuran. Somali genealogies, however, are a part of politics and not of logic. At one time, we have seen, this relationship of inclusion was inversed and a distinction was made between ‘non-Balada’ and ‘Balada’ Ajuran. This distinction was now abandoned as ‘artificial’. As genealogical relationships, real or fictional, were taken as the basis of political divisions, the distinction between Balad and other Ajuran may indeed have been of a different kind: ‘Balad’ being more usual for the Kenyan Ajuran and ‘Ajuran’ for the newcomers. This distinction seems largely to coincide with that between Walemügge and Waghle. The district reports, however, suggest that Garen at one time must have been administratively treated as ‘non-Balada’ although they are genealogically ‘Walemügge’ and according to the report itself ‘partly Boranised’. The alien origin of Ido Robleh’s patrilineage is of special importance since those Ajuran who had been tiriso of the Boran for many generations were seen as pagans or semi-pagans and unfit to lead by the newcomers. The Gelbaris, the segment to which Ido Robleh’s lineage was associated among the Walemügge, are even regarded as unfit for intermarriage by some other Somali who – contrary to the equality preached by the Prophet – are proud of their long genealogical strings of supposedly impeccable Muslims. The story goes that a long time ago a gang of young men went on a raid. Hungry from a long march they found a dead camel which – as it had not been slaughtered – was ritually unclean (bakhti). They ate it nevertheless but some of them agreed among themselves that those who tarried there the longest should henceforth be avoided by the others. Thus the Gelbaris earned their name: Those Who Slept by the Camel. There are similar stories about other groups like the Ribi, a hunter group among the Walemügge, or the Gelible among the Degodia.68 All these stories ascribe a social stigma to a breech of Islamic food avoidances. This may be another way to say that they are the most recently incorporated pagan groups. Informants would, however, reject such interpretations as incompatible with the theory that all Somali descended from the Quraish, the tribe of the Prophet. The history of Ajuran is one of the conflict and coalescence of two diverse elements which represented two varieties of Somali culture, one a Boranized Proto-Rendille-Somali culture and the other modern Islamic Somality. The  The name does not occur in Lewis (1955), where it would certainly have been mentioned if it had any wider currency in Somalia. 67

 For a wider regional perspective on food avoidances and social status see Braukämper (1984); Freeman and Pankhurst (2001). 68

36

Pax Borana

latter gained the upper hand and this was what brought about the decline of the Boran-centered alliance. The cultural re-orientation of the Garre, Sakuye and parts of the Boran themselves towards Islam and a more Somali-like way of life appears to be a consequence rather than a cause of this decline. Even today,69 however, we cannot say that the Worr Libin alliance is quite and only a matter of the past. Gudrun Dahl (1979: 233) describes how much the med’ic, the skin bracelet of a sacrificial animal that conveys the honour of jallab, a gada official, is valued by modern Muslim businessmen and politicians among the Waso Boran. Such med’ic are sent in a limited number at the time of the Boran installations of gada officials to all former allies so that they can use them for the installation of their own officials. Although the allies today decide for themselves who is going to be a jallab, the ritual sanction of the office still derives from the Boran qallu who sends the skin rings. Thus, the Sakuye still derive the political authority of their jallab from the Boran. Also Maalin Adan Ido Robleh of the Ajuran is such a jallab70 and thus, in a way, is still what his grandfather was: a ritual client of the pagan Boran.  I have left this text from my Habilitation thesis, submitted in 1985, unchanged. ‘Today’ refers to the early 1980s. For more recent turns in alliances and a re-emergence of the ‘Worr Libin’ as a political reality, the reader should refer to Schlee and Shongolo (2012). 69

70

 Halakke Guyyo, August 1980, Kalacha.

37

2 Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

The term ‘non-Proto-Rendille-Somali elements’ can have a variety of meanings. In the case of the killer complex, which is shared by all Proto-Rendille-Somali (PRS)-derived ethnic cultures, it means ‘not particularly PRS’, because it has a much wider distribution in the area. The same applies to the subject of the section below, the calendric cycles of seven. These are shared by all ex-PRS groups and distinguish them from some of their neighbours but, of course, they have a much wider distribution. There is no need to tell an English-speaking audience that in the European cultures, among others, we also find weeks which are made up of seven days. Other elements, like those of the Boran political system or the Samburu youth culture, are non-PRS in the sense of not having been part of the early PRS-complex of cultural features. They may be old elsewhere, but among the Cushitic speaking camel pastoralists of northern Kenya there are accretions which are younger than the PRS features they all share. They stem from neighbouring cultures just outside the PRS area and have affected different ex-PRS groups in various degrees.

THE KILLER COMPLEX Among the complexes of cultural elements which predate the advent of Christianity and Islam and which are characteristic of the interwoven cultures of the Horn of Africa is the Verdienstkomplex described by Jensen, i.e. the achievement complex or complex of merit.1 A close association and interdependence of ideas and practices concerning the fertility of fields, livestock and women on the one hand and those concerning the killing of enemies or large beasts on the other hand – the complementarity of giving birth/begetting on the one hand and killing on the other – is at the centre of this complex. We start with the example of a group which culturally and linguistically occupies a very central position among the Cushitic groups of the KenyanEthiopian borderland whose interrelations we are trying to disentangle. The Hor (Arbore) who live north of Chew Bahir (Lake Stephanie) in southern Ethiopia speak a language which is closely related to Rendille by origin2 but has incorporated many Oromo (Boran) loanwords. Interethnic clan relationships point to both the Boran and the Rendille.3 Also the ideas the Hor hold about the  Jensen (1948, 1950, 1960b). Braukämper in his English summary calls it meritorious complex (2001: 209). 1

 Hayward (1984).

2

 Schlee (1989a, 2008a).

3

39

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

link between killing and fertility are typical for the area and resonate with the ideas of their neighbours. In a recent thesis on the Hor (Arbore) who live north of Chew Bahir (Lake Stephanie) in southern Ethiopia, Tadesse Wolde Gossa (1999) has shown that there are propitious and less propitious categories of enemies. The propitious ones are those whose genitals the Hor appropriate for securing fertility. The killer who brings home the trophy might enter into special relationship with a ritual leader, a qawot, who acquires it from him because he cannot fill his function unless he is in possession of genital trophies from an appropriate origin. The qawot is ritually responsible for the order of the Hor world: for the growth of grain and the health and fertility of livestock and people. The Rendille also cut the genitals of slain enemies. There are rituals to honour killers who come home with such trophies – girls heap their beads on them, they are feasted and celebrated, and their mothers’ brothers have to give them a female camel, as Mu‘allim Mukhtar describes it for the Garre in the account below. Meerat, killer’s songs or songs of triumph (the Boran equivalent is gerars) are sung by them, some of which have been rendered and translated in my Rendille monograph (Schlee 1979). After a transition period, the killers proudly wear their bead necklaces and sets of copper wristlets. But this transition period itself, until the ‘throwing away of the genitals’, is a dangerous period and the way it is ended has elements of purification. Also the killer status to which this transition period leads is honoured and desired but it is not propitious in the sense of being conducive to health, wealth, and fertility. Among the Rendille the aggressive and reproductive aspects of manhood are not as closely interwoven as in the ideologies of some other North-east African peoples. As among the Hor, also for the Rendille there are people whom to kill is more or less propitious than to kill others. The category of enemies who are good to kill in the Hor sense, i.e. whose killing conveys a blessing, is however missing among the Rendille. Killing different kinds of enemies corresponds only to different degrees of unpropitiousness. Many stories circulate about killers who have got mentally confused or have nightmares. It is especially dangerous to kill someone of one’s own ‘blood’.4 Because of the widespread interethnic clan relationships among the former PRS peoples, killing a Boran or a Turkana is less risky from this angle than killing a Somali or a Gabra. Among the Boran, killing is closely associated with fertility. The two are linked through the gada system. The Ethiopian monk Bahrey (Guidi 1973 [1907]) reports about the warriors of the sixteenth century ‘Galla’ that they were not allowed to shave their heads before they had killed an enemy. He believed their fierceness to be motivated by the suffering inflicted on them by lice. Later, the killings wore more of a collective than an individual duty, although the individual acquisition of the killer status and the right to wear ivory bracelets (arbore) remained desirable. It was essential that the newly inaugurated uncircumcised warrior age-set, raab, had to go to a war expedition, to take livestock and to kill male enemies. The warrior stage was followed by circumcision and marriage. Killing had to precede the reproductive stages (Cerulli 1933; Haberland 1963; Legesse 1973). In northern Kenya the periodic threat emanating from these once-in-eight-years ritual wars might have led many PRS people to submit to Boran hegemony. No longer being enemies they would then no longer be qualified as a source of trophies (Schlee 1989a: 38, 125f).  For the Rendille notion of ‘blood’ and more on pseudo-biological (ethno-biological) categories which the Rendille apply to their social world, see Schlee (1994a). 4

40

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

Because of its importance within this frame of ideas, killing an enemy is surrounded by a complex of rituals and beliefs which is known as the killer complex. Some young Oromo writers who are not acquainted with anthropological terminology have taken offence at my use of this term (see Schlee and Shongolo 1995: 10 and Baxter, Hultin and Triulzi 1996: 17–19 for a discussion). They appear to have taken the term ‘complex’ in a psychological sense, as if I was speaking of mad killers. The Oromo and Somali do seem to have their share of bloodthirsty madmen, just as the Germans, the English or whoever you name, but I was not speaking of those. Psychology is not my strength; I prefer to deal with the directly observable, interactive aspects, the sociological side so to say, of words, deeds and symbols. I was speaking of the killer complex as a complex of cultural features regarding killers and killings just as Herskovits (1926) was speaking of the cattle complex as a complex of cultural features surrounding cattle.5 These Oromo writers also appear to believe that I wanted to blame the Boran in particular for having the killer complex. A closer look at my writings would soon have revealed the contrary: this complex is shared by the Rendille and many others (Schlee 1979, 1989a: 60). I even describe it as characteristic of the Proto-Rendille-Somali, a wide-spread ancestral and definitely non-Oromo culture of the area (Schlee 1989: 89). To illustrate this point I here render the account of a Garre interlocutor. For the Garre, Mu‘allim Mukhtar6 denies that the castration of bodies was practised. In other aspects, however, the institution he describes is very similar to its Rendille or Gabra equivalent: Gosti tiyya ijjeeftu hamtu. Nami shanille yo duule yokhaan wo ijjeesa yokhaan ufí duu. Namba wan [No.1] abuyya iti kenna. ‘Daada-te [Swahili], anná d’alle, intal tiyya d’alte,’ iti kenna, haisum7. Abba isa ka ejji kan, abeeran ka ejji kan, akhaakhun ka ejji kan, isi malle, isa iti hidda. My tribe are bad killers. Even five people, when they go on a raid, they either kill something or die themselves. First the MoBr (mother’s brother) gives [a camel]. ‘My sister, [i.e.] I gave birth to him, my girl gave birth to him!’ he gives it to him, to his nephew. While his [the killer’s] father is standing [i.e. waiting], his FaBr (father’s brother) is standing, his grandfather is standing, only [the MoBr], he ties it to it [a rope as an insignium to the neck of the camel he is about to hand over to his SiSo (sister’s son).8 Alaa sun eegi nama kennani, aki hala kalasimme taanííni. Jabdu ammó. Hala guyyá ijjesa sun kennan, fuud’isa ta yo ini hori fuud’u hin kénnan. Meerille him báátu. This she-camel, when she has been given to the man, is not like this kalassime camel [a form of borrowed camel,]. But she is strong [difficult, i.e. many rules have to be observed when dealing with her]. A camel cow that has been  Herskovits himself gave rise to some confusion, however, by stressing the psychological implications of these cultural traits, an emphasis which led to misinterpretations by development agents of pastoral economies as being guided by irrational principles (Schlee 1989b). The term also invited student jokes about mad cows long before the term Mad Cow Disease became popular in the media. 5

 Mu‘allim Mukhtar ‘Usmaan, Garre, Tuf, June 1980. Cited also briefly in Schlee (1989a: 60). The following paragaphs were removed from there because of the publisher’s insistence that the book needed to be shortened and streamlined. It is, however, also perfectly clear from the published text that the killer complex was not limited to the peoples of Oromo ancestry. 6

 Cf. the term eysim in Rendille (Schlee 2009: 128–130).

7

 This rope is called heraar in Rendille and hada by the Gabra, just like the rope tied to the neck of a herd sire. 8

41

Islam & Ethnicity in Northern Kenya & Southern Ethiopia given [lit: they give] on the occasion [lit: day] of that killing, is not given away like the stock taken by [given to, like maal,9 kalassime or an outright gift] somebody when he collects stock. She does not go out [i.e. is not paid] as meer [allocation to a bride, mahér]. Halaa sun gaal isa keesat horata malle, hin kénnan. Hín horat. Sooriyo qallate, ini horate, baafate bitate malle; fuud’ intal isá, ta niiti isá, him báásu. Akki d’ibbi imbáásu. Jabdu. He only keeps that she-camel in his camel [herd for breeding], she is not given away. He breeds her. He may sacrifice her for sooriyo, breed her, take her out and sell her;10 for marrying his girl, his wife, he does not pay her [i.e. use the camel for payment]. By no means does he pay her. She is strong. ‘Mu‘allim Mukhtar, aanan nag’ naan jed’ani. Waan sun ammum tanaalle Borana qab. Duub min tokko daqani, min nam garaa jabbaa, ka gesigi nam abbá duulá fakkata, man isá daqaní jemo dawatan duub, jemo d’ira. They tell me: ‘Mu‘allim Mukhtar, pour milk’.11 The Boran even today have such things. So they go to one house, the house of a man with a strong belly,12 a gesigi, a man who is similar to a ‘Father of War’,13 so they go to his house and sing the jemo, the jemo of men. Uf uf uf uf uf. Gurba kaan ka gaaf kaan hala iti kennan, halaaf isaalle wólti faarsan. Onnen qabate, nam kan min kan keesa, ol cee, surri mina ol bae, qabani. Uf uf uf. Uf uf uf uf uf.14 This boy to whom that day they gave the she-camel, they praise him together with the she-camel [in a song]. The state of wildness gets hold of these people in the house, they jump up, go up through the roof of the house, one grabs them [holds the wild people firmly]. Uf uf uf. Guyya sun faarsani, d’awataní. Duub nam chuf ruhin hammate, gesigi tae. Bori dirama duub duulan. That day they praise, sing. Then everybody is sick at heart,15 has become gesigi.16 So, the next morning, they go to war. El Waqi duulaní, kuuno Arsi, Maan jed’ani, Maan Gammad, galaan sadí, afur qabataní, kuun galaan duub Arsi duul daqani. Ji afur, ji sadí, ji shan, akana duub al jir.

 Another form of camel loan, cf Schlee (1989a: 56ff).

9

 An aiti magah of the Rendille or a Gabra sarma would never be sacrificed or sold. Otherwise the camels referred to by these terms are precise equivalents of the camel a MoBr among the Garre had to give his SiSo for killing an enemy. 10

 i.e. into the storage containers, because in a few days time the jemo is to be held at his house. Mu‘allim Mukhtar takes himself as a fictitious example of a host at such an occasion. 11

 Gara jabba (B) and ur miige (R) – ‘strong belly’ mean fearlessness and strength, also a high resistance against curse and social stress. The symbolism in English is similar: ‘He has not got the guts to do this’, ‘he can’t stomach it’, etc. 12

 Abba duula – ‘Father of War’ is the title of the Boran war leader. In Boran society, this is one of the offices provided by the gada (age-set or generation-set) system. 13

 He imitated the shouts or moaning sounds of somebody raving or undergoing a violent state of possession. 14

 Ruhi stems from Sw. roho – ‘soul, spirit, breath, throat’. Mu‘allim Mukhtar literally says that the ‘ruhi has become bad’. 15

42

 Gesigi had been equated above to the qualities of an abba duula.

16

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups They go to war from El Waq, way over there, in Arsi, it is called Maan, Maan Gammad, they cross three, four rivers, there, behind the rivers, they go to raid the Arsi. So they stay out something like four, three or five months. Nadd’een chufta maan? – aanan ind’úgtu,17 irrá chitani. ‘Duuli gale?’‚ Hin gálle.’ Aanan ititu naqan. What about all the women? – They do not drink milk. They abstain from it.18 ‘Has the raiding party returned?’ ‘It has not returned.’ They pour the milk into the sour milk [container]. Nam akanum duulu, akkanum duulu, intal tokko challa d’alle, ilmaan hin d’allín, ganni chuf duula, abbá duula, duuba ka Kuranjo-a, worr Banna Salale, ka att beetu ‘Abdi Mammo Salale, Banasa Duubo Diibe, ‘Abdi maqaan hed’atte, walaale d’agagáme,’ tan dawatani. Intal tokkot ilmaan isa dunya gubbaat hafe. [One] man who went to war in this manner, begot19 only one daughter, he did not beget sons, every year he goes to war, a war leader, he is from Kuranjo,20 the people [lineage] of Banna Salale, of which, as you know, they sing: ‘ ‘Abdi Mammo Salale [. . .] ‘Abdi with the many names, he got them mixed up [. . .]’.21 One girl remained on earth of his children. Nami sun gann isa ka d’ibbi tokko due. Intal sun ammatan ilmaan isiin d’alte, yo atti si agarsisaniille, fuul ilmaan isiin d’alte. Ilmaan guddo d’alte, fuul beeka, ammum tanaalle. Duub worri kun yo sadí afur taeelle yo woliin jir, nami kuunki tokko duul itti d’ufe, akum ree fachaasani kaan fachaas. Ammaalle worri kun hín jir, ilmaan isaalle. Worr akasi hín jir. The man died at an age of a hundred years. That daughter now, the sons she gave birth to, if they had shown you him [the old man], [his] face is [like] the children she gave birth to. She had many children, I know them, even now. Now, those people, and be they only three or four, if they are together, and a thousand men come on a raid, they scare them away like smallstock is scared away. Even now such people and their children exist. Such people exist. 17

 In Boran and Gabra speech this form would be simplified to ind’úddu (g + t > dd).  Literally: cut [themselves] off from it.

18

19

 dalle – ‘gave birth, begot’.  Tuf and Kuranjo are the two major divisions of Garre.

20

 Abdullahi Shongolo has collected a fuller version of this killer’s song (gerars) which Mu‘allim Mukhtar here briefly alludes to, and also a genealogy which goes back to ‘Abdi Mammo Saralle (Salale). The genealogy runs as follows: 21

Hajji Abdi, the interlocutor, brother to a wealthy Moiale businessman, son of Edin, an interpreter during the colonial period, in the 1950s son of Ahmad, son of Kiti, son of Abdi, the person in question, son of Mammo, who in earlier times fought against Arsi and Ejji, son of Saralle. Abdi Mammo Saralle lived around 1845 and used to fight the Arsi. These are the words of the gerars: Abdi Mammo Saralle

Abdi Mammo Saralle

baanata D’uubo Liibe

the hope of D’uubo Liibe [his mother]

Abdi Godaanna baalio Indo Gorbe

Abdi Godaana the glory of Indo Gorbe

Abdi maqaan heddate

Abdi’s names became many

Wallaalan dagagame.

the ignorant mixed them up

43

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

The reason for quoting here the last paragraph is not that it exceeds in realism, but that it gives a vivid image of the Garre ideal of warriorhood. The visible sign of one’s achievement as a warrior is the sarma camel, which is praised together with its owner in songs, which then give a strong incentive for all those who do not yet own a sarma, to go to war: ‘They feel sick and leave next morning.’ An ironic aspect of the killer complex and the ideas associated with it is that modernist Oromo nationalists of the same type as those denying its existence and criticizing the present author for his use of the term make use of ideas associated with the imagery of penises appropriated as trophies in their own political rhetoric. The highly politicized Oromo concept for ‘development’, misooma, has replaced the older term guddina (literally: ‘becoming big’, ‘grow’) in the course of the 1990s. Older meanings of the word stem misa/misu range from ‘fertile, tilled’ to ‘penis’. In the first sense it has an association with the mobilization of communal labour which was needed before the onset of the rains in hoe-cultivation areas. In the latter sense it occurs in gerars, killer’s songs, in which it is used for the trophy taken from the slain enemy and for the manliness of the killer. Ideas of achievement in the modern context do not seem to differ radically from older ideas (Zitelmann 1999: 381–87). The close association between fertility and killing, already noted by Cerulli, still pervades the modern use of the Oromo language.

THE ORIGIN OF THE CYCLES OF SEVEN

44

A conspicuous feature of the groups derived from the PRS is their elaborate calendar based on cycles of seven days and homonymous years. This calendar contrasts with the neighbouring Oromo and Nilotic systems of time-reckoning. This locally makes it a PRS marker, an indicator that in this area the users of this calendar and followers of the rules associated with it derive this part of their culture from the PRS. If we widen the comparison, however, and look for similarities across Africa and the Eurasian continent, we find that ideas associated with this calendar resonate with ideas from far and wide, ideas which may be very old and have had much time to spread around the globe. From this perspective, the cycles of seven may be classified as PRS but not specifically PRS. The oldest written records of mankind stem from a Semitic population in Mesopotamia and date from three thousand years BC. They are written in Sumerian, a Pre-Semitic language which by then had already died out as anyone’s first language and had evolved into a language of writing and erudition, comparable to Latin in the European Middle Ages. The first urban civilization from which we have written records was already fully developed and even old – it had a classical language which had already died out – by the time it left these traces in writing. The Babylonian culture, as we may call by a generalizing name its local variants in a succession of small and large states over the subsequent millennia, had astronomy, the observation of celestial bodies, as the time-giver that made it tick. State organization and worship revolved around periods obtained from the observation of the night sky. Also astronomy is already fully developed in the oldest records we have of it (Winckler 1907). The week of seven days has come down to us from this Babylonian culture. Like its historically less successful competitor, the week of five days, it is not an arbitrary list of names. It is a system. The position of each name in it is determined by a combination of observational data and set rules.

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

The days of the week are equated with Gods and these are equated with the Sun, the Moon and the five planets known to the Babylonians: the seven celestial bodies which (appear to) move across the background of fixed stars. These seven entities can be ordered according to the length of their periodic movement in relation to the Earth, i.e. the time span between their appearances at a given point seen from Earth to the next time they appear in the same spot. Table 2.1   Seven celestial bodies and their periodic movements in relation to the Earth

Planets

Days

Moon

27. 25 days

Mercury

88 days

Venus

25 days

Sun (i.e. Earth!)

365.25 days = one year

Mars

387 days

Jupiter

11 years plus 315 days

Saturn

28 years plus 167 days

If one devotes each hour of the day to one of these seven Gods, as the Babylonians did, and if one orders them according to the length of their circuits relative to the earth, one has the same God in hour 1, hour 8, hour 15, and hour 22. Then one has the next one in the list for hour 23, the one after that for hour 24, and the third one after the God taken as the starting point for hour 25, the first hour of the subsequent day. Each day therefore starts (and is named after) a given God. This relationship between celestial bodies and the order in which the days of the week succeed each other can be visualized by a heptagram in the shape of a seven-pointed star:

Figure 2.1  Heptagram illustrating the days of the Gods (Source: Winckler 1907: 67)

The lines of this heptagram lead us from the Sun to the Moon (from Sunday to Monday), from there to Mars (Tuesday is mardi in French and has been named after the Germanic equivalent of Mars). From there we proceed to Merkur (mercredi, Wednesday, the day of Wotan, the Germanic interpretation of Mercurius) and from there to Jupiter/Tonar, Thor and via Venus/Freya (vendredi, Friday), to Saturnus (Saturday).

45

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

The days of the week and their prognostic functions have played a great role in the reconstruction of the Proto-Rendille-Somali complex. In the analysis of the age-set systems of the Gabra and Rendille it has become clear that much of them is modelled after the seven-day week, by the seven-year cycle linked to it and multiples thereof which form the different sectors of the generation-set cycle (Schlee 1989a: 54–92). In Africa one finds cycles comparable to the week which comprise different numbers of days, but the seven-day week is widespread in Asia and, via Europe and its expansion in the period of industrial modernity, has become a global phenomenon. How can such a widespread cultural element be of any diacritic value? It is of such value in North-east Africa, because it is different from the Boran cycle of ayaana, it also establishes a difference between Rendille/Gabra Malbe type gada systems which are based on cycles of seven years and the Boran system which is based on cycles of eight years, and it helps to distinguish them from certain Nilotic age-set systems which are not based on any rigid counting of years at all but which adjust to demographic pressures or take the flowering of certain plants as a clue that the time has come for new initiations.22 It might therefore be of interest to pause a while and think about where these cycles of seven have come from and how they might have got to this region. We know that the seven-day week is of ancient oriental, namely Sumerian, origin. It has diffused from there to Europe, where its units have been re-named according to the pantheons of the different branches of the Indo-Europeans. It is plausible that – any evidence for an alternative origin being absent – this same system of time reckoning has also diffused into north-eastern Africa and become a cultural possession of the PRS peoples. The question remains: when? In trying to answer this question, we cannot pin down particular centuries, but we can discuss two alternative gross hypotheses: Hypothesis 1: The cycles of seven represent an Arabic-Islamic influence. Hypothesis 2: The cycles of seven have already been part of North-east African cultures before Islam made its influence felt.

Arguments supporting Hypothesis 1 First argument. The seven-day week is not common to Cushitic and not even common to all peoples of Eastern-Lowland Cushitic speech. Even such close neighbours of the PRS as the Boran have a completely different time reckoning system. According to Baxter (personal communication) Boran do, surely, also use the seven day week and also name years and the eight year cycle is seven years plus one but after all, their historical time reckoning is then based on multiples of eight, not of seven. The Boran week – a cycle of 27 named days (ayaana) – is quite different from the seven-day week and not even a multiple of it. If the seven-day week was very old in the area, one would expect it to have been more generally accepted. Second argument. The names of the days of the week are of ultimately Arabo-Islamic derivation. As the Arabic and Rendille names have been given elsewhere (Schlee 1979: 85), I here only list the very similar Gabra forms. Together with these, I list two sets of names, one for boys one for girls, who are born on the respective days.

 Orchardson (1961: 12) explains that Kipsigis initiations could not take place in the second year after the flowering of a forest plant called setitot which flowers once in eight or nine years. 22

46

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups Table 2.2  Gabra forms of the names of days of the week and the respective names for boys and girls born on each day23

Day of the week

Boys’ name

Girls’ name

Alsinin (Monday)

Mammo (Mahmud, Muhammed)

Midín

Talasa (Tuesday)

Isaako (Is’haaq)

Talaso

Arba (Wednesday)

Ali (‘Ali)

Arbe

Kamis (Thursday)

Umuro (‘Umar)

Kamme

Gumat (Friday)

Adan (Adam)

Gumato

Sabdi (Saturday)

Abudo (‘Abd...)

Sabdio

Ahad (Sunday)

Ibrai (Ibrahim)

Aad

It is obvious that the boys’ names correspond to the Arabic names given in brackets and that the girls’ names are derived from the Arabic numerals, which form the names of the days of the week, with the exception of ‘Friday’, the Arabic equivalent of which means ‘the day of assembly’. These names are by no means the only personal names in use among the Gabra. There are others, of Boran origin, which, as it is usual in many African societies, refer to the circumstances of birth, like Diramo for a girl born in the morning or T’ura (Latecomer) or Dullac (The Old One) for boys for whom their mother had to wait for a long time, be it because of a series of miscarriages or other reproductive mishaps, etc. This association of Arabic names for the days of the week with Muslim boys’ names may strengthen the hypothesis of an Arabo-Islamic origin of the time-reckoning system itself, although no such link between personal names and names of the days of the week is known to me from anywhere in the Muslim world outside North-east Africa. The names of the days themselves are often used as names for children (Khamis for someone born on a Thursday, etc.), but the list of equivalences with names of the prophets of Islam or with Muslim holy men is something I have come across only here. I would be grateful for information about any such equivalence found in other countries.

Arguments supporting Hypothesis 2 First argument. One might argue that the cycle of seven days itself is older in PRS culture than the Arabic names given to the days, because there is an alternative set of names and associated myths. Tablino (1980: 79f) gives an almost full set of these names which now have widely fallen into disuse. The only gap in the list is Thursday.24

 Table includes some material from Tablino (1980). In his laudable efforts to analyse Gabra personal names Tablino fails to point out that this set of seven boys’ names is of Arabo-Islamic derivation. Instead he maintains that ‘l’esame morfologico-semantico delle singole forme resta da compiere’ – the morphological examination of individual semantic forms remains to be done. (Tablino 1980–81: 77). 23

 I have departed from Tablino’s transcription in order to render my own ideas about vowel length. 24

47

Islam & Ethnicity in Northern Kenya & Southern Ethiopia Table 2.3  Gabra set of names of days of the week and their associations

Day of the week

Association

Monday

ayaan ree

‘the day of the small stock’

Tuesday

ayaan worabesa

‘the day of the hyena’25

Wednesday

ayaan arba

‘the day of the elephant’

Friday

ayaan dabela ayaan dikira

‘the day of the dabela elders’ or ‘the day of the dikir chants’

Saturday

ayaan loon

‘the day of the cattle’

Sunday

ayaan gaala

‘the day of the camel’

26

While the association of Friday with ritual activities reminds one of Islam, the remainder of this list is not only un-Islamic, but may even sound slightly heterodox and blasphemous to Muslim ears. The Rendille also refer to Sunday as the seeri gaal, the ‘day of the camel’, Saturday as seeri lolyo, the ‘day of the cattle’, and Monday as seeri adi, the ‘day of smallstock’. If asked why, they respond that these species of domestic animals have first been seen on the eves of these days and spent the first night in the settlement on the respective days. There is a strange association between Tuesday and the ability to fight, boys born on a Tuesday are believed to be tough and to have the stamina to emerge victorious from any fights they get involved in. In European mythology, as we know, Tuesday, and its Latin equivalent, the day of Mars, like its derivations in the Romance languages, is the day of the God of War. This may be a strange co-incidence or due to an old far-distance diffusion of an idea – who knows? If the Rendille and the European beliefs about Tuesday derive from a common source, this source must be pre-Islamic. But can the Islamic world with all the pre-Islamic heritage which is handed down within it, be a bridge of diffusion between Babylon and the Rendille? Features attributed to other days of the week have been discussed elsewhere (Schlee 1989a: 55f). These are about what to do or not to do with camels on a given day of the week. They are very specific and so closely tied to the PRS camel complex that we would not expect them elsewhere. But this remains to be seen. As to data of this kind from the Muslim world, we have collections from its two geographical extremes. Westermarck (1926 Vol. I and Vol. II) has encountered innumerable beliefs about the propitious or unpropitious character of certain days of the week in Morocco. We can only cite some of them for illustration: ‘Charms for good purposes may be written with success on a Friday, Monday, or Thursday, those for wicked purposes on a Tuesday, Wednesday, or Saturday, particularly towards the end of the months when the nights are as black as the object aimed at’ (1926 Vol. I: 217). The timing of the harmful charms remind us of the negative associations of the last Wednesday of the month for the descendants of the PRS who believe that boys born on this day will inadvertently cause the death of their seniors, and who avoid a number of activities on such a day.27 But Wednesday as such has by no means exclu Tablino here translates in the plural ‘il gionro delle iene’ for reasons unknown to me

25 26

 Note the homonymity with the corresponding name based on an Arabic numeral.

 See also below, the section on ‘Islam, modern Somality and the borders of PRS culture in the Horn’. 27

48

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

sively negative associations, especially if there is moonlight.28 In the Hiaina, churning butter starts on a Wednesday (ibid.: 245). ‘The Ait Wäryäägher29 consider Wednesday, like Monday, to be good for the beginning of work of any description. At Tangier and among the Iglíwa Wednesday is the best day for the washing of clothes. The latter say that a person who starts on a journey on that day is sure to return’ (Westermarck 1926 Vol. II: 43). This latter belief is in rather direct contradiction to the PRS avoidance to move camp on a Wednesday. On the other hand, people in Tangier believe that sleeping after the ’asar prayers is particularly bad on a Wednesday (Westermarck 1926 Vol. I: 301). Days of the week are associated with certain dead sheikhs, some of whom are of a rather localized importance, others with jinn. Approaching these saints or curing diseases caused by these jinn should therefore be done on the respective days (ibid.: 329, 333f, 498f). Enough examples! The general picture becomes clear: these beliefs are manifold, often localized, and for any vague correspondence to PRS beliefs one can find as many differences. For the other geographical extreme of the Muslim world, Eastern Turkestan/Xinjiang we have data compiled by Ildikó Bellér-Hann.30 Apart from unlucky days of the month, the first of each month being the most critical, we find here beliefs about propitious and unpropitious days of the week, or of good or bad luck attached to specific actions on certain days of the week, like the one that says when a woman combed her hair on a Wednesday her husband would die before long. To comb one’s hair on a Saturday and Sunday involved great risk. Even dying on certain days could have negative consequences. While to pass away on a Friday was considered fortunate, on a Wednesday it was thought to be highly inauspicious, because on this day the gates of Hell stand open. Falling ill on a Saturday and staying in bed on this day was also considered dangerous, and people tried to stay on their feet on this day. (Jarring 1979: 15)

There again, the inauspicious nature of the Wednesday vaguely reminds us of the last Wednesday of the Moon which is regarded as bad to be born on by the descendants of the PRS. But as an indicator of cultural relatedness this superficial similarity does not stand closer scrutiny. Here the combination with a phase of the moon is missing, which seems to be essential to the PRS belief. The beliefs recorded by Bellér-Hann’s sources about propitious and unpropitious days are too numerous to be cited here. They mostly follow the pattern which also prevails among the Rendille and Gabra, namely that days are not good or bad as such but good or bad for specific activities. To attribute a common source to these beliefs one would, however, need agreement of the specific contents of these beliefs. Instead, one finds a multiplicity of beliefs of this kind with a wide margin of variation. It is, however, interesting to note that one of her sources explains the different character of the days by reference to the deities of Graeco-Roman antiquity. This is a clear indication that the spread of Islam has been the vehicle by which pre-Islamic beliefs have been transported to China, since the peoples of Xinjiang, where these beliefs have been recorded, are all Muslims.  In the section below on ‘Islam, modern Somality and the borders of PRS culture in the Horn’ combinations of days of the week with lunar phases are also discussed. In the present context we try to limit the discussion to days of the week as such. 28

 Diacritic for vowel length in the original replaced by duplication, diacritic for ghain on /g/ replaced by /gh/. 29

 These are included in a fuller version in her Habilitation thesis (Bellér-Hann 2006).

30

49

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

A more comprehensive description of the auspicious/inauspicious days of the week is given in the Sa‘atnamä originating from the northern region inhabited by the Taranchi. Since it was published in 1897 – together with a Russian translation – (Pantusov 1897), the original must have been produced before this date at least. It seems to be a more or less codified text with astrological and religious explanations for the nature of the days, and as such it complements the information given by Molla ‘Abdul Qadir from Yarkand, whose text must have reflected knowledge passed down or circulated orally among the population. According to this source, Sunday as the first day of the week was generally considered inauspicious for building, for starting new jobs, to greet great men, to sow seed, and to shave one’s head. The ringing of the right ear at night for a man meant that at night he had to avoid his wife; if cutting his nails meant that the devil would whisper to him, taking a full bath that he would meet trouble and his life will be short. Putting on new clothes on this day meant that he would find peace and good things, but cutting new clothes could mean that he would never be free from sadness. Putting these new clothes on would entail unpleasantness between husband and wife. Moving house could mean that he would suffer losses. Monday was considered good for preparing tools for agricultural work, for building and for meeting great people. It was also considered inauspicious to have one’s hair shaved on this day and to cut one’s nail. Taking a full bath meant that ones property will increase. The ringing of the ear on such a day meant the arrival of a message and the ringing of the ear at night signified the arrival of good news. The cutting of new clothes on this day could entail happiness but for some it could mean mourning; if he puts them on then his property might suffer. Moving house on this day also meant accumulating a lot of property in the future. Tuesday was under the protection of Mars and therefore it was auspicious for preparing for battle. But shaving one’s hair could lead to sorrow and the ringing of ears to loss. It was inauspicious for cutting one’s nails and taking a full bath could cause illness. Cutting a new dress on this day meant burning in fire, but wearing such clothes could entail the fulfillment of the wearer’s wishes. Moving house however meant becoming rich. The ringing of the ears predicted loss, but if this happened at night, it was interpreted to mean that the person would have many children. Wednesday was protected by Mercury, and this day was auspicious for starting things, for getting news, for meeting great people. Cutting nails could bring glory and fame and having one’s head shaved meant relief from sadness and grief. A full bath taken on this day anticipated an abundance of wealth and the cutting of new clothes power. Putting them on could bring the fulfillment of wishes. The ringing of the ears during the day meant taking advantage of thieves, but at night it meant deprivation of friends and/or of bread. Thursday was protected by Jupiter and this day was considered auspicious for contracting marriage, for worship and for having one’s hair shaved. The ringing of the ears during the day could indicate the arrival of a close friend, at night it signified power. Cutting nails however could bring decline and if a person took a full bath he risked becoming a crook. Cutting clothes on this day meant success in science and putting them on could anticipate receiving jewels. Wearing new clothes on this day could bring wealth, while moving into a new house on this day could mean that one’s property would suffer damage.

50

Protected by Venus, Friday was considered auspicious for getting married and going to war. Shaving one’s head was recommended and the cutting of nails could bring high esteem and attention. The ringing of the ears during the daytime indicated sadness, but the same sign at night promised that

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups hard things would become easier. Similarly, an improvement of things was expected if one took a full bath. Similarly, the cutting of a new dress and wearing it were also auspicious. However, moving house on this day was supposed to result in malice emerging between relatives. Protected by Saturn, Saturday was considered particularly auspicious for planting trees and digging irrigation canals and ponds. Having one’s hair shaved off was inauspicious and similarly the cutting of nails could result in illness. A full bath taken on this day would ensure long life and power. Ear ringing on this day indicated receiving a message from far away, but the same thing happening at night of the same day meant that the person’s wife had made friends with a stranger. Moving house on this day meant the arrival of guests, and the cutting and wearing of new clothes entailed that for long the person would not be relieved from unpleasantness. The putting on of new clothes could also bring sorrow.31

Apart from explicitly connecting Sunday and Monday to the Sun and the Moon respectively, this source has adopted the entire Babylonian pantheon of Gods/Planets in its Roman guise. It leaves us to speculate on which way it has diffused to the Turki and how it has managed to pass the filters of Islamic monotheism on its way there. Had we come across these beliefs anywhere near the Horn of Africa we might see them as a possible source of the Rendille association between Tuesday and toughness as a fighter. In conclusion we might say that while the seven-day week as such in its local manifestations seems to stem from a common source and to be of a venerable age, the details of the beliefs associated with it vary enormously, both within the Islamic culture(s) and beyond, and no clear patterns of derivation or historical diffusion emerge. Second argument for Hypothesis 2. The cycle of seven days is linked, as we have seen, to a cycle of homonymous years which, being solar, have nothing to do with the Muslim calendar which has chosen to ignore the solar year and to apply the name ‘year’ (sannatun) to the cycle of twelve lunar months instead. One author derives the solar year and the associated custom of dab shid from Iranian origins and assumes that they were established on African soil during the Persian occupation of Zeila in the fifth century (Hunt 1951: 9; Lewis 1955: 62). The one thing we know for sure about this element of the calendric system is that it is non-Islamic, the other one is that it is old among the Somali: Cerulli (1926: 6) quotes an inscription on a Mogadishu tombstone which allows the conclusion that the 22 August 1365 fell into a Saturday year. If we extrapolate the present-day Rendille solar calendar, which is based on seasons, backwards, we come to the conclusion that the year from the spring rains 1365 to the spring rains 1366 should have been a Sunday year. Possibly the old Somali have counted the year from some other point of the solar cycle, or they just have counted 365 days without adjustments to astronomical or meteorological events, as the Gabra do. In this latter case, the difference between any given day of this cycle and any given date of the Gregorian calendar, which reckons with intercalary days, would grow at a rate of twenty-five days per century. This would result in dates which are 150 days later in the Gregorian year six centuries ago. The third possible explanation of this disagreement would be that somebody, be it the stonecutter or the PRS or Cerulli or myself, has made an error of one year. Anyhow, the system as such seems to have been in use in Mogadishu in AD 1365, and to have found its expression in an inscription side by side on the same tombstone with a date of a lunar hijra year which gave the chronological clue to Cerulli.  From Pantusov’s (1897) translation of the Sa‘atnamä’, cited in Bellér-Hann (2006).

31

51

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

The reader may have noticed that I lean towards Hypothesis 2, the pre-Islamic origin of the cycles of seven, but the evidence supporting it is not conclusive enough to reject definitely the alternative Hypothesis 1, of an Islamic origin. But if pre-Islamic . . . what does that mean? The terminus ante quem would be the arrival of Islam in the area, the terminus post quem the invention of the system by the Sumerians. This leaves a window of thousands of years for the cycles of seven to reach what is now northern Kenya. European researchers have often been struck by the similarities of practices of East African pastoralists and practices described in the Old Testament (e.g. Jensen 1960a). These comprise libations of milk and markings with sacrificial blood and other practices of the PRS which, however, are in no way specific to them. Merker (1910 [1904]) has even derived the Maasai from the lost tribes of Israel.32 Of course, traditions about how peoples from all over the world originate from the lost tribes of Israel, popular or idiosyncratic as they may be, are regarded with scepticism by academic historians, but the similarities are there. Below, in Chapter 3 under the heading ‘Acceptance and Rejection of Christianity and Islam among the Rendille in Northern Kenya’, we shall come back to questions of cultural relatedness, because these certainly have an effect on acceptance and rejection of beliefs and practices. There we shall also see why direct Jewish or Hebrew influences and such influences mediated through Ethiopian Christianity or the Falasha of Ethiopia appear unlikely. The Old Testament is unique as a document of the ancient world in its richness of information and it is certainly the best known such document. But there is no reason to assume that the ancient Jews were alone or that they were the only people of their kind. The practices described in the Old Testament may derive from yet older practices which have not only been inherited by the Jews but also by their neighbours. We do not know how different the peasants, smallstock herders and nomads of South-west Asia in ancient times were from the peasants, smallstock herders and nomads of North-east Africa. These adjacent parts of two continents certainly formed one region where people moved back and forth and there is no reason to assume a high level of difference. The romantics who have looked for the lost tribes of Israel in north-eastern Africa may have missed the point only narrowly. The similarities which we hesitate to attribute to ancient Jews coming to Africa can possibly be attributed to some ancestral cultural forms shared by Jews and other Semites and Cushites.33  Merker was a Jewish officer of the German Schutztruppe in Tanganyika. As a Jew he might have earned himself respect easily as an intellectual or a businessman, but to be a military person was by no means customary for a Jew in Germany and by no means easily accepted by the conservative (i.e. racialist) corps of officers. To find the Maasai, who were reputed to be the most warlike society of East Africa, to be of Jewish origin might have helped him to reconcile his Jewish identity with that of a military officer in the German Kaiserreich. 32

 These linguistic groupings do not allow any differentiation between different Semitic languages as being closer or not so close to Rendille and Boran. The Cushitic (and Chadic, Berber, Omotic, etc.) languages are cognates of the Semitic languages. The Afroasiatic language family consists of all these branches and comprises them all. Arabic and Hebrew are so closely related to each other that it is futile to ask whether one of them is more closely related than the other to any of the other branches of the Afroasiatic family. This needs to be explained in response to a critical reader who remarked ‘why not Hebrew?’ whenever I pointed to an Arabic etymology. In spite of the assumed old age (c. 10,000 years) and the enormous geographical spread of the Afroasiatic family, the similarities between Rendille (and the other Lowland East Cushitic languages like Oromo and Somali) and Semitic languages like Arabic (and by implication also Hebrew) are so considerable, not so much lexically as in the domain of grammar, that when I tried to teach the conjugation of Arabic verbs to my Rendille-speaking children I found it useful to point to the identical consonantal structures of the derivational morphemes. I am sure one could make the same experience when trying to teach Hebrew to Rendille. The ancient relationship manifests itself clearly. Recent Semitic loanwords in Boran and Rendille, however, tend to stem from Arabic and, to a much lesser extent, from Amharic and Geez (Schlee 1994c), and for the Muslim Somali, anyhow, Arabic is the privileged source of loanwords. 33

52

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

NEW CULTURES FOR OLD In Identities on the Move (Schlee 1989a) I have tried to reconstruct the PRS complex of cultural features which is the oldest discernible layer underlying the cultures34 of the Rendille, Gabra, Sakuye, Garre and possibly the Arbore and some Hawiyya and other Somali. In order to do so, I have abstracted from the Boran Oromo culture which most of these groups share to various degrees. Many of the ethnicities under study have been treated as varieties of Oromo, formerly ‘Galla’, by other scholars and not without reason. That earlier book stresses the importance of an old cultural stratum and examines how it split up, diffused and diversified. It studies centrifugal tendencies. If I had concentrated on the culture which is newer in the area (although it may be old elsewhere), namely the Boran culture, I might have written a study of an integrative, centripetal process: a number of groups starting to share a new expansive culture and acquiring a new common identity as cultural and political allies of the Boran. In order to do justice to this aspect of the matter, I want to discuss at least some of the features of the present-day cultures which derive from Boran influences. As one people of PRS origin, the Rendille have undergone a number of influences from another non-PRS group, the Nilotic Samburu. Also in this case, newer cultural units cut across older ones. Finally we shall have to consider various strata of Islamic influence.

Becoming Oromo

To start with the Boran35 influences, we cannot overlook that the most outstanding among these, of course, is the language. The Gabra, Garre, Sakuye, Ajuran and neighbouring groups today speak the Boran dialect of Oromo.36 (The Ajuran and a part of the Garre in addition to this speak different dialects of Somali.) It has been mainly on the ground of language that administrators, anthropologists and historians have always divided the peoples of the lowlands of the Horn into the two broad categories ‘Galla’ and ‘Somali’. Language has barred the view from other sub-systems of culture whose comparison would have led to quite different categorizations. It has become clear from what has been said in this chapter and the preceding one that there is much more to culture and history than just language, and this will become even clearer in the second half of this book. Nevertheless we should not forget that linguistic links are strong ones and that language, apart from being a means of communication, is the major tool of the human mind. The tool of an artisan, although made by hand may make  The concept of culture and more specifically ‘cultures’ in the plural has been criticized for implying discrete and isolated units. From what follows it will become clear that here I am not speaking of discrete and isolated units. 34

 The term ‘Boran’ itself has sometimes acquired a wider meaning due to the assimilatory force of the people originally meant by it. Goto (1972: 1, 4) includes the Gabra, Sakuye and Waat (‘Watta’) under this label. 35

 In the Language and Dialect Atlas of Kenya (Heine and Möhlig 1980: 57), Heine speaks of a Sakuye dialect, spoken by 4,369 persons, a Boran dialect, spoken by 34,086 persons, a Gabra dialect, spoken by just as exactly known a number of people, etc. According to our own observation, Boran and Sakuye, who often grow up and grow old in the same villages and have done so for generations, speak the same dialect of Oromo, namely the Boran dialect, and so do many other groups. The numbers Heine gives stem from the 1969 census and refer to tribal selfclassification, i.e. to genealogical or pseudo-genealogical affiliation, not to linguistic units. In a similar way Heine deals with Somali. 36

53

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

the hand in turn by training the grip and shaping the palm. In a similar way language, the tool of our mind, may have an intricate Humboldtian relationship of mutual stimulation with other aspects of the human mind. Language is something very close to us. More than the political links which vary in strength and are sometimes absent, it may be the community of the language that links Oromo speakers together and makes them nam Orma, Oromo people. The rich oral literature, the treasure of proverbs (Shongolo and Schlee 2007)37 and the sophisticated poetry are inexhaustible sources of intellectual inspiration and pleasure to anyone who is lucky enough to be a speaker of this language. The second sphere in which Oromo – in our case mainly Boran – culture transcends the framework of the original Oromo is political organization. We have already noted above (Chapter 1) that many groups of PRS origin have belonged to a political, ritual and military alliance, the Worr Libin, which was headed by the Boran. One can acquire features of a political organization in two ways: one is by participating in it, the other by copying it. The Alganna phratry (grouping of clans) of the Gabra, the Sakuye and the Ajuran (in certain phases) may be seen as an extension of the Boran polity: their hayyus and jallabas, or age-set spokesmen, are installed either directly by the qallu of one or the other Boran moiety or derive their authority from him by the fur ring of a sacrificial animal which the qallu sends to them as a sign of installation, although the actual choice of a person may be made by others. The other Gabra, who independently install their own age-set officials for themselves, copy the Boran terminology and many aspects of their political institutions. Their hayyus have the boku or sceptre and the whip (licho) as symbols of power in the same way as the Boran abba gada and hayyus. Also their jallabas are installed into office in a similar fashion to those of the Boran. The division into eebiftu (praying) or qallu lineages, and lineages which are not eebiftu and therefore eligible as hayyu, is also shared by the Boran (the worr qallu as opposed to the worr boku) but is probably not derived from the Boran culture but the common original possession of both the PRS and the Boran. The Rendille who show few Boran influences also have this division, and even Muslim Somali claim lineage-specific powers to heal, pray for, curse and miraculously harm or kill others, although such powers are hardly compatible with orthodox Islam. Even among the Rendille, an independent people outside the Worr Libin alliance who have kept their original Somaloid language, the Boran influence – although significantly weaker – should not be underestimated. The Rendille use the Oromo term jaldab,38 but, their political system being quite distinct, with a different meaning. It does not denote a specific office but a senior elder, an ‘old gentleman’ in general. Like the Gabra, the Rendille use certain Boran-type milk vessels along with the Boran names for these. R: madal, B: madala; R: jijo, B: cico are milk containers made of plant fibres which are plaited or better sewn with a needle. This is by no means a marginal feature because of the ritual functions of these vessels in the soorio and almado ceremonies and their association with elderhood.39

37

 For similar work on Rendille proverbs see Schlee and Sahado (2002).

 Probably derived from some earlier stage of Boran since in modern speech the word has been phonetically simplified to jallab. 38

54

39

 Schlee (1979: 102, 113, 115f, 120, 123, 296f, 445).

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

Samburu elements in Rendille culture

Muslim traditions claim that the early Rendille were Muslims and were forced by the Samburu to pierce their earlobes and to abandon Islam.40 While there is no evidence which confirms that Islam ever was a core element of Rendille culture, the second part of this tradition may contain some truth. While other Cushites either make small holes in their ears, or only the right ear like the Gabra, just large enough to pass an earring through it, or do not pierce their ears at all, like the Muslim Somali, many Nilotes, e.g. the Samburu, the Maasai, the Nandi and traditionally also their Bantu neighbours (e.g. Kikuyu), widen earlobes by successively larger inlets until, in elderhood, when these plugs have been replaced by small metal earrings, the earlobes dangle down almost to the shoulders.41 The name for the perforated ivory discs worn as ear plugs by circumcised warriors, il-kaba is Maa and probably borrowed from the Samburu (il-kamba).42 Another feature shared with Nilotes and probably derived from them is the removal of the two lower middle incisor teeth. The two features – pierced and enlarged earlobes and the removal of two and only two lower incisors43 – is shared with the Rendille not only by the Samburu but by the whole Maasai/Nandi cluster.44 Apart from the Rendille, to my knowledge, no Cushites share these features. This strengthens the hypothesis of a Nilotic origin of this custom. Both the piercing of the earlobes and the extraction (more precisely, levering out) of the lower incisors are delayed until puberty in the case of twins and breech born children. The piercing of the earlobes in such cases is done with a ‘bleeding arrow’, made and blessed by a blacksmith, instead of the commonly used acacia thorns. This applies equally to Rendille and Samburu, and possibly others. This is not to say that all manipulations on the body are common to the Rendille and Samburu and borrowed from the latter. Some very conspicuous features like the incisions around the navel which make the skin contract over it and leave a very small hole45 and the three pairs of circular ornamental scars on the belly46 are typical for Rendille men and exclusive to them. Circumcision – the socially most important form of ritual surgery – is a universal practice among the former PRS and must have been an original possession of the Rendille but its particular form is shared with the Samburu. Both groups close the wound in a particular way by pulling the incised skin of the penis shaft over the glans47 while other groups solve the problem in different ways. The Gabra fix the skin with thorns of Acacia tortilis. The Rendille form of circumcision and customs associated with it may be borrowed from the Samburu. The institution as such seems to be of Cushitic (or, on an  Schlee (1989a: 100, 102, 107, 113, 115–22, 222f).

40 41

 For the graphic illustration see Schlee (1979: 165).

 Nowadays some warriors have fashioned certain plastic items into ear plugs which are indistinguishable to the eye from the ivory ones. Uncircumcised boys are only allowed to wear wooden plugs. 42

43

 Other Nilotes may remove more, or only one.

 Cf. Huntingford (1969: 37) about: the Nandi (52), the Kipsigis (62), the Dorobo of the North Tinderet Forest (73), the Elgeyo (125), the Maasai (15, 89), and the Suk [Pokot] (15, 89) who, although they belong here linguistically, are culturally more similar to the Karamojong cluster (except in the matter of circumcision). 44

 Cf. Spencer (1973: 43f).

45

 For a pictorial illustration cf. Schlee (1979: 154).

46

 For a pictorial illustration cf. Schlee (1979: 159f).

47

55

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

earlier time level, Afroasiatic) origin and borrowed by some Nilotes, while later elaborations on this theme may have been borrowed back from the Nilotic Samburu by the Cushitic Rendille. Proceeding outwards from the surface of the body to its cover, the attire and hair-dress, we note that many items belonging to the warrior fashion, the expressions of the culmination phase of male vanity among the Rendille,48 are of Samburu origin and have Samburu names. The name for ‘red ochre’: il-karia (ol-karia, pl. il-karian49) which mixed with fat is liberally applied to hair and shoulders, is of Maa origin, and so are the names for the various types of beads, e.g. in-gorowo, which are worn as a necklace, in-geri50 for the long chain of beads worn across the chest, il-katar51 for beads worn as a bracelet, in kantarre, which are worn around the ankles, im-bageti, for small red beads in general. The long plaited hair, halhal, is referred to by a Rendille word (the Samburu would say il-maasi), but is an original feature of Samburu warriors, which has, according to Spencer (1973: 144) been copied by the Rendille as recently as the 1940s. As the Samburu warriors in a certain period had adopted the coiffure of the Turkana and later dropped it again (Spencer 1973: 152), so the Rendille warriors, no less receptive to foreign influences than the European youth culture, are presently at the height of a Samburu fad. To which extent elements of this fashion will become stable elements of Rendille culture is difficult to say. Samburu youth culture, apart from bodily ornaments, has affected the institution of premarital concubinage. Targén, a ram which is slaughtered by a warrior at the parental home of his girl-friend, is a Samburu institution which had been adopted at the same time as the hair-dress. This custom has replaced labardús, i.e. milk which has been given to the mother of the girl. The word and the object, labardús (Samburu: il-bardús (?)), are likewise of Maa origin. Here we have a case of one Samburu custom replacing another Samburu custom among the Rendille. Apart from the terminology of the youth culture, a number of other words are common to Maa and Rendille. It is not clear in every case who has borrowed these from whom. At an early time, linguistic borrowing has occurred by Nilotes from Cushites, probably Somaloids,52 a process which has resulted in the similarities which were used to justify the name ‘Nilo-Hamites’. Later borrowing from the Samburu is obvious in the case of certain words which retain the Maa prefixes en- and ol- as in- and il- in Rendille: In-geri Il-karengéet (pl. ilkerengeetó)

type of bead ‘town’

A number of Swahili words seem to have reached the Rendille via Samburu, like il-baánga il-dáwa il-kaássi il-túba il-tukáan

(Sw. panga) (Sw. dawa) (Sw. kazi) (Sw. chupa) (Sw. duka ex Arab. dukanun)

‘machete’ ‘medicine’ ‘work’ ‘bottle’ ‘shop’

 In the mid-seventies, when mirrors were still rare among the Rendille, it has often happened that warriors admired themselves for a great length with narcissistic pleasure in the rear-view mirror of my motorcycle, applying minor corrections to their hair-dress or red ochre make-up. 48

 Cf. Mol’s Maa dictionary (Mol 1978).

49

 This is the Rendille pronunciation. Mol gives en-keri, pl. in-kerin as the Maa form.

50

 Mol: ol-kataar, pl. il-kataari -Maa.

51

56

 Cf. Fleming (1964: 90f).

52

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

One key ideological concept, ‘respect’ (towards one’s elders, towards those whom one avoids sexually) is expressed by a Samburu word, in-kanyit. En-aisugi,53 R. eysugi, ‘snuff’, is an important item of the youth culture. It is therefore plausible that the Rendille have borrowed it from the Samburu. The following similarities, some of which may be coincidental, are not so easy to explain. The question whether all of them are cognates and, if so, the direction of borrowing can only be ascertained by closer linguistic comparison, involving also the distribution of these words in other Nilotic languages on the one hand and other Cushitic languages on the other. Table 2.4  Comparison of words in Maa and Rendille with an English translation

Maa

Rendille

English

pa-supen54

subeén

‘young ewe’

ol-turrur

úrur

‘gathering, crowd’

a-ipiru

biir-nán

‘to twirl, to kindle (fire)’

ol-koroi

káraw

‘colobus monkey’

a-sik

sig-nán

‘to rub’

suuji

suúj

‘ugly’

o-sina

sina

‘grief, sorrow’

Somebody more familiar with Maa may notice more such similarities. While I find the above similarities difficult to explain, in the case of Maa and Rendille kinship terminologies it is rather obvious to me that the Rendille have combined two terminologies – a Cushitic one largely shared with the Somali and Boran speakers and the Nilotic Maa terminology – to form a highly complex new terminology with a larger number of terms used to make more semantic distinctions than in either of the two parental systems. In other words: they have combined two smaller terminologies into one larger one. This matter has been examined in detail elsewhere (Schlee 1994b, 2009). Among the bilingual Ariaal, a transitional group between the Rendille and Samburu, much more of the Samburu culture can be found. Some Ariaal can be said to be not only bi-lingual but also bi-cultural. With the numerous similarities to the Samburu, not only of the Ariaal but also of the ‘white’ Rendille, i.e. those who participate in the gaalgulamme ceremony,55 in mind, a perspective like that adopted by Spencer in his 1973 book appears fully justified. Spencer deals with ‘symbiosis’, exchange and integration between Rendille and Samburu. The perspective I have taken in Identities on the Move (Schlee 1989a) is neither better nor worse but simply different: I look in the opposite direction, north and north-east, and therefore discover similarities with other Cushites, with whom Spencer hardly deals at all.  This and the following Maa terms are rendered in the spelling of Mol’s dictionary.

53

 Ehret (1971: 166, 1974: 90) reconstructs *supeni for Proto-Maasai (cf. Vossen 1982: 98). Rottland (1982: 432, 464) reconstructs *supein for Proto-Kalenjin and Proto-Kalenjin-Omotik. This does not mean that the word necessarily is ultimately of Nilotic origin since it is saben-ti (Keene and Spitler 1976: 152) or sabéen-tii (Abraham 1964: 213 in Somali. Ehret lists this word as an Eastern Cushitic loan in Southern Nilotic History (1971: 110). 54

 ‘White Rendille’ is a straight translation of the Rendille term Ren’dillehi dakhan. It is used to distinguish these Rendille ‘proper’ (Spencer 1973) from the Ariaal who participate in Samburu type ilmugit ceremonies for their warrior age-set, rather than joining the Rendille for their big gaalgualmme gathering (Spencer 1973; Schlee 1979). 55

57

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

While according to the set of features discussed in earlier sections of this chapter we may group the Rendille with other cultures deriving from the ancestral PRS stratum, more recent mutual borrowing and the fluency of transition between them let the Rendille/Ariaal/Samburu/Maasai appear as a loose unit. Different criteria thus lead to different classifications.

Islam, modern Somality and the borders of PRS culture in the Horn

Islam has experienced two maxima of extension. The first maximum was in the early sixteenth century, the second one can be witnessed now or, rather, we are still moving towards it. Between these two maxima there was a marked recession of Islam. Ethiopian Christianity in the north and Oromo monotheism (which shares roots with the religions of the Book but is not derived from them) in the south expanded for three centuries at the expense of Islam, until the latter in the early nineteenth century started to regain its present vigour. The first expansion dates back, with some ups and downs, ultimately to the time of the Prophet. The ahaadith tell us that there was a hospitably accepted Muslim community in exile in Ethiopia even before the conquest of Mecca: a longer Islamic tradition than that of the Holy City. In the course of the centuries, Muslim centres developed not only to the north of the Ethiopian highlands but also on their eastern and southern fringes, which here interests us above all. The Christian Empire thus was limited to the northern highlands west of the Rift Valley. Even this heartland was overrun in the early sixteenth century in a jihad which, for the first and only time, combined the Muslim forces of Ethiopia under the leadership of one man, namely the Harari Ahmad Grañ.56 There was a temporary ‘conquest of Abyssinia’ or Futuh al-Habasha as the relevant chronicle is called. The Grañ wars themselves possibly were the indirect cause for the later recession of Islam (and in some areas also of Christianity) because the pagan Oromo penetrated into the regions which had been depopulated by them. At least such a connection is postulated by a number of authors (e.g. Huntingford 1955: 19; Abir 1968; Markakis 1974: 15f; Braukämper 1978: 127). Oromoization did not always mean paganization but led in some places to Islamic syncretism. For example, the old Muslim centre of pilgrimage with the sanctuary of Sheikh Hussein in Annajina in Bale temporarily became the centre of rather unorthodox and heterogeneous practices and even today is in the force fields of both the reinvigorated orthodox Islam and possession cults which have their roots not only in Islamic Sufism but also in the traditional belief in animal spirits (ayaana).57 Old Islamic traditions were re-interpreted and adjusted to new relationships of ethnic dominance. The historical model of the semi-mythical Sheikh Hussein probably was a Somali or Somalized Arab from Merka and lived in the twelfth century, if not earlier (Andrzejewski 1972: 2, 1975; Braukämper 1980c: 1–4). Ravenstein, after notes by Wakefield, who in his turn apparently interviewed travelling Oromo and Somali on the coast between 1865 and c. 1882 about the interior (Ravenstein 1884: 259), has something quite different to say about Sheikh Hussein: he was a Somali from Harrar who left his native city in order to convert the Galla. He translated the Qur’aan into Galla etc. All this would transfer Sheikh Hussein into the sixteenth century or later, because  Grañ is Amharic for ‘left-handed’. His synonymous Somali name is Ahmad Gure (Axmad Guurey). His patronym was A. ibn Ibrahim al-Ghazi. He lived from 1506 to 1543 (Trimingham 1952: 85; Ullendorf 1973: 69; Markakis 1974: 69). The ‘ñ’ stands for the palatalized ‘n’ (/ny/). The Hispanic spelling is preferred here, because alternative spellings (Gragne, Grany) invite mispronunciations. There is no end vowel but a glide. 56

 Cf. about Sheikh Hussein: Andrzejewski (1972, 1975; Braukämper (1980c); Cerulli (1936 Vol. II, part I; 1957 Vol. II: 127–140); Savoia (1932: 69–80). 57

58

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

before that there were no Oromo (Galla) in Bale.58 Braukämper (1980a: 3) quotes even grosser anachronisms. Apart from overlaying and re-shaping the old Islam, Oromoization in many instances superseded it. Still in 1884, when further north59 and further south60 Islam had been on the advance again for decades, Ravenstein (1884: 271) reports that in the interior of the Horn, in what is now the Kenyan/Ethiopian borderland, Muslim Somali have become pagans under Boran influence. This fits in with oral traditions about the Gabra Miigo, certain Ajuran sections and others. Apparently the Qur’aanic revelation in the three and a half centuries between c. 1550 and 1900 did not strike open ears in this area, although, on the other hand, there is no question of religious persecution of Muslims. Rather there seems to have been sublime pressure of adjustment.61 It was not before the twentieth century that Islam made decisive advances also in northern Kenya and southern Ethiopia, i.e. in those parts of the Horn which are most remote from the coast. When Muslim Daarood and Degodia Somali pushed the pagan Boran from the wells of Wajir (c. 1916–1932), the Garre, who were close associates of the Boran and even the Walemügge Ajuran (who had remained Muslims in name only if at all) were alienated from the Boran and brought back into the Somali fold. Finally, in the 1930s, entire Oromo speaking tribes like the Sakuye and Tana Orma converted to Islam. Today even the Gabra Miigo who throughout the colonial period have been referred to as pagans in all reports, regard themselves as Muslims (at least when they hit town), although this does not seem to influence their actual ritual life very much.62 Even the Boran, who were always regarded as the stronghold of Oromo traditionalism, today comprise many Muslims. Already some time ago Baxter (1966) noted this with regard to the southern Boran who live separated from the main ethnic body on the Ewaso Ngiro, the ‘Waso Boran’.63 In the meantime it has become true also for many Boran around Marsabit and Moyale. The border town Moyale which, ethnically, has a strong Boran imprint, now has an Islamic rhythm of life. The only ethnic groups of Cushitic speech in this area among whom Islam has not gained a foothold – at least not among the pastoral segment – are the Rendille, the Gabra Malbe, the Dasanech at the northern end of Lake Turkana and possibly the Arbore on Lake Stephanie (Chew Bahir). Rendille who leave the pastoral sector become nominal Christians often only at school while they are subject to the pressures of the Italian missionaries (who are anything but

 Braukämper (1980a: 6) did not notice that the relevant passage in Paulitschke (1896 Vol. II: 71) which he quotes sentence by sentence is a translation of Ravenstein (1884: 264). The fault is Paulitschke’s because he cites Ravenstein in the previous paragraph, but not in this one. 58

 Trimingham (1952: 114ff).

59

 I am thinking here of the south-west migration of Muslim Daarood Somali (Turnbull 1955) at the expense of the Warr Day Oromo. 60

 Ravenstein (1884: 271) refers to these Somali as immigrants, while in fact these paganized groups were old residents of the area who had remained there while others withdrew from the expanding Oromo. There also were Somali immigrants (Pankhurst 1965: 38). These often were ‘Ejji’, i.e. northern Somali like Isaaq and Daarood (my interviews with Waako D’iriba and Doti Qalla Rasa, Marsabit und Bubisa, April 1980). Although these also did not always exceed in their constancy in keeping Muslim practices, most of them nevertheless remained nominal Muslims. 61

 This statement reflects my impressions from 1978–1980. For a more recent ethnography see Fekadu Adugna (2009). 62

 Baxter (personal communication) saw Muslim fundamentalists among the Waso Boran in 1983 who forbade mothers to sing lullabies and urged everybody to wear underpants and be puritanical. ‘They thrive in the camps of destitutes.’ 63

59

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

over-scrupulous in their methods in my view), while later and in other fields of urban activity they tend to become Muslims.64 This pendulous movement of Islam is one of the factors which make the historical interpretation of the cultural picture of the interior of the Horn rather difficult. The first expansion of Islam in many coastal areas laid the foundation of a continuous Islamic tradition. In the interior of the Horn, by contrast, Islam was in a certain period pushed back and gave room to pagan ethnic cultures which were, however, interspersed with Islamic splinters. Braukämper (1980b: passim, 1978: 126f, 1980c: 9) mentions such cultural elements among the Hadiya and Sidama. Further south, in my own research area, the Gabra Miigo and Garre, apart from their PRS roots seem to be heirs of this early Islam who have allowed this heirloom to be temporarily submerged. The Gabra and even the Rendille show, although to a much lesser degree, certain apparently old Arabo-Islamic features. We have already seen that Islam in none of these cases affected the cultural core which is PRS. The calendar in its basic functioning, the entire PRS camel complex and clan exogamy are non-Islamic. In a summary way one can say that PRS culture was marginally Islamized. Marginal here is intended to mean: generally weakly, and more strongly on the eastern margins, i.e. among those who were to evolve into the Garre, Gabra Miigo and the Ajuran, in so far as the last belong here at all.65 Whether or not those Somali who live outside northern Kenya and southern Ethiopia once had PRS-like cultures is not easy to ascertain, because today another cultural complex, the ‘modern Islamic Somality’, is very dominant among them. As the last cultural layer we have to note just this recent Islamic expansion which is often tied to the Somali as its ethnic medium. In overview, we thus have to distinguish: 1.   the early Islamic expansion, 2.   the non-Islamic ethnic cultures based on it (Hadiya, Sidamo, etc.), 3.   those ethnic cultures which are not derived from the early Islamic expansion but are only marginally Islamized (e.g. the PRS culture), 4.   the recent Islamic expansion. None of these four forms exists in a state of purity, not even modern Islam of the most rigid shape. Even the most orthodox Somali from the areas, which have been converted the first, cannot have been Muslims for longer than 1,300 years while they have been Cushites for many thousand years. A common Cushitic substratum, often closely linked to the language, has been preserved, e.g. in formulae of oaths and other expressions of the subconscious or everyday religion even among Muslims. Let us start with Arabo-Islamic elements which I would attribute already to the PRS cultures because they have acquired too much local colouring in the course of time to be attributed to the recent Islamic wave of the past few decades. One of the first differences between the Rendille and the Somali language which comes to one’s mind is that Somali has incorporated a great deal of Arabic loanwords in the technical, religious, moral, political and social spheres. In Rendille these are missing. Rendille thus makes the impression of an ‘original’ Somaloid language, one that has preserved the state before the heavy Arabo-Islamic influence of the last millennium. But this is not entirely true. There are a number of Arabic loanwords and not  See Chapter 3, section ‘Acceptance and rejection of Christianity and Islam among the Rendille in Northern Kenya’. 64

 We shall see below (this section) that the Walemügge half of the Ajuran exhibit more PRS features, the Waqle half fewer (cf. Schlee 1989a: 209–230). 65

60

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

only in those semantic fields where we would expect them. Arabic cajuz means ‘old woman’, Somali hajuus has the same meaning (for c  h more examples can be found). It is used with a great deal of reverence. In Rendille, however, hajuus means ‘slut’: it has undergone a change of meaning by acquiring a negative connotation. Changes of meaning (if the basic meaning remains the same and there is no doubt about the identity of the word) and changes of shape (if corresponding to regular sound shifts) are indicators that a word has had some time to change and develop in a language and has not been borrowed recently. In the above example the fact that the word does not belong to the religious sphere, like many recent loans in the speech of Islamized Rendille, underlines this impression. Somali, moreover, today is not important as a source of Arabic loanwords in Rendille; most recent Arabic loans get into Rendille via Swahili and have a distinctly Swahili touch: siasa – ‘politics’, adabu – ‘punishment’, (ku)dharau – ‘to despise’. I give a few more examples of Arabic loans in Rendille which look old to me, and leave it to the discretion of the reader whether he/she shares this impression. Table 2.5  Examples of Arabic loanwords in Rendille

Rendille alál

Arabic ‘own, full property’

halal

‘lawful, ritually clean’

(also in Gabra and Sakuye usage of Oromo, cf. Schlee (1989a: 54–91) (h)araam ‘mean, greedy’

haramun

‘forbidden, unclean’

huggum

haqqun

‘right, claim, privilege, duty’

sama, yasumu

‘to fast’

‘custom, law’

a-soom-a ‘I fast, spend the day hungry (especially about animals which cannot go to pasture)

(Cf. the names of the months Ramadan and Shawwal which are Soom ‘fasting’ and Soomfur or Furam ‘breaking the fast’, although no Rendille within living memory has fasted or broken the fast at the respective times, except for recent converts. Unlike the case of the names of the days of the week, which are Arabic (cf. above: ‘The origin of the cycles of seven’) there is no older set of names for the months. We therefore have to reckon with the possibility that the months whose names have Arabo-Islamic form or content (Soom, Furam, Harrafa, Daga) have been called by these names already for a long time.) The call to the communal evening prayer (which neither in gestures nor words resemble the Muslim salat) is just ‘hooo’ in Rendille. The cry is repeated twice or four times according to the seniority status of the settlement. Among the Gabra, however, we are struck more often by culturally completely integrated, Gabra-ized customs and concepts which at the second glance remind one of Arabo-Islamic culture. Their call to prayer definitely echoes the intonation of the adhan, although it is not in Arabic. It is not in their own language, Boran, either, but in some Somaloid dialect no longer spoken by the Gabra! Kaaalayo! Kaaalayo!

Kaaalayo! Kaaalayo!

61

Islam & Ethnicity in Northern Kenya & Southern Ethiopia Waaqu bursisa (Waaq us roobo) Waaqu bursisa (Waaq us roobo) Sher Waaqi da Come! Come! [. . .] May God protect us during the night (?) (if said in the morning: may God rain (?)) May God protect our borders / the Law(?)

We have come across this language (Schlee 1989a: 71f in connection with almodo) as the language of the dikir, the ritual chants. It is called Af Daiyo, ‘Daiyo language’, by the Gabra.66 It consists of no longer fully understood ritualized phrases in some Somaloid language and is full of Arabic quotations with Islamic content. The frequent invocation of the Prophet, which is disliked by the more rigorous reformist brands of Sunnism who fear that this may be understood as divinization, reminds one of Sufism. Metaphors like ‘divine light’ etc. do the same. All this is no longer understood by the Gabra. We do not know when all this penetrated into the culture of the Gabra or their PRS ancestors, but it must have been before c. 1550 because these Arabo-Islamic elements are linked with a Somaloid speech which pre-dates Oromoization and because under the Boran hegemony the political climate did not favour the spread of Islamic culture. A camel brand which consists of straight lines on the side of the neck of a camel is called alif by the Gabra, the name of the first letter of the Arabic alphabet. A similar use of this word has been reported among the Somali.67 In comparing North-east African cultures it is difficult not to come back again and again to their calendars. As much has been said above about the PRS calendar and how it differs from other calendars already, I now want to reconsider only some features which are especially suited to distinguish between different waves of the spreading Somaloid and Somali cultures: the propitious and unpropitious days and months. The existence of prognostic calendars may correspond to a functional need: calming the fear of the future, denying the psychological recognition of the truth that the ever present possibility of sudden death and misery is part of the human condition, and to ease decisions which are beyond our empirical insights. The form of prognostic calendar, however, is entirely cultural. No empirical grounds can be given why any prognostic calendar should be better then any other. That is why mutually contradicting systems of such beliefs have been co-existing peacefully for millennia. If one such system had actually worked, it would have replaced all others long ago. This arbitrariness of prognostic calendars make them a wonderful example of conformity with the Formkriterium, criterion of form, which was first applied by Frobenius and first formulated by Graebner (1911). It makes these calendars as suitable for historical reconstruction as language. We speak a language because of our need to communicate and we speak a particular language because of our history. The two cultures we want to compare from this angle are PRS culture and ‘modern Islamic Somality’,68 as it is represented above all by the northern  Afan nama gaaf nami bae Af Daiyo. ‘The language of people when people first emerged was the Daiyo language.’ (Jirima Mollu, Bubisa, February 1980). 66

67

62

 Abraham (1964: 10).  ‘Modern’ because it has been on the advance in the south since the nineteenth century.

68

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

Somali. About these northern Somali we know that they have been Muslims for many centuries. Elements of culture which are linked to a complex of features characteristic of these northern Somali can therefore be said to belong to an Islamic stratum.69 Table 2.6  Comparative chart of names of months in six languages Rendille

Gabra

Garre

Sondéer I

Somdeera I

Rajab Rajab/ Soomdeera I

Wajir Somali Isaaq Rajab

7 Rajab

Sondéer II

Soomdeera II

Sha’abaan/ Sha’baan Soomdeera II

Shab’an

8 Sha’baan

Soom

Soom

Soom

Soom

9 Ramadaan

Soon

Arabic

[fasting:

]

Furám

Furám

Furám

Soonfur

Soonfur

10 Shawwaal

Dibiyál

Didial

Didial

Sidataal

Sidatál

11 Dhu l qi’da(ti)

Harráfa

Arrafa

Arrafa

‘Arrafo

Arrafa

12 Dhu l hijja(ti)

Dága

Yaka

Zaka

Dago

Seko

1 Muharram

Ragarr I

Ragarra I

Safar

Safar

Safar

2 Safar

Ragarr II

Ragarra II

Moulid

Rabii’ al Awal

Mawlid

3 Rabii’ ul awwal

Haytikelee

Faite

Jibor I

Rabii’ al Aakhir

Rajal I

4 Rabii’ uth thaani Rabii’ ul aakhir

Haytibooran I

Jibor I

Haytibooran II Jibor II

Jibor II

Jumaada l Awal

Rajal II

5 Jumaadaa l awlaa

Jibor III

Jumaada l Aakhir

Rajal III

6 Jumaadaa l aakhira(ti)

To avoid confusion by the use of the names of months in four different languages, I number the months of the hijra year from 1 (Muharram) to 12 (Dhu-l Hijja) and apply the same numbers also for the corresponding lunations in the terminologies of the Rendille, Gabra and Somali, even though these do not reckon the year from a particular month but base it on empirical observation of the solar revolution. The list further above thus starts with the seventh and ends with the sixth month. (I decided to leave the two Sondér on top of the list because of their prominent position in the PRS ritual.) Now it is interesting to compare in which months one is not allowed to marry. Among the Isaaq and probably among other northern Somali these are the months 2, 4, 5 and 6. (Safar, Rajaló I, II, III). The Rendille whom we here  In our limited context of time and space we do not need to worry about ultimate origins. An ‘Islamic’ cultural feature can, of course, be pre-Islamic Asian and have moved with Islam into Africa, it can be pre-Islamic African and have been incorporated into an Islamic culture one thousand years ago in Africa, or it can be Islamic in the narrow sense by being linked to the orthodox doctrine. We do not make these distinctions here. We only distinguish northern Somali culture as one complex from PRS culture. 69

63

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

take as an example for a group of PRS origin do not marry in the months 2, 3, 5, 6 and 9. The overlap, i.e. those months in which neither the Isaaq nor the Rendille may marry, consists of the months 2, 5 and 6. For two reasons this overlap may not be interpreted as a similarity and much less as an indicator of relatedness of the two systems: 1.  It is not high enough to exclude chance as its cause, 2.  the unpropitious months in the two cases correspond to different grids of classification, i.e. they are unpropitious for different reasons. The months 4, 5 and 6 have the same name in the Isaaq dialect (Rajal) so that one can generalize: all Rajaló (pl. of Rajal) are unpropitious. The months 2 and 3 in Rendille are both called Ragarr and the months 5 and 6 both Hay ti booran, plural: Hayo hi borbooran – ‘grey months’. Neither of these categories is congruent with any of the Somali categories.

64

A striking difference between the two systems is the value attached to the ninth month, Ramadan. For Muslims this month is governed by strict rules but nevertheless richly blessed, while it appears dangerous and unpropitious to Rendille, especially to the Odoola among them. Also the third month with the birthday of the Prophet (Maulid) for the Rendille belongs to the unpropitious category Ragarr. So much for the purportedly Muslim origin of the Rendille. In the fourth month all Rendille may marry, unless they are the only sons of their parents. The fourth month is called Hay ti kelee, ‘the single month’, because it is framed in by two pairs of two months which share one name each: a danger that goes out from the singular. Being used to a calendar in which months (or the artificial subdivisions of the solar year which they call ‘months’) have individual names, Western readers might be surprised to see the number of cases in which names in these lists are used for two or even three successive months. A wider view across spaces and times reveals, however, that this is by no means unusual. The oldest known names of months in ancient Mesopotamia referred to double months. Periodically, for reasons which will become clear shortly, a thirteenth month was added to the 2x6, to adjust the cycle of lunar months to the solar year (Winckler 1907: 55–68). Much later the Muslims abhorred the idea of having two time-givers, Moon and Sun. This idea did not harmonize with their emphasis on Unity and Monotheism. But their precursors, the ancient Babylonians, did try to make the cycle of six double months and the solar year match, although, of course, the former is short by 10.2 days. This necessitated the insertion of intercalary months once in three years or so and that, in turn, implied that either one name needed to be used for three successive months or a month with a special name was fit in. That means that both the use of names for pairs of months and for uneven numbers of months was acceptable to the ancient Babylonians. Double months were also known to the ancient Romans: they just had a set of six names to name twelve months. When they changed to naming months individually rather than in pairs, they ran out of names after the first six and had to use numerals for the remainder. Of these numerals those from September (‘seventh’) to December (‘tenth’) have survived to be found in our modern European languages (Winckler 1898: 191). The names of July and August, of course, go back to a Roman dictator and the first Roman emperor. The pattern of having double months thus may be very ancient, but a comparison of the above lists of names shows that the distribution of terms

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

for pairs or triplets over the cycle of twelve is not the same in the different language or dialect communities. If they derive from the same old source, they must have had plenty of time to evolve apart from each other. The implications for cultural history which can be derived from the comparison of the names of months are very similar to the ones we have reached above (‘The origin of the cycles of seven’) from the names of the days of the week and the beliefs associated with them. The patterns are old: both the seven-day week and the habit of naming months in pairs date back to Babylonian times. The details differ between the variants compared: ancient and modern Europeans (to the extent that some of these beliefs have survived among them), Muslims, and the heirs of the PRS culture differ in most of the fears and predictions they derive from something happening on a given day of the week. In a similar fashion the ascription of names and ritual meaning to a given month differ from one society to the other. Winckler (1898: 324–350) discusses from which earlier calendar the Muslim cycle of twelve lunar months (the hijra ‘year’) was derived in the reform which took place in the year 10 A.H. (anno hegirae). But this reform was a systematic change from a lunar year, which was periodically adjusted to the solar year, to a purely lunar form of time reckoning which was not concerned with the continuously widening gap to the solar year. Any month or pair of months the name of which was derived from month(s) of the older calendar would thereafter continuously change its position within the solar year. The names would no longer refer to the months to which they would have referred in the old system. Of the Babylonian heritage little more than a certain tendency to name months in pairs has come down to the Rendille/Gabra/Somali name sets for months. The way in which these pairs are distributed over the list of twelve no longer reflects the Babylonian pattern. In the Rendille, Gabra, and Isaaq lists, there are uneven numbers of months between the pairs: something which could have occurred if the pairs had maintained the position which they had in the early state when the whole cycle of twelve was subdivided into six pairs. In the case of the cycles of seven found among the PRS we have come to the tentative conclusion that they have not been derived from Islamic models but from pre-Islamic ones. We come to the same conclusion with even more certainty with regard to the cycle of twelve names for the Rendille and Gabra months: they pre-date Arabic/Islamic influence in basic features. The correspondences are few: Soom and Furam (Soonfur, Soom fur) point to Arabic: soom in form and meaning: it is the ‘fast’ of Ramadan, Furam/Soonfur meaning ‘opening, untying the fast’. Harrafa is another name for Dhu l hijja also in Arabic. The other names have no Arabic equivalents in form or meaning. The distribution of pairs does not follow the Arabic pattern: two of the three pairs do not share positions with the paired months in the Arabic list. In an older solar year of the Arabs, Muharram and Safar shared the name Safar (Winckler 1898: 325), but even that pair has no correspondence in any of the name lists under comparison. Also an earlier application of the name ‘Ramadan’ to a pair of months as has been claimed on a hypothetical basis (Winckler 1898: 326) would not fall in the same place as any of the paired months in the modern calendars. We can conclude that the cycles of twelve months in their present shape cannot be as old as the seven-day week. While the week, the cycle of seven days, has been unrolling for thousands of years in an undisturbed

65

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

sequence, the different cycles of twelve months show signs of renaming and re-ordering. After discussing days and months we will now have a look at the days within a month. One feature that is not distinctive but common to all Somali, all PRS peoples and even the Arabs is the abhorrence of the second Wednesday after the full moon, i.e. the last Wednesday of the month. One would not migrate on such a day nor sacrifice an animal, nor should a son be born. I have already quoted some beliefs about this day in Identities on the Move (Schlee 1989a) and in the section above (The origin of the cycles of seven), and therefore give only one more example, the one of an Ajuran man whose son was born on such a day: The father said: ‘The boy will kill me. I do not want to have him.’ He therefore took the Qur’aan and read the Qur’aan until the boy became foolish. Everybody hated him, but the boy still talked like a reasonable person. ‘My father hates me’, he said. The same day he got lost, like somebody who has fallen into a hole. One does not like these Wednesday people.70

A difference between the two prognostic calendars is, however, that the Isaaq also believe the third of the moon to be unpropitious while the Rendille do not single out this day in any way. The Isaaq believe that a child born on such a day would later lose three marriage partners by early death and that only the fourth would reach a normal age.71 Somali normally know some examples of people to whom this has happened while counter-examples are not quoted. For the problem of how to draw a line (or lines: iso-elements by analogy to isoglosses and isohyets) between those Somali who exhibit a substantial part of the PRS complex and those who are ‘modern’ Somali, these distinctions are very useful. For the Ajuran, who as a whole cannot be subsumed easily under such headings as PRS-derivates or ‘modern Islamic Somali’, the calendar of fate is one of the criteria which allow cultural differentiations inside this heterogeneous ethnic body. The Waqle-Ajuran share the Isaaq beliefs about the third of the month and also regard the fifth as negative. Further they avoid migrating on a Tuesday. This is not so among the other big Ajuran section, the Walemügge. Although they associate the Tuesday with fighting and war (as the Rendille do and the ancient Romans did, after all it is the day of Mars), they do not attach to it an unambiguously negative value nor do they link it with any avoidances. All Ajuran avoid, if possible, migrating on a Friday. This finds its explanation in the Islamic religious practice. No explanation can be found in Islam for the avoidance of moving on a Sunday.72 This PRS avoidance, which is shared by all Ajuran, is linked to the belief that Sunday is the day of the camel. Other rules resulting from this belief have been described above. The Isaaq do not have any such rules or beliefs about Sunday. On the other hand, there are features which are common to all Ajuran and  Muhammad cUmmar Hussein, Gelbaris, interview in Buna, 6 October 1984. Also Mahad cAli Cirresa and Tuulich Abba Kiyyo, interview in Bute, 8 October 1984 with reference to the Walemügge section of Ajuran and Mu‘allim Muhammad, interview 18 May 1984 in Koiya, and others with reference to the Waqle section confirm that this day is regarded as unpropitious. 70

 Interview Woris Hassan Musa, 1 July 1984, Korr.

71

 Some Arabs are, however, said to avoid shaving on a Sunday. Stefan Reichmuth, personal communication. 72

66

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

also shared by the Isaaq, but not by the PRS. They do not marry in the months 4, 5 and 6, for which they, too, have only one name although a different one from the Isaaq.73 This comparison of the days and months regarded as unpropitious reveals that the Ajuran in general are more like the PRS than the Isaaq are, and that inside the Ajuran the Walemügge share more PRS elements than the Waqle. We thus have a continuum with the PRS and the Isaaq at the extremes: PRS

Ajuran Walemügge

Isaaq Waqle

This impression is confirmed by a comparison of the practices of camel keeping and by the migrational history of the Ajuran, as reconstructed from oral sources, both of which are discussed in Schlee (1989a: 209–28). In fact, the Walemügge section has lived for a couple of centuries under the Boran hegemony, an umbrella they shared with peoples of PRS origin with whom they were closely allied (and from whom some Ajuran groups may in fact directly derive), while the Waqle section lived in an Islamic Somali environment in the hinterland of Mogadishu in whose history they played a role (Lewis 1955: 47). We here have two types of evidence, Kulturgeschichte74 and oral traditions, which are both regarded with a great deal of suspicion by proper historians (the former more than the latter) and which here confirm each other, regardless of the raised eyebrows of those who only believe in ink and do not trust their own eyes and ears. I hope to have shown that those who are willing to use cultural evidence for historical reconstruction can find suitable criteria for discerning different cultural strata. To give a fuller picture which would complete our jigsaw puzzle and cover the entire Horn would require the cooperation of more scholars, some of whom might have to overcome some methodological prejudice.75 The cluster of peoples compared here with regard to their prognostic calendars is partly Islamized. We may therefore ask whether all or some of these beliefs are part of an Islamic heritage. The belief that certain lunar dates bring good or bad fortune is widespread in the Muslim world. To illustrate this, we again look at Westermarck’s Ritual and Belief in Morocco. For Morocco, Westermarck has recorded the following beliefs: the Tsuul believe that a woman who ‘goes on a Wednesday night when the moon is full to another person’s house and takes from there some ashes and bran of barley, which she then puts underneath her own milk-jar’, will churn much butter (1926 Vol. I: 249). In Aglu, where a solar time reckoning appears to exist, it is believed that one should not start on a journey on the last Wednesday of a month of the solar year (1926 Vol. II: 43). These examples show that combinations of days of the week with lunar dates are also attributed propitious or unpropitious characteristics outside the Horn of Africa. But if one looks at the actual contents of these beliefs, similarities are either absent or so vague that it is not possible to identify channels of  In Boran: Mili Kara, Mili Jiddu, Mili Eege, in Somali: Mili Hore, Mili Dehe, Mili Dambe, i.e. the First, Middle and Last Mili. Cf. above (Table 2.6 Comparative chart of names of months) the Garre terminology which also has a common name for these three months. 73

74

 Formal history of civilisation.  Cf. Schlee (1985) for a more detailed methodological discussion.

75

67

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

diffusion. Moreover, local variations appear to be so high that Westermarck gives a very detailed local or tribal range of validity for each belief he cites. It is therefore impossible to speak of ‘Islamic’ beliefs in a generalizing sense. As we have seen above (‘The origin of the cycles of seven’), some such beliefs appear to have Islamic sources of inspiration, others not, and none of them is shared heritage all across the Islamic world.

LOCALIZED SUFISM Before the shifta emergency of the 1960s, the Sakuye must have been difficult to distinguish from the camel-keeping Gabra. They had the same pastoral economy with the same camel-oriented rituals derived from the ProtoRendille-Somali culture. As most of their clans had split from the Rendille in or about the seventeenth century, and most of the remainder has Garre roots, this is not surprising. The main difference to the Gabra Malbe was that the Sakuye had become Muslims a couple of decades before, but their adherence to Islam must have been rather nominal. As has been mentioned in the ‘Introduction’, the herds of the Sakuye were machine-gunned by Government forces in the shifta war against Somali separatists, and the surviving animals perished in the ‘keeps’. The Sakuye themselves were dispersed. Many of them had fled to Somalia. That a northern cluster of the Sakuye76 gathered again around Dabel, east of Moyale, is partly owed to a charismatic leader whose parish or jamaa the Sakuye have become. Abba Ganna (‘lord of the spring rains’, ‘father of the year’, ‘father of abundance’) was of the clan Jiriwa of Miigo, but that had nothing to do with his status.77 This remarkable man started his career as a type of hermit, settling apart from the other Sakuye in a place of power, as he put it, but of course unlike a Christian hermit, with his wives. Muslims do not believe in the spiritual benefits of celibacy. Already before the Second World War he settled at Dabel where there was a deep well cut into the rocks by the ancient, semi-mythical Warr Daya (Schlee 1989a: 35, 37f, 42f, 73, 95, 156, 226) and there were the ceremonial grounds, the jila grounds, of the Sakuye. Each of his wives had a mobile house or rather tent of mats (man dasse), but unlike the other Sakuye he moved very little. Even then he thought of agriculture as one of the potentials of that country. Other Sakuye settlements joined him seasonally and on festive occasions. He himself stayed throughout the year. Even during the shifta emergency of the sixties, when all Sakuye fled, he refused to move. Only when his herd of cattle, which was in a satellite camp (for), was taken and the herdsman killed, he moved into the ‘keep’ of Moyale, one of the three guarded camps where the Sakuye and their livestock were brought. After the emergency those Sakuye who had survived the camp were first expelled from the Eastern Province and even moved by force to the North Eastern Province (which is ethnically Somali and religiously Muslim) in the notorious trek of death, but later the remainders of the Sakuye were told to settle at Dabel again. They got some food aid, collected honey, and Abba Ganna told them about agriculture. In the course of the years the Sakuye even acquired some animals again, mostly smallstock. 76

68

 There are also Sakuye at the Waso River.  These paragraphs refer to the period 1984–1991. Abba Ganna died in 1994.

77

Non-Proto-Rendille-Somali Elements of Modern Ethnic Groups

Abba Ganna went on pilgrimage to the sanctuary of Sheikh Hussein in the Bale region of Ethiopia three times.78 The Islamic sect of the followers of Sheikh Hussein forms the wider religious context of the jamaa of Dabel. While Sheikh Hussein is recognized as a saint also by many orthodox Muslims, not all of his followers are recognized as orthodox. He played a role in the ayaana cult in which autochtonous beliefs in animal spirits and spirit possession are mixed with the veneration of Muslim saints. Abba Ganna in a way became a living local proxy for the dead and remote Sheikh Hussein. He attracted visitors of various ethnic origins from far and wide. Most of them expected healing from mental disorders and other diseases for themselves or the sick accompanying them. They came with gifts of coffee beans (bun), miraa (a stimulant, catha edulis), tea, sugar, incense (qumbi) or, if they were wealthy, even camels and cattle. Such gifts were often distributed by Abba Ganna among the poor. In his turn, Abba Ganna received support from his jamaa when he went on pilgrimage to Annajina (which, when he became an old man, he no longer did) or, in form of labour, for the communal projects he initiated. These communal projects comprised the digging of two wells, for which Abba Ganna’s wisdom provided the knowledge of the suitable sites (he also once, on a visit to the Waso area, gave the southern Sakuye a hint where to dig for water), and the surrounding of the vast traditional ritual sites of the Sakuye with raw stone walls. These sites now comprise also guest huts, and they have a new name and a new meaning: gamme awlia, ‘tonsure of the spirits of the saints’. It is into these well guarded enclosures that Abba Ganna, once he no longer went on pilgrimage, retired from time to time for periods of seclusion of some weeks or months. On suitable nights determined by Abba Ganna the inhabitants of Dabel gathered on a free space near Abba Ganna’s well fenced and tidy yard, took off their shoes and sung and danced to the sound of drums. These gatherings were led by some elders of the jamaa and by Abba Ganna’s grown-up sons whose special status was marked by their long loose hair which stands up in a manner fashionably called ‘Afro look’ in the west. The texts of the songs were conventional Muslim phrases of invocation like Laa ilaahu illa-llaah, Muhammad yaa rasuulu-llaah or Assalaam ‘aleykum [meaning:‘aleyka] habiibullaah etc. The rhythm accelerated and some participants entered a state of possession, moaning and throwing themselves onto the ground or walking around through the crowd apparently without noticing their human obstacles. Bodily contact between men and women was no longer avoided and for this reason more prudish orthodox Muslims regarded these meetings as abominations and orgies. Personally, I never saw any body contacts which appeared to me intentional and sexually motivated in such contexts. In the healing sessions drums, songs and possession were used likewise and were attributed therapeutic qualities. Social life in Dabel was intensified by the identity of the residential community and the jamaa of Abba Ganna. In 1984 the Muslim festival of ‘Id ul-Hajj (‘Id ul-Harrafa) was held together in the gamme awliya while in other Muslim communities the sacrificial slaughters take place in every individual household. I met no Sakuye at Dabel who criticized Abba Ganna or doubted his special gifts and authority. Even the government appointed chief said that he would and could not reject a request by Abba Ganna. Conflicts with the official sphere  Cf. above under the heading ‘Islam, modern Somality, and the borders of PRS culture in the Horn’ and Schlee (1990a, 1992d) on the Gabra. 78

69

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

so far did, as far as I know, not occur, be it because Abba Ganna did not meddle much in day-to-day affairs or because of the overwhelming respect he enjoyed. The Sakuye of Dabel and the social and religious changes they have undergone or effected are an example of how much clanship and the egalitarian and balanced authority distribution are linked to the nomadic way of life and the camel culture. Once this complex disappears, new social forms and spiritual contents need to be sought. The Sakuye of Dabel have found both. The following chapter tries to further illustrate this point.

70

3 Modern Trends

THE SAKUYE COMMUNITY OF DABEL AND THEIR POSSESSION CULT In its mainstream varieties, Islam, or Sunni Islam at least, presents itself as a religion which does not strive for any unusual states of consciousness, one eminently practicable for those with no more than average religious gifts. Its practice consists in the fulfilment of five central duties: the most important one is the ritual prayer (salaat), which must be performed five times a day, and is characterized by set sequences of movements and synchronized text elements. In carrying out these prayers, correct performance is all-important. Directly afterwards or at other times, one may also address freely-formulated prayers to God, but this is not actually required of the faithful. Fervour is not essential. Another central duty, reciting the creed (shahaada), is absolved upon entering the Islamic faith, and since it is repeated several times a day in the call to prayer and the ritual prayers, demands no particular effort either. The remaining duties are conditional ones. The pilgrimage, the hajj, need only be made if one has the material means for it, and the poor tax (zakat) is similarly dependent on one’s income. Fasting only takes place during a specific month and only if one is physically fit enough and various other conditions are given. In those places where Islam comes into contact with other belief systems and has to present itself plausibly, like for example on its southern periphery, where I have conducted field research for many years, it is eager to show that it is a religion which is simple to practise. God wants to make it easy for people to come to Him, and has sent a clear, uncomplicated message via His prophet, together with simple-to-follow rules by which to live; the Qur’aan also provides a wealth of evidence in support of this. In what follows, however, I am going to exercise great reserve as regards Qur’aan exegesis, since, as the many strains of Islam demonstrate, vastly differing convictions can be derived from various parts of the Qur’aan. As I cannot lay claim to the status of being a Qur’aanic scholar myself, I am neither able nor entitled to take part in these debates. The views I am going to present here have been shaped by the observation of religious practices. This includes numerous discussions with both simple adherents of the faith and dignitaries, the many Friday sermons I have listened to over the years and the observation of the everyday life among Somali and Oromo-speaking Muslims in north Kenya. In the spirit of the title of Geertz’ (1968) book, Islam Observed, what I am concerned with here is Islam as it is lived, argued, presented and taught, and not in theological constructs that I or others have set up. While for sociologists and anthropologists this is undoubtedly the usual perspective, it nevertheless needs to be made explicit sometimes to avoid misunderstandings. Much of this chapter is based on case histories, on actual incidents

71

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

which show how beliefs inform arguments and practices. No attempts at systematization of these beliefs in the sense of a Sakuye theology are made, but some sociological conclusions will be attempted at the end of this section. When it comes to Islam, there is a good deal that requires no explaining to a Christian public, since these two sister-religions, with Judaism as a third, have a large part of their contents in common. However, there are some important differences in the central dogmas. Unlike in Islam for example, the Christian creation myth leads straight into the Original Sin, in which one is, from the Christian perspective, inescapably embroiled and from which one can only be freed through God’s grace, or sola gratia, as Luther says. Partaking in sacraments and abiding by the Law is perceived differently by divers branches of Christianity in its conditional character for obtaining this grace, but in all of them it is the original sin which needs to be forgiven. In Islam, things are more in one’s own hands: there is no hereditary sin – Muslims are responsible for their own doings and omissions. The counterpart to Original Sin, the sacrificial death of God’s son, is similarly absent. If the central Christian message is boiled down to the basics, the way it comes across to someone hitherto wholly untouched by it – for instance a north Kenyan ‘heathen’ subjected to missionary efforts, then it goes something like this: because of the theft of a fruit, all humankind is punished with a hereditary curse and eternal damnation; however, a second (and assuredly more serious) misdemeanour, namely the murder of God’s only son, leads to an immediate pardon – by none other than the father of the murdered victim, by whom earlier, in reaction to the theft of nothing more than a fruit, humankind had been so cruelly punished.1 The ring of caricature is unintended; far be it from me to want to offend anyone’s religious sensibilities. However, the defamiliarizing effects of this simplified portrayal of the central Christian mystery may serve, I believe, to show how difficult it is to get across to non-Christians. Islam, by comparison, is easy. One has to accept the unity of God, accept the fact that He makes His rules known through His prophets and live by those rules. Although Islamic theology teeters on edge of the same epistemological abysses as its Christian counterpart, and engages with the same complexities, by contrast popular Islam, as represented and practised by the faithful en masse, has simple articles of faith and practicable rules. As a result, many believers feel it is altogether too sober, that religious experience is too tightly controlled, and that the believer is denied a sense of closeness to God. In short, it lacks the element of ecstasy. Good Friday and Easter (sacrificial death and resurrection) are absent in Islam, because the prophet cIsa, that is, Jesus, does not die but (like Elias in the Old Testament) is taken up into Heaven. Similarly, there is no equivalent of Pentecost. While Islam does acknowledge the existence of spirits (ruh and jinn), possession by spirits or by the Spirit, the Holy Ghost, plays no role whatsoever in mainstream Islam. Being filled with rapturous religious emotion, the transports of experiencing God’s proximity, of religious ecstasy, are more or less unknown in the Islamic mainstream. This need for ecstasy is amply catered for by Sufism, the Islam of the ‘dervish’ orders. The sobriety of official Islam would seem to explain the need for Sufism as an ecstatic counter-movement. Nothing that has been said so far is new; it has all been said many times before. In fact, the notion that the Sufic orders supply those very emotions  ‘If someone touched my child, I would not let him get away with it,’ was the response of a Rendille elder to an Italian missionary who tried to explain how the death of the Son of God actually led to salvation rather than the annihilation of mankind. 1

72

Modern Trends

denied in mainstream Islam is stated quite explicitly by Evans-Pritchard (1949) in the introduction to The Sanusi of Cyrenaica. Another widely accepted and not implausible theorem frequently articulated in this regard is that ecstatic religion is somehow connected with deprivation and marginality. One example of this is from Lewis in his book Ecstatic Religion (1978 [1971]). This idea will also be confirmed in the following. Although I am aware that the radical deconstruction of everything our precursors have said up to now is one way to instant fame, unluckily for me my empirical findings in Dabel seem to confirm certain well-established theories, at least in their general outlines. I can only hope that their particular local shapes are of some interest. To summarize the history of Dabel: we recall from the preceding chapter, that in the shifta emergency, immediately after the independence of Kenya in the 1960s, many Sakuye escaped the forced sedenterization in ‘keeps’ and the massacres rife in Kenya by fleeing to Somalia. At the start of the 1970s, scattered remnants of the group settled in Dabel, one of their holy places, which in earlier times had had to be visited regularly in connection with promotion ceremonies in their now defunct age-grade system (naturally also a pre-Islamic institution). Their camels having fallen victim to army machine-guns, they were no longer mobile and would have to settle somewhere – so why not in the holy place? The topography was favourable, and with the help of the holy man, Abba Ganna, points were located where water was dug for and found. The new wells were sufficient for the needs of a small, permanently settled community. With the help of a mission church, initially at least (later they withdrew in frustration, since Muslims are highly resistant to conversion), the Sakuye now tried their hand at farming as a way of sustaining themselves. Being situated in the foothills of the Ethiopian plateau rather than on the hot lowland plain, that is, in a region that received a certain amount of rainfall from time to time, in some years they were lucky. Another important means of securing a living was poaching, an arduous and risky business. In the attempt to smuggle two or three leopard-skins out of the country, Sakuye have been known to fall victim to the sheer vastness of the distance to be covered, and to have died of thirst on the way.2 They also collected honey and incense to sell in the border town of Moyale. In the absence of camels, the old camel-related rituals had lost their point; and the more or less nominal affiliation to Islam, no more than one or two generations old here, apparently no longer met the religious needs of the poverty-stricken and marginalized population. It was replaced by an ecstatic variant, the Sufic Islam of the Husayniyya order together with a cult of possession that had become interwoven with it, and also clearly displayed non-Islamic origins. The following pictures may illustrate the setting of this chapter. I shall also elaborate a bit on techniques of transport and locomotion, not only because they are an important part of everyday life, but also because they form the experiential background for metaphors about spirit possession. Dabel lies in a depression between two elevations. In earlier times, when regular pilgrimages to the tribe’s holy places had to be made as part of agegrade promotion ceremonies, these elevations played an important role in various traditions related to the stations of the pilgrimages. Today3 parts of  Conversation with Diqa Abdul Ali, 5 February 1992, Dabel.

2

 The ethnographic present refers to the period 1980 to1991, during which I carried out longer and shorter periods of field research among the Sakuye, mainly at Dabel. Revisiting the location and the theme would be interesting. Global factors like Mission Churches and reformist Islam were present in the area even then, but they may have changed both their scale and their shape since. 3

73

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Figure 3.1  Dabel from a nearby elevation (Photo © G. Schlee, 1991)

Figure 3.2  Sakuye elder in a nomadic hamlet (Photo © G. Schlee, 1991)

74

Modern Trends

Figure 3.3 Somali hut under construction (Photo © G. Schlee, 1989)

Figure 3.4  Somali hut frame covered with mats (Photo © G. Schlee, 1989)

75

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

these older traditions have mingled with elements of the Islamic hajj, yielding the new practices of the local Husayniyya cult.4 The forms of settlement and the shapes of houses on the other hand have nothing whatsoever to do with Sakuye traditions. Rectangular houses with corrugated iron roofs were introduced here by the government, which housed the district police (AP, Administration Police) in dwellings of this nature, presumably adapting European examples, while round houses with conical grass roofs are derived from the Ethiopian tukul. The walls in both cases are of interwoven branches and twigs, which are subsequently plastered with earth. Through the slow reconstitution of their herds after their destruction, today most Sakuye5 are in the position to pursue a nomadic way of life once more, though only on a seasonal basis, with Dabel as a centre to which they return repeatedly, due to its permanent water supply, and because of the handful of alternative complementary sources of income, such as the precarious, rainfall-dependent cultivation of field crops and small-scale trade. ‘Tethered nomadism’, i.e. nomadism revolving around a fixed centre, is a term which describes this form of mobile livestock husbandry with great accuracy. In the nomadic hamlets found in a radius of only about 30 kilometres from Dabel, the Sakuye have a variant of semi-spherical mat tents corresponding exactly to those of the neighbouring Somali. This construction can be erected and taken down again in less than two hours; the tents are loaded onto packcamels when the group moves on. To transport them on the swaying backs of the camels, the various parts of the house are tied together in such a way that they support each other. The mats, which are used as padding for the animal’s back, also serve as supports. There are virtually no parts that only serve to make transport easier without having other functions as well. For instance, there are no saddles – whether pack saddles or riding saddles. This is undoubtedly because Somali and Sakuye, no matter how many camels they might have, do not ride the animals, but walk behind them. For Arabs, this is most disconcerting: Arabic herdsmen allow themselves to be carried on camel-back when the animals are grazing, they herd them from this vantage-point, they more or less live up there, sometimes sleep there as well. Conversely, the Sakuye were disconcerted by the fact that, in view of the enormous distances that always had to be covered, I had broken in a few camels in order to spare myself sore feet and keep myself fresh for the interviews. To be honest, my attempts at breaking camels in were not always an unqualified success. A riding accident provided me with some insights into local healing practices which will be described further down in this chapter. The additional fact that I occasionally used a whip on my riding camels aroused further concerns among the Sakuye as to whether camels that had been subjected to such a flogging could still be eaten, or if they had become haraam. Draught oxen, which they had seen in Ethiopia and which had been beaten in similar fashion, gave rise to similar qualms. Riding they regard as un-Islamic. I argued that the major prophets of Islam were riders – Jesus on an ass (on Palm Sunday) and Muhammad on camel-back, but with little success. It is always the respective local practices that people regard as Islamic. Reading the Qur’aan does not do much to shake these beliefs, since much is

 The adherents of the cult themselves do not regard these practices as originating from traditional elements, but as coming from the direct divine inspiration of the holy man, Abba Ganna. (Interview Khadr Abba Gama, field diary 1992: 31ff). 4

 I here speak of the northern Sakuye from the Marsabit District. The Waso Sakuye (Isiolo District), with whom there is very little contact, are not being taken into consideration. 5

76

Modern Trends

Figure 3.5  Migrating nomadic hamlet in Mandera District (Photo © G. Schlee, 1990)

Figure 3.6  At the well (Photo © G. Schlee, 1991)

77

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Figure 3.7  Hussein, a Sakuye Chief. Here the chief is intent on tracking a lion. In his clothing, Islamic elements (the kofia), western ones (his T-shirt) and above all, military ones (khaki trousers) are combined. Before becoming Chief, he was a Senior Sergeant with the Administration Police. (Photo © G. Schlee, 1991)

recited and little understood. Translation and explanation are reserved for the higher levels of learning. The Sakuye’s practice of walking and the absence of a culture of riding may also explain why they use different metaphors from us when describing possession. The spirit does not ‘sit on’6 its victim, but slithers into it. The experience is thus less one of possession than of being filled. The Boran, who speak the same language, and have a long tradition of riding – also to war – nevertheless employ the same metaphorical language. Dabel acts as a stake, the kind to which one tethers an animal that would otherwise wander away; the nomads always come back, as if tethered to this central place. One explanation for this must lie in the presence of the deep wells in Dabel. In Dabel too, the Chief is to be found; though a government official, he was chosen from a lineage that traditionally provided office-bearers. In general, money income is very limited. Many youngsters cannot go to school because their families cannot afford the school fees; others cannot because they are obliged by the familial division of labour to herd the animals. Without much purchasing-power, people fall back on the home-made or the affordable. One example is this shirt (figure 3.12) which evidently stems from a consignment of FDJ (East German youth organization) shirts which was dumped here, since it wasn’t needed after the German Democratic Republic was dissolved. Though one may have heard more than enough about the phenomenon of the world market, in its details and ramifications it still has the power to fascinate. The local hierarchy has two peaks. At one of them we find the Chief, at the other one Abba Ganna, ‘the father of the spring rains’, the holy man mentioned earlier as having been able to trace reserves of ground water. The 78

 Possession is a derivate of sedere, ‘to sit’ (Latin).

6

Modern Trends

Figure 3.8  Often, the Chief is found somewhere around the well. With the elders, he arranges the purchase of diesel for the pump, supervises the watering cycles and collects contributions. The whole thing is a little reminiscent of the hydraulic theory of the formation of states. The perceptible degree of centralization, at any rate, is undoubtedly dependent in part on the fact that the Chief controls the water. (Photo © G. Schlee, 1991)

Chief refers to him as ‘the other department’. He is a healer and soothsayer with a clientele that comes from far afield, while in Dabel itself he has gradually been elevated to a position reminiscent to that of a sacred king. Seventy years old and very corpulent, he normally remains hidden from the gaze of his visitors. He is addressed through a curtain and through an intermediary, who repeats everything that is said to him and who also repeats all the holy man’s answers in a louder voice for the benefit of the visitors.7 He lives with his wives (whose number is kept constantly at the permitted four by marrying a young girl whenever another wife dies or has been divorced), as well as some of his numerous offspring, in a sprawling homestead separate from others. On the one side this abuts onto an area where great, round stones have been placed – the place of spirits; on the other side, the settlements of his living assistants are situated. The spirits Abba Ganna communicates with are subsumed under the term awliya (Ar., plural of waali], that is, the Arabic word for ‘holy ones, saints’ – a word that is also found in the Qur’aan, i.e. in a revelatory context that pre-dates the emergence of the Islamic cult of saints and is translated there as the ‘friends’ of God.8 Here, though, the concept is a very broad one: it does not only denote devout Muslims, nor even only human spirits or souls. Abba  Here, public preachers often have a ‘loudspeaker’, someone who repeats everything they say more loudly. However, given Abba Ganna’s seclusion, in addition to providing acoustic clarity, a middleman of this nature also has the function of increasing the distance between the dignitary and the visitor. 7

 Sura X, V, 62 in the translation by A. Abdullah Yusuf Ali (1989 [1981]); V, 63 according to the Marmaduke Pickthall (1984 [1930]) count, and translated as ‘friends’ in both cases. This also conforms to the local interpretation (interview with Khadr Abba Ganna and Sheikh Aliow, 23 January 1992). 8

79

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Figure 3.9  The anthropologist's camp. I divided my time between the village of Dabel and the nomadic satellite settlements. My camp – this is the urban variant, with my desk in the middle – looked pretty similar in both cases. (Photo © G. Schlee, 1991)

Figure 3.10  The anthropologist’s tent (Photo © G. Schlee, 1991)

80

Modern Trends

Ganna communicates with both Muslim and non-Islamic spirits, and is reputed to have close relationships with animal spirits as well. The indigenous Cushitic language, Oromo, has another concept that denotes entities of this kind: ayaan. The meaning of this word ranges from days of the Oromo calendar and creatures associated with these days, to all possible forms that God may take (Bartels 1983). Once, Abba Ganna foretold that it would rain in Dabel, which meant there was no reason to go away. Those who nevertheless did so found they had no peace at their new settlement; lions circled around the camp at night, roaring but never actually killing any animals. To escape the threat, the rovers returned to the vicinity of Dabel. Abba Ganna explained that what had frightened them had not been real lions, but messengers (Bergano) of God, sent to shepherd the Sakuye back to the best place again. Here then were spirit animals as the vehicle of God’s will, and of Abba Ganna’s prophecy.9 One can be possessed by spirits too, which is not always necessarily a state to be avoided. Abba Ganna’s followers, particularly the female ones, find many reasons – in the Muslim calendar, to which they have added yet more fitting occasions of their own10 – to induce possession by means of drumming and intensive staccato clapping, accompanied by the repetitive chanting of simple Arabic formulas like the shahaada, the two articles of faith: la illaha illa-lah Muhammadar-rasulullah.

The clapping is done in a vigorous way, not just with the hands, but with a movement of the arms starting at the shoulders. In combination with the ejective chanting it may lead to hyperventilation. On the few occasions I practiced participant observation, I did not fall into trance but my palms started to hurt rather soon. On these occasions, called hadar, which continue until the early morning with undiminished intensity, often a number of women will be seen rolling on the ground, grunting and groaning. The soloists are usually men, with women singing the refrain. The men ‘direct’ the proceedings or beat the drum. I have never observed men in a trance; while it may occur, it would appear to be much less common for men. The men who are directing the possessed women bend menacingly over them from time to time and cry out rabbi sodaat! – Fear the Lord! (Rabbi is easily identified here as a word that Oromo has taken over from Arabic.) Anyone who wonders why the poor, possessed women should be roared at so harshly is of course asking the wrong question. It is not the women who are being addressed, but the spirits that have come to reside in their bodies; they are being admonished not to be too rough with their involuntary hostesses. These women who are so willing to enter ecstatic states and indeed actively provoke them nevertheless all have painful experiences of possession behind them. Entry into the cult group always happens in a situation of crisis and illness. This also applies to Abba Ganna himself, as well as his ‘officers’, both male and female. The process always begins with the illness, then it is overcome thanks to successful negotiations with the invading spirit, with which – and this is the recovery phase – viable arrangements for mutual coexistence can be made.  Mohamed Raasa in conversation with neighbours, 10 September 1990, 10 kilometres southwest of Dabel. 9

10

 Eve’s Day, New Moon.

81

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Figure 3.11 Another urban element is this teenager whom the Chief had made his third wife the year before – standing here in front of her new house with its cement plastering and corrugated iron roof, smartly dressed, contented and great with child. (Photo © G. Schlee, 1991)

Figure 3.12  FDJ (Freie Deutsche Jugend) was the East German youth organization. This FDJ shirt found its way to the Sakuye after the dissolution of the GDR and German unification. (Photo © G. Schlee, 1991)

82

Modern Trends

Figure 3.13  Ali Jarso and the ethnographer (Photo © G. Schlee, 1992)

Abba Ganna is a virtuoso at this. He has mastered his spirit, a particularly great spirit, which is superior to other spirits,11 and thus he can also communicate with other spirits of all kinds. He has power over spirits. One of the legends that has grown up around him during his lifetime tells that he once announced that he wanted to banish all the spirits (ayaana) and drive them into the sea. Other experts in spirit healing begged him to give up the idea, however, since how would they make a living otherwise?12 Today, Abba Ganna does not do much spirit healing any more. He has left this to his deputies, e.g. Corporal Ali Jarso and his wife, whom even Ali Jarso himself says is equal to him in ability, though ‘Corporal’ (Koblo) is much too modest a rank for both of them. With their skills in the handling of spirits, Ali claims, they deserve a title like Sergeant Major. ‘Corporal’ however is his actual rank as a retired police officer, and so he has kept the title, despite his higher standing in the business of keeping spirits under control. He and his wife deal successfully with any kind of madness, simply by calling upon God. Others who might erroneously consider themselves equally gifted would immediately get cold feet as soon as a case of possession became a little violent.13 The spirit with which his wife is possessed is a companion (wahel, B) of Abba Ganna’s spirit, and for this reason, the latter would hate to be without Ali Jarso and his wife. The expertise of these specialists consists in establishing what kind of spirit it is that has taken possession of someone, what it likes and how it can be placated. Clearly, there are no general panaceas when it comes to curing possession. The appropriate procedures depend entirely on the respective spirit. For example, some spirits are revolted by blood and others have an insatiable appetite for it.  Diqa Abdul Ali, 29 January 1992.

11

12

 ibid.  Ali Jarso, 7 February 1992, Buley Jaldes.

13

83

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

I will begin by describing a case of spirits that did not want any blood. I myself appear in this story, which places it firmly in the accredited methodological tradition of participant observation. After having been thrown, high and far, by a frisky camel, I struggled for a long time with sciatica, which severely restricted the mobility of my left leg. Since it was out of reach, modern medicine was not an option; I decided to look for alternatives, with the idea that I could at least make the most of my unhappy state by learning something about Sakuye medicine. So I sought the advice of my neighbours. They decided blood needed to be let, since evidently a congestion of bad blood had built up at the place affected by the fall. In lieu of a scalpel, a razor blade was attached to a stick. Fortunately I had a new, clean one for the purpose. Many tasks demanding a high degree of skill and dexterity are, also here, the domain of women. In the nomadic hamlet in which I was staying however, there wasn’t any woman who was a suitable candidate to carry out the bloodletting. Those who knew the technique were unable to assist because they had a history of spirit possession. If the problem is that the spirit could transfer itself to me during the operation, I said, don’t worry. But clearly that was not the problem. We will return presently to the connection between blood and spirit possession. After the horn was removed, the bleeding quickly stopped. I am illustrating this technique in such detail, because it has been forgotten in the West. Yet it was a universally known therapy there as well until the nineteenth century. Later, I reported to Abba Ganna how my helpers had had to search for a woman who had not been possessed for a long time, since no possessed woman was prepared to carry out the operation, and I asked him what being possessed actually had to do with bloodletting. I sat on one side of the curtain, while Abba Ganna lay on his bed behind it. At the end of the curtain, placed so that she had a view of both sides, sat one of his younger wives, who acted as speaker. She repeated everything Abba Ganna said for my benefit, and repeated everything that I said to him – speaking on these occasions in a very clear and accentuated manner, and correcting my grammar mistakes, so that through her I sounded much more competent than in the original. He questioned me on theological topics. Even as a child I had had problems with the idea of the sacrificial death of Christ, and the consumption of his body and blood, whether symbolic or otherwise, and had refused to be confirmed. I happened to mention to Abba Ganna in the course of the conversation that I had never received communion and had no intention of ever doing so. He returned abruptly to this issue when I asked him about the relationship between possession and bloodletting. ‘When you suck at the horn, you may get blood in your mouth’, he explained, ‘but the spirit that inhabits the person is just as disgusted by your blood as you are by the blood of Jesus’.14 This, then, is what he had made of my theological scruples: that I was revolted by blood. In an Islamic context this makes good sense too, since the consumption of blood is forbidden in Islam, as well as the ingestion of any human substance whatsoever, as human beings clearly belong to the category ‘unclean’, together with dogs, pigs and monkeys, their metabolic cousins.15 This revulsion at cannibalism or anthropophagy, that is, disgust at any human substance, seems by the way to explain why even lax 14

84

 See a similar statement by Teepo Hane, field diary (1991, III: 24).

15

 Field diary (1991, IV: 45, 30).

Modern Trends

Figure 3.14 Here my neighbour, Teepo Hane, can be seen constructing an instrument in lieu of scalpel. (Photo © G. Schlee, 1991)

Figure 3.15 Finally, this woman was prepared to travel the approximately 6 km from Dabel to treat me for the equivalent of 25 cents U.S. at the time. (Photo © Abdullahi Shongolo, 1991)

85

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

Figure 3.16 My back was slit open, and a horn with a hole in the tip was placed over the wound; then a vacuum was created by sucking air through the hole. (Photo © Abdullahi Shongolo, 1991) Figure 3.17 After this the hole was covered, and blood allowed to run into the horn for a while. Here the impression left by the horn on the skin can still be seen. (Photo © Abdullahi Shongolo, 1991)

86

Modern Trends

Muslims and non-Muslims among the Cushites of northern Kenya who have no qualms about consuming the blood of ungulates, if they cut their finger, would never dream of sticking the affected digit in their mouths. Observing that Europeans do in fact do this, can lead to bizarre theories on their part, which cannot be explored further here.16 In another case though, involving the young wife of a wealthy man, we encounter a spirit that could not get enough blood17 – albeit blood allowed by the categories of the pre-Islamic belief system, the blood of pastoral animals. After the birth of a child, this woman was apparently particularly susceptible to possession. After the birth of her first child, I was told, she fell victim to possession, after the middle child not, but after the birth of the youngest, which was born during my field research, it was very bad. This time, the spirit, upon being asked by experts, solicited one sheep, one goat and seven hadar dances of the kind previously described. Recently, her stepson told me, when all other attempts to treat her had failed, and her condition was causing particular concern, she had got up and gone to Abba Ganna. Shortly afterwards she returned in a completely normal state. Treatments had already been undertaken, I was informed, for which her husband had paid seven heifers. But the spirit demanded blood. So the blood of a slaughtered animal was put in a bowl in front of the invalid, who drank it, with a cloth covering both head and bowl. After this, the patient went on to imbibe one and a half kettles of the liquid, several litres in all. When the spirit was finally satisfied and left her, and she was lucid once more, she was asked whether she didn’t feel awfully full. No, not in the least, since she hadn’t drunk anything at all. It was the spirit that had done the drinking. Also according to this logic, the patient could not be reproached for consuming blood, although this is forbidden by Islamic law. The fact that the spirit desires blood is a sign that it must be a heathen Boran spirit, also known as Odola, and an unusually greedy one too. Enough about the stepmother of my interlocutor. His real mother is also possessed, but by a Muslim spirit with far more modest needs. In addition to the occasional declamation of dikir, it is content with a cooking pot full of coffee every fortnight and a piece of a small ruminant once a year. So, to paraphrase Goethe, the spirits nourish their majesty laboriously on sacrifice and prayer (or if not their majesty their shadowy existence).18 There is a final ritual which, I was told, is rather expensive, but worth it, since repeated treatments for recurring possession work out even more costly in the long run. After all, in addition to fulfilling the spirits’ wishes, there are also the specialists in attendance, whose strength has to be kept up with food, tea and in a recent development – since there are only a few of them, and their services are in great demand – even miraa (the stimulant, catha edulis). If the spirit demands a song, then an entire group of singers and dancers has to be provided with food and drink. However, this final ritual, known as seer, does, I am assured, guarantee freedom from relapses.  Moh. Raasa, field diary (1991, III: 28).

16 17

 Field diary (1992, II: 37f).

 Apologies are due to the spirit of the Dichterfürst (Prince of Poets) for distorting two lines of his Prometheus: 18

Ich kenne nichts Ärmeres

I do not know anything more miserable

Unter der Sonn als euch, Götter.

/under the sun than you, Gods.

Ihr nähret kümmerlich Von Opfersteuern Und Gebetshauch Eure Majestät

/You nourish your majesty laboriously on sacrificial taxes and the breath of prayers

87

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

88

To return to the illness of the younger wife of the wealthy man. She had never believed in ayaana, he told me. However, during a visit to Nairobi, she developed a serious cough. Since it was feared that she might have TB, she was X-rayed in Kenyatta Hospital. For the same reason, others from her immediate environment were examined, and he too, still at secondary school at that time, was also X-rayed. Nobody however showed any sign of TB. She then used Vicks VapoRub for the cough, but it caused a severe skin irritation. The doctors said it was an allergy, and gave her an ointment to alleviate the discomfort. But this ointment in its turn also caused burn-like symptoms. Ayaana people then told her she was possessed by a spirit, and he didn’t think much of (modern) medicine – as is generally the case with spirits. In Nairobi, gatherings to cure spirit possession could not be conducted openly, and so the woman was taken to some Garre, who convened such a gathering secretly in a room. Hardly had she entered the room, when her spirit revealed itself. After that, she requested to be taken to Dabel at once, in order to receive competent treatment. After all these case histories and impressions (called ‘anecdotal evidence’ by sceptics), let us now try to adopt a more general perspective and to arrive at some conclusions. Far from being a ‘traditional society’, the Sakuye have undergone recent rapid change as a result of massive violence. The Dabel community as I observed it around 1980 and 1990 had only been constituted as a result of the shifta emergency of the 1960s. Both from the photographic essay which is part of this chapter and from the verbal descriptions it has been made clear that outside forces (like the global trade in second hand clothing) have left their traces. Development, government, outside connections, the borehole, the road to Moyale play an important role. In the perception of the chief, there is a rather modern division between the secular and the religious: Abba Ganna is from the ‘other department’, as the Chief once put it. The ecstatic Islam of the possession cult to which much of this chapter has been devoted is not the only variety of Islam present. Variations exist within families. One of Abba Ganna’s sons is well versed in the scriptures, he is even able to converse fluently in a somewhat bookish sort of Arabic, and is, though relaxed and friendly and not austere, as sober a personality as one can be. His brother has the appearance of a long-haired Sufi. So a rather syncretistic kind of belief in spirits, in which awliya fades into ayaana and ayaana into awliya, coexists with Islamic learning. To those outside, however, the Sakuye defend their own mixture of things. There is much scepticism of Saudi Arabian money donated to mosques in Moyale, donations which are felt to have strings attached. ‘Wahhabi’ influence is resented and at Dabel the Sakuye prefer to get along with their own meagre resources rather than accepting petrodollars. So there are global links and global oppositions. Below the global level we find regional connections of a considerable reach. Dabel can be seen as modelled after Annajina in Bale, Ethiopia, and clients and patients of Abba Ganna are said to come from as far as Mombasa in the other direction. Itinerant long-haired preachers and healers are frequent visitors. Sufism, of course, also has its global dimension, as can be illustrated by the Burhaniyya order in Germany or the Murids in New York, but at Dabel I have not come across such truly global connections, only far-reaching regional ones. Other elements are clearly local. There is an ambivalence about the consumption of blood. It is forbidden by Islam but until a generation or two ago was a regular part of the pre-Islamic Sakuye diet. It still is a regular part of the diet of non-Muslim pastoralists like Rendille or Maasai. We have seen that the spirits which possess Sakuye often demand blood, and we cannot exclude

Modern Trends

the possibility that this desire of the ‘spirit’ reflects a desire of the possessed person. Sakuye (like Kenyan pastoral Somali) only need slight excuses to drink the blood of slaughtered animals. A ‘medical’ reason like flu or weakness during a pregnancy is enough to transform blood from an unlawful food item into a lawful medicine. Sakuye Islam here and in many other examples of lax observance appears tolerant. In the light of the connection between ‘purity and power’ and the dynamics of ‘rigidification’ of rules as they gain wider acceptance and following them is no longer a mark of distinction (Schlee 2008a), we can interpret this as a compromise. Islam, which has been accepted by the Sakuye not so long ago, cannot be too demanding without risking loss of adherents. To be a good Muslim by Sakuye standards does not require too much effort. It can be expected that once the Muslim way of life and Islamic ritual practice are generally accepted, religious elites may raise standards of purity for themselves to remain distinct from the mass. Of course, there are also counter-tendencies to this, as Islam is propagated as an easy and practicable religion for everyone. Local figures of thought or modes of interpretation, however, do not perceive or present themselves necessarily as local or limited in their validity. In this chapter we have seen how questions around the Christian Eucharist, part of the experience from Europe reported by a European, have been interpreted in the light of the ambivalences Sakuye feel about the consumption of blood. Global, regional and local as analytical (‘scalar’) tools need to be distinguished from self-perceptions as global, regional or local. These classifications may coincide or not. Throughout the period of poverty following the disaster of the shifta emergency, the Sakuye of Dabel have made conscious efforts, reminiscent of social engineering, to preserve their community as distinct from others. They agreed not to give their daughters to non-Sakuye for marriage, as no Sakuye was in the position to pay a proper bride price, otherwise all girls would have left the community to become wives of higher bidders from outside. With the loss of their camels the camel oriented PRS rituals like sooriyo and almado had become meaningless. Islam gained importance not only as a provider of meaning; the particular brand of ecstatic Islam practiced by a part of the community even established Dabel as a (sub)centre of a regional ritual network.

ACCEPTANCE AND REJECTION OF CHRISTIANITY AND ISLAM AMONG THE RENDILLE IN NORTHERN KENYA In 1966, Paul Baxter published an essay entitled ‘Acceptance and Rejection of Islam among the Boran of the Northern Frontier District of Kenya’; the title of this section openly acknowledges its paternity.19 Apart from that, the similarities are of a rather general nature: like Baxter, I am concerned  A German version of this section was published as Schlee (1982b). When reading it again in 2007 and 2008 I found it still to be adequate for the 1980s. To revise it after 25 years would have meant to change a text which in the meantime had turned into a historical document. I have therefore confined updates and views from later perspectives to the footnotes, so that the different time layers are mirrored in clearly distinguishable layers of text. Basically, what I wrote in 1982 still holds true for the present situation. History has not taken any sudden U-turns. But developments discernible then, of course, have continued and become more conspicuous. The global framework, in which the Rendille like all of us are situated, has changed considerably. Secularism, which has never had a strong basis in Kenya because an impecunious state has left much of education to the churches, is on the retreat worldwide and all sorts of crusaders and holy warriors, thought to be long distinct species in 1982, have reappeared on the scene. 19

89

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

with the question of the theological compatibilities and incompatibilities of different religious systems. Like him, I intend to examine which social constellations facilitate the transition from one faith to another, and to explore the political connotations of belonging to one or the other world religion, or alternatively, to a traditional one. When one goes into detail however, the case of the Rendille differs far too greatly from that of the neighbouring Boran for me to adopt more than general impulses from Baxter. One example of these differences is that the Rendille have never, or rather only sporadically and as it were erroneously,20 been subjected to harassment by Muslim Somali militias. Nor are there any nomadic Somali herdsmen in the neighbourhood of the Rendille; the few Somali found there are traders and are not, in the main, descended from the established pastoral nomadic Somali tribes in Kenya. Instead, their forebears tend to be isolated individuals from north Somalia who ended up in Kenya in the course of colonial history, like for example veterans of the King’s African Rifles. When Baxter, to cite a further difference, sees in the military and economic superiority of the pastoral Somalis an explanation for the Islamization and Somalization of the southern Boran, in the case of the Rendille one is confronted with a situation in which these factors fall away, and yet Islamization is continuing to spread. Yet another difference from the Boran is that amongst marginalized Rendille people, i.e. those not participating in the nomadic pastoral economy of the Rendille due to poverty or for some other reason, Christian missions and missionaries do play a certain role. Since there is already a monograph on the Rendille,21 I will merely give a brief résumé of the socio-economic situation into which Christianity and Islam step. In pre-colonial times, the population development of both herds and people seems to have been of a cyclical-catastrophic nature. Thanks to the sterilizing effects of the sun, disease is uncommon and infant mortality is low.22 When there is good grazing, little stands in the way of human and animal reproduction. However, it appears that from time to time during a drought the pastoral system would break down, whereupon famine and loss of livestock, together with the long periods of intensified raids by neighbouring ethnic groups these led to, took their toll. According to the Rendille philosophy of history, which is closely allied to their cyclical age-grade system, fortune and misfortune recur every 84 years. Drought, famine and livestock raids, accompanied by ritual killings with specific honours for the killer, still occur today.23 However, the ultimate consequence of this misery – death – is often prevented by famine relief provided by state and church. When it rains again and the milk flows in streams, then even the poorest has a full belly. Thus it is only the yearly dry periods – and precautions against them – as well as the non-materialization of the rainy season which drive the Rendille out of the pastoral sector from time to time. One main alternative occupation Rendille tend to take up is work as night watchmen in Nairobi, where house-owners  Both the Rendille and the Somali believe that Rendille and Somali are brothers, and that anyone who kills a member of the ‘brother’ ethnic group will go mad. This belief is connected to the one that Rendille and Somali share the same ‘blood’ (cf. Schlee 1994a, 2008a). 20

21

 Schlee (1979).  The Rendille did not escape the smallpox epidemic of the 1890s, however.

22

 This was written in 1982. From a later perspective one might add that new factors have emerged which lead to conflict (see chapter 5 in this volume, and Schlee & Shongolo 2012), so that it is not just a question of old types of disruption still occurring today. Food security under modern conditions should be provided by market exchange and money deposits, but fair market conditions which permit this have not emerged. So we still have cycles of disasters. 23

90

Modern Trends

would rather engage people from remote areas of the country for security duties than members of the local populace. In view of the warrior traditions of their people, the Rendille also favour the callings of soldier or policeman (AP: Administration Police). In the dry seasons, older and poorer members of the Rendille population gather near the mission stations. Shepherds and herdsmen then become the flocks of the missionaries. This migration away from the nomadic pastoral sector should not lead us to assume that the population which pursues the traditional way of life of the Rendille is in any way numerically weakened or on the decline. On the contrary, today there are more Rendille engaged in nomadic pastoralism and upholding traditional culture than in the entire period of time over which we have an historical overview.24 Those parts of the population which migrate to towns merely represent the numerical surplus of the pastoral sector, which, in the traditional system, would have fallen victim to famine or the sword. In any attempt to establish statistics pertaining to religions, which due to the complete lack of numerical records can only be done here as rough impressions, it is nonetheless noticeable that all the Rendille who live in mat tents and are nomadic herdsmen follow the traditional belief system and take part in the collective annual rituals (and others related to certain lunar months) carried out for the well-being of people and animals. The highly elaborate agegrade ceremonies, the marriage rites, as well as rituals performed to pass on to new-borns clan-specific powers to curse and bless, are also carried out within this group, which forms the majority of the Rendille population. Christianity and Islam are religions for Rendille who are non-participants in the pastoral economy. There are several hundred adherents of the respective religions, although the number of Christians appears to be inflated by people receiving charity whose actual beliefs are unknown. Since it is not up to me to make normative judgements, however, all those who describe themselves as Muslim or Christian are regarded as such. Christians who have managed to return to the nomadic pastoral sector, possibly with the help of a missionary, after a phase of extreme poverty, bear a slight social stigma and are inclined to reticence about their Christianity. Muslims who are converted outside the pastoral sector and later join a Rendille settlement tend to have a better reputation than the Christians or ex-Christians. However, due to the distribution of meat as a form of reciprocal neighbourliness, and due to the collective life-rhythms of  Nothing is known about population numbers before the smallpox epidemic at the end of the nineteenth century. The statement that there are more Rendille engaged in nomadic pastoralism and upholding traditional culture than ever before needs to be qualified from a present perspective. Most Rendille have become almost sedentary around the schools and dispensaries built by the missions. Few have moved into fixed rectangular houses provided by the missions like in Sambamba Village, Laisamis, and some have taken up agriculture on Marsabit Mountain where they face violent opposition from competitors for land and market shares. Most are still nomads according to Scholz’ definition which puts the emphasis on the absence of fixed dwellings (Scholz 1991: 30f), but they have given up moving their houses on loading camels. They just move from time to time to a clean place in the neighbourhood, for hygienic or for ritual reasons, carrying their belongings themselves or using donkeys. The mobility of the satellite herds has not been adversely affected by this development (cf. Schlee 1991c, 1998a). The decline of the number of speakers of the Rendille language, predicted by some (Heine and Möhlig 1980) has not occurred. Outside the Rendille core areas around Korr and Kargi, many Rendille continue to be linguistically absorbed by Samburu speech communities, but due to population growth the number of Rendille who remain Rendille speakers is also growing. ‘Sloughing off’ a ‘surplus population’ seems to be widespread among pastoralists. For the Turkana example see McCabe (2004: 148, 188). 24

91

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

the Rendille, they find it difficult to observe religious rules of food avoidance and periods of fast.25 Briefly then, the nomadic pastoral majority of the Rendille rejects both Christianity and Islam, although Islam enjoys a somewhat higher social prestige. As a rule of thumb, these two world religions will be found wherever there are no camels. The world religions grow in proportion to the number of Rendille leaving the livestock economy: they grow continuously, without the numbers of adherents of traditional belief systems decreasing. The population grows, and the surplus only is divided up by traditionalism, Christianity and Islam. At a later stage I will recount some actual cases which will further illustrate and specify these relations. However, in order to assess what ‘conversion’ means for the Rendille on the level of religious belief, let us begin by comparing the three theological systems as they appear to the Rendille. In other words, we will first deal with the theological aspects of conversion before returning to the sociological ones. This method of theological analysis takes the logical demands of the belief system as its starting point and proceeds deductively, whereas the sociological analysis takes the inductive route, beginning with specific case stories and generalizing from there. The Rendille are monotheists, and use the same word, Waakh, for ‘God’ and ‘sky’. The ‘heathen terrors’ from which the early missionaries wanted to free Africans are not to be found amongst them. Instead, the Rendille have a belief which gives them courage. A belief in witchcraft and sorcery does exist, especially amongst women, who accuse each other of involvement in such activities. Amongst men, tensions are expressed in a different idiom: amongst younger men as violence, amongst older as curses, uttered as both a collective punishment or as individual revenge for damage or insults. Harmful magic is unacceptable and may lead to the expulsion of anyone suspected of it, usually a woman. The curse however is a socially accepted speech act, and is in fact nothing else but a specific type of prayer in which God is asked to kill someone. Thus God, if he hears the prayer, becomes the final perpetrator of the killing, and God cannot be brought to justice. In this way, the curse becomes a moral deed and its success is its justification. After all, were the curse not justified (the criteria of justification being in God’s hands), God would not have helped bring about its success. Being outraged at a curse is therefore just as futile as being outraged if the rains do not come: Waakh enenyet male-tossati – God does not negotiate (with human beings).26 Despite all these threats, witchcraft, harmful magic, violence and even curses, it is nevertheless possible to call upon God. One can even set curse against curse. Then ‘two stones are thrown against each other; one is smashed and one remains’.27 In war, mutual curses of this type are normal. The counter-curse, however, assumes that one’s own right to live is just as strong as that of one’s opponent; if someone feels guilty towards his adversary, then the former will quickly submit to the latter; that is, he takes  A Muslim Rendille who read this paper points out that converts to Islam are also exposed to some ridicule as ‘those who wash their behinds’ or they are categorized as Somali (‘Dafara’). There are cases in which Muslim Rendille actually have joined Somali communities to the east. In one case this had a fatal outcome. When stock of this Somali community was raided by Ariaal, by way of revenge Somali simply got hold of the one Rendille living among them and cut his throat. 25

 Actually, every Rendille, like every Christian or Muslim, does indeed negotiate with God, and the Latter, by occasionally answering prayers positively, appears to signal His readiness to negotiate; but the proverb would seem to ignore this. 26

92

 See the Rendille text in Schlee (1979: 358).

27

Modern Trends

him by the leg and begs for his blessing. Curses specific to certain clans and the belief in the innate powers of particular people tend to complicate these relationships somewhat; the business of uttering curses will be returned to below. In relation to the concept of God, it has been shown thus far that ultimately God represents a protective force in conflicts between people. As for the cosmological aspect of this idea of God, it has already been mentioned that, to a large extent, it corresponds to ‘sky’. In many contexts Waakh can be equally well translated as ‘God’ or as ‘sky’; and for the Rendille, although of course they distinguish between the two aspects, in many contexts the concept does indeed have both denotations. Even when one of them is clearly foregrounded, the other remains. Waakh a gugaha – God/ the sky is thundering, Waakh a dea – God/the sky is raining, Waakh mirisse keene – God/the sky has brought clouds, Waakhi cheletet – literally, God/ the sky of yesterday, i.e. the last rains, the last rainy season. Waakh hos soo bahche means, the rain made the grass sprout (made it come out), or else, God made the grass sprout; normally however it means both at once. ‘God created the people’ is expressed in exactly the same way. The word ‘grass’ is merely replaced by the word ‘people’: Waakh enenyet soo bahche, literally ‘God made the people come out’. In fact, there are several myths pertaining to the origins of clans which involve people climbing out of holes in the ground. The analogy and linguistic parallels to the rain, which causes life to sprout forth from the earth, are obvious.28 The counterpart to the fertilizing power of God/the sky, the rain, coming from above, is of course – as in many old belief systems – the earth which, suitably enough, takes the female gender in Rendille. The complementarity of earth is reflected in numerous expressions. An example is this half-joking ritual dialogue, a fertility charm: Are the camels pregnant while they are still suckling their young? Yes, they are. Who said that? God/the sky. And? The earth. When? Today. When? Today. Are the women pregnant while they’re still breastfeeding? Yes, they are. Who said that? God/ the sky. And? The earth. When? Today.  Schlee (1978) examines the social, cosmological and mythological implications of the verb ‘to come out’ in Rendille in more detail. 28

93

Islam & Ethnicity in Northern Kenya & Southern Ethiopia When? Today. etc.29

Sky and earth are equally often invoked in curses, often at the same time,30 as they are in name-giving ceremonies.31 Prayers spoken with ‘a cool head’ however, i.e. without anger and without the joking element of the fertility ritual, are always addressed to God/the sky, and never to the earth. Rain means the constant renewal of God’s creative power;32 as regards survival, then, the cosmic/aetiological implications of this concept of God are also enormous. It is God who gives or withholds rain, and thus life. So the nomads, following the green vegetation that results from localized rainfall, follow God: Waakhe raahan – they follow the rain/sky/God. Behaving in a way pleasing to God and the performance of collective rituals is necessary for there to be rain, and for the well-being of beasts and human beings. At a later stage I will discuss the role of God and the relationships He has to different Rendille clans in the Rendille social system.33 At this point though I would like to pause and compare the Rendille concept of God as outlined so far with the corresponding Islamic and Christian ideas. The Rendille and Islamic views of God as a refuge from witchcraft and magic are identical in many respects. The Qur’aan also recognizes the existence of magic – not as being effective in its own right, but only in that God allows it. In confrontations with heathen sorcerers, God helps his prophets to triumph by bestowing them with more powerful magic (cf. Suras II, 102, VII, 113–16, X, 77–82, XX, 58–73). In Sura CXIV God is called upon as a refuge from all kinds of evil, including the evil which emanates from women who blow on knots. Blowing on knots in a rope is a harmful form of magic practised to this day by Arab women.34 Since it acknowledges the existence and effects of occult practices, Islam enjoys a much stronger position with the Rendille and others35 than modern Christianity does. Under the influence of the European Enlightenment (a non-Christian current of thought!) proponents of the Christian faith dispute the existence of sorcery and witchcraft. However, for a person who is genuinely afraid of evil spells, being told the perceived threat is all in the mind  Original text in Schlee (1979: 170).

29

 Rendille text in Schlee (1979: 419).

30

 Schlee (1979: 162).

31

 If God/the sky is regarded as the male entity and the earth as the female, then rain and sperm become analogous to a certain extent, and the creation idea approaches that of procreation. The debate between Christianity and Islam over the nature of Christ, and whether he was created like Adam (cf. Sura III, 47, 59), and thus only the child of God in a metaphorical sense as Adam and his successors were, or whether he was the son of God in some special way – that is, whether he was begotten of God, engendered in some form, would thus acquire a different slant. However, this is an extension of the creation/procreation logic on my part, not a view subscribed to by the Rendille. Indeed, they would always say, Waakhe soobahche, – ‘God has caused to exist/emerge’, and never Waakhe delle – ‘God has begotten/borne’. The sky god, who renews life with rain, is only one aspect of God. In addition, as in the two world religions referred to here, there is an aspect of God that cannot be directly apprehended with the senses, one who judges, punishes, forgives and intervenes in the lives of human beings in numerous and mysterious ways. 32

 This is also the central focus of my monograph Das Glaubens- und Sozialsystem der Rendille (‘The Belief System and Social System of the Rendille’) (Schlee 1979). 33

 Personal communication from J. Abun-Nasr.

34

94

 Parkin (1970: 223 ff).

35

Modern Trends

is not much help. He or she is going to look elsewhere for more effective remedies, and not infrequently, these are purportedly found with Muslim Sheikhs. The modern European features of Christianity, then, make this particular faith harder for the Rendille to accept. Ethiopian Christianity is beyond their horizons, so Christianity is equated with European civilization, as presented daily in missionary work in schools and hospitals. Apart from a handful of American Protestants, this work tends to be carried out by Italian Catholics, who have established a network of colourful concrete churches all across northern Kenya and are particularly active in running schools. In fact, they are so dominant, that schoolchildren are astonished to learn that Jesus was a Jew and not an Italian. Whereas Islam has strong Somali associations, Christianity is associated above all with Europe.36 The implications for the concept of God are that God becomes even more distant, moving out of everyday life and into the Beyond, and that rationalistic categories such as coincidence and medical diagnoses replace curses, blessings and magic as explanations. As far as basic theological categories are concerned, the idea of the Trinity is especially difficult for the Rendille to accept.37 The Trinity implies the godhead of Jesus, conceived of the Holy Ghost; the virginity of Mary, the mother of Christ, serves as the central indicator that a conception of this kind took place. However, quite apart from the question whether or not one accepts this virginity as a plausible fact, the point of virginity as a theological postulate is not comprehensible to Rendille. Since the Rendille allow premarital sexual relations and attach no importance whatsoever to virginity, the train of thought associating chastity, purity and consecration is simply non-existent with them. Accordingly, Catholic clergy with their ideals of chastity are objects of derision for the Rendille, particularly when these ideals are pursued without success. Nevertheless, the complications inherent in the Christian Trinity aside, one cannot help being struck by the correspondences between the Rendille belief system and that of the Abrahamic religions – particularly if one contrasts them with other African belief systems in which spirits and ancestor cults play a far greater role. In the case of the culturally closely related Maasai,  Industrial modernity and modern technology certainly have a fascination for Rendille. The connection between notions of modernity developing in Africa and the mission churches has been explored by Comaroff and Comaroff (1997) and by Donham (1999). These analyses certainly are of great value for the cases they describe and might be valid also for other cases – Falge (2006) has applied them to the Nuer). I hesitate, however, to speak of the progress of modernity among the Rendille or of the development of a Rendille vernacular modernity, because modernity has not really been a model of success among the Rendille. That many Rendille owe their lives to modern influences like famine relief and Western medicine does not make them modern persons. Those among them who have actually acquired skills and knowledge associated with modernity by going to school have done so with mixed success. There are some Rendille university graduates, even one Ph.D. holder in economics who works with the World Bank. But there are many who find that public sector employment, mostly as a teacher, is not an economically comfortable position. Educated people often find their illiterate brothers who engage in pastoralism wealthier than themselves and then question their earlier career choices. A more serious problem is the hundreds of unemployed school leavers. A recent (2008) rise in armed robberies has been connected with these. Recent developments are seen as negative by many Rendille. People impoverished by paying school fees then find the school leavers in whom they have invested unemployed and themselves as recipients of food aid around the mission stations. Rendille luhlo waaye – ‘the Rendille have lost their legs’ is their assessment of the effects of ‘modernity’. 36

 Not only for them: in this regard, one only has to consider the suspicion voiced on the part of Islam that this is in fact polytheism. 37

95

Islam & Ethnicity in Northern Kenya & Southern Ethiopia

similar observations have led to speculations about a possible Hebraic origin.38 Although such hypotheses have not attained currency in academic debates, the similarities which occasioned them remain historically fascinating. Ethiopian-Christian influences, which were felt much further north still, in southern Ethiopia, and at a later stage than Islam, can probably be ignored, since they never reached as far as the Rendille’s neighbours. Historical connections to the Falasha Jews who evolved in the Judaized northern Ethiopian culture among those who rejected Christianity39 are unlikely for the same reason. To definitely exclude such influences is, however, not possible, especially since it has been postulated that many pastoral ethnic groups living in East Africa were located further north in earlier periods. Here, however, I am going to restrict the discussion to just two hypotheses: 1. The similarities between the Rendille belief system and that of the monotheistic world religions is due to the common origin of the ‘Abrahamic’ religion (that is, an earlier stage of these world religions) and the belief systems of pastoral peoples in East and North Africa. 2.  These similarities are a result of relatively recent Islamic influences.

Why this historical problem should be of any significance here is quite simply because, if the second hypothesis should prove valid, the whole question of acceptance and rejection of Christianity and Islam could be reduced to the somewhat banal assertion that societies that were exposed to strong Islamic influences previously are more susceptible to Islamization than to Christianization now. Pre-emptively though, it can already be said at this point that, despite the dizzying historical distances involved, the first hypothesis is the likelier. The argument against hypothesis 2 is that in spite of the considerable theological correspondences, there are virtually no Islamic elements whatsoever in the ritual practices of the Rendille.40 Both their large-scale collective rituals and their smaller household ones have been observed and described in the greatest of detail,41 yet no sign of Arabic formulas, nor of bowing in prayer, nor any other feature originating in Islam’s distinctive rituals, was discerned. The Rendille do not so much as carry leather pouches embroidered with verses from the Qur’aan (recent converts aside), although these are very common amongst only marginally Islamized traditionalists in the Sudanese contact zone. In the cases of ethnic groups that have been exposed to Islamic influences, it is frequently found that ritual forms and outward fashions are adopted long before the articles of faith. With the Rendille though, the content of their religious beliefs is strongly reminiscent of the ‘high religions’, whereas the forms of religious practice differ markedly. A counter-argument might run that the Rendille’s northern neighbours, the Oromo-speaking Gabra, have similar rituals to the Rendille, but that these certainly do have Arabic formulas running through them. Nevertheless, the overall findings, together with comparisons with other ethnic groups, show

38

 Merker (1910).

 For a discussion of Falasha origins and what it meant to be ‘Jewish’ in Ethiopia in different periods see Quirin (1992). 39

 Cf. above, Chapter 2, the sections on ‘The origin of the cycles of seven’ and ‘Islam, modern Somality and the borders of PRS culture in the Horn’. 40

96

41

 Schlee (1979).

Modern Trends

that this ritual complex and that of the Rendille have a common – non-Islamic – origin. The Gabra’s Arabisms are isolated appropriations.42 Apart from belief and ritual, further Arabic elements in Rendille language and culture are quickly accounted for. Although the names of the days of the week are Arabic,43 intermingled remnants of another nomenclature can still be traced: ‘Day of the Cattle, Day of the Camel, Day of the Small Ruminants’, all of which are associated with specific rules of behaviour with regard to these animals and their products.44 The names of three months are of Arabic derivation (Soom, Harafa, Daga [