African Indigenous Religions and Disease

Document Sample
African Indigenous Religions and Disease Powered By Docstoc
					African Indigenous Religions and
        Disease Causation
Studies of Religion in Africa
     Supplements to the Journal of
           Religion in Africa

                      Edited by
                  Paul Gifford
   School of Oriental and African Studies, London

                    Deputy Editor
                 Ingrid Lawrie
        College of the Resurrection, Mirfield

                   VOLUME 28
African Indigenous Religions
  and Disease Causation
 From Spiritual Beings to Living Humans

            David Westerlund

             LEIDEN • BOSTON
                    This book is printed on acid-free paper.

           Library of Congress Cataloging-in-Publication Data

                       LC Control Number: 2006047581

Mériot, Sylvie-Anne.
    [Cuisinier nostalgique. English]
    Nostalgic cooks : another French paradox / edited by Sylvie-Anne Mériot ;
translated by Trevor Cox and Chanelle Paul ; pref. by André Grelon.
        p. cm. — (International studies in sociology and social anthropology,
ISSN 0074-8684 ; v. 97)
    Includes bibliographical references and index.
    ISBN 90-04-14346-7 paperback : alk. paper 1. Restaurant management—
France—History. I. Title. II. Series.

TX945.M49613 2006

                           ISSN     0169-9814
                           ISBN-13: 978 90 04 14433 0
                           ISBN-10: 90 04 14433 1

 © Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands.
   Koninklijke Brill NV incorporates the imprints Brill Academic
         Publishers, Martinus Nijhoff Publishers and VSP.

 All rights reserved. No part of this publication may be reproduced, translated,
   stored in a retrieval system, or transmitted in any form or by any means,
  electronic, mechanical, photocopying, recording or otherwise, without prior
                      written permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Brill
provided that the appropriate fees are paid directly to The Copyright Clearance
      Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.
                          Fees are subject to change.

                          printed in the netherlands

Preface ..........................................................................................   vii

Introduction ..................................................................................       1
Chapter One                Ethnographic Background ..............................                    25
Chapter Two                Heavenly Beings among the San ..................                          41
Chapter Three              God in Maasai Thought ................................                    65
Chapter Four               Sukuma Spirits of Ancestors ..........................                    85
Chapter Five               Kongo Spirits or Nkisi .................................... 103
Chapter Six                Yoruba Divinities ............................................ 121
Chapter Seven              Living Humans among the San
                           and Maasai .................................................... 149
Chapter Eight              Witchery among the Sukuma, Kongo
                           and Yoruba .................................................... 165
Chapter Nine               Factors of Continuity and Change .............. 189
Appendix                   Notes on Natural Causation of Disease ...... 209

References .................................................................................... 217
Index .............................................................................................. 235

This is the end of a long research journey. It started in the 1980s
during my five years as a research fellow in a project, ‘African folk
models and their application’, at the Department of Cultural Anthro-
pology, Uppsala University. This was an interdisciplinary project,
and as a historian of religion I co-operated with anthropologists and
other scholars. One of the early major publications that emanated
from the project was the collective volume Culture, Experience and
Pluralism: Essays on African Ideas of Illness and Healing (1989), which I
co-edited with the project’s director, Anita Jacobson-Widding. My
present study can be seen partly as an expanded and updated ver-
sion of the long essay ‘Pluralism and Change: A Comparative and
Historical Approach to African Disease Etiologies’ from that volume.
   Having studied history and comparative religion at Stockholm
University, I found employment for some years in a department of
cultural anthropology to be a new and valuable experience. Although
my approach to the study of religions was, and still is, mainly his-
torical, I did learn a great deal from the encounter with anthropo-
logical colleagues. For the opportunity to participate in the Africa
project, as we usually called it, I am particularly indebted to Anita
Jacobson-Widding. I would also like to express my thanks to the
other colleagues at her department in Uppsala.
   In 1988 I was employed in the history of religions at the Faculty
of Theology in Uppsala, where I stayed until I returned to Stockholm
in 2003 and started my present work at the new Södertörn University
College. The Faculty of Theology in Uppsala was another good loca-
tion for African studies. In particular, the co-operation with schol-
ars of mission such as the late Carl F. Hallencreutz has been beneficial.
Moreover, Uppsala offers good library resources for research on
Africa. In particular, the University library, Carolina Rediviva, and
the library of the Nordic Africa Institute have rich collections.
   Much of the material for this book has been gathered during
several research visits to European mission archives and African
universities. The archive of the White Fathers in Rome has been
particularly important, but valuable and seldom utilized sources in
Catholic and Protestant archives in Leipzig, Berlin, Chevilly (outside
viii                            preface

Paris), Birmingham and Brussels have also been used. During my
stays in East Africa (Dar es Salaam, Nairobi and Kampala), West
Africa (Ibadan, Ife and Jos) and South Africa (Cape Town and
Durban), I have co-operated with several scholars. For assisting me
in various ways I am especially indebted to Cuthbert Omari, Leuben
Njinya-Mujinya, Jacob Olupona, Umar Danfulani and Abdulkader
   Since the early 1990s I have travelled frequently to Bayreuth in
Germany, where I have used the excellent Africa collections of the
University library. As the material is available on open shelves it is
also very convenient to use. For facilitating and making my stays in
Bayreuth enjoyable, I wish to express my thanks to Ulrich Berner,
the chair-holder in the study of religions (Religionswissenschaft) at the
University of Bayreuth, and all his colleagues. For financial support,
I am, finally, indebted to the Swedish Research Council.

David Westerlund
Stockholm June 2004

                          Ce qu’il nous faut aujourd’hui reconnaître, c’est
                          que le pluralisme ne survient pas de l’extérieur à
                          une société quelle qu’elle soit, mais qu’il lui est tou-
                          jours déjà inhérent.1
                          Uniformity on which model-makers rely is an
The idea of writing this historical and comparative study of African
etiologies of disease grew out of my research in the 1980s and early
1990s on some previous works on African indigenous religions.
Focusing particularly on studies by African scholars of religion, I dis-
cussed in a number of historiographically oriented works the significance
of certain theological and ideological presuppositions.3 In the 1960s
and 1970s leading researchers like Mbiti and Idowu, as well as a
host of others, presented overgeneralized and over systematized accounts
of African religion (in the singular). Though much of their criticism
of previous western research on African religions, related partly to
other kinds of presupposition, was well founded and very important,
their theologically and nationalistically inspired work was weak in terms
of historical and cultural contextualization.

                            Issues and approaches

Since the 1980s, there has been a growing number of more locally
based studies on African religions by African, as well as by western,
scholars of religion.4 Considering this recent tendency in research on
African religions, carried out by scholars of religion to replace ‘com-
parativism’ with ‘localism’, can we avoid a return to an overempha-
sis on comparative perspectives as a reaction to a similar overemphasis

     Hountondji 1980: 233.
     Wijsen & Tanner 2002: 186.
     See especially Westerlund 1985, 1991 and 1993. Cf. Flood 1999.
     One good example of such works, which take the issue of contextualization
seriously, is J.K. Olupona’s book on certain aspects of the religion of the Ondo
Yoruba (Olupona 1991).
2                                   introduction

on the study of religion in a local context? One of the main ques-
tions here is how locally oriented studies may be combined with a
comparative perspective, which is crucially important to the history
of religions, and to the study of religions in general. Before the
approach of this book is briefly outlined, however, some notes on
the use of the concept of religion seem essential.
   In an article entitled ‘Histories of Religion in Africa’, Louis Brenner
has for his purposes defined religion ‘as the field of cultural expres-
sion that focuses specifically on communication and relationship
between human beings and those (usually) unseen spiritual entities
and/or forces that they believe affect their lives’.5 Such a definition
serves my purposes too. I agree with Brenner that religions should
not be seen as self-contained systems or ‘world-views’ and that it is
important to focus on issues of heterogeneity and pluralism.6 Here
the significance of intra-religious as well as inter-religious plurality is
strongly emphasized. For the structuring and analytical focus of the
book, however, the question of inter-religious diversity is of prime
   Even though scholars may disagree strongly, and largely for ideo-
historical reasons, about the old issue of the influence of material
factors on non-material, including religion, and vice versa, it is common
knowledge that there is some connection or ‘co-variation’ between
religion and other dimensions of culture. It can be argued, for exam-
ple, that religions among hunting-gathering peoples tend to have
certain structural similarities that differ from other such similarities
found among agricultural peoples with more complex and centralized
socio-economic and political features. Thus, it is important to study
religion in a wider cultural context.7 However, it should be stressed
that, like religions, cultures are heterogeneous and changing entities.8

      Brenner 2000: 164.
     Ibid., 143. In a previously published study on African disease etiologies (Westerlund
1989), I used ‘pluralism’ as a key concept. See also, e.g., Platvoet 2002: 505.
      Ecological factors are another aspect of significance for the differing forms of reli-
gion which, however, will not be in particular focus in this work. For an interesting,
broad and recent study of religion and the environment, see Tanner & Mitchell 2000.
      As remarked by Wijsen & Tanner (2002: 6 f.), culture is a term that implies
the submergence or at least the subjugation of individuals and small group identi-
ties: ‘descriptions and definitions of culture seem to be characterised by collective-
ness, as if groups and masses of people look and behave the same way’. Cultures
of peoples studied in this book have been identified and depicted mainly by out-
siders, although particularly in recent decades intellectual insiders have contributed
to this work too.
                                   introduction                                        3

Thus, we should be wary of the risk of exaggerating aspects of homo-
geneity and orderliness. While arguing for the significance of con-
textualizing the study of African—and other—religions, I hold that
religion cannot be fully understood by references to non-religious
factors or dimensions, be they social, political, economic, psychological,
intellectual or any other. Hence the issue of religious beliefs should
be taken seriously.
   For the comparative purposes of this book, a systematic selection
of five different peoples or ethnic groups from various parts of Africa
has been made. Whereas the San of southern Africa have a tradition
of hunting-gathering and non-centralized socio-political relations, the
Maasai of East Africa are pastoralists and have a slightly more cen-
tralized and complex socio-political structure. Then there are three
peoples with a mainly agricultural base, the Sukuma of East Africa,
the Kongo of Central Africa and the Yoruba of West Africa, who
differ—from less to more—in terms of cultural complexity.9
   Following conventional usage, the ethnic labels San, Maasai,
Sukuma, Kongo and Yoruba will also be used when referring to
religious and medical conceptions among these peoples.10 It should
be underlined at the outset, however, that this usage does not imply
a simple connection between ethnicity and religion. As will be empha-
sized several times, there may be, for instance, historical and regional
reasons for questioning such a connection. It would be a functionalist
fallacy to conceive of ethnic groups as islands, isolated in time and
space. There is a well documented potential for certain phenomena,
such as myths, divination and witchery eradication movements, to
cross ethnic and other boundaries. It should be remembered, also,
that to a considerable degree ethnic, or ‘tribal’, identities are colonial
inventions. Moreover, since there may be important differences
between various sub-groups and regions of the peoples concerned, I
have tried to specify as much as possible on which sub-groups and
areas the main emphasis lies. The importance of individual variations
should be kept in mind too, although such differences are difficult

       More details about cultural differences between these five peoples will be pro-
vided in chapter 1.
       Some scholars argue that, partly because ethnic labels like these are tainted
with foreign notions of territorial homogeneity, they should be abandoned. However,
it is difficult to avoid such nomenclature, particularly in studies based on ethnographic
material of the past. It should be noted, also, that to a large extent such ethnic labels
have become self-designations of the peoples concerned. See further, e.g., Hersak
2001: 615.
4                                  introduction

to observe in historical and comparative studies like this, which is
based on archival material and literature rather than field research.11
   Being essentially a (religio-)historical study, this work covers a
period of somewhat more than a century. Although there are a few
references to the late nineteenth century, it is basically a work on
the twentieth century, with a main emphasis on the latter half of
that period. This focus is partly directed by the paucity of material
from earlier periods, but it is primarily the result of a conscious
choice. An important reason for concentrating specifically on etiologies
of disease was that I felt, and still feel, that the British anthropologist
Evans-Pritchard was right when, in his classic book Nuer Religion, he
concluded that the test of what is the predominant motif in a religion
is usually, or perhaps always, to what sickness and other problems
are attributed, and what steps are taken to avoid or eliminate them.12
   In an essay on recent changes in anthropological approaches to
the study of African misfortune, Susan Reynolds Whyte concludes
that a basic shift has occurred:
     Affliction, which was once dealt with in monographs on African religion
     and cosmology, now seems to belong to the realm of medicine and
     medical anthropology. What we knew as divination now appears to
     be diagnosis; what we analysed as ritual is termed therapy. The victim
     of supernatural forces is called the patient, and his or her relatives the
     therapy managing group. Ritual specialists have been discovered—by
     both development aid organizations and the African press—to be ‘tra-
     ditional healers’. One is tempted to speak of the medicalization of
     African religion.13
While rightly concluding that studies of misfortune focusing on religious
and social dimensions have not outlived their usefulness, but are con-
tinuously an absolutely central problem,14 Whyte remarks that medical

      It should be added, though, that even in field reports scholars surprisingly seldom
provide names and other information about their informants or interviewees. One
of the reasons for this anonymity is the safeguarding of them. However, it renders
the critical use of sources more difficult; and even in those cases where scholars
feel that anonymity is necessary, it is possible to elaborate on individual variations
and provide information about the (anonymous) informants.
      Evans-Pritchard 1956: 315. See also, e.g., Wijsen & Tanner (2000), who argue
that the religious practices of the Sukuma ‘originate from the misfortunes they expe-
rience in their lives’.
      Whyte 1989: 289. See also Last (1984: 14) who argues that the medicalization
of the vocabulary marks a redirection of interest away from causes of illness and
other kinds of misfortune towards the patient and his or her body.
      This is stressed by, among others, Good (1987: 14 f.).
                                  introduction                                      5

anthropology emphasizes two dimensions that have often been missing
in earlier anthropological works on religion. One is the analytical
autonomy of the suffering body and mind, which often serves as a
particularly compelling motivation for thought and action, whereas
the other dimension necessitated by medical anthropology is an appre-
ciation of larger historical processes. No single ‘ethnomedicine’ exists
in isolation, and all modes of dealing with affliction must be con-
textualized. Interpretations of sickness can be assessed as keys to reli-
gious transformations of the most basic order.15 This may provide a
meeting place for historians of religion who focus on processes, that
is, historical studies of African religions, and some scholars who spe-
cialize in medical anthropology.
   Like African indigenous religions, African medical ideas and practices
are not self-contained systems but rather characterized by inclusiveness
and pragmatism. Therapeutic fields can be extremely diverse, and
an ethnic group rarely, if ever, has a single coherent set of medical
practices.16 While concentrating primarily here on religious aspects
of African etiologies of disease among the five peoples concerned, a
multi- and interdisciplinary openness is crucially significant in this
field of study. As argued by C.M. Good, there seem to be few realms
of inquiry that demand a greater commitment to interdisciplinary
perspectives.17 Such a commitment is one of the characteristics of this
book. The multi- and interdisciplinary interests are manifested in the
structuring of the theme, in the literature used—which is derived
from several disciplines—as well as in the eclectic analyses.18 While
the main focus is religio-historical, insights from the disciplines of his-
tory and ethnography/anthropology in particular are important too.

      Whyte 1989: 299 f.
      In a recent book, Whyte (1997: 224) has argued that ‘some of the most inter-
esting new ethnographies of experience highlight the indeterminate nature of life
and show the problems of capturing what they are trying to study in analytical
concepts and systematic representations’. See further, e.g., Pool 1994: 17 and
Feierman 1995: 78.
      Good 1987: xiv. For some important works that focus on the interrelationship
of illness and culture, see Kleinman 1980; Hahn 1995; and Dahlin 2002: 43 ff.
      The eclectic, non-dogmatic and exploratory character of the discussion is partly
inspired by, among others, Droogers (1985) and Fabian (1985).
6                                  introduction

                                  Some key concepts

Following G.M. Foster, most non-western disease etiologies can be
accounted for by two basic principles, termed ‘personalistic’ and ‘nat-
uralistic’. If illnesses are thought to be caused by the active, purposeful
intervention of human and suprahuman agents, they are personalis-
tic, while naturalistic etiologies refer to natural forces or conditions.19
I find this typology useful, even though, as I have argued in a pre-
vious study,20 there appears to be no rationale for differentiating, on
grounds of principle, ‘non-western systems’ from western ones. All
medical systems are rooted in specific historical and cultural con-
texts, and the two principles discussed by Foster are found in the
west too. There are, for instance, western Christians who believe in
evil spirits as agents of disease. The personalistic principle can also
be found in ‘scientific’ medicine.21 Biomedicine has a naturalistic
character, but a more personalistic orientation is found in social med-
icine and medical psychology. Curative medicine, on the one hand,
focuses on biology and the individual, while prevention and public
health, on the other, are more concerned with social groups.22 ‘To
simplify somewhat, the biomedical paradigm tells us that, for example,
tuberculosis is “caused” by Myobacterium tuberculosis, whereas the behav-
ioral science paradigm tells us that tuberculosis is “caused” by poverty
and malnutrition.’23 However, human beings are simultaneously cul-
tural and biological creatures, and the two dimensions necessarily
interact, in the west as well as in Africa and elsewhere.
   For the purposes of this study on African indigenous etiologies of
disease, I will split Foster’s personalistic category into two. In order
to designate three ideal types of etiology, and, in accordance with the
definition of religion referred to above, I will use the concepts religious
(supra-human), social (human) and natural (mainly physical). The
first category presupposes a belief that human beings are influenced
by or dependent on certain suprahuman or spiritual entities or powers,
who/which may or may not be conceived of in personalistic terms.
Among many African peoples, spirits of ancestors play important roles.

         Foster 1976: 775 ff. Cf. the earlier study by Rivers (1924: 7–8).
         Westerlund 1989: 178.
         On the issue ‘why medicine cannot be a science’, see Munson 1981.
         See further, e.g., Janzen & Feierman 1979: 242.
         Romanucci-Ross, Moerman & Tancredi 1983: viii.
                                  introduction                                     7

Here I follow the religio-phenomenological convention of assigning
these spirits, the ‘living dead’ in Mbiti’s terminology, to the world
of religion. They are certainly human beings, and their living relatives’
relationship to them has both religious and social functions. Yet, as
spiritual beings, ancestors are qualitatively different from living humans.
Thus, I agree with, among others, Brain and disagree with Kopytoff,
who argues that African ancestors are above all elders to be understood
in terms of the same category as living elders.24
   Social, or human, causation here refers to relations between living
human beings, which in Africa often entail a supranormal dimension.
For want of better words, I will use the terms witchcraft, sorcery
and witchery when referring to certain important phenomena in this
category.25 A better precision can be achieved by utilizing indigenous
African concepts, which will also be done, but in a comparative
study like this analytical terms are indispensable. The classical dis-
tinction, made by Evans-Pritchard, between witches with an inherent
supranormal power and sorcerers who use various techniques or
‘medicines’ in order to harm others, is (only) sometimes applicable,
as in the case of the Azande studied by Evans-Pritchard himself.26
The term witchery will be used when there are references to both
witchcraft and sorcery.27 Another important example in the category
of social causation of disease is the curse. As will be shown later,
curses, as well as blessings, are often seen as powerful means for
influencing the state of health of human beings.
   The third category, natural or mainly physical causes of disease,
refers to the effects of entities of nature, such as insects, germs,
natural substances, forces or conditions like certain food, lack of
sleep, cold weather or a destroyed equilibrium of some basic elements

      Brain 1973; Kopytoff 1971. While anthropologists have concentrated mainly
on the social function of the belief in and cult or veneration of ancestors, scholars
of religion have largely accentuated the religious function.
      Cf., e.g., MacGaffey 1980: 304 and Axelson 1983/84.
      Evans-Pritchard 1937.
      It is interesting to note that, although African scholars of religion, for good
reasons, have replaced many terms used by western scholars to designate various
aspects of African religions and cultures, they have continued utilizing the concepts
of witchcraft and sorcery. These words are also common in popular usage in Africa.
In particular, the term witchcraft is employed when references are made to different
manifestations of the harmful use of supranormal power. As remarked by Geschiere
(1997: 14), scholars would isolate themselves from daily discussions in the societies
concerned should they stop using terms like witchcraft and sorcery.
8                                 introduction

in the body. With the exception of the breaking of taboos regarding,
for instance, food and sexuality, in which case the force that makes
people sick is impersonal but not physical or biological, the natural
aspects of African ideas of illness causation have often been overlooked
in works by, among others, scholars of religion and anthropologists.
Natural causation of disease is not a main focus of this study either,
but there is at least a brief appendix on that. If limited, the notes
on such causation are important in order to avoid distorted or one-
sided depictions of African etiologies of disease.28
    As will be seen, my ‘etic’ categorization of disease etiologies may
correspond more or less directly to ‘emic’ conceptual categories of
causality.29 By contrast, the distinction made by some scholars in
medical anthropology between ‘disease’, the specialist diagnosis of
sickness, and ‘illness’, the personally experienced state of being sick,
which may or may not coincide with disease, seems to presuppose
the consideration of a biomedical system. In any case, I have not
found this distinction very useful for the purposes of this study. Hence
disease and illness, as well as sickness and affliction, will be used as
interchangeable terms.
    It should be emphasized repeatedly that the etiological categories
utilized here are ideal types and that, in practice, there are many
overlappings. Ideas of causes of illnesses, thus, tend to be multifactorial
rather than unifactorial. In other words, religious and social disease
explanations often overlap observations of natural causes. Moreover,
it is common to think in terms of ‘ultimate’ causes beyond more
directly observable ones, especially in cases of incurable and serious
illnesses or when a natural therapy proves ineffective. Then questions
of ‘how’ are supplemented with the questions ‘why now’ and ‘why
me’. The answers usually involve references to spiritual beings or
other living humans. Multiple causal modes and combinations co-exist
both in African societies and in the thoughts of individuals.30

      Cf. Pool (1994: 1) who, in my view simplistically, argues that there is an
overemphasis on naturalistic etiologies and practical activity, which he sees as bio-
medically determined constructs imposed on African cultures.
      See also, e.g., Yoder 1981: 241 and Feierman 1985: 79.
      In my essay ‘Pluralism and Change’ (Westerlund 1989), I discuss critically some
different views advanced by Horton (1967) and Augé (1986). In another study
(Horton 1982), the former revised some of his previous ideas about the ‘closed’
predicament of ‘African traditional thought’. Concerning this discussion, see further,
e.g., Gillies 1976: 391 f.; Janzen 1981: 188 ff.; Last 1981: 389 ff.; Comaroff 1982:
60 f.; Feierman 1985: 76; Morris 1987: 306.
                                introduction                                   9

   The main disposition of the book will follow the categorization of
disease etiologies. Thus, I will first present religious aspects of illness
causation among the five peoples concerned. Secondly, the issue of
living human agents of sickness will be treated. Thirdly, there will
be an eclectic historical and comparative analysis of possible reasons
for a gradual major shift from spiritual beings to living humans as
causes of afflictions—indicated already by the sub-title of the book.
Considering the gradually increased significance of human agents of
disease, this analysis will pay much attention to socio-historical factors
as well as to wide socio-economic and political contexts. Finally,
notes on natural causation will be presented in an appendix.

                        Notes on Method and Material

Although, given the scope, the use of sources and literature for this
study has to be quite selective, there is a striving for Grundlichkeit. As
argued by Pouwels, a wise historian draws upon as wide a range of
information as possible in his or her historical interpretations.31 In
particular, comparison of much material is an important way of deal-
ing with the inherent limitations of fieldwork studies, which are usually
based on information from a very restricted number of informants
and often depict an ‘ethnographic present’.32 The bulk of the material
used here is scholarly literature, published as well as unpublished,
and written by non-African and African scholars with more or less
detailed knowledge of the peoples concerned.33 However, much mate-
rial also comes from several European mission archives, especially
the ethnographically most valuable archive of the White Fathers in
Rome. Besides, there are many works written by authors who have
not been scholars but, for example, missionaries and travellers.
   When using such a variety of material with the main purpose of
coming as close as possible to insider or ‘emic’ positions, it is essential
to pay attention to the historical method of source criticism. Hence
criteria such as nearness, dependency and bias will be considered.
Since there is a functional view of the material, in principle no difference
will be made between non-scholarly works and scholarly literature

      Pouwels 1987: 17.
      See also Widlok 2001: 367 and Wijsen & Tanner 2002: 186. Cf. Sullivan 2001.
      For a study on the increasing significance of scholarly works by Africans on
religions in Africa, see Platvoet 1996.
10                              introduction

as sources. In particular, the problem of bias or presuppositions of
the authors who have produced the indirect accounts of African
indigenous conceptions of disease causation will be repeatedly discussed.
All kinds of accounts or discourses are dependent on certain ideo-
historical contexts, and one of the advantages of drawing on as wide
a range of material as possible is that this makes more or less hidden
presuppositions more easily discovered.34

The different groups of San figure to a varying degree in scholarly
works. Whereas the main groups have been treated extensively, much
less attention has been devoted to the smaller ones. The Kung and
Nharo, who are most in focus here, belong to the former category.
In particular, much has been written on the Kung.
   Among the older sources, those by Passarge, Vedder, Dornan,
Bleek, Lebzelter and Schapera have been especially useful.35 One of
the problems with the works by Passarge and Dornan is that they
are often unclear as to which groups of San their information concerns.
Much of the information in the geographer Passarge’s book Die
Buschmänner der Kalahari (1907) is derived from an Nharo (Aikwe) man,
who was Passarge’s servant when he travelled in South Africa in the
late 1890s. Passarge makes frequent use of literature too, but that
material is not clearly differentiated from the oral material, and pre-
cise references to the various written sources are lacking. Not sur-
prisingly, his descriptions of the San are coloured by the prevailing
views of them as being ‘uncivilized savages’. Regarding their language,
for instance, he says that it gives such a strange impression ‘that it
can hardly be recognized as a human language’.36
   The attitudes of Dornan and Vedder, who were both missionaries,
are somewhat more sympathetic. Dornan writes primarily about San
in the Kalahari and other parts of what is now Botswana. Some of

       Cf. MacGaffey 1981: 263 and McKenzie 2002: 111 ff. For an important his-
toriographic and religio-historical study in this field, which focuses on African
material, see van Rinsum 2001. Cf., e.g., the broader work Braun & McCutcheon
2000. See also note 3 above, where some of my own contributions in this area are
listed. Further, in recent years the significance of ‘reflexivity’ has been strongly
emphasized among anthropologists. Cf. Widlock 2001.
       Passarge 1907; Vedder 1912 and 1937; Dornan 1925; Bleek 1928 and 1929;
Lebzelter 1928 and 1934; Schapera 1930.
       Passarge 1907: 17.
                                introduction                                 11

his information stems from Tswana informants, and he is frequently
vague in his references to sources. Hence expressions such as ‘I have
heard of . . .’ abound in the work by Dornan.37 Unlike Passarge and
Dornan, as well as many others of the older authors, Vedder refers
specifically to one particular group of San, namely the Kung of pre-
sent-day Namibia. Vedder maintained contact with some Kung for
a period of about ten years. As concluded by Thurner, his publica-
tions are ‘good sources with much information’.38
   Like her father, W.H.I. Bleek, Dorothea Bleek was one of the
best-known and most prolific specialists on the San. She was interested
primarily in linguistic studies but made substantial ethnographic
contributions as well.39 Viktor Lebzelter spent about two years
(1926–1928) in the field and was supported by, among others, German
and Finnish missionaries. He studied several groups of San but con-
centrated particularly on the Kung in the area of Grootfontein in
present-day Namibia. In addition to his often brief direct contacts
with Kung, he used written sources of differing value, particularly
accounts by officers in the German colonial troops.40 The cautious
remark by Schapera that Lebzelter ‘cannot be entirely freed from a
suspicion of prejudice in favour of the monotheistic doctrine associated
with the name of Pater W. Schmidt and his school of disciples’ seems
appropriate, although Lebzelter was not a dogmatic follower of
Schmidt.41 Schapera’s own standard work, The Khoisan Peoples of South
Africa, is an impressive and most valuable survey of the older material.
He used virtually all the relevant writings that were available before
the publication of that book in 1930. A renowned ethnologist, his
summary is critical and rich in terms of interesting reflections.
   Among more modern works, those by Marshall, Lee, Köhler,
Guenther, Heinz, Barnard, Katz and Thurner have proved particularly
valuable.42 Marshall, who together with her family and others studied
the Kung of the Nyae Nyae region in eastern Namibia, collected

     Dornan 1925.
     Thurner 1983: 88.
     The work that is of greatest interest here is the book on the western Nharo
(Bleek 1928).
     Cf. Lebzelter 1934: 3.
     Schapera 1930: 191. Cf. Thurner 1983: 96.
     Marshall 1962 and 1969; Lee 1967, 1968 and 1984; Köhler 1971, 1978 and
1978/79; Guenther 1975, 1975/76, 1979, 1986 and 1999; Heinz 1975; Barnard
1979 and 1988; Katz 1982; Thurner 1983.
12                                 introduction

her field material mainly in the 1950s. At that time these Kung still
lived fairly isolated from Bantu and white people. Her essays on
‘!Kung Bushman Religious Beliefs’ (1962) and ‘The Medicine Dance
of the !Kung Bushmen’ (1968) are exceedingly informative. Like
Marshall, Lee is one of the most knowledgeable and prolific anthro-
pological specialists on the Kung San. The works by Lee used in
this study are based on fieldwork carried out among the Dobe Kung
in the 1960s and 1970s. The Dobe area is in the Kalahari Desert
of north-western Botswana, not very far from the Nyae Nyae region
on the other side of the border. Another scholar who has co-operated
with Lee and studied the Dobe Kung is Katz, a psychologist whose
book Boiling Energy: Community Healing among the Kalahari Kung (1982)
is a most informative study, although the author spent only about
three months in the field. A new collection of San folklore is Living
Legends of a Dying Culture: Bushmen Myths, Legends and Fables (1994),
which has been edited and illustrated by Fourie. An important recent,
and more general, work on the San is the historically oriented book
The Bushmen of Southern Africa: A Foraging Society in Transition (2000),
jointly written by Smith, Malherbe, Guenther and Berens.
   In recent decades the outstanding German linguist Köhler has
studied intensively the language and culture of the Kxoe San in the
Caprivi strip of Namibia. In comparison to the Nyae Nyae and Dobe
Kung, the Kxoe have been more influenced by surrounding Bantu
cultures. Hence it is interesting to compare their conceptions of dis-
ease causation to those of the Kung. Another example for comparison
is provided by Heinz, who has spent much time among the Ko (Xo)
San in south-western Botswana. The studies of Guenther and Barnard
concern the Nharo San. The anthropologist Guenther has studied
hunting-gathering as well as farm Nharo in the Ghanzi district of
western Botswana. His works are much concerned with religious and
historical issues. Thus, his studies are of great interest here. Whereas
Guenther’s material refers to the eastern Nharo, Barnard’s studies
concern, in particular, the central Nharo.43
   Regarding the San, the only archival material referred to in this
book are some diaries written by M. Gusinde in the early 1950s,

      For more detailed information about most of the scholars referred to here, as
well as about several others, see Thurner 1983: 65–102. Thurner’s book Die trans-
zendenten und mythischen Wesen der San (Buschmänner), published in 1983, is an extensive
work based on religio-ethnological analyses of historical sources.
                                introduction                                 13

now kept in the archive of the Anthropos Institute in St Augustin
near Bonn in Germany.44 Like Lebzelter, Gusinde was supported by
Father Schmidt. Much of his field material, which was collected par-
ticularly among Kung and Kxoe (Hukwe), has been published, and
some of these publications have been used here.

For various reasons, the Maasai have fascinated a great many scholars.
Hence a good number of sources is available.45 However, few of
these are specifically concerned with religious issues, and even fewer
of them deal primarily or exclusively with etiologies of disease. In
gathering written material on Maasai ideas on agents of illness, it
has thus, virtually throughout, been necessary to extract such infor-
mation from sources which deal primarily with other topics or which
are more general in terms of contents.
   It seems appropriate to begin these notes about important sources
on the Maasai with a reference to Merker’s classical study Die Masai:
Ethnographische Monographie eines ostafrikanischen Semitenvolkes (1904). This
is a voluminous work, based on several years of field experience,
which has frequently been cited by other scholars. Yet it must be
read with great caution. As indicated by the title of the book, Merker
held that the Maasai were a Semitic people who had immigrated to
East Africa. To some extent this theoretical bias distorted his pre-
sentation of the Maasai religion. In Merker’s view, this monotheistic
religion was completely different from the religions of neighbouring
peoples, and so very similar to Judaism that a Semitic origin had to
be assumed.46
   Another substantial classical study is Hollis’s The Masai: Their
Language and Folklore (1905), although this has not been used very

      I have, however, come across some published but rarely used material, such
as the studies by Mogg (n.d.) and Estermann (1949), in mission archives too.
      For a valuable bibliography of older material, see Mol 1978: 177–190.
      Although Merker exaggerated the similarities between Maasai religion and
Judaism, it cannot be denied that they do have several elements in common.
However, these can probably best be understood in a religio-typological perspec-
tive. Some authors have argued that Merker’s exaggerated views are explained by
contacts with informants who for a long time had been influenced by Christianity.
See, e.g., Johnston 1915: 482 and Voshaar 1979: 320.
14                                  introduction

much here.47 More important to this study of agents of disease is
Fokken’s essay ‘Gottesanschaungen und religiöse Überlieferungen der
Masai’ (1917). Fokken, who spent many years as a Lutheran mis-
sionary among Arusha and pastoral Maasai in German East Africa,
was one of the best-informed early students of Maasai religion and
culture. Some other missionaries have also made significant contri-
butions to the study of the Maasai. In the archive of the Leipzig
mission I excerpted unpublished and published missionary material
of relevance to my work. In addition to the essay by Fokken, the
studies of Blumer and Buchta proved to be of particular value. Even
though it is a compilation, and coloured by its clear theoretical bias,
Father Schmidt’s overview of Maasai religion in the seventh volume
of Der Ursprung der Gottesidee (1940) is a useful work too. It summa-
rizes the most important early sources and contains a valuable crit-
ical discussion of these sources.48
   Regarding more recent material, the most substantial works are,
paradoxically, also the least accessible because they are unpublished.
Two unpublished doctoral theses by Jacobs and Galaty, respectively,
both of which are based on extensive field research and deal with
the pastoral Maasai, have been quite useful, although neither of them
is concerned primarily with religious issues. To the best of my know-
ledge, the most comprehensive study of Maasai religion is another
unpublished doctoral thesis, written by Voshaar, who before writing
his thesis spent twelve years as a Catholic missionary among Maasai
in Kenya. In 1977 and 1978 he, moreover, carried out proper field
research during a year in Kenya and Tanzania. Eight of those twelve
months were spent among pastoral Maasai and agriculturalist Arusha
in Tanzania.49 Voshaar’s thesis contains some theological discussions,
but it has primarily a descriptive, phenomenological character. It
comprises a wealth of new information. Further, much new mater-
ial on Maasai religion is also presented in a fourth major unpub-
lished work, written by Olsson, a historian of religion.50 This material,

     Hollis’s book is largely a collection of texts (stories, proverbs, riddles and songs).
His pioneering collection has subsequently been supplemented with many others.
See, e.g., Fuchs 1910; Fokken 1914; Massek & Siddai 1974; Olsson 1975; Hauge
1979; Olsson 1982a; Kipury 1983.
     Schmidt 1940: 325–333.
     Voshaar 1979: 5 f.
     Olsson 1975, 1982a and 1982b.
                                 introduction                           15

largely in the form of texts, has been collected during some periods
of field research among pastoral Maasai in Kenya.
   My account of the Arusha derives largely from unpublished sources
too. In addition to the earlier works by German Lutheran mission-
aries, referred to above, there are some very interesting theses and
other studies by mainly Maa-speaking students and scholars at the
Lutheran Theological College Makumira, near the town of Arusha.51
In this category, the works of Kimerei, Benson, Marari and Landei
are especially valuable.52 With the exception of, in particular, some
works by the anthropologist Gulliver,53 there are few substantial pub-
lications available on the culture and religion of these agricultural
   In 1989, Hurskainen, Olsson and Århem simultaneously published
thought-provoking essays on Maasai ideas of illness and healing; and
six years later Sindiga published a more descriptive account of ‘Maasai
traditional medicine’.54 These are important examples of the few spe-
cialized studies in this field. Hurskainen’s more extensive study on
the Parakuyo Maasai in Tanzania and Spencer’s work on the Matapato
Maasai of Kenya have been valuable, even though they are not pri-
marily concerned with religion or concepts of disease.55 This may
also be said about Peron’s more recent book L’Occidentalisation des
Maasaï du Kenya.56 By contrast, Hauge’s study of several groups of
pastoral Maasai in Kenya and Tanzania deals mainly with religious
issues, but its quality is somewhat uneven.57 In 1998 Voshaar pub-
lished an important book, Between the Oreteti-Tree and the Tree of the
Cross, which is a study of the indigenous Maasai religion and, in
particular, its encounter with Christianity.58

The archival material that has been used in the account of Sukuma
religion and disease causation is found in the archive of the White
Fathers in Rome. For historians of religion this is perhaps the most

       Some of these studies concern pastoral Maasai in Tanzania too.
       Kimerei 1973; Benson 1974; Marari 1980; Landei 1982 and n.d.
       Gulliver 1963 and 1969.
       Hurskainen 1989; Olsson 1989; Århem 1989; Sindiga 1995.
       Hurskainen 1984; Spencer 1988.
       Peron 1995.
       Hauge 1979.
       Voshaar 1998.
16                              introduction

valuable of all mission archives. Many members of this congrega-
tion, which was founded in 1868, have been excellent specialists on
African religions, languages and cultures. From the very beginning,
the mission strategy of the White Fathers encouraged intensive stud-
ies of indigenous conditions in the mission areas of East Africa and
elsewhere. Among other things, the missionaries were asked to doc-
ument their findings in diaries and annual reports.59 This material
has been important for the study of the older period, that is, the
late nineteenth and early twentieth centuries. From a religio-histor-
ical point of view, however, the most valuable sources in the archive
of the White Fathers are the answers to the ethnographical ques-
tionnaire of the early 1950s. This questionnaire, Table d’enquêtes sur
les moeurs et coutumes indigènes, was prepared by the international lead-
ership of the congregation. It was compiled by J. Mazé and sent out
to the members in the various mission fields in Africa. In addition
to the answers to the questionnaire from missionaries in Sukumaland,
there are several unpublished manuscripts on Sukuma religion and
culture. Father Hendriks, in particular, produced detailed accounts.
Most of the material of the White Fathers is drawn from northern
Sukumaland.60 In the University library of Dar es Salaam several
manuscripts, as well as published works, by Cory, a well-known gov-
ernment sociologist in Tanganyika during British colonial rule, are
deposited. Some material from these Cory collections has been used
in this book.
   There are few specialized studies of religion and disease causation
among the Sukuma. In the late 1960s, however, two doctoral the-
ses in this field were presented at the Catholic University of America.
C.R. Hatfield Jr completed his work, ‘The Nfumu in Tradition and
Change: A Study of the Position of Religious Practitioners among
the Sukuma of Tanzania, E.A.’, in 1968, while M.B. Reid’s study,
‘Persistence and Change in the Health Concepts and Practices of
the Sukuma of Tanzania, East Africa’, was presented one year later.61
A more recent unpublished work, with a certain theological bias, is

     See further Westerlund 1986a: 13. It is not always clear who the authors of
the various diaries are.
     For some more information on the archive of the White Fathers, see Westerlund
1986a: 14 ff.
     Another unpublished doctoral thesis which has been useful, although it is not
primarily concerned with religion and etiologies of disease, is E.A. Welch’s ‘Life
and Literature of the Sukuma in Tanzania, East Africa’, from 1974.
                                introduction                                  17

A.M. Ng’weshemi’s ‘Mulungu: God in Sukuma Belief and Its
Implications for the Church’ (1990). This work is based largely on
interview material.
   Of published material, the numerous studies of the anthropologist
and colonial civil servant R.E.S. Tanner are particularly important.
Other significant published works include Millroth’s Lyuba: Traditional
Religion of the Sukuma (1965) and Brandström’s doctoral thesis, ‘Boundless
Universe: The Culture of Expansion among the Sukuma-Nyamwezi
of Tanzania’ (1990). An important contemporary specialist on the
indigenous religion and Christianity in Sukumaland is the Dutch mis-
siologist F.J.S. Wijsen, who has written, among other things, a mono-
graph on popular religion and evangelization there.62 In the early
twenty-first century, he published two important books with Tanner.
The first, Seeking a Good Life: Religion and Society in Usukuma, Tanzania
(2000) focuses primarily on religious aspects, whereas in the other,
‘I Am Just a Sukuma’: Globalization and Identity Construction in Northwest
Tanzania (2002), religion is seen in a wide(r) historical and cultural
   Though published in 1967, R.G. Abrahams’s book, The Peoples of
Greater Unyamwezi, Tanzania, is still a valuable introduction to the
larger cultural region in which the Sukuma are found.63 Since the
present book concentrates on the Sukuma, and particularly north-
ern Sukuma, there are only a few references to the Nyamwezi proper.
Comparative information on the Nyamwezi is based mainly on the
classical works by Fr Bösch, a White Father, and the Moravian mis-
sionary W. Blohm,64 but some references to more recent articles will
also be made.65

Concerning the Kongo, there is much material of missionaries and
missionary ethnographers available. Both Protestant and Catholic
sources and literature are used in this study. Most of the Protestant

      Wijsen 1993.
      A classical overview of the Sukuma proper is D.W. Malcolm’s book Sukumaland
(1953). Like Tanner, Malcolm was a colonial civil servant.
      Bösch 1930; Blohm 1933. According to Brandström (1990, chapter 1: 9), the
studies of Bösch and Blohm remain, in terms of ethnographic detail and insight
gained from living and working among the people, unsuperseded until today.
      Tcherkézoff 1985; Blokland 2000.
18                               introduction

material utilized here has been produced by representatives of the
Swedish Covenant Church. Of particular significance are the ethno-
graphic contributions made by Karl Laman, who was a missionary
during a period of almost three decades (1891–1919). The archival
material of Laman and other missionaries from the Swedish Covenant
Church is now deposited with the National Archive in Stockholm.
Here, however, it is basically some published material in Swedish
and English that has been used. Much of Laman’s material is found
in the four volumes of The Kongo, which were published between
1953 and 1968 by Lagercrantz in his series Studia Ethnographica
Upsaliensia. These volumes are based on Laman’s ‘Etnografiska anteck-
ningar’ (ethnographic notes), which in turn are based on a huge col-
lection of notebooks (cahiers), written in the Kongo language by
Congolese assistants who were independent of each other. When
editing The Kongo, Lagercrantz used the Swedish material and, as a
result, the work of the Africans was not very well presented. For
instance, information about time and place of the original material
is often lacking. Such shortcomings notwithstanding, The Kongo is a
valuable collection of very detailed ethnography, which focuses pri-
marily on the Sundi.66 In 1990, when this work was out of print, a
new collection of Laman’s Kikongo texts, translated and edited by
Wyatt MacGaffey, was published.
   Among Catholic missionaries specializing on the Kongo, the Jesuit
van Wing was perhaps the most important ethnographer. His clas-
sical study Études Bakongo: Sociologie—religion et magie (1959) was first
published in two separate volumes, the first of which appeared in
1921 (on sociology) and the second in 1938 (on religion and magic).
Another classic work by a Catholic scholar is Hagenbucher-Sacripanti’s
Les Fondements spirituels du pouvoir de Loango, République populaire du Congo
(1973), which focuses, however, particularly on the Vili and the
Yombe in the north-west. In addition to published works such as
these, material from the archives of the Jesuits in Brussels and of
the Holy Ghost Fathers in Chevilly, near Paris, has been used. Not
only unpublished sources such as diaries and letters but also, and

     Janzen (1972: 325 ff.) argues that colonial ideas of homogeneous ‘tribes’ may
be a cause for the harmonization of the information. This is possible, but it should
be remarked that Lagercrantz did not have access to the original cahiers when he
edited Laman’s notes for publication (Söderberg & Widman 1973: 351). See fur-
ther, e.g., Axelson 1970: 32; Westerlund 1986: 10, 12.
                                  introduction                                    19

more importantly, articles in Catholic journals and magazines are
found in these archives. The Jesuits published, for instance, Missions
Belges de la Compagnie de Jésus and Revue Missionaire des Jésuites Belges.
Among periodicals in the archive of the Holy Ghost Fathers that
proved particularly useful was Les Missions Catholiques, which contains
a large number of articles on the Loango area. Another Catholic
mission periodical that has some material on the indigenous Kongo
culture and religion is the Missions en Chine et Congo,67 published by
the Sceut Fathers.
    In more recent decades a good number of scholarly works on
Kongo religion and concepts of disease have appeared. Among African
scholars, two of the most important are the sociologist T.K.M. Buakasa
and the historian K. Mahaniah. In particular the former’s book
L’Impensé du discours: ‘kindoki’ et ‘nkisi’ en pays Kongo (1973) and the
latter’s La Maladie et la guérison en milieu Kongo (1982) are of much
interest.68 Largely due to the missionary work of the Covenant Church
there are several Swedish scholars who have carried out fieldwork
among the Kongo people. In 1979 Ragnar Widman, who organized
the fine mission archive of the Covenant Church in Sweden, pub-
lished his Ph.D. thesis in the history of religions, Trosföreställningar i
Nedre Zaire från 1880-talet. It provides an informative overview of
Kongo religion, based on a vast amount of material, including many
little-known Swedish and African works. The study by Widman, a
former missionary, has a religionist and Christianizing tendency. A
more radical Christian theology of continuity is represented by Å.
Dalmalm, another former missionary of the Covenant Church, who
in 1985 published her Ph.D. thesis L’Église a l’épreuve de la tradition:
La Communauté évangélique du Zaire et le kindoki. This work is based on
field material from the Manianga area, which is about midway
between Matadi, the coastal area of Zaire, and the capital Kinshasa.
Even though it focuses on the attitudes of the Zairean sister church
of the Covenant Church, her discussion of the problem of kindoki,

      In 1915 it was renamed Missions de Sceut.
      Buakasa’s book focuses primarily on the Ndibu, but he stresses the similarities
between the various Kongo sub-groups. Mahaniah gathered his material both in
villages and in Kinshasa. Among other works of special interest here are the two
unpublished theses by P.-A. Pambou (1979–80; 1980–81) and L. Dimomfu’s long
essay ‘L’Art de guérir chez les Kongo du Zaire: Discours magique ou science medi-
cale?’ (1984).
20                               introduction

or witchery, is of interest also to scholars outside the field of mis-
sion studies or church history, historians of religion and anthropol-
ogists, for example.
   A different point of departure characterizes the works of the Swedish
anthropologist Anita Jacobson-Widding. Her major study of Kongo
culture and religion, Red-White-Black as a Mode of Thought: A Study of
Triadic Classification by Colours in the Ritual Symbolism and Cognitive Thought
of the Peoples of the Lower Congo is a structuralist-inspired work on the
symbolic meaning of colours. Although it draws largely on the ethno-
graphic material of Laman and other Swedish missionaries, it does
not focus on processes of change. A more historical approach is
found in the works by the American anthropologists John Janzen
and the above-mentioned MacGaffey. Like Dalmalm’s L’Église a
l’épreuve de la tradition, Janzen’s book The Quest for Therapy: Medical
Pluralism in Lower Zaire (1978) is based on field material from the
Manianga area. It is a pioneering work on medical pluralism in an
African context. MacGaffey’s volume Religion and Society in Central
Africa: The Bakongo of Lower Zaire (1986) is a broad study that deals
primarily with the nineteenth century. Another important volume by
MacGaffey is Kongo Political Culture: The Conceptual Challenge of the
Particular (2000), where he holistically integrates perspectives of different

Since Yoruba culture and religion have interested a very large num-
ber of scholars, there is an abundance of material available. The
material used here refers to virtually all parts of Yorubaland, but
there is a certain preponderance of works on Oyo and Egba Yoruba.
Many of the studies are based on information collected in and around
the cities of Ibadan, Abeokuta and Ife. The oldest sources utilized
here are from the archive of the Church Missionary Society.
Missionaries from this society, who worked mainly among the Egba
in or outside Abeokuta, wrote journals that occasionally include infor-
mation on religion and disease causation.69 Some of the people
associated with the Church Missionary Society also presented their
experiences and knowledge in published form. Two important early

       For a table of CMS missionaries and others, see McKenzie 1997: 561.
                               introduction                                 21

examples are Samuel Adjai Crowther’s Experiences with Heathens and
Mohammedans in West Africa (1892) and James Johnson’s Yoruba Heathenism
(1899). In general, the unpublished and published works of the early
Anglican missionaries are characterized by very critical and polem-
ical views of Yoruba religion or ‘heathenism’. Their theology of reli-
gion was a theology of discontinuity, which coloured their presentations
of Yoruba religion.
   The main title of Stephen Farrow’s book from 1926, Faith, Fancies
and Fetich or Yoruba Paganism, indicates that he wrote in the same
polemical vein. In this book he criticized the work of another early
scholar, Richard Dennett’s classical Nigerian Studies or the Religious and
Political System of the Yoruba (1910), which was reprinted in 1968, for
having an anti-Christian bias.70 This is a questionable criticism,
although Dennett’s book contains some evolutionist speculations and
depicts the Yoruba as markedly ‘superior’ to their neighbours. Further,
some useful material on Yoruba religion and medicine is found in
Samuel Johnson’s The History of the Yorubas: From the Earliest Times to
the Beginning of the Protectorate. This book was edited in 1921 by Johnson’s
brother, Obadiah, who rewrote and edited the copious notes of his
brother, which had been forwarded to England in 1899, two years
before Samuel’s death.
   Several of the early authors who published studies on Yoruba reli-
gion, culture and history were Yoruba themselves, who through their
contacts with the Church Missionary Society had received a west-
ern-style education. In 1948, another Yoruba writer, Jonathan Olumide
Lucas, whose book was published by this society, issued his classical
The Religion of the Yorubas: Being an Account of the Yoruba Peoples of Southern
Nigeria, Especially in Relation to the Religion of Ancient Egypt. As indicated
by the sub-title of this work, Lucas speculated about an Egyptian
origin of the Yoruba. Yet his book contains a great deal of inter-
esting information on Yoruba religion and ‘magic’.
   The archival material on the Oyo Yoruba that has been used here
is made up of the answers to the questionnaire of the White Fathers.71
As a rule, Catholic Fathers who produced these answers in the 1950s
were less conservative or polemical in terms of theology of religion
than were the earlier missionaries of the Church Missionary Society.

       Farrow 1926: 5.
       Table n.d.b.
22                                  introduction

In comparison to the material of the White Fathers in Sukumaland,
however, the Oyo material is much less detailed and of more uneven
   In more recent decades, scholars of religion, anthropologists and
others have produced a large number of publications on Yoruba reli-
gion and medicine. There are also many important unpublished
works written by Nigerian scholars in Departments of Religious
Studies, especially at the Universities of Ibadan and Ile-Ife. Many
Nigerian scholars of religion have been inspired by the prolific pio-
neer E.B. Idowu, particularly by his books Olódùmarè: God in Yoruba
Belief (1963) and African Traditional Religion: A Definition (1973). One of
the scholars influenced by Idowu is J.O. Awolalu, perhaps Idowu’s
best-known disciple, who in 1979 published his book Yoruba Beliefs
and Sacrificial Rites. The works of Idowu and Awolalu, as well as of
others on whom they have exerted an influence, are theologically
coloured and based on a theology of continuity. Idowu’s studies in
particular are also characterized by a nationalistic fervour. In gen-
eral, Idowu’s and Awolalu’s works tend to present a unified or har-
monized picture of Yoruba religion.72
   More specific references to one particular group are found, for
instance, in T.M. Oladapo’s Ph.D. thesis, ‘Traditional Healing among
the Ife: A Yoruba Group in Nigeria’ (1984).73 Like Oladapo’s study,
P.A. Dopamu’s extremely detailed Ph.D. thesis, ‘The Practice of
Magic and Medicine in Yoruba Traditional Religion’ (1977), pre-
sented at the Department of Religious Studies, University of Ibadan,
is based primarily on field material. Dopamu collected his material
in several localities in Oyo as well as in Ondo and Ogun states.
Another broad and informative Ph.D. thesis, presented at the same
Department, is O. Olukunle’s ‘Witchcraft: A Study in Yoruba Belief
and Metaphysics’ (1979).74
   As will be shown in chapter six, works by anthropologists like
Morton-Williams, who outlines the religious features of the Oyo

     See further Westerlund 1985: 26–38, 53 f.
     The author died before the time of presentation to the board.
     At the time of my visit to Ibadan in October and November 1989, S.A.
Osunwole’s Ph.D. thesis Healing in Yoruba Traditional Beliefs Systems (Institute of African
Studies, University of Ibadan, 1989) was not yet available. However, I had access
to some of his earlier papers, of which the one entitled ‘Witchcraft and Sorcery:
A Study in Yoruba Beliefs and Medicine’ (Osunwole n.d.) proved particularly
                                 introduction                                   23

Yoruba,75 and Buckley, whose work was concentrated on the city of
Ibadan,76 differ essentially from most studies of Nigerian scholars of
religion. Buckley’s book Yoruba Medicine (1985) gives the impression
that suprahuman and human agents of disease are of little significance.
His presentation of the encyclopaedic knowledge of two Yoruba doc-
tors or herbalists is certainly very interesting and elegant. Yet one
wonders how representative these informants are. No matter how
important and previously overlooked the herbalist aspects of Yoruba
medicine may be, other studies show that there is much more to it
than that.
   Other specialized works on Yoruba medicine and ideas of illness
include Disorder Among the Yoruba: A Report from the Cornell-Aro Mental
Health Research Project in the Western Region, Nigeria,77 Raymond Prince’s
long essay ‘Indigenous Yoruba Psychiatry’ and Una Maclean’s book
Magical Medicine: A Nigerian Case-Study.78 The report Psychiatric Disorder
among the Yoruba is based on material from an area around Abeokuta.
Most of the informants, who belonged to the Egba group, were
Muslims and Christians, but almost none of them had completely
rejected the indigenous Yoruba beliefs, and many clung to these as
‘the paramount truths’.79 Prince’s informants lived in Abeokuta,
Ibadan, Ile-Ife and Ijebu-Odi or in villages close to these towns,
while Maclean gathered her material basically in Ibadan. Another
important study based on research in Ibadan is George E. Simpson’s
detailed work Yoruba Religion and Medicine in Ibadan (1980). Whereas
Prince’s and Maclean’s informants were mainly religious and med-
ical experts, the majority of Simpson’s informants were ordinary peo-
ple, most of whom were Christians and Muslims.80
   One of the best-known western specialists on Yoruba religion and
culture is Ulli Beier. His numerous publications in this and other
Nigerian fields include the recent Auf dem Auge Gottes wächst kein Gras:
Zur Religion, Kunst und Politik der Yoruba und Igbo in Westafrika (1999).

      Morton-Williams 1964.
      Buckley 1976, 1985a and 1985b.
      Leighton et al. 1963.
      Prince 1964; Maclean 1971.
      Leighton et al. 1963: 35.
      In 1979, some valuable but shorter specialized studies of Yoruba medicine and
disease conceptions were published simultaneously by J.A.A. Ayoade, R. Braito &
T. Asuni and N.M. Wolff in the collective volume African Therapeutic Systems, ed. by
Z.A. Ademuwagun et al. Among other studies of particular interest here, those by
Asuni (1976) and Maclean (1976) may also be mentioned.
24                            introduction

Another well-known specialist on Yoruba religion is Peter McKenzie,
whose large-scale work Hail Orisha! A Phenomenology of a West African
Religion in the Mid-Nineteenth Century (1997) is an important example
of the ethnographic and religio-historical usefulness of missionary
archives, in this case primarily of the (Anglican) Church Missionary
Society and secondarily of the Methodist Missionary Society. As indi-
cated by the sub-title of McKenzie’s book, however, it deals with a
period before or at the outset of the time frame of this study.81

      For an interesting methodological debate about McKenzie’s book between
J. Cox and McKenzie himself, see Cox 2001 and McKenzie 2002.
                                  CHAPTER ONE


                            Every tribe believes that its own habits and values
                            are uniquely reasonable; ours is no exception.1
                            The idea of homogeneity is one that stems in
                            part from anthropological studies based on field
                            work restricted by time, expense and the relatively
                            static nature of relationships with a small number
                            of informants.2


In scholarly works on some related groups of people in southern
Africa, ‘traditionally’ with a hunting-gathering mode of subsistence,
the term ‘San’ has frequently been employed. For want of a better
one I have used, and will continue using, that term in this book.
The ‘San’ or ‘Bushmen’ do not have any general or collective name
for themselves—both concepts are, or were originally, pejorative des-
ignations of outsiders.3 The word Bushmen was first used by Dutch
settlers, whereas San is a Khoi (‘Hottentot’) term, which may be
translated ‘aborigines’ or ‘settlers proper’.4 In Botswana, one of the
Bantu terms for San, ‘Basarwa’, is officially employed, despite its
derogatory connotation.5
   Although the short stature, click languages6 and traditions of hunt-
ing and gathering have constituted criteria for the habitual use, by
scholars and other outsiders, of various overarching labels such as
San, Bushmen and Basarwa, it should be stressed that we are here

    MacGaffey 1997: vii.
    Wijsen & Tanner 2002: 187.
    Guenther 1979: 102; Thurner 1983: 60.
    Dornan 1925: 40; Schapera 1930: 31; Lee 1984: 9.
    Guenther 1986: 9; Smith et al. 2000: 65.
    Since this is a comparative study of five peoples, diacritical marks of click sounds
are not included. Likewise, there will be no diacritical marks in the rendering of
Yoruba terms later on. For information about San clicks, see e.g. Lee 1984: xvii
and Thurner 1983: 63.
26                               chapter one

dealing with a cluster of peoples who differ considerably in physical
as well as in cultural and linguistic respects. Some tall and dark-
skinned San in Angola, Namibia and Botswana have been referred
to as ‘black’ in contrast to the short and ‘yellow’ San.7 Gusinde8 and
Thurner, 9 among others, have repeatedly pointed to the great
significance of Bantu and other foreign influences on the San, par-
ticularly among the groups in the northern parts of the San area,
which stretches from the Cape Province in the south, though the
Cape San are now virtually extinct, to south-eastern Angola and
western Zambia in the north. H.-J. Heinz rightly argues that, since
‘there are more differences between Bushmen tribes than there are
resemblances’, scholars should always specify which groups of San
they are referring to.10
   On linguistic grounds, the San peoples or ‘tribes’ have been divided
by scholars into three main clusters or groups. The northern cluster
comprises, among others, the Kung and the Heiom (Heikum) of
Namibia and Botswana as well as Angolan groups such as the Okung.
In the central main group we find, for instance, the Naron or Nharo,
also referred to as Aikwe, and the Kxoe, or Hukwe, of Botswana
and Namibia. The southern cluster, finally, include—or rather included,
since most of the San in the southern areas have been almost com-
pletely exterminated—the Ko (Xo) of southern Kalahari, the Xam
(Kham) of the Cape Province and other groups in the Transvaal,
Lesotho and south-eastern Namibia.11
   Economically, the San still include some full-time hunter-gatherers,
although these are becoming increasingly rare.12 Other San are farm-
ers and herders who mix their new farming and herding activities

     Lee 1984: 10. See also Widlok 2001: 361.
     E.g., Gusinde 1952: 392, 401, 403; 1965: 38.
     Thurner 1983: 63 f., passim.
     Heinz 1975: 19. See also, e.g., Smith et al. 2000: 67 f.; Widlok 2001: 364.
     Here I have relied mainly on Hirshberg (1975: 389) and Thurner (1983: 62).
See also, e.g., Schapera 1930: 31–38.
     The following introductory notes on the culture of the San, as well as on the
Maasai, Sukuma, Kongo and Yoruba later on, will not be documented in detail.
These notes are included here primarily for the service of those readers who are
not specialists in African cultural studies. For more details on the culture of the
San, see e.g. Schapera 1930; Hirschberg 1975; and Smith et al. 2000; and for more
specialized studies of the Kung and Nharo, see e.g. Marshall 1976, Lee 1984,
Guenther 1986 and Guenther 1999. An introductory account of western writings
on the San is found in an article by Guenther (1980), whose title, ‘From “Brutal
Savages” to “Harmless People”’, neatly summarizes the conclusions of the article.
                         ethnographic background                                  27

with hunting and gathering. There is also a growing number of farm
and migrant labourers. While, in the hunting-gathering type of cul-
ture, men are responsible for the hunting, it is mainly women who
gather all kinds of edible vegetarian foodstuff and insects. In the
1980s the San who lived by hunting and, more importantly, by gath-
ering constituted a few thousand people only out of a total of some
tens of thousands of San.13 Among the Kung there was at that time
still a considerable number of hunter-gatherers, whereas the great
majority of the Nharo occasionally or temporarily only returned to
hunting and, more often, to gathering. Regardless of their main mode
of subsistence, most San now live in very poor economic circum-
stances and are largely being exploited by other people.
    Socially, the ‘band’ or ‘camp’ has been, and often continues to
be, a vital unit. A band consists of some tens of people, most of
whom are closely related. As among the other peoples studied in
this book, kinship is an important organizing principle of social life.
However, friends and in-laws can be included in the bands too, and
the group structure may be very flexible. Many scholars have depicted
San social relations, including relations between the sexes, as egali-
tarian. Accordingly, the San have a reputation for not being very
competitive. There is still frequently a marked emphasis on sharing
within the bands. Polygamy is not unknown but is very unusual.
    So-called tribes, such as Kung and Nharo, are essentially linguis-
tic groups, not social or political units. The bands are autonomous,
and institutionalized political offices have been rare. What little polit-
ical authority there may be is usually vested in older men rather
than in headmen or chiefs. Land and hunting territory is not owned
by individuals but by the bands. However, there are no special
boundary marks between the areas occupied by different bands, and
neighbouring bands can come together or split apart for various rea-
sons, like the availability of food and water.
    The Kung in Namibia and Botswana are the San group or ‘tribe’
that will be in particular focus in this book. Secondarily, the Nharo
in Botswana will be studied in some detail too, whereas occasional
references only will be made to other groups. A substantial number
of the Kung are still hunter-gatherers, and this group of San has

      It is impossible to give an exact number. In 1984, Lee’s professional guess was
‘close to 50,000’ (Lee 1984: 9), while a more recent estimate (Smith et al. 2000:
65) says ‘105,000 or so’. Cf. also Guenther 1986: 25.
28                                chapter one

been studied extensively for a long period of time and by several
different scholars. Since the Nharo have been more strongly influenced
by the cultures of neighbouring peoples, it is interesting to compare
their conceptions of disease to those of the Kung. Although there is
less source material available on the Nharo than on the Kung, the
former also belong to those groups of San whose cultures are fairly
well documented.
   In the modern historical period covered here, the conditions of
the San have changed more or less dramatically. Dependence on
other peoples, colonialism and incorporation into modern state appa-
ratuses have ended in an almost total extinction of the hunting-gath-
ering mode of life. In recent decades particularly, drastic impairments
have befallen the Kung in Namibia. Many of them were recruited
into the South African army, and unemployment, drunkenness and
violence have become increasingly common aspects of life. In this
situation there is a widespread breakdown of the custom of sharing.
Even though considerable changes have occurred among Kung in
Botswana too, these have been less drastic than those in Namibia.
Among the Nharo the changes have been most extensive in the
domain of economics, whereas the social features of life have retained
much of the typical band pattern.


The term ‘Maasai’ refers to some groups of people in southern Kenya
and northern Tanzania. Whereas some scholars use this concept to
designate all who speak the Nilotic Maa language,14 others prefer to
leave out some ‘peripheral’ groups of Maa-speakers, for example
Samburu in Kenya and Arusha in Tanzania. The word ‘groups’
here corresponds to the Maa word iloshon (sing. olosho), which is fre-
quently translated ‘sections’ or ‘tribes’. These sections have their own
names, territories, peculiarities of dress, hair-styles, weapons and so
on. They also have their own observance of ceremonies and have
not been politically united. The dominant section in the northern

      In its strictest sense, the word ‘Maasai’ means ‘speaker of the language Maa’
(Beckwith & Saitoti 1980: 18). Maa is an introductory particle to questions, state-
ments and speech in general, which draws polite attention to the speaker. As an
introductory particle it entails attitudes such as respect, exitement, politeness and
tenderness (Voshaar 1979: 17). See also Voshaar 1998: 56 f.
                         ethnographic background                                  29

parts of the area inhabited by the Maasai is Purko, and in the south
it is Kisonko (Kisongo). As pointed out by Spencer,15 most writers
have ‘overlooked the possibility of variation between the sixteen or
so tribal sections’. Even recent scholars have tended to note that the
Maasai do this or that, rather than noting, for instance, that the
Purko Maasai do this or the Kisonko Maasai do that.16
   More often, a broader distinction between pastoral Maasai and
agricultural Maasai is made. Although this seems to be a funda-
mental and important distinction, which I will attempt to follow here,
it should be remarked that it is not a precise one. Agriculturalists
have livestock, and pastoralists occasionally have to supplement the
pastoral food products with vegetable foods—their alleged dislike for
agricultural produce has sometimes been exaggerated.17 Whereas my
presentation of ‘pastoral Maasai’ is based on material from most of
the mainly pastoral sections, the presentation of ‘agricultural Maasai’
is restricted basically to the Arusha section in Tanzania.18 In 1969,
P.H. Gulliver estimated the rapidly growing number of the Arusha
to be about 100,000;19 and in a more recent study the pastoral
Maasai were estimated at about 300,000.20
   Economically, cattle are the most important asset to pastoral Maasai.
However, cattle play a central role not only economically but also
culturally. According to a well-known myth, cattle belong to the
Maasai by divine right, and raids to bring home ‘straying’ cattle

      Spencer 1988: 2.
      Spencer’s book (1988) is unusual but worthy of imitation in that even the title
indicates that it is a study of one particular section, the Matapato.
      Voshaar (1979: 27) even holds that Maasai would be ‘quite horrified’ to learn
that they are supposed to abhore plant food. He adds that the eating of such food
is ‘quite common now’, and that it was ‘not uncommon formerly’. See also, e.g.,
Orr & Gilks 1931: 22; Beckwith & Saitoti 1980: 29; Hurskainen 1984: 119; Sicard
1999: 81. Jacobs (1965: 30 ff.) notes that the term ‘ilOikop’ has been used by ‘pure
Maasai’ or ‘Maasai proper’ as a derogatory designation for Maa-speakers who have
engaged in agricultural work, fishing and hunting-gathering. The word ilOikop con-
tains the root of the term enkop (ground, earth), but it can also be used with ref-
erence to dead or murdered people (Mol 1978: 52, 108; Voshaar 1979: 26;
Berg-Schlosser 1984: 150). See also Voshaar 1998: 57 f.
      As will be shown, there are in terms of religion and disease causation certain
important differences between pastoral Maasai and Arusha. Although the differences
between the various sections of pastoral Maasai are clearly less significant, it is
regrettable that, in most cases, the sources do not provide enough information for
more detailed regional comparisons.
      Gulliver 1969: 229.
      Århem 1987: 4. Cf., e.g., Kronenberg 1979: 180; Berg-Schlosser 1984: 152.
30                             chapter one

among neighbouring peoples have been a common feature of Maasai
history.21 While sheep and goats are often slaughtered for food, cat-
tle are reserved for rituals and ceremonial feasts. Milk is consumed
on a daily basis, and during the dry season, when milk yields are
low, it can be mixed with fresh blood from cattle. However, the
consumption of blood has become increasingly rare. During the wet
season, when sufficient grass and water can be found in most parts
of Maasailand, the herds of the semi-nomadic Maasai are dispersed
over large areas; but during the dry season they are concentrated
near relatively few permanent sources of water.
   In the patrilineal social system there are two moieties, which are
divided into clans. The clans, in turn, are divided into sub-clans.
Apart from determining marriageability—exogamy now being prac-
tised basically on the sub-clan level—kinship is, however, of little
significance. Kin ties are usually less important than age peers or
friends. The age-group organization plays a central role, although
its importance was even greater in the past. The main divisions
within this organization, which concerns the men only, are between
young boys, warriors and elders. When a boy reaches adolescence,
he is circumcised and becomes initiated into warriorhood. About
seven years later, junior warriors are promoted to senior warriors,
and after another seven years or so a final major ceremony is held,
marking the beginning of elderhood.22
   Boys are responsible for the routine herding activities. Warriors,
who used to live in separate villages, are expected to protect the life
and property of the community members. In the past, but rarely
nowadays, warriors fought enemies and carried out cattle raids. Elders,
who are not seldom polygynous, wield the political, economic and
religious power in the community. Some early western observers held
that the iloibonok (healers or religious leaders), who will be studied in
more detail in chapters three and, especially, seven, were chiefs and
headmen.23 This was, however, a misunderstanding of their position
and function. True, a few outstanding iloibonok have been politically
influential, but that has been because of the weight of their per-
sonalities rather than because of their position as iloibonok. In principle

     See, e.g., Hauge 1979: 79.
     Like warriorhood, boyhood and elderhood may be sub-divided into junior and
senior grades, which are periods of about seven years each.
     E.g., Krapf 1857: 440. Cf., e.g., the more recent work by Waller 1995.
                       ethnographic background                               31

at least, there is no indigenous political authority above that of the
   Even though women are not members of age-groups, they may
be said to be attached to the various groups in that they are daugh-
ters, wives and mothers of men in their various groups.24 Unlike
men, women usually marry at a very young age. Whereas boys are
circumcised during communally strictly regulated periods with inter-
vals of a few years, circumcision of girls is performed on an indi-
vidual or private basis when they reach puberty. After circumcision,
in the form of clitoridectomy, girls are eligible for marriage.25 Women,
who are much dominated by men, do most of the work at home.
Gathering material for houses as well as building them, fetching
water and firewood, milking and distributing food and, above all,
bearing and taking care of children are some of their duties.
   The land-oriented economy of the Arusha is the main difference
from their cattle-oriented neighbours. As a settled and recognizable
group, the Arusha community is about 200 years old. Small groups
of Maa-speakers, who were refugees from Maasai internecine wars
during the early nineteenth century, came to settle on the south-
western slopes of Mount Meru, where they established a new seden-
tary, agricultural community. To some extent their basically Maasai
social, political and religious system gradually became influenced by
the influx of particularly Meru and Chagga immigrants, who came
there as refugees or captives.26
   Nevertheless, Arusha as well as pastoral Maasai have a reputation
for being ‘conservative’. The latter in particular are frequently described
as culturally exclusive and reluctant to accept changes. In terms of
development efforts the Maasai were largely ignored by the colonial
authorities. The ‘capitalist’ rulers of post-colonial Kenya as well as
the ‘socialist’ establishment of Tanzania have made strong attempts
to initiate significant changes in the lives of the Maasai, but in both
cases with meagre results. The reluctance to abandon the ‘traditional’
way of life is seen in the religious sphere too. Despite protracted

     Voshaar 1979: 62.
     For a detailed account of female circumcision, see Sicard 1999: chapter 6.
     The Arusha language or dialect is quite similar to other Maa dialects, even
though the pronunciation, in particular, has been influenced by Bantu languages
(Hohenberger 1958: 16). See further, e.g., Voshaar 1998: chapter 2 (61–99).
32                                 chapter one

efforts by various Christian missions, very few pastoral Maasai as
well as Arusha have converted to Christianity.27


The directional term Sukuma, which means north, is used as a self-
designation of a people who inhabit an area south and south-east
of Lake Victoria, mainly in the Mwanza and Shinyanga Regions of
north-western Tanzania. They speak a Bantu language, and are, lin-
guistically as well as culturally, closely related to their southern neigh-
bours, the Nyamwezi. Today the Sukuma may number more than two
million. They constitute the most numerous ethnic group of Tanzania.28
Sukuma may be seen as a generic name for people with different
origins, and it is important to keep the provisional nature and con-
textual relativity of such an ethnic label in mind.29 Moreover, like
many other ‘tribal’ identities, ‘Sukumaness’ is a fairly recent inven-
tion. While people now speak about themselves as Sukuma, in the
mid-twentieth century or earlier they referred to themselves by using
clan names.30
   Although my material refers to various parts of the large Suku-
maland, at least the bulk of the archival sources are from the
north. In comparison to Sukuma and especially Nyamwezi further
south, the northern Sukuma have been less influenced by outside
factors such as Islam and Arabs. During colonial rule great num-
bers of men in the southern area around Tabora worked as porters

      The information on Maasai culture is based on, among others, the following
works: Gulliver 1963; Jacobs 1965; Gulliver 1969; Ho et al. 1971: Galaty 1977;
Voshaar 1979; Donovan 1982; Berg-Schlosser 1984; Århem 1985; Århem 1987;
Voshaar 1998. There are many brief introductions to Maasai culture available, for
example Sankan 1971; Salvadori & Fedders 1973; Kronenberg 1979: 171–177;
Beckwith & Saitoti 1980; Kipury 1983: 1–9.
      According to the last, or latest, national census report in which ethnic affiliation
was recorded (1967), the Sukuma numbered one and a half million. However, in
the early twenty-first century, the figure six million(!) was given by Wijsen & Tanner
(2002: 1).
      This has also been stressed by, among others, Malcolm (1953: 43) and Brandström
(1990: chapter 1: 2). For instance, it was not until the colonial period that Sukuma
and Nyamwezi became established and perceived of as two distinct and separate
groups. In some old writings the term Bagwe, which was not a self-designation, was
used instead of Sukuma. See, e.g., Barthelemy 1905: 285; Gass 1973 [1919]: 385;
Table n.d.a.: 2.
      Wijsen & Tanner 2002: 1.
                      ethnographic background                             33

and were used to being on the move. For a long time Muslims and
particularly Christians have tried to convert the Sukuma, but most
of them have been slow to respond. According to Wijsen and Tanner
(2000), about 85 per cent of them are followers of the indigenous
religion, while the others are either Christians (13 per cent) or Muslims
(2 per cent).31 In general, the Sukuma have had a reputation of being
‘conservative’.32 It may, certainly, be argued that there are impor-
tant differences between various regions, social groups and individ-
uals in Sukumaland.33 The issue of cultural variability is strongly
stressed by Wijsen and Tanner.34 However, in comparison to the
culture(s) of the (even) more highly dispersed San people(s) of south-
ern Africa, the culture of the Sukuma people seems to be somewhat
less heterogeneous.
   In 1897 Father Brard reported that the Sukuma were sedentary
farmers who cultivated sorghum, potatoes, manioc, groundnuts and
beans. Their tobacco was ‘excellent’, he felt. Brard added that there
were many sheep and goats. Partly as a result of serious epidemics
there were few cattle, although their number was on the increase.35
Later the number of cattle grew to such an extent that modern
scholars have referred to the Sukuma as agro-pastoralists. Yet culti-
vation is still the main subsistence activity, even though animal hus-
bandry is of great economic and social significance, and cattle are
not very fundamental to the religious and symbolic system.36 Currently,
sorghum and maize are the main food crops, but millet, potatoes
and cassava are important too. Cotton is the predominant cash crop.
The old hoe-cultivation is now increasingly being replaced by ox-
   The Sukuma have a tradition of living dispersed in small settle-
ments, at least prior to the Tanzanian villagization process in the
mid-1970s, which introduced some changes in this respect. Individual
homesteads or clusters of homesteads formed neighbourhoods that
varied in size from two or three to a few hundred. Neighbours col-
laborate in a wide range of activities such as house-building, agricultural

     Wijsen & Tanner 2000: 10.
     See, e.g., Table n.d.a: 278; Itandala 1983: 16, 34.
     Tanner 1967: 1; Brandström 1990: chapter 5: 30.
     Wijsen & Tanner 2002: 186 ff., passim.
     Brard 1897: 156.
     Cattle are the most important means for accumulating wealth and are fre-
quently used, for instance, as bridewealth in marriage transactions.
34                             chapter one

work and rituals. People often move from one neighbourhood or vil-
lage to another, and kinsfolk may be spread over a wide area. Budugu
is a core concept in the Sukuma social system. It is an all-embrac-
ing concept that refers to the collectivity of living and dead kin as
well as to the state of being a kinsman, ndugu (plur. budugu). Kinspeople
may not marry each other. With the exception of some chiefly fam-
ilies, the Sukuma are patrilineal. Today kinship relations are less
important than they have been.
    When the institution of chieftainship was introduced into Suku-
maland, possibly about 300 years ago, it did not wholly replace the
significant gerontocratic organization of the society. Hence elders and
associations of elders continued to wield great influence. The chiefs,
batemi (sing. ntemi), came as strangers, and their rule was not auto-
cratic. Elders and neighbourhood headmen were largely responsible
for the choice of chiefs and for controlling their actions. Inefficient
chiefs could even be expelled from their chiefdoms. There were also
associations of younger men who to some extent could resist or coun-
terbalance the authority of the chiefs. Besides, the numerous Sukuma
chiefdoms, governed by various chiefly dynasties, were not central-
ized but autonomous units.
    In the late 1890s Brard described the Sukuma chief as a ‘primus
inter pares’.37 During colonial rule, however, when chiefs became depen-
dent on foreign, centralized regimes, the local balance of power
shifted in favour of the chiefs. They were now able to add to their
primarily ritual functions important secular duties as, for example,
judges and collectors of new taxes. Both the distribution of chiefly
power and the selection of the chiefs ceased to be dependent for its
functions on the people of the chiefdom. Shortly after the attain-
ment of independence in Tanzania (1961), another important change
occurred. In 1963 the new government abolished chieftainships alto-
gether. Although to a limited extent certain ritual and other func-
tions have continued to exist, and at times even been revived, the
chiefs have lost most of their power or have entered the new admin-
istrative and political system on local or national level.38

     Brard 1897: 155.
     Among sources or literature that have been used for the overview of Sukuma
culture, the following may be mentioned: Table n.d.a.; Cory 1951; Malcolm 1953;
Schans 1955; Abrahams 1967; Tanner 1970; Itandala 1983; Wanitzek 1986;
Brandström 1990; Tanner & Wijsen 2000; Tanner & Wijsen 2002.
                        ethnographic background                                 35


In the early 1480s Portuguese sailors ‘discovered’ the Kongo king-
dom, and because of this early and continued intrusion the history
of the Kongo people is much better documented in written sources
than is the history of other peoples in sub-Saharan Africa. Northwest
of the Kongo area of western Central Africa was the less well-known
Loango kingdom, formed by the Vili, with a closely related type of
culture and religion. Both these areas and Bantu groups of people
became subjected to a long-standing western and Christian mission-
ary influence. Furthermore, large numbers of Africans were taken as
slaves. Catholic missionaries belonging to several orders, including
Jesuits and Holy Ghost Fathers, confronted the indigenous religion
and attempted to win converts. In the late nineteenth century, when
modern colonialism started, missionary endeavours expanded very
considerably with the coming of various Protestant denominations,
including Baptists and members of the Swedish Covenant Church
(Svenska Missionsförbundet). In addition to the Catholic Church, the
Swedish Covenant Church is of special interest here because several
of its missionaries among the Kongo people, or Bakongo (sing. Mukongo),
produced many written accounts of the indigenous culture and reli-
gion. The best known and most prolific of these Swedish mission-
aries was Karl Laman. Established in 1885, the Congo Free State
favoured a policy of assimilation of the Congolese to the same civil
status as Europeans while, as of 1908, the Belgian Congo with its
policy of ‘indirect rule’ created a more plural society divided between
African and European sectors. Whereas Europeans controlled the
bureaucratic and industrial institutions, Africans provided unskilled
and semiskilled labour. In the 1960s the independent nations of Zaire
and Congo were established. Today the majority of the Kongo peo-
ple live in the province of Lower Congo in the Democratic Republic
of Congo (formerly Zaire and, before that, Belgian Congo). Smaller
groups live in the Republic of Congo (previously the French colony
of Moyen Congo) and in Cabinda, a detached part of Angola (for-
merly a Portuguese colony). Altogether the Kongo people may num-
ber more than four million.39

     See further, e.g., Axelson 1970: 203 ff.; Vansina 1975: 652 ff.; MacGaffey
1986: ix, 16, 40 f. Vansina 1975 is a useful, brief introduction to the history and
culture of the Kongo people.
36                              chapter one

   The Kongo speak various dialects of the KiKongo language.40
They are divided into a number of sub-groups or major clans.41
Much of the material used here refers to the Sundi and the Bwende.
Some sources and literature concern the Vili and Yombe in the
north-western part of the greater Kongo area. However, other sub-
groups have not been left out, and in much of the material it has
not been made clear which groups are presented. This vagueness is
certainly a problem. Moreover, the Kongo have intermarried and
largely interacted with and been influenced by neighbouring peo-
ples.42 Yet many scholars, including the above-mentioned missionary
ethnographer Laman and contemporary scholars like the sociologist
T.K.M. Buakasa and the anthropologist Wyatt MacGaffey, argue
that there is a basic uniformity or common cultural core that unites
the different sub-groups.43 MacGaffey, furthermore, stresses the sim-
ilarities between the Kongo society and a host of neighbouring Bantu
societies in Congo, Zaire and Angola.44
   The Kongo live in both forest and savannah areas. The tradi-
tional economy is based mainly on agriculture, while hunting and
fishing play a subsidiary role. In addition to hunting and fishing,
men are responsible for slashing and burning, various crafts and
tending orchards of palm and fruit trees. Hoe cultivation is prac-
tised by women who work hard in the fields but also do the cook-
ing and raise the children. Root crops like manioc, groundnuts,
maize, bananas and vegetables are important food crops. As domes-
tic animals, goats, sheep, fowl and dogs may be found. Pigs and tse-
tse-resistant cattle also provide some protein. Important examples of
craft specializations are weaving, basketry, pottery and, in particu-
lar, blacksmithing.45
   Unlike the other peoples studied here, the Kongo have a matri-
lineal social system. Thus, lineages comprise matrilineally traced
descendants of a common ancestress. Lineages have jural rights and

     The etymology and original meaning of the word ‘Kongo’ is uncertain (Widman
1979: 29 f. Cf. MacGaffey 1986: 23).
     See, e.g., Janzen 1978: 12; Widman 1979: 24 ff.; Dalmalm 1985: 55, 58.
     Vansina 1975: 651; Dalmalm 1985: 17.
     Buakasa 1973: 11; MacGaffey 1986: 22 f. See also, e.g., Janzen 1972: 12;
Jacobson-Widding 1979: 23.
     MacGaffey 1986: 23.
     Vansina 1975: 649, 661 ff.; Janzen 1978: 13; Mahaniah 1979: 212; Jacobson-
Widding 1979: 23 f.
                         ethnographic background                                 37

obligations. As there are sub-clans, there are also sub-lineages. Leading
members of a lineage may live on or near its land, but villages can
have members of several descent communities. Today many of the
Kongo live in urban areas. It has been estimated, for instance, that
almost half the population of Kinshasa are Bakongo.46 Dalmalm has
stressed the ‘hierarchical’ aspects of Kongo society and the significance
of seniority for authority and power.47 MacGaffey, however, has also
emphasized certain ‘egalitarian’ aspects and the fact that cadets sooner
or later may become elders. He remarks that it is from a male per-
spective that Kongo society can be seen as egalitarian, with no per-
sistent social stratification or concentration of power.48
   A significant hierarchical element of the Kongo tradition is the
former existence of sacred kingship and chiefship. In addition to their
politically leading roles, kings and chiefs had important religious func-
tions. The centralized state was dismantled in 1665 by the Portuguese.
Then autonomous chiefdoms existed until the end of the nineteenth
century when new colonial states were established. During the colo-
nial era there were still chiefs, but they gradually lost many of their
old functions.49 What has remained, and still largely remains, is the
basic system of Kongo government, the ad hoc interlineal commit-
tees of elders who co-operate to regulate marriage, divorce, funer-
als and disputes. Yet, ‘the image of chiefship still haunts this utterly
decentralized system’, and ‘people talk as though there were hierar-
chies and chiefs’.50

      Mahaniah 1980: 591.
      Dalmalm 1985: 59.
      MacGaffey (1986: 24–34) discusses these issues in a historical perspective and
focuses on the situation of the nineteeth century. ‘Because menstruation, regarded
as polluting, was specifically incompatible with most titles and priestly functions,
women only acceded to important roles in the process of social reproduction’ (ibid.,
30 f.). See further, e.g., Buakasa 1973: 12 ff.; Vansina 1975: 664 f.; Jacobson-
Widding 1979: 28 ff.; Janzen 1979: 16, 20 ff.; Mahaniah 1980: 591.
      For a recent specialised study of Kongo political culture, focusing on mater-
ial from the early twentieth century, see MacGaffey 2000.
      MacGaffey 1986: 101. See further, e.g., Vansina 1975: 669; Mahaniah 1979:
210, 214, 219; Jacobson-Widding 1979: 85, 92; MacGaffey 1986: 17, 80, 102, 195
ff.; MacGaffey 2000: 224 ff.
38                             chapter one


Of the five peoples studied in this book, the Yoruba-speaking peo-
ple are by far the most populous. Reliable recent statistics are lack-
ing, but the estimates vary from some millions to about twenty million
or even more. There is no clear-cut answer to the question of who
is a Yoruba. As remarked by Daryll Forde, among others, the Yoruba
unity is a unity of language rather than of culture.51 In terms of cul-
ture, there is much diversity. The Yoruba dialects are a cluster of
the Kwa group of languages, and the Oyo dialect is the model for
the written language, or ‘standard Yoruba’.52
   The Yoruba-speaking groups of people live primarily in south-
western Nigeria, but some Yoruba speakers are found in other West
African countries, including Benin and Togo. Certain similarities
between Yoruba culture and some classical cultures, like the Egyptian
one, caused some early scholars to ponder over an eastern origin,53
and some migrant legends point to the east as well. There are immi-
grant groups that have come from the east and intermarried with
the Yoruba. However, cultural as well as physical and linguistic fea-
tures indicate that the neighbours in the forest belt of West Africa
are the most closely related peoples. According to Yoruba myths,
the city of Ife is the cradle of the Yoruba people. Among the Yoruba,
urban residence is a pre-colonial heritage, and the urban centres still
trace their origin to Ife, which is regarded as the centre of Yoruba
culture and religion. Myths of origin tell, further, that the Yoruba
once formed a single political entity. Hence there is a mythical con-
cept of political as well as cultural unity. There is no firm histori-
cal evidence, however, that the Yoruba ever formed a single political
unit. In pre-colonial time the Yoruba referred to themselves by using
sub-group names like Oyo and Egba.54

      Forde 1951: 1.
      Yoruba is a tonal language. However, the tones—high, middle and low—will
not be rendered here. See, e.g., Abraham 1958: x.
      E.g., Dennett 1910; Lucas 1948.
      The comprehensive term ‘Yoruba’, which may be derived from a Hausa/Fulbe
nickname meaning ‘cunning’, was adopted by missionaries of the Anglican Church
Missionary Society and is now used by the Yoruba themselves. See, e.g., Forde
1951: 1 and Bascom 1969a: 5.
                        ethnographic background                                 39

   The indigenous Yoruba economy is based on sedentary hoe farm-
ing, craft specialization and trade. Cassava, yams, maize and bananas
are main food crops, and cocoa is a particularly important cash crop.
Farming is primarily men’s work, while many women engage in
trade, which is a well developed system with a network of markets.
Although people live in cities, they have farms, which surround cities.
Secondary dwellings may be built on the farms, however, particu-
larly if these are far away from the primary dwellings in the cities.
In the highly complex Yoruba economy, specialized crafts include,
among other things, weaving, dyeing, ironworking, brasscasting, wood-
carving and leatherworking. The techniques of various crafts are
known only to small groups of professionals and are often protected
as secrets by religious sanctions. The Yoruba are perhaps best known
to the world for their art: woodcarvings; the portrait heads known
as ‘Ife bronzes’ in particular have made Yoruba artists world famous.
   The Yoruba kinship system is in practice strongly patrilineal, and
rank or hierarchy is based on seniority among men. Clans and sub-
clans are thus headed by the oldest male members. Clan members
are considered blood relatives, and marriages between them are for-
bidden. Facial scarification is a way of marking clan membership.
Although clans are important social and legal entities, the Yoruba
do not have a narrowly kinship-oriented social structure. Like the
economy, the social pattern is very complex and varied. Several
groups and societies cut across kinship lines. For instance, there are
secret societies, clubs for young people and separate societies for
women and men. As is shown in chapter six, groupings may also
be based on religious criteria.
   Politically, the traditions of sacred kingship have been important,
and in certain ways they still are, although kings as well as chiefs
have lost much of their political influence.55 In his presentation of
the indigenous political structure of Ife, W. Bascom points out that,
previously, the king was isolated in the palace, appearing in public
only once a year.56 New kings from the royal patrilineal clan were
selected by town and palace chiefs. Beneath these were other chiefs
in a hierarchic system of intermediaries. Urban centres of various

      The current significance of kingship traditions is exemplified, for example, in
Olupona’s study of the Ondo Yoruba (Olupona 1991). For another important case
study of Yoruba sacred kingship, see Pemberton & Afolayan 1996.
      Bascom 1969a: 29 ff.
40                                 chapter one

Yoruba groups still have a king as supreme authority or at least as
a symbol of such authority. Contemporary Yorubaland is a complex
mix of old and new. Under British colonial rule, until 1960, the
Yoruba were subjected to the system of indirect rule, which meant
that the indigenous Yoruba structure remained largely intact. Since
the British hindered immigration of white settlers into this part of
Africa, the Yoruba did not come under strong influence from a large
alien population. Today the great majority of the Yoruba are Muslims
and Christians. However, indigenous Yoruba traditions, such as those
associated with the system of sacred kingship, are still quite impor-
tant even to Muslims and Christians. In particular, such traditions
survive among Sufi Muslims and members of the indigenous African
churches. The Yoruba have, in general, a reputation for being suc-
cessful in terms of co-operating across (formal) religious boundaries,
and it is not very unusual to find Yoruba families in which some
members are Muslims, others are Christians and yet others are ad-
herents of the indigenous Yoruba religion. In the current Nigerian
situation, characterized partly by strong religious tensions, the long-
standing Yoruba tradition of co-operation and ‘sharing’ across reli-
gious borders seems to mitigate the effects of ‘fundamentalist’ Muslim
and Christian inroads in Yorubaland.57

     For this ethnographic survey of the Yoruba culture I have drawn on a num-
ber of sources, some of which may be mentioned here. In the series Ethnographic
Survey of Africa, Forde published his short study The Yoruba-Speaking Peoples of South-
Western Nigeria in 1951. A more detailed, and still very useful, survey is Bascom
1969a. In German, Schultz-Weidner’s chapter ‘Die Ostatlantische Provinz’ (Schultz-
Weidner 1979) in the second part of Die Völker Afrikas und ihre traditionellen Kulturen
(H. Baumann, ed.), includes a presentation of the Yoruba culture. See also, e.g.,
Eades 1980 and Beier 1999.
                                 CHAPTER TWO


                            The sky is the dwelling place of all the divine beings
                            and spirits in !Kung belief.1
                            Severe internal pain and sickness which make one
                            feel genuinely ill all over . . . are called ‘sickness of
                            the sky’.2


The religion of the Kung and other San differs from most, if not
all, other indigenous African religions in that, as a rule, it does not
include any spirits of nature. Hence spiritual beings or divinities are
not associated with trees, hills, rivers and other parts of the earth.3
Nor are the ancestors believed to be inside the earth, as is the case
in some religions of agriculturalist peoples. The abode of the deities
and spirits of the dead is above the sky that holds the sun, moon
and stars. When referring to this abode, I will here use the term
‘heaven’. Among the Nyae Nyae Kung the category of diseases that
are caused by heavenly beings is designated ‘sickness of the sky’ (kwi
naa). Such heavenly diseases can manifest themselves in grave inter-
nal ailments of which people are aware, but they can also exist in
a person without that person knowing it. This religious category of
illness is distinguished from the other category, which includes mild,
localized ailments, visible on the surface of the body, common aches
and minor injuries.4
    Before elaborating on the heavenly beings and their significance
as agents of disease, some general remarks on San religion should
be made. In addition to the significant religious differences between
various groups of San, referred to in the Introduction, many scholars

      Marshall 1986: 171.
      Marshall 1969: 370 f.
      Cf. Lebzelter 1928: 409.
      Marshall 1969: 370 f.
42                               chapter two

have stressed the importance of individual variations within different
groups.5 There are remarkable variations concerning religious con-
ceptions as well as with regard to intensity of belief. Similarly, ver-
sions of mythological and other tales may vary greatly.6 In one of
his studies on Nharo San, Guenther has concluded:
     One fundamental feature about Nharo belief, and one [that is] shared
     with the belief patterns of Bushmen in general, is its multifarious,
     inchoate and amorphous quality. There is wide variation in the accounts
     provided by different Nharo individuals when they describe the appear-
     ance and qualities of a supernatural agent. Nharo supernaturalism
     seems to be a confusing tangle of ideas and beliefs, marked by
     contradictions, inconsistencies, vagueness and lack of culture-wide
One of the reasons for the individual, amorphous and fluid charac-
ter of San religion is the absence of priests or other religious func-
tionaries rendering systematic and intelligible the complex of beliefs
and transmitting these ‘packaged’ beliefs to other individuals and
generations.8 Another reason is the lack of any formal religious edu-
cation of San children, who are socialized primarily within multi-
age peer groups rather than by adults, thus developing their own
idiosyncratic notions about religion. A third factor is the condition
of mobility which isolates individuals and groups from one another
over months or even years, resulting in localized religious traditions
and the diversification of the cognitive world. Moreover, it must not
be forgotten that most San groups have lived in more or less close
contact with neighbouring peoples, which has contributed to a shat-
tering of their religious ideas. Paradoxically, however, such influences,
in combination with decreasing mobility, new forms of education
and other factors, have also led to a reversed process of increasing
religious coherence and unity, which will be exemplified later in this
chapter. Even though Christian missionary work among the San has

    For some examples, see Dornan 1925: 162; Bleek 1928: 25 f.; Lebzelter 1934:
56; Gusinde 1965: 38; Barnard 1979: 72; Barnard 1988: 229 f.; Tanaka 1980: 110;
Lewis-Williams 1981: 77; Katz 1982: 29; Guenther 1999: 58 ff.; Smith et al. 2000: 65.
    See further, e.g., Lebzelter 1934: 24; Vedder 1937: 420 f.; Guenther 1979: 106;
Marshall 1986: 172; Fourie 1994: 9.
    Guenther 1986: 216.
                      heavenly beings among the san                                 43

met with great obstacles, the influence of Christian ideas on San
religion is another factor that must not be overlooked.
   In studies of this religion, the highly abstract concept of ‘religion’
and the tendency in much scholarly work to present African religions
as well-structured, logical, coherent and unified become particularly
problematic. Certainly, some elements of San religion are less varied
and more agreed-upon than others, and my presentation will
concentrate on these. I will also point out at least some of the
variations concerning such central elements. Nevertheless, it should
be borne in mind that much of the variety and richness of San
religion will necessarily remain concealed. Another general charac-
teristic of this religion is that it has little bearing upon economic and
social life. Several scholars have stressed the extraordinary paucity
of communal rituals, which contrasts vividly with many overgener-
alized presentations of African religions as being particularly rich in
terms of such rituals. Guenther even says that paucity of ritual is
‘the remarkable feature about traditional Bushman religion’ (my italics).9
As a rule, initiation rituals and trance dances are the only elaborate
communal rituals.10 The former type of ritual will not be studied
here, but the trance dance will be dealt with extensively in a special
section, since the religious agents of disease figure very prominently
in it. These agents are, firstly, God and a lesser deity or, in Thurner’s
(1986) terminology, das transzendente Hauptwesen and das transzendente
Nebenwesen. Secondly, there are the spirits of dead humans. God, the
lesser deity and the spirits of the dead are the ‘heavenly beings’ who
will now be presented in some detail.

                                   God the Creator

Among the Kung there are several names for God. In the Nyae
Nyae region, Marshall recorded seven divine names,11 some of which
(for example, Hishe, Huwe and Gauwa) have been reported from other

      Guenther 1979: 111. See further, e.g., Dornan 1925: 162; Bleek 1928: 27;
Heinz 1975: 27; Guenther 1986: 251 f.
      Initiation of girls is of particular importance. In a study of Nharo, Bleek (1927)
assumes that the whole initiation ceremony for boys, which is less elaborate, has
been borrowed from Bantu neighbours long ago. See also Guenther 1999: 167,
182; Smith et al. 2000: 85.
      Marshall 1962: 223 ff.
44                                chapter two

areas too,12 and one earthly name (Gao na, Old Gao), which may also
be the name of a male human being. Marshall says that God gave
himself his names to praise himself.13 In his account of the Dobe
Kung, Lee reports that God is called Gangwanana (cf. Gauwa above),
which he translates ‘big big god’, and that there are other names
too.14 Concerning Nharo, Barnard says that God is usually referred
to as Nadiba, which otherwise means ‘sky’.15 Guenther uses the name
Neri, which also indicates some association with the sky.16 ‘Nharo
language poetically reflects this association: the words for sky and
cloud are Neri ki (“Neri face”) and Neri oo (“Neri hair”) respectively.’
Other groups of San have other names for God.17 A number of
these, apparently, are the result of influence from other African reli-
gions as well as from Christianity.18 Not surprisingly, Father Schmidt
held that the most common ones were those that could be trans-
lated ‘Lord’ (Herr) and that, albeit less frequent than the name ‘Lord’,
the name ‘Father’ (Vater) was important and ‘altbuschmännisch’.19
Even though Schmidt may have laid too much stress on the significance
of such names, other sources make it clear that they do exist.
Regarding the Nharo, for instance, God is sometimes called xuba,
which means ‘lord’ or ‘master’ in a religious as well as in a secular

      E.g., Dornan 1925: 148 ff.; Schapera 1930: 182 f.; Gusinde 1963/65: 37.
      Marshall 1962: 223. Barnard (1988: 223) suggests that the names mentioned
by Marshall may be regarded as attributes of God. However, the origin of some of
the names was completely unknown to Marshall’s informants.
      Lee 1984: 107.
      Barnard 1988: 221. Cf. Bleek 1928: 25.
      Guenther 1986: 219.
      For a survey of some of these, see Schapera 1930: 177–195. One much dis-
cussed example, outside the scope of this study, is the name Kaggen of the Xam in
South Africa. This is also the name of the mantis, an insect that feeds on other
insects and clasps its prey in forelimbs held up as if in prayer; and some early
scholars drew the erroneous conclusion that San identified God with this insect.
For some contributions to this discussion, see Stow 1910: 131 ff.; Bleek 1929: 305 ff.;
Schapera 1930: 177 ff.; Gusinde 1965: 37, 39; Holm 1965: 45 ff.; Lewis-Williams
1981: 119. Schmidt (1973) is a more specialized study of this issue. Although the
discussions have concerned the Cape San, in particular, that is not the only group
with a name for God that is also the name for mantis. With regard to the Kung,
who are of special concern here, see Lee 1984: 107. Cf. Marshall 1962: 222.
      E.g., Schapera 1930: 190; Gusinde 1963/65: 37.
      Schmidt 1933: 685. See also Lebzelter 1934: 54 f.
                     heavenly beings among the san                               45

sense;20 and Kung often call him father, especially in supplication.21
In a sense, human beings may be regarded as his ‘children’. There
are examples, also, from some San groups, among others Kung and
Nharo, of the belief that God has a divine wife and divine chil-
dren.22 These figure primarily in myths, however, and will not be
studied in this work, which deals with the living religious context.23
   It is not easy to describe the nature of God among San. Informants
and scholars differ a great deal in their accounts of his attributes.
Virtually all seem to agree, however, that he is the creator and the
highest being.24 In a comprehensive study, based on early material,
Schapera concludes that God is ‘the creator, the source of the
beneficent rain and of good luck in hunting and food gathering, and
the protector of his people from illness and danger’.25 These princi-
pal conceptions, which Schapera felt were common to most of the
San of his time, have been reported by later scholars too. In par-
ticular, the ideas of God as a giver of good things have been stressed
by researchers influenced by Schmidt and his theory of Urmonoteismus.26
The conclusions drawn by Schmidt and his disciples have been sup-
ported in a more recent work by Tanaka, who thus says that God
represents all that is good; for instance, he provides food, causes
rain, makes plants grow and saves people from starvation and drought.27

      Barnard 1988: 221.
      Marshall 1962: 227, 246. With the exception of the androgynous Kxyani of
Kxoe (Köhler 1978/79: 20 ff.), it seems that God is anthropomorphically depicted
as ‘male’ among all groups of San (Thurner 1983: 391).
      E.g., Schapera 1930: 177; Marshall 1962: 225 f.; Barnard 1988: 221.
      The reports about God’s wife and children have been questioned by Schmidt
(1929: 297) and Gusinde (1963/65: 37; 1965: 41), who regard San religion as
‘monotheistic’ (Schmidt 1933: 684; Gusinde 1965: 40). According to Gusinde, the
notions of wife and children are the result of foreign influence. Cf. Lebzelter 1928:
      On God as the creator, see Dornan 1925: 149; Schmidt 1933: 686 f.; Wilhelm
1954: 162; Gusinde 1963/65: 38; Heinz 1975: 20; Köhler 1978/79: 20 f.; Tanaka
1980: 110.
      Shapera 1930: 195. See also Vedder 1912: 414; Schebesta 1923/24: 115;
Dornan 1925: 149; Fourie 1928: 104. Cf. Seyffert 1913: 204 f.
      Lebzelter 1928: 407; Lebzelter 1934: 54 ff.; Gusinde 1963/65: 37 f.; Gusinde
1965: 40 f. Schmidt himself (1933: 686 f.) concluded: ‘Dass das Höchste Wesen ein
gutes Wesen sei, ist die allgemeine Aussage aller Stämme und aller Quellen. Von ihm
kommt, wenigstens in letzter Linie, alles Gute, das die Menschen erfahren, und er
sendet nur Gutes.’
      Tanaka 1980: 110 f.
46                               chapter two

   In more recent works the good and helpful qualities of God are
stressed too, although not as strongly and consistently as in the stud-
ies by Schmidt and others. Rather, God’s nature is depicted as more
or less ambivalent. In this book, the word ‘ambivalent’ is normally
used to refer to the morally mixed nature of deities and spirits, des-
ignating them as both constructive and destructive, both good and
evil, or, in other words, as spiritual beings in non-dualistic religions.
Marshall and Lee indicate that ambivalence may be an important
characteristic of God among Kung, particularly as he is portrayed
in myths.28 Among Nharo, however, this feature seems to be less
pronounced. Guenther says that good is attributed to him,29 whereas
Barnard points out that his goodness, or his moral ambiguity, varies
with the degree to which the concept of God assumes identity with
the other, more evil, elements of the spiritual world.30 Based on the
analyses of her impressive amount of material from various periods
of time, Thurner has drawn the conclusion that God is ambivalent
but that the good aspects are predominant.31 Her conclusion coin-
cides with my own interpretation of the material available to me.32
Transcendental attributes such as ‘omniscience’, ‘omnipotence’ and
‘omnipresence’ appear primarily in late sources,33 which may indi-
cate some changes over time. Even in studies by early scholars of
the ‘Vienna school’—Schmidt, Schebesta and Lebzelter—such attrib-
utes appear sparsely, if they appear at all. With regard to the Kung,
Marshall says that, in contrast to former beliefs, they now (1962)
believe that God is all-powerful.34 Scholars of San disagree on the
issue whether, or to what extent, the ‘traditional’ concepts of God
have been influenced by Christianity, but that question need not be
discussed in detail here.35

     Marshall 1962: 228–238; Lee 1984: 107. See also, e.g., Fourie 1994: 6.
     Guenther 1979: 106.
     Barnard 1988: 224.
     Thurner (1983: 391) says, regarding the character of God, ‘dass er häufiger
ausschliesslich gut als gut und böse, das heisst willkürlich und unberechenbar, oder
indifferent ist’.
     Cf. further, e.g., Bleek 1928: 25; Schapera 1930: 185; Köhler 1978/79: 20 ff.;
Katz 1982: 30; Guenther 1986: 218 f.
     E.g., Silberbauer 1972: 319; Thurner 1983: 382, 386, 391 f.
     Marshall 1962: 234. See also, e.g., Köhler 1978/79: 23 f.
     Hirschberg (1975: 393), for example, argues that such influences cannot be
excluded, while Thurner (1983: 392) presents the opposite conclusion.
                     heavenly beings among the san                              47

   While offerings or sacrifices to God are rare or non-existent, prayers
are more or less common.36 Offerings have been mentioned by
Schebesta, Schmidt and Gusinde,37 but the great majority of schol-
ars either flatly deny that any offerings are made,38 or simply do not
say anything about it. As a rule, prayers are said by individuals in
a non-formalized way. When praying, they ask or thank God for
various things such as success in the hunt and recovery from seri-
ous illnesses.39 In relation to the Kung, Marshall reports that they
may all express their concerns, anxieties, hopes and thankfulness at
any time or place and without any special postures.40 God is not far
away or indifferent to humanity. Rather, he has become increasingly
involved in the lives of Kung. Among the Nharo, God appears to
be somewhat more remote.41 Guenther even speaks of him as an
‘otiose’ being, although he mentions that people occasionally do pray
to him.42 Earlier accounts of Nharo seem to indicate that God was
less remote or more frequently addressed in prayer.43 However, the
evidence is too slender and the variations between different Nharo
individuals too great to warrant any far-reaching conclusions in this

                 The Lesser Deity and the Spirits of the Dead

As mentioned above, Gauwa is one of the names for God among
Kung in the Nyae Nyae region. Paradoxically, this is also the name
of the lesser deity, who shares all the divine names of God except
the earthly one, Gao Na. Among Kung, ‘one shares in some mystical

     Here it is, again, important to call attention to differences between various
groups and individuals of San. Cf., e.g., Heinz’s account of Xo (1975: 22, 36), who
pray frequently, and Köhler’s study of the Kxoe area (Köhler 1978/79: 21), where
prayers are very seldom said.
     Schebesta 1923/24: 115; Schmidt 1933: 688; Gusinde 1963/65: 38 ff.; Gusinde
1965: 40.
     E.g., Vedder 1912: 414; Heinz 1975: 27; Köhler 1971: 324.
     For some accounts of prayers among San, see Schebesta 1923/24: 123; Bleek
1928: 25; Lebzelter 1928: 407, 412; Schapera 1930: 183; Schmidt 1933: 687;
Lebzelter 1934: 11; Gusinde 1963/65: 38 ff.; Gusinde 1965: 40; Heinz 1975: 22;
Köhler 1978/79: 21; Thurner 1983: 392; Heinz 1986: 31.
     Marshall 1962: 244, 246 f.
     Barnard 1988: 224.
     Guenther 1986: 65, 214, 219.
     Bleek 1928: 25; Schapera 1930: 183.
48                               chapter two

way the entity of the person for whom one is named’.44 The word
Gauwa is used by Nharo too.45 The Dobe Kung call the lesser deity
Gangwa matse (small Gangwa) to distinguish him from the big Gangwa.
Barnard holds that among Khoisan peoples this name is almost uni-
versally the term used for the lesser deity or for the evil aspect of
God.46 It is, furthermore, used to designate the spirits of the dead.
Among the Nyae Nyae Kung these are, thus, called gauwasi.47 The
Dobe Kung call them gangwasi,48 and among the Nharo they may
be referred to as Gauwani.49
   Hence the names suggest that the three categories of heavenly
beings are closely connected with each other. Some of the statements
about God and the lesser deity seem to indicate that they were once
one single being.50 Indeed, some San do not think of Gauwa as a
separate being.51 The ‘two beings’ are sometimes regarded as two
aspects of the same divine being.52 On the other hand, there are
studies that depict a dualist or almost dualist system. One example
is the aforementioned account by Tanaka, according to which the
lesser deity represents everything that is evil or bad and, conse-
quently, is blamed for all kinds of misfortune.53 More than half a
century earlier, Dornan referred to him as ‘the evil spirit’ or ‘demon’.54
Several scholars have applied the term Satan or Devil.55 In some
cases informants influenced by or knowledgeable about Christianity
have used these terms in talks with the scholars concerned.56

      Marshall 1962: 225.
      Guenther 1979: 106.
      Barnard 1988: 226.
      Marshall 1962: 241.
      Lee 1984: 107.
      Guenther 1979: 110. Nharo use the term kwe gau gau too (ibid.; cf. Guenther
1986: 218). See further, e.g., Schapera 1930: 186 ff., 193.
      Concerning the Xam, Lewis-Williams (1981: 122) contends that there was pre-
viously one ‘capricious’ being only.
      E.g., Bleek 1928: 26; Barnard 1979: 72; Lee 1984: 107.
      Marshall 1962: 238; Thurner 1983: 393; Barnard 1988: 226. Cf. the case of
the Kxoe, who believe in one divine being, Kxyani, who has a male and a female
aspect (Köhler 1978/79: 20 ff.).
      Tanaka 1980: 110 f.
      Dornan 1925: 150. Similar statements about this deity as an evil or bad being
have been made by, among others, Fourie (1928: 104) and Wilhelm (1954: 163).
      E.g., Bleek 1928: 26; Schapera 1930: 188; Guenther 1986: 224. Cf. Fourie
1994: 6.
      Lebzelter 1928: 408, 410; Schapera 1930: 188; Guenther 1986: 224. Cf. Lee
1984: 107.
                   heavenly beings among the san                          49

   However, in most cases such a translation seems misleading. Even
though there is enough evidence to conclude that, to some San, he
was or is conceived of as entirely evil, he has more frequently been
thought of as an ambivalent being.57 An interesting hypothesis was
presented by Schmidt, who interpreted the lesser deity as the prog-
enitor of humanity.58 As a representative of humanity he cannot be
entirely evil. Such a conception of him must, thus, be due to for-
eign influence. Like other humans created by God, he is subordi-
nated to the creator, and there is no worship of him. More recent
sources support the description of the lesser deity as subordinated to
God.59 Among the Nharo, as described by Guenther,60 some regard
the former as the son of God, while others see him as God’s ser-
vant. Marshall holds that Kung increasingly think of the lesser deity
as dependent of God.61 Being subservient to God, he should carry
out orders. Yet he preserves considerable independence and insti-
gates his own affairs. Even though the emphasis is on evil-doing,
good deeds can be expected of him. The paradox of subservience
and independence has been observed among, for instance, Nharo
too. Among them, he is sometimes described as a jealous rival of
God, a very unpredictable and capricious being.62 Nharo are one of
the groups with myths that depict the lesser deity as a trickster or
culture-hero. For instance, Nharo may say that he brought fire but
also knowledge of death.63
   Although he may be associated with the earth, he is more often
conceived of as a heavenly being. His heavenly dwelling is separate
from that of God. According to Nharo, as well as to Ko, God’s
dwelling is in a higher region than that of the lower deity.64 Kung
associate the former with the east, whereas the latter is said to live
in the western region, where the sun sets.65 As was briefly mentioned

     See further Lebzelter 1928: 407; Schmidt 1933: 689; Marshall 1962: 239;
Heinz 1975: 22; Guenther 1979: 108; Thurner 1983: 394; Guenther 1986: 219,
     Schmidt 1933: 689 f.; Vedder 1937: 431.
     Heinz 1975: 23; Thurner 1983: 393.
     Guenther 1979: 106.
     Marshall 1962: 239.
     Guenther 1986: 222 f., 248.
     Guenther 1979: 108 f.; Thurner 1983: 395. Cf. Gusinde 1965: 41. See also
Smith et al. 2000: 80.
     Guenther 1979: 107; Heinz 1975: 22. Cf. Bleek 1928: 26.
     Marshall 1962: 233, 240.
50                              chapter two

above, there are stories about a wife and children of God. With his
interpretation of the lower deity as the progenitor of humanity in
view, Schmidt finds it puzzling that there is no first ancestress at his
side.66 He felt that this could be due to insufficient knowledge of this
being at that time. In fact, more recent works show that there are
stories, primarily of a mythic character and significance, about a wife
and children of the lower deity.67 Modern sources, furthermore, sup-
port Schmidt’s note that there is usually no worship of the lesser
deity. The paucity of examples concerning prayers to this being is
in marked contrast to the numerous examples of such worship of
God. Prayers to the lesser deity seem to occur primarily among
groups, such as some Kung of southern Angola and northern Namibia,
where he is regarded as a master of animals. In most cases, how-
ever, it is God who has features of a master of animals, although
this concept seems to be much less important among San than among
hunting-gathering peoples of northern Eurasia.68
   In order to shed further light on the human characteristics of the
lesser deity, it is essential to consider his relationship to the spirits
of the dead. Both his name and certain attributes appear to suggest
that he is closely allied to these spirits. When a death has occurred,
the dead person has usually been buried in a contracted position,
like a foetus or a person asleep, without any elaborate ceremonies
and rites of mourning.69 Concerning Kung, for instance, Marshall
says that burial has no effect upon the status in the afterlife.70 Thus,
whether dead Kung are buried properly bound in deep round graves,
or scratched into shallow trenches, or not buried at all and eaten
by beasts, makes no difference in that respect. Nor does the deceased
person’s manner of living have any bearing on his or her where-
abouts in the afterlife. Only a very few sources report that ‘good’
people go to God, whereas ‘bad’ people go to the lesser deity.71 In
the case of the farm Nharo studied by Guenther, ideas about a good
and a bad place for the dead, and the importance of a person’s

     Schmidt 1933: 690.
     Marshall 1962: 226 f.; Heinz 1975: 22; Katz 1982: 29; Guenther 1986: 221.
     Thurner 1983: 386 f.
     For some information about funerals, see Passarge 1907: 110 ff.; Vedder 1912:
413; Bleek 1928: 35; Schapera 1930: 160–166; Woodburn 1982: 199 f.; Guenther
1986: 281 ff.; Heinz 1986.
     Marshall 1962: 243.
     E.g., Heinz: 24. Cf. Fourie 1928: 104.
                     heavenly beings among the san                              51

deeds while living, a Christian influence is clearly discernible.72 Such
influence is, however, unlikely in the case of the Ko San. The Ko
studied by Heinz had no idea or concept of Christian beliefs.73 In
comparison to hunting-gathering Nharo, farm Nharo have slightly
more elaborate mortuary practices too.74
    Guenther has the impression that the belief that there is no exis-
tence whatsoever when you die, a belief still held by some Nharo,
is ‘more truly Nharo’.75 Other sources on Nharo, however, seem to
contradict this impression.76 It should be noted, also, that the great
majority of reports on San conceptions of the afterlife, including the
early ones, give evidence of belief in a continued existence.77 For
instance, ‘the Nyae Nyae Kung believe strongly and vividly in the
existence of spirits of the dead, the gauwasi, who live immortal lives
in the sky with Gao na’. Spirits are like air and cannot normally be
seen by living humans. God gives the spirits of the dead everlasting
life.78 Although they grow older, he rejuvenates them before they are
very old.79 In some early sources it was erroneously concluded that
San worship heavenly bodies, particularly the moon.80 Certainly, the
moon is important in San mythology and symbolism. For example,
a very widespread myth about the origin of death says that the moon
wanted human beings to die and return again, as he does himself.
The hare was instructed to deliver this message but distorted it,
telling human beings that they would die and not return again, and
ever since death has existed on earth.81 As pointed out by Barnard,
however, there is no evidence that San did, or do, worship the

      Guenther 1979: 110; Thurner 1983: 327.
      Heinz 1975: 33.
      Guenther 1986: 282. Cf. further Schapera 1930: 168.
      Guenther 1986: 246 f.
      E.g., Bleek 1928: 26; Barnard 1979: 71.
      Seyffert 1913: 201; Bleek 1927: 54; Schapera 1930: 171; Vedder 1937: 431;
Wilhelm 1954: 162; Gusinde 1965: 40; Heinz 1975: 23; Hirschberg 1975: 394;
Köhler 1978/79: 24.
      Marshall 1962: 241. Cf. Guenther’s account (1986: 241 ff.) of Nharo concep-
tions of spirits or souls. This issue will be further studied below.
      Marshall 1962: 241 ff. Cf. Lebzelter 1934: 6 f., 11, 13 f. The quality of per-
petual life is one of the reasons why God is associated with the moon, which is
said to have the capacity of rejuvenation or death and rebirth too.
      E.g., Bleek 1927: 305; Schapera 1930: 172 ff.
      Schapera 1930: 160; Schmidt 1933: 693; Guenther 1986: 245 f.
      Barnard 1988: 220.
52                              chapter two

   In the case of the Nyae Nyae Kung, the spirits of the dead are
taken, by spirits of humans who have died earlier, first to the west-
ern part of heaven, where the lesser deity lives, before being carried
further to the east, to the place where God lives. Apparently, most
sources indicate that the spirits of the dead have their dwelling near
or with God.83 Yet they are by no means bound to a particular loca-
tion but move around frequently, on earth as in heaven.84 Although
they are usually invisible, they may occasionally be seen, particularly
by healers.85 Like God, the spirits of the dead are often depicted as
anthropomorphic beings.86 They may, thus, be envisaged in much
the same way as living humans.87 Among the Kung and Nharo, for
example, the spirits of the dead eat and live together as spouses.88
   The character of the spirits of the dead is ambivalent or ambigu-
ous. As in the case of the lesser deity, however, the negative or evil
traits are the predominant ones, and they are said to be most active
at night.89 Hence it is not surprising that reports about a certain
amount of fear of the spirits recur frequently in the sources.90 Although
they may act independently, they appear more often to be regarded
as servants of God and the lesser deity.91 The fact that they are mes-
sengers or minor characters may mitigate the fear felt by human
beings, and people speak less reluctantly and more frequently about
the spirits than about God and the lesser deity.92 Among the Nharo,
in particular, the conceptions of spirits are more diverse than my
brief account can possibly demonstrate. Barnard writes about a special

      E.g., Lebzelter 1928: 407; Vedder 1937: 431; Köhler 1978/79: 24.
      Schapera 1930: 193; Marshall 1962: 242; Heinz 1975: 23. Cf. the studies of
Barnard (1979: 71 f.) and Guenther (1986: 245) on Nharo. According to Marshall
(1962: 243), those Kung who have committed suicide live with the lesser deity in
the west.
      E.g., Bleek 1927: 54; Marshall 1962: 242; Barnard 1979: 71.
      The mythic form of the lesser deity seems more often to be theriomorphic
(Thurner 1983: 391).
      Lebzelter 1928: 109; Vedder 1937: 431.
      Marshall 1962: 243; Barnard 1979: 71. Unlike Nharo spirits of the dead, as
reported by Barnard, Kung spirits of the dead do not beget children, according to
Marshall. They are, thus, believed to remain children in the world of spirits. Cf.
further Heinz 1975: 23.
      Bleek 1928: 26; Schapera 1930: 193; Marshall 1962: 244; Barnard 1979: 71;
Lee 1984: 107; Barnard 1988: 226.
      See, e.g., Vedder 1912: 413; Seyffert 1913: 201; Bleek 1928: 26; Wilhelm
1954: 161; Heinz 1975: 231; Barnard 1979: 72.
      Marshall 1962: 241; Köhler 1971: 321; Köhler 1978/79: 24.
      Marshall 1962: 242, 244; Heinz 1975: 23.
                    heavenly beings among the san                              53

category called Ka-je-m Go-dzi (Old Fathers), who were the original
Nharo to whom God gave their land and culture, and who are not
feared but may be regarded as ‘ancestors’ in the good sense of the
word.93 This may be compared to Guenther’s account of the kwe gau
gau, whom some of his Nharo informants regarded as ‘ghosts of
recently deceased people’, staying around the graves of their dead
bodies for a while, before continuing their journey to God, while
others referred to them simply as ‘Gauwa’s things’.94 These spirits
or ghosts are usually quite harmless.95
   As in the case of the lesser deity, there is normally no organized
cult or veneration of the spirits of the dead. Accordingly, the term
‘ancestors’ is avoided here. Older as well as more recent studies show
clearly that prayers or offerings to the deceased are exceptional.96
Among others, Schapera concluded that the San did not have any
organized ‘family or tribal ancestor worship’ or any form of religious
practice in which the spirits of the dead were regularly invoked or
propitiated.97 Similarly, Marshall and Barnard have concluded, con-
cerning Kung and Nharo respectively, that the concept of having
special relations with their own ancestors are lacking.98 Marshall adds
that the spirits of the dead, who have their own supplies of food
and implements, do not want anything from living humans.99 Hence
there is no point in offering them anything.

                  The Heavenly Beings as Agents of Disease

Heavenly beings figure prominently in the important trance dance,
which has been described and discussed in many different sources
from various periods of time. Even rock paintings indicate that this
dance, which has also been labelled, for example, ‘medicine dance’
or ‘therapeutic dance’, was an essential element in San religion many
thousands of years ago.100 There are certain differences between
      Barnard 1979: 72.
      Guenther 1986: 218 f.
      Cf. further Köhler 1978/79: 25.
      See, e.g., Stow 1910: 133; Seyffert 1913: 211; Dornan 1925: 148; Köhler 1971:
324 f.; Heinz 1975: 24. Cf. Schmidt 1933: 691; Holm 1965: 137; Heinz 1986: 25;
Fourie 1994: 7. No doubt, Passarge (1907: 107) was badly informed on this issue.
      Schapera 1930: 171, 395.
      Marshall 1962: 241; Barnard 1979: 72.
      Marshall 1962: 243.
      Vinnicombe 1976: 310 ff.; Lewis-Williams 1981: chapter 7.
54                               chapter two

various groups of San in terms of performance and contents of the
trance dance. Yet the similarities are more conspicuous, and there
seem to have been few changes over time. As noted by Lewis-
Williams, there is even a marked similarity between southern San
dances, evidenced in rock paintings, and current dances performed
by Kung and other San groups.101 As a background to the presen-
tation of the heavenly beings as agents of disease, I summarize briefly,
and with special reference to Kung and Nharo, the major features
of the trance dance.102
   A trance dance usually starts at dusk and frequently goes on until
dawn the following day. The central purposes are prevention and
healing of serious afflictions, but there may be important elements
of recreation, entertainment and enjoyment too.103 Moreover, such
a dance is an event when healers gather and transmit to other partici-
pants important information about how things are in the spiritual world,
and how people in this world would do best to relate to them. Dances
are not necessarily restricted to members of one particular band, nor
is participation based on kinship ties. The number of participants and
spectators, sometimes including Bantu people or other ‘outsiders’, may
range from a few dozen, or even less, to a few hundred.104
   The trance dance takes place around a fire, either inside or outside
the encampment. It can be planned but may also begin spontaneously,
usually by women who start singing and clapping. Most women and
girls sit at the fire, very close together, while the men who join in,
and occasionally a few women, dance in either direction around them.
After some time of intensified dancing the trance comes to some of
the dancers, and then they start treating the other participants.105

       Ibid., 76.
       It should be mentioned that, in addition to the healing in this communal
dance, serious diseases may also be treated in dances or rituals for individuals,
although these are less significant and less frequent (e.g., Bleek 1928: 29; Heinz
1975: 28; Barnard 1979: 72 f.).
       E.g., Marshall 1962: 248; Lee 1967: 33; Guenther 1986: 253. See also Gall
2001: 240.
       Katz (1982: 38) reports about the Kung that small dances, with people from
one camp only, involved about 15–20 persons, whereas the large ones, which were
not restricted to people from one camp, could attract up to 200. Among farm
Nharo, according to Guenther (1975: 162), up to 300 spectators may watch a trance
dance. See also, e.g., Marshall 1969: 353 ff.
       For more detailed accounts of the trance dance, see especially Lee 1967; Lee
1968; Marshall 1969; Guenther 1975; Guenther 1975/76; Barnard 1979; Katz 1982;
Guenther 1986: 253–263; Guenther 1999: chapter 8. Among older sources, I have
                     heavenly beings among the san                                55

   A key concept in connection with the trance dance is the concept
of num (among Kung) or tsso (among Nharo).106 Num (or tsso) has
been described as a ‘spiritual energy’107 or ‘potent therapeutic sub-
stance’,108 which normally resides in the stomach. Kung hold that it
was created by God, who gave it his own power.109 He does not
wield or command it, but he can stop it from working any time he
wishes. Not only human beings have num; it exists in many other
things too. Num is always strong, and sometimes so strong that it is
dangerous. If ordinary mortals come too close to God, for instance,
his num would automatically kill them. ‘It seemed to me that num
had several attributes similar to those of electricity. Like electricity,
num is powerful and invisible, capable of beneficent effects, but highly
dangerous if too strong.’110 In the trance dance, num is activated, and
when it is ‘boiling’, it rises up the spine to the head.111 This ‘boil-
ing energy’ (kia), to use the title of the book by Katz,112 is a strong
healing power, which exudes from the body in the form of sweat.113
   The state of trance is induced by the repetitive dancing, the singing
and the fire.114 Hence num resides not only in the dancers or heal-
ers but also in the songs and in the dance fire. The songs, which
are named after ‘strong’ things such as certain animals, usually have
sounds rather than ordinary words and complex melodies.115 Kung,
among others, believe that num songs are created by God and revealed
in dreams and visions.116 Like the songs, the dance fire augments

found Lebzelter (1934: 48 ff.), Mogg (n.d.: 6 f.) and Gusinde (MS) particularly inter-
esting. Mogg’s study was probably published in the mid-1930s. Detailed descrip-
tions of the trance dance are found in Gusinde’s field notes from, e.g., 8 October
1950. See also Fourie 1994: 8; Smith et al. 2000: 78 ff.
       Cf., e.g., the terms gaoxa (Marshall 1962: 248) and tco (Köhler 1971: 318 f.).
       Katz 1982: 34.
       Guenther 1975: 162.
       E.g., Marshall 1962: 248; Katz 1982: 92.
       Marshall 1969: 350 f.
       Katz 1982: 41. See also Lee 1984: 109.
       Katz 1982.
       See further, e.g., Barnard 1979: 75; Lee 1968: 44; Lee 1984: 109; Shostak
1981: chapter 13; Katz 1982: 42, passim; Guenther 1986: 243 f.; Smith et al. 2000:
       Lee 1968: 36; Katz 1982: 94. Psychologically, physical exertion and food
deprivation may be important factors for understanding the inducement of trance
(Guenther 1975/76: 49; Katz 1982: 94).
       Barnard 1979: 74; Lee 1984: 111; Lewis-Williams 1981: 83.
       Marshall 1969: 366 f.; Katz 1982: 123.
56                            chapter two

the concentration of num at a trance dance.117 Unlike many shamans
in South America, for instance, Kung and Nharo healers normally
do not make use of intoxicants.118 Katz reports that, in rare cases,
a drug may be given to some Kung with difficulties in experienc-
ing trance.119 Likewise, there are few, if any, special paraphernalia,
although the use of rattles, wound around the legs of the dancers,
and sticks have been reported.120
   In the state of trance, healers may perform feats like touching the
fire, eating burning coals, seeing at a great distance and obtaining
special information. They claim that if their own num is hot enough,
they will not be burnt by the fire.121 At least in the past, some pow-
erful healers were believed to be able to transform themselves into
animals, lions for example.122 When healers are in deep trance, a
comatose or sometimes unconscious state, and said to be ‘half dead’
or ‘dead’, their souls are believed to leave their bodies to travel to
the world of spiritual beings. This ‘shamanistic’ feature of out-of-
body travel is an important element in San religion.123
   One important reason for a healer’s travel in extra-corporeal form
is the diagnosis of ‘soul loss’, which is another striking similarity
between San religion and religions among hunting-gathering peoples
in North America and northern Eurasia. Taking the souls or spirits
of living humans away is one of the various means used by the spir-
itual beings to bring diseases and other types of misfortune; and the
trance dance is the occasion when they are most likely to bring
afflictions. Since it normally occurs at night, when the spiritual beings
are very alert, they are attracted by the event.124
   In order to grasp the idea of soul loss, an examination of San
conceptions of ‘soul’ or ‘spirit’ is needed. Both of these terms have
been used in the literature on San to designate a part of human
beings that is normally considered to be invisible and immortal. Here
the slightly different ideas of Kung and Nharo will be compared.

      Marshall 1969: 357; Lee 1968: 44.
      Marshall 1969: 372; Guenther 1986: 269 f.
      Katz 1976: 293.
      E.g., Marshall 1969: 358. Cf. Lewis-Williams 1981: 78.
      Heinz 1975: 29; Barnard 1979: 75; Katz 1982: 100.
      Dornan 1925: 152; Holm 1965: 87 f.; Lee 1967: 34 f.
      For some more details, see Lebzelter 1934: 54; Marshall 1962: 250; Holm
1965: 86; Lewis-Williams 1981: 77; Shostak 1981: 292 f.; Katz 1982: 43, 100;
Guenther 1986: 271; Fourie 1994: 8; Smith et al. 2000: 80.
      Marshall 1962: 244; Marshall 1969: 350; Katz 1982: 43, 112 f.
                      heavenly beings among the san                   57

When a Kung person dies, spirits of the dead come to take that
person’s ‘spirit’ ( gauwa)—this is the term used by, among others,
Marshall—by pulling it through the head of the corpse. Likewise, it
is through their heads that the spirits of the healers in trance leave
their bodies temporarily to encounter the spiritual being, who lurks
in the shadows around the dance fire, and it is through the heads
that the spirits return. The spirit that is taken from the corpse when
a person dies is distinct from ‘life’ (toa), which is put inside the body
of a person, or animal, and held there by God. It exists in the vital
organs of the body, including the head, but not in the arms and
legs. Whereas wounds in the former parts can be lethal, somebody
may even lose a limb yet survive. Eventually ‘life’ dies in the body
and stays there, but the spirit does not die.125
   This clear distinction between ‘spirit’ and ‘life’ does not seem to
exist among the Nharo studied by Guenther.126 According to his
report, the spirit or ‘soul’ (i )—Guenther uses the latter term—is an
integral part of the body and imparts the body with life, thought
and feeling. The soul derives originally from God and becomes ‘the
quintessential substance of the body’. Diffused throughout the body
by the blood, the ‘soul substance’ is found in its strongest concen-
tration in the vital organs of the body. In particular, it is associated
with the heart and the brain. It is alert and active when the body
is awake as well as asleep, and it can temporarily leave the body.
For those Nharo who believe in an afterlife, the soul (i ) is the part
of the body that survives death and returns to God.127 Guenther
reports that tssa Neri (‘steal from Neri’) is a common cause of dis-
ease in the Nharo conceptual system. When God is inattentive, the
lesser deity may take a soul of a living human away from God’s
sphere of influence into his own, and that person will then become
sick. Tssa Neri is the explanation for cases of relapse, too; it is typ-
ical of the lesser deity to thwart God in this fashion, snatching the
soul of the sick person who had just then, by the good grace of
God, seemed to be recovering. One particular affliction caused by
the lesser deity’s contest with God is madness. In defiance of the
latter, the former will place a second soul into a person, which is

        Marshall 1962: 242. See also Fourie 1994: 7.
        Guenther 1986: 242 f.
        Ibid., 245.
58                              chapter two

at odds with the other soul, derived from God at the time of birth.
The madness is a result of the struggle between the two souls.128
   When the healers encounter the spiritual beings in the trance
dance, they often lack the respectful attitudes they normally have.
Kung, among others, do not argue with God and the lesser deity
in daily life and may fear to utter their ordinary names, using expres-
sions such as ‘the one in the east’ or ‘the one in the west’ instead.
The attitudes observed in trance dances can be more open and
straightforward. Although more cautious means like cajoling or plead-
ing with the spiritual beings occur in trance dances as well, the heal-
ers often become aggressive and swear at them. Sometimes burning
sticks are hurled at them and indecent invectives are used.129 When
a healer’s soul leaves his or her body to try to rescue the lost soul
of a seriously sick person, the healer enters into a very close con-
tact with the spiritual world, and the struggle becomes intense.
According to Katz, ‘this struggle is at the heart of the healer’s art
and power’.130
   Besides soul loss, another major reason for serious illnesses is intru-
sion. In that case some object, originating from a spiritual being, is
believed to have entered the body of a person. Apparently, the intrud-
ing object is most often thought of as a miniature arrow.131 From
the Kung area, for instance, Marshall and Shostak, among others,
report invisible arrows shot by the spirits of the dead.132 Guenther,
on the other hand, says that while some Nharo conceive of the
arrows, which in this case are shot by the lesser deity, as invisible
and immaterial, others say that they are concrete objects.133 Whether
invisible or not, such arrows or other things that have made people
seriously ill must be removed by the healers, who can achieve that
by, for instance, sucking or drawing them out.134

       Ibid., 223 f.
       E.g., Marshall 1962: 227 f.; Marshall 1969: 248; Guenther 1986: 271.
       Katz 1982: 43. On their returns from out-of-body travels, the healers may
describe the heavenly realm, sometimes in great detail, and recount their fight for
the sick one’s soul. Cf. Guenther (1986: 261) who says that Nharo healers do not
share their experiences after their travels in trance.
       Cf., e.g., the case of the Kxoe area, where the spiritual beings use spears
(Köhler 1962: 319).
       Marshall 1962: 244; Shostak 1981: 291.
       Guenther 1986: 272.
       E.g., Marshall 1969: 369; Katz 1982: 105. In the late 1930s, Vedder (1937:
429) reported that some healers could suck out, for example, some thorn or piece
of bone, take it out of the mouth, show it to people around and then throw it into
                     heavenly beings among the san                                59

   Important as these methods may be, diseases and deaths caused
by spiritual beings can be brought about in other ways too. Among
Kung, for example, they can influence wild animals to attack peo-
ple, and they can be the cause of a person’s falling from a tree.
Lightning can be used by the spirits of the dead to kill people when
God and the lesser deity so command.135 Like lightning, a whirlwind
is a ‘death thing’ in Kung belief. A whirlwind is called gauwa a,
which means ‘Gauwa smell’. The spirits of the dead come in whirl-
winds too, but neither they nor the lesser deity can be seen or heard
in these winds. People who encounter such winds must try to get
away, lest they will sicken or die.136 This association of whirlwinds
with the lesser deity is also found among Nharo.137
   Sometimes ‘sickness’ seems to be a very abstract notion. A per-
son may or may not be aware of having sickness.138 Yet even those
who are unaware of being sick are treated by the healers in the
trance dance. A healer in trance may ‘absorb’ a person’s affliction
into his or her own body and let it run its full course there. Although
this can be painful for the healer, his or her ‘hot’ state of trance
makes it possible to endure and survive.139 With regard to Kung,
Katz has drawn the following conclusion:
     Sickness is more an existential condition or level of being than a par-
     ticular illness or symptom. Everybody has some sickness, and so every-
     body who is at a dance is given healing. In most persons, this sickness
     remains incipient, neither serious nor manifested in symptoms. In some
     persons, the sickness is actualized into what Westerners would call an
     illness. Persons who are ill get especially intensive and extensive heal-
     ing. Num is for prevention as well as treatment.140

the fire. One of the Kung healers interviewed by Katz (1982: 82) said that the ‘lit-
tle things’ he removed from a man with a sick stomach were like twigs.
       Marshall 1962: 244. Cf. Vedder 1912: 413.
       Marshall 1962: 239.
       Guenther (1986: 220) translates a ‘breath’. According to Bleek (1928: 26), the
wind is called by the name of the lesser deity ‘when it is strong and hurls’. Cf.
Schapera 1930: 194; Barnard 1988: 226. In her essay on San in Angola, Bleek
(1927: 54) hypothesized that God and the lesser deity ‘verkörperte Naturkräfte sind,
ersterer wohl der Wald oder das Wachsen des Waldes, letzterer Wind und Regen’.
       In the latter case the term ‘disease’ would perhaps be somewhat inappropri-
ate, since it seems to imply an awareness of not being at ease (dis-ease).
       Marshall 1962: 244; Marshall 1969: 349, 370 f.; Guenther 1975/76: 49.
       Katz 1982: 102.
60                                chapter two

There are no different words to distinguish ‘potential sickness’ from
sickness that is specific or manifested in symptoms. In both cases the
term xai is used. Hence the distinctions are created by the context
of usage.141
   As a rule, it does not seem possible to differentiate clearly between
the spiritual beings as agents of disease and death, although some
general tendencies may be discerned. Among the Kung, it seems
that the spirits of the dead are the primary direct causes of serious
afflictions. Yet it must be remembered that they are often com-
manded by the more powerful beings. God is seldom a direct cause
of disease, but he may send diseases and death through the lesser
deity and the spirits.142 Marshall even states that his being the cause
of sickness and death was a reason for his giving himself many
names.143 Being all-powerful, he is ultimately responsible for death.
According to Marshall, the lesser deity was previously, when the idea
of God’s omnipotence was less significant, thought of as the death-
giver.144 He is now conceived of more as a messenger, although, like
the spirits of the dead, he can act independently as well. If they are
in a benevolent mood, they may be placated by the healers in the
trance dance. That is unusual, however, and the healers must then
try to drive them away instead. Sometimes they succeed in doing
so, but sometimes they lose the battles with these spiritual beings.
God is confronted by the most powerful healers only, whose souls
travel to his home, at the greatest risk to their very lives.145 According
to Guenther, the spiritual agent most responsible for disease among
Nharo is the lesser deity.146 Yet some diseases may stem from God
too. For instance, he can send deformed children to a mother. God
is associated with death too. When death is sent by God, however,
it is a ‘good death’, that is, one that takes a person’s life quietly,
without pain, and at an advanced age. A death caused by the lesser
deity is, instead, a ‘bad death’, that is, a death of pain, spasms, kick-
ing and crying. On the whole, he is here more of an envious oppo-
nent of God than a messenger. Paradoxically, however, the lesser

        Ibid., 53.
        Lee 1967: 134; Katz 1982: 102; Woodburn 1982: 200.
        Marshall 1962: 223.
        Ibid., 234.
        Ibid., 250 f.; Marshall 1969: 350; Katz 1982: 43, 105.
        Guenther 1975: 162; Guenther 1986: 271.
                      heavenly beings among the san                               61

deity also heals illnesses, and holds the medicinal plants of the veld.
Most Nharo healers, therefore, attribute their powers of trance healing
to this deity.147
   Among the Ko, for instance, the lesser deity seems to be the most
common agent of disease too. God, on the other hand, is more a
source of healing than of illness and death. If a person who has
been very sick recovers, it is an indication that there has been a
regular conflict between God and the lesser deity and that God has
prevailed. Yet, as among Nharo described by Guenther, the lesser
deity of the Ko is not an entirely bad being. Thus he is believed to
give medicines to healers and to intercede, occasionally, with God
for a person’s life.148 With regard to the Kxoe, Köhler says that dis-
eases and deaths are associated primarily with the female aspect
(weibliche Gott) of the androgynous God, Kxyani. The ‘male God’
(männliche Gott) usually does not send illnesses. As messengers of the
‘female God’, the Herrin der Krankheit, the spirits of the dead can cause
deaths, but when they act independently they occasion illnesses only.149
   Why, then, do the heavenly beings cause diseases and deaths? The
basic San answer to this question appears to be some reference to
their ambivalent and unpredictable nature, which was described
above. In works by scholars of the Vienna School, the moral aspect
of the San thoughts about the reasons for diseases and deaths has
clearly been exaggerated.150 Lee mentions that one of his informants
once argued that the major agents of disease among the Kung, that
is the spirits of the dead, did not bother those persons who behaved
themselves, but he later reversed himself.151 Apparently, the idea of
afflictions as punishment for misbehaviour was not a predominant
one among those Kung studied by Lee. From the Nyae Nyae Kung
area, Marshall reports that the concept of ‘sin’ as an offence against
spiritual beings is vague. One person’s wrongdoing against another

       Guenther 1986: 222 ff., 253. Among the Central Nharo studied by Barnard
(1979: 71 f., 75), the significance of the spirits of the dead as agents of illness and
death is more strongly emphasized. A few of their healers claim to derive their
healing power from God or the lesser deity. More often, however, these healers
associate this power with the spirits of the dead.
       Heinz 1975: 22 f., 34, 36.
       Köhler 1971: 321; Köhler 1978: 37; Köhler 1978/79: 21, 34.
       See especially Gusinde (1963/65: 38), who states that God punishes evil per-
sons by serious diseases and early deaths. Cf. Lebzelter 1934: 11.
       Lee 1984: 107 f.
62                               chapter two

is corrected or avenged within their social context and not by God,
who punishes people for his own reasons, which can be quite obscure.
Burning bees, for instance, is something that displeases him inten-
sively. Since he is very fond of bees and honey, he may send sick-
ness and death upon people who burn. The somewhat unpredictable
nature of God is exemplified by the story about the hunter who died
after eating the meat of a gemsbok he had killed, because God had
changed himself into that particular gemsbok. The informant who
told Marshall this story said that God might regret having done such
a thing—killing a man who was just hunting to feed his children—
and, to make amends, he might particularly favour that man’s son.152
   Shostak provides a Kung example of how the spirits of the dead,
as it were, may punish the punished. A spirit can exercise his or
her power against the living if, for example, a certain woman is not
being treated well by her husband and other people. In such a case
the spirit may conclude that no one cares whether she remains alive
and, for that reason, ‘take her to the sky’.153 Moreover, the spirits
of the dead may bring sickness simply because they are longing for
the living. The process of death can, thus, be seen as a struggle
between two loving sets of relatives, one living and the other dead,
each wanting the individual for themselves.154
   Regarding Nharo, Guenther says that the reason why God some-
times sends deformed children to women is to determine if they will
take care of them; and he is believed to be pleased if they decide
to keep such children.155 One reason why the lesser deity, among
Nharo, may become offended, is the presence of such non-San things
as shiny metallic objects, for example, knives and flashlights, and
tobacco. Therefore, such things are concealed at a trance dance.156
Moreover, the ‘fairly physiological’ Nharo notion of ‘soul’ (i ) is a
part of the explanation why some people become sick and die more
easily than others. For instance, the soul is ‘smaller’ in children than
in adults, and it is ‘smaller’ in women than in men.157 Like Kung
and other San women, Nharo women can become healers, but most

        Marshall 1962: 245.
        Shostak 1981: 291.
        Lee 1984: 109.
        Guenther 1986: 242. This applies to twins, also. Cf. Marshall 1962: 245.
        Guenther 1975/76: 49.
        Guenther 1986: 242. See also, e.g., Drennan 1937: 248; Tanaka 1980: 115.
                    heavenly beings among the san                             63

healers, and particularly the most renowned and powerful ones, are
men.158 Among Kung, for instance, the search for the power called
num is of greater importance to men than to women;159 and in the
state of trance or ‘half-death’ the men, temporarily and paradoxi-
cally, become particularly exposed and vulnerable.160
   Even though a variety of San explanations to the problem why
the heavenly beings cause diseases and deaths have been rendered
in the sources, there is clearly no systematic or coherent ‘theologi-
cal answer’ to that question. Like other aspects of San religion, this
issue is characterized by flexibility and fluidity. Besides, Lee’s con-
clusion that the Kung do not spend much time in pondering such
religious or philosophical problems seems to be applicable to other
groups of San as well.161 In the above account of the spiritual beings
as agents of disease and death only a few references have been made
to the older material. One reason for that is simply that the older
sources do not contain much information about the religious or cul-
tural aspects of San medicine.162 Another reason is that as far as
they do provide such information, it does not, as a rule, differ
significantly from the accounts presented in the more recent sources.163
Like the performance of the trance dance, the religious content of
this dance has, by and large, been invariable throughout the period
of time considered here. To the extent that more significant changes
have occurred, these have largely resulted from increased influences
from non-San ideas and practices.164

      Shostak 1981: 298; Guenther 1986: 264.
      Katz 1982: 44.
      Marshall 1962: 250 f.; Katz 1982: 43, 105.
      Lee 1984: 109.
      E.g., Nolte 1986; Lübbert 1901.
      Cf. further, e.g., Dornan 1925: 150; Schapera 1930: 188, 194; Schmidt 1933:
689; Lebzelter 1934: 44, 46; Wilhelm 1954: 73.
      The results of these influences will be discussed in chapter 9.
                              CHAPTER THREE

                    GOD IN MAASAI THOUGHT

                          An Geister glauben die Maasai nicht.1
                          God, who created the world and human beings,
                          who gives and maintains life, is also the ultimate
                          cause of disease and death.2


Many of the scholars who have studied the Maasai religion have
emphasized its ‘uniqueness’ in the African context. In his classic work
from 1910, Merker even held that the difference between the Maasai
and neighbouring peoples is greater in the religious field than in any
other field.3 Fully four decades later, Reusch depicted the Maasai as
‘ausgesprochene Monotheisten’ (typical monotheists),4 and more
recently, H.-E. Hauge has repeatedly claimed that their religion is
‘one of the purest monotheistic religions of the world’.5
   Not only the theocentric and monotheistic features of pastoral
Maasai religion but also, for instance, the lack of such elements as
sacred buildings, religious images, priesthood, spirits of ancestors and,
as a rule, even of belief in life beyond have been apprehended as
evidence of the unique character of this religion.6 Certainly, it seems
that some of these characteristics, such as the monotheism and the
common disbelief in the hereafter, are particularly pronounced among
pastoral Maasai. By and large, however, their religion is similar to
the form of religion that is found among other, especially non-Muslim,

      Fuchs 1910: 114.
      Århem 1989: 80.
      ‘Nirgends zeigt sich bei einem Vergleich der Ethnographie der Masai mit der-
jenigen der ihnen benachbarten, um sie herum wohnenden Völker eine so tiefe
Kluft, wie auf dem Feld der religiösen Anschaung.’
      Reusch 1946: n.p.
      Hauge 1979: 42, 59. See also, e.g., Merker 1904: 196. Cf. Schmidt 1940: 424
f.; Olsson 1982b: 9; Berg-Schlosser 1984: 170.
      E.g., Olsson 1982b: 1; Olsson 1989: 238.
66                               chapter three

pastoral peoples in Africa.7 Apparently, it is only when it is com-
pared to religions of non-pastoralist, particularly agricultural, peoples
that the religion of the pastoral Maasai stands out as exceptional.
   As in the case of San religion, it is, furthermore, important to
take into account the individual and flexible traits of Maasai reli-
gion. While some Maasai believe staunchly in God, others are not
religious believers at all, and there is a wide range of differing kinds
of religiosity between such extremes. More than a century ago, Fischer
stated in a travel report that the religious practices of the Maasai
were in no way prominent;8 and about a hundred years later Berg-
Schlosser concluded that there is a low emphasis on religion in
Maasai culture. Among many Maasai, he said, ‘a kind of skeptical
“agnosticism” seems to prevail’.9 Jacobs reports that even among ‘tra-
ditional elders’ he found many agnostics. Some even questioned the
existence and power of God altogether.10 Further, there is a great
deal of variation, for example, in the telling of myths and other sto-
ries.11 As stressed by Voshaar, ‘THE story and THE version do not
exist’.12 Stories and beliefs are not codified or expressed in system-
atic treatises.13 There are no controlling theological authorities or
priests who have formulated such treatises. An outsider’s creation of
a ‘systematic theology’ would, thus, distort the reality of Maasai
   In addition to the problem of individual variations, the possibility
of some basic differences between the various sections (ilosho) and
smaller parts of them should be kept in mind. As remarked by
Spencer, the tendency of many scholars to refer to ‘the Maasai’,
rather than to some specific localities, could account for many of

      For some comparative notes, see Hauge 1979: 160 f.; Kronenberg 1979: 160 f.;
Westerlund 1986b: 13 f.
      Fischer 1884: 72.
      Berg-Schlosser 1984: 170 f.
      Jacobs 1965: 323. Cf. Marari (1980: 13 f.) who says that elders sometimes cor-
rect, for instance, warriors who occasionally dispute the existence of God. Cf. also
the article by Johnston (1915: 483) who probably exaggerates the importance of
western influence. In a study of the Turkana, another people with a culture and
religion similar to that of the Maasai, Gulliver (1951: 229 ff., 251) concluded that
most Turkana did not resort a great deal to the religious side of life but had, rather,
an apathetic attitude to religion. He emphasized, also, that the degree of religious
commitment varied clearly from one individual to the other.
      E.g., Voshaar 1979: 44 f., 118 ff.; Marari 1980: 28.
      Voshaar 1979: 107.
      Ibid., 118.
                           god in maasai thought                                 67

the contradictions between different sources. While realizing that the
comparative perspective of this study, as well as the nature of most
of the sources, does not render any greater precision possible, I will
at least attempt to avoid speaking in overgeneralized terms about
‘the religion of the pastoral Maasai’.
   Maasai religion differs from San religion in that, among other
things, it is richer in ceremonies and rituals. Above all, there are
elaborate ceremonies connected with initiations within the age-group
system, such as the circumcision ceremony, emurata.14 The major age-
group ceremonies involve features like ritual head-shaving, continual
blessings, slaughter of animals, paintings of the face or body, singing,
dancing and feasting.15
   Another difference between Maasai religion and San religion is
that, in the former, leadership functions are more strongly domi-
nated by men. As will be exemplified later, elders are the most
prominent authorities in religious affairs. It seems that, as a rule,
women are ‘more religious’ than men in that they pray, or sing
songs of prayer, and make offerings more frequently.16 However, rit-
ual leadership functions and the right to make important religious,
as well as political, decisions are vested in the elders.17

                                 God the Creator

The most important name, or the proper name of God, among the
Maasai is enkAi. The word enkai can be translated ‘heaven’, or ‘sky’,
and ‘rain’ too.18 As suggested by the name, God is associated particularly
with the heavenly realm, yet he is not identified with it. He may also

      For some information about the significance of circumcision, see Widenmann
1895: 302 f.; Merker 1904: 60 ff.; Hauge 1979: 13 ff.
      See further, e.g., Mol 1978: 41; Voshaar 1979: 86 ff.; Beckwith & Saitoti 1980:
30 f.; Århem 1985: 21 ff., 24 ff.; Ndoponoi 1986: passim.
      E.g., Blumer 1927: 77; Ndoponoi 1986: 19. Blumer (1927: 77) adds that, among
those Arusha whom he studied, prayers of men were usually shorter than those of
      Voshaar 1979: 50 f., 82 f.; Århem 1985: 25. The supranormal role of elders
and of healers, who are also believed to possess a special spiritual power and func-
tion as religious experts and leaders, will be studied particularly in chapter 7. On
the great significance of natural causation among Maasai, see the appendix.
      See, e.g., Decken 1871: 25; Fokken 1917: 240; Mol 1978: 75, 131, 145; Voshaar
1979: 109 f.; Olsson 1982b: 1 f.; Voshaar 1998: 131. Here I follow Olsson’s custom
of using enkAi for ‘God’ and enkai for ‘sky’ and ‘rain’.
68                              chapter three

be said to be omnipresent.19 Although it is very common, it is not
uncontroversial to refer to God as ‘he’.20 Voshaar and A. Hurskainen
hold that the feminine form enkAi can be understood in connection
with aspects of fertility, such as the significance of rain.21 Despite the
feminine form, however, ‘Maasai generally tend to think vaguely of
God as a “male” entity as far as they conceive of him in personal
categories at all. God is great and powerful, and these are qualities
that the Maasai consider distinctive of the male sex but never asso-
ciate with female beings.’22 Hauge reports that his informants always
spoke of God as ‘he’.23 Yet Maasai do not seem to be very inter-
ested in or concerned with this issue. Confronted with the problem
of gender, they may say that enkAi is only a ‘name’, although it is
the ‘basic one’.24 EnkAi is a spiritual being, and thus nonphysical,
and cannot be specified as ‘female’ or ‘male’.25
   Moreover, in addition to this name, there are several other ‘female’
and ‘male’ expressions for God.26 In the term Pasai the gender prefix
is omitted and substituted by the particle pa-, which is a common
element in a number of essentially religious terms of respectful address.
It can be translated ‘receiver of prayer’ or, literally, ‘the one who is
prayed to (beseeched, worshipped)’.27 With reference to the expres-
sions enkai narok (‘the black god’) and enkai nanyokie (‘the red god’),
some early scholars erroneously concluded that Maasai believed in
two gods.28 These colour designations seem to indicate the different
colours of the sky during the rainy seasons and the dry seasons.
While enkai narok is associated with dark clouds that bring rain, enkai
nanyokie refers to the red colour of the sky at sunrise and sunset

      E.g., Hauge 1979: 17 f.; Olsson 1982b: 10 f.
      Voshaar (1979) is exceptional in that he consistently writes ‘She’ instead.
      Voshaar 1979: 110 ff.; Hurskainen 1984: 175; Voshaar 1998: 133 f. This may
be compared, for instance, to Marari’s statement that Maasai relate God’s care for
his creation to a mother’s care for her children after birth. Marari 1980: 19 f. Cf.
Olsson 1989b: 4 f. On the issue of God as a provider of fertility, see also Magesa
1997: 88.
      Olsson 1982b: 3.
      Hauge 1979: 17.
      Olsson (1982b: 4) stresses that if we assume that the basic meaning of enkAi
is ‘sky’ or ‘rain’, i.e. something inanimate, then the question of the feminine gen-
der prefix in enkAi would turn out to be a pseudo-problem.
      See, e.g., Merker 1904: 196; Marari 1980: 14.
      E.g., Fokken 1917: 243; Kimerei 1973: 43; Marari 1980: 22 ff.
      Kimerei 1973: 43; Hauge 1979: 18; Olsson 1982b: 6 ff.
      E.g., Hollis 1905: 264; Burns 1908: 169: Dallas 1931: 40. Cf. Johnston 1915:
                           god in maasai thought                                  69

during the dry season.29 ‘The black god’ brings rain and, conse-
quently, grass to cattle and prosperity to people, whereas ‘the red
god’ may bring famine and death. The latter expression is also a
metaphoric designation for the warriors, whose bodies are painted
with red ochre and who are connected with aggression and death.30
   As symbolized by the expressions enkai narok and enkai nanyoki, God
is the creator and giver of everything, the master of life as well as
of death. He is the sustainer of the world as of all living things and
actively involved in daily life. What human beings need, he pro-
vides. Maasai are, above all, provided with cattle. Through the gift
of cattle, God has established a special relationship between himself
and the ‘people of cattle’, although he is God of all peoples.31 God
is like a father and mother of all creatures; and he ‘appears to have
many of the highly respected attributes associated with extreme age,
only more so’.32 Hence he is, for instance, not only wise but omni-
scient, not only powerful but almighty and not only old but eternal.
Moreover, he is described as good and merciful. Although he is
immanent and said to be everywhere, on earth as in heaven, he is
also transcendent and associated particularly with all things above.33
The association of God with heaven, heavenly bodies and celestial
phenomena does not preclude the depiction of him in personal and
anthropomorphic terms. For example, anthropomorphic elements are
common in mythical accounts of God. Among other things, he dwells
in the same camp as primeval man, he converses and mingles with
human beings, and he ascends to heaven via a rope. ‘This is, however,
seldom understood in exegetical conversational settings as literally
true. According to the transmitters themselves, the significance is that

       Voshaar 1979: 110; Marari 1980: 22. Cf. Hauge: 1979:18 and Voshaar 1998:
      Beckwith & Saitoti 1980: 26; Marari 1980: 22 ff.; Olsson 1982b: 14 ff.
      In a sense, Maasai have thought about themselves as the ‘chosen people’ of
God. See, e.g., Merker 1904: 196; Schmidt 1940: 324; Berg-Schlosser 1984: 172.
      Spencer 1988: 49. On the role of God as creator, sustainer and provider of
all things, and particularly of cattle, see further Burns 1908: 170; Fuchs 1910: 96;
Fokken 1917: 248; Blumer 1927: 76; Hauge 1979: 21; Voshaar 1979: 116, 125;
Marari 1980: 23, 42; Århem 1987: 22.
      See further, e.g., Merker 1904: 196; Berthold 1927:5; Blumer 1927: 75; Hauge
1979: 18; Voshaar 1979: 112; Beckwith & Saitoti 1980: 26; Marari 1980: 15 ff.;
Henschel 1983: 152; Landei n.d.: 21 ff.; Voshaar 1998: 139, 142. Cf. Spencer 1988:
49. Words like omniscient, omnipotent, omnipresent, immanent, transcendent, good
and merciful have been used particularly by Christian scholars. See further Westerlund
1985: 32.
70                             chapter three

God and humans stood in a closer relationship primevally than they
do presently.’34 What Maasai conclude from mythical accounts is,
thus, that the first human being had a direct experience of God and,
as a consequence, knew his nature.35
   Among pastoral Maasai and agricultural Arusha, prayers and
offerings to God are common. The important cult of God is sub-
stantiated by early as well as by more recent sources.36 For instance,
Hauge reports that people pray to God every day, usually in the
morning but often in the evening too. In connection with their
prayers, women frequently make libations of milk.37 Prayers may be
said and offerings be made by individuals, families or communally
by larger groups of people. The prayers of women are often in the
form of prayer songs, which may be sung in connection with dancing.38
Even though there are no fixed places for worship, the presence and
power of God is associated especially with certain places or things
such as trees and mountains. Trees are seen as one of the most powerful
manifestations of God.39 The tall oreteti tree, a fig tree, is of parti-
cular significance. Oreteti trees, like mountains, rise up towards
heaven and are thus, in a sense, close to God. They function as a
kind of medium by which prayers and offerings reach him. God, as it
were, hears prayers through such entities, as he hears through heav-
enly bodies.40
   In their prayers, Maasai ask God for the good things in life. Among
other things, they ask for offspring, health and a long life. Prayers
are also said in order to thank God for such things.41 According to
Voshaar,42 that which human beings can do and must do is not the
subject of prayer. Prayers and offerings are means of establishing
contact or communication between the worshippers and God. Sacrifices
may be offered for a variety of reasons such as to restore health, to

     Olsson 1982b: 31.
     Ibid., 51.
     See, e.g., Decken 1871: 26; Johnston 1886: 417; Fokken 1917: 243 ff.; Blumer
1927: 76; Le Roy 1928: 344; Schmidt 1940: 375 f.; Hauge 1979: 36 f.; Voshaar
1979: 126 ff.; Peron 1995: 52.
     Hauge 1979:19. See also, e.g., Fuchs 1910: 96; Voshaar 1979: 138.
     Fuchs 1910: 110; Hauge 1979: 23; Voshaar 1979: 138 ff., 245.
     E.g., Reusch 1953/54: 88; Jacobs 1965: 139; Århem 1989: 82; Magesa 1997:
     Olsson 1982b: 22–27; Olsson 1989: 241 f. See also Voshaar 1998: 163 f.
     Fokken 1917: 244, 246; Schmidt 1940: 390; Voshaar 1979: 133, 135.
     Ibid., 127.
                         god in maasai thought                             71

ward off evil, to return gifts to God as well as to ask for propitiation.
Moreover, sacrifices accompany all major religious ceremonies.43
Although animals are rarely slaughtered for food, there are thus a
number of occasions when they may be slaughtered for ritual

                            Other Spiritual Beings

Because of the monotheistic and theocentric character of Maasai reli-
gion, there is little or no room for other spiritual beings. Like Fuchs,45
who was quoted at the beginning of this chapter, other early authors,
in particular, emphasized the lack of such beings among Maasai.46
Merker’s biased ideas about ‘guardian angels’ were soon refuted in
Fokken’s better-informed study of Maasai religion.47 There is, however,
some evidence that various spiritual beings have played a more
significant role in more recent decades. As will be shown later, a
spread of the phenomenon of spirit possession, which is a new phe-
nomenon among Maasai, has recently occurred in Tanzania.
   In addition to these modern spirits of possession, who are foreign
spirits, there is also an indigenous Maasai concept of spirit, oloirirua
(pl. iloiriruani ), which refers to an evil type of spiritual being. Mol
uses not only the word ‘spirit’ but also the term ‘devil’ when he
translates this concept into English.48 This reflects a Christian usage
of the concept.49 There seems to exist a certain fear of the evil spirits
or iloiriruani.50 It should be stressed, however, that there is no cult
of them and that, by and large, they are of minor significance.51 In
folk stories some fearful ‘monsters’ called inkukuni appear frequently.
Like iloiriruani, they are not worshipped.52

     E.g., Schmidt 1940: 394; Hauge 1979: 35 ff.; Århem 1985: 10.
     For more information on sacrifices among Maasai, see Hauge 1979: 25–38;
Voshaar 1979: 218–252.
     Fuchs 1910: 114.
     E.g., Thomson 1885: 259; Johnston 1886: 417; Le Roy 1892: 144.
     Merker 1904: 196; Fokken 1917: 243.
     Mol 1978: 149. Cf. Hauge 1979: 41.
     Voshaar 1979: 216.
     E.g., Bleeker 1963: 119 f.; Hauge 1979: 41; Donovan 1982: 21.
     For some more information, see Johnston 1915: 481; Voshaar 1979: 216. Cf.
Merker 1904: 202; Hollis 1905: 265.
     E.g., Hauge 1979: 41 ff.; Kipury 1983: 19. Cf. Baumann 1894: 164.
72                             chapter three

   As mentioned above, pastoral Maasai do not normally believe in
an afterlife.53 Merker’s statements about a wonderful ‘paradise’ for
the souls of ‘good’ people and an empty desert for ‘evil’ ones,54 which
have been echoed by some other authors,55 seem to be a result of
Christian influence. None of Fokken’s numerous informants were
able to confirm these statements.56 ‘Dead people are dead, and there
is no theology about a life beyond, be it on high or down below.’57
The physical centre of a person’s life as well as ‘spirit’ or ‘soul’ is
the heart. The word oltau can be translated ‘heart’, ‘mind’, ‘spirit’
or ‘soul’.58 In a sense, thus, the spiritual dimension of human beings
is considered organic; and all spiritual faculties are believed to die
with the body.59 As a rule, pastoral Maasai have not buried their
dead. They have been left in the bush to be devoured by wild ani-
mals.60 The absence of proper graves reflects the lack of eschato-
logical ideas and the aversion to digging the soil.61 Århem adds that,
by carrying the corpses to the bush, people take them back to the
domain of God, from where they once came. The wilderness here
‘takes on a sacred connotation’; it is ‘the sacred space of God as
opposed to the secular and social space of the homestead’.62
   It should be remembered, however, that some burials do take
place, and that some people are believed to continue their existence
after they have died. Even early sources give evidence for the belief
that important or leading persons, particularly healers, iloibonok, enjoy
such an existence.63 Above all, those people are also the ones who
have been buried.64 Perhaps the fear or reluctance to mention names

      This has been observed by a great number of scholars. See, e.g., Thomson
1883/84: 259; Fuchs 1910: 113; Fokken 1914: 37; Leakey 1930: 206; Bleeker 1963:
118; Hauge 1979: 43 f.; Saibull & Carr 1981: 117.
      Merker 1904: 197.
      E.g., Weishaupt 1930: 224.
      Fokken 1917: 247.
      Voshaar 1979: 123.
      Mol 1978: 81.
      Olsson 1989: 240.
      Baumann 1894: 163; Fuchs 1910: 111; Hauge 1979: 54. See also, e.g., Thomson
1885:259; Hollis 1905:304; Burns 1908: 170. Cf. Decken 1871: 25.
      E.g., Laube 1986: 116.
      Århem 1989: 83.
      Hollis 1905: 308; Fuchs 1910: 114.
      Baumann 1894: 164; Schmidt 1940: 398. Cf. Decken (1871: 25) and Johnston
(1886: 416 f.) who write about the burial of small children too.
                          god in maasai thought                                 73

of dead ancestors may not only indicate a fear of death, or of recall-
ing sorrow, but also a more general fear of the dead as such.65
   Johnston held that, at the beginning of this century, the belief that
great persons could not entirely cease to exist was ‘a gradually grow-
ing belief ’;66 and there is evidence that this belief has continued to
grow in some areas. Commenting on the belief in annihilation,
reported by early authors, Hurskainen writes about the current sit-
uation: ‘Whatever is the truth of the beliefs of the 19th century
Maasai, the Parakuyu believe in the continuity of life after death.’67
Above all, such a belief as well as elaborate mortuary rituals and
burials have become prominent features of Arusha religion.
   In many ways the cult or veneration of ancestors among the Arusha
resembles that among neighbouring Bantu peoples such as the Meru
and the Chagga. More than half a century ago, Blumer stressed that
Arusha people were much more concerned with spirits of ancestors
than with God, although he added that it was difficult to say whether
these spirits were subordinated to or put on a par with God him-
self.68 According to Kimerei’s interpretation, ancestors are ‘interme-
diaries’ between human beings and God.69 There are shrines at the
graveyards where people pray and make offerings to their ances-
tors.70 Shrines are established at the graves of fathers; and other
ancestors are also approached through the dead fathers, who are the
most powerful and active of them all. Rituals at the shrines are con-
ducted by living fathers, and sons are utterly dependent on the rit-
ual prerogatives of their fathers.

                        Religion and Disease Causation

Since natural factors are the most important causes of illness for
Maasai, there is less to say about spiritual beings as agents of disease
and death in this chapter than elsewhere in this part of the book.

      Krapf 1857: 441; Voshaar 1979: 303; Berg-Schlosser 1984: 162. See also
Sindiga 1995: 101.
      Johnston 1915: 481 f.
      Hurskainen 1984: 180. Cf. Hauge 1979: 49, 51.
      Blumer 1927: 78.
      Kimerei 1973: 64. This is the typical interpretation of the role of ancestors
found in works by Christian scholars influenced by a theology of continuity. See
further Westerlund 1985: 34 ff.
      Kimerei 1973: 66.
74                             chapter three

As appears from one of the quotations at the beginning of this
chapter,71 however, God is the ultimate cause of illness and death.
This has been demonstrated in a number of other sources too.72
Galaty concludes that ‘God himself is thought to operate at levels
of ultimate principles behind all fundamental processes in life’, even
though ‘this monad of fundamentality does not adequately explain
the particularities of individual life events’.73 In particular, it seems
that serious diseases that may lead to death and death itself are fre-
quently believed to be occasioned by God. This has been reported
in works on pastoral Maasai as well as on Arusha.74
   A disease or death immediately caused by God is referred to by
the use of the term enkea. Such a disease or death is distinguished
from illnesses and deaths ultimately caused by God but mediated by
healers, iloibonok, who themselves derive their powers from God.75
Thus, enkea represents the unmediated and destructive intervention
in the lives of living humans. This may be compared, for instance,
to the view of birth as a result of a creative intervention by God.76
As implied in the concept enkea, disease is conceived of as a kind of
death. An enkea is a serious affliction which, in a sense, is a fore-
taste of death. Mild ailments like headache or stomach-ache, which
are not serious or mortal, may generally be referred to by the use
of another term, emoyan. If a person is weaker than is considered
normal, and this weakness is believed to be caused by blood loss, it
is regarded as something distinct from disease proper. Such weak-
ness is cured not by medicine but by the consumption of large
amounts of blood and fat. Notions of strength and weakness are usu-
ally described in the idiom of blood. Hence a strong person is ‘full
of blood’, whereas a weak person is said to ‘lack blood’.77
   Merker seems to indicate an element of predestination when he
says that a human being dies when the lifespan, ‘predetermined’ by
God, of that person has come to an end.78 To the best of my

      Århem 1989: 80.
      E.g., Hollis 1905: 334 f.; Peterson 1985: 175; Olsson 1989: 236.
      Galaty 1977: 320.
      E.g., Fokken 1914: 32; Marari 1980: 46; Landei 1982: 59. Galaty (1977: 321)
says, however, that any event may be explained by ‘the will or caprice of God’.
      The latter type of illnesses and deaths will be studied in chapter 7.
      Århem 1989: 80 f.
      Ibid., 84.
      Merker 1904: 196.
                            god in maasai thought                    75

knowledge, however, this statement has not been confirmed in other
sources. It may be compared, for example, to Fokken’s conclusion
that a human being does not die until God ends the course of life
of that individual.79 Merker himself says that, in the hope of recov-
ery, a dying person may sacrifice a bull to God.80 In a sense, how-
ever, death and disease are beyond human comprehension, since
nobody can predict the actions of God.81 When causing afflictions
and death, God may use natural elements such as lightning and
strong winds. If somebody is hit by lightning, people may say that
the ‘red God’ has hit that person.82 As is the case with San, among
whom destruction caused by whirlwinds is associated with the lesser
deity, Gauwa, the ‘red’ aspect of enkAi appears in destructive winds.
Earthquakes and the eruption of volcanoes are other means by which
God may make his presence known in a destructive way.83
   Olsson says that it is in certain types of discourse, like myth and
liturgy, that God is seen as the ultimate cause of events in cosmos,
nature, society and history. On such levels of discourse, or in such
speech situations, observable facts such as lightning and volcanoes
are thought of as being employed by God as ‘instruments’ or ‘meth-
ods’, impukonot (sing. empukonoto). In other words, particular events or
states, like ailments, can be discussed at different levels of discourse
involving an empirically observable cause or an unobservable and
intentional dimension.
   When Maasai discuss origins of diseases they are, thus, often heard
to apply arguments according to the following lines of thought. First
they try to find out what may be the empirically observable cause,
and if it is considered a sufficient cause they try to find some rem-
edy. Although the established cause may also be regarded as neces-
sary, it may be that in some ensuing speech situations it is no longer
regarded as sufficient. The empirically observed cause is then regarded
as guided by some unobservable entity, or, more precisely, guided
through a non-empirical or non-concrete dimension of it. The ‘cause’
is employed as an instrument by some ‘guiding cause’.84

       Fokken 1914: 32.
       Merker 1904: 192.
       Århem 1989: 80.
       Fokken 1917: 241.
       Marari 1980: 15 ff.
       Olsson 1989: 236.
76                               chapter three

   It is emphasized by Olsson that empirical and non-empirical causes
should not be seen as alternatives. As appears from the quotation
above, non-empirical causes are rather thought of as being comple-
mentary to empirical ones. This is in accordance with the situation
among other peoples studied here and, as remarked in the intro-
duction, the employment of terms such as ‘natural’ and ‘religious’
causes need not imply that these are alternatives.
   Voshaar holds that all objects that are used in daily life have a
meaning and efficaciousness beyond their immediate purpose. Hence
‘a club is not just a club’. Things like milk, grass and fire are sym-
bolically meaningful and efficacious of themselves and are used in
various rituals. Other things, for example tobacco containers and
walking sticks, are imbued with an extra meaning and power through
blessing and consecration by elders. For instance, after the birth of
a baby, objects for personal or family use are blessed by the spittle
of elders. In part, this is done in order to protect the mother as well
as the child from sickness and death.85
   Mythical Maasai versions of the origin of disease and death tell
that the first beings had to suffer these evils because they had dis-
obeyed God by eating the fruit of a certain tree.86 Disease and death
are henceforth implicit in the human condition. However, they may
also be interpreted as God’s punishment for the transgressions of
human beings here and now. According to Merker, elephantiasis is
caused by incest, and malformed or stillborn babies are also the
result of ‘sins’ punished by God.87 Much more recently, Spencer has
demonstrated that any deformity of a baby may, continuously, be
regarded as an indication of ‘some past sin dogging either parent’.88
   There is in all matters concerning birth and infancy, not only
among human beings but among all living things, an element of
providence expressed as the prerogative of God. To take any life in
embryonic form or at a young age is regarded as a sin, engoki.
Misfortune will surely follow such killing. For instance, persistent

      Voshaar 1979: 199 ff.; Voshaar 1998: 147.
      Merker 1904: 265; Galaty 1977: 475; Århem 1989: 79.
      Merker 1904: 51, 175.
      Spencer 1988: 42. Merker (1904: 51) held that malformed babies were undoubt-
edly killed immediately after being born. Spencer (1988: 42) says that Matapato
Maasai are reluctant to discuss family tragedies and dispose of dead children unob-
strusively. The general absence of congenital disabilities, he continues, ‘inevitably
raises questions regarding the fate of defective babies’.
                           god in maasai thought                         77

misfortune of some elders may occur because they have killed women
as warriors, not knowing that the women were pregnant.89 There is
a notion that an infant is a foetus, olkibiroto, as long as it is fed wholly
on the mother’s milk. It is not until milk from the herd is intro-
duced into its diet that the ‘foetus’ becomes a ‘person’, oltungani.
During the very first period of life, when the child is particularly
weak, dependent and vulnerable, it is also thought of as being espe-
cially close to God.90 There are other occasions, also, when the con-
cept of engoki, or sin, is applicable. On the whole, evil done when
life is ‘weakest’ is engoki. For example, elders and people in situa-
tions of rites de passage, that is, when a former state of life is left and
a new state has not yet been entered, are vulnerable; and evil done
to such categories of people may result in immediate sickness or
death of the perpetrator. The punishment may even affect such an
evildoer’s family and animals.91 As mentioned above, trees are seen
as particularly powerful manifestations of God. According to Hauge,
there are certain places where the trees are never cut down, or the
branches never broken, for fear of some misfortune. Maasai believe
that God will punish a person for such disrespect by sending some
kind of disease upon that person or upon his or her family or cattle.92
    In studies of Arusha, in particular, the connection between sins
and illnesses as divine punishments have been stressed. Landei argues
that if somebody has done something wrong, God will sooner or
later punish that person, and the punishment may not only affect
the individual but the corporate group as well.93 Likewise, Kimerei
says that sickness is a consequence of sin and adds that almost all
suffering may be attributed to the sins of human beings.94 It should
be remarked, however, that some of Kimerei’s informants were
Christian converts. Moreover, the emphasis—or overemphasis—on
disease as punishment for sins may be seen as an example of the
Christian theological framework of interpretation that characterizes
the works of Kimerei, Landei and some others.
    It seems that in the early work of Merker (1904) there is a sim-
ilar exaggeration of the role of God as a guardian of morals. Merker

       Spencer 1988: 39.
       Mol 1978: 70; Spencer 1988: 41.
       Voshaar 1979: 190 ff.
       Hauge 1979: 19.
       Landei 1982: 57 ff.
       Kimerei 1773: 52, 55. Cf. Landei 1982: 58.
78                              chapter three

says that since human beings are ‘weak and sinful’, God occasionally
has to punish them.95 A wider comparison of sources indicates that
such punishments are not generally believed to be common. It appears
that, as a rule, wrongdoers are dealt with by elders rather than by
God himself. Hauge even holds that, although God may be invoked
to take measures against transgressors, he is in general ‘morally
indifferent’ and is not expected to intervene automatically.96 Similarly,
Århem states that he has not heard of diseases that have been
attributed directly to moral misconduct or transgression of norma-
tive rules.97
   About half a century ago, Schmidt hypothesized that Maasai morals
had ‘degenerated’. In his ‘inverted’ evolutionist perspective, Maasai
monotheism was no longer as ‘pure’ as it once was, nor was God
expected to act as a guardian of morals to the same extent as for-
merly. According to Schmidt, things like cattle raids and the ‘licen-
tious’ sexual morals were earlier forbidden by Maasai laws, and God
severely punished transgressors.98 Other accounts of Maasai religion
do not warrant any such conclusions about significant changes con-
cerning the role of God as a guardian of morals. Schmidt’s, as well
as Merker’s, conclusions are theoretically biased, and Hauge’s words
about God as ‘morally indifferent’ appear to be yet another exag-
geration. As indicated by the above-mentioned expressions ‘the black
God’ and ‘the red God’, for instance, God is ambivalent. Hence he
is considered to embody the dual qualities of good and evil, creation
and destruction, within himself.99 In his study of the Matapato Maasai,
Spencer speaks about ‘a dualistic perception of God, protective on
the one hand and castigating on the other’.100
   Clearly, the sources are conflicting and do not allow any precise
conclusions about the issue of diseases and death as divine punish-
ments. On this question, as well as in other cases, the significance

       ‘Er tut es dann durch Krankheit, Dürre oder Viehseuchen.’ Merker 1904: 196.
       Hauge 1979: 40. Cf. ibid., 47.
       Århem 1989: 84.
       Schmidt 1940: 428 ff. According to Schmidt (ibid., 431): ‘Der Verfall der
Sitten geht zwar Hand in Hand auch mit dem Verfall des Glaubens und der Kultur,
aber der erstere geht beteutend schneller vor sich und kann schon bedenklich weiter
gediehen sein, wenn auch der Glaube und noch mehr der Kult noch auf
bemerkenswerter Höhe stehen.’
       E.g., Galaty 1979: 321; Marari 1980: 22; Hurskainen 1984: 174 f.; Peron
1995: 53.
       Spencer 1988: 48.
                            god in maasai thought                                  79

of variations between groups and individuals must be kept in mind
too. These considerations notwithstanding, it appears that, in most
cases, God is not thought of as a moral guardian who metes out
punishments in the form of diseases, deaths and other forms of mis-
fortune. He does not often seem to be involved directly as a cause of
illnesses and deaths, except in the case of serious afflictions and when
people harm certain living things, such as infants and trees, that
stand in a particularly close relationship to him. Nevertheless, his
role stands out as very important in that he is also indirectly involved
in the actions of elders and healers who, as will be elaborated in
chapter seven, wield the power of cursing which is an important
cause of disease. Elders and healers stand in a special relationship
to God and use their supranormal power in conjunction with him.
    Prayers and sacrifices to God often exist in connection with cases
of disease; and it seems that they may be directed to him even when
he is not thought of as the direct or immediate cause of the disease
in question.101 In cases when God is believed to be the direct cause
of an illness, he is called upon in a special manner to bring redress.102
When people are wasted by serious illnesses such as epidemics, large
numbers of worshippers may be involved and several animals may
be slaughtered near mountains or oreteti trees.103

                   Other Spiritual Beings as Agents of Disease

Since the evil or bad spirits called iloiriruani can enter a person and
cause mental disturbances, Hauge suggests that they may be referred
to as ‘disease agents’.104 Voshaar, on the other hand, hesitates to use
the term ‘spirit’ at all and says concerning oloirirua:
        It is something, not further specified, that enters a person and which
        may be said to possess that person. One who has oloirirua in him, has
        something in him that makes him do mad things. He utters strange
        sounds, throws off his clothes, walks around shouting. He is like a
        drunk. In fact, one who is habitually drunk, might eventually drink

       E.g., Johnston 1915: 482; Schmidt 1940: 382; Voshaar 1979: 245.
       Voshaar (1979: 250 f.) holds that the sacrifices are reminders of life rather
than gifts or substitutions for life. God himself is confronted with the facts of life:
‘See what life is, do this to us no more’. Cf. Hauge 1979: 37.
       Kimerei 1973: 56, 59; Hauge 1979: 25, 35. Kimerei (1973: 10 f.) and Hauge
(1979: 35) emphasize the element of reconciliation in sacrifices.
       Hauge 1979: 41. Cf., e.g., Bleeker 1963: 120 f.
80                              chapter three

        himself into a state where he is said to have oloirirua. The state is
        different from being stupid (emodai ), from being mad or simpleton (emu-
        noi ), from being agitated, from being depressed etc. However all these
        states, that have known causes, may be the beginning of a process that
        eventually induces oloirirua.105
Although more research is needed to elucidate the vague and little
observed phenomenon of oloirirua and its deleterious effects on human
beings, it appears that it has no direct relation to the modern spread
of spirit possession among some Maasai in Tanzania. In the new
context of possession, pepo or olpepo, a word of Bantu origin, is prob-
ably the most common name for ‘spirit’, and it is sometimes used
as a genre name for a group of various spirits.106 According to Benson,
pepo possession first occurred during a brief period at the time of
German occupancy. Due to great problems of drought and famine,
some Maasai had to flee to Bantu areas in the coastal region; and
when they returned they ‘brought’ with them some foreign spirits.
After that, isolated cases of possession by such spirits were reported,
but it was not until the last two or three decades that this modern
phenomenon became common or widespread.107 Spirit possession
among Maasai has also been spread through the activities of numerous
healers, waganga, from the coast, who have extended their activities
to the Maasai plains. The main function of these healers is to appease
the foreign spirits, who cannot be appeased by Maasai healers, iloi-
bonok. However, the new healers are instrumental in spreading the
phenomenon too. Moreover, spirit possession among Maasai has been
influenced by neighbouring peoples such as Gogo and Chagga.108
   When possession spreads like an epidemic, it becomes an impor-
tant cause of disease. The most frequently occurring initial symp-
toms of possession are chronic headaches, stomach pains, pains in
the back and limbs, dizziness, lack of strength, lethargy and faint-
ing. If somebody has pains in any part of the body, and if these
pains do not disappear through ordinary medical treatment, indige-
nous or biomedical, there is reason to suspect that there are spirits
involved. Furthermore, possession is quite often linked with fear,
stress and anxiety.109 Persons who are possessed by pepo can act in

        Voshaar 1979: 216. Cf. Berntsen 1979: 289.
        Ibid., 266; Hurskainen 1989:145.
        Benson 1974: 171.
        Hurskainen 1989: 142.
        Peterson 1985: 174; Hurskainen 1989: 143.
                           god in maasai thought                      81

a way similar to that of people who have oloirirua. Hence they may
shiver, run away in a mad way, throw off their garments and so
on.110 Perlitz reports, for instance, that during a Christian service for
possessed Maasai women, who came there to be baptized, one of
the women danced naked in the church, threw herself to the ground
and shouted.111
   With a few exceptions only, it is Maasai women who become pos-
sessed.112 Some of the spirits of possession are animal spirits or spir-
its of nature, others are ‘human’ spirits representing particular ethnic
groups, for instance Mgogo or Mwarusha, and yet others, such as
Shetani and Jini, may be of Islamic origin. The possessed women
behave in different ways, which are said to be typical of the vari-
ous spirits. A lion spirit, for example, gives the host a dangerous
power.113 However, since possession brings diseases and other types
of suffering, it is not something that people try to attain, and those
who become possessed aim at being liberated from their spirits of
possession. Possession is contracted through contacts with non-Maasai
people, and possessed Maasai must consult non-Maasai healers or
Christian churches for the purpose of exorcism. Since contacts with
non-Maasai healers, and the pepo songs and incense they use in their
healing ceremonies, involve risks of contracting new spirits, the heal-
ing provided by the churches is nowadays regarded as the most
effective way of coping with possession.114 As a consequence, there
is a tendency among Maasai to think of the Christian baptism pri-
marily as a ‘medicine’ for certain illnesses caused by spirit posses-
sion.115 While other diseases are treated by Maasai healers, modern
hospitals or dispensaries, or by the use of medicines available to the
patients themselves, spirit intrusion thus requires an entirely different
kind of conceptualization and treatment. ‘Spirit possession is clearly
classified as a malady to be distinguished from other ailments.’116
   Pepo has become a name for sickness which does not respond to
biomedicine or indigenous types of healing; and Peterson concludes
that Maasai men have thus far willingly brought their possessed wives

        Voshaar 1979: 266.
        Perlitz 1973: 14.
        Peterson 1985: 175; Hurskainen 1989: 145 ff.
        Hurskainen 1989: 144 f.
        Voshaar 1979: 266; Hurskainen 1989: 147.
        Peterson 1985: 176.
        Hurskainen 1989: 147.
82                             chapter three

to the churches—but not the well ones. In a strongly male-oriented
society, he adds, this may leave the impression ‘that Christianity is
for women, and sick ones at that’.117 Here the Catholic Church differs
from the Lutheran Church, however, in that Catholics have insisted
that a whole homestead be baptized rather than one sick member.118
   Since pastoral Maasai do not normally believe in an afterlife, dead
people are only in exceptional cases held to be able to influence the
conditions of living humans. Unlike ordinary people, healers (iloibonok)
may be believed to continue their existence as invisible and power-
ful spirits who can harm the living.119 Like living healers, as will be
shown in chapter seven, dead healers can cause diseases and other
types of misfortune.120 Among the Arusha the situation is much
different in that spirits of ancestors and cult of these spirits are promi-
nent features. The importance of the spirits of ancestors as agents
of disease is strongly emphasized in early works by scholars includ-
ing Fokken and Blumer.121 A similar picture emerges from the more
recent study by Kimerei, although he holds that the ancestors are
intermediaries between God and human beings.122 When ancestors
are believed to be the causes of various illnesses, Arusha pray and
make sacrifices to them or, in the interpretation of Kimerei, through
them to God.123
   With regard to the Parakuyo, Hurskainen says that their belief in
some kind of continuity of life in the hereafter is manifested in similar

       Peterson 1985: 177.
       Voshaar 1979: 267.
       Hauge 1979: 51 ff.
       Ibid., 56.
       Fokken 1917: 242; Blumer 1927: 78. According to Fokken (1917: 242), the
belief in a more general influence of such spirits is fairly important even among
pastoral Maasai, ‘Die Erde gilt als Wohnsitz der Geister der Verstorbenen. Von
ihnen hält man das Schicksal der Menschen abhängig. Diese Ansicht finden wir
bei den ansässigen Ackerbauvölkern stärker ausgeprägt als bei dem Hirtenvolk der
Masai. Doch auch letztere glauben an ein Eingreifen der Geister der Verstorbenen
in das Leben der Hinterbliebenen, wenn auch nicht so sehr zu ihrem Unheil, so
doch zu ihrem Segen.’ It seems likely that the pastoral Maasai studied by Fokken
were more influenced by Arusha and other agriculturalist groups than were pas-
toralists in other parts of Maasailand.
       Kimerei 1973: 56, 64.
       Ibid., 56, 66. See also Fokken (1917: 239) and Blumer (1927: 79). Kimerei
(ibid., viii), who was a careful adherent of the indigenous Arusha religion before
joining the Evangelical Lutheran Church of Tanzania, reports that he once took
part in an important ‘veneration service’ for ancestors, performed on his behalf,
when he was at the point of dying of an unknown illness.
                           god in maasai thought                    83

rituals of crisis. A person who suffers from a disease or some other
kind of misfortune is supposed to go to the place where his father
was buried, or left to be eaten by animals, and bring honey beer,
milk and tobacco as offerings. In a prayer the dead is asked to leave
the suffering in peace. If the cause of misfortune is thought to be
misconduct against a deceased father, people may travel long distances
to make offerings.124 The central role of deceased fathers is empha-
sized even more strongly in Gulliver’s work on Arusha. He says that
supplications are invariably addressed primarily to the father, even
when it has been divined that another ancestor is the cause of the
illness. The dead father is believed to be able to contact and influence
all earlier male and female ancestors, including those whose names
are forgotten, but who can still affect the living.125 There are, thus,
no shrines specifically for earlier ancestors which could have general
significance for a wider group.126 A woman’s agnatic ancestors may
bring diseases or other kinds of misfortune upon her and her children.
In such a case, her husband arranges and prosecutes the matter with
her agnates—she may not even attend the rituals. It seems, however,
that at least older women are far more likely to be affected by their
husband’s ancestors than by those of their natal agnatic kin.127
    The supranormal power of a dead person can harm those living
agnates who ignore their obligations. Yet the sufferer need not nec-
essarily be the actual offender. In case of an unusually serious and
persistent illness, the head of the afflicted family initiates a special
ceremony of supplication to the ancestors at the shrine of his home-
stead. Serious diseases or misfortunes may not be occasioned only
by the nearer ancestors of the afflicted person or persons but may
derive from other ancestors too. In such cases all members of the
maximal lineage should attend and participate, and special pieces of
the slaughtered ox are offered to all the lineage dead. In practice,
some lineage members who live far away may not attend; but mem-
bers of the inner lineage always come, wherever they live.128

        Hurskainen 1984: 180 f.
        Gulliver 1963: 83.
        Ibid., 78.
        Ibid., 142 f.
        Ibid., 98 f.
                              CHAPTER FOUR


                         The religion of the Sukuma, in its practical aspects,
                         is composed almost entirely of a direct ritual rela-
                         tionship with the spirits of their ancestors.1
                         Il faut noter tout d’abord qu’en principe chaque
                         famille ne s’occupe que des mânes de ses ancê-
                         tres. . . . C’est à eux en effet qu’elle attribue dans
                         la plupart des cas un malheur survenu, maladie,
                         contrariété, sterilité des époux, etc.2


In many comparative accounts of indigenous African religions, ‘ances-
tor worship’ has been depicted virtually as ‘the religion’ of Africa,
although there are many African peoples who pay little or no atten-
tion to the spirits of the dead. However, in certain agriculturalist
and agro-pastoralist cultures without highly centralized socio-political
systems, spirits of ancestors have the central role in the religious sys-
tem. The indigenous Sukuma culture is one example of that type of
society in which the cult or veneration of ancestors is the most promi-
nent feature. Belief in God and a few other spiritual beings is, cer-
tainly, a part of Sukuma religion, but ‘what constitutes the very focus
of the ritual life’ among the Sukuma is the relationship between the
living and the dead, which is visibly manifested in the cult or ven-
eration of the ancestors.3
   Even though it seems that Sukuma religion has been somewhat
less heterogeneous than the religions of the San and the Maasai, the
importance of variations has been stressed by several scholars who
have specialized in the Sukuma too.4 There is no systematic theology

     Tanner 1967: 5.
     Gass 1973 [1919]: 392.
     Brandström 1990: chapter 6: 10. See also, e.g., Malcolm 1953: 50; Abrahams
1967: 77; Hatfield 1968: 16. Cf. Hendriks 1952b: 43.
     See, e.g., Table n.d.a.: 447; Tanner 1967: 1; Hatfield 1968: 13; Ng’weshemi
1990: 3; Wijsen & Tanner 2002: 64, passim.
86                              chapter four

or consistent cosmology, and there are considerable differences in
terms of doctrines or beliefs.5 Regarding behaviour or ritual life there
may be important variations too. ‘It is impossible to say that the
Sukuma do this or believe that; only the broadest principles may be
common to them as a whole.’6
   In 1910–11 some White Fathers reported that the Sukuma did
not show a great interest in religion, at least not in the religion of
the Christian missionaries.7 Yet, according to some other missionaries,
there were ‘no sceptics’ among the Sukuma.8 This is questionable.
As remarked by Wijsen and Tanner, it seems more likely that religious
and secular attitudes, which is an analytical rather than an emic
differentiation, have co-existed even before the colonial period. On
the whole, they stress the individual character of Sukuma beliefs—
or disbeliefs—and practices. Being individually or familially rather
than communally based, Sukuma religion may be described as a ‘do
it yourself ’ religion.9 Although Christian leaders have complained
that materialistic or secular values have increasingly influenced the
Sukuma, such leaders have concluded that many of them have clung,
or still cling, to indigenous beliefs and practices.10
   During the period of study here, as will be elaborated in chapter
eight, witchery has been clearly more prevalent among the Sukuma
than among the San and Maasai. One of the functions of some
Sukuma secret societies, for example Bagika, is to combat witches
and sorcerers. Societies like Bagika and Bagoyangi, one of those soci-
eties whose members are snake-charmers and specialists in treating
snake-bites, may also have medical functions.11 Like rituals of secret
societies, the few more open communal rituals associated with, for

      Cory 1960: 14; Tanner 1959a: 124; Tanner 1967: 27; Balina, Mayala &
Mabula 1971: 11.
      Tanner 1967: 2 f.
      Rapports 1910–1911: 158.
      Gass 1973 [1919]: 394. Cf. also Bösch’s account of the Nyamwezi (1930: 39),
among whom he had not found a single ‘atheist’ or ‘blasphemer’.
      Wijsen & Tanner 2002: 54 ff., 61, 177. See also, e.g., Wijsen & Tanner 2000:
21, 42. Rituals carried out by chiefs for the benefit of a whole community consti-
tute an exception.
      See, e.g., Table n.d.a: 446; Tanner 1967: 2; Ng’weshemi 1990: 1.
      Table n.d.a: 146, 524; Millroth 1965: 146 ff. In Sukuma history, secret soci-
eties and associations for the different sexes as well as for age-groups, neighbours
and other categories of people have played an important role, although they have
now lost much of their significance. See further, e.g., Table n.d.a.: 146 ff.; Cory
1960: 17 ff.; Juma 1960: 27 ff.; Abrahams 1967: 63 ff.; Wanitzek 1986: 199.
                       sukuma spirits of ancestors                             87

instance, chieftainship and ceremonies connected with the agricultural
cycle of the year have diminished considerably or virtually disappeared
altogether,12 thus strengthening the individual and situational character
of Sukuma religion. As a rule, religious practices take place when
people feel there is a need for them.
   The concept mhola, which may be translated ‘health’, is a term
referring to the desirable state of life in a most comprehensive sense.
Thus, not only health—in a limited sense—but also prosperity, peace
and all the good things of life belong to the realm of mhola. It is the
‘cool’ state of peace and good relations between the living as well
as between the living and the dead. The state of mhola must be
attained, maintained and, time after time, reattained. Busatu, illness,
is a less comprehensive term than mhola. There is no inclusive expres-
sion for the state of ‘non-mhola’. However, illness and natural calami-
ties like drought and subsequent famine (nzala), for example, are
spoken of in terms of ‘not being’ or ‘lacking’ mhola.13

                               Spirits of Ancestors

In the 1870s, when the White Father Livinhac taught the Sukuma
about the ‘immortal souls’ of human beings, one ‘negro’ responded
that death was the end of everything.14 Five years later, another
White Father, Nathalie, reported that Sukuma people believed in
‘souls’ but not in punishments or rewards in the afterlife.15 In several
respects Nathalie’s account of ancestors in Sukuma belief and practice
is similar to later accounts by scholars.
   Moyo is a Sukuma word that has frequently been translated ‘soul’.
It is associated with the heart, the breath and the shadow of a human
being. Moyo is the principle or spirit of life, which is centred in the
heart and maintained or manifested by the breath. The shadow is
also a sign that a person is living, and dead people or corpses are
said to lack shadows. When a human being dies, moyo ceases to

     Abrahams 1967: 61; Tanner 1967: 35; Wanitzek 1986: 205 f.; Wijsen & Tanner
2002: 56.
     Brandström 1990: chapter 6: 2. See also Wijsen & Tanner 2000: 46; Blokland
2000: 15.
     Livinhac 1879: 172. That there were some Sukuma who held this view was stated
by, among others, some White Fathers in the early 1950s too (Table n.d.a.: 421).
     Nathalie 1884: 186.
88                               chapter four

exist.16 Within a human being, however, there is a spiritual entity,
associated with the blood, which survives death. Thus, unlike animals,
humans continue existing after they have died. The surviving spirit
of human beings, or spirits of ancestors, are usually called masamva
(sing., isamva).17 An isamva is an immaterial or incorporeal but also
individual or personal power of great importance.
   Like Father Nathalie, other authors have concluded that there is
no ‘heaven’ and ‘hell’ in the Sukuma afterlife.18 Hence ‘good’ and
‘bad’ persons do not exist in different abodes after death. As in other
religions, Sukuma conceptions of the afterlife tend to be vague and
conflicting. In some respects the life of ancestors is similar to the life
of living humans. For instance, the social position is not changed in
the land of spirits, and ancestors are offered food products which
they enjoyed in life. Spirits of ancestors may be associated with the
earth and the graves as well with the air above.19 It cannot be
excluded that the notions of some Sukuma, who believe that dead
people go to God, are influenced by Christian or other new ideas.
Wherever the spirits of ancestors may be thought to exist, they stay
in the world of spirits and are not reincarnated.20
   There is a close interrelationship between living and dead. Living
humans relate to their own deceased agnatic as well as cognatic
relatives.21 However, a wife is outside the sphere of her husband’s
ancestors, and her husband does not have to fear her ancestors.22
In a sense, spirits of ancestors depend upon children and grand-

      Table n.d.a: 419, 468; Tanner 1959a: 109 ff.; Millroth 1965: 114 f. Cf.
Hendriks 1952b: 1. From the Nyamwezi area, Bösch (1930: 47 f.) reported that
moyo was associated with the heart or the stomach but seldom with the head or
brain. See also Blohm 1933: 151 f.
      Table n.d.a: 420 f.; Hendriks 1962: 2; Tanner 1967: 15. Apparently, there is
some uncertainty concerning the proper meaning of this term. According to Gass
(1973: 394), however, it is ‘without doubt’ derived from the verb kusama, move away
(‘déménager’). Cf. Hendriks 1952b: 3. Brandström (1990: chapter 6: 10) says that
masamva refer to the active power of the ancestors. In some areas the concept miz-
imu is used instead of masamva; and special categories of ancestors may have spe-
cial names. See, e.g., Table n.d.a: 428; Reid 1969: 44.
      Cf. Ng’weshemi (1990: 28), whose account is influenced by Christian conceptions.
      Since, with the exception of dominant chiefly lineages, Sukuma lineages have
not been and are not in any way localized, the land is not considered as belong-
ing to the ancestors (Tanner 1959a: 123).
      Table n.d.a: 421; Hendriks 1952b: 6 f.; Tanner 1958: 225; Tanner 1959a:
123; Millroth 1965: 116 f.; Tanner 1967: 20. See also Bösch 1930: 38, 50 f.
      Augustiny 1923–24: 169; Hendriks 1952b: 3; Malcolm 1953: 68; Millroth 1965:
121 f.; Brandström 1990, chapter 6: 11.
      Table n.d.a: 469; Brandström 1990, chapter 6: 12.
                        sukuma spirits of ancestors                               89

children for their ‘survival’, although deceased children and adults
without offspring may be included in general terms when masamva
are invoked. Descendants are, thus, of vital importance.23 Ancestors
of three or four generations can be remembered and honoured by
descendants. People of high social status, such as chiefs, remember
and invoke a greater number of spirits than do ordinary people.24
Chiefly ancestors do not influence their own descendants only but
also others in the chiefdom territory.25 In a sense, a chief was identified
with his chiefdom. There are ancient chiefs, batemi, who have been
honoured for more than two centuries.26 In a chief ’s ancestry, the
progenitor of the lineage may be mentioned in recitations at propi-
tiation ceremonies.27 It should be emphasized, however, that during
the last few decades not many people have been concerned with
activities relating to chiefly ancestors, and the cult or veneration of
ancestors has ceased to be a part of Sukuma political life.28
   This cult or veneration is manifested in a rich variety of, among
other things, sacrifices, invocations and dances. There are simple
libations of milk and beer as well as more complex rituals with
sacrifices of fine and healthy goats, sheep and cattle in large feasts.
Spirits of ancestors may be invoked as individuals or as collective
groups. In order to be accepted by the spirits, not only the correct
performance of a ritual but also the right intentions of those who
perform the ritual is important. When people invoke ancestors, they
often invoke God as well. Heads of families and elders have the
main responsibility for and play a big part in the cult or veneration
of ancestors, although their importance and the respect for them
have decreased in recent decades.29 People do not necessarily have
to honour the masamva at fixed places, even though the rituals or

      Tanner 1958: 225; Hendriks 1962: 2. Cf. Tanner 1959a: 111.
      Tanner 1967: 13; Hatfield 1968: 166; Balina, Mayala and Mabula 1971: 45.
      Abrahams 1967: 77; Brandström 1990, chapter 6: 12. Twin-ancestors as well
as ancestors of famous rainmakers and great healers may also have a territorial
      Hendriks 1952b: 3; Millroth 1965: 176.
      Tanner 1958: 225.
      See further, e.g., Bösch 1930: 44; Tanner 1955: 274; Tanner 1959a: 115.
      Like some other authors influenced by Christian theology, Ng’weshemi (1990:
19) interprets the role of ancestors as intermediaries between God and living humans,
and he prefers to say that the ancestors are ‘revered’ rather than ‘worshipped’.
Father Hendriks (1962: 7) concluded that Sukuma prayers to ancestors resembled
Christians’ addresses to saints who were asked to intercede with God. See also
Wijsen 1993: 108; Tanner & Mitchell 2002: 56; Wijsen & Tanner 2002: 55. For
more information about the cult or veneration of Sukuma ancestors, see Nathalie
90                                chapter four

ceremonies are frequently performed at graves. Cult activity may
also be localized at small spirit houses built in honour of ancestors
and constructed from specific trees with symbolic meanings.30 Moreover,
animals like cows, goats and sheep are kept in remembrance of
masamva. Such dedicated animals are separated from other animals
herded by a family. Dedication of animals, as well as erection of
spirit houses or shrines, are often the result of some illness or other
misfortune caused by spirits of ancestors. A consecrated animal may
not be sold or slaughtered and is kept until it dies.31 In addition to
animals, virtually all kinds of inanimate objects may be consecrated.
Some of these ritual paraphernalia are articles that ancestors possessed
in their lifetimes, but other objects such as arms, sticks, skins of
animals, necklaces, pearls and many other things can be dedicated
too. In time of illness, for instance, people bear their venerated
   In 1897, Father Brard reported that sacrifices to ancestors were
common.33 Nowadays, more than a century later, the frequency of
such sacrifices is much lower. As early as in the 1950s, Tanner con-
cluded that many ceremonies in honour of ancestors had virtually
disappeared, and that others were markedly on the decline. Even in
the more out-of-the-way parts of Sukumaland, regular everyday rituals
directed to spirits of ancestors were no longer carried out.34 In 1967,

1884: 187; Augustiny 1923–24: 169; Hendriks 1952b: 8, 15, 19 f., 38 f.; Tanner
1958: 229; Tanner 1959: 120; Millroth 1965: 162; Tanner 1967: 13, 39 ff.; Hatfield
1968: 65; Steeves 1990: 67; Wijsen & Tanner 2002: 62.
     Augustiny 1923–24: 169; Hendriks 1952b: 6, 16 ff., 32; Tanner 1959: 118;
Millroth 1965: 160 f.; Tanner 1967: 32; Ng’weshemi 1990: 19; Wijsen & Tanner
2002: 55. Cf. Tanner 1958: 225. In a study of the Nyamwezi, Tcherkézoff (1985:
73) compares the spirit houses to altars.
     Table n.d.a: 271 ff., 617 ff.; Hendriks 1952b: 13, 21 f.; Tanner 1959: 117 ff.;
Tanner 1967: 33; Tcherkézoff 1985: 74. Among the Sukuma there is a great variety
of taboos, particularly for women, and only some of them are associated with
ancestors. Detailed information about taboo conceptions are found in, for instance,
Gass 1973 [1919]: 440 ff., passim; Table n.d.a: 626 ff.; Kirwen 1974: 163.
     Gass 1973 [1919]: 395, 397; Hendriks 1952b: 15, 23 f.; Tanner 1958: 230;
Hendriks 1962: 4; Tcherkézoff 1985: 72. The sacred objects referred to here, which
are usually called shitongelejo (sing., kitongelejo), are not amulets or charms. Amulets
may be but are not necessarily connected with the veneration of ancestors and they
are even more numerous than the shitongelejo. See further Table n.d.a: 577; Abrahams
1967: 72; Tanner 1967: 29; Wijsen & Tanner 2002: 59.
     Brard 1897: 156.
     Tanner 1958: 229; Tanner 1959: 116.
                       sukuma spirits of ancestors                             91

Tanner drew the following conclusion with regard to the cult or
veneration of ancestors:
     Now both the educated and uneducated persons neglect their ances-
     tors for years and consider propitiating them only when they are in
     trouble, so that the cult of the ancestors has changed from maintaining
     their goodwill by regular rites as was done in the past to the present
     intermittent recognition of their powers to harm and the ceremonies,
     individual rather than collective, necessary to recover their goodwill.
     It is their interference rather than their benevolence that occasions the
Sickness now seems to be the most important reason for ceremonies
of propitiation. Hence it is largely in the context of disease problems
that the construction of spirit houses, the consecration of various
objects, sacrifices, initiation of individuals into secret societies and so
forth must be seen. Through the various means of propitiation, health
can be regained.36

                         Ancestors as Agents of Disease

It seems that spirits of ancestors may be thought of as agents of all
kinds of diseases as well as of infertility and other misfortunes, possibly
with the exception of death. There are conflicting reports with regard
to ancestors as causes of death. Whereas in some accounts it is con-
cluded that death can be occasioned by them,37 there are other reports
stating that at least ordinary ancestors are not associated with deaths.
This may reflect the variety of beliefs held by different Sukuma indi-
viduals and groups. According to Hendriks, ancestors cannot themselves
occasion death, but they may obtain help from witches or sorcerers,

     Tanner 1967: 24. See also Tanner 1970: 24. See further Wijsen & Tanner
2002: 55.
     For some more information on propitiation in connection with illnesses and
other types of misfortune occasioned by ancestors, see Nathalie 1884: 188; Gass
1973 [1919]: 403, 408, 410; Hendriks 1952a: 41, 44; Hendriks 1952b: 33; Table
n.d.a.: 474, 496; Malcolm 1953: 23; Tanner 1958: 227 ff.; Tanner 1959: 116; Tanner
1967: 24, 33, 36 f.; Hatfield 1968: 63; Balina, Mayala and Mabula 1971: 20;
Tcherkézoff 1985: 59 ff.; Steeves 1990: 81. An account of the various aspects of a
specific sacrificial ceremony of propitiation is found in, for example, Tanner 1959:
     See, e.g., Tanner 1967: 17; Balina, Mayala and Mabula 1971: 8; Ng’weshemi
1990: 19, 39; Steeves 1990: 69.
92                                 chapter four

who are agents of death.38 As emphasized by Tanner, moreover,
ancestors are ultimately dependent on God and do not have a direct
control over the forces of nature. Power over life and death is the
prerogative of God but not of ancestors.39 The French quotation at
the beginning of this chapter shows that, in 1919, Gass and others
concluded that ancestors were seen as the main reason for maladies
and other kinds of misfortune; and, even though there has been a
decline in terms of beliefs and practices relating to ancestors, there
are still many Sukuma who relate to these spirits as important agents
of disease. To be ‘seized’ by an isamva is an undesirable experience
associated with some misfortune, such as physical or mental illness.40
   It appears that, in general, there are no specific diseases that are
occasioned by the actions of ancestors. As a rule, the symptoms of
a certain illness do not indicate if it is an ancestor at work. However,
‘social’ symptoms such as lack of respect for daily family life may
do so. Moreover, if a child is doing poorly in the days immediately
after its birth, that would easily be taken as a sign of ancestor action.41
According to Hendriks, there are a few diseases whose names indi-
cate that they are connected with ancestors. For example, ihugi is
the name of an affliction, caused by an ancestor, which makes the
sick person furious or even mad. Some masamva are particularly angry
and malignant because they died a violent death. They may have
been killed in time of war, of famine or of witchery, or they may
have drowned or been eaten by a wild animal.42 In most cases, how-
ever, divination is needed in order to find out not only which ances-
tor is at work but, first of all, if it is an ancestor at all. Likewise,
problems of fertility may be caused by spirits of ancestors but can
also have human and natural causes.
   There are some diseases that may not be occasioned by ordinary
ancestors. In particular, epidemic diseases are found in this category.

      Hendriks 1952b: 8.
      Tanner 1967: 23. See further Augustiny 1923–24: 168, 170; Millroth 1965:
183; Balina, Mayala and Mabula 1971: 46.
      Hatfield 1968: 59 f.; Reid 1969: 186; Steeves 1990: 69. According to Tanner
(1959a: 108), sickness may be regarded as the outward expression of the illness of
moyo. Concerning the Nyamwezi, Tcherkézoff (1985: 85 ff.) says that when sickness
comes, the ‘soul’, that is, the vital principle, diminishes, and if it does not stop, the
afflicted person may eventually die.
      Tanner 1959a: 114; Reid 1969: 77 f.
      Hendriks 1962: 1 f.
                          sukuma spirits of ancestors                               93

The masamva cause illnesses of their descendants but not of people
who are not members of their own families. For instance, a string
of white beads and a copper bracelet worn on the left wrist may
tell us that the bearer is afflicted by paternal ancestors, while a string
of black beads worn on the right wrist informs that a consecration
has been made to maternal ancestors.43 As will be elaborated in
chapter eight, diseases caused by witchery may have somewhat more
specific symptoms than have afflictions occasioned by spirits of ances-
tors. Hence it may also be concluded from certain symptoms that
the disease is not the result of the intervention of an ancestor.
   Unlike ordinary ancestors, spirits of chiefs can be the cause of epi-
demic illnesses. Chiefly ancestors may also harm their own descen-
dants on an individual basis, but, as representatives of a whole
territory, they can influence all the inhabitants of that area. Previously,
chiefs attempted to control, by ritual means, deadly diseases like
smallpox, plague, measles, cholera and rinderpest. Epidemics sent by
chiefly ancestors, as well as their withholding of rains, were lethal
threats to large groups of people.44
   There are many reasons why spirits of ancestors may cause dis-
eases and other types of misfortune. According to Tanner, ‘Sukuma
believe that misfortune in all its forms is the result of sin which turns
back on its perpetrator and spreads through his family’.45 The use
of the term ‘sin’ may be somewhat misleading here. In another work,
however, Tanner elucidates his use of this concept. He holds that
there are ‘no absolute values’ in Sukuma morals and that moral
issues are qualified by relationship, locality and the circumstances
surrounding an incident. A person’s primary ties are to his or her
family and tend to diminish with increasing kinship distance. ‘The
average Sukuma would seem to grade his ideas of right and wrong
according to the degree of relationship which he has to the person
in the context.’46

        Hendriks 1952b: 8; Table n.d.a.: 471; Steeves 1990: 69; Brandström, chapter
5: 1.
      Hendriks 1952b: 8; Ng’weshemi 1990: 20 f.; Steeves 1990: 80. See also
Tcherkézoff 1985: 60; Blokland 2000: 32. There is some evidence that, in the past,
ailing chiefs could be killed by ritual officers. Since, in a sense, a chief was identified
with his chiefdom, he had to be healthy. If he was seriously ill, the community was
affected too. See further Tanner 1959a: 115; Millroth 1965: 176 f.; Abrahams 1967:
76 f.
      Tanner 1958: 225.
      Tanner 1967: 22.
94                              chapter four

    Needless to say, this feature is not peculiar to Sukuma, although
it tends to be less pronounced in cultures where kinship ties are not
quite as important as they are among the Sukuma. Many scholars
emphasize the great importance that Sukuma attach to correct behav-
iour within the community of living and dead family members.
Lineage values must be observed and ancestors must be honoured,
lest diseases or other problems should follow. If grave faults, such
as incest between close kin, are committed, disaster may threaten
not only human fertility but even the fertility of the land. Seeing
that they influence their own descendants only, ancestors cannot be
depicted as upholders of Sukuma custom and morals in general.47
    Another way of angering ancestors, which may result in maladies,
is failure to name children after them. An ancestor can demand
more than one child to be named after him or her. Name-giving is
usually done soon after birth, but names may be changed later in
life because of illness which has been attributed by divination to a
dereliction of filial duty. The head of the lineage will invoke the
ancestors and asperse the afflicted person at a formal ceremony.
After this ceremony, however, there is no need for him or her to
use the new name in everyday life, as the formal ceremony and the
mention of the name therein is sufficient to alleviate the troubles.48
    If ancestors are not properly remembered in the prayers and
offerings of their descendants, the spirits will occasionally cause mis-
fortune until such remembrance is made. Masamva may also affect
living relatives in order to air past grievances and to demand satis-
faction. Such grudges can be based upon past events which may be
unknown to the living.49 A particularly unpredictable and dangerous
category of ancestors are those who died in unusual circumstances
and whose corpses were mishandled or lost. The absence of a corpse
is a breach of lineage unity, which makes the lineage members more
exposed to sickness. A large proportion of all ceremonies carried out
to reduce illness are directed to the unpredictable ‘wandering spirits’.50

      Concerning the origin of death, Sukuma share with many peoples a mytho-
logical tradition of blaming the first woman. There are various versions of what
‘deadly sin’ she committed. One of them tells that she broke a clay pot, which she
was forbidden to break or destroy. See, e.g., Table n.d.a: 422, 430, 434, cf. 423,
435, 471; Tanner 1959: 122; Millroth 1965: 199 f.; Balina, Mayala and Mabula
1971: 34; Welch 1974: 177; Steeves 1990: 69; Brandström 1990, chapter 6: 18.
      Tanner 1959: 112 ff.; Balina, Mayala and Mabula 1971: 34, 41.
      Hatfield 1968: 63, 77; Steeves 1990: 66; Wijsen & Tanner 2002: 63.
      Tanner 1958: 222. See also Tcherkézoff (1985: 71) who speaks about such
‘aerial’ elements causing illness among the Nyamwezi.
                       sukuma spirits of ancestors                              95

   A certain amount of unpredictability is a characteristic of all ances-
tors. Like living humans, spirits of ancestors are not good or bad but
good and bad. As ambivalent beings, ancestors may be contented
agents of blessings as well as discontented agents of misfortune. Thus,
the surviving relatives ask their masamva to assist in providing good
harvests, many children, good health and other good things in life.
Since the activities of ancestors are more powerful than those of
their descendants, the spiritual assistance is important in order to
obtain a good life on earth.51 As emphasized repeatedly by Tanner,
however, the masamva are ultimately dependent on God, even though
prayers and offerings are addressed directly to the former and in-
directly only to the latter.52
   Since ancestors are able to harm their descendants, there is some
amount of awe of them; and, as indicated above, it seems that, at
least in recent decades, masamva are generally asked to abstain from
evil rather than to bring about good. Yet the amount of fear should
not be exaggerated. The feelings of living humans towards their
ancestors are a complex mixture of affection and awe. In a sense,
they are still one with the living members of their families rather
than dominating outsiders who aggressively control the living.53 When
communicating with their living relatives, ancestors can appear in
dreams or possess people. Possession, particularly of women, seems
to be a common phenomenon among the Sukuma. A possessed
person, or ‘medium’, acts as a mouthpiece of the possessing spirit,
and there are different varieties of glossolalia. People who become
possessed are both ordinary persons and healers, bafumu (sing., nfumu).
Moreover, an ancestor can reveal himself or herself in the form of
an animal, usually a snake, which must be well treated.54
   In divination, it is the living who actively seek to find out which
ancestor is the cause of a particular disease and what complaints or

      Hendriks 1952b: 8; Hendriks 1962: 3; Hatfield 1968: 60; Balina, Mayala and
Mabula 1971: 7; Steeves 1990: 68. Cf. Reid 1969: 44.
      See, e.g., Tanner 1959a: 116; Tanner 1967: 23.
      See further, e.g., Gass 1973 [1919]: 402; Table n.d.a: 474; Hendriks 1952b:
15; Tanner 1959: 122; Balina, Mayala and Mabula 1971: 20. Cf. the accounts by
Hendriks (1952b: 8), who stresses the element of fear, and Tanner who, in some
places (e.g., 1959: 121; 1967: 39), concludes that there seemed to be no such feel-
ings towards ancestors. In another work, Tanner (1967: 21) inconsistently argues
that ‘(ancestors) are feared rather than venerated’.
      Nathalie 1884: 188; Gass 1973 [1919]: 41, 443; Bösch 1930: 49; Table n.d.a:
471, 623; Hendriks 1952b: 9, 12 f.; Hendriks 1962: 2; Millroth 1965: 119, 123 f.;
Wijsen 1993: 113 f.; Wijsen & Tanner 2000: 73; Wijsen & Tanner 2002: 64.
96                                chapter four

wishes that spirit may have.55 Simple and common forms of divina-
tion are known to ordinary people, while more advanced cases are
referred to specialists. Bafumu are highly specialized in various forms
of divination and healing.56 Many of them are members and lead-
ers of secret societies. Bamanga (sing., manga), the majority of whom
seem to be women, are diviners who often practise as healers too,
and whose contact with and possession by masamva is of vital impor-
tance.57 Although there are some bafumu who work with medicines
or instruments, and do not rely on spiritual support, most of them
are dependent on the assistance and power of masamva. The more
successful ones are said to have a special ritual power or personal
charisma called busebu.58 Much of the energy of the bafumu is devoted
to the task of maintaining or re-establishing harmony between living
humans and ancestors, and their attention to people in need is very
personal. Yet the bafumu, most of whom are men, should not be seen
as some exclusive priestly class. In principle at least, anybody can
become an nfumu. Nowadays, Tanzanian authorities require licences
and certificates of people who want to practise as bafumu. It appears,
however, that most of them prefer to remain outside the scrutiny of
administrators at all levels.59
   A detailed discussion about the functions of the cult or venera-
tion of ancestors is outside the scope of this study. Yet a few of the
conclusions drawn here should be stressed. Ancestor veneration helps
to keep families together. It is closely associated with the Sukuma
social system and serves as a kind of social cement. Ancestors are

      The position and signs of certain objects as well as entrails of animals are used
in divination. Among the Sukuma, chicken divination is the most common instru-
mental form (Table n.d.a: 474; Hatfield 1968: 116 ff.).
      Bufumu is an abstract concept, indicating the ‘magic component’ in a medi-
cine or, more generally, the complexes that result from any human being’s spiri-
tual links with his or her ancestors (Cory n.d.b: 1). When referring to the term
bufumu, Hatfield (1968: 90 ff.) uses the expression ‘supernatural power’ and says that
a person’s power and influence is attributed to his or her bufumu. We may com-
pare here the San concept of num. Bufumu is, moreover, the name of one of the
most important Sukuma secret societies. However, bafumu are not necessarily mem-
bers of that society (Malcolm 1953: 68). According to Tanner (1967: 42) and
Ng’weshemi (1990: 21), the term nfumu may be derived from the verb kufumbula,
‘to discover’.
      Bumanga is the name of another important Sukuma society.
      Wijsen & Tanner 2000: 29. The term is used for fever too.
      See further Gass 1973 [1919]: 412; Hendriks 1952b: 9; Table n.d.a: 152 ff.,
503; Millroth 1965: 140 ff.; Tanner 1967: 42 ff., 49; Hatfield 1968: 145, 151, 153,
287; Reid 1969: 95, 104; Pedersen 1977: 64; Wijsen & Tanner 2000: 45.
                        sukuma spirits of ancestors                               97

interested in the successful continuation of their lineages, and their
presence sanctions the moral behaviour of individuals within specific
families. It was indicated above that to a certain extent ancestors
seem to have an intermediary function, since God is invoked in con-
nection with prayers and rituals directed toward masamva. In the fol-
lowing section, the role of God in Sukuma religion and disease
causation will be presented.60

                            God in Sukuma Thought

Among the Sukuma, there is, with certain local variations, a great
number of words for God the creator. Many of the names indicate
attributes of God. The most common names seem to be associated
with the sun too. This link between God and the sun is found among
many African peoples. Lyuba, Liuba and Liova are some of the Sukuma
forms referring to God as well as to the sun. The prefix li, which
is found in several names for God, is augmentative, thus adding a
meaning of grandeur or greatness. Another common name is Liwelelo,
which refers to God and to the universe. It is used to address the
omnipresent God, who is immense and puissant. God is not identified
with the sun and the universe. These entities should rather be seen
as symbols or manifestations of him. According to Brandström, Liwelelo
is the Sukuma-Nyamwezi term for ‘the totality of the universe’. It
designates the Unknown, ‘something beyond human control’.61 God
is the ruler of the sun and the universe. Other names indicate different
aspects, intentions or facets of God’s character. For example, there
are several names for God as creator. Nowadays, non-Sukuma names
such as Mungu and Mulungu are frequently used too.62

      For more discussions of the function of ancestor veneration among the Sukuma,
see Hendriks 1952b: 4; Table n.d.a: 447; Tanner 1959a: 115, 120; Tanner 1967:
26; Ng’weshemi 1990: 24; Steeves 1990: 67.
      Brandström 1990, chapter 5: 22; chapter 6: 9.
      See further, e.g., Nathalie 1884: 186; Gass 1973 [1919]: 385; Table n.d.a: 412
ff., 507; Tanner 1956a: 45; Cory 1960: 15; Millroth 1965: 19 f., 46, 95 ff.; Balina,
Mayala and Mabula 1971: 2; Ng’weshemi 1990: 8 ff. One of the ‘localized’ names
for God in Sukumaland is Ngassa. This name is associated with water holes, rivers
and, above all, Lake Victoria. It designates God as the ruler of this great lake. See
Tanner 1967: 5; Balina, Mayala and Mabula 1971: 5. Cf. the accounts on the
Nyamwezi in Bösch (1930: 26 ff.) and Schönenberger (1961).
98                                chapter four

   As creator and master of the universe, God is without beginning
and end.63 He is the Supreme Being and, consequently, pre-eminent
over lesser spirits. Since he is the major power in the hierarchy of
beings, everything else is contingent upon him. Ultimately, he is the
master of life and death, the manifestation of power in its totality.64
If symbolically ‘male’, God is incorporeal and invisible. Unique and
unlimited, he has no family and no particular living place, although
in invocations he is associated with the heavenly realm.65 In comparison
to Maasai, for instance, Sukuma seem to be much less concerned
with the immanent presence of God. Hence transcendence is a pri-
mary characteristic of God in Sukuma thought. In some scholarly
accounts, he is even depicted as a deus otiosus.66
   There is a significant element of ambivalence in the nature of
God, although several authors conclude that he is regarded as pri-
marily good.67 Tanner, for example, holds that God ‘is above the
petty influences of man’ and that his character is ‘primarily good’.68
Yet he does not seem to be directly concerned with particular cases
of good or bad actions. In other words, his interest in the moral
system is theoretical rather than concrete, even though he may be
described as an ultimate judge.69 Since God’s transcendence and non-
involvement in human affairs are pronounced features of Sukuma
religion, it is not surprising that there is little or no cult of God.
Both early and later sources show that Sukuma have seldom prayed

      Nathalie 1884: 186; Table n.d.a.: 416; Welch 1974: 175.
      Table n.d.a.: 416; Tanner 1956a: 45 f., 51; Millroth 1965: 99; Hatfield 1968:
54; Balina, Mayala and Mabula 1971: 4.
      Table n.d.a: 412, 417; Millroth 1965: 99; Tanner 1967: 6.
      See, e.g., Cory 1960: 15; Tanner 1967: 7; Hatfield 1968: 55, 76. Unlike these
anthropologists, some scholars of religion have described God as actively involved
in the affairs of human beings (e.g., Millroth 1965: 205; Ng’weshemi 1990: 18, cf.
7). Although, as will be seen, there are certainly some examples of belief in such
involvement, their view seems to be exaggerated and to some extent influenced by
Christian theology. Cf. Wijsen 1993: 112; Wijsen & Tanner 2002: 62.
      See, e.g., Table n.d.a.: 417; Millroth 1965: 104; Ng’weshemi 1990: 38.
      Tanner 1967: 7. Cf. Brandström (1990, chapter 6: 9) who stresses more strongly
the ambivalent nature of God: ‘(Liwelelo) is both an ultimate value and a non-value.
It is an ultimate value in that it covers everything and it is a non-value in the sense
that it represents neither plus nor minus, neither good nor bad.’
      Table n.d.a.: 448, cf. 417; Tanner 1956a: 448; Millroth 1965: 100; Balina,
Mayala and Mabula 1971: iii. Cf. Ng’weshemi 1990: 12, 26. According to Tanner
(1967: 7), Sukuma view God as being ‘very largely indifferent to man’s activities’.
                        sukuma spirits of ancestors                               99

and sacrificed directly to God. Likewise, temples or shrines have not
been built in his honour.70
    In a sense, however, there is an ‘indirect’ cult of God. As indicated
previously, God may be invoked when people address their ancestors.71
In some exceptional cases, furthermore, people can pray and sacrifice
directly to God himself. Such a cult is usually in the form of peti-
tions for something, such as rain and children. If there is a threat
of war or before starting a long and risky journey, people may also
turn directly to God. Of particular interest here is the evidence for
prayers and sacrifices in connection with cases of epidemics and other
serious diseases.72 People may ask God to heal them from such seri-
ous afflictions.73
    Unlike spirits of ancestors, God is not often seen as an agent of
disease. However, there are instances of his involvement in cases of
illness and death. In principle, it appears that any case of disease
and death may be attributed to God. While ancestors tend to be
seen primarily as agents of fairly benign afflictions, diseases caused
by God are usually more severe. When attempts to heal sick people
by addressing ancestors, who have been associated with their maladies,
fail, God may be seen as the real cause of their problems.74 If God
decides to take a person away, there is no medicine that can stop
the process of dying. Primevally, God was not only the giver but
also the terminator of life; and, ultimately, he is still the master of
life and death. In the final analysis, thus, health and disease are
under his control.75
       It is generally recognised that a man cannot be sick without the con-
       nivance of the Supreme Being, irrespective of the attitude of the ances-
       tor spirits . . . Should a man have a prolonged illness before he died,
       it would be said that the Supreme Being cannot be beaten but that

      Brard 1897: 156; Gass 1973 [1919]: 387; Cory 1960: 15; Tanner 1967: 8;
Brandström 1990, chapter 6: 10; Wijsen & Tanner 2002: 174. Concerning the sim-
ilar situation among the Nyamwezi, see Bösch 1930: 39; Schönenberger 1961: 949.
Cf. Ng’weshemi 1990: 14 f.; Wijsen 1993: 108.
      Table n.d.a: 440 f.; Cory 1960: 440 f.
      Nathalie 1884: 189; Bösch 1930: 39 ff.; Table n.d.a.: 443 f., 507; Millroth
1965: 125; Ng’weshemi 1990: 16 f. Cf. Wijsen 1993: 158.
      Table n.d.a.: 444; Ng’weshemi 1990: 17.
      Gass 1973 [1919]: 386; Table n.d.a.: 447; Tanner 1956a: 49, 52; Tanner 1967:
9 f.; Ng’weshemi 1990: 12; Wijsen & Tanner 2002: 55.
      Gass 1973 [1919]: 385 f.; Table n.d.a.: 418; Balina, Mayala and Mabula 1971:
3, 43; Ng’weshemi 1990: 11. Concerning the Nyamwezi, see Bösch 1930: 36; Blohm
1933: 189. It seems that death may not necessarily be seen as something entirely bad.
100                              chapter four

      his ancestors were very strong . . . and if he should recover, that the
      intentions of the ancestor spirits have been agreed to by the Supreme
Tanner even speaks about a sense of ‘fatalism’, or predestination,
among the Sukuma. God dispenses luck, or bad luck, although no
particular system of giving or withholding luck can be discerned.77
Similarly, Brandström concludes that God may be brought in to
describe ‘wonder, the unexpected and the unexplainable’.78 According
to Brandström, God is ‘beyond dualism’ and ‘all anthropomorphic
characteristics’.79 Other sources indicate that, as death was first sent
as a punishment, God may at times punish evildoers by sending ill-
ness and death, even though he is not generally seen as the upholder
of the moral system.80 Moreover, Tanner reports that illness can be
interpreted as a warning from God, and in such a case there is good
hope of ultimate recovery.81

                              Other Spiritual Beings

Among the Sukuma there are few, if any, spiritual beings apart from
God and the spirits of ancestors. There are certain beings associated
with lakes, mountains, hills and other parts of nature that could be
referred to as spirits of nature. It seems, however, that such beings
have a very marginal position in Sukuma religion.82 As a rule, they
stand aloof from human beings and are not objects of cult.83 Scholars,
and apparently the Sukuma themselves, vary a great deal in their
conceptions about the nature spirits. A basic question is whether they
are seen as separate or individual beings at all, and if they were

According to Tanner (1956a: 47), there is a popular saying that without death
people would not have enough room to live. Death may, thus, be regarded as a
‘clearing away’ by the Supreme Being so that all ‘may live well’.
      Tanner 1956a: 50. See also Tanner 1967: 10; Steeves 1990: 63.
      Tanner 1956a: 51. See also Wijsen & Tanner 2002: 62.
      Brandström 1990, chapter 6: 10.
      Ibid., chapter 6: 25.
      Perhaps this idea of disease and death as punishments from God has been
somewhat exaggerated by scholars of religion. See, e.g., Bösch 1930: 37 f.; Table
n.d.a: 418; Millroth 1965: 104, 183; Ng’weshemi 1990: 7, 12, 24.
      Tanner 1956a: 49.
      Brard 1897: 152 f.; Hendriks 1952b: 43 ff.; Cory 1960: 90; Pedersen 1977: 62;
Wijsen & Tanner 2002: 59. Millroth (1965: 111 f.) reports a guardian of animals too.
      Tanner 1956a: 45; Hatfield 1968: 47.
                        sukuma spirits of ancestors                               101

once human beings, or if they are conceived of as a kind of emanation
from the Supreme Being.84 At least the most frequently described
‘spirit of nature’, Ngassa, is usually depicted as a manifestation or
localized function of God rather than as a separate and indepen-
dent being. According to Gass and others, Ngassa is the designation
for God as the master or guardian of Lake Victoria.85 Before embarking
upon their voyage across the lake, sailors may sacrifice to Ngassa,
who may also be invoked in connection with the veneration of
ancestor spirits.86
   Spirits of nature are of very minor, if any, significance as agents
of disease and death. The sacrifices to Ngassa are made in order to
avoid death by drowning and other problems. In the main, how-
ever, it seems that when other spiritual beings than ancestors are
thought of as causes of disease, possession and various attacks on
people, those spiritual beings are of foreign origin, such as the Swezi
spirits.87 Other foreign spirits, like mashetani, majini and mapepo, have
entered Sukumaland through contacts with Muslims. Such spirits
have increasingly become the best available explanation for personal
misfortune which divination does not attribute to ancestors or witchery.88

      For some differing accounts, see Hendriks 1952b: 44 f.; Tanner 1956a: 45 f.;
Millroth 1965: 107 ff.; Hatfield 1968: 49; Steeves 1990: 63 ff.
      Gass 1973: 386. See also Tanner 1956a: 46; Tanner 1967: 5; Welch 1974:
183; Pedersen 1977: 62. Cf. Millroth 1965: 107 f.
      Table n.d.a.: 481; Millroth 1965: 109; Welch 1974: 183. Cf. Brard 1897: 152 f.
      Millroth 1965: 109, 113; Abrahams 1967: 78. With regard to the Nyamwezi,
Blohm (1933: 173) pointed to the influence of foreign spirits. Cory (1960: 15 f.)
held that, although spirits of nature did not normally take any interest in this world,
they could occasionally be induced by powerful ‘magicians’ to cause evil. He men-
tioned Simungala as an example of a particularly dangerous ‘demon’ that could
cause smallpox, plague and other epidemics. However, the observance of taboos,
the avoidance of provocation and the use of protective medicines were considered
sufficient to keep away the displeasure of such spiritual beings.
      Wijsen & Tanner 2002: 61.
                               CHAPTER FIVE

                     KONGO SPIRITS OR NKISI

                          Le fétichisme . . . c’est à peu près à cela que se
                          résume toute leur religion.1
                          Every nkisi has its own affliction that it imposes,
                          and it is used also in the treatment of the same.2


As indicated by the first quotation above, as well as by many other
writings from the late nineteenth and early twentieth centuries, so-
called fetishism was at that time the predominant element in the
Kongo religion.3 Time and again western missionaries and colonial
rulers launched attacks on the concepts of nkisi. ‘Fetishes’ (nkisi or
minkisi), that is, sacred objects associated with certain spirits, were
destroyed in bonfires. The American anthropologist Janzen concludes
that, as a result of the ‘crusade’ against nkisi between 1890 and 1930,
which was joined by African prophets like Simon Kimbangu, region
by region in the Lower Congo responded to the call by abandoning
or burning most of their nkisi.4
   The Kongolese world or cosmos is divided into one visible and
one invisible part. The visible world of the living is above the earth,
and the invisible world below is inhabited by ancestors and other
spiritual beings. During the nineteenth century, according to MacGaffey,
‘a sense of the dead as moving, by a series of transformations in the
afterlife, through a hierarchy of increasingly remote but also more
powerful and functionally less specific positions in the other world,
apparently corresponded to the existence in the real world of hierarchies

    Extrait 1895: 576.
    Laman 1962: 69.
    See also, e.g., De Hert 1895; Marichelle 1898: 595; Le Scao 1908: 333; Sadin
1910: 132 ff.
    Janzen 1978: 50. See further Axelson 1970: 47, 142; Hagenbucher-Sacripanti
1973: 31; Mahaniah 1973: 227; Janzen 1979: 211; Widman 1979: 147; Pambou
1980–81: 42.
104                              chapter five

and titles’. After the nineteenth century, however, ‘the collapse of
real political structures above the village level entailed the collapse
of the spiritual hierarchy, so that the notion of ranked classes of
spirits is now poorly articulated’.5
   In addition to the belief in and cult of nkisi and ordinary ancestors,
the Kongo religion, as well as other indigenous religions in the matri-
lineal Bantu belt of western Central Africa, also contains belief in,
but little or no cult of, the Supreme Being or God. The uniformity
of the structure of religion in this area has been stressed by, among
others, Vansina, Jacobson-Widding and MacGaffey.6 As pointed out
by Pambou, however, there is no systematic theology and dogmat-
ics.7 Moreover, as early as in 1910, Sadin observed in an essay on
the Kongo religion that religious specialists were ‘very often’ scepti-
cal about the power of the nkisi spirits.8 The long history of mis-
sionary presence among the Kongo means that some particulars of
the indigenous Kongo religion have been influenced by Christianity.
Yet, as emphasized by Widman, it is difficult or sometimes impos-
sible to see in what ways and to what extent this influence has
affected the indigenous religion.9
   As in the other cases studied in this book, disease etiologies among
the Kongo are related to religion but also to human and natural
factors. ‘Disease’ (mayeela) is a broad category of afflictions of all
kinds, or conditions perceived as such. The ‘why me?’ question is
important. Thus, people want to know why afflictions are distributed
as they are and rarely admit chance as an explanation.10 While com-
mon and benign afflictions may have natural or physical causes, spir-
itual and human agents tend to be suspected in more serious cases.
Considering the importance of the latter agents of illness, social
groups are largely involved in the process of healing or therapy.11

      MacGaffey 1983: 129. See also Janzen 1978: 23; Mahaniah 1979: 218; Jacobson-
Widding 1979: 332 f. Cf. Mahaniah 1982: 30 f., 48.
      Vansina 1975: 671 ff.; Jacobson-Widding 1979: 25 f.; MacGaffey 1986: 185.
      Pambou 1980–81: 38. See also Bouquet 1969: 23; Thiel 1974: 641. Vansina
(1975: 574) remarks that myths are not very elaborate.
      Sadin 1910: 135.
      Widman 1979: 39.
      MacGaffey 1983: 148.
      See further, e.g., Mahaniah 1973: 229 f., 231 ff., 237 f.; Janzen 1978: 8, 67,
73, 224 f.; Mahaniah 1982: 16, 63 ff., 67 ff.; Dimomfu 1984: 76.
                            kongo spirits or      NKISI                        105

                               Spirits of Ancestors

Even though, as indicated above, the nkisi were the predominant
spiritual beings in the Kongo religion in the late nineteenth and
early twentieth century, ordinary spirits of ancestors were also impor-
tant. Since the cult of nkisi, or very ‘old’ spirits of ancestors, is related
to the cult or veneration of ordinary, or ‘young’ ancestors, it is rea-
sonable to present the latter and their role as agents of disease first.
   In addition to the life-essence, or principle of life (moyo), which is
associated with the blood and the heart, human beings have an ele-
ment, or ‘soul’, that survives the death of the body.12 Van Wing
neatly expressed the significance of ancestors by saying that ‘the dead
are the living par excellence’.13 For two to three generations, ancestors
(bakulu, sing. nkulu) are still lineage members and interact closely with
their relatives on earth. A special category of ancestors, associated
with witchery, are the bankuyu or minkuyu (sing. nkuyu) or ‘ghosts’.
Although in some dialects, and for some people even within the
same dialect area, one of these Kongo terms—bakulu or bankuyu—
may indicate the alternative concept, ‘bakulu refers most often to pri-
marily benevolent spirits with some known or implied genealogical
relationship to the living, whereas min’kuyu are usually malevolent,
unattached ghosts who are not supplicated but exorcised’.14
   Similarly, there is some uncertainty and variation concerning the
term basimbi (sing. simbi ), which refers to another category of spiritual
beings, who are also somehow related to the nkisi. According to some
Kongolese, the simbi spirits are distant ancestors, who have lost their
individuality, while others seem to believe that they are a special
class of beings in the underworld, created by Nzambi, the Supreme

      Nsala and mwela are Kongo terms for the ‘soul’. The complex Kongo views of
human beings have been elaborated by, among others, Laman (1962: 1 ff.) and—
largely following Laman—Jacobson-Widding (1979: 307–323). See also, e.g., Proyart
1819: 151; van Wing 1959: 376; Janzen 1978: 158; Widman 1979: 106 ff.; MacGaffey
1986: 50 ff., 135.
      van Wing 1959: 250.
      MacGaffey 1986: 64. MacGaffey (ibid.) adds that ‘this apparent vagueness is
characteristic of all status designations, including kinship terms, and many related
expressions’. The vagueness in the use of the terms bakulu and bankuyu may be illus-
trated by comparing the accounts of, for instance, Laman (1916: 199, 205 f.),
Widman (1979: 92) and Dalmalm (1985: 64). See also Jacobson-Widding 1979: 94
ff., 103 ff.; Pambou 1980–81: 39. Cf., e.g., Hersak (2001: 615, 622), who writes
about the Vili and Yombe in the north-western Kongo region.
106                              chapter five

Being or God. They may also be associated with certain parts of
nature, such as rivers and streams, and thus be conceived of as local
spirits or spirits of nature. According to some people, moreover, the
basimbi are the spirits of those who died a violent death. Like ordi-
nary ancestors, the basimbi are generally benevolent, although they
can be capricious and ambivalent.15
   The life of ancestors in their invisible world is similar to that of
living humans, but in most respects the world of the dead inverts
the latter. Hence, for example, to the ancestors six a.m. is nightfall.
Since dead people have been conceived of as white, missionaries and
other white people from the west have been thought of as return-
ing ancestors. In a sense, the grave is the house of the deceased in
the afterlife, and cemeteries are usually situated in mixed forests that
grow up on the sites of abandoned villages.16 Yet ancestors are not
tied to their graves but also associated with the wider forest, the
underground and the ocean or other forms of water. Certainly, intel-
lectuals of indigenous Kongo society may provide analytical accounts
of ancestors and other otherworldly forces. MacGaffey points out,
however, that individual informants are chiefly aware of the unique
reality of their own experiences. Therefore, their answers to questions
regarding, among other things, the whereabouts of the ancestors tend
to be anecdotes, which may seem contradictory, rather than coherent
analytical summaries.17
   Since ordinary ancestors, or bakulu, are generally benevolent and
peaceful, they do not belong to the most important agents of disease.
As a rule, they protect rather than harm their living relatives.18 Yet
certain afflictions may be occasioned by bakulu. For instance, they
can cause minor problems such as headaches.19 Laman and Jacobson-
Widding stress, in particular, the significance of paternal ancestors.

      In the seventeenth century, according to MacGaffey (2000: 212), a hierarchy
of simbi spirits corresponded to a hierarchy of political domains. For more details
on basimbi, see Laman 1916: 206; Laman 1962: 33 ff.; Buakasa 1973: 239; Jacobson-
Widding 1979: 114 ff.; Widman 1979: 190; Mahaniah 1982: 34; Lagercrantz 1983/84:
69; MacGaffey 1983: 133, 145; Dalmalm 1985: 65; MacGaffey 1986: 63, 68, 76
ff., passim. On the issue of regional variation, see e.g. Hersak 2001: 615.
      Sadin 1905; Vansina 1975: 672; Jacobson-Widding 1979: 98; Widman 1979:
89, 116 f., 134; MacGaffey 1986: 45 ff.; MacGaffey 2000: 206.
      MacGaffey 1986: 88 f.
      Laman 1916: 207; van Wing 1959: 379; Mahaniah 1982: 231. Cf. MacGaffey
2000: 222.
      Janzen 1978: 178; MacGaffey 1986: 67.
                          kongo spirits or     NKISI                      107

According to Laman, the paternal power (kitata) is the key element
in the ancestor cult or veneration of the matrilineal Kongo society.
As will be elaborated in chapter eight, living fathers, who do not
have jural power over their own children, have an extraordinary
power of cursing the children if they do not respect and pay due
homage to their fathers. A father’s curse can result in illness and
other types of misfortune.20 Likewise, dead fathers can punish descen-
dants who misbehave. For instance, a son can become passive,
depressed and even impotent. Impotence in a wide sense is a char-
acteristic affliction caused by (living or dead) fathers. Correspondingly,
infertility or various kinds of passiveness can affect a daughter.
Whenever somebody feels that he or she is not successful and has
lost vitality, the effect of the paternal power is easily suspected.
Ordinary maternal ancestors, by contrast, usually do not occasion
   To some extent bakulu superintend the social and moral order.
Ancestors are thought to be like the elders they recently were. Hence
they are very conscious of their due and likely to punish disrespect.
In general, the punitive tendency is considered justifiable, although
there may also be unacknowledged motives and hidden hostilities.22
In order to please the ancestors it is important to provide a proper
cult. The Kongo cult or veneration of ancestors is essentially directed
to deceased fathers and other paternal relatives, that is, members of
the father’s lineage. Fathers are honoured through gifts. Palm wine
and other offerings are brought to the burial places, and the fathers
reciprocate by protecting and blessing their descendants. The cult of
chiefly ancestors is particularly elaborate. More serious afflictions,
such as epidemics, can be caused by deceased chiefs, who then have
to be supplicated and appeased.23
   Like the bakulu, the basimbi do not belong to the most important
agents of disease. According to T.K.M. Buakasa, the latter may also
play a certain role as guardians of morals and punish transgressions.24

     Laman 1962: 46.
     Jacobson-Widding 1979: 384 ff.
     van Wing 1959: 379; Laman 1962: 49; Janzen 1978: 178; MacGaffey 1983:
145 f.; Dalmalm 1985: 175.
     See further Laman 1916: 209 f.; Jacobson-Widding 1979: 111 ff., 128; Widman
1979: 126 ff.
     Buakasa 1973: 240 f.
108                             chapter five

They influence the fertility of women, as well as of fields. Abnormally
born children, particularly twins and albinos, are regarded as a sort
of incarnation of simbi spirits.
   Twins are regarded as an affliction in two senses: literally, because
it is a great burden on their mother to nurse them and to try to
keep them alive, and mystically, in that twins, as bisimbi, use their
supernatural power (kindoki, ‘witchcraft’) to bewitch each other and
other people who incur their displeasure. The principal purpose of
the rites of the cult is to keep the children themselves and other
bisimbi connected with them happy. A subsidiary purpose is to cure
certain diseases that may afflict members of a lineage into which
twins have been born and that has thereby been revealed as particularly
susceptible to simbi influences.25
   By contrast to the generally benevolent bakulu and basimbi, the
bankuyu are malevolent and extremely malicious. Hence they can
cause serious afflictions, and some diseases may require the exorcizing
of bankuyu.26 Van Wing reports that human flesh is to these noxious
beings, who are condemned to anonymous wanderings, what pork
is to ordinary Kongo people, that is, the choice dish.27
   There is evidence that in the late eighteenth and early nineteenth
centuries the cult or veneration of bakulu was, in general, a very
important element alongside the cult of the nkisi in the indigenous
Kongo religion.28 During the modern era of colonial rule, however,
the significance of the ancestors, and particularly of chiefly ancestors,
declined gradually. Nowadays bakulu are rarely invoked and seldom
suspected as agents of afflictions. Even bankuyu, unlike living ‘witches’
or ‘sorcerers’ (bandoki ), are rarely depicted as such agents.29

                                 Nkisi Spirits

While the bakulu are ‘young’ and the basimbi are ‘old’ ancestors, the
nkisi (or minkisi ) form a third category of even older ancestors, although
the borderline between basimbi and nkisi is sometimes blurred. Some

     MacGaffey 1986: 85 f.
     Laman 1916: 206; Laman 1962: 21 f.; Jacobson-Widding 1979: 103 f.; Mahaniah
1982: 31.
     van Wing 1959: 291; MacGaffey 1986: 73.
     See, e.g., Sadin 1910: 139; van Wing 1959: 348, cf. 423.
     Mahaniah 1979: 225; MacGaffey 1986: 71 ff. Cf. Widman 1979: 142.
                          kongo spirits or     NKISI                      109

of the latter have existed from the very beginning of the Kongo
people. In addition to the great number of minor and local nkisi,
there are more than one hundred major ones. Among them are the
founders of the Kongo sub-groups or clans.30 In order to glorify and
mark the great significance of the highest nkisi, the concept of Nzambi,
the Supreme Being, may be used as a symbolic designation for them.
Thus, they are closely associated with God. Some nkisi are conceived
of as male, others as female; yet others are sexless.31 These spirits
can be connected to all sorts of places on land, in water and in the
sky. They can also be divided into different ‘families’ with specialized
   Nkisi spirits are, moreover, associated with certain sacred objects,
‘fetishes’, of different kinds and sizes.33 The nkisi objects are produced,
often in the form of sculptures that look like human beings or, more
rarely, like animals. Pots, baskets and similar things may also be
used. The objects have medicines attached, sometimes in a hole in
the stomach area and sometimes in a medicine bag. Religious specialists
known as banganga (sing., nganga), for whom healing is an important
activity, know and teach how to produce nkisi objects and what pre-
scriptions must be followed. There are certain taboos that an owner
of a nkisi object must observe. The medicines are chosen for both
symbolic and pharmacological reasons. For instance, there may be
heads of snakes in particularly dangerous nkisi, and hair of an albino
provides an albino’s special power. However, there may also be herbal
medicines in the nkisi object which, supplemented by more such medi-
cines, can be used in the treatment of diseases. In the nkisi cult, thus,
‘religion’, ‘magic’ and ‘science’ merge. Nkisi objects in the form of
human beings, which may be due to Catholic influence, have bent
legs, symbolizing the vitality and strength of the spirits. In certain
areas some sculptures have nails or other sharp things hammered
into them. This may also be a late innovation, possibly influenced
by crucifixes.34 According to, among others, Laman and Lagercrantz,

     Laman 1928: 18, 20; Thiel 1974: 638; Janzen 1978: 45; Widman 1979: 147;
MacGaffey 1990: 4.
     Laman 1917: 293; Widman 1979: 183; Jacobson-Widding 1983: 387; Lagercrantz
1983/84: 68.
     Laman 1928: 20; Vansina 1975: 672; Janzen 1978: 46; Lagercrantz 1983/84:
67, 70.
     See further Laman 1962: 67; van Wing 1959: 383; Buakasa 1968: 154 f.;
Dalmalm 1985: 65; MacGaffey 1990: 5 f.
     Cf. MacGaffey 2000: 99.
110                             chapter five

the nails are meant to hurt and thus activate the nkisi spirit.35 Moreover,
a nkisi can be ritually ‘killed’ by driving nails into it.36
   In addition to being important agents and healers of disease, the
nkisi have a number of different functions. For instance, they can
bring good luck in hunting, detect thieves, protect against bandoki,
give fertility to women and good harvests. There are nkisi for different
ages, for men and women, for war, different trades and so forth. In
part, they can be seen as guardian spirits of individuals and groups.
Previously, a chief in a village could compose a great nkisi for the
protection of all the inhabitants in the village. Like human beings,
however, the nkisi are more or less ambivalent and can thus harm
as well as protect. According to Laman, the later nkisi have more
specialized functions than the oldest ones.37 The nkisi can intervene
in every aspect of life, be it in the social, political, legal or religious
sphere. A nkisi can operate of its own accord, or by order of a nganga
or ndoki, but its sphere of action is restricted by its inherent quali-
ties and specialization.38 For instance, there are nkisi for different
ages. Some are primarily for men, others mainly for women. Male
nkisi tend to be more violent than the female ones, who are more
calm and peaceful. As indicated above, there are three great groups
of nkisi. Their medicines and power derive from these respective
spheres, although it is rare for a nkisi to have all its medicine from
only one of these spheres. The families of nkisi are organized in a
way that is similar to the system of matrilineal lineages among the
Kongo, even though they are less elaborate than the latter.39
   As indicated by the second quotation at the opening of this chap-
ter, a nkisi normally has its own sickness that it imposes, and it is
used also in the treatment of the same malady.40 Some nkisi can
cause several illnesses. Conversely, the same kind of illness can be

       Laman 1962: 90; Lagercrantz 1983/84: 81. See also, e.g., MacGaffey 2000:
     MacGaffey 1986: 159. For more information on nkisi objects, see Laman 1928:
16, 19, 25 f.; Laman 1962: 67 ff.; van Wing 1959: 384, 410; Buakasa 1973: 177,
227; Jacobson-Widding 1979: 139 ff., 152; Widman 1979: 153 ff., 183, 201 f.;
Lagercrantz 1983/84: 71, 77 f.; Dalmalm 1985: 65; MacGaffey 1986: 138 f., 140,
142 ff.
     Laman 1962: 71, 100. See further, e.g., van Wing 1959: 385, 394; Buakasa
1968: 156, 163; Jacobson-Widding 1979: 139; Widman 1979: 187 ff.
     Laman 1962: 75.
     Laman 1962: 71 f.; Buakasa 1973: 236 f.; Widman 1979: 185 f.
     See also, e.g., Sadin 1910: 133; Jacobson-Widding 1983: 387.
                          kongo spirits or     NKISI                     111

brought by different nkisi. Virtually all sorts of diseases, including
very serious and chronic ones—even death, can be occasioned by
nkisi spirits. In general, afflictions brought by female spirits tend to
be less serious than those that are caused by male ones.41 Apparently,
it is more difficult to see any major differences between earth, water
and air nkisi in terms of afflictions caused by them.42 A malady is
often called by the name of the nkisi that provokes it. Based on inter-
views with seven informants in 1980, Mahaniah made a list of illnesses
that may be caused by nkisi and other spiritual beings. This list shows
that the variety is great.43 In the following, some examples will illus-
trate the multiplicity of afflictions caused by various nkisi spirits.
    Nkondi is a great nkisi who previously was invoked in connection
with peace settlements and other forms of agreements between, for
instance, clans and villages. A breach of a contract could result in
illness. Nkondi could attack culprits of his own accord or after being
induced to do so. Like chiefs, thus, Nkondi maintained public order.
Nkondi nkisi are important also because they belong to those that
can destroy bandoki. On account of different specializations and com-
positions there are several Nkondi variants. For example, one attacks
the organs of respiration, while another spreads small scabby sores.
A Nkondi nkisi can also cause epilepsy and a range of other afflictions.44
    Another major nkisi who subdues bandoki, as well as bankuyu, is
Kula. According to Laman, there are, apart from general medicines,
among other things a hen’s foot, a hoof of an antelope, a claw of
a giant lizard and parts of several venomous snakes. These repre-
sent the qualities of the animals in question, such as swiftness, pugnac-
ity and ability to kill quickly. To Kula belong, also, small packets
of medicines to be used as protection against bankuyu spirits. Among
the Bwende, such medicines were previously sold in market places.
The composition of Kula objects involves much work, which often
takes place towards the evening, when bankuyu spirits become active.
Nkisi Kula catch, kill or chase away bankuyu or bandoki during their

     Butaye 1899: 310; Laman 1928: 21; van Wing 1959: 385; Hagenbucher-
Sacripanti 1973: 109; Mahaniah 1979: 241; Pambou 1979–80: 18 f.; Dalmalm 1985:
     Cf. Lagercrantz 1983/84: 88.
     Mahaniah 1982: 39–46. See also Laman 1962: 70.
     Laman 1928: 75 ff.; Laman 1962: 78, 86 ff.; Mahaniah 1982: 127 ff.; MacGaffey
2000: 100.
112                             chapter five

nocturnal journeys. Kula, who can be male or female, are the most
common nkisi for treating illnesses caused by bankuyu and bandoki.45
   Another very well-known major nkisi is Bunzi, who in some areas
is called Lemba. There is a variety of afflictions that can be caused,
and healed, by Bunzi. It can attack in the belly, chest, eyes and legs.
When someone suffers from dyspnoea, stitch, chest pains or anxiety,
it may be suspected that this nkisi has entered into the body of the
sufferer. Swollen stomach, limbs or eyes as well as coughing trouble
are characteristic problems brought by Bunzi.46 There are several
variants of this nkisi. From the Yombe area, for instance, Hagenbucher-
Sacripanti reports that Bunzi may be the cause of rheumatism and
arthritis. However, it also offers protection against thieves and other
evildoers. Bunzi, who is stronger than the majority of nkisi, is above
all resorted to when somebody has had constant bad luck in his
hunting or fishing.47
   Nakongo is one of the very oldest nkisi. According to Laman, he
      may be as magnificent as Nkondi, give good luck and happiness; but
      his chief function is to afflict people with illness by way of revenge for
      crime, and to be invoked by the sick person. Nakongo gives people
      hernias, makes their limbs crooked, their bodies swollen and gives them
      scabs and blisters between toes and fingers.48
The attacks may also be directed towards the chest and produce a
stubborn cough. For instance, thieves and adulterers may be visited
by Nakongo. As in the case of Bunzi, among others, there are sev-
eral variants of nkisi Nakongo. One of these, Makongo Banga, causes
madness and ignorance, while another, Makongo Mpanzu, seizes
people by the side of the chest and ‘squeezes’ to produce a lethal
cramp. Like Makongo Banga, Makongo na Mvangu occasions var-
ious kinds of insanity.49
   Mbumba is a famous and feared nkisi who provokes, among other
things, blood blisters, swollen and aching legs, severe diarrhoea and
stomach cancer. Yet, like other nkisi, Mbumba is ambivalent and

     Laman 1928: 81, 237 ff.; Laman 1962: 78, 92 ff.; Jacobson-Widding 1979:
233 f.
     Laman 1926: 215 f.; Laman 1962: 105 ff., 113.
     Laman 1962: 107 f.
     Ibid., 144.
     Ibid., 142–148; Jacobson-Widding 1979: 241.
                            kongo spirits or      NKISI                        113

can, for instance, make women who have had fertility problems
pregnant. Moreover, this nkisi protects against theft. For example,
people who steal fruit from trees that are protected by Mbumba will
immediately be attacked by this spiritual being.50
   Another important agent of disease is the Nkita family. Dimomfu
says that more than fifty illnesses may be treated by healers who
specialize in this family, although only a few of these afflictions are
of particular significance.51 Delivery problems and sterility, as well
as boils and sores all over the body, are important categories of
Nkita afflictions. Another such category is bodily malformations such
as harelip and hunchback. Nkita may also be the name of small ani-
mals entering human bodies, causing swellings and aches.52 Like the
Nkita nkisi, to whom they are related, the Kimpi nkisi can occasion
a high number of maladies, including miscarriage and insanity.53
   Another nkisi concerned with reproduction problems is Funza, who
caused sterility, miscarriage and was the reason for malformed humans,
animals, plants and inanimate things like stones. Laman observed,
however, that already at his time this nkisi had fallen into oblivion.54
While some nkisi, like Funza, lost their significance, others could
become increasingly important, and new ones from other peoples
were introduced. Particularly in urban areas, such as Léopoldville,
even nkisi from far away, like West Africa, could be taken over and
become worshipped by the Kongo.55
   Since there are hundreds of more or less important nkisi among
the Kongo, only some of them, and their role as agents of disease,
can be presented here. Mvutudi and Mwanza belong to the type of
nkisi that have the ‘shamanistic’ function of recapturing lost souls of
human beings. Mbumba is an ancient, well-known and feared nkisi.
Like most of the oldest nkisi, he is less specialized than the younger
ones. Among other things, he causes stomach problems, tumours—

       Laman 1928: 242 f.; Hagenbucher-Sacripanti 1973: 113 ff.; Widman 1979:
      Dimomfu 1984: 48.
      According to Jacobson-Widding (1979: 201), it is the Nkita nkisi who transform
themselves into little animals. Cf. Laman 1962: 132.
      van Wing 1959: 411; Laman 1962: 132; Buakasa 1968: 158 ff.; Buakasa 1973:
158 f., 169 ff.; Jacobson-Widding 1979: 201; Lagercrantz 1983/84: 88.
      Laman 1962: 78. See also Hagenbucher-Sacripanti 1973: 134 f.
      One important example, mentioned by van Wing (1959: 385 f.), is Muyeki. This
strong nkisi protects its worshippers and makes them invulnerable to their enemies.
114                              chapter five

including cancer—blood blisters, swollen and aching legs and severe
diarrhoea. On the other hand, Mbumba gives fertility and may be
used generally as a means of protection against the influence and
evil deeds of dangerous spirits and bandoki. Mbola and several other
spiritual beings belong to an nkisi family that is much feared because
of the large malignant sores with which it afflicts its victims. As a
beneficent power, Mbola can discover bandoki, thieves and other crim-
inals as well as hinder them in their activities.56 Nsakula is used
mainly as a protective nkisi, which may be worn on the chest. There
are many spirits in the Nsakula family, and they may occasion
headache and aching hips or legs. Moreover, they can prevent, but
also give, rain. According to Laman, Mayimbi is everywhere regarded
as a great nkisi belonging to a family of ‘smashers’ (bankosi). In cer-
tain areas it consists chiefly of two large sculptures, man and wife,
with small children (i.e., little sculptures), fashioned like human beings.
If a village is ravaged by an epidemic, the Mayimbi spirits are invoked
and sacrifices are made. Severe epidemics are provoked by the male
Mayimbi, while the female one may occasion less serious epidemics.
When the male Mayimbi attacks a village, he comes with great
power, smashes down trees, palms and other things, and the ail-
ments of the people are generally manifested in the chest. In some
tracts, however, Mayimbi is a little nkisi in a shell that gives good
luck, although it may also cause epilepsy.57
   In the account above some answers to the question why nkisi cause
diseases have been provided. A fundamental reason is their morally
dual nature: they protect and heal, but they also destroy and bring
afflictions. Each nkisi has this dual tendency and capacity. As they
are spiritual beings, the intelligence and power of the nkisi are superior
to those of humans. The anger of a nkisi can be roused because of
the malice within the nkisi itself.58 In many cases nkisi spirits act as
guardians of morals and react to crimes by sending afflictions. Breach
of taboo rules associated with specific spirits is another important

      Laman 1962: 78, 100 ff., 119, 138.
      Ibid., 124. In addition to Laman, several other scholars have provided exam-
ples of Kongo nkisi and their capacity to cause afflictions. By and large, however,
other examples show the same characteristics. For more details, see e.g. van Wing
1959: 388–418; Hagenbucher-Sacripanti 1973: 129–140; Buakasa 1973: 163–181;
Jacobson-Widding 1979: 200–219; Widman 1979: 187–193.
      Buakasa 1973: 17, 235; Janzen 1979: 210; Widman 1979: 194.
                           kongo spirits or      NKISI                       115

reason for afflictions inflicted by them.59 A living human’s cult rela-
tionship to a specific spirit is often initiated when the spirit seeks an
‘owner’ by causing his or her illness. If it is a serious malady, the
best way of becoming healed is to find a nganga who can teach the
patient how to put together the nkisi object. The patient then enters
into the service of this particular nkisi spirit and can help others
whom it may afflict.60
   In principle, anybody can become a nganga, although the pre-
scriptions and taboos connected with this profession deter most peo-
ple. In practice, male banganga are more common than female ones.
Yet the latter may be as powerful and influential as the former. In
addition to the possibility of being recruited through illness, it is also
possible to become a nganga after being an apprentice to a master
nganga. Patrilateral succession (i.e., from father to son) is the most
common mode of apprenticeship. A woman can become a nganga,
for example, after being obliged to purchase a treatment for a sick
child. She may then begin practising it on others. Occasionally,
elderly or rich people become banganga for the sake of gaining per-
sonal honour. As a rule, the recruitment of a nganga entails different
modes, although illness and successful healing by a nkisi may be the
most important means of embracing the profession.61
   There are various kinds of banganga, and they tend to be highly
specialized. The category that is of particular interest here is the
nganga nkisi. The banganga who specialize in nkisi can be differentiated
further, depending on which nkisi they serve. Hence they can be
called nganga Lemba, nganga Kulu and so forth. Among other categories
of banganga specialists, the nganga mbuki, herbalist or doctor, and nganga
ngombo, diviner, may be mentioned.62 It should be remarked, however,
that the distinction between these banganga and the nganga nkisi is not
always clear in practice. It is interesting to note, also, that a Christian
priest may be called nganga Nzambi (‘priest of God’). Since the banganga

      van Wing 1959: 400; Pambou 1979–80: 21 f.; Dalmalm 1985: 77. Janzen
(1979: 121) reports a case of somebody who was hit by affliction because he had
stepped on a malevolent nkisi, planted by a vengeful person out to destroy him.
      Lagercrantz 1983/84: 68 f. The most powerful nkisi are also the most expen-
sive (MacGaffey 2000: 112).
      van Wing 1959: 421; Laman 1962: 72, 173, 175; Janzen 1978: 196 f.; Jacobson-
Widding 1979: 68, 145; Lagercrantz 1983/84: 75.
      A few details about the former will be provided in the appendix. On Kongo
divination, see e.g. Laman 1962: 173; Mahaniah 1973: 231; Janzen 1979: 210.
116                              chapter five

may be concerned with a great variety of problems, not only regard-
ing religion and healing but also, for instance, law and teaching, the
term is difficult or impossible to translate into English. When a term
like healer is used, it should, therefore, be remembered that the art
of nganga covers much more than issues of illness and healing.63
   In 1910, the Jesuit missionary Sadin drew the conclusion that the
banganga, whom he called féticheurs, had great authority.64 Their impor-
tance is evidenced also by the fact that Christian missionaries have
regarded them as the major obstacle to the conversion of the Kongo
people to Christianity. In his historical study Culture Confrontation in
the Lower Congo, Axelson concludes that the banganga and their func-
tions proved to be ‘the most vital and long-lived social institution’.65
An essential reason for the authority of the nganga is the special kin-
doki power and intelligence ‘of the night’, that is, of the world of
spirits, that he or she has access to. Although it is possible for ban-
ganga to use this power in a destructive way, which allegedly does
happen, they are generally expected to use it for constructive or
healing purposes. Therefore, they are of great significance in the
struggle against the bandoki, who constantly try to harm other people.66
The ambivalent and extraordinary power of kindoki makes it possible
for the banganga to perform similar feats to the bandoki. For example,
some banganga may transform themselves into animals. A nganga Kula
is able to see the bankuyu, attack and exorcize them when they have
taken possession of somebody’s body.
   There are reports that some banganga perform miracles, predict
future events or conjure up their nkisi spirits or dead humans (bakulu).
Among other things, banganga may cut themselves in the tongue with-
out showing any pain, although blood flows; or, like San healers,
they may dance through fire, the flames licking their bodies but caus-
ing them no harm.67 In many ways prophets (bangunza) in indigenous
Christian churches in the Kongo area, such as the well-known Church
of Jesus Christ on Earth by the Prophet Simon Kimbangu, are similar

     According to van Wing (1959: 418), the word nganga means faiseur (maker,
manufacturer). See further, e.g., Janzen 1978: 195; Widman 1979: 157 ff.; Batukezanga
1981: 60.
     Sadin 1910: 135.
     Axelson 1970: 266. See also, e.g., Widman 1979: 157.
     Buakasa 1973: 152; Hagenbucher-Sacripanti 1973: 144. The issue of kindoki
and bandoki will be discussed in detail in chapter 8.
     Laman 1962: 18, 175, 177, 181.
                        kongo spirits or    NKISI                     117

to the banganga. For instance, like the latter, the prophets heal sick
people, perform miracles and struggle against the bandoki. To some
degree, thus, the bangunza may be said to have replaced the banganga.
The most explicit difference between the former and the latter, it
seems, is that the prophets charge no fees, although they may accept
gifts, while the banganga normally charge fees for their services.68
   When the banganga treat patients they use ‘religious’, ‘magical’ and
‘natural’ means. The medicines used are normally not the same as
those found in the nkisi bag. In some cases the patient can only be
healed by exorcizing an nkuyu spirit that has taken possession of him
or her, while in other cases, as among the San, the soul (nsala) of
the patient must be recaptured by the nganga.69 Normally the patient
knows the nganga, and their relation to each other is not impersonal.
Moreover, there is a large group of other people supporting the
patient. Janzen uses the term ‘therapy managing group’. Selection
of therapy, in co-operation with specialists, and assistance through-
out the process of healing is largely the responsibility of kinspeople,
although nowadays increasingly friends, job peers, coreligionists and
neighbours of the patient substitute for them. Janzen’s material indi-
cates that the previously pivotal role of the diviner, nganga ngombo,
as chief diagnostician has largely been taken over by the therapy
managing group.70
   At the beginning of the period studied here, nkisi spirits consti-
tuted the main focus of cult activity among the Kongo. The cult of
nkisi consists of, among other things, invocations, prayers, sacrifices
and dancing. For instance, a nkisi may be asked to attack an enemy
or to perform miracles. Prayers may also be said, and sacrifices made,
in order to soften the wrath and destructive inclination of a nkisi. In
addition to sacrifices of hens and other animals, there are offerings
of food, like maize, and beverages, such as palm wine. The most
powerful nkisi may have special houses or shrines or can be placed
on altars. Normally, however, the nkisi object is kept in the house
of its owner. A nkisi becomes angry if it is not properly treated and
provided with blood from game that has been killed, corn at harvest
time and so forth. A nkisi may be inherited when the owner dies,

     Laman 1962: 75; Bouquet 1969: 43; Dalmalm 1985: 92; MacGaffey 1986:
244. Cf. Janzen 1978: 155.
     Laman 1962: 77; Jacobson-Widding 1979: 152.
     Janzen 1978: 4, 131, 224 f., 229. See also, e.g., van Wing 1959: 237.
118                            chapter five

but the new person may not use it until he or she has been dedi-
cated and initiated by a nganga.71
   Gradually, however, people have lost their belief and interest in
the nkisi cult. In 1910, the Holy Ghost Father Marichelle concluded
that he and other missionaries were doing their best to wipe out
‘fetishism’.72 Rapid Christianization, political transformations, the
spread of biomedical practices and other factors of change during
the period in focus here have led to a current situation where most
of the old cults have been abandoned. In recent decades, the nkisi
have almost disappeared as agents of disease. Many banganga, there-
fore, survive by reorienting some of their activities and, not least,
because of the continuous—and even increased—problem of kindoki,
which will be elaborated in chapters eight and nine.73

                               God the Creator

Among the Kongo, as well as among several other neighbouring
peoples, the name of the Supreme Being, or God the creator, is
Nzambi. The etymology of this name is unknown, but it signifies
someone who is higher, stronger and more powerful than other
beings. It also denotes something incomprehensible and mysterious
or, in short, divine. As mentioned previously, the name Nzambi can
be used not only when speaking about God but also with reference
to the most important nkisi spirits, who are close to the Supreme
Being. This is a way of honouring as well as inducing fear of them.
Nkadi a Mpemba is a name that became used by Christians as a des-
ignation for the Devil. It refers to something cruel or evil, and it
seems likely, as remarked by MacGaffey, that in pre-Christian thought
Nkadi a Mpemba, the ruler of bandoki, would have been the bad side
of Nzambi.74
   Father Butaye reported in 1899 that everybody knew the Supreme
Being and that he is the ultimate lord of life and death.75 In the

     Laman 1962: 71, 80; Jacobson-Widding 1979: 131; Widman 1979: 197 f.;
Lagercrantz 1983/84: 83.
     Marichelle 1910: 238.
     Janzen 1978: 147; Janzen 1979: 211; Lagercrantz 1983/84: 70; Dalmalm 1985:
91, 172, 202 f.
     MacGaffey 1986: 108. See further Laman 1917: 288, 292 ff.; Vansina 1975:
672; Jacobson-Widding 1979: 349 f.; Widman 1979: 81 ff. Cf. Proyart 1819: 143.
     Butaye 1899: 309.
                         kongo spirits or     NKISI                     119

Kongo area, as in other parts of Africa, God is associated with the
heavenly realm. Nzambi is the creator of heaven but also of the earth
and its beings.76 Considering the long history of Christian influence
in the Kongo area, it is particularly difficult to depict the pre-Christian
ideas and attributes of God. However, the following is an attempt
to render some of the main characteristics of Nzambi. Although the
Supreme Being is sexless, or may have both male and female per-
sonae, he is often referred to as ‘male’.77 Some authors have strongly
stressed the morally positive qualities of the Supreme Being, such as
goodness and righteousness, as well as superior characteristics like
omnipresence, omniscience and omnipotence. In order to stress that
God is the greatest, highest and most powerful, attributes like Mpungu
may be added to the name Nzambi.78
   Widman accentuates not only the unique greatness of God but
states also that Nzambi is actively involved in the lives of human
beings. In other words, Widman argues against the common idea of
Nzambi as a deus otiosus. The Supreme Being is not only a symbol of
justice who has provided laws but also punishes those humans who
break these laws.79 However, in Widman’s theologically oriented work,
the uniqueness and active role of Nzambi are exaggerated. As pointed
out by, among others, Widman’s compatriots Laman and Dalmalm,
the character of Nzambi is essentially ambivalent, although the emer-
gence of Nkadi a Mpemba reflects a dualistic tendency that has been
strengthened by Christian influence.80 Jacobson-Widding, another
compatriot of Widman, concludes that, although concerned with jus-
tice, Nzambi does not execute judgements.81 As shown above, ances-
tors and nkisi spirits have a role as guardians of morals. For Christian
Kongo people, however, God has largely taken over this role.82
   As in the Sukuma religion, the Supreme Being is not a major agent
of disease. However, illnesses resulting from old age and incurable
maladies can be caused by Nzambi.83 Paradoxically, the expression

     Laman 1917: 271; Jacobson-Widding 1979: 337; Widman 1979: 67.
     Widman 1979: 58; MacGaffey 1986: 79.
     See, e.g., Sadin 1910: 131; Jacobson-Widding 1979: 345 ff.; Widman 1979:
53, 57, 74, 76. The name Nzambi Mpungu is used by several other Bantu peoples
     Widman 1979: 62 f. See also, e.g., van Wing 1959: 302 ff.
     Laman 1917: 273; Dalmalm 1985: 66 f.
     Jacobson-Widding 1979: 347.
     See also Dalmalm 1985: 175.
     van Wing 1959: 231; Mahaniah 1982: 82.
120                           chapter five

‘illness of God’ (kimbevo kia Nzambi) may refer to benign afflictions
with brief duration. Yet such afflictions have natural causes and do
not involve spiritual beings in the treatment process. Nor do they
concern problems of social relations.84 There are various versions of
the myth that death was caused by disobedience, usually of a woman.
Also, before human beings disobeyed God, heaven could be reached
by them.85 When somebody dies, it may be said that Nzambi took
him or her. Deaths caused by him are usually ‘natural’ deaths of
old people. If God has decided to end somebody’s life, nobody can
do anything about it. He can ‘eat’ the soul of a human being.86 In
general, unexplained events calling for no specific action can be
attributed to him.87
    Since Nzambi is not very actively involved in the lives of human
beings, his cult is quite limited. Some older works indicate that there
was no cult at all, neither public nor private.88 In general, the belief
in and cult of God is seldom mentioned in the old material of mis-
sionaries. This may reflect a theology of discontinuity. As remarked
by MacGaffey, ‘the statement, often made, that there was no cult
of Nzambi is not entirely true’, although there are no special priests
and sacred objects of God in the indigenous Kongo religion.89 Widman
exaggerates the occurrence of prayers and sacrifices to Nzambi,90 but
as a force in human affairs he may serve as a last resort. When all
else fails, prayers can be addressed to him. Nzambi is universal, and
his cult, if thin and residual, is also universal. His only special respon-
sibility is death itself, ‘the eventual goal of all careers, the universal
merging of all particularities’.91 It should be noted, however, that
while the importance of ancestors and nkisi spirits gradually decreased
during the period studied here, the significance of Nzambi instead
tended to increase.92

     Buakasa 1973: 148; Janzen 1978: 8; Dalmalm 1985: 77. Concerning the ill-
ness category kimbevo kia Nzambi, see further the appendix.
     For details, see Widman 1979: 71 ff. See also, e.g., Laman 1917: 291 f.;
Jacobson-Widding 1979: 338.
     Laman 1917: 281. See further Butaye 1899: 309; Laman 1917: 274, 283; van
Wing 1959: 252; Buakasa 1973: 148; Widman 1979: 69 f.
     MacGaffey 1986: 78.
     See, e.g., Sadin 1910: 131; Laman 1917: 272; Boucher 1918: 143.
     MacGaffey 1986: 78. See also Widman 1979: 133.
     Widman 1979: 77 ff.
     MacGaffey 1986: 78 f. Cf. Widman 1979: 87.
     MacGaffey 1986: 102.
                                 CHAPTER SIX

                            YORUBA DIVINITIES

                            The extraordinary richness of Yoruba religion lies
                            in the profusion of its orisa, in the facility with
                            which in the past an orisa has formed and gath-
                            ered about itself a cult-group.1
                            In practice, as far as my observations go, people
                            do not pay much attention to their Orisas except
                            in times of misfortune and sickness.2


Of all the religions studied here, Yoruba religion is the most com-
plex. There are, in addition to the hundreds of divinities, orisha, in
the pantheon, ancestors and other spirits, traditions of sacred king-
ship, belief in reincarnation and, in the cultic domain, various priest-
hoods, elaborate rituals and festivals as well as sacred places and
buildings. For a long time this exceedingly rich and varied religion
has evoked interest among a great number of scholars in religious
studies, anthropology and several other disciplines.
   In works by Nigerian scholars of religion the influence of E.B.
Idowu is often clearly discernible. According to Idowu (1963), his
disciple J.O. Awolalu (1979) and some others, the several hundreds
of divinities, orisha, in the Yoruba pantheon are intermediaries between
God and humans. Hence God is the ruler, and the divinities are his
ministers or servants. Each of them is governor of a certain depart-
ment and constantly controlled by God.3
   An influential anthropological model of Yoruba religion has been
presented by Peter Morton-Williams (1964). Based on material from
the Oyo area, he produced a threefold model of what he called the
‘Yoruba cosmology’. In this model, particularly the ‘male’ God,

      McKenzie 1976: 197.
      Prince 1964: 95.
      Idowu 1963: 61 f.
122                             chapter six

Olorun, and the divinities are associated with the heavenly realm,
while the ‘female’ earth goddess, Onile, and some spirits have their
domain in the earth. Assisted by deities, God created the world, the
middle zone, where human beings live and are influenced by spiri-
tual beings that pass freely between different zones. According to
Morton-Williams, thus, the Oyo Yoruba cosmos is made up of heaven
and earth enfolding an island-like world.4
   Morton-Williams referred to certain oral texts such as myths and
praise songs as evidence for his ‘new model of the Yoruba cosmos’.5
As remarked by several scholars, however, there is no evidence to
show that this abstracted model is a shared cosmology.6 Different
cult groups have different ‘cosmologies’ or ‘theologies’, and the Oyo
representation rendered by Morton-Williams is not found in the same
form among other Yoruba-speaking peoples. There is among the
Yoruba much variability between groups and individuals as well as
between regions or towns.7 E.T. Lawson, for example, writes about
two parts of the cosmos, heaven and earth, rather than three, but
fails to specify which Yoruba peoples he refers to.8 In his unduly
generalized scheme, the earth is the abode of human beings, ani-
mals and people who indulge in practices of witchery. The earth
goddess is not included in this model.
   Although the role of this deity undoubtedly varies a great deal,
Morton-Williams is probably right in criticizing several scholars of
religion for underestimating her importance in Yoruba religion. That
does not necessarily mean, however, that the hierarchic model of
Idowu, Awolalu, Lawson and others is inadequate. It certainly appears
to be influenced by Christian theology and thinking.9 Yet it is inter-
esting to note, for instance, that several of the CMS missionaries,
who did not champion a theology of continuity, in the late nine-
teenth century reported that people conceived of the divinities as
servants of God.10 About 150 years ago, when Crowther prepared
his Abeokuta mission, he was informed by orisha worshippers in
Freetown that the deities were inferior beings who were commis-

      Morton-Williams 1964: 244 ff.
      Ibid., 243.
      E.g., McKenzie 1976: 198 f.; Hallgren 1988: 12.
      Bascom 1969: 97; Eades 1980: 118; Buckley 1985b: 187.
      Lawson 1984: 57. Cf. Eades 1980: 121.
      Westerlund 1985: 34, 44–60.
      E.g., Crowther 1892: 10; Johnson 1899: 45.
                                yoruba divinities                                  123

sioned by God to superintend various matters on earth.11 His theology
of discontinuity was clearly reflected in his views on the powerless-
ness and ‘vanity’ of the deities of the ‘heathens’.12 Hence the hier-
archic model of Yoruba religion, which corresponds to the hierarchic
political and social system, has been presented by many authors with
different biases and must, therefore, be considered well documented.13
    In general, the possible influence of Christianity and Islam on
Yoruba religion should not be forgotten. It is often difficult to draw
a dividing line between indigenous Yoruba and Christian or Muslim
beliefs and practices; and the Yoruba have a long-standing tradition
of co-operation and ‘sharing’ across religious borders in, for exam-
ple, religiously mixed families. In the case of the ‘heathen’ infor-
mants of CMS missionaries in the nineteenth century, however, it
is very unlikely that there was any considerable Christian or Muslim
influence. A problem with the hierarchic model, as with the trian-
gular model of Morton-Williams, is that it is too static and tends to
make Yoruba religion appear more systematic and coherent than it
is.14 Although there are certain regularities in the diversity, it must
be remembered that ‘Yoruba religion, no less than the rest of Yoruba
life, is a field for discussion and debate, and not at all one for dog-
matic unanimity’.15 J.S. Eades is probably right in concluding that
Yoruba religion is less concerned with ‘a systematic and logically
coherent set of beliefs’ than ‘with the problems of the individual in
this world’.16
    Like Sukuma and Kongo, Yoruba usually explain serious and pro-
tracted illnesses by referring to suprahuman or human causes.17 God,
ancestors and other spirits may be referred to as agents of disease.
However, the most important of the spiritual causes of disease are
the divinities. Thus it is the latter category of spiritual beings that
will be in particular focus in this chapter.

      Crowther 1892: 10.
      Ibid., 28.
      Some early writers compared the orisha to the saints of Christianity (Verger
1966: 27 f.). See further, e.g., Johnson 1921: 26; Awolalu 1970: 24; Dopamu 1977:
101; Simpson 1980: 2; Larsen 1983: 25 ff.; Olukoju 1997: 1.
      This criticism is applicable, also, to the generalization of Beier (1980: xi), who
claims that the deities are ‘aspects or facets of the same divine force’ rather than
subordinated messengers of God.
      Buckley 1985b: 187.
      Eades 1980: 128.
      E.g., Dopamu 1977: 392; Ayoade 1979: 49; Osunwole 1989: 10 f.
124                               chapter six

                                 God the Creator

Among the several Yoruba names for God the most important ones
seem to be Olodumare and Olorun. There is some uncertainty
regarding the meaning of the word Olodumare, although it is often
translated ‘Almighty’.18 In Idowu’s classical study of Olodumare, this
name is said to indicate that God is the head or lord of all in heaven
and on earth, absolutely unique and beyond comparison.19 The name
Olorun, which owes much of its popularity to Christian and Muslim
influence, is a more common name for God. Yoruba Christians and
Muslims have adopted this name, which can be translated ‘owner
(or lord) of heaven’.20
   The attributes of Olorun are similar to the attributes associated
with God among the other peoples studied here. In general, it seems
that there are more differences concerning the extent to which God
is the object of (direct) cult in different African religions than in terms
of ideas about his nature. Hence God is thought of as creator, the
Supreme Being who is immortal and unchanging. Particularly scholars
who are influenced by a Christian theology of continuity have couched
Yoruba conceptions of God in terms like omnipotence, omniscience
and omnipresence,21 but such concepts may also be found in earlier
and theologically more conservative works.22 Apparently, there is some
Christian as well as Muslim influence on current Yoruba concepts
of God, although it is difficult or impossible to establish to what extent
and in what way this influence may have altered older ideas of God.23
   Concerning the cult of God there may also be some Christian
and Muslim influence. Early sources indicate that God was not wor-
shipped directly, even though he could be thought of as the ulti-
mate receiver of sacrifices and prayers.24 According to Farrow, God
is too exalted to be approached with the familiarity shown towards
the divinities and too high and distant to be offered sacrifices and

      Table n.d.b.: 1; Awolalu & Dopamu 1979: 37 ff.
      Idowu 1963: 36.
      Parrat 1974: 6; Awolalu & Dopamu 1979: 39.
      Idowu 1963: 40 f.; Awolalu 1979: 14 f.; Awolalu & Dopamu 1979: 49 f.
      E.g., Farrow 1926: 27 f. Dennett (1910 [1968]: 67, 72), whom Farrow (1926: 5)
criticized for having an ‘anti-Christian bias’ and lack of knowledge of the Yoruba
language, held that at the time when he published his book (1910), ideas about
Olorun had already changed.
      Cf. Verger 1966: 40.
      E.g., Lakaumi 1885; Johnson 1921: 26. Cf. Table n.d.b.: 2 f.
                             yoruba divinities                            125

prayers.25 Apparently, there is still often little or no direct cult of
God.26 Yet there is evidence that, at least partly because of Christian
and Muslim influence, even adherents of the ‘traditional’ religion
now turn directly to God more frequently than previously, although
the direct cult of God may have been exaggerated by some scholars.27
   In criticism of other scholars, Idowu strongly emphasizes that moral
sanctions are not vested solely in other beings than God and holds
that God is the real source and norm of the moral values of Yoruba
religion.28 According to Awolalu, God punishes human beings for
evil deeds and rewards them for good deeds.29 Similarly, another
Nigerian scholar of religion, O. Olukunle, argues that illnesses and
other types of misfortune can be sent by God as punishment for
wrong conduct and that even a child may have to pay for the ‘sins’
of its parents.30 However, the idea of disease as a punishment for
moral wrongdoing must not be exaggerated. Only rarely, it seems,
is illness interpreted as a result of direct intervention by God himself.
   Without indicating any reasons why God may cause afflictions,
J.A. Ayoade reports that a disease from God ‘is identified by its per-
sistence, resistance to time-hallowed medication, and its ultimate
incurability’.31 The idea that there is no antidote for a sickness or
other misfortune, like barrenness, that is occasioned by God is also
stressed by N.H. Wolff.32 Accidental deaths or deaths of young peo-
ple may be caused by God, although that is only one of several pos-
sible explanations.33 In a sense, even ‘natural’ diseases may be thought
of as diseases from God. When afflictions are explained by references
to germs in the body of the sick person, the germs are not con-
ceived of as invasions from outside the body but rather as substances
placed by God within the body from birth.34 In that perspective, dis-
eases and death are, like health and life, ultimately from God.35

     Farrow 1926: 30.
     Morton-Williams 1964: 246; Bascom 1969: 79; Lawson 1984: 57.
     E.g., Idowu 1963: 142 f.; Awolalu 1979: 16 f.; Comstock 1979: 10 f.
     Idowu 1963: 151, 154.
     Awolalu 1979: 15.
     Olukunle 1979: 124 ff.
     Ayoade 1979: 49.
     Wolff 1979: 130.
     Table n.d.b.: 154 f.
     Buckley 1985a: 33.
     The significance of bodily germs as causes of disease has been stressed—and
probably exaggerated—by Buckley (1985a). See further the appendix.
126                              chapter six


The ‘earth-owner’, Onile, is a goddess or spirit of great importance
because she provides the crops, and her cult is very widespread.36
Onile, or simply Ile (the personified earth), receives special sacrifices
at the time of planting and harvesting, and as a conceptual coun-
terpart of God the Father, she may be addressed as ‘Mother’. While
Morton-Williams refers to Onile as the ‘Earth Goddess’ and says
that she is asserted by those Yoruba who worship her to be coeval
with God and, thus, to have existed before the other divinities, orisha,37
other scholars refer to this being as a particularly important spirit
of nature.38 The problem of the status of Onile need not be discussed
in detail here. As concluded by Hallgren, however, it is question-
able if Morton-Williams’s picture of a divine polarization, heaven/earth
and male/female, is common in Yoruba religion.39
    More important in this context is the observation that Onile is of
little or no significance as an agent of disease. She is a spiritual being
who is associated with health, fertility and life rather than with ill-
ness and death. In this respect Onile differs from several of the
divinities who are important agents of disease, barrenness and death.
All deities can cause illness and other kinds of misfortune, but some
of them are much more feared than others, while certain divinities
are hardly ever associated with disease and death. Before the role
of specific deities is studied, however, some general observations
should be made.
    In many works on Yoruba religion and culture, specific numbers
of the orisha have been presented. Such numbers may vary from a
few hundred to more than one thousand, although four hundred
and one seems to be the most commonly accepted sum.40 The last
number of this sum (i.e., 401) is a reference to the sacred ruler. As
remarked by Idowu, the numbers should be regarded primarily as
symbols of the plurality and indefiniteness of the divinities. In prac-
tice, there are great differences between regions, groups and indi-
viduals in terms of how many and which deities are worshipped.

     Ojo 1966: 168; Adeniyi 1984: 63.
     Morton-Williams 1964: 245 f. Cf. Bascom 1969a: 92.
     E.g., Awolalu 1979: 45; Adeniyi 1984: 63.
     According to Hallgren (1988: 66), ‘it seems more likely to be an esoteric and
philosophical vision’.
     E.g., Morton-Williams 1964: 246; Wolff 1979: 131; Simpson 1980: 1.
                            yoruba divinities                            127

   The divinities have links with nature and natural forces, that is,
with hills, rivers, winds and so on, but also with the human sphere.
They are associated with historical events as well as with cultural
and economic activities. Many divinities are also patrons of towns.
Like town deities—and unlike ancestors—deified human beings are
tied to localities rather than to descent groups.41 Most of the divini-
ties are worshipped locally, but the most important ones are known
and worshipped virtually all over Yorubaland. ‘The principal orisa
are each the head of a hierarchy of lesser (usually more localized or
specialized) orisa, much as both high officers of state and also vassal
kings head hierarchies of lesser officials.’42 As indicated by this quo-
tation, the parallels between the orisha hierarchies and the indige-
nous Yoruba political system are obvious and have been stressed by
other scholars too.43
   Physical and psychiatric afflictions, as well as barrenness and death,
can be occasioned by deities.44 For instance, some divinities are
important in that they cause and control epidemics.45 The role of
the deities as agents of disease is one of the reasons why human
beings enter into a cultic relationship with them. Important deities
have their own priesthoods, which are organized into hierarchies of
ranked offices. Temples or shrines are erected for the worship of
such deities, and the priests serve as mediators between the human
and suprahuman worlds. People may also maintain one room in
their houses in honour of one or more divinities, who may be wor-
shipped individually at home. Locality, profession and descent are
factors that determine which deities are worshipped, but worship is
also based on individual choice. Sacrifices range from simple liba-
tions to those that involve the killing of big animals. Prevention and
recovery from illness as well as protection from death are some
important motives for sacrifices. The head of the sacrificial animal
may touch a patient’s head and chest, while prayers are offered that
the disease will pass from the afflicted person to the animal. Such
petitions, and petitions for material blessings, are important elements

     E.g., Ojo 1966: 159 ff.; McKenzie 1976: 197; Lawson 1984: 63, 66. See also
Dennett 1910 (1968): 71; Lucas 1948: 119; Eades 1980: 119.
     Morton-Williams 1964: 245.
     E.g., Idowu 1963: 48 f.; Eades 1980: 119.
     Table n.d.b.: 3, 7, 155; Prince 1964: 95; Wolff 1979: 129.
     Morton-Williams 1964: 251; Awolalu 1970: 31.
128                               chapter six

of Yoruba prayers. The most elaborate type of rituals are the annual
orisha festivals, which may involve a large proportion of the population
of a town as well as cult members from elsewhere. At such occa-
sions possession and trance phenomena are common. Music, par-
ticularly in the form of drumming, and dancing are important elements
in festivals, and different deities have various rhythms. In the state
of trance and possession, the deities use the possessed worshippers
as media of communication. Pierre Verger compares the festivals to
theatrical performances or operettas, but with the ‘actors’ in a state
of trance.46
   The most significant and dreadful disease agent among Yoruba
divinities is Shopona. Yet, like other deities, Shopona is not exclu-
sively evil and destructive. Rather, he is ambiguous and gives fertil-
ity and life as well as diseases and death.47 A myth tells that Shopona
sprang from the body of Yemoja, the female deity of the Ogun
river.48 R.E. Dennett reported another tradition, which stated that
Shopona was a very wicked boy who once, after beating several of
his townspeople to death, was taken by his parents to a doctor who
taught him the use of bad and poisonous drugs.49 With these he
then killed even more of his fellow citizens. After his death he was
deified, and people started worshipping him. The earth is his ele-
ment, and he is sometimes referred to as the king of the earth.50
   Shopona and his entourage of minor spiritual beings are associ-
ated with smallpox and insanity.51 Smallpox is usually referred to
simply by this name, Shopona, but it may also be called ‘hot earth’
(Ileegbono).52 Here the term ‘smallpox’ includes a much wider range
of diseases than is included in this concept as used in a western con-
text. The diseases caused by the Shopona family include many fevers

      Verger 1969: 64. For documentation and more detailed information on the
cult of Yoruba divinities, see Johnson 1899: 38 f.; Farrow 1926: 105; Lucas 1948:
177; Morton-Williams 1960: 34; Idowu 1963: 116, 129–139; Verger 1963; Morton-
Williams 1964: 251 f.; Prince 1964: 95; Verger 1969; Prince 1979: 117 f.; Eades
1980: 120; Simpson 1980: 64 ff.; Lawson 1984: 55; Hallgren 1988: 58 ff.; Olukoju
1997: 9 ff.
      Maclean 1971: 38; Buckley 1985a: 133.
      Lucas 1948: 112; Ajose 1957: 270.
      Dennett 1910: 231.
      Simpson 1980: 37; Larsen 1983: 27.
      E.g., Leighton et al. 1964: 116; Buckley 1985a: 98. The name Shopona may
be used as a generic name for the family of smallpox deities or spirits (Prince 1964:
      Buckley 1985a: 99.
                               yoruba divinities                           129

(particularly with delirium), rashes, carbuncles, boils and psychoses.
In addition to psychoses, other psychiatric disturbances can also be
caused by Shopona. Apparently, several different kinds of serious or
fatal illnesses, as well as famine, may be associated with Shopona.
Occasionally, even less serious ailments such as chickenpox can be
attributed to this deity. Some Yoruba believe that if other deities
are offended, they send Shopona to punish the offenders.53
   Since Shopona disfigures and slays people, or makes them insane,
he is much feared. Because of the fear, people avoid using his real
name. Instead, names such as ‘Father’ and ‘Lord’ may be utilized.
Similarly, people frequently use euphemisms when describing the
afflictions caused by Shopona.54 He is depicted as a short-tempered,
cruel and irascible divinity.55 Shopona is such a dangerous deity that
he must be kept away from the towns. Of all Yoruba divinities, he
seems to be the only one who is hardly ever worshipped in towns.56
He is associated with the forest, and that is also where his sanctu-
aries have been erected.57 Since the worship of Shopona was for-
bidden by the British colonial regime, it became difficult to build
shrines to him and worship openly. More or less secretly, however,
some cult of the deity continued in Nigeria, and some worshippers
also went to Benin (Dahomey) to perform the annual festival as well
as some lesser rituals. Probably, there are still some sanctuaries in
Nigerian forests, and some people may have small shrines in their
houses.58 About three decades ago, Prince even concluded that
Shopona was ‘still very active in many parts of Yorubaland’.59
   The reason why the cult of Shopona was outlawed by the British
regime was that his priests were accused of deliberately spreading
smallpox and enriching themselves from the property of the victims.60
Allegedly, the priests disseminated the smallpox infection by means
of liquids or dried scabs.61 According to the reports by, among others,

       Prince 1964: 95, 105; Maclean 1971: 38; Buckley 1985a: 99, 131 f.
       Bascom 1969a: 92; Lucas & Hendrickse 1971: 35; Larsen 1983: 28.
       Simpson 1980: 37.
       Buckley 1985a: 130.
       Farrow 1926: 57; Lucas 1948: 113; Bascom 1969a: 92.
       Lucas 1948: 112; Maclean 1971: 39; Simpson 1980: 40.
       Prince 1964: 105.
       Farrow 1926: 57; Ajose 1957: 270.
       Maclean 1966: 132; Bascom 1969a: 91.
130                            chapter six

Dennett and Lucas, the priests had the following materials which
they used in spreading the disease:62
– a calabash containing some portion of the corpse of a smallpox
– a pot with black liquid collected from the corpse of the victim or
  made up of water with which the victim’s rashes had been washed
– a vessel of black powder compounded with dried scabs.
The liquid or the powder was, allegedly, thrown during night time
by the Shopona priests in front of the houses of the targeted victims.
   Naturally, the priests denied the charges levelled against them and
claimed that the purpose of the Shopona cult is to prevent the spread
of smallpox and to aid in the recovery of those suffering from the
disease.63 Yet burials of Shopona victims are performed secretly, and
without normal funeral rites, by the priests. Such burials take place
in the forest or bush.64 Like Shopona himself, his victims belong to
the wilderness, where they are isolated from healthy people and
treated by the priests. Various myths and taboos inculcate the impor-
tance of isolation, disinfection and other preventative measures. The
treatment also includes sacrifices to Shopona. When sufferers die,
their belongings may be handed over to the priests as fees for the
   During a smallpox epidemic, people avoid festivities with dancing
and drumming, probably because of fear that such activities would
attract Shopona into the town.66 When the annual ceremonies for
Shopona take place, worshippers sing, dance and beat the drum for
several days. At such occasions, moreover, offerings of food as well
as sacrifices of animals are made, and prayers are addressed to
Shopona. In such prayers, for instance, the divinity is asked to pre-
vent the death of children, wives and husbands until the next annual
ceremony and to give children to those who are infertile. Worshippers
may promise sacrifices should their prayers be answered.67

     Dennett 1910: 231 f.; Lucas 1948: 113.
     Bascom 1969a: 91.
     Idowu 1963: 98; Bascom 1969a: 91; Buckley 1985a: 103. At least formerly,
Shopona victims were buried at the shrines of the deity (Maclean 1971: 39).
     Ajose 1957: 270 f.; Prince 1964: 96.
     Bascom 1969a: 92; Buckley 1985b: 195 f. Cf. Idowu 1963: 98.
     Simpson 1980: 40 f.
                                yoruba divinities                                 131

   Like the cult of other deities, the cult of Shopona differs a great
deal from one place to another. Even in a single town there may
be several quite distinct cults of this divinity.68 Among other things,
various cult groups may observe distinctive taboos.69 However, cer-
tain taboos seem to be of particular significance. There are strong
taboos against benniseed (sesame) and palm-kernel oil. Benniseed is
one of Nigeria’s commercial exports, but little, if any, is grown by
the Yoruba.70 Palm-kernel oil is offered to Shopona only in delib-
erate attempts to antagonize him.71 People who break his taboos and
offend him are punished by the ruthless divinity, and his wrath may
even affect innocent people.72 According to Idowu, Shopona holds
absolute sway over the earth, and his will must be accepted.73
   When the deity possesses a worshipper, he is offended by people
who whistle and laugh.74 ‘The wind is what afflicts people with small-
pox, and sounds such as whistling and whispering, which remind
men of the wind, are believed to be dangerous when Sonponno(!) is
abroad.’75 The deity works like a whirlwind and can attack anyone.76
Hence he can victimize initiated as well as uninitiated persons. Some
people have Shopona in their lineage and if they neglect some ritual,
they may be taken ill.77 When others, who do not have Shopona as
a lineage deity, become victims of this agent of disease, they are not
necessarily initiated into the Shopona cult as part of their treatments.
R. Prince’s material, which concerns psychiatric afflictions, in particular,
indicates that it is only after there have been several recurrences of
an illness caused by Shopona that the babalawo ( priest, diviner) rec-
ommends initiation into the cult of the divinity.78
   Orunmila, or Ifa, is a deity who differs very much from Shopona.
Ifa is everything that Shopona is not, a quiet, peaceful god, con-
cerned with maintaining the culture, health and well-being of the

       Buckley 1985a: 98 f.
       Ibid., 187.
       Bascom 1969a: 91 f.
       Buckley 1985a: 108.
       Bascom 1969a: 91.
       Idowu 1963: 95, 97.
       Bascom 1969a: 91.
       Buckley 1985a: 112.
       Leighton et al. 1963: 104.
       Lucas 1948: 112; Simpson 1964: 96; Maclean 1966: 139.
       Prince 1964: 105. This observation apples to other orisha cults as well.
132                              chapter six

town and its inhabitants. Yet, at times, he transforms himself into a
vicious and vindictive god with properties precisely analogous to those
of Shopona.79 Like this deity and other orisha, thus, Orunmila is
ambivalent or ambiguous and may cause disease and other types of
misfortune, although he is more concerned with health than dis-
ease—hence a healer rather than an agent of illness.80 First and fore-
most, Orunmila is a deity of divination, but he is also a great source
or promoter of fecundity, prosperity and health.81 He is worshipped
in all parts of Yorubaland and is one of the most important divini-
ties.82 Awolalu believes that there is an Orunmila shrine in almost
every ‘traditional home’.83 As a deity of divination, he has great wis-
dom and power. People may consult him on virtually all occasions
when they are in trouble and before important actions are taken.84
   According to Bascom,85 the divination system known as Ifa pro-
vides the most direct access to God apart from prayer. Through Ifa,
the intentions of other orisha can be revealed too.86 In the very com-
plicated system of Ifa divination, palm nuts from a special palm tree
and stories or verses are used.87 The verses include incantations,
which are of vital significance in Yoruba medicine. A minor but
important part of medicines used to cure illnesses contain such incan-
tations. The hidden knowledge of a medical incantation reveals the
power of the ingredients utilized in a specific medicine. The incan-
tation shows how the visible or known qualities of medicinal ingre-
dients are useful in relation to a special problem.88
   Through Ifa divination, thus, the diviner or priest of Orunmila,
babalawo (father of secret things), who must learn a vast number of
verses by heart, reveals important secrets and mediates between
human and divine. The knowledge of Ifa divination stems from

      Buckley 1985a: 133.
      Ibid., 128 ff.; Simpson 1980: 7.
      Farrow 1926: 40; Idowu 1963: 78; Hallgren 1988: 30.
      Idowu 1963: 76; Awolalu & Dopamu 1979: 80.
      Awolalu 1979: 23 f.
      Bascom 1969a: 80; Awolalu & Dopamu 1979: 80; Simpson 1980: 7.
      Bascom 1969a: 80.
      Morton-Williams 1964: 248.
      There are many studies of Ifa divination available. For two classical works,
see Bascom 1969b and Abimbola 1976. Brief introductions can be found in, e.g.,
Johnson 1899: 19 ff.; Awolalu & Dopamu 1979: 147 f.; Simpson 1980: 73 f.; and
Lawson 1984: 67 ff. For some details on the palm tree of Ifa, see Buckley 1985a:
113 ff.
      Bascom 1969b: 61; Buckley 1985a: 98, 140, 144.
                             yoruba divinities                              133

Orunmila himself, and the babalawo transmits and interprets the wishes
of God and orisha to humankind and prescribes sacrifices.89 The deter-
mination of the correct sacrifice is necessary to secure a favourable
resolution of the disease or other problem confronting a client. A
sacrifice of some kind is almost always a part of the proscribed
actions after divination.90
   Closely associated with Orunmila is Eshu, a divine trickster.
According to Morton-Williams’s account,91 God has sent Orunmila
and Eshu into the world as ‘a pair of divine mediators’. E.B. Idowu
refers to Eshu as the right-hand divinity of Orunmila,92 and Bascom
uses the term ‘trinity’ when he writes about these deities and God.93
Eshu is an important extension of God’s power, a divine messenger,
and some traditions tell that he was the one who taught Orunmila
the secrets of divination. A small part of each sacrifice prescribed
through Ifa divination is set aside for him to ensure that he will
carry the rest to God, for whom most sacrifices are destined. Such
sacrifices are deposited at the shrines of Eshu. Sacrifices to the orisha
are made at their own shrines, but again something is set aside for
Eshu, so that he will not cause the client trouble.94
   Eshu is one of the most powerful deities, and, since he serves God
and the orisha by troubling people who offend or neglect them, all
kinds of diseases and other evils may be associated with him.95 The
idea of Eshu as a servant of God is stressed strongly by, among
others, E.O. Olukoje.96 However, whereas Orunmila may be referred
to as a being of light and revelation, designations such as ‘the being
of darkness’ and ‘the anger of the divinities’ refer to Eshu. He is
regarded as a spiritual being who magnifies petty misdeeds into dire
offences.97 Terms like dreadful, malicious and unpredictable point to
the evil and dangerous traits of his character, and in their prayers

     Bascom 1969a: 80; Lawson 1984: 68; Buckley 1985a: 112, 130.
      Bascom 1969b: 60; Lawson 1984: 69. The favourite sacrificial animal of
Orunmila is a she-goat (Bascom 1969b: 65). For more information on the types of
sacrifices that are performed in the cult of Orunmila, see Simpson 1980: 9 ff.
     Morton-Williams 1964: 248.
     Idowu 1963: 80.
     Bascom 1969a: 80.
     Dennett 1910: 78, 94 ff.; Bascom 1969b: 60, 65; Olukoje 1997: 1 ff. Cf. Lawson
1984: 60 f.
     Bascom 1969a: 79; Simpson 1980: 17.
     Olukoje 1997: 3.
     Dennett 1910: 95; Morton-Williams 1964: 248.
134                              chapter six

people often ask him not to trouble them with illnesses and other
problems.98 They may also pray to God for protection against the
anger of Eshu. Like other divinities, Eshu has his own worshippers
and priests, even though all worshippers must pay attention to him;
and, in the annual ceremonies, rituals of protection from evil are a
central theme.99
   With reference to the evil and destructive aspects of Eshu’s char-
acter, Christian Yoruba have compared him to Satan or the Devil.
However, that comparison is not adequate because Eshu is not an
entirely evil being. A trickster, he has evil and good qualities. Hence
he is, for instance, a creative bringer of health as well as an agent
of disease. His ambiguous or elusive character is reflected in the
many names, such as punisher and rewarder, that are used to por-
tray him.100
   Another important deity who may act in a destructive way, although
he has creative properties too, is Shango, the divinity who controls
lightning and thunder. According to some traditions, he is the brother
of Shopona.101 Like the latter, Shango is a powerful and dangerous
deity. Although the cult of him is widespread, it is of special significance
in the Oyo Yoruba area, where he was once a king before his ascen-
dance to heaven.102 Ifa divination may indicate that a sick person
should undergo initiation into the Shango cult for cure; and wor-
shippers pray and sacrifice to him to protect themselves from illness
and death.103 Unlike Shopona, however, he has little to do with sick-
ness. As the deity of lightning, he causes injuries and deaths rather
than specific diseases.104 With his ‘thunderstones’ he harms or kills
people and destroys their houses.105
   While Shopona priests have been accused of spreading smallpox,
Shango priests have been believed to have the power to direct light-
ning. When lightning has struck, and people have died as a result,

       Idowu 1963: 81; Bascom 1969a: 79.
       Bascom 1969a: 79; Simpson 1980: 19.
       Idowu 1963: 80, 83, 85; Lawson 1984: 60 f.; Hallgren 1988: 30 f., 33. See
also Dennett 1910: 78; Bascom 1969a: 79; and Olukoje 1997: 5 ff.
       Bascom 1969a: 91; Isola 1977: 120; Larsen 1983: 28.
       Idowu 1963: 90 f.; Morton-Williams 1964: 255; Simpson 1980: 20. Cf. Farrow
1926: 47.
       Simpson 1980: 25 f.
       Morton-Williams 1960: 35; Buckley 1985a: 139.
       Bascom 1969a: 84; Lucas & Hendrickse 1971: 35; Simpson 1980: 20.
                             yoruba divinities                             135

Shango priests arrange sacrifices, purification rituals and burials at
tremendous costs or confiscate property.106 Previously, the Shango
cult was a powerful corporation and not fully under the jurisdiction
of the local states outside Oyo. Succession to the highest ranks in
the cult was vested in certain, mostly royal, Oyo lineages; and other
Shango priests who were resident in various kingdoms had to come
to Oyo for the final stages of their initiatory training and to be
equipped with the paraphernalia of priesthood.107
   Just like other worshippers, the Shango worshippers have special
taboos to observe, and breaches of these may be punished by the
deity. Moreover, the moral expectations of Shango are very important.
He expects honesty, guards morality and castigates evildoers. Thus,
death by lightning may be interpreted as a punishment for serious
moral transgressions. Sorcerers, troublemakers and others who offend
Shango are in danger of being injured or killed by him.108 Idowu
writes about the actions of this divinity as a manifestation of God’s
wrath,109 and Isola concludes that he is ‘the sworn enemy of liars,
thieves and witches’.110 Consequently, Shango worshippers who have
been maltreated by other people may ask him to avenge the wrongs.111
   According to Hallgren,112 the morally dual nature of Shango, his
mother Yemoja and his wives, who are river deities too, is related
to their connection to water, an element that may give as well as
take life. Correspondingly, the duality of Ogun, another son of Yemoja
and the deity of iron, steel and war, is reflected in the constructive
and destructive capability or various uses of metal implements.113
Blacksmiths, hunters, barbers, circumcisers, taxi drivers and many
others may worship Ogun, whose cult is very widespread and pop-
ular. In addition to occupation, membership of certain lineages can
be a reason for worshipping this divinity.114

      Idowu 1963: 92; Simpson 1980: 20. Cf. Maclean 1971: 37.
      Morton-Williams 1964: 255. In a sense, every Oyo king, alaafin, was Shango.
Once on the throne, they incarnated this deity, and when they died, they were
deified and became Shango (Isola 1990: 97).
      Bascom 1969a: 84; Maclean 1971: 37; Isola 1977: 120.
      Idowu 1963: 89.
      Isola 1991: 95.
      Simpson 1980: 20.
      Hallgren 1988: 34 ff.
      Lawson 1984: 61 f.; Hallgren 1988: 36.
      Idowu 1963: 86 f.; Bascom 1969a: 82; Simpson 1980: 30. For some infor-
mation about the important role of Ogun at the present century, see e.g. Dennett
1910: 123 ff. and Farrow 1926: 51.
136                               chapter six

   A Yoruba can swear an oath by kissing a piece of iron in a court
or may make a covenant in the name of this god. An oath made
before Ogun is very serious and may be compared to a Christian’s
swearing on the Bible. A person who breaks such an oath can, there-
fore, be severely punished by the deity. Illness and even death may
follow.115 The breaking of taboos and negligence of cultic responsi-
bilities, such as failure to provide annual festivals, are other reasons
for punishment.116 For instance, Ogun can be the cause of injuries
and deaths in road accidents.117 Because of his association with metals,
drivers of motor vehicles often carry a representation of Ogun as an
amulet to prevent accidents and ensure their own safety.118 Hence
Ogun is also believed to provide protection. In annual ceremonies
the worshippers ask him to protect them from diseases, and he is
expected to grant them health, fertility and life.119
   As the mother of nature deities, Yemoja, a mighty water divinity,
has already been mentioned. She is greatly venerated and worshipped
almost all over Yorubaland. Numerous Yemoja sanctuaries are found
on the banks of the river Ogun. A provider of health and fertility,
Yemoja is one of the most important Yoruba deities.120 However,
she is associated with illness and death too. For instance, she can
cause stomach problems for those who offend her, or kill them by
drowning.121 In the old CMS material there is an interesting account
by Crowther.122 According to him, a celebrated female ‘impostor’,
backed by the chiefs and other prominent men in Abeokuta, had
drawn people away from the CMS dispensary. After performing cer-
tain ceremonies in the Ogun river, the ‘impostor’ said that all who
drank of the water would be healed from all sorts of afflictions.
Crowther was astonished to observe her success in terms of attract-
ing people to her ‘pool of Bethesda’. All kinds of sick people—lepers,
epileptics, deaf, dumb, blind and so forth—came at all hours of the
day to drink of the ‘healing water’. This account by Crowther is but

        Idowu 1963: 88; Simpson 1980: 29; Hallgren 1988: 36.
        Bascom 1969a: 83; Simpson 1980: 30.
        Bascom 1969a: 83; Simpson 1980: 30; Hallgren 1988: 36.
        Lawson 1984: 62; Hallgren 1988: 36.
        Simpson 1980: 30; Hallgren 1988: 37.
        Farrow 1926: 46; Ojo 1986: 165; Hallgren 1988: 33.
        Bascom 1969a: 69.
        Crowther 1855.
                               yoruba divinities                      137

one example of how important the belief in and cult of deities were
in the nineteenth century.
   The deity of the river Niger is Oya, one of the wives of Shango.
If worshippers offend or neglect her, she can afflict them with a
throat disease, which may be fatal. Like Yemoja, she can kill peo-
ple by drowning too.123 No less dangerous is Oshun, the deity of the
Oshun river and another wife of Shango. This goddess fights by
causing dysentery, stomach-ache and menopause, in addition to
drowning.124 Like Yemoja and other divinities, however, Oya and
Oshun are ambivalent and sources of health and fertility too.125
   Obatala (Orisha-nla, Orishala) belongs to an important circle of
‘white’ deities. Their whiteness is usually interpreted as a symbol of
purity and high morality.126 In their presentations of Obatala, who
is one of the divinities worshipped throughout Yorubaland, Idowu
and Awolalu use the term ‘arch-divinity’.127 He is a vice-regent of
God on earth and has a special creative power. Above all, God has
appointed him to shape the forms of human beings. He is like a
sculptor who moulds babies, while life itself comes from God.128 As
a moulder of infants, Obatala is responsible not only for normal but
also for abnormal forms or appearances. For instance, hunchbacks,
cripples, dwarfs and albinos may even be regarded as sacred to him.
They become worshippers of him and remind people of his exis-
tence. Children born in a caul or with the umbilical cord wrapped
around the neck may also be sacred to Obatala.129 It seems, how-
ever, that abnormality has been interpreted in other ways too. Farrow
saw the existence of abnormal humans as signs of the displeasure of
the deity.130 According to information gathered by G.E. Simpson,131
abnormal characteristics of human beings are usually understood as
a way of punishing the mother for wrongdoing. For example, dis-
paraging talk about Obatala or violation of some food taboo may
be the cause. One of Simpson’s informants, a babalawo who was also

        Ojo 1966: 164; Bascom 1969a: 87 f.
        Prince 1964: 96; Bascom 1969a: 90; Simpson 1980: 27.
        Bascom 1969a: 90; Hallgren 1988: 34.
        Idowu 1963: 73 f., 81; Bascom 1969a: 81; Hallgren 1988: 28.
        Idowu 1963: 71; Awolalu 1979: 21.
        Dennett 1910: 83; Idowu 1963: 71; Simpson 1980: 3.
        Bascom 1969a: 81; Lawson 1984: 59; Hallgren 1988: 28.
        Farrow 1926: 43.
        Simpson 1980: 3.
138                               chapter six

a devotee of Obatala, added that the divinity can use worms to
attack people who displease him.132 First and foremost, however,
Obatala is associated with what Hallgren refers to as ‘the good things
in life’, that is, health, fertility and so on.133
   In many narratives Oduduwa (Odudua, Odua), another white
deity, is depicted as the wife of Obatala. However, both these divini-
ties may be thought of as either male or female. As the progenitor
of the Yoruba and the first ruler of Ile-Ife, from whom subsequent
kings descended, Oduduwa is an important deity. Today most Yoruba
seem to think of Oduduwa as male.134 Yet, despite his importance
as the founder of the royal dynasty, Oduduwa does not have a
prominent role as a divine agent of disease.135
   Less important, or more local, divinities may also act as agents of
disease and death. Here, however, only a few examples can be given.
Ibeji is the deity of twins and has a reputation for being cruel, trou-
blesome and stubborn. If parents of twins fail to appease him, he
may try to kill the children. Hence he is worshipped by such par-
ents, as well as by twins themselves.136 Egbe is a goddess whose main
concern is small children. She can attack such children when they
sleep, but she is also capable of healing children, and parents may
be advised to worship her.137 Nature deities, such as Olokun, the
great sea divinity, the river god Erinle and wind deities, are often
dangerous agents of disease and death.138 Two examples of divini-
ties in whose character the constructive and beneficial aspects pre-
dominate are Osanyin, the most important deity of medicine, and
Okebadan, the city divinity of Ibadan. Osanyin, who is closely
associated with Orunmila, is believed to possess more knowledge
than any other deity of the use of plant materials to cure illnesses.

       Cf. Buckley 1985a: 33.
       Ibid., 6; Hallgren 1988: 28 f.
       Dennett 1910: 73 ff.; Lloyd 1960: 223; Hallgren 1988: 29. See also Farrow
1926: 44 f.; Idowu 1963: 12, 15; Bascom 1969a: 80 f.
       In this context it is of interest to note that, unlike some other African peo-
ples with sacred kings, the Yoruba apparently did not believe that the health of
the sacred king affected the prosperity of the king’s town, and he was not killed if
he was seriously ill (Lloyd 1960: 228). Cf., e.g., Parrinder 1974: 69 and Olupona
1991: 59.
       Farrow 1926: 58; Simpson 1980: 44 f.
       Simpson 1980: 47.
       See, e.g., Dennett 1910: 110; Farrow 1926: 60; Lucas 1948: 151–174; Bascom
1969a: 88 ff.; Simpson 1980: 59; Larsen 1983: 30 ff.
                               yoruba divinities                                139

Among other things, Okebadan prevents evil spirits and pestilence
from descending on the city, and people expect healing from her.
Another divinity closely attached to Orunmila is Ela, who is often
invoked during worship to come and bless offerings to other divini-
ties. As a spiritual being of truth and rightness, Ela is opposed to
the evil works of Eshu.139
   According to Morton-Williams,140 the anger of the divinities is not
roused by moral shortcomings of human beings. As a rule, the mate-
rial presented here supports his conclusion. Rather than moral trans-
gressions, the breaking of taboos, neglect of rituals and failure to
placate the deities may rouse their displeasure and result in diseases
and other types of misfortune. The frequency and severity of the
appearance of the various divinities as agents of illness is much related
to their differing characters. Although they are all ambivalent, some
of them very seldom, if ever, act as such agents, whereas others,
most particularly Shopona, are very prone to cause afflictions and
other hardships.141
   Through disease, or some other misfortune, a divinity may indi-
cate that it is making demands on an individual or a group of peo-
ple. After the occurrence, or after several occurrences, of some illness,
a babalawo may recommend initiation into the cult of a certain deity.
Sometimes former patients are also instructed to join a cult to pre-
vent relapsing into the illnesses from which they previously suffered.142
However, a patient who, nevertheless, relapses or who is not healed,
despite the involvement in the cult of a certain deity, may appeal
to another one for relief. There is, in other words, a degree of choice,
and disappointed devotees can approach new divinities. In a sense,
like ‘Big Men’ in society, deities have a reciprocal relationship with
their followers and can be made ‘bigger’ or ‘smaller’ by the atten-
tion of their supporters.143
   As indicated by the quotation from a work by Prince at the begin-
ning of this chapter,144 divinities have lost much of their significance

      Idowu 1963: 103 f.; Simpson 1980: 16 f., 42, 56. Cf. Ogunsakin-Fabarebo
1998: 14.
      Morton-Williams 1964: 247.
      See further, e.g., Prince 1964: 105; Maclean 1966: 139; Bascom 1969a: 91;
Buckley 1985a: 188.
      Morton-Williams 1964: 251; Prince 1964: 105, 113.
      Barber 1981: 725, 731 f., 736. According to Barber (ibid., 725), this idea is
found in several West African cultures where roles are achieved rather than ascribed.
      Prince 1964: 95.
140                               chapter six

in recent decades. Prince concluded that it was only in times of ill-
ness or some other misfortune that people paid much attention to
their divinities. The findings of Simpson, who says that ‘the real test
of the strength of traditional beliefs comes in times of trouble’, support
Prince’s conclusions.145 In such times, according to Simpson, ‘a large
percentage of the Yoruba, educated and uneducated, consult a babal-
awo or other traditional leader for guidance’.146 Otherwise, the belief
in the powers of the divinities has decreased sharply, possession is
sometimes openly ridiculed, particularly by young people, and civil
officials no longer support ‘traditional rituals’ as they used to do.
Simpson’s material indicates that among those who still worship
Yoruba deities, highly educated people, men and people in cities are
   It appears that even in the area of disease causation the deities
are no longer a major factor. Although Buckley seems to have exag-
gerated the insignificance of religious factors, his conclusion that ill-
ness is generally not related to the activities of spiritual beings is by
and large supported by other scholars.148 In particular, Simpson con-
cludes that, so far as his ordinary informants are concerned, the only
orisha of importance in connection with illness is Shopona.149 When
ordinary persons are compared to healers, some difference can be
discerned. A healer is more likely to attribute diseases to religious
causes than is a non-healer. According to Simpson, this is so ‘because
of the healer’s greater familiarity with powerful forces, his greater
preoccupation with such forces, and his personal and professional
stake in healing and ritual’.150 However, in Prince’s study of 101 psy-
chiatric cases investigated at ‘native treatment centers’ as early as in
1961–62, it is shown that even the healers there referred much more
seldom to religious than to human and natural causes.151

       Simpson 1980: 144 f.
       Ibid., 145.
       Ibid., 81, 121, 133 f., 143 f. When asked by Simpson about their personal
attitudes toward the retention of the indigenous Yoruba religion, almost 50 per cent
of 271 interviewees in the Ibadan area said that they believed it would disappear
within a few decades (ibid., 145).
       Buckley 1985a: 98.
       Simpson 1980: 109.
       Ibid., 108.
       Prince 1964: 96.
                               yoruba divinities                                141

                      Ancestors and Other Spiritual Beings

In comparison to Sukuma ancestors, Yoruba ancestors are much less
important as agents of disease. While the divinities, orisha, figure very
prominently in the CMS material from the late nineteenth century,
ancestors are almost totally absent there. However, since spirits of
ancestors have played and still play a certain role as agents of dis-
ease, at least a brief account of their position in Yoruba religion is
needed here. What is generally referred to as ‘soul’ conceptions
among the Yoruba are very complex and varied. In view of differences
in terminology between various areas, it becomes particularly difficult
to generalize. Nevertheless, it seems possible to reproduce some rel-
atively consistent views.152
    A person is regarded as both a physical and a spiritual being. The
most important spiritual elements are usually referred to as emi (breath)
and ori or eleda (head), respectively.153 Emi is the vital principle, that
is, the power that gives life to the body and distinguishes a living
person from the dead. Whereas emi is associated with the chest and
lungs,154 ori refers both to the head and the brain, which enables
people to think, and to an ‘invisible and intangible’ entity. The ideas
of this entity are complex and partly contradictory, but it appears
that, at least according to some Yoruba, some part of it exists as a
sort of spirit double in a heavenly realm. In order to improve one’s
life on earth, one should give offerings to one’s own ori once a year.155
    Ori, furthermore, refers to the ‘destiny’ of a person, which is given
by God. Yet there is an element of flexibility in that an individual
is believed to choose between various alternatives, including different
characters or temperaments. Besides, sacrifices and the assistance of
divinities, particularly Orunmila, influence the content of life on earth
in a beneficial way. Conversely, the earthly ori, or destiny, may be
altered for the worse by evil beings, like witches, as well as by the
actions of human beings themselves.156 Hence it is not a question of

       The following account is based primarily on the works by Eades (1980: 121 f.),
Simpson (1980: 63 f.) and Lawson (1984: 67 f.). Some other works of interest are
Prince (1964: 93), Bascom (1969a: 71 ff.) and Olukunle (1979: chapter 6).
       Cf. Bascom 1969a: 71.
       Ibid. Cf. Olukunle 1979: 169.
       Particularly offerings of goats, fowls and sheep are made for that purpose
(Simpson 1980: 64).
       Idowu 1963: chapter 13; Bascom 1969a: 72 ff., 80; Hallgren 1988: 60 ff.
142                              chapter six

immutable predestination, even though the content and length of life
in some way and to a certain extent is thought of as predetermined.
   The attainment of a ripe old age may be seen as a good indication
that one has fulfilled one’s destiny. Concerning the whereabouts of
the dead in the hereafter, there are again varying beliefs. There are
some notions about the netherworld as a residence for dead humans.
More often, however, these are associated with the heavenly realm—
to some extent probably because of Muslim and Christian influence.
In particular, the importance of heaven as a residence for dead
humans has been emphasized, and perhaps overemphasized, by the-
ologically influenced Christian scholars like Idowu and Awolalu.157
The Yoruba word for the invisible world of spiritual beings, which
is often translated ‘heaven’, is orun. There is a ‘good heaven’, orun
rere, as well as a ‘bad heaven’, orun buburu (or buruku). According to
Idowu and Awolalu, among others, there is a final judgement, and
depending on the deeds of human beings on earth, God decides that
‘good’ people will enter the good heaven, while ‘sinners’ will be sent
to the bad one. Orun is generally believed to be near God and the
orisha. Some people believe that, for instance, a dead babalawo lives
with Ifa, while dead albinos live in the garden of Orishala.158
   However, to be near God and the divinities in the good heaven
is not a final goal. Belief in rebirth is an important feature of Yoruba
religion. This belief is reflected in the tradition of giving children
names such as Babatunde (father returns) and Yetunde (mother returns).
Yet some spiritual elements of reborn persons continue to exist in
the world of spiritual beings as well. Thus the term ‘partial reincarna-
tion’, suggested by Awolalu, may be an appropriate designation.159
In general, reincarnation, or rebirth, seems to be a much longed-for
goal. Yet, not all people reach that goal. Bascom reports that ‘cruel’
people and those who have committed suicide can never reincarnate.160
The moral dimension is stressed even more strongly by Awolalu,
who says that only the ‘good souls’ can become reincarnated, while

       Idowu 1963: 196 ff.; Awolalu 1979: 57 f.
       Bascom 1969a: 75 f. For more details on Yoruba conceptions of the here-
after, see Hallgren 1988: 67 ff. See also, e.g., Idowu 1973: 188; Dennis 1979: 15;
Lawson 1984: 62 f.; Beier 1999: 90.
       Awolalu 1979: 36. See further Buckley 1976: 411; Eades 1980: 122; Hallgren
1988: 62 ff.
       Bascom 1969a: 76.
                               yoruba divinities                                143

the ‘wicked’ ones are not allowed to reunite with their relatives on
    The notion about abiku souls or spirits is of special interest here,
since it may be referred to as an explanation of why a child becomes
sick and dies. If an abiku is born in a child on earth, it soon leaves
for heaven again, because it wants to reunite quickly with its heav-
enly playmates or companions; and when the abiku returns, the child
dies. If a woman gives birth to a succession of children who die in
infancy, it may be divined that an abiku is at work. In such a case,
special rites are performed when the next child is born in order to
break the attachment of the abiku to its heavenly companions. Abiku
children are given special names, like Aiyedan (‘life is good’)—imply-
ing that the child should stay to enjoy it. Normally, the circumci-
sion and scarification of an abiku child is postponed until it appears
likely that it will survive.162
    Apart from offerings to one’s own ori, or ‘spirit double’, offerings
to humans are mainly, if not exclusively, directed to one’s parents.163
According to Bascom,164 children sacrifice to their dead parents at
their graves on the day that these had sacrificed to their own ‘heads’
when alive, whereas sacrifices to grandparents are made only when
a diviner or priest, babalawo, for some particular reason tells the
grandchildren to make such a sacrifice.165 In a sense, ancestors are
still regarded as family members, and they continue to influence the
lives of their own families on earth.166 If properly served, they have
powers for their lineage members and their wives of bestowing chil-
dren, health, protection and prosperity.167
    In general, they are considered benign,168 although they may cause
illnesses and other types of misfortune too. In Prince’s interviews

       Awolalu 1979: 59.
       Morton-Williams 1960: 35; Leighton et al. 1963: 80; Eades 1980: 122 f.
Oladapo reports in a study (1984: 110 f.), which is based on interviews with about
200 healers in the Ile-Ife area, that some illnesses that are very difficult to diag-
nose and name may be interpreted as effects of diseases suffered in a previous life.
       Forde 1951: 29; Morton-Williams 1960: 37; Simpson 1980: 64.
       Bascom 1969a: 72.
       In accordance with his general emphasis on the moral dimension, Awolalu
(1970: 25) states that only ‘good ancestors’ are remembered and invoked. Like divini-
ties, these ancestors are ‘intermediaries’ or means of reaching God himself.
       Dopamu 1977: 103.
       Morton-Williams 1964: 247; Awolalu 1970: 27.
       Barber 1981: 729.
144                               chapter six

about the causes of psychiatric illness, ancestors were never men-
tioned. Sometimes, however, physical diseases, as well as ‘unrest of
mind’ and lack of prosperity, were attributed to them.169 It appears
that there are often moral reasons for afflictions sent by ancestors.
Their role as guardians of morality has been stressed not only by
theologically inspired scholars like Awolalu,170 but also by more sec-
ularly oriented anthropologists like Morton-Williams, who concludes
that the ancestors are concerned with the good reputation of their
descendants.171 Moral shortcomings of descendants rouse the dis-
pleasure of ancestors. Other reasons why ancestors cause illnesses
may be failure to sacrifice and lack of proper burials.172
   Funeral ceremonies and rituals should be performed as soon as
possible so that the dead person does not cause illness or some other
misfortune. Although the very elaborate funeral rituals cannot be
studied in detail here, it should be stressed that many of them are
intended to ensure that the deceased will be born again. In the pre-
ceding section it was noted that people who die of smallpox or by
lightning are buried in special places by the priests of Shopona and
Shango respectively. Other people who may be buried in sacred
groves of specific divinities are lepers, albinos, hunchbacks, women
who die in pregnancy and others who have met death in special
ways. In normal cases, a deceased is buried in a grave that is dug
under the floor of or outside his or her house.173
   In Yoruba history, cult groups such as egungun, ogboni and oro have
had important political, social and judicial functions.174 The egungun
masquerades represent ancestors but assume a role that cuts across
descent-group boundaries. Prominent men with lineal descendants
may hope to have egungun masks bearing their names and to have
songs sung in their praise, while masqueraders robed from head to
foot in a variety of dresses parade through a town. Important male

      Prince 1964: 94.
      Awolalu 1970: 25.
      Morton-Williams 1964: 247. See also, e.g., Horton 1971: 41.
      Prince 1964: 94.
      Bascom 1969a: 66 ff.; Awolalu & Dopamu 1979: 258–263; Adeniyi 1984: 70.
Bascom (ibid., 66 f.) gives several examples of precautions that must be taken. When
bathing the corpse, for instance, the head must not nod at anyone, lest that per-
son will die too. The fine clothes are put on backward so that the deceased will
know its way back to earth when the time for rebirth comes; and, in order to avoid
the dead person returning as a leper, no cloth with red colour can be used.
      See, e.g., Farrow 1926; Forde 1951: 18; Morton-Williams 1964: 252.
                                yoruba divinities                                  145

spirits of ancestors are ritually present in these masked dancers. There
is a dual significance of the egungun cult in that it involves a com-
memoration of individual ancestors as well as representing the col-
lective dead acting on behalf of the community as a whole on special
occasions such as festivals.175 Simpson reports that, unlike ordinary
ancestors, those who are represented in the egungun cult may afflict
living humans not only with physical ailments but also with mental
illness. Some people believe that egungun spirits function as messen-
gers for dangerous deities in a malignant way too.176 There are
different classes of egungun, and the elder and most powerful one is
also the most feared.177 Egungun masks must not be touched by unini-
tiated people, particularly women, and violation of that taboo may
be considered extremely dangerous.178 In general, however, the atti-
tudes towards egungun spirits, like the attitudes towards other ances-
tors, are characterized as much, if not more, by love and affection
as by fear or awe.179
    The ogboni used to be the politically most important cult group or
secret society. The members of this society, mainly highly influential,
elderly men, worshipped Ile, the earth goddess, whom they regarded
as superior to all orisha or divinities. By settling civil disputes, deal-
ing with criminal charges and curbing the power of kings and chiefs,
they were supposed to maintain law and order. Sometimes ances-
tors are associated with Ile and her earthly realm, but there are also
certain vengeful spirits, who seem to be regarded as manifestations
of her power, who may punish misdeeds on the part of the members
of the ogboni. During the initiation ritual a new member must swear,
among other things, not to reveal the secrets of the cult. Violation
of this and other taboos may cause disease and even death.180

       Morton-Williams 1960: 36; Eades 1980: 123 f.; Lawson 1984: 63. See also
Beier 1999: 85–104.
       Prince 1964: 95.
       Bascom 1969a: 93; Awolalu & Dopamu 1979: 230; Beier 1999: 93.
       Formerly, it was believed that such a violation could even result in death
(Hallgren 1988: 56; Beier 1999: 94).
       Eades 1980: 123; Hallgren 1988: 56. As will be shown in chapter 8, the egun-
gun cult, as well as other cults studied here, have important functions in the strug-
gle against anti-social activities. Some senior members of the egungun cult are believed
to be able to identify witches, who among the Yoruba are almost invariably women.
       For documentation and more information on the Ogboni society, see e.g.
Dennett 1910: 32; Morton-Williams 1964: 245, 248; Awolalu & Dopamu 1979: 225
ff.; Comstock 1979: 7; Simpson 1980: 60 f.; Hallgren 1988: 64 f.
146                              chapter six

   The oro society is another secret association, which previously was
politically powerful in south-western Yorubaland, especially among
the Egba in Abeokuta. Oro is a deity, but the members of this purely
male society invoke and sacrifice to spirits of ancestors too. For
instance, when there is an epidemic, the ancestors are summoned
and sacrifices are made at special groves.181 The oro cult is partly
intended to check the spread of diseases and to deal with evildoers
in the spiritual realm as well as among living humans.182 A partic-
ularly important ritual object in the oro cult is the bull-roarer, which
has an awe-inspiring sound and which is highly dangerous for wrong-
doers and uninitiated people.183
   While the male element is predominant in the egungun, ogboni and
ori cults, the female aspects are more pronounced in the gelede of
western Yorubaland. This is another masquerade cult in which men
masquerade as women, some of whom look fat and grossly preg-
nant.184 Although gelede members seem to be concerned primarily
with the issue of fertility, men join the cult because of their own
impotence or because their wives are barren. Diseases caused by
witchery may be a reason for joining gelede too.185 The masks rep-
resent dead ancestors active in the service of the living, bringing the
latter fertility and health. However, if an initiate refuses to dance
under the mask or defects from the cult, he might well be punished
with illness.186
   In recent times the significance of the ancestors, as well as of soci-
eties such as egungun, ogboni, oro and gelede, has waned a great deal.
Simpson, for example, concludes that belief in the powers of not
only the orisha but also the ancestors ‘has decreased sharply’.187 About
three decades ago, it was reported that among the Egba in the

      See, e.g., Lucas 1948: 120 ff.; Morton-Williams 1960: 37; Awolalu & Dopamu
1979: 227 ff.
      Like Shopono and Shango priests, the priests of Oro may be called to make
atonement after certain deaths. Bascom (1969a: 93) reports that this happens when
a man who has suffered from elephantiasis of the testicles dies.
      Morton-Williams 1964: 256; Simpson 1980: 53 f.; Adeniyi 1984: 57.
      Both egungun and gelede are associated with a deity called Amaiyegun who
taught people how to make and use the costumes that mask their wearers (Bascom
1969a: 94 f.).
      Gelede herself was a witch when she lived on earth, and members of the cult
propitiate witches by sacrifices.
      Prince 1964: 109; Bascom 1969a: 95; Hallgren 1988: 57 f.
      Simpson 1980: 81.
                             yoruba divinities                              147

Abeokuta area lineage obligations were increasingly neglected and
that village-oriented deities were more important than lineage ances-
tors.188 The ogboni and oro societies have lost much of their influence
to modern institutions like courts and local administrations; and they
are now largely defunct. In 1914 a revised ogboni association called
the Christian Ogboni Society was organized, the name of which was
later changed to the Reformed Ogboni Fraternity, but this associa-
tion does not attract much support either.189
   In addition to divinities and ancestors, there are some other spir-
itual beings. These spirits, ebora, lack institutionalized priesthoods and
are more local and less powerful than the deities, orisha.190 Apparently,
the spirits are principally associated with natural phenomena and,
like deities and ancestors, they are ambivalent. Although, in general,
such spirits do not play an important part in human affairs, they
may occasionally be thought of as agents of disease or some other
misfortune.191 For instance, whirlwind spirits and forest spirits can
harm people.192 Some scholars write about ‘evil spirits’, anjonu, that
may be dreaded and can cause mental as well as physical illness.193
According to Wolff, these spirits are thought of as a kind of evil
spirits ‘responsible for bringing disease and misfortune to people’.194
   Like other African peoples, Yoruba conceive of ‘health’ as an
inclusive entity. It is not simply a physical, mechanistic absence of
symptoms.195 The Yoruba term alafia, which sometimes is translated
‘health’, is the sum total of all that is good and that humans may
desire—‘an undisturbed harmonious life’.196 According to Awolalu,
alafia is very similar to the Hebrew concept of shalom, which refers
to those things that make for harmony, joy and wholeness in all

       Leighton et al. 1963: 51, 220. Cf. ibid., 289 f.
       Awolalu & Dopamu 1979: 227; Simpson 1980: 149.
       Adeniyi 1984: 84 f.; Buckley 1985a: 98.
       Morton-Williams 1964: 245; Awolalu 1979: 45; Barber 1981: 729.
       Ojo 1966: 172; Dopamu 1977: 635.
       See, e.g., Wolf 1979: 127 and Oladapo 1984: 144, 174.
       Wolff 1979: 130. Sometimes the terms spirits and divinities are not clearly
distinguished. Simpson (1980: 103) reports, for instance, that smallpox is widely
attributed to ‘meeting evil spirits’, that is, to meeting Shopona.
       Lowery-Palmer 1980: 214.
       Awolalu 1970: 21.
148                               chapter six

respects. ‘From the Yoruba point of view, alafia is incomplete or dis-
rupted when there is no totality about it.’197
   In this chapter, as in the preceding ones, the significance of the
relationship of human beings to God, divinities, ancestors and other
spiritual beings for health and well-being in general has been stud-
ied. The following chapters of the book will deal with the impor-
tance of the interrelationship between living humans.

     Ibid., 22. Cf. Buckley (1985a: 66) who, in his more ‘naturalistic’ account of
Yoruba ideas of health and illness, refers to health as an ordered structure of the
body. For more information on the views presented by Buckley, see the appendix.
                              CHAPTER SEVEN


                          The role of ‘witch’ is not an institution of Bushmen
                          Maasai religion is . . . characterized by a general
                          lack of belief in witchcraft and other forms of ‘super-


The lack or paucity of living human agents of disease is a charac-
teristic feature of hunting-gathering San as well as of pastoralist
Maasai. This is amply documented in older as well as in more mod-
ern sources. As a rule, it is only among San and Maasai who, usu-
ally in recent time, have been influenced by other peoples that living
humans have become significant as agents of disease.
   Regarding San in general, Guenther says that ‘social diseases’,
mainly witchcraft and sorcery, were introduced by the black peo-
ples, and that the San’s own ‘religious system’ has ‘always been
devoid’ of these ‘black arts’.3 It is not surprising to find that Father
Schmidt concluded that ‘Schwarzzauberei’ was of little or no signifi-
cance among Kung and other San groups.4 However, his conclusion
has been supported by other older scholars who do not seem to have
been influenced by his theory of Urmonoteismus;5 and, in works from
the most recent decades, other scholars have stressed the lack of
indigenous conceptions of witchery too.6

     Lee 1967: 37.
     Berg-Schlosser 1984: 170.
     Guenther 1975/76: 47.
     Schmidt 1933: 609.
     E.g., Maingard 1937: 285; Esterman 1949: 31; Drobec 1953: 135. Cf. Passarge
1907: 109.
     Marshall 1962: 249; Silberbauer 1972: 320; Woodburn 1982: 203. Cf. Smith
et al. 2000: 86.
150                             chapter seven

   Among pastoralist Maasai, human agents of disease and death are,
and have been, of slightly greater importance than among hunting-
gathering San. However, these agents are primarily, or almost exclu-
sively, healers and elders who may use their special power to inflict
harm on other people. For this purpose they use the method of
cursing or certain witchery practices. As will be seen, there are impor-
tant differences between these human agents and the ‘witches’ and
‘sorcerers’ found among the agriculturalist peoples that will be stud-
ied in the following chapter. Like Lee in the quotation above, Marshall
says that the Kung do not have ‘sorcerers, witches or witch doc-
tors’.7 This has been said about pastoral Maasai society as well,
although to some Maasai, people with ‘evil eyes’, who phenomeno-
logically or functionally may be compared to ‘witches’, do exist.


It has been remarked that certain features of the healing systems of
hunting-gathering peoples such as San are strikingly similar to witch-
craft phenomena.8 The ability to travel out-of-body, seize souls,
change shape into animal form and handle fire are some examples.
In societies where witchcraft is a prevalent problem, as will be
exemplified later, there may be a belief that not only witches but
also, for instance, healers and chiefs, are able to perform such supra-
normal feats. As far as belief patterns are concerned, San have in
their ideas of the supranormal power called num (Kung) or tsso
(Nharo), as it were, a seed for the development of witchcraft should
the conditions conducive to such development arise. Although the
power of num is normally used for healing purposes, it is an ambiva-
lent power, like God himself, from whom it ultimately derives, and
can thus cause diseases as well as provide health.
   Katz reports a healer called Toma Zho, ‘a healer in transition’,
once in a trance dance pointing his finger across the dance fire at
another dancer, who immediately fell over. By doing so, Toma Zho
caused his num to become dangerous, a ‘death thing’.9 Even from
earlier periods there are, in fact, examples of healers who have used
their power for negative or evil purposes. Like spiritual beings, healers

      Marshall 1962: 249.
      Zuesse 1985: 233.
      Katz 1982: 178, 263. Cf. Guenther 1975: 164.
                                  living humans                                   151

have sometimes been thought of as being capable of shooting small
invisible or visible arrows, thus causing diseases.10 In some accounts,11
the supranormal ‘weapon’ of the healers has been referred to as the
‘Bushman revolver’. Writing on the Nharo, Barnard confirms Bleek’s
information about ‘grass arrows’ but adds, again in conformity with
Bleek’s description, that accusations of ‘evil medicine’ within the com-
munity are rare.12
   Another means by which a healer has been believed capable of
harming other human beings is by transforming himself or herself
into a dangerous animal. While, at least for the southern San, the
antelope was the symbol of the socially beneficial healer, the lion
symbolized the antisocial possibilities of trance.13 In the past, accord-
ing to Lee, San regarded a lion that had attacked a human being
as a human healer-turned-lion.14 However, since such attacks are
very rare, there is little reinforcement of the belief in the malevo-
lence of trance dancers.
   Although Barnard may be right in criticizing Guenther for being
too categorical in saying that it was not until recently that witchery
was introduced among San,15 some other scholars appear to have
exaggerated the historical significance of this problem. Since it is
particularly in older sources, written by scholars such as Vedder,
Seyffert and Dornan, that witchery is said to be important,16 it could
be asked whether it was indeed a more prevalent problem at the
beginning of the twentieth century than in recent decades.17 In view
of the contradictory evidence found in other and usually more reli-
able sources, it seems more likely, however, that older scholars like
Vedder and others were not well informed about this issue.

      E.g., Bleek 1928: 28; Schapera 1930: 196.
      E.g., Vedder 1937: 430; Wilhelm 1954: 166.
      Barnard 1979: 71. The ‘grass arrows’ are imitations of real arrows. They are
made of grass and have small pieces of horn attached as arrowheads. When Nharo
say that such arrows ‘kill’ (ku), the phrase is not to be taken literally. Rather, they
cause a crick in the neck. Cf. Bleek 1928: 28.
      Holm 1965: 68 f.; Lewis-Williams 1981: 97.
      Lee 1967: 35. Cf. Heinz 1975: 29.
      Barnard 1979: 70 f.
      E.g., Vedder 1912: 411; Seyffert 1913: 204; Dornan 1925: 144, 153.
      The statements of the authors referred to in the preceding note that most or
virtually all deaths were believed to be caused by witchery or ‘Zauber’ can be com-
pared, for instance, to the report by Woodburn (1982: 203) who says that Kung,
among others, show little or no interest in the causes of death and repudiate the
idea that death is caused by the supranormal actions of other members of their
own society. Cf. further Schapera 1930: 214; Hirschberg 1975: 394.
152                             chapter seven

   To some extent, also, the question of regional variations must be
taken into consideration. One San people among whom witchery
seems to be more significant than among, for instance, Kung and
Nharo, are the Kxoe, although even they practise it much less than
neighbouring Bantu peoples.18 Kxoe healers who use the supranor-
mal power of tco (cf. tsso among Nharo and num among Kung) for
evil purposes are much feared. In most cases, however, evildoers
who use supranormal powers are believed to be Mbukushu, a neigh-
bouring Bantu people, who allegedly have secret medicines which
are more powerful than Kxoe medicines.19 This seems to be the case
in the Nharo and Kung areas too, where San fear the possibility of
witchery attacks from outsiders more than from their own healers.20
According to Köhler, Kxoe can become ill because of something he
calls Tierzauber. Without the interference of human enemies, people
can become ‘bewitched’ by wild animals. It is by the activation of
the power of tco that an animal, or plant, can ‘bewitch’ a human
being and thus make that person ill. For example, big warts or pro-
tuberances in the face of a person may be associated with the evil
influence of a warthog.21
   In recent decades the contacts between San and black peoples
have become intensified. Further, San have become increasingly
incorporated into the process of modernization. The new situation
has brought about significant changes in the conceptualization of dis-
eases and their causes. For instance, the eastern farm Nharo studied
by Guenther now compartmentalize sickness into three categories.
First, ‘Bushman disease’ is represented by the sickness tssa Neri, which
is treated by the trance dancer. Second, ‘Bantu disease’ consists of
witchery, and its treatment is in the hands of the Bantu healer.

      Köhler 1971: 319.
      Köhler 1978/79: 28 f.
      E.g., Barnard 1979: 71; Woodburn 1982: 203; Guenther 1986: 60.
      With reference to examples such as this, Köhler (1978: 54) draws the following
conclusion regarding ‘Tierzauber’ among Kxoe: ‘In den meisten Fällen lässt sich
nachweisen, dass dem Tierzauber das Motiv der Ähnlichkeit zwischen einem Körperteil
des Tieres, dem Tier als Ganzem in Gestalt und Farbe sowie Eigenarten und
Gewohnheiten des Tieres einerseits und den Krankheitssymptomen des Menschen
andererseits zugrunde liegt.’ On the issue of ‘Tierzauber’ I have found little or no
information in other sources. One exception is Wilhelm (1954: 168), who holds that
Kung ascribe ‘magische Kräfte’ to certain animals. Another cause of disease that
is seldom mentioned in the sources on San is the curse. Heinz (1975: 26) reports
that Ko may curse each other, but these curses do not involve suprahuman powers,
and they know that their curses ‘have little meaning’.
                                living humans                                153

Third, ‘European disease’ consists essentially of organic, communi-
cable illnesses such as tuberculosis, which are treated by European
or other biomedically trained practitioners. Although a few of the
eastern Nharo have begun to practise ‘a Bushman version of witch-
craft’ called tssoku, ‘poisoning’ (cf. tsso), ‘Nharo witchcraft’ is squarely
based on the Tswana complex, to which many Nharo have access
through ties of intermarriage. In the hands of an Nharo, witchery
practice is somewhat vague or spurious and not very much feared.22
   The fear of witchery practised by non-Nharo is, on the other
hand, ‘always most intense, bordering on terror’. The new witchery
practices have highly disintegrative effects on individuals as well as
on the settler community as a whole. As argued by Guenther, the
effects would probably have been less serious or destructive had the
Nharo been able to deal with the problems of witchery within their
own conceptual system.23 Nharo lack their own defences and must,
thus, rely on Bantu healers in order to handle these problems.
   A similar situation exists also among the Dobe Kung studied by
Katz and Lee. When Kung suspect that witchery of an unfamiliar
black person is the cause of an illness, they do not attempt further
healing, because they believe that such witchery is too strong for
their num.24 Although they are struggling to accommodate the witch-
ery beliefs of their black neighbours, the Kung therefore have to
consult black healers. Conversely, black people are also impressed
with Kung healing techniques and frequently ask Kung healers to
treat their sick. According to Lee, Kung are less impressed with
‘European’ theories of disease causation than with those of their
black neighbours; but they have easily accepted the efficacy of
‘European’ medicines, particularly antibiotics.25
   One of the changes that have occurred among Kung is the intro-
duction of a new dance, the drum dance, which is performed by
women. The women dance and enter trance, while the men here
play a supportive role, beating the drum. The use of drums is the
result of influence from the Mbukushu.26 The drum dance is not pri-
marily a healing dance as such but rather a dance for introducing

     Guenther 1975: 164; Guenther 1986: 218, 240 f. See further, e.g., Widlok
2001: 361.
     Guenther 1986: 60 f., 65 f.
     Katz 1982: 55, 103.
     Lee 1984: 116 f.
     Cf. the early account of the use of drums in San dances in Lloyd (1911: 355).
154                             chapter seven

women to kia and allowing them to go deeper into it.27 Another
change is the professionalization of healers and the introduction of
payment for their treatment.
   This change has taken place among, for example, Nharo too.28
Moreover, there has been a recent rise in trance dance performances,
and the ritual has become more elaborate and esoteric. Guenther
has observed some revitalizing effects generated by the trance dance,
including the establishment of a sense of San-wide ethnic identity.
Among farm Nharo, religion has gained ‘foremost prominence’, and
the healer has become ‘a powerful rallying symbol that represents
the San people and their culture’.29 According to Guenther, religion
has not only become more prominent but also more standardized
and coherent.30


Since healers, iloibonok, are the primary human agents of disease
among Maasai, a brief presentation of this category of people is
needed. An important characteristic of healers is that they usually
belong to a special clan. According to some mythical accounts, the
first healer was sent by God at a time when the Maasai were already
a populous group. This healer, who is often called Kidong’oi, a name
which means ‘tail’, was adopted by the ancestors of the Aiser clan,
and healers therefore normally belong to the Enkidong’ family or
sub-clan of that clan.31 This sub-clan is the only one among pastoral
Maasai organized along lines of a segmentary lineage system and
possesses a long genealogical tradition.32 Although the Enkidong’ fam-
ily has gained ‘an extraordinary prominence’ in Maasailand, there
are some healers in various parts of this area who do not belong to
this family.33

      Lee 1984: 113 ff.; Katz 1982: chapter 9.
      Guenther 1986: 262 ff. See further, e.g., Smith et al. 2000: 86 f.
      Guenther 1975/76: 50 ff.
      Guenther 1986: 286 ff.
      See further, e.g., Fischer 1882/83: 63; Fokken 1917: 249 f.; Blumer 1927: 80;
Schmidt 1940: 323; Bernardi 1959: 163; Sankan 1971: 73 ff.; Saibull & Carr 1981:
18 f., 55; Århem 1989: 81; Waller 1995: 28.
      Jacobs 1965: 321.
      Galaty 1977: 279, 296.
                               living humans                               155

   There are different versions of the way in which Kidong’oi came
to the Maasai. For instance, according to some accounts he was sent
from heaven, whereas other versions indicate that he emerged from
the earth. He may be depicted as a semi-divine and semi-bestial
being—with a tail, as indicated by his name.34 Some Maasai hold
that he was of Kikuyu origin, while others suggest that he came
from the Oromo further north. All seem to agree, however, that he
was not a Maasai.35 Although the genealogy of Enkidong’ healers
can be traced back to the middle of the seventeenth century, it has
been suggested that they arrived later.36 Whatever the exact time of
arrival, it seems that the number of these healers is still quite low
in comparison to the number of healers among neighbouring peo-
ples.37 Not all members of the Enkidong’ sub-clan function as heal-
ers.38 However, in addition to consulting the Enkidong’ iloibonok, or
the few proper Maasai healers, Maasai are known to consult heal-
ers from other ethnic groups.39 This was exemplified in the account
of spirit possession above, but that is not the only case when for-
eign healers are consulted.
   Members of the Enkidong’ sub-clan have certain powers and gifts
or talents, provided by God, which other human beings do not have.
Because of the supranormal power of Enkidong’ healers, other peo-
ple regard them as particularly awe-inspiring, and there is a certain
fear of them. In their different localities they, therefore, tend to live
apart from others. As representatives of God they are, in a sense,
marginal men.40 In particular, the Enkidong’ healers have a gift of
prophecy or clairvoyance and know how to fabricate and use ‘rit-
ual medicine’.41 They are believed to have a power of knowing what
has happened and what will happen.42 Supranormal medicines, or
amulets, are called intasimi (sing. entasim) and differ from ordinary

      Galaty 1977: 280–288; Århem 1989: 81.
      Jacobs 1965: 321; Voshaar 1979: 331.
      E.g., Voshaar 1979: 208; Hurskainen 1984: 182.
      Hauge 1979: 59.
      Jacobs 1965: 322.
      E.g., Galaty 1977: 306 f.; Hauge 1979: 59; Hurskainen 1984: 184.
      Jacobs 1965: 321 f.; Marari 1980: 43; Århem 1989: 81. Blumer (1927: 80)
even claims that the most important of them are thought of as divine, on a par
with God himself. This, however, seems to be an exaggeration.
      Fokken 1917: 251; Sankan 1971: 75; Galaty 1977: 302 f. For a special study
of the prophetic role of Maasai healers, see Waller 1995.
      Galaty 1977: 295; Waller 1995: 32 f.
156                             chapter seven

medicines (ilkeek) in that it is not the physical properties that count
but the supranormal qualities.43
   There are different types of iloibonok. They have various skills, and
some are more powerful and influential than others.44 However, none
of these healers must be confused with the ilabaak (sing. olbaani ), who
are (secular) doctors or curers who operate, vaccinate, set broken
bones and know the healing properties of ordinary medicines.45 As
agents of God, the iloibonok have not only creative or constructive
but also destructive supranormal powers. Hence they have the power
to cause disease and death through, for instance, cursing as well as
the power to heal and ensure fertility.46 Their curses, ildeketa (sing.
oldeket), are considered stronger than those of most common Maasai
and, according to Galaty, are perhaps used with less restraint.47
   Among other people who may curse in order to inflict illnesses
or other misfortunes upon others, the elders are the most important
category. However, curses as well as blessings can be used by a vari-
ety of people, sometimes including women, who may curse their own
children.48 In other cases, curses by women are not considered
effective.49 Apart from the specialist oloiboni, the most potent curses—
as well as blessings—one can receive are from individuals who stand
in certain determinate relationships to oneself, such as the father or
grandfather or mother’s brother. A curse is an act that passes between
friends rather than strangers and is not an anonymous but a pub-
lic act. Members of a superior age-set are like ‘fathers’ to members
of an inferior one. Hence elders wield a curse threat over warriors.50
   The belief in the power of cursing, or blessing, is associated with
legitimate authority, which underpins the social order.51 In particular,
curses of elders are considered effective on account of the right order

      Voshaar 1979: 211; Århem 1989: 76. The term intasimi may also refer to the
very rituals in which amulets are fabricated as well as to rituals in general (Mol
1978: 42; Olsson 1989: 238). See further, e.g., Blumer (1927: 81) and Berntsen
(1973: 80).
      Bernardi 1959: 162 f.; Jacobs 1965: 321 ff.; Henschel 1983: 156.
      Jacobs 1965: 322; Voshaar 1979: 203 f.; Sindiga 1995: 97 f. For some infor-
mation on other types of specialists, see Voshaar 1979: 204 ff. See also Waller
1995: 43.
      Galaty 1977: 321; Århem 1989: 81; Sindiga 1995: 100.
      Galaty 1977: 310.
      E.g., Saibull & Carr 1981: 39; Spencer 1988: 48; Landei n.d.: 44; Peron 1995:
52; Waller 1995: 29; Voshaar 1998: 147.
      Voshaar 1979: 187.
      Galaty 1977: 309 f., 370; Voshaar 1979: 187; Landei 1982: 56.
      Spencer 1988: 219.
                                 living humans                                   157

of things and are used or threatened to be used for the benefit of
the community or family.52 The ultimate agent of the curse is God
himself, and appeals are made to him to punish an alleged offender.53
The power of cursing can be misused, however, and especially the
iloibonok are believed to use their power occasionally for their own
private benefit, allegedly because they are greedy and power-seeking.54
It seems that any kind of disease and death may be caused by a curse.55
Although a curse is usually given by word of mouth, the intention,
anger or ill-will of a person may be enough to inflict harm upon
another individual.56 However, diseases and deaths caused by curses
or negative intentions appear to be rare.57 Concerning the relation-
ship between a Matapato father and his children, Spencer says that:
       In practice, it is the blessing that is constantly uttered, whereas the
       curse is largely a rhetorical threat. These threats are generally seen at
       most as attempts to intimidate children into submission, since even in
       the heat of anger no reasonable father would want to undermine his
       own family and inflict misfortune. Some maintained, however, that in
       extreme situations the death curse really is intended.58
Beyond the protective care of parents, the mystical knowledge and
power of elders at large takes on a more formidable aspect; and the
morality of respect for age is seen, in the final analysis, as a matter
of life and death.59 Yet elders seem to show great restraint before
they resort to the pronouncement of curses. During his more than
twelve years among Maasai, Voshaar came across very few instances
of curses; and, with regard to the Arusha, Gulliver says that material
compensations for injuries or deprivation of rights are more com-
mon than the use of supranormal means.60

      Voshaar 1979: 213.
      Gulliver 1963: 286; Kimerei 1973: 19; Olsson 1989: 236. The words of a for-
mula of curse are like ‘negative words of prayer’ (Kimerei 1973: 30).
      Voshaar 1979: 213; Spencer 1988: 220. Cf. Landei (1982: 61) who says that
a curse cannot harm an entirely innocent person.
      Kimerei 1973: 27 f.; Galaty 1977: 321; Peterson 1985: 175. From the Arusha
area, Gulliver (1963: 288 ff.) reports on ritual oaths of various kinds that may
occasion serious illnesses and death too.
      Voshaar 1979: 187; Landei 1980: 21; Ndoponoi 1986: 17.
      E.g., Gulliver 1963: 286; Voshaar 1979: 187.
      Spencer 1988: 48.
      Voshaar 1979: 189; Gulliver 1963: 288. Cf. Galaty (1977: 310) who concludes
that ‘while many curses are actually taken lightly, since they function more as abuses
than imprecations, many are exceedingly serious and require a ritual removal by
a counter blessing of the curser’.
158                                chapter seven

   A special role is played by the blacksmiths (ilkunono) since they are
believed to command a curse over all Maasai in a general rather
than in a specific sense, much as a member of one age-set holds the
power of a curse over all members of a reciprocally inferior age-set.
In other words, all blacksmiths hold such power over all other Maasai.
The blacksmiths are said to wield this power because all possessions
and actions depend on them, for nothing important can be done
without the implements of iron.61 Blacksmiths are despised and con-
sidered inferior; they live separately and form an endogamous group.62
The source of the contempt for the blacksmiths is their fundamental
ascendancy over the Maasai based on the relationship of maker and
user, giver and receiver, specialist and ordinary citizen. The black-
smiths are a small minority and their habitations usually clustered
in a few locations; but through their crafts they are implicitly involved,
at a basic level, in all Maasai functions and activities.63
   Like healers (iloibonok) and elders, blacksmiths consider themselves
in accord with God in their use of the curse, because the power in
general as well as the efficacy in particular depends upon the mediation
of God. Maasai, however, do not link their curse with divine mediation
and intrinsic justice but see the curse as vindictive and illustrative
of the blacksmiths’ envy of them. For major as well as minor diseases
and other misfortunes the blacksmiths are ready scapegoats.64 Little
mentioned by Maasai but stressed by the blacksmiths themselves is
the power of healing inherent in them. Like other persons with a power
of harm, they have an inverse power of assistance or healing too;
and they prefer to be considered as men of constructive rather than
destructive powers and aims. The most common afflictions treated
by blacksmiths are disease and barrenness. When treating people,
they use charms or ritual medicines which are unique to them.
Blacksmiths even claim that iloibonok are blessed and healed by them.65
   Curses can be accompanied by certain acts sometimes loosely
referred to as ‘sorcery’ or ‘witchcraft’, esakutore or esakutote.66 The

      Galaty 1977: 371 f. See also Waller 1995: 33.
      Fokken 1917: 248; Berg-Schlosser 1984: 157; Laube 1986: 119.
      Galaty 1977: 382.
      Ibid., 375 f.
      Ibid., 379 ff. Like blacksmiths, members of other ethnic groups such as Gogo
and Rendille are outsiders whose supranormal power and ritual curse may be much
respected and feared ( Jacobs 1965: 322).
      Voshaar 1979: 213; Spencer 1988: 220; Århem 1989: 81. According to Mol
(1978: 172), esakutore refers to ‘witchcraft (in general)’, whereas esakutote means ‘witch-
craft (in particular)’; the verb, ‘to practise witchcraft’, is asakut.
                                 living humans                                  159

practising of witchery, which can cause disease and death, may
require that the agents possess something of the victims, such as
fingernails or hair or the dirt of their footprint.67 Outside of the vil-
lage, such substances are mixed with other ingredients, picked up
with the leaves of certain trees and flicked in the inauspicious direc-
tion of the victims themselves.68 Acts of witchery may be performed
by members of the Enkidong’ dynasty from which all sorcery is
assumed ultimately to derive.69 Galaty and Spencer seem to agree
that the most common situation in which sorcery occurs is that of
competition or jealousy between iloibonok.70 Several scholars stress,
however, that these healers practise witchery in rare cases only.71
    In addition to the iloibonok, there is another group of people, called
ilkuyatik, who possess certain supranormal skills for good or evil. Very
little has been written on these persons. Voshaar briefly mentions
their existence but does not say why they are believed to have spe-
cial powers.72 It seems that the ilkuyatik, who are more associated
with witchery than are the iloibonok, previously had the place of the
latter. Hence they were probably healers who were pushed away by
the emerging iloibonok. At least in Tanzania, there are still some ilkuya-
tik who practise healing, although they fear the stronger power of the
    The idea of ‘sorcerer’ or ‘witch’, olasakutoni,74 who harms people
simply out of evil motives or for profit, does exist, even though it
appears to be quite insignificant.75 According to Galaty, ordinary
Maasai virtually interchange the terms of oloiboni and olasakutoni and
primarily fear the former.76 Among Matapato, the stereotype of the

      Galaty 1977: 312, 371; Peterson 1985: 175; Århem 1989: 81.
      Galaty 1977: 312; Voshaar 1979: 213 f. Galaty (1977: 312 f.) adds that bewitch-
ing can also occur through the use of thoughts and songs or by the assistance of
an animal, often a fox, to convey the ‘medicine’ to the village of the victim. See
also Hauge 1979: 58.
      Jacobs 1965: 322; Spencer 1988: 221; Århem 1989: 81; Waller 1995: 31.
      Galaty 1977: 312; Spencer 1988: 221.
      E.g., Jacobs 1965: 323; Hauge 1979: 59; Hurskainen 1984: 184. Cf., e.g.,
Fuchs (1910: 129), Berthold (1927: 5) and Landei (n.d.: 41) who hold that iloibonok
do not use their supranormal power for evil purposes and in order to cause harm.
      Voshaar 1979: 214. Cf. Hauge 1979: 56.
      This information is based on personal communication with A. Hurskainen. Cf.
Jacobs 1965: 323.
      Cf. Mol 1978: 172.
      Galaty 1977: 306; Hauge 1979: 57 f. Cf. Mol 1978: 172.
      Galaty 1977: 306. Cf. the apparently misinformed accounts by, among others,
Fischer (1882/83: 72) and Dallas (1931: 41).
160                               chapter seven

sinister sorcerer is usually not a stranger but a jealous brother or
co-wife.77 Sorcery, which requires only technique and no moral
justification to be effective, is like poison in the hands of a malicious
person. People are expected to adopt a lifestyle that would attract
neither the jealousies of sorcerers nor the suspicions of others that
they might be sorcerers themselves. Sorcerers deliberately seek to
harm their victims and may pose a threat to strong people as well
as weak, although elders claim that their power to curse is superior
to the power of any known sorcerer. Elders themselves disclaim any
detailed knowledge of the techniques of sorcery, since that would
raise awkward questions as to how they know.78 However, although
the competitive and jealous iloibonok are the most common suspects
of sorcery, elders who adopt too remote a life-style, and cut them-
selves off from their age mates, may also be suspected of practising
sorcery.79 Concerning the Arusha, the written information on witch-
ery problems is scarce, but some reports seem to indicate that such
problems are, or have been, more prevalent among these settled
Maasai than among the semi-nomadic, pastoral Maasai.80
   In the category of human causation of disease there is, further-
more, a belief in people with ‘eyes’, inkonyek. Such a belief is confirmed
by early as well as by more recent scholars.81 At the beginning of
this century, Merker reported that the belief in the ‘evil eye’, that
makes human beings and animals sick, was widely spread (‘allgemein
verbreitet’);82 and seventy-five years later Hauge, likewise, concluded
that there is ‘a general belief that some persons can cause harm by
staring at someone with an “evil” eye’.83 People with ‘eyes’ can peer
right into living things and ‘see’, for instance what they have eaten
or the sex of a foetus. The natural desires for things like food and

      Spencer 1988: 45.
      Ibid., 219 ff.
      Ibid., 225 f.
      Fleisch 1936: 438; Hohenberger 1936: 1; Gulliver 1963: 289. It is possible that
the missionaries Fleisch and Hohenberger exaggerated the importance of witchery.
On the other hand, the conspicuous absence of accounts of witchery in the recent
works by indigenous authors like Kimerei, Landei and Marari may be understood
in the light of their different tendency not to blacken but rather to ‘uplift’ the reli-
gion of their own people. In general, African scholars of religion have paid little
attention to witchery phenomena. See further, e.g., Westerlund (1985: 30, 36).
      E.g., Fischer 1982/83: 72; Koenig 1956: 96; Ndoponoi 1986: 17; Sindiga 1995:
101; Voshaar 1998: 155 f.
      Merker 1904: 203.
      Hauge 1979: 58.
                                living humans                             161

children can be dangerous when they well up in such people. Their
hidden desires do not harm their own families and herds, but oth-
ers are at risk. When they look at an adult person or a cow, the
‘poison in their eyes’ may cause faintness or mild illness. However,
small children and calves are more vulnerable and can become seri-
ously ill or even die.84
   The ‘evil eye’ is inherited and does not necessarily imply malev-
olent intent. If a person with ‘eyes’ gives a blessing by spitting in
the direction of any possible victim, the effect of the ‘poison’ is neu-
tralized. However, there are malcontents and strangers who pass
through an area who may wish to conceal their power. Unknown
guests at local celebrations or feasts are, therefore, politely asked to
spit. Moreover, a pregnant woman is expected to keep her body well
covered in order to protect the foetus when there is a stranger in
the vicinity. Partly because of the fear of ‘eyes’, children hide behind
their mothers’ skirts when visitors are coming. Even dangerous ani-
mals are sometimes said to have ‘eyes’.85
   In comparison to, among others, Kisonko Maasai to the south,
Matapato less often fear abnormal people with ‘eyes’ who pose a
general threat. Rather, Matapato suggest that the real danger to
mothers and children comes from normal Matapato who nurture
grief for a child that died or a child that was never even conceived.
The belief in ‘eyes’ is an expression of the idea that inner feelings
are assumed to emanate from within a person and affect the inti-
mate social environment. It is, therefore, important that people with
sad memories or unfulfilled desires should spit to annul the harm-
ful effects that might stem from their suppressed feelings.86 A baby
who has been harmed by the ‘evil eye’ of a person can only be
healed by the same person. Hence iloibonok cannot be of help in such
a situation.87 In addition to hiding a baby, however, there are other
preventative measures that may be used to avoid illnesses caused by
people with ‘eyes’. Various kinds of amulet are believed to make the
‘evil eye’ ineffective. Ultimately, such objects are provided by God
himself, not by human beings.88

       Spencer 1988: 43.
       Hauge 1979: 58; Voshaar 1979: 195 f.; Spencer 1988: 43 f.; Voshaar 1998:
       Spencer 1988: 44; Voshaar 1998: 155 f.
       Hauge 1979: 58; Spencer 1988: 43.
       Voshaar 1979: 206 f.
162                              chapter seven

   In certain respects the changes during colonial and post-colonial
times have led to a decrease of the influence of the iloibonok. For
instance, they no longer foretell favourable times for warfare and
cattle raids, nor do they prepare war amulets for warriors. An impor-
tant change of particular interest here is the introduction of bio-
medical services to the Maasai. Whereas the iloibonok formerly provided
virtually all the medical services for people and livestock, local dis-
pensaries or regional hospitals and veterinary facilities now provide
important supplementary services. As observed in the account of the
recent phenomenon of spirit possession in chapter three, Maasai heal-
ers now have to compete with Bantu and Swahili healers too.89
   Paradoxically, however, the position of iloibonok has been strength-
ened, and their number has increased. Even some women now
become healers. One reason for the strengthened position of iloibonok
is that they have gradually suppressed the ilkuyatik. Similarly, they
have gradually replaced rain-makers too.90 Moreover, the consulta-
tion of biomedical practitioners does not necessarily mean that the
expertise of the iloibonok is rejected. Rather, as remarked by Galaty,
it often happens that individuals will simultaneously consult the local
medical clinics as well as iloibonok, believing that both methods together
may help where one alone would fail.91 Galaty concludes that heal-
ers of the Enkidong’ family have gained ‘an extraordinary promi-
nence in Maasai-land, to the near exclusion of any competition’.92
Because their power has expanded, other specialist groups have
become redundant or at least very peripheral in the hierarchy of
consultants. By condensing the supranormal functions in one insti-
tution, iloibonok may have rendered the system of supranormal expla-
nations of events more coherent. According to Galaty, it is perhaps
the overlapping quality of functions that explains both the ascendancy
of the iloibonok and the growing power of their office through the
last century and the ‘incredible competition’ between various healers.93
   Voshaar emphasizes that, while the significance of healers has
increased, the authority of elders has begun to crumble under the

       Benson 1974: 115; Berg-Schlosser 1984: 168 f.; Hurskainen 1984: 186 f.
       Galaty 1977: 277.
       Ibid., 326.
       Ibid., 279.
       Ibid., 319 f.
                            living humans                            163

pressure of modern changes in the fields of, for instance, knowledge
and education. He holds that, particularly in the southern parts of
Maasailand, some iloibonok, whose motives are ‘greed and power’,
are successful in creating fear through esakutore and the threat of their
curses.94 A certain increase of the use by healers of their supranor-
mal power for evil purposes is also reported by Hauge. He stresses,
however, that ‘evil medicine-men’ are still scarce and that, on the
whole, there are fewer healers among Maasai than among neighbouring
peoples like Chagga and Luo. According to Hauge’s old informants,
people who previously wanted to harm anyone consulted healers
from such neighbouring groups, because ‘evil medicine-men’ and the
practising of ‘evil magic’ were ‘extremely rare in former days’.95
    Admittedly, the sources are more or less vague on the issue of
changes in the field of disease causation. Yet it seems clear that, per-
haps with the exception of the upsurge of spirit possession in some
Maasai areas in Tanzania, there have been no drastic changes; even
though the increased importance of witchery, and particularly of the
iloibonok as agents of disease, is of somewhat greater significance
among Maasai than are the similar changes among San.

       Voshaar 1979: 268.
       Hauge 1979: 59.
                             CHAPTER EIGHT

                   KONGO AND YORUBA

                         The conception of the sorcerer (nlogi ), who under-
                         stands and secretly practices bad magic knowing
                         that it is immoral, is a very real presence in every
                         Sukuma community.1
                         [Among the Kongo] the social sources of illness
                         are seen to be more pernicious than ever and the
                         branches of medicine that deal with witchcraft and
                         magic more necessary than ever.2
                         While good relations have to be maintained with
                         the orisa and the ancestors, the greatest dangers
                         probably lie in the activities of the witches. Witchcraft
                         beliefs are still almost universal among the Yoruba,
                         despite the growth of education and the world


These quotations concerning, respectively, Sukuma, Kongo and
Yoruba indicate that belief in witchery is still very strong or even
stronger than before. Among the Sukuma there is no clear
differentiation between ‘witchcraft’ and ‘sorcery’. Hence the Sukuma
concept bulogi may be translated ‘witchery’.4 In the literature, bulogi
has also been referred to as ‘black magic’ and ‘poison’, although
‘witchcraft’ and ‘sorcery’ seem to be the most commonly used terms.
People who practise bulogi are called balogi (sing. nlogi).5 Some accounts
of their activities stress the aspect of witchcraft, that is, an inner

    Tanner 1956b: 437.
    Janzen 1979: 209.
    Eades 1980: 125.
    See, e.g., Nathalie 1884: 188; Brard 1897: 157; Welch 1974: 200.
    According to Tanner (1956b: 437), the term bulogi comes from a verb mean-
ing ‘to fear’. Cf. the concept mitunga, as discussed by Hendriks 1952b: 44.
166                              chapter eight

extraordinary power used for evil purposes.6 Others emphasize their
character of sorcery techniques that in principle are available to
everybody.7 There is much dread and hatred of balogi. Symbolically,
bulogi is associated with the colour black. While the blackness of life-
giving clouds of rain and fertility is good, the blackness of witchery
and death is bad. Mhola, the desirable state of everything, is threat-
ened by the activities of balogi.8 These may or may not be associ-
ated with ‘medicines’. Bugota, or buganga, is a Sukuma concept that
has often been translated as medicine. It represents an impersonal
power within human reach and can be manipulated by human beings.
The concept is comprehensive and refers, for instance, to headache
powders and injections for bilharzia as well as to protection against
evildoers and medicines for revenge. In bulogi, the power of bugota is
utilized aggressively in order to harm people.9
   Regarding the Kongo, missionary accounts from the late nine-
teenth and early twentieth centuries frequently contain references to
witchery beliefs and practices. Judging by these accounts, witchery
was at that time an important phenomenon among the Kongo;10
and, as exemplified by the quotation concerning the Kongo at the
beginning of this chapter, later reports by missionaries and scholars
show that during the twentieth and early twenty-first century it was
continuously significant.11 There are many similarities between the
Sukuma phenomenon of bulogi and the kindoki of the Kongo. The
concept of kindoki, like bulogi, may refer to both mystical and tech-
nical means (‘witchcraft’ and ‘sorcery’).12 Since it appears, however,
that ‘witchcraft’, in Evans-Pritchard’s sense, is the most important
aspect of kindoki, the following account will focus on that aspect.

      ‘On emploiera le terme “nogi” pour signifier que quelqu’un possède un pou-
voir secret et extraordinaire de jeter un mauvais sort qui aura comme effet la mala-
die ou la mort de quelqu’un’ (Table n.d.a.: 521).
      ‘Les balogi sont les artisans de la Magie Noire, magie considerée comme néfaste’
(Gass 1973 [1919]: 411).
      Brandström 1990, chapter 5: 20. The word bulogi comes from the verb kuloga,
which means to fear. Wijsen & Tanner 2002: 67. Cf. Bösch’s account of the
Nyamwezi (Bösch 1933: 224).
      Cory 1949: 13 ff.; Hatfield 1968: 83; Brandström 1990, chapter 6: 12 f.
      Augouard 1885: 103 ff.; Cambier 1890: 367; Dewilde 1894: 397; Butaye 1896;
Baltus 1898: 77; Le Scao 1908: 332; Le Scao 1915: 429 ff.
      See also, e.g., Boucher 1928: 147; Struyf 1933: 400 ff.; van Wing 1959: 377;
MacGaffey 1983: 151; Dalmalm 1985: 75.
      MacGaffey 1986: 185. Cf. Dalmalm 1985: 85.
                                   witchery                                167

    As a morally ambivalent supranormal power, kindoki can be utilized
for both positive and negative purposes. Whereas, among others,
chiefs and healers (banganga) use, or at least are normally supposed
to use, this power with good intents, bandoki (‘witches’/‘sorcerers’) act
in evil and destructive ways.13 As utilized by the latter, kindoki is the
antithesis of ideal governance. ‘The contrast between competent
authority and the disastrous consequences of unbridled envy, anger,
and injustice is a recurring theme in the exploration of causality of
illness.’14 Afflictions caused by kindoki belong to the category of ‘illness
by man’.15 Another means by which living humans can provoke ill-
nesses, both among the Kongo and the Sukuma, is through curses.
    As explanations of disease, references to human agents are of great
significance in Yorubaland too.16 Among the Yoruba, there is a con-
ceptual differentiation between ‘witchcraft’ and ‘sorcery’. A common
term for ‘witch’ is aje, and a ‘sorcerer’ is usually called oso.17 Collectively,
witches and sorcerers are sometimes referred to as ‘children of the
world’.18 Belief in witchery seems to be almost universal in Yorubaland.
Regarding ‘magic’ in general, P.A. Dopamu says that the demand
for it is ‘endemic’.19 Simpson reports that, among his more than
three hundred informants, there was no one who did not believe in
the existence of witches; and Adeniyi argues that no other indige-
nous belief is ‘more ingrained’ than the belief in such beings.20 In a
similar way, many other scholars confirm that witchery beliefs are
very real and prevalent among Yoruba people, regardless of sex, age
and occupation. In addition to the particularly serious activities of
‘witches’ and ‘sorcerers’, the use of the curse, epe, is a significant
non-physical means of causing illness among the Yoruba too.21

       See, e.g., Jacobson-Widding 1979: 68; Mahaniah 1979: 209, 214.
       Janzen 1978: 24.
       Ibid., 67, 73.
       See, e.g., Leighton et al. 1963: 38; Prince 1979: 115; Oladapo 1984: 191.
       Lucas 1948: 283; Dopamu 1977: 27, 41; Wolff 1979: 127. Cf. Olukunle 1979:
     Eades 1980: 121; Lawson 1984: 66. Cf. Wolff 1979: 128.
     Dopamu 1977: 669.
     Adeniyi 1984: 32.
     See further, e.g., Maclean 1971: 32; Dopamu 1977: 154, 263, 288, 660;
Olukunle 1979: 237; Eades 1980: 125; Simpson 1980: 127, 135, 137; Buckley 1985a:
11; Osunwole n.d.: 3.
168                               chapter eight


In an article entitled ‘The Ingredients of Magic Medicine’, Cory
concludes that ‘the medical knowledge and magical practices of the
Sukuma-Nyamwezi tribal group are acknowledged as the leading
school in a considerable part of the territory’.22 There is a great con-
cern for bugota in various contexts, such as in the secret societies,
special-purpose associations and ritual guilds, as well as in the every-
day life of Sukuma people.23 In 1956, Tanner thought that it would
probably be impossible to find anyone among the Sukuma who did
not believe in the existence of balogi. Yet he held, also, that there
was ‘a broad basis of scepticism to all their mystical knowledge’ that
prevented any particular conception in their religion gaining domi-
       It must not be inferred that the community is sorcerer-ridden and that
       the average Sukuma is constantly avoiding or suffering from bad magic.
       It is rather that witchcraft is one of several answers to misfortune which
       have to be sifted and calculated by the many magicians . . . who are
       consulted by almost everyone in trouble.24
Among the Sukuma there are numerous stereotypes about balogi,
which are well known in many other cultures too. For instance, they
can make themselves invisible and transform themselves into ani-
mals, including lions and crocodiles. In particular, they are associ-
ated with hyenas. Balogi co-operate and meet each other in nocturnal
covens. In lonely places they sing and dance naked together. Moreover,
they commit gruesome crimes such as incest and infanticide.25 Balogi
can instigate all kinds of evil. Illnesses and deaths are frequently
caused by them. For example, meningitis, nervous disorders, mental
afflictions, venereal diseases and sterility may be attributed to the
actions of balogi. Afflictions caused by them often have symptoms
related to swellings and internal pain. It is common, also, to see
incurable maladies as a result of their evil machinations. In general,
however, there are no specific diseases that are occasioned by balogi

     Cory 1949: 13.
     Brandström 1990, chapter 6: 13.
     Tanner 1956b: 437 f. Almost half a century later, he still claimed that ‘virtually
the whole Sukuma population’ believed in bulogi. Wijsen & Tanner 2000: 49.
     Brard 1897: 157; Gass 1973 [1919]: 430 ff.; Table n.d.b.: 522; Tanner 1956b:
439; Reid 1969: 54. Regarding the Nyamwezi, see Bösch 1930: 224 f.; Blohm 1933:
148 f.
                                   witchery                                  169

only. If a person whose illness has been caused by ancestors, God
or natural causes does not get relief, the problem of witchery may
be considered.26 Above all, balogi are the prime instigators of acci-
dents and deaths. When people die, the causes are seldom believed
to be natural or physical only. In particular, stillborn babies and
people who have died at an early age are easily seen as victims of
bulogi. Accidental deaths are more prone to bulogi causation than are
   Balogi may or may not use medicines when harming other peo-
ple. The composition of medicines is based, in part, on symbolic
principles. The ingredients are associated with the victim and the
desired result. Incantations may help activate the supranormal power
needed. It is, further, essential to obtain not only something of the
would-be victim, or something that he or she has used, but also to
know the full name of that person. As indicated above, another
means by which a nlogi can harm or kill a person is by transform-
ing himself or herself into a wild animal. There are other supra-
normal means of attacking and harming people as well. However,
the simple poisoning of somebody’s food or drink may also be
regarded as an act of balogi.28
   Extraordinary power is associated not only with balogi but also
with batemi, chiefs, and bafumu, healers. According to Cory, the ‘magic
powers’ of Sukuma chiefs represented the fundamental source of their
authority. Although the power of batemi is an ambiguous one, it is
normally used for, and meant to be used for, beneficial ends.29 Because
of the supranormal power of chiefs, balogi composing aggressive med-
icines may try to take, for instance, some food from the table of a
chief. In order to avoid misuse of their saliva, chiefs have avoided
spitting.30 Although both the nlogi and the mfumu have extraordinary

     Table n.d.b.: 2, 150, 585 f.; Tanner 1956b: 438 f., 442; Abrahams 1967: 78;
Reid 1969: 73, 77; Wijsen 1993: 148; Wijsen & Tanner 2002: 65. See also Bösch
1930: 224 f.; Blohm 1933: 150.
     Nathalie 1884: 187; Brard 1897: 157; Barthelemy 1905: 286; Gass 1973 [1919]:
429; Table n.d.a.: 18, 123, 521; Hendriks 1952b: 8; Tanner 1956b: 439; Welch
1974: 200; Ng’weshemi 1990: 28.
     For more details, see Cory 1949: passim; Table n.d.a.: 521, 527 ff., 582 ff.;
Wright 1954: 71; Tanner 1959a: 113; Tanner 1967: 18; Reid 1969: 74; Tanner
1970: 20; Balina et al. 1971: 47.
     Cory 1951: 74. See also, e.g., Brandström 1990, chapter 6: 20; Blokland 2000:
     Table n.d.a.: 274.
170                            chapter eight

power, they are very different. Unlike balogi, bafumu are respected
persons who use, or are supposed to utilize, their special power for
good and constructive purposes. Bafumu may well be suspected of
using their force and knowledge in a wrong or destructive way, and
anti-social medicines may be submerged in the covert parts of their
profession. Yet it seems rare to find healers among people accused
of practising witchery.31 As a rule, bafumu are leading opponents of
balogi. An important function of secret societies, in which the former
play leading roles, is to disclose or identify the latter and to provide
remedies. When initiated into secret societies, novices are taught
about the world of witchery and how to avoid its evil effects.32
   Bulogi is seen both as something inherited and as an acquired
knowledge or power. Hence people can seek after knowledge of
witchery in a conscious and deliberate way. Yet there is also an
assumption that balogi originally carried with them some latent dis-
position for evil. While some authors, like Tanner, stress the former
aspect, others, such as Millroth and the White Fathers who in the
early 1950s compiled the answers to the questionnaire, emphasize
the view that people were born balogi. In part, this seems to be a
reflection of the fact that the Sukuma themselves differ in this matter.33
It appears that all people may be accused of practising bulogi. Possibly,
accusations of witchery are directed mostly against successful and
rich people, particularly those who do not share their wealth with
the less fortunate. Balogi are both male and female, although the lat-
ter seem to be more frequently accused than the former. In partic-
ular, co-wives in polygamous households are exposed to witchery
accusations. Women before as well as, in particular, after menopause
may be charged with practising witchery. Not only close kinsfolk but
also affines and neighbours may bewitch each other. Balogi are seen
as social and moral deviants with harmful powers rather than as
people with physical and mental abnormalities, and there is much
fear of them. Moreover, the fear of being accused is an important
means of social control in the society.34
     Gass 1973 [1919]: 411 f.; Hatfield 1968: 104; Welch 1974: 210. On the
Nyamwezi, see Bösch 1930: 224 and Blohm 1933: 148.
     Gass 1973 [1919]: 413; Cory n.d.b.; Table n.d.a.: 542, 544; Hendriks 1952a:
31 ff. See also Bösch 1930: 225 f.
     Table n.d.a.: 523; Millroth 1965: 145. Cf. Tanner 1956b.: 442 and Tanner
1970: 22 f. See further Welch 1974: 203.
     Table n.d.a.: 523; Tanner 1956b: 440, 442; Abrahams 1967: 78; Tanner 1970:
22; Welch 1974: 188, 197, 208, 211; Wijsen 1993: 152; Wijsen & Tanner 2002:
66, 134.
                                  witchery                                 171

    Since terrible things may happen because of the greed, envy, anger
and desire for revenge of balogi,35 they have been, and may still be
severely punished. Old sources by missionaries have many references
to killings of balogi among the Sukuma. These missionaries mention,
also, that chiefs had the main responsibility for the conviction of
people accused of practising bulogi, and that the bodies of these peo-
ple were not properly buried. Some people who were accused took
their refuge in mission stations, where missionaries tried to protect
them. Since the colonial authorities prohibited the killing of balogi,
some of these were cleaned out and chased away from their homes.36
Although balogi may still be killed, forced migration has increasingly
become the ‘solution’ to the problem of bulogi.37
    There is some evidence that the problems of witchery have increased
in recent decades. In 1956, Tanner concluded that ‘old men can
still remember that before the German occupation witchcraft was
rare’.38 Tanner, as well as Gass, noticed that the believed increase
of balogi led, also, to a rising number of bafumu.39 In 1971, some
indigenous writers held that even most of the baptised Christian
Sukuma believed firmly in bulogi;40 and two years earlier Reid con-
cluded that balogi had become more important than ancestors as
causative agents of disease. In other words, there has been a ten-
dency of restructuring, that is, to attribute illnesses that were once
considered ancestor caused to witchery accusation.41 The Witch Murders
in Sukumaland by Tanner is a report about killings of accused balogi
in 1962. Many women, who were charged with the killing of sev-
eral people, were beaten to death with branches by crowds of men.42

     On such motifs, see Tanner 1956b: 437, 440 and 1970: 19. See also Bösch
1930: 224.
     Nathalie 1884: 187, 191; Livinhac 1888: 334; Hirth 1894: 460; Brard 1897:
157; Hirth 1899: 174; Barthelemy 1905: 286.
     See further Bösch 1930: 231; Table n.d.a.: 524; Hendriks 1952b: 45; Schans
1955: 30; Tanner 1956b: 438; Abrahams 1967: 42, 61; Welch 1974: 188; Wijsen
& Tanner 2002: 67. Whereas Bösch (1930: 228) writes about ordeals, fire tests and
poison cups among the Nyamwezi, Schans (1955: 26) claims that the use of poi-
son was not indigenous among the Sukuma, though it may have been used to some
extent because of influence from neighbouring peoples.
     Tanner 1956b: 443. See also Wijsen 1993: 115, 142. Cf. Wijsen & Tanner
2002: 59.
     Ibid.; Gass 1973 [1919]: 435.
     Balina, Mayala & Mabula 1971: 46.
     Reid 1969: 186 f. See also, e.g., Wijsen 1993: 115, 142; Tanner & Mitchell
2002: 131; and Wijsen & Tanner 2002: 55.
     Tanner 1970: 7 f.
172                              chapter eight

In his report about these events, Tanner stresses very strongly the
almost universal belief among the Sukuma in bulogi.43 Even though
there has been a general tendency of increase in terms of witchery
problems, there are certain fluctuations in the intensity of these issues.
In the mid-1980s, there was a new height of the problem of bulogi.
Hence it was perhaps not too much of an exaggeration when Steeves
spoke, in 1990, about a ‘constant threat’ of spiritual dangers such
as witchery.44


Bandoki, as well as other people with kindoki, have a kundu (or sev-
eral makundu). A kundu is a special gland, usually found in the stom-
ach, which enables the ndoki to ‘eat’ (dia) the vital essences or power
of other people. It is believed that, through autopsy, the kundu (or
makundu) of a dead ndoki can be found.45 Most accounts depict a
belief in kundu as a real bodily organ. However, some scholars stress
the notion of kundu as a symbol of evil.46 A one-sided emphasis on
the realistic or the symbolic understanding of kundu appears to be ill-
founded. Rather, there is both a symbolic language and a belief in
kundu as a real gland in the body.47 Like chiefs and banganga, bandoki
are said to possess ‘four eyes’ or ‘night knowledge’. Thus, they can
move about in the world of spiritual beings and deploy powers
derived from it in the world of living humans. Prophets (bangunza)
in indigenous churches may also have night knowledge or intelli-
gence.48 The feats of bandoki are similar to feats told about balogi

      ‘It would now be almost impossible to find anyone who does not believe in
the witch’s existence as one of the basic elements in their social life—a living real-
ity rather than an abstract idea’ (Tanner 1970: 19).
      Steeves 1990: 88 and personal communication with Bill Arens in 1990. On
the issue of curses, which are of much less significance among the Sukuma than
among the Maasai, see Ng’weshemi 1990: 27.
      For more details, see van Wing 1959: 345; Laman 1962: 216; Hagenbucher-
Sacripanti 1973: 144; Jacobson-Widding 1979: 54; Widman 1979: 164 f.; Pambou
1979–80: 23 ff.; Jacobson-Widding 1983: 382; MacGaffey 1986: 163 f.
      The symbolic interpretation is emphasized by, among others, Buakasa (1973:
142) and Axelson (1983: 23 ff.).
      See further Dalmalm 1985: 205 ff.
      Buakasa 1973: 29, 138 f.; Mahaniah 1982: 32; MacGaffey 1983: 140; MacGaffey
2000: 204. Cf. Dalmalm 1985: 85. To some degree, all elders also have special
power (MacGaffey 2000: 222).
                                  witchery                                  173

among the Sukuma. For instance, bandoki can transform themselves
into animals, leave their bodies and fly away.
     It is very difficult for the banganga to catch kundu people or bandoki, as
     they may hide anywhere at all and pass into any body or shape. The
     nsala [soul] of someone without kundu cannot leave his body completely
     to enter another shape, but this is a feat of the ndoki.49
Kindoki is an important explanation for evil, and it is frequently the
answer to the question why somebody has become ill. Other kinds
of misfortune may also be due to kindoki.50 With the exception of
sporadic maladies, which appear and disappear quickly, virtually all
sorts of illnesses may be caused by bandoki.51 Thus, physical diseases
as well as mental disorder can belong to this category. According to
A. Jacobson-Widding, particularly lingering sicknesses, internal prob-
lems and all conditions characterized by lack of appetite, faintness
and a gradual decline are due to the activities of bandoki.52 A gen-
eral characteristic of maladies occasioned by bandoki is that they tend
to be serious.53 The evil machinations of a ndoki can also cause prob-
lems of reproduction, such as sterility and miscarriage.54 In many
cases, the attacks of bandoki lead to the death of the victims. Early
missionary accounts very often provide examples of such deaths.55
In 1959, van Wing claimed that nine out of ten deaths were con-
sidered to be caused by kindoki and that domestic animals were also
killed by bandoki.56 In particular, kindoki is suspected in virtually all
cases of premature deaths.57 A ndoki may ‘eat’ a victim in any place
and at any time.58 When the ndoki ‘consumes’ the victim’s inner
essence, the latter sickens or dies.59

     Laman 1962: 217. See also, e.g., Sadin 1910: 138; van Wing 1959: 365, 373;
Laman 1962: 225–234; Buakasa 1973: 147; Pambou 1979–80: 23; Dalmalm 1985:
     Le Scao 1915: 431; van Wing 1959: 421 f.; Widman 1979: 178 f.; MacGaffey
1983: 135; Dalmalm 1985: 85, 90.
     Buakasa 1973: 148.
     Jacobson-Widding 1983: 382.
     See further Mahaniah 1982: 39–46.
     van Wing 1959: 358, 365; MacGaffey 1986: 162.
     See, e.g., Augouard 1885: 103 f.; Dewilde 1894: 379; Baltus 1898: 77; Butaye
1899: 311.
     van Wing 1959: 368, 371, 377. Cf. Vansina 1975: 673.
     Jacobson-Widding 1983: 382.
     Laman 1962: 222.
     Ibid., 219; MacGaffey 1986: 161.
174                             chapter eight

       Witches supposedly kill by stealing the souls of their victims and either
       eating or imprisoning them . . . Imprisoning the soul causes the visible
       body to sicken and eventually die, but victims can be rescued by heal-
       ers able to discover where the soul has been taken and to force or
       negotiate its return.60
As among the Sukuma and many other peoples, those among the
Kongo who were accused of being bandoki could previously be forced
to drink poison, in the Kongo case made of bark from a tree called
nkasa. This was the most common and feared ordeal. It was described
and strongly criticized in many early reports by missionaries.61 In
case of guilt, the poison was believed to pierce the kundu, which was
then excreted.
   Failure of the poison to identify a suspect who by other criteria
was ‘obviously’ a witch could be attributed to the suspect’s ingenu-
ity in hiding his kundu inaccessibly in his body or outside it alto-
gether, in the care of simbi spirits. Convicted witches were subsequently
killed by other means. Exonerated suspects danced triumphantly
naked, received heavy compensation from their accusers and often
took new names.62
   How, then, have bandoki been unmasked, and what kind of people
have been accused among the Kongo? While the bandoki themselves
are supposed to recognize each other, even if they live in different
villages, it can be difficult for others to know who they are.63 Certain
banganga and nkisi are specialists in revealing bandoki. A victim may
also recognize a ndoki in his or her dreams, and the ndoki can sub-
sequently be unmasked by a nkisi through its nganga. Moreover, cer-
tain bodily peculiarities, such as red eyes or deformed limbs, may
arouse suspicions of kindoki.64 Envy, greed, lechery and nervousness
can be seen as ndoki characteristics. However, it is very difficult to
establish certain types of ndoki personalities.65
   The kundu power can be inherited. It is also possible to be initiated
at a later stage in life. Accidentally, a person may become ‘contam-

     MacGaffey 1986: 162. Cf. van Wing 1959: 372.
     See, e.g., Augouard 1885: 105 ff.; Butaye 1896; Circulaire 1926: 1 f.; Struyf
1933: 401. For information on other types of ordeals, see Widman 1979: 172 f.
     MacGaffey 1986: 166. See also van Wing 1959: 253 ff.; Pambou 1979–80:
37; Dalmalm 1985: 82.
     Laman 1962: 222.
     Ibid., 220 f.
     van Wing 1959: 371; Buakasa 1973: 144; Jacobson-Widding 1979: 63; Jacobson-
Widding 1983: 383.
                                    witchery                                    175

inated’ by kindoki. One way of becoming a ndoki is to eat, unsus-
pectingly, meat that is in fact human flesh. Even if this meal may
be consumed ‘at night’, that is, in one’s dreams, it can nevertheless
be binding.66 Both women and men can be suspected and accused
of being bandoki.67 In various ways, bandoki tend to be ‘outsiders’.
They may be exceptionally rich and talented, live longer and have
more children than others. People who have become too rich and
successful are easily suspected of being bandoki. Suspicions of kindoki
may thrive where there is rivalry for power. Kindoki is a way of man-
ifesting the competitiveness of Kongo society, and accusations may
be provoked by someone’s feelings that he or she has been out-
smarted.68 People who are outsiders in the sense of being, for instance,
handicapped or too poor may also be suspected of practising kindoki.69
   The Kongo frequently say that bandoki attack only members of
their own lineages.70 Increasingly, however, fellow-workers, class-mates
and others have also been accused, particularly in urban areas. ‘In
modern practice, anybody may be accused of bewitching anyone
else.’71 Although killings of bandoki can still, illegally and secretly,
occur, they may now instead have to move from their home areas.72
First, however, attempts to make people confess and achieve recon-
ciliation should be made.73 In the twentieth century several kindoki
eradication movements, such as the Munkukusa of the early 1950s,
have been active in the Kongo area; and prophets in indigenous
churches became important helpers in the fight against the spread
of bandoki. At times, people have also relied, ‘to a far greater extent
than missionaries realized’, on two institutions that replaced the

      MacGaffey 1986: 164. See also van Wing 1959: 370; Buakasa 1973: 145;
Widman 1979: 171; Dalmalm 1985: 89.
      Hagenbucher-Sacripanti (1973: 146) stresses the significance of kundu inheri-
tance from mother to child, and Janzen (1978: 77) says that old, postmenopausal
women are often suspected of envy toward younger, fertile clanswomen. Cf. Jacobson-
Widding (1979: 58) who argues instead that women are seldom bandoki.
      MacGaffey 1986: 161. See further Buakasa 1973: 138 f.; Jacobson-Widding
1979: 53, 63, 79; Mahaniah 1982: 85 f.; Jacobson-Widding 1983: 384; MacGaffey
1983: 141, 143, 146.
      Buakasa 1973: 144; Dalmalm 1985: 88.
      Jacobson-Widding 1979: 61; Dalmalm 1985: 88.
      MacGaffey 1986: 164. See also Dalmalm 1985: 89.
      At the time when there were slaves among the Kongo, the rich and power-
ful could put forward slaves to take the poison ordeal for them. See further Augouard
1885: 106; Le Scao 1908: 331; MacGaffey 1986: 38.
      Dalmalm 1985: 80 f.
176                              chapter eight

poison ordeal: the Christian eucharist and surgery. ‘Both were regarded
as fatal to those who submitted to them without confessing their
witchcraft.’74 As will be discussed in the next chapter, witchery issues
are much involved in political contexts.75
    Diseases caused by kindoki are an extreme manifestation of ten-
sions and conflicts. However, ill will, envy, lies, malevolent intrigues
and backbiting may suffice to cause an illness. Even gossip, which
is a destructive force that erodes community harmony, can occasion
maladies and other kinds of misfortune. Since ‘conflict may cause
illness’, reconciliation—sometimes enforced—is vitally important.76
Conflicts can concern, for example, dissatisfaction with inheritance
distribution, wedding gifts and harvests. Nowadays particularly serious
contradictions may derive from conditions of labour migration,
distribution of wages and other sources of prestige.77 As among the
Maasai, for instance, blessings and curses are important causes of
health and disease respectively. For the Kongo, the curse of fathers
is of particular significance. Unlike a maternal uncle, who has jural
power over his sister’s children, a father only holds ‘mystical’ power
(kitata) to punish his own children. His curses, and blessings, can be
pronounced when he is alive as well as in the afterlife.78 A father’s
curse can have more or less serious effects. Thus, minor afflictions
such as headaches as well as, for instance, infertility may be the result.79
When curses are uttered by lineage heads, whose power of cursing
is particularly strong, the consequences may be even more serious.80


Among the Yoruba, aje (witchcraft) is an inherent power and, some-
how, witches are usually believed to ‘keep’ this special power in their
stomachs. More rarely, it is associated with their eyes. Some people

      MacGaffey 1986: 167 f. See further, e.g., Widman 1979: 203–232; MacGaffey
1983; MacGaffey 2000: 99.
      See also MacGaffey 2000: 225 f.
      Janzen 1978: 95, 99, 144, 205; Batukezanga 1981: 58. Cf. Dalmalm 1985: 78.
      Mahaniah 1973: 233; Janzen 1978: 145.
      Mahaniah 1973: 235; Jacobson-Widding 1979: 52, 72, 76; Widman 1979: 118
ff.; Jacobson-Widding 1983: 385. In some parts of northern Kongoland, where res-
idence is patrilocal, fathers have jural power over their own children. Nevertheless,
there may be a fear of kitata even there.
      Janzen 1978: 178; Widman 1979: 203.
      Jacobson-Widding 1979: 74.
                                   witchery                                   177

believe that the witches obtain their special power from God, whereas
others hold that it is derived from the divinities, orisha.81 Since the
power is ambivalent, it may be exercised for good and protective,
as well as for bad or destructive, purposes. Normally, however, it is
believed to be used solely for evil and anti-social ends. Unlike
‘sorcerers’, witches need not use ‘magical’ objects or techniques.82
Whereas osho (sorcery) is basically a male activity, the great major-
ity of witches are women, particularly old women. Most of those
who openly claim to be witches are barren women, past childbear-
ing age. Other women in unfortunate circumstances, who are unable
to achieve their wishes, are easily suspected of using witchcraft.83
Several scholars write about the belief in a real ‘witch cult’, composed
mostly of elderly women in league with each other, which is highly
secret and much feared.84 The secret guild, or ‘assembly meeting’ of
witches at night, is called ajo.85
   As a rule, the power of witchcraft is inherited, that is, it passes
from mother to daughter. However, it may also be given to non-
relatives, and it can even be purchased.86 Although a person may
not necessarily be aware that she (or he) has the power of witch-
craft, most witches are believed to know that they are witches.87
Yoruba ideas of witches are similar to those found in Sukumaland
as well as in many other parts of the African continent and else-
where. For instance, they are credited with the power of transformation
into animals such as birds or cats. When a witch has metamorphosed
herself into an animal, she falls into a deep sleep or becomes
unconscious. If anything happens to their familiars, some evil or even
death will befall the witches’ human bodies. Weird cries of noctur-
nal birds are often regarded as signs of the presence of witches.88

      According to Maclean (1971: 41), the power of witchcraft is given by Eshu,
and Buckley (1985a: 100 f.) reports that Shopona and witches may co-operate with
one another.
      Lucas 1948: 283; Dopamu 1977: 31, 36, 596; Simpson 1980: 76 f., 92. Cf.
Awolalu 1970: 80.
      Morton-Williams 1960: 37 f.; Leighton et al. 1963: 46 f.; Olukunle 1979: 192.
      E.g., Prince 1964: 91; Awolalu 1970: 29; Olukunle 1979: 105.
      Lucas 1948: 283; Dopamu 1977: 145. On women and witchcraft, see further,
e.g., Lucas 1948: 284; Prince 1964: 89; Bascom 1969a: 95; Maclean 1971: 41;
Olukunle 1979: 100, 192; Eades 1980: 125.
      Dopamu 1977: 31, 146; Eades 1980: 125; Osunwole n.d.: 5.
      Simpson 1980: 81; Osunwole n.d.: 4.
      Lucas 1948: 284; Maclean 1971: 42; Olukunle 1979: 100, 139.
178                               chapter eight

Moreover, witches are believed to have an insatiable appetite for sex
and a great craving for human flesh and blood.89
   Witches attack ethereal bodies or souls, and when somebody is
attacked that person’s physical body is weakened or destroyed.90 Such
attacks are not necessarily caused by personal enmity against the vic-
tims, but witches may harm others simply because they are full of
hatred and express their hatred by causing misfortune and death.91
A great variety of diseases, including serious ones, and deaths can
be occasioned by witchcraft. In 1960, Morton-Williams reported that
witchcraft was believed to be the most frequent cause of death,92
and about three decades later S.A. Osunwole concluded that nearly
all kinds of misfortune may be associated with the activities of
witches.93 According to Awolalu, witchcraft is considered the most
important factor that causes the abnormal to happen and brings
about disharmony in society.94 Similarly, Maclean and Oladapo refer
to witchcraft as a prime cause of illness and disaster.95
   When the following symptoms appear, attacks of witches are often
suspected: ‘bad dreams (especially of falling or being chased), sleep-
lessness, restlessness, worries, sterility, stomach pains, headache, and
skin sores’.96 However, witchcraft is a cause of many types of symp-
toms, and any of those listed in this quotation may also be the result
of some other cause. Besides, it must not be forgotten that most dis-
eases have multiple causes.97 Illnesses that are particularly likely to
be attributed to witchcraft are those that in the west would often be
referred to as ‘psychosomatic’. Likewise, psychiatric illnesses and prob-
lems are frequently associated with the machinations of witches.
Hence afflictions such as chronic ulcers, skin diseases, sterility, impo-
tence and insanity easily raise suspicions of witchcraft.98 P.A. Dopamu

      Bascom 1969a: 95; Maclean 1971: 43; Eades 1980: 125. For more examples
of the characteristics and alleged abilities of witches, see Olukunle 1979: 186 and
Simpson 1980: 75 ff.
      Olukunle 1979: 172; Adeniyi 1984: 67.
      Morton-Williams 1960: 36.
      Ibid., 35.
      Osunwole n.d.: 8.
      Awolalu 1970: 29.
      Maclean 1971: 43; Oladapo 1984: 191. See further, e.g., Lucas 1948: 284;
Leighton et al. 1963: 38, 148; Awolalu 1979: 87.
      Leighton et al. 1963: 104.
      Simpson 1980: 108.
      See, e.g., Prince 1964: 92 f., 96 f.; Maclean 1971: 41; Simpson 1980: 99 f., 108.
                                      witchery                                      179

stresses the significance of serious or protracted illnesses that are per-
ilous and extraordinary, or unusual maladies that bring wonderment
and bewilderment.99 Afflictions that are difficult or impossible to diag-
nose, like unusual blindness, are often in the category of witchcraft
causation too.100
   Furthermore, witchcraft can aggravate or intensify diseases caused
by non-human factors and can spoil the power of medicine.101 Illnesses
that defy a simple or normal therapy, as well as serious afflictions
in general, are readily interpreted within the causative framework of
witchcraft.102 Likewise, diseases that strike suddenly or violently are
easily referred to this framework.103 Other afflictions that may be
found in this interpretative context are rheumatism, epilepsy, ele-
phantiasis, malformation, retardation, leprosy, measles and small-
pox.104 The possibility of witchcraft is almost always considered when
men are impotent, women barren and when there are delayed or
prolonged pregnancies or other problems with the reproductive sys-
tem.105 As a woman grows old, and still has not become pregnant,
she may even be suspected of being a witch herself. In such a case
she is considered to be ‘so imbued with evil witchcraft power that
her own body is infected’.106 Witches can destroy their victims ‘in
stages’ until, ultimately, death occurs.107 While deaths of old people
are usually seen as natural, untimely as well as sudden deaths are
often regarded as consequences of witchcraft.108 In particular, the
death of children brings the issue of witchcraft to the fore.109
   Like Sukuma, Yoruba formerly tried witches by ordeals, and those
who were found guilty could be punished by death. For example,

       Dopamu 1977: 392.
       Oladapo 1984: 110; Osunwole n.d.: 8.
       Prince 1964: 91 f.
       Olukunle 1979: 126 f.; Osunwole n.d.: 8.
       Olukunle 1979: 140; Oladapo 1984: 110.
       Simpson 1980: 108; Oladapo 1984: 110; Osunwole n.d.: 8. The case of epi-
demic diseases like smallpox is particularly noteworthy because, as shown in chap-
ter 6, such diseases have often been associated with the role of divinities like Shopona.
This may indicate an important change, which will be discussed further later on.
It should be remarked here, however, that a deity like Shopona may also be thought
of as acting through witches on earth. See also Maclean 1976: 306.
       Olukunle 1979: 127; Wolff 1979: 129; Adeniyi 1984: 32.
       Wolff 1979: 126.
       Olukunle 1979: 172.
       Table n.d.b: 154; Adeniyi 1984: 32.
       Morton-Williams 1960: 35; Olukunle 1979: 211.
180                             chapter eight

an accused person could be given something poisonous to drink. It
seems, however, that Yoruba ‘witches’ were rarely brought to trial;
more often, they were executed secretly by certain cults.110 Oro and
Egungun, which were presented in chapter six, were such cults that
formerly executed witches, for instance by hanging. Probably, witches
were considered too dangerous to be apprehended by ‘secular’ author-
ities. Some people believed that the bull-roarers of the Oyo society
were enough to destroy witches.111 Under special circumstances, for
example when pestilence was threatening the inhabitants of a town,
oro rites could be performed in order to summon the ancestors col-
lectively and to drive away witchcraft from the town. Hence the oro
cult was believed to cleanse a town from witchcraft.112
   Apparently, it has occurred among the Yoruba, as among the
Sukuma and other ethnic groups, that persons accused of witchcraft
have been lynched by furious people. Other means of punishing
Yoruba witches are whipping, heavy fines or expulsion from their
communities.113 Unlike oro, gelede is a society in which witches have
been propitiated by sacrifice rather than executed. Diseases and other
misfortunes attributed to witchcraft are one important reason why
people have joined gelede. More recently, witchfinding cults and
Aladura churches have become significant institutions for protection
against witchcraft.114 In addition to seeking the assistance of such
institutions, people protect themselves by using amulets and per-
forming certain rites. If necessary, sacrifices to witches can be made.
Moreover, diviners and healers may provide ‘medicines’ for protec-
tion; and some of them, who are particularly skilled in so-called
magic, are believed to be capable of depriving witches of their spe-
cial power.115 Under normal circumstances, ancestors effectively guard
their descendants against attacks from witches.116
   Although old women form the majority of witchcraft suspects, any
person with anti-social behaviour may be suspected. It seems that
polygamous homes provide a particularly fertile ground for witchcraft
accusations, which are often made against co-wives or wives of other

        Lucas 1948: 284; Bascom 1969a: 40.
        Bascom 1969a: 93 f.; Olukunle 1979: 177; Simpson 1980: 53.
        Morton-Williams 1960: 37; Morton-Williams 1964: 247; Simpson 1980: 53.
        Lucas 1948: 284; Olukunle 1979: 177.
        Prince 1964: 109; Bascom 1969a: 95; Eades 1980: 125.
        Maclean 1971: 43; Eades 1980: 125; Simpson 1980: 80.
        Awolalu 1970: 27.
                                witchery                               181

men in a compound.117 According to Eades, such accusations ‘are
clearly related to the tensions arising from polygyny and the wife’s
subordination to more senior wives in the husband’s compound’.118
To accuse a rival or an opponent of witchcraft may be an easy way
of stigmatizing her (or him). Above all, witches attack kinspeople and
neighbours, but they are also able to operate beyond the local level.119


By definition osho (sorcery) is something evil and destructive, but a
person who at one time uses ‘magical’ techniques in order to harm
others may at another time use such techniques for good purposes.
The term oloogun may be employed with reference to a ‘herbalist’ as
well as to a ‘sorcerer’, although there is often a tendency to reserve
this term for the latter category and use the concept onishegun for
the former. Like a herbalist, a priest or babalawo may utilize his
power and knowledge of, among other things, incantations for evil
purposes.120 Oogun, which is frequently translated ‘medicine’ or ‘magic’,
is a mysterious or supranormal power in medicines that derives ulti-
mately from God and that can be used by humans for good or evil
ends. As is the case with oloogun, however, the term oogun is usually
employed in the bad sense. A special power inherent in, for instance,
names and in the spoken word is also recognized. Sorcerers are
known to utilize powers such as these but, like witches, they may
derive their malevolent forces from other sources too, the exact nature
of which is not known to outsiders. In sorcery, furthermore, acts and
expressions are combined with the force of will or thought. Particularly
sorcery that operates from a distance should be backed by will-
   Yoruba techniques or methods of sorcery follow principles famil-
iar from, for instance, previous Sukuma examples in this chapter,
and need not be treated in great detail here. Yet some specific
Yoruba cases may be offered. Some cult groups involve their deities

      Olukunle 1979: 129; Eades 1980: 125; Osunwole n.d.: 10.
      Eades 1925: 125.
      Olukunle 1979: 130; Osunwole n.d.: 10 f.
      Dopamu 1977: 31, 40; Wolff 1979: 128; Oladapo 1984: 17; Ogunsakin-
Fabarebo 1998: 9.
      Dopamu 1977: 34 ff., 42, 500, 551, 579 f., 592 f.; Wolff 1979: 127. Cf.
Farrow 1926: 116; Lucas 1948: 269, 276; Ayoade 1979: 50 ff.
182                             chapter eight

in attempts to injure enemies. In particular, followers of Shopona
appeal to him and use his emblem, a carved wooden figure, which
is prepared and placed according to special instructions, in order to
inflict mental illnesses.122 Names of certain divinities, like Orunmila
(Ifa) and Eshu, appear often in incantations, which indicate the power
that specific ingredients possess when they are combined, and which
may be used in connection with certain acts of sorcery. In principle,
incantations are part of the Ifa corpus and are, therefore, the eso-
teric knowledge of healers.123 Now, healers are not normally expected
to indulge in acts of sorcery. As among other peoples studied in this
book, it may occur, although it is usually believed to be rare.124
   Like healers, members of societies such as ogboni and oro are not
supposed to use their powers for anti-social ends, even though it
may happen. For example, the symbols of the ogboni society, which
are metal images of human figures known as edon, have been uti-
lized in acts of sorcery. In the past, these cultic objects were kept
secret from the eyes of the uninitiated, and they were supposed to
be used with the sanction of the whole ogboni society rather than
capriciously by individual members. If sacrosanct objects such as edon,
dresses of egungun or oro bull-roarers are touched by uninitiated peo-
ple, some believe that serious diseases or even death may follow.125
   In the 1920s, Farrow concluded that the most terrible exercise of
the power of oogun was apeta, which he called ‘invocation-shooting’.126
This type of sorcery was subsequently described by, among others,
Lucas and Dopamu.127 A sorcerer who aims at killing somebody by
using apeta makes a mud image of the intended victim, which is
treated symbolically or ‘magically’. At night the sorcerer sets this
image up, calls the name of the foe three times and then shoots at
the figure with a miniature bow and arrows. The wounds or dis-
eases that follow may soon bring the life of the victim to an end,

       Simpson 1980: 37. See also, e.g., Ogunsakin-Fabarebo 1998.
       Dopamu 1977: 114; Buckley 1985a: 146, 159. Cf. Olukoju 1997.
       Olukunle 1979: 108. Cf. S. Johnson 1921: 121. The ‘double’ capacity of
Yoruba healers is reflected, for instance, in the double meaning of the word oloo-
gun, as indicated above. For some more information on the various categories and
skills of healers among the Yoruba, see Table n.d.b.: 200; Maclean 1966: 132; Braito
& Asuni 1979: 188; Wolff 1979: 127 ff.; Oladapo 1984: 17 f.
       Lucas 1948: 278; Awolalu & Dopamu 1979: 226, 229.
       Literally, apeta means ‘call and shoot’ (Dopamu 1977: 189).
       Lucas 1948: 272; Dopamu 1977: 189.
                                   witchery                                  183

unless effective countermeasures can be applied.128 Apeta exemplifies
some of the important symbolic aspects of Yoruba sorcery, such as
the significance of names, numbers and colours. To know a person’s
name is, in a sense, to have some power over that person. The
power of names and the utterance of other words of power form a
large and important part of Yoruba sorcery and medicine.129 The
name of the intended victim of apeta is called three times because
three is one of the numbers that is believed to have a mysterious
power.130 The fact that the mud image is set up at night can be
related to the significance of black as a symbol of diseases, death
and evil in general.131
   The Yoruba practice of sorcery very often involves the use of
human body parts like skulls and bones. Substances from victims,
such as hair and clothing, as well as things like footprints and shad-
ows, are also frequently utilized or ‘treated’ in acts of sorcery.
Sometimes a sorcerer needs to be in close contact with a victim,
while other techniques work from a distance.132 Dopamu’s volumi-
nous thesis shows elaborately what an enormous variety of techniques
and methods there are in Yoruba sorcery. As emphasized by Dopamu,
the ways by which sorcery can be employed ‘cannot be exhausted’—
it may be used to procure virtually any end that cannot be achieved
by other means.133 Although sorcery is an activity, based on princi-
ples such as similarity, contagion, contact and transferability, that
aims at a kind of distillation of certain inner forces or essences, it
must not be seen as an alternative to religious and natural practices.
As among other peoples studied here, religious measures, practices
of sorcery and the use of herbal medicines are largely intertwined
among the Yoruba.134 When medicines are prepared, analogical and
other ways of thinking are not incompatible with empirical testing;

       Many other examples of sorcery techniques among the Yoruba are found in,
for instance, Lucas 1948: chapter 14 and Dopamu 1977: passim.
       Prince 1964: 90; Dopamu 1977: 42, 551, 559 f.; Ayoade 1979: 51, 54.
       Dopamu 1977: 517. According to Dopamu (ibid.), the following numbers are
also believed to have a special power and appear frequently in acts of sorcery: 1,
7, 9, 10, 200, 201, 1000 and 2000.
       For some information on the symbolism of the basic colours black, red and
white, see Dopamu 1977: 527 ff. and Buckley 1985a: chapters 3–5.
       Lucas 1948: 274, 277 f.; Prince 1974: 89 f.; Dopamu 1977: 215.
       Dopamu 1977: 228, 289. See also, e.g., Ogunsakin-Fabarebo 1998.
       Verger 1971: 50 ff.; Dopamu 1977: 67, 85; Ayoade 1979: 50 f., 54.
184                          chapter eight

and a sorcerer may well combine his (or her) ‘magical’ preparations
with ‘natural’ measures, like attempts to put poison in the food of
the victim.135
    Like witchcraft, sorcery can be the source of a great variety of
diseases and deaths. Sorcery may be the cause of almost any affliction,
and it is frequently held that dangerous types of sorcery, like apeta,
have occasioned the death of many people.136 Barrenness and other
problems with the reproductive system as well as markedly psycho-
somatic afflictions, like skin and stomach troubles, seem to be some-
what less associated with sorcery than with witchcraft.137 Like witchcraft,
however, sorcery is very often referred to in cases of psychiatric dis-
eases.138 The material from Prince’s study of more than 100 cases
diagnosed by Yoruba healers at indigenous treatment centres shows
that sorcery and witchcraft are believed to be a much more common
cause of psychiatric illnesses than are religious and natural causes.139
In addition to inflicting mental maladies, acts of sorcery may cause
people to commit suicide. Generally, sudden and unexpected, or vio-
lent, deaths and diseases are often attributed to sorcery.140
    Perhaps the most striking general characteristic of illnesses associ-
ated with sorcery is that they tend to be serious.141 The following
list, extracted from various sources, provides some examples of such
afflictions: cancer, rheumatism, elephantiasis, blindness and other eye
problems, tuberculosis, smallpox, measles, leprosy, epilepsy and inabil-
ity to urinate.142 As is the case with witchcraft, sorcery can worsen
maladies that originally were due to other causes.143 In such cases a
modern biomedical treatment is often combined with consultations
of indigenous healers, since illnesses caused by sorcery and witch-
craft, many believe, belong to a category of afflictions that cannot
be cured by modern medicine. A combination of different types of

      Awolalu 1970: 29 f.; Buckley 1985a: 160 f.
      Leighton et al. 1963: 38, 148; Dopamu 1977: 196; Osunwole n.d.: 8.
      For some examples, see Dopamu 1977: 207; Oladapo 1984: 140; Simpson
1980: 108; Osunwole n.d.: 8.
      See, e.g., Dopamu 1977: 166 ff.; Simpson 1980: 107; Osunwole n.d.: 8.
      Prince 1964: 96.
      Awolalu 1970: 25, 30; Oladapo 1984: 110.
      Dopamu 1977: 392; Oladapo 1984: 110; Osunwole n.d.: 8.
      Farrow 1926: 119; Lucas 1948: 277; Awolalu 1970: 30; Dopamu 1977: 207,
215; Simpson 1980: 107 f.; Oladapo 1984: 110; Osunwole n.d.: 8.
      Dopamu 1977: 392.
                                    witchery                                   185

treatment is, on the whole, very common, partly because it is usu-
ally assumed that most diseases, as well as other types of misfortune,
have multiple causes.144
    For protection against acts of sorcery a great variety of charms
are available. The use of charms or amulets in order to upset evil
forces, as well as for other reasons, is ubiquitous in Yorubaland.145
Amulets are usually obtained from healers or diviners, and some are
made with the assistance of divinities, thus becoming, as it were,
storage cells for some of the power of deities.146 If a person is too
powerfully guarded, because of the use of amulets, sorcerous counter-
measures and divine assistance, a sorcerer may instead try to strike
down the victim’s son.147
    In principle, sorcery can be practised by anybody. One reason
why there are fewer women than men among sorcerers, as well as
among healers, is that menstruation is believed to destroy the power
of medicines.148 The practice of sorcery reflects jealousies and hos-
tilities in society. It appears that conflicts and quarrels concerning
property and relations between the sexes give rise to most acts of
sorcery. As a rule, the persons involved are relatives or neighbours
who have been in close contact with each other. However, sorcery
is not normally used against members in a person’s nuclear family.
Sorcery is often seen as a weapon of the weak against the strong,
of the poor against the rich or of the unfavoured against the favoured,
but it may also work in the reverse direction.149 Formerly, alleged
sorcerers ran the same risk as people accused of being witches; they
could be executed by societies such as oro and egungun.150

       Simpson 1980: 108 f., 113, 130 f. A special case of affliction caused by sor-
cery was reported by Farrow (1925: 125), who held that a sorcerer can give birth
to and stimulate the growth of small ‘snakes’ and ‘insects’ inside the body of a foe
which occasion various diseases. This may be compared to the work by Buckley
(1985a: 32 f.), who says that such entities are placed by God within the body from
birth, and that it is natural reasons, such as too much sweet food, sexual activity
or alcohol, rather than human or suprahuman agents, that cause these entities to
become ‘too strong’ or ‘powerful’ in the body and therefore give rise to maladies.
See also the appendix.
       Lucas 1948: 279; Simpson 1980: 85.
       Morton-Williams 1964: 249 f.; Simpson 1980: 85. For some examples of var-
ious types of charm, see Lucas 1948: 279 ff. and Simpson 1980: 85.
       Morton-Williams 1960: 36.
       Buckley 1985a: 147.
       Prince 1964: 89; Dopamu 1977: 270, 272, 284.
       Bascom 1969a: 93 f.
186                            chapter eight

   Among the Yoruba, the curse (epe) is thought of as a powerful
means of inflicting illness or even death.151 However, scholars vary
somewhat in their assessment of how important it is. For instance,
Prince reports that curses are among the commonest causes of psy-
chiatric disorder,152 while Simpson says that curses are rare.153
Apparently, their significance varies a great deal from one Yoruba
group to another.154 In areas where curses are uttered, there tends
to be a strong fear of their harmful effects. Like acts of sorcery,
methods of cursing are employed mainly by men, particularly older
men, and are frequently associated with quarrels over women or
land.155 According to Simpson, some healers can be paid to curse
malefactors.156 A curse is an emphatic utterance, or command, made
in the presence of the victim, in which the operator of the curse
expresses which illness or other misfortune should befall the victim.157
   Yoruba curses are combined with the use of sorcerous ‘medicines’,
usually contained in some animal horn, and a person who curses
somebody may lick the medicines before uttering the curse. If a vic-
tim is powerfully defended, for instance because of divine protection
or the use of charms, the intended effect can be delayed or there
may be no effect at all.158 There is some evidence that, without being
‘strengthened’ by a proper curse, an evil wish may be transmitted
by a mere look. Also, a quarrel can in itself have harmful effects.
For example, it may cause complications in connection with deliv-
eries. In such cases confessions are important in order to counter-
check the harmful effects of quarrelling.159 The beneficial counterparts
of curses are blessings.160 The method of blessing, ape, may be used
for healing and other constructive purposes. For instance, it can

       Table n.d.b.: 155; Buckley 1985a: 142.
       Prince 1964: 91.
       Simpson 1980: 82.
       See, e.g., Lucas 1948: 276 and Dopamu 1977: 124.
       Prince 1964: 91; Dopamu 1977: 236; Buckley 1985a: 141.
       Simpson 1980: 82. Like Maasai healers, iloibonok among others, Yoruba peo-
ple who can curse are also able to tell what may happen in the future (Buckley
1985a: 142a: 142 f.).
       Prince 1964: 91; Dopamu 1977: 236; Simpson 1980: 82.
       Prince 1964: 91; Simpson 1980: 84; Buckley 1985a: 143. For detailed infor-
mation on the preparation of medicines used in combination with curses, see Lucas
1948: 276 and Dopamu 1977: 230, 233 ff., 245.
       Table n.d.b.: 22; Dopamu 1977: 593.
       Cf. Buckley (1985a: 144) who speaks about ‘prayers’.
                                   witchery                                  187

contribute to stopping the flowing of blood after an accident. Ape
may also be used by a healer when he discharges a patient who has
been mentally ill.161
   There is much evidence showing that, as among the Sukuma and
Kongo, witchcraft, sorcery and curses are still prevalent problems
among the Yoruba, and several sources even indicate that these prob-
lems have increased. Old sources such as the archival material of
the Church Missionary Society and Johnson (1899) remarkably seldom
mention these phenomena. Whereas much is said about the importance
of the divinities, orisha, very little is reported about witches and sor-
cerers. Since the old Christian missionaries as a rule were highly
critical of Yoruba ‘heathenism’, it does not seem likely that they
would have abstained deliberately from writing about witchery had
it been a very important problem. It is more likely that they would
have tended to exaggerate its importance in an attempt to blacken
the ‘pagan’ religion.
   Writing on witchcraft, sorcery and curses as causes of illness in
the 1960s, Prince reported that ‘many informants . . . emphasized that
these factors were on the increase’ and that healers unanimously
agreed that the practice of sorcery was ‘extremely common’.162 Dopamu
asserts that ‘the Yoruba believe they can achieve anything by their
magic’, which they use from the womb to the grave.163 Among oth-
ers, Dopamu and Oladapo agree that the significance of witchery
has increased in recent decades.164 As far as the belief in and prac-
tice of witchery is concerned, several scholars conclude that there
are no major differences between highly educated people and peo-
ple with low education or between city-dwellers and villagers.165 While
the victims of witchery used to be confined to relatives and neigh-
bours within compounds and villages, they are now increasingly found
beyond such confines.166 Correspondingly, there is an enlargement
of scale in terms of which people are accused of practising sorcery

      Dopamu 1977: 229 f.; Simpson 1980: 82; Buckley 1985a: 144.
      Prince 1964: 89.
      Dopamu 1977: 263, 288.
      Ibid., vii; Olukunle 1979. See also, e.g., Leighton et al. 1963: 38; Adeniyi
1984: 32; and Oladapo 1984: 191, 196. Cf. Simpson 1980: 110.
      See, e.g., Olukunle 1979: 221, 223, 237; Eades 1980: 125; Simpson 1980:
124, 134 f.
      Olukunle 1979: 229; Osunwole n.d.: 11.
188                            chapter eight

and, in particular, witchcraft. It is still primarily old women who
may be accused of being witches, but younger women and even men
are increasingly being suspected.167 In chapter nine some of the pos-
sible reasons for the change and increase of witchery problems will
be discussed.

       Olukunle 1979: 231; Simpson 1980: 75.
                               CHAPTER NINE


                          Using witchcraft . . . for sociological diagnosis is use-
                          ful so long as one does not reduce African medi-
                          cine to the expression and palliation of social
                          En Afrique postcoloniale, le rapport entre sorcel-
                          lerie et politique se renouvelle sans cesse, pas seule-
                          ment sur le plan local, mais aussi au niveau national
                          et même international.2


The decline of belief in spiritual beings and the subsequent increase
of living humans as agents of disease, which have been exemplified
in previous chapters, have been observed in many other parts of
Africa too. For instance, in a collected volume on witchery, the edi-
tors Middleton and Winter point out that ‘it is commonly held by
people in a large number of African societies that the practice of
secret maleficent acts is on the increase’.3 Concerning the Ndembu,
an agricultural Bantu people in northern Zambia, Victor Turner
concludes that illnesses and other types of misfortune no longer bind
a group of people together in veneration of ancestors sanctioning
the moral order. Since diseases are increasingly attributed to witch-
ery, which in the past caused death only, misfortune now tends to
break a group instead. In a study of another agricultural people, the
Cokwe or Chokwe in south-western Zaire, P. Stanley Yoder simi-
larly remarks that the importance of redressive rituals concerned with
ancestors has diminished and that they are far less frequently invoked
now than before.4

      Feierman 1985: 106.
      Geschiere 1996: 82.
      Middleton & Winter 1963: 20.
      Yoder 1981: 242.
190                              chapter nine

   Yet another example of the increased significance of religious dis-
ease causation and the increased importance of human agents is
found in Barnes-Dean’s work on the agro-pastoral Lugbara in west-
ern Uganda. Barnes-Dean concludes that, previously, most maladies
among the Lugbara were said to be sent by patrilineal ancestors.
Only when this possibility was discounted did people turn to seek
the cause of illness in witchery practices. Nowadays, however, such
practices are a predominant feature of the Lugbara medical system.5
More recently, scholars such as Stadler, Geschiere and Colson have
reported from various parts of Africa about a proliferation of witch-
ery problems.6 Their examples, and many others, clearly support the
above-mentioned conclusions drawn by Middleton and Winter.7 On
a rapidly increasing scale, witchery reproduces itself hand-in-hand
with modern changes.8 According to Geschiere, the rampant anxi-
ety about witchery among Africans in many parts of the continent
now triggers ‘a desperate search for new protections to contain novel
and therefore all the more frightening witchcraft threats’.9
   In addition to the shift from religious to human agents of disease,
an increased significance for the category of natural causation has
been observed in many parts of Africa too. In this book both of
these changes have been exemplified particularly in the chapters
regarding the Sukuma, Kongo and Yoruba peoples. In a more gen-
eral discussion, Whyte remarks that individualistic treatment of per-
sons by medicines—both ‘African’ and ‘western’—seem to be gaining
ground. This change, she argues, is one of the reasons for the shift
in anthropological research on African misfortune from religion to
medicine.10 Even though this chapter, like the preceding ones, will
not focus primarily on this aspect of change, but rather on some
possible causes for the gradual shift from religious to—in a wide
sense—social causation of illness, it is important to bear in mind.

     Barnes-Dean 1986: 344, 351.
     Stadler 1996: 87; Geschiere 1997: 7, 21; Colson 2000: 334 f. See also, e.g.,
Comaroff 1993; Whyte 1997; Douglas 1999; Geschiere 2000. In some areas AIDS
has contributed strongly to the fear of witchery. See e.g., Yamba 1997; Colson
2000: 353.
     See further, e.g., Westerlund 1980: 143 ff., 149 ff.; Barbee 1986: 77; Chavunduka
1986: 69; Multhaupt 1987: 455; Swantz 1989: 286.
     Ciekawy & Geschiere 1998: 3.
     Geschiere 1997: 216.
     Whyte 1989: 294.
                   factors of continuity and change                           191

   Particularly since the early 1990s, the increasing problems of witch-
ery has also caused a strongly renewed scholarly interest in this issue.
In their introduction to the collective volume Modernity and Its Malcontents:
Ritual and Power in Postcolonial Africa, the Comaroffs stress that, despite
the predictions of modernization theory and historical materialism,
the world has not been reduced to sameness and that ‘there are, in
short, many modernities’.11 In studies of what Geschiere refers to as
the ‘modernity of witchcraft’,12 the old functionalist approach, focus-
ing on order and internal integration, and the tendency to study
witchery in local contexts are too limited. There is clearly a need
for a wider scope that takes into consideration broader socio-eco-
nomic and political processes and not only the micro-politics of,
among other things, kinship, gender relations and morality.13 Hence,
the discussion in this chapter will be guided by such an ambition.

                                Modern Influences

Thanks to the missionary efforts of Christians and Muslims, as well
as other new factors of change, the adherence to indigenous African
religions was largely weakened during the period studied in this book.
Although the rapid expansion of Christianity and Islam in Africa is
not the object of study here, it should be noted that this expansion
has been more successful among the Sukuma and, in particular, the
Kongo and Yoruba than among the hunting-gathering San and pas-
toral Maasai. Christian and Muslim leaders have, sometimes relent-
lessly, fought against indigenous African beliefs and practices. To
some extent the pre-Christian and pre-Islamic religions have man-
aged to survive in altered forms within new religious contexts, par-
ticularly in Sufi Islam and in indigenous or African-initiated churches
(AICs). There are several such churches among the Yoruba and
Kongo. Sufism is well represented in Yorubaland but only as tiny
minorities among the other peoples studied here. Whereas leaders

      Comaroff 1993: xi.
      Geschiere’s monograph with this title (Geschiere 1997) is an important con-
tribution to the study of politics in Africa and exemplifies that modernization does
not exclude cultural heterogeneity.
      Ciekawy & Geschiere 1998: 1. For a study of older and too limited approaches
to the study of witchery in Africa, see Multhaupt 1990.
192                                chapter nine

and other adherents of AICs and Sufi groups have handled witchery,
more orthodox Christians and Muslims have tried, if largely unsuc-
cessfully, to root out these phenomena. Among charismatic Christians,
for instance in Pentecostal churches, who strongly believe in the exis-
tence of the Devil and evil spirits, there may also be ways of dealing
with problems of witchery which are similar to those found in AICs.14
   The Dutch Reformed Church, which has been in the forefront of
the fight against San beliefs and rituals, has reported some revivals.15
By and large, however, the San have a reputation for being quite
resistant to missionary influence.16 In Maasailand Christians, as well
as Muslims, have had even less sway over people, and only a few
per cent of the Maasai have converted to the new religions. Thus,
it is easy to find many Maasai settlements without a single Christian
or Muslim.17 Nowadays a sizeable number—according to one recent
estimate 12 per cent18—of the Sukuma are Christians, although in
general Sukuma people have been more hesitant to convert than
have, for instance, the Kongo.19 In Sukumaland, as elsewhere, mis-
sionaries and African priests—as well as political leaders, especially
in the upper echelons—have fought against pre-Christian beliefs and
practices.20 Among the Kongo this combat has existed for more than
500 years, since the start of the Catholic Christian mission in the
late fifteenth century.21 As of the late nineteenth century, the religious
struggle intensified and missionaries increasingly destroyed nkisi objects
or brought them to western museums. In the early 2000s little remains
of the mediating hierarchy of nkisi spirits and ancestors.22 Among the
Yoruba, both Christianity and Islam have contributed to undermining

       For an example of a study of Pentecostalism and occult forces, see Meyer
1998. See also, e.g., Ciekawy & Geschiere 1998: 8 and Colson 2000: 334.
       Katz 1982: 253 f.
       See, e.g., Marshall 1962: 221; Barnard 1978: 78.
       Hauge 1971:1; Voshaar 1979: 269 ff. See also, e.g., Donovan 1982: 175–199.
       Wijsen & Tanner 2000: 10.
       No precise figures can be given, since during the latest decades no official sta-
tistics on religious or ethnic affiliations has been provided. Information about such
affiliations is a politically sensitive issue. Therefore, it is usually not included in the
most recent census reports from African countries.
       See further, e.g., Reid 1969: 125–157.
       For a detailed historical study of culture confrontation in the Lower Congo
area, see Axelson 1970.
       Mahaniah 1982: 169 f.; MacGaffey 1986: 246. For some information on indige-
nous churches and witchery eradication movements among the Kongo, see e.g.
Mahaniah 1982: 131–160.
                  factors of continuity and change                           193

belief in the orisha. As in other areas, new schools run by Christian
and Muslim organizations, as well as by colonial and post-colonial
governments, have played an important part in challenging indige-
nous African religions.23
   Besides, more secular-oriented modern education and health care
have contributed largely to challenging African indigenous religions
and concepts of disease causation. The spread of bio-medical ideas
and practices is one of the factors that have furthered the impor-
tance of non-religious and especially natural etiologies and therapies.
Modern external influences, religious as well as non-religious, increase
the pluralism that already existed. Thus, in principle, these influences
do not add anything new. For instance, Islamic and Christian ideas
of jinns or evil spirits fit into the category of religious causation, while
biomedicine with its secular basis is narrowly concerned with the
category of natural and biological causation. As a rule, neither
Christian and Islamic nor biomedical disease etiologies pay much
regard to social aspects, even though indigenous African ideas of
human causation may survive within new religious frameworks. As
stressed by Feierman, corporeal individualism, which is ‘cultural, not
natural or objective’, pervades biomedical knowledge.24 While the
plural and flexible character of African medical systems apparently
is one of the reasons for the great adaptability to biomedical as well
as to new religiously based etiologies and treatments, it seems that
the influence of biomedicine tends to decrease the indigenous plu-
ralism or multidimensionality in these fields.
   It is interesting to compare here the British historian Terence
Ranger’s more general discussion on religion, development and iden-
tity in Africa, in which he convincingly challenges the organic model
of society and religion.25 Before the advent of modern colonialism,
there was not, in his view, an organic collectivity but a creative and
resilient pluralism. Commenting on the current situation, he argues
that ‘the real identity crisis in Africa is not found in changes from a
single traditional “frozen” identity to a bewildering pluralism. The

      See further, e.g., Simpson 1980: 109 f.; Apter 1993: 115. For an interesting
study of the survival among adherents of the indigenous Yoruba religion, as well
as among many Christians and Muslims, of certain festivals and rituals associated
with the sacred kingship traditions in the Ondo area, which constitute a kind of
local civil religion, see Olupona 1990.
      Feierman 1985: 109.
      Ranger 1987.
194                                chapter nine

real identity crisis is exactly the other way round. It is produced by
the change from a creative pluralism to single frozen identities.’26
One may clarify by stressing that there have been attempts, by polit-
ical and religious hierarchies, to create new identities. In practice,
however, this has not lead to the disappearance of the creative and
resilient pluralism.

                                  San and Maasai

The enlarged significance of natural causation, ‘herbalists’ and ‘sec-
ular medicines’, or what Feierman calls ‘commoditized popular
healing’,27 can partly be seen as an aspect of a wider process of
secularization and individualization.28 Paradoxically, the increased
weight of social causation, particularly in terms of witchery, may also
be understood as a part of this process. Among hunting-gathering
San and pastoralist Maasai, who have experienced much less of this
increase compared to other peoples studied in this book, there has
been a stronger ‘resistance’ to modern influences derived primarily
from the west. In particular, the pastoralist Maasai have become well
known because of their ‘conservatism’. P.H. Gulliver, among others,
points out that in their severe environment there is relatively little
scope for experiments with new possibilities,29 while Berg-Schlosser
argues that the ‘conservatism’ of the Maasai can be attributed to
neither ecological conditions nor a lack of contact with the outside
world. He maintains, further, that it has been of relatively little im-
portance whether attempts to initiate significant changes in their lives
have been made by colonialists, Christian missions or post-colonial
governments of Kenya and Tanzania. In Berg-Schosser’s view, the
Maasai mode of living clearly implies a certain ‘cultural preference’.30
Whatever the reasons for the alleged ‘conservatism’ of hunting-
gathering San and pastoralist Maasai may be, and a detailed dis-
cussion of this issue on a general level is beyond the scope of this

     Ibid., 156.
     Feierman 1985: 75. See also, e.g., Oyebola 1981: 87.
     Chavunduka and Last (1986: 263) argue that the most obvious reason for the
pre-eminence of herbalism in new associations of healers is the legal restraints against
practising anything that might be construed as sorcery.
     Gulliver 1969: 238 ff.
     Berg-Schlosser 1984: 157.
                  factors of continuity and change                        195

study,31 those San and Maasai whose mode of living has been more
radically changed need to be brought into the discussion about pos-
sible factors of continuity and change.
   As pointed out in chapter seven, among San and Maasai, includ-
ing Nharo and Arusha, respectively, who have been subject to par-
ticularly far-reaching socio-economic and other forms of upheaval,
human causes of disease have become increasingly important. It is
primarily among such groups that the supranormal powers associ-
ated with the healing properties of San healers and Maasai iloibonok
are more and more being used in the destructive context of witch-
ery. While the Arusha transfer from a semi-nomadic pastoral to a
sedentary agricultural mode of living occurred about 200 years ago,
the Nharo since the nineteenth century gradually through contact
with black and white settler groups have undergone even more rad-
ical socio-economic changes.
   Nharo farmers now live in a state of economic dependence on
white and black farmers. According to Guenther, it is the intense
and pervasive social conflict prevailing among Nharo farmers that
has created a fertile ground for the introduction of Bantu-derived
witchery; and it is due to the permanent oppression and deprivation
that their religion has become less world-removed and disengaged.32
Likewise, the marked rise of trance dances is caused mainly by the
increased existential stress, and trance dancers have become profes-
sionalized authority figures. Despite such important changes, Guenther
remarked in 1986 that the band, the fundamental structure of Nharo
social organization, was still basically intact. This, he argued, is
because of its flexible or fluid character; the necessity to maintain
band features such as sharing in an acculturative situation marked
by poverty; and finally because Nharo cultural traditions have been
maintained and even revitalized because of the activation and elab-
oration of religion.33
   In an article from 1992, Guenther discusses the issue of why witch-
ery in general is ‘not a Bushman thing’. His main answer is that by
and large there has been little or no interpersonal conflict that could
have formed a breeding ground for witchery suspicions, accusations

     For some discussions of changes and lack of changes among the San, see Lee
1984: 129–145 and Katz 1982: 251 ff. Cf. Barnard 1988: 217.
     Guenther 1979: 103, 112, 118.
     Guenther 1986: 288–295. See further Guenther 1999: chapter 8.
196                                chapter nine

and practices. For instance, there have been few, if any, tensions
over rights to land and other property. Likewise, problems such as
political rivalry, gender antagonism, conflicts between women in
polygynous households and generational tensions, which tend to be
common in witchery-ridden societies, have largely been absent from
hunting-gathering San settings. Moreover, among groups like Kung,
free vent has been given to anger and ill-will through talking, shout-
ing, singing and dancing. The trance dance, which may serve partly
as a kind of community healing, can be a particularly effective ludic
mechanism of conflict resolution. Another mechanism for open ex-
pressions of tensions is the joking relationship, which is a typical fea-
ture of San groups.34 Moreover, withdrawal is a common solution
to conflicts.
   Among the Nharo, by contrast, the seeds have been sown for
socio-economic inequality, gender tensions, conflicts over property
and similar problems. Most Nharo now live in an incipient form of
sedentary village society, with its rivalries, tensions and ressentiments.
In this situation the notion that the two existential stress factors of
disease and interpersonal conflicts, which have increased in tandem,
are causally linked has become entrenched. Yet even Nharo farm-
ers continue to categorize witchery problems as a ‘black custom’ and
use Tswana concepts for designating such issues. By this exercise,
according to Guenther, they can deflect tension and strife away from
themselves and onto a neighbouring people who already have a
ready-made, established witchery complex, thus lending more credence
to their notion that it is ‘not a Bushman thing’.35
   In previous chapters it has been shown that among the Maasai,
natural causes and treatments of diseases were of great significance
even in pre-colonial times and that, in comparison to hunting-gath-
ering San, living humans have been only slightly more important as
agents of disease. There are certain structural resemblances between
these two peoples that differentiate them from the Sukuma, Kongo,
Yoruba as well as other sedentary and mainly agriculturalist peo-
ples. While the San social structure is characterized by the band,
the age-set system is a marked feature of the Maasai. However, in
both cases there is relatively little of socio-economic inequality that

        Guenther 1992: 92 f. On joking relationships among the Kung, see Lee 1984:
64 f.
        Guenther 1992: 99 ff. See further, e.g., Smith et al. 2000: 86.
                    factors of continuity and change              197

could cause interpersonal conflicts and eventually lead to accusations
of activities causing disease or other misfortunes. Neither San nor
Maasai are politically centralized, though their societies have a cer-
tain gerontocratic character. The power of cursing controlled by
Maasai elders, which may cause diseases, tends to have a socially
stabilizing rather than destabilizing effect. By and large the increas-
ing types of conflict that characterize African village and urban con-
texts are rare in the semi-nomadic Maasai settlements.
   However, there is one particularly important difference between
hunting-gathering San and pastoralist Maasai. While gender relations
among the former are characterized by equality, or only quite lim-
ited inequality, such relations have a markedly patriarchal character
among the latter. Thus, tensions and conflicts between men and
women as well as between women in polygynous households are
more common among the Maasai than among the San. It is true
that there are a few examples of witchery used by, for instance, co-
wives. Yet, as shown previously, this is not a highly developed con-
cept among the pastoralist Maasai, and it is not primarily associated
with women. Nor is the possession of ‘evil eyes’ thought to be mainly
a female phenomenon.
   However, the recent development of spirit possession among some
Maasai women in Tanzania, which is manifested largely in various
symptoms of disease, may partly be interpreted as a reflection of
changes related to the gender inequality among the Maasai. Since
possession is ‘foreign’, like the ‘black art’ of witchery among the
Nharo, and consequently cannot be treated by iloibonok but requires
a new kind of conceptualization and treatment, it is to some extent
outside the control of Maasai men. As stressed by Hurskainen,36
Maasai women do have certain outlets for expressing suppressed
aggressions. For example, women as a group are allowed to chase
and beat their husbands for a special period of four days in a rit-
ual arranged for a man preparing for the circumcision of his first
child. However, such institutionalized and controlled situations for
channelling emotions offer only temporary relief, and the men’s con-
trol awaits the women after the end of the ritual period.
   Spirit possession, by comparison, offers women another kind of
outlet, which may bring them more than temporary change. At least

       Hurskainen 1989: 148.
198                              chapter nine

those Maasai women who do not only consult Bantu healers, but
eventually become members of Christian communities, are no longer
as much controlled by non-Maasai men as they used to be. They
may take leadership roles themselves and learn new things like read-
ing and writing. It would certainly be interesting to pursue the dis-
cussion of spirit possession among the Maasai, particularly in some
feminist perspective. However, since it is not a case of certain living
humans causing diseases of other human beings, such a detailed dis-
cussion is beyond the scope of this study.

                           Sukuma, Kongo and Yoruba

Among the Sukuma, Kongo and Yoruba, as well as among other
sedentary agriculturalist peoples, there are a number of factors that
seem to have contributed to the increase of living humans as agents
of sickness and, simultaneously, to the decrease of the significance
of spiritual beings. Whereas the Sukuma used to live mainly in fairly
small settlements before the post-colonial villagization process, the
long-term effects of which were limited too, the Kongo and Yoruba
have a long history of village life, and in the case of the latter even
of urban life. In recent decades the process of urbanization has accel-
erated, particularly in areas where the latter peoples live.37 Whereas
indigenous religions have become weakened in the processes of vil-
lagization and urbanization, political struggles, economic rivalry and
social stratification have increased, apparently ‘fertilizing’ the ground
for witchery ideas and practices.38 The retreat of suprahuman beings
as agents of disease has created, as it were, a vacuum that has been
filled, in part, by supranormal living humans.
   Previously, ambivalent spirits and divinities also lent important aid
in the struggle against human evildoers. Partly, the increasingly
‘demonic’ use of a supranormal power, which in itself is ambivalent
like the power of God and other suprahuman beings, is a result of
the breakdown of political institutions that, among other things, had
the task of controlling accusations of witchery. Some political changes,

      The significance of urbanization for the increase of witchery problems has been
stressed by, among others, Olukunle 1979: 221.
      The importance of economic factors is underlined by, e.g., Simpson 1980: 81.
For a South African example of the significance of villagization for the increase of
witchery accusations, see Stadler 1996: 89.
                   factors of continuity and change                          199

both at local and national levels, seem to have made people more
unprotected than before.39 For instance, the early suppression of the
nkisi system in Kongoland made it more secret and negative. It was
not replaced by any comparable institution, and increasingly it became
associated with kindoki.40 Among the Yoruba, associations such as oro,
which partly had the role of controlling evildoers, gradually lost most
of their influence or disappeared.
   Since chiefs used to have particularly important roles in the con-
trol of witchery, colonial and post-colonial reforms to change and
abolish the institution of chieftainship made it more difficult to han-
dle problems that people believed were caused by witchery. If chiefs
no longer took action against evildoers, people increasingly felt they
had to do it themselves. The event in Sukumaland in 1962, when
a series of killings occurred when women accused of witchery were
beaten to death by crowds of men, is an extreme example of this.41
In his research report The Witch Murders in Sukumaland, R.E.S. Tanner
ends by writing:
       The Sukuma fear and hatred of witchcraft which had been controlled
       traditionally by the chief and, in the colonial period, by administra-
       tive action, came to the surface as an expression of local tensions
       increased by the widening social and political distance between ruler
       and ruled.42
During the latest decades there have been many more cases of killings
of, mainly elderly male and above all female, balogi. According to
Tanner and Wijsen, these killings were partly connected to the emerg-
ing divisive inequalities caused by the substantial growth of the cot-
ton industry in Sukumaland. Large sums of money from this cash
crop production were controlled principally by older male land-
holders, although much of the cultivation and harvesting of their
cotton was done by their dissatisfied women.43

      See, e.g., Tanner 1967: 54 f.
      Mahaniah 1973: 244.
      Tanner 1970. See also, e.g., Olukunle (1979: 269) who writes about similar
events among the Yoruba.
      Tanner 1970: 39 f. Cf. Wijsen & Tanner 2002: 136. Concerning the chang-
ing role of chiefs among the Kongo, see e.g. Janzen 1978: 24 and Mahaniah 1979:
225 ff. Similarly, as indicated in chapter 1, among the Yoruba, kings and chiefs
have lost much of their political power, although they may still play a significant
religious and cultural role.
      Wijsen & Tanner 2002: 135 ff.
200                             chapter nine

   As remarked by Multhaupt,44 the colonial attempts to uproot witch-
ery were one of the means to legitimize colonialism. Accounts of
witchery could be used to stress the superiority of western rational-
ity and civilization. In more recent decades the political fight against
ideas and practices of witchery has been seen as an important part
of the process of modernization. The colonial ban on the poison
ordeal had quite far-reaching consequences. Many people viewed
this measure as an attempt to protect evildoers against retaliation by
their innocent victims.45 Deprived largely of its normal cultural con-
trol by colonial interferences, witchery became a disruptive and dan-
gerously anti-social force. The use of protective medicines, another
important means of witchcraft control, also came under attack from
colonial and post-colonial political leaders—as well as from mis-
sionaries and other Christian and Muslim leaders. Like chiefs, heal-
ers who provided such medicines were strongly criticized and often
hindered in carrying out what they thought were useful anti-witch-
ery practices.46 In most colonies, it was possible to convict virtually
any healer of witchery and to send that person to jail. As a result
of the colonial policies, many healers abandoned their activities at
the public level.47
   In response to radical changes bringing insecurity and anxiety,
many witchery cleansing movements have arisen in several parts of
Africa where people have conceived of witchery as a serious prob-
lem.48 Although such movements are not exclusively linked to the
colonial and post-colonial period of African history, this is a time
when they have been thriving. Some of the leaders have been
Christians or Muslims. Leaders of cleansing movements treated not
only individuals but also communities, thus trying to purify whole
groups of witchery problems. Even though such leaders have been
criticized by politicians, their means of controlling witchery has not
been a crime. Thus, it has provided people who believe in witchery
with an at least fairly acceptable means of dealing with witchery

      Multhaupt 1987: 445.
      Middleton & Winter 1963: 21. See also, e.g., Geschiere 1997: 15 and Douglas
1999: 181.
      See further, e.g., Lee 1976: 114 ff. and Janzen 1978: 24, 206. Cf. Tanner
(1970: 23) who writes that alongside the decrease of chiefly controls among the
Sukuma, the numbers of healers increased.
      Feierman 1995: 87.
      Middleton & Winter 1963: 24.
                   factors of continuity and change                  201

fears, suspicions and accusations.49 Many leaders of cleansing move-
ments have been innovative people opposed to colonial or post-colo-
nial authorities and collaborative chiefs no longer able to handle
problems of witchery; and their activities frequently manifested new
syntheses and values rather than social breakdown.50
   While cleansing movements attempt to solve the problem by puri-
fying the practitioners of witchery of their evil intent after confes-
sions, the goal of ‘witch hunting’ campaigns is to find and kill such
people. In recent decades such hunting has occurred increasingly
among rural and urban populations alike.51 The above-mentioned
case from Sukumaland is an important example of this serious prob-
lem. Although the purpose of cleansing movements is not to kill
human beings, there may be some violence, such as beating and
burning, involved in the attempts to make accused people confess.52
Thus, it is sometimes difficult to differentiate clearly between cleans-
ing and hunting movements. Since the aim of both types is to erad-
icate the problem of witchery, even though the means are totally
different, the term eradication movement may be used for referring
to both.
   Among the Kongo the previously mentioned Munkukusa of the
early 1950s was one of several examples of eradication movements.
In the Munkukusa ritual a mixture of grave dirt and palm wine was
employed as kundu-piercing medication. This mixture was placed in
a trench forming the diameter of a cosmographic circle. Each suspect
jumped over this trench after swearing an oath of the Nkondi type.
Then he or she hammered nails into a wooden cross, in place of
the Nkondi statue, and threw away various things symbolizing improper
wealth resulting from kindoki.53 Munkukusa followers resorted to the
graves of non-Christian ancestors for assistance in the fight against
bandoki but not to Christian ancestors since they were thought likely
to be soft on bandoki.54 Prophets, bangunza, who among other things
have tried to control witchery and facilitate group reconciliation,
were largely seen by colonialists as a political threat and frequently

       See further, e.g., Lee 1976 and Colson 2000: 343 f.
       Cf. Beidelman 1970: 354 f. and Multhaupt 1990: 38 ff.
       Stadler 1996: 108; Yamba 1997: 203; and Colson 2000: 334.
       See, e.g., Douglas 1999: 184 f.
       Dalmalm 1985: 92 f.; MacGaffey 1986: 187; MacGaffey 2000: 99.
       MacGaffey 1986: 170.
202                            chapter nine

imprisoned or exiled. The best-known of the prophets, Simon Kim-
bangu, was sentenced to death in 1921 and then imprisoned for life
in Katanga, where he died in 1951. In post-colonial times the official
attitudes to bangunza, as well as banganga, have been more tolerant.
However, in Kongoland as elsewhere, the contemporary official inter-
est in African medicine concerns, as it did during colonial times,
basically the herbal aspects, whereas what is seen as ‘magical devices’
can be vehemently opposed.55
   An interesting example from the Yoruba area is the Atinga move-
ment, which, like the Munkukusa in Kongoland, started its activi-
ties in the early 1950s. Like several other eradication movements it
was not restricted to one single ethnic group—it originated in the
mid-1940s in the (then) Southern Gold Coast from where it spread
east. Not only adherents of the indigenous Yoruba religion but also
Christians and Muslims embraced the new cult. Atinga, a foreign
deity, was seen as an angel by many Christians. Some people who
became possessed by Atinga claimed to have the power to recog-
nize witches and to discover where sorcerers kept their harmful
objects. Witches who were pointed out were asked to confess and
thus be cleansed of their aje (witchcraft). The followers of the Atinga
movement were mainly young people opposed to old orisha cults and
frequently attacked several orisha shrines. Eventually, however, it was
prohibited by law, and at least some of the destroyed shrines were
rebuilt.56 According to Apter, the development of a cocoa economy
intensified the existing etiology of witchcraft, and the Atinga cult had
a strategic value for the rising commercial elite. The ‘new men’ could
finance the cult from trade income and persecuted, in particular,
female traders and the predominantly female orisha cults rather than,
for instance, the Oro.57

                         Social Relations and Status

According to Mary Douglas, problems of human causation of mis-
fortune are likely to be marginal, if present at all, among peoples

     Janzen 1978: 51 ff., 60 ff., 208 ff. Concerning the Sukuma, see e.g. Hatfield
1968: 285 ff.
     Simpson 1980: 79 f.
     Apter 1993: 120 f. See also, e.g., Multhaupt 1990: 210 ff.
                    factors of continuity and change                               203

with sparse and irregular social contacts. She mentions hunting-
gathering San and pastoralist Nuer as examples of such peoples with
low and irregular levels of social interaction. Conversely, issues like
witchery and the evil eye are likely to be important where human
beings press closely upon one another and the intensive social rela-
tions are ill defined.58 In a similar sociological way, Middleton and
Winter argue that social tensions and competition arise more fre-
quently in societies or situations where status is achieved than in
those where it is ascribed.59 The fact that, in modern African soci-
eties, social relations are increasingly associated with achieved rather
than with ascribed status is apparently one of the reasons why human
causation of disease and other types of misfortune is on the increase.60
   The data on the peoples studied in this book to some extent sup-
port these contentions, and the discussion above can partly be seen
in this light. Among these peoples, hunting-gathering San have the
lowest and most irregular level of social interaction, and among them
witchery and other kinds of human causation of disease are of lit-
tle or no significance. Although such causation does exist among pas-
toralist Maasai, it is of minor importance there too. Even among
the Arusha, the social system is basically ascriptive: that is, social
status and relationships are designated by birth, age and place of
residence. Achievement is certainly possible but tends to be limited
by a prescribed framework. Of particular importance are the age
institutions, which for Arusha, as well as for pastoralist Maasai, play
a vital and fundamental role in their social and individual lives.61
   Unlike the above-mentioned anthropological works by Douglas,
Middleton and Winter, however, this book focuses largely on processes
of change. Among the San and Maasai, the incipient increase of
competition for enhanced social status and income seems to be most
clearly manifested by the tendency towards professionalization of
healers or trance dancers and iloibonok, respectively. By their con-
spicuous expansion of power the iloibonok have rendered other spe-
cialists redundant or at least peripheral.62 In one of his studies on

      Douglas 1970: xxx ff.
      Middleton & Winter 1963: 18.
      For a broader discussion of social anthropological theories on witchery in Africa,
see Multhaupt 1987. See also, e.g., Nadel 1952.
      Gulliver 1969: 232 f.
      Galaty 1977: 319.
204                              chapter nine

the Nharo, Guenther even writes that a successful trance dancer can
be assigned the potential role of a charismatic political leader with
far-reaching authority.63 Increasingly, the ambivalent supranormal
power of San and Maasai healers is being used not only for con-
structive purposes but also for evil machinations. Among the farm
Nharo, where witchery ideas and practices have developed more
extensively than among the hunting-gathering Kung, the Tswana
concept of kgaba has become used as a kind of negative counterpart
to tsso. As the healing power of tsso can be utilized not only by more
or less professionalized healers but by virtually anybody, all Nharo
are now also potential users of kgaba, the negative aspect of the supra-
normal power.64
   In a sociologically oriented study of hunting-gathering Kung, Lee
argues that they project the blame for malevolence to forces outside
the social body and regard healing power as being derived from liv-
ing humans. ‘They seek within the social body for benevolent pow-
ers.’ This serves to bind together the living in a common front against
‘hostile external forces’.65 While this is largely the case, it should be
remembered that the suprahuman beings are not entirely hostile.
Like the power of num, which ultimately derives from the ‘external’
force of God himself, they are rather ambivalent. Hence, it is somewhat
simplistic to argue that the therapeutic dance or trance performance
‘can be regarded as a drama in which the stresses and tensions of
social life are transformed into a common struggle against external
sources of malevolence’.66
   There is hardly any society where the social structure is purely
ascriptive. During the period studied here, however, the principle of
achievement has been more important among the Sukuma, Kongo
and Yoruba than among the San and Maasai. For instance, in 1933,
Blohm concluded that if a Sukuma man had an extraordinarily rich
harvest, or if a hunter was exceptionally successful in his hunting,
they risked being accused of witchery.67 In addition, the change

     Guenther 1975: 165.
     Ibid., 164; Guenther 1962: 85–90.
     Lee 1968: 51. Cf. Woodburn 1982: 207.
     Ibid., 53. Cf. Douglas’s (1970: xxx f.) broader discussion of witchery in terms
of outsiders and internal enemies, respectively. When witches are conceived of as
outsiders or external enemies, accusations reaffirm group boundaries and solidarity.
     Blohm 1933: 148 f.
                   factors of continuity and change                            205

towards a pattern that is increasingly dominated by the principle of
achievement has been more rapid among the first mentioned three
peoples, and particularly among the Kongo and Yoruba.
   Concerning the socially fairly stratified society of the Kongo,
MacGaffey in a general statement concludes that people there ‘have
always been individualistic entrepreneurs anxious to make a profit’.68
He argues that the distinction between public and private ends reflects
the struggles of a highly politicized society where authoritative allo-
cations of personal and corporate rights—supposedly fixed by descent,
kinship and ancestors—in fact have been constantly subject to polit-
ical challenges, particularly on the part of chiefs, elders and other
powerful people, supposedly the guardians of the legitimate order.
‘Denunciation of witchcraft and magic has always been part of an
effort to establish central authority and social discipline.’69 In another
study, MacGaffey stresses that jealousy and competition are com-
mon features among the Kongo and that they spend much time in
the hope of outsmarting others.70
   Of all the peoples focused in this study, the Yoruba is the most
socially stratified. Even at the beginning of the period concerned
here, that is, the end of the nineteenth century, they formed highly
stratified and complex urban societies. While some roles were ascribed,
there was also a significant amount of struggle for improvement.
From that time onwards, competition, as well as problems of witch-
ery, have increased. For instance, in his book on Yoruba religion
and medicine, Simpson concludes that rivalry and socio-economic
ambition, especially with reference to the education of children, are
widespread among the Yoruba today.71
   In general, family bonds are stretched to breaking points by new
inequalities. This is an important micro-political background to the
modern problems of witchery in many parts of Africa.72 While kin-
ship continues to be an essential factor in studies of such problems,73
other factors like changing relationships between people of different

     MacGaffey 1986: 174.
     Ibid., 175.
     MacGaffey 1983: 146.
     Simpson 1980: 81.
     Geschiere & Fisiy 1994: 183. Cf. Douglas 1999: 187 ff.
     This has been emphasized in, e.g., Geschiere & Fisiy 1994: 325 f. and Geschiere
1997: 9.
206                              chapter nine

generations and sexes, neighbours, work-mates and members of com-
peting religions—one dominant, the other suppressed—have become
increasingly significant.74 Accusations of witchery may also be involved
in personal vendettas.75 Witchery discourses can serve to bridge the
gap between the familiar realities of the domestic community and
the large-scale processes of change that have opened up new possi-
bilities for enrichment but that also impose new forms of depen-
dency, which are particularly frightening since they appear to be
impersonal.76 In the following section some examples of the widen-
ing political scene of witchery will be given.

                      Towards a Wider Political Context

There is ‘a general feeling in many parts of Africa—and certainly
not there alone—that witchcraft is reproducing itself hand-in-hand
with modern changes, and on a rapidly increasing scale. It is in par-
ticular this increase of scale that makes the occult all the more fright-
ening and uncertain.’77 The language and phenomena of witchery
now appear in broad regional, national and even in transnational
contexts. Like colonial rulers, new political elites condemn witchery,
but privately they are much involved in consultations of specialists
in these fields. It appears that the preoccupation with witchery haunts
political and other leaders as much as others. Various African media
make it clear that witchery permeates everyday conversation about
politics, the pursuit of power, and the complex interdependence of
rural and urban life.78 The role of ‘magic’ and ‘witch-doctors’ in,
for instance, local and national soccer teams is well-known. Increasingly,
witchery is no longer monopolized by specialists but can, in princi-
ple, be used by anyone. As a consequence, virtually anybody may
be suspected too.79

     For some examples, see Multhaupt 1990: 32; Stadler 1996: 87 ff.; Douglas
1999: 182 ff.
     See, e.g., Stadler 1996: 107; Douglas 1999: 183.
     Geschiere 1997: 24 f.
     Ciekawy & Geschiere 1998: 3.
     See further, e.g., Comaroff & Comaroff 1993: xxvi; Geschiere & Fisiy 1994:
323, 330; Geschiere 1996: 85; Geschiere 1997: 1, 3, 6; Schatzberg 2000: 34; Bernault
& Tonda 2000: 5; Geschiere 2000: 19.
     Geschiere & Fisiy 1994: 334; Bernault & Tonda 2000: 7; Schatzberg 2000: 40.
                    factors of continuity and change                             207

   As touched on above, of the five peoples analyzed in this volume,
the San and Maasai have been somewhat less affected by the mod-
ern enlargement of scale than have the Sukuma, Kongo and Yoruba.
Comparatively few San and Maasai have been involved in struggles
for power in modern political institutions, even though such bodies
certainly do influence them too. By contrast, many Sukuma, Kongo
and Yoruba people have been active as politicians and in other lead-
ership roles within modern structures. For instance, MacGaffey has
concluded that politicians, in particular, have taken the operation of
invisible forces of witchery seriously and that the belief in such forces
is an essential factor in politics at the national level.80 In his recent
book Kongo Political Culture, he writes:
       When Laurent Kabila’s miscellaneous army displaced Mobutu in 1997,
       some of its units conceived their task as that of rounding up witches
       to purify the country of corruption . . . ‘Witchcraft’ (that is, kindoki ) is
       not primitive thought ‘surviving’ in modern times and ‘adapting’ to
       the stresses of modernity, but is itself a mode of modernity.81
In much research on the peoples focused on here, as well as on
many other African peoples or societies, there has often been a ten-
dency to see them as more distinct and isolated than they are and
have been.82 Hence, there is a need for more research on the wider
political and socio-economic contexts with which they interact. Another
problem in many studies is the tendency to see witchery as a con-
servative force, that is, as something that people use in order to
defend themselves against changes. As remarked by Geschiere, how-
ever, there is an important challenge to explore the possibilities
offered by, among other things, witchery discourses to gain control
over modern changes.83 This means, also, that there is a need to go
beyond the common focus on witchery accusations, which tends to
relate witchery to the reproduction of social orders. Since it can be
dangerous to publicly accuse people in power, it is not frequently
done. Therefore, scholars who carry out field research need to ‘venture
into the more vague spheres and try to make sense of the turmoil

       MacGaffey 1997: vii f.
       MacGaffey 2000: 226 f.
       This point has also been stressed by, e.g., Tanner & Mitchell 2002: 22.
       Geschiere 1997: 14 f.
208                              chapter nine

of rumors or the highly ambivalent and elusive role of the healers
and how they affect social relations’.84
   In this final chapter the important increase in the role of living
humans as agents of disease, who have largely replaced spiritual
beings in this role, has been discussed historically and comparatively
with a focus on socio-economic and political issues. It can hardly be
denied that witchery, in a wide sense, is an eminently social phe-
nomenon. However, in much theorizing about witchery, aspects of
change have not been given enough consideration. While socio-eco-
nomic and political analyses of continuity and change are focused
on here, it should not be forgotten that there are other important
dimensions of witchery and similar phenomena too. For instance,
psychological aspects as well as issues of meaning and evil have not
been studied in depth.85 Thus, such dimensions are further exam-
ples of desiderata for future research on African etiologies of disease.

     Ibid., 219.
     Among others, Middleton and Winter (1963: 1) too onesidedly argue that beliefs
in witches and sorcerers ‘are social, not psychological, phenomena and must so be
analysed’. Cf., e.g., Prince (1961 and 1964) who in a valuable way deals with psy-
chological and psychiatric aspects of witchery among the Yoruba. On the issue of
witchery as an explanation for evil, see, e.g., Douglas 1999: 189 ff.


                           Popular African medicine has strong pragmatic ele-
                           ments and gives weight to natural explanation.1


In studies from the late nineteenth and early twentieth centuries it
was often reported that natural causation and treatment of diseases
played an important part in San medical ideas and practices.2 G.W.
Stow put it thus:
      The Bushmen certainly are acquainted with a number of very valu-
      able medicinal plants; some of them are specifics in the cure of sev-
      eral diseases which have frequently baffled the skill of the most eminent
      medical practitioners; and it is a matter of astonishment that no effort
      has been made to discover such important secrets. Thus they were
      able to effect certain cures in cases of snake-bite, taenia, dysentery,
      and calculus, besides the rapid removal of gonorrheal affections.3
Some later scholars, also, refer to the extensive knowledge about
plant medicines and provide examples of natural causes and methods
of curing.4 In modern works by anthropologists and scholars of reli-
gion, the occurrence of such references is not prominent.5 This may
partly be because of shifting interests of scholars, but it may also
indicate a deterioration of knowledge and skills in the natural and
biological aspects of San medicine.6

     Feierman 1985: 106.
     E.g., Nolte 1886; Lübbert 1901; Werner 1906: 258.
     Stow 1910: 125.
     E.g., Dornan 1925: 142, 162; Bleek 1928: 29; Schapera 1930: 215; Maingard
1937: 287 ff.; Drobec 1953: 124–31; Wilhelm 1954: 173 ff.
     For some short examples, though, see Marshall 1969: 370; Köhler 1971: 323;
Barnard 1979: 69 f.; and Smith et al. 2000: 70.
     Heinz’s statement that the Ko, a southern San language group in Botswana,
‘do not possess great medicinal knowledge’ is an interesting case in point (Heinz
1975: 30). In his article Heinz states that the Ko collect only eight or nine medi-
cinal plants of which they use the roots, stem or leaves to counteract certain types
of ailment. This information contrasts sharply with statements made by earlier
210                                 appendix


With regard to the Maasai, the issue of natural or physical causa-
tion is of particular importance. Early and modern sources seem to
agree that, for Maasai, natural causes are of paramount significance.
There is indeed some evidence that the ideas of such causes were
even more prevalent previously than during the last few decades. In
1904, for instance, Merker drew the conclusion that, to the Maasai,
these were the most important reasons for ill-health;7 and twenty-
three years later Berthold presented similar conclusions.8 In a more
recent study, Olsson says that ‘to the best of my knowledge, pas-
toral Maasai prefer to regard disease as physical influence upon, or
organic perturbation in the bodies of men and animals’.9 This is also
confirmed by Sindiga who, in an important article from 1995, stresses
the wide knowledge of herbal medicines. From an early age, both
boys and girls start acquiring knowledge of such medicines.10
    Ilabaak (sing. olabaani ) form a category of medical practitioners,
who may be referred to as (secular) doctors or curers and who must
not be confused with the iloibonok, whose art of healing has a religious
dimension. In the medicine used by ilabaak, as well as by ordinary
Maasai, it is the physical properties that count but, unlike those of
iloibonok, not any supranormal qualities. Moreover, the doctors perform,
among other things, operations, bone-setting and vaccinations too.11

scholars such as Stow (1910: 125) and Dornan (1925: 142, 162). Heinz says that
‘almost all’ of the Ko medical activities are restricted to ‘exorcising dances’. He
does not discuss historical issues here, but it may be asked whether there may have
been, concomitant with a decrease of knowledge of natural causes and remedies,
an increased significance of the religious aspects of Ko medicine. This question may
be asked, also, with regard to Guenther’s material on the recent resurgence of
Nharo religion.
      ‘Die Entstehung von inneren Krankheiten führen die Masai, im Gegensatz zu
den Negervölkern, nie auf das Tun böser Geister und nur selten auf einen gegen
den Erkrankten von einem seiner persönlichen Feinde bereiteten Zauber zurück.’
(Merker 1904: 174).
      Writing on the Maasai doctor, he argued appreciatively: ‘Seine Kunst ist eine
Kunst, Zauberei spielt keine Rolle in seinem Gewerbe, sondern eine geradezu
verblüffende Kenntnis der Anatomie, der Funktionen der verschiedenen Organe,
der heilenden Eigenschaften von Pflanzen, Erden, ja selbst von Stoffen aus Lebewesen’
(Berthold 1927: 5).
       Olsson 1989: 235.
       Sindiga 1995: 95.
       See, e.g., Jacobs 1965: 322; Galaty 1977: 298 f.; Voshaar 1979: 203 f., 211;
Sindiga 1995: 97 f.
                                   appendix                                  211

Several authors have provided more or less detailed accounts of the
wide medical knowledge and use of a great variety of natural reme-
dies, particularly herbal medicines, and diseases thought to be cured
by these.12 One Maasai way of increasing anatomical knowledge has
been by studying the anatomy of cattle.13 An example of the treat-
ment of patients suffering from venereal diseases, which at times
have also been a great problem among the Maasai, is the custom
of enforcing quarantine on active sufferers from syphilis.14 The preva-
lence of venereal diseases is, in part, a consequence of liberal views
on sexual relations, although the ‘promiscuous’ life of Maasai war-
riors has been exaggerated by some authors.15 It is interesting to
note, also, that health reasons have been given for the removal of
the two central lower incisors. If an ill person keeps his or her mouth
tightly closed, food and drink can be poured in through the hole.16


Among the Sukuma, most diseases that are thought to have natural
causes are minor and transitory. Such maladies do not require
divination. According to Reid, congenital illnesses are also consid-
ered to be caused naturally.17 Although fertility problems are often
attributed to ancestors, even serious afflictions such as sterility and
impotence may well have natural causes.18 From the Nyamwezi area,
Bösch reported that deaths might be attributed to natural causes,

      Baumann 1894: 162 f.; Merker 1904: 174 ff., 340 ff.; Hollis 1905: 335 ff.;
Fuchs 1910: 122; Berthold 1927: 5 ff.; Jacobs 1965: 139 ff.; Sankan 1971: 59 ff.;
Galaty 1977: 298 f.; Voshaar 1979: 323 f.; Landei 1982: 44 f.; Hurskainen 1984:
204 ff.; Sindiga 1995: 101. Cf., e.g., Nüssler 1931: 172; Buchta 1932: 266. For an
extensive study of ethnobotany among the Mukogodo Maasai, see Brenzinger, Heine
& Heine 1995.
      Saibull & Carr 1981: 75.
      McKay 1950: 454 f. According to McKay, this ‘excellent custom’ was, unfor-
tunately, discontinued among those Purko Maasai he studied when European doc-
tors started treating syphilis. See also Koenig 1956: 97.
      E.g., Thomson 1883/84: 252. In recent decades the problem of venereal dis-
eases, including AIDS, appears to have worsened (Voshaar 1979: 317 f.).
      Leakey 1930: 187. A Maasai theory about internal worms has been reported
by Buchta (1939: 228), but neither he nor, as far as I know, any other scholar has
elaborated on this interesting theory.
      Reid 1969: 71 f., 80.
      Table n.d.a.: 2; Reid 1969: 78.
212                                 appendix

although in most cases the balogi, witches/sorcerers, were held
   There is extensive knowledge about hygiene and household remedies.
If an affliction is not too serious, people first try some self-therapy
at home.20 Bösch claimed that at his time the Nyamwezi lay peoples
knew more about anatomy than did their western counterparts.21
Some remedies or medicines for minor ailments, for instance emetics
and laxatives, are known by virtually all Sukuma.22 In addition to
the various healers, bafumu, who are specialized in dealing with
suprahuman powers and causes of illness, there are other bafumu, or
‘secular doctors’, who treat naturally caused afflictions.23 Even a mod-
ern medical doctor can be called mfumu.24
   There is a great knowledge of herbal medicines. ‘Most adult Sukuma
can name and identify hundreds of plants, shrubs and trees with their
uses, most of which will be medicinal.’25 Hatfield says that there is an
amazing variety of substances used in medicines,26 and Tanner con-
cludes that the Sukuma have ‘an enormous pharmacopaeia’, probably
extending to more than a thousand plants, each of which may have
several uses.27 In another work, Tanner says that the bafumu are char-
acterized by an ‘almost frenetic activity to seek for greater knowl-
edge’.28 Often the ‘natural’ efficacy of plant medicines merge with the
symbolic or ‘magic’ potential of the same, or added, ingredients.29

     ‘Les indigènes eux-mêmes admettent bien qu’on peut mourir de mort naturelle,
mais pratiquement ils rendent presque toujours les balogi responsables d’un mort’
(Bösch 1930: 225).
     Table n.d.a.: 113 ff., 136 f.; Tanner 1959: 119; Reid 1969: 230.
     Bösch 1930: 229. See also Cory n.d.a.: 1.
     Table n.d.a.: 151.
     A curer who treats such afflictions can be called mfumu naguji or ngota. See fur-
ther Tanner 1967: 43 and Ng’weshemi 1990: 22. Cf. Hatfield 1968: 110.
     Brandström 1990, chapter 6: 9. At least external surgery is practised. See fur-
ther Cory n.d.a.: 2; Table n.d.a.: 167, 170, 590; Hatfield 1968: 111. Cf. Tanner
1959: 120 f. Concerning the Nyamwezi, see Bösch 1930: 286 ff., 293.
     Wijsen & Tanner 2002: 47.
     Hatfield 1968: 111.
     Tanner 1957: 119. Some lists of medicines are found in, for instance, Bösch
1930: 289 ff. and Table n.d.a.: 591 ff. See further Augustiny 1923/24: 171; Bösch
1930: 286; Table n.d.a.: 173, 589 f.
     Tanner 1967: 48.
     Many examples are rendered in the long lists of remedies in Table n.d.a:
155–178. The concept of medicine, bugota, implies both ‘natural’ and ‘magical’
aspects (Hatfield 1968: 83; Steeves 1990: 70 f.). See also Bösch 1930: 286 and
Table n.d.a.: 586. Unlike organized cult activities and divination, which have
                                    appendix                                    213


As among the Sukuma, for instance, diseases that have natural
causes—‘illnesses of God’—are generally mild or benign afflictions
which respond readily to therapy and are not related to particular
disturbances in the immediate social relationships of the sufferer.
Common and simple problems like headache, stomach-ache, cold,
fever or accidental wounds may belong to this category. Moreover,
diseases or deaths of old people or neonates are often considered to
be natural. Bouquet and Janzen, among others, report that there is
much empiricism and experimentation in Kongo medicine.30 However,
if an affliction does not respond to symptomatic treatment, then it
is suspected to be caused by spiritual beings or living humans.31
   Minor ailments are frequently treated at home and, as remarked
by Dalmalm, all Congolese are potential curers.32 A specialist on the
treatment of afflictions with natural causes is called nganga mbuki (or
nganga buka). Such a curer, who in English is often referred to as a
‘herbalist’, uses medicines (bilongo) which are not connected to nkisi
objects and spirits. While missions and colonial rulers condemned
banganga who dealt with problems of nkisi and kindoki, they tended to
accept the nganga mbuki. As a result, banganga increasingly tried to
adopt discreet identities as ‘herbalists’. In independent Zaire the
campaign of ‘authenticity’ included, among other things, a certain
exaltation of traditional medicine.33 The treatment of diseases with
natural causes may involve, in addition to the use of herbal medi-
cines, scarifications, incisions, bone setting, steam baths, massage and
head-pack applications. There is an extensive knowledge of plant
medicines, but indigenous methods of surgery are not very advanced
and nowadays largely ignored.34

diminished, medicines seem to flourish. Especially in towns there has been a pro-
liferation of individual practitioners who practise medicine without much indige-
nous training and equipment (Hatfield 1968: 299; Tanner 1969: 288).
       Bouquet 1969: 36; Janzen 1978: 198.
       Janzen 1978: 22, 48, 127. See further, e.g., Mahaniah 1973: 229; Mahaniah
1980: 593; Pambou 1980–81: 59; Dimomfu 1984: 26 ff.; Dalmalm 1985: 77.
       Dalmalm 1985: 79. See also Janzen 1978: 64; Pambou 1980–81: 60; Dimomfu
1984: 24.
       Van Wing 1959: 419; Mahaniah 1973: 229; Janzen 1978: 45, 53, 57; Janzen
1979: 214 f.; Pambou 1979–80: 44; Dalmalm 1985: 63. Cf. Dimomfu 1984: 12 ff.
       For some more details, see Bouquet 1969: 33 f.; Janzen 1978: 179 ff.; Jacobson-
Widding 1979: 152; Janzen 1979: 208; Pambou 1979–80: 45 f.; Batukezanga 1981:
62; Mahaniah 1982: 23, 54 ff.; Lagercrantz 1983/84: 85; Dimomfu 1984: 26.
214                                   appendix


Like the Maasai, Sukuma, Kongo and probably most African peo-
ples, the Yoruba distinguish between the roles of ‘curers’ and ‘healers’.
Although there is, in practice, some overlapping between the different
types of experts, an onisegun is basically a doctor or curer who deals
primarily with naturally or physically occasioned afflictions, whereas
a babalawo, or some other priest or diviner, is an expert on religiously
and socially based diagnoses and therapies. A great multitude of
minor ailments can be treated at home or by an onisegun or some
other type of curer. Associations for herbalists have existed in
Yorubaland since the nineteenth century, and there has recently been
a proliferation of such associations.35 In some works, lists of various
herbal medicines are provided.36 Medicinal treatments, as well as
theories of causation, frequently combine the physical or pharma-
cological element with religious and social aspects.37
   The significance of symbolic features such as colour, smell and
flavour is discussed in detail by Buckley.38 Yet he emphasizes that
the importance of symbolism must not be exaggerated. ‘Folk medi-
cine’, he says, is primarily an instrumental system of techniques for
dealing practically with certain human ills.39 Although the represen-
tativeness of Buckley’s few informants may be questioned, it is inter-
esting to note that, according to them, not only minor afflictions but
the great majority of illnesses are due to natural causes. For instance,
a wide variety of diseases, such as dysentery, dropsy, yaws (= frambesia,
a contagious skin disease) and some psychiatric problems, may be
caused, wholly or at least partly, by invisible worms in the body. In
addition to such worms, there are germs that occasion sicknesses;
and impure or abnormal blood may be the cause of, among other
things, rheumatism and sterility. Particularly the significance of worms
and germs in the body, which is discussed in detail in Buckley’s

      See further, e.g., Table n.d.b.: 200 f.; Ajose 1957: 269 f.; Leighton et al. 1963:
113; Bascom 1969a: 70; Maclean 1971: 75, 82 f.; Verger 1971: 50, 52; Asuni 1976:
4; Maclean 1976: 315; Braito & Asuni 1979: 188; Wolff 1979: 127; Simpson 1980:
93 ff.; Oyebola 1981: 87, 89.
      E.g., Table n.d.b.: 201–206; Verger 1971: 50–55; Simpson 1980: 150 ff.
      Margetts 1965: 116; Maclean 1971: 83; Ayoade 1979: 49.
      Buckley 1985a: 45 ff., 53 ff. 70 ff., 86 ff.
      Ibid., 108.
                                    appendix                                    215

Yoruba Medicine (1985), has been observed by many other scholars too.40
According to Buckley’s informants, the worms and germs within the
body contribute to the health of the body. However, too much sweet
food, sexual activity or alcohol can make them ‘too strong’ or ‘too
powerful’, and then they cause diseases.41
   Other natural causes of illness are, for instance, faulty diet, hered-
itary factors, hemp smoking and other toxins. It is interesting to
observe that, among Simpson’s informants, ‘non-healers’ spoke more
often about natural causes than did the ‘healers’, who were more
familiar with the powerful suprahuman and human forces and who
had a personal stake in healing.42 Perhaps the corresponding great
familiarity with and personal stake in ‘herbalism’ of Buckley’s infor-
mants are part of the explanation why Yoruba medicine, in their
view, is a basically naturalistic system.43 Besides herbal treatment,
there are indigenous methods used to perform, for instance, cir-
cumcision, scarification and bone setting. ‘Traditional’ surgery, how-
ever, is largely carried out by Hausa experts who travel around
western Nigeria practising their trade.44 When there are epidemic
diseases, like smallpox, isolation and measures of disinfection are
means of dealing with the problems.45
   While many scholars, such as Dopamu, Simpson and Oladapo,
emphasize the importance of rituals, incantations and ‘magic’ that
are combined with natural types of treatment, Buckley stresses pri-
marily the ‘underlying rationality’ of indigenous Yoruba medicine.
‘Within Yoruba medicinal practice there is a strong element of crit-
ical appraisal akin to, although not identical with, the spirit of scientific
enquiry to be found in the West.’46 Consequently, Buckley’s infor-
mants and others ‘constantly subject their medicinal knowledge to
empirical criticism’.47

     Prince 1964: 89; Lucas & Hendrickse 1971: 35; Dopamu 1977: 458; Wolff
1979: 128; Simpson 1980: 98, 100 ff.; Oladapo 1984: 127; Buckley 1985a: 25 ff.
     Ibid., 32 f. Cf. Simpson (1980: 3) who says that, according to one of his infor-
mants, the deity Obatala may attack people with worms.
     Simpson 1980: 108.
     Cf. further Leighton et al. 1963: 110, 113; Prince 1964: 88 f.; Maclean 1971:
125; Ayoade 1979: 49; Wolff 1979: 128; Simpson 1980: 100–109; Oladapo 1984:
     Asuni 1976: 4; Simpson 1980: 97.
     Ajose 1957: 270.
     Buckley 1985a: 159.
     Ibid., 161

Abimbola, W. 1976. Ifá: An Exposition of Ifá Literary Corpus. Ibadan: Ibadan University
Abrahams, R.G. 1967. The Peoples of Greater Unyamwezi: Tanzania (Nyamwezi, Sukuma,
  Sumbwa, Kimbu, Konongo). Ethnographic Survey of Africa, East Central Africa, Part
  17. London: International African Institute.
Adeniyi, M.A. 1984. ‘The Concept of Spirits in Yoruba Land: Ede as a Case Study’.
  MA thesis, Institute of African Studies, University of Ibadan.
Ajose, O.A. 1957. ‘Preventive Medicine and Superstition in Nigeria’, Africa 27 (3),
Apter, A. 1993. ‘Atinga Revisited: Yoruba Witchcraft and the Cocoa Economy,
  1950–1951’, in J. Comaroff & J. Comaroff (eds.), Modernity and Its Malcontents:
  Ritual and Power in Postcolonial Africa, 111–128. Chicago & London: Chicago
  University Press.
Århem, K. 1985. The Symbolic World of the Maasai Homestead. Working Papers in
  African Studies 10. Uppsala: African Studies Programme, Department of Cultural
  Anthropology, University of Uppsala.
——. 1987. Milk, Meat and Blood: Diet as a Cultural Code among the Pastoral Maasai.
  Working Papers in African Studies 35. Uppsala: African Studies Programme,
  Department of Cultural Anthropology, University of Uppsala.
——. 1989. ‘Why Trees are Medicine: Aspects of Maasai Cosmology’, in A. Jacobson-
  Widding & D. Westerlund (eds.), Culture, Experience and Pluralism: African Ideas of
  Illness and Healing, 75–84. Uppsala Studies in Cultural Anthropology 13. Stockholm:
  Almqvist & Wiksell International.
Asuni, T. 1976. ‘Modern Medicine and Traditional Medicine’. Unpublished paper,
  Fortbildungsseminar für Ärzte und medizinisches Pflegepersonal der Arbeits-
  gemeinschaft für Entwicklungshilfe, 14–27.11.76, Ibadan, Nigeria.
Augé, M. 1986. ‘L’Anthropologie de la maladie’, L’Homme 26 (97–98), 81–90.
Augouard, R.P. 1885. ‘Mœurs et coutumes de la région de Stanley-Pool’, L’Écho
  des Missions d’Afrique 2 (7), 103–108.
Augustiny, J. 1923–24. ‘Sukuma-Texte (gesammelt und übersetzt von J. Augustiny),
  V’, Zeitschrift für Eingeborenen-Sprachen 14, 164–172.
Awolalu, J.O. 1970. ‘The Yoruba Philosophy of Life’, Présence Africaine (73), 20–38.
——. 1979. Yoruba Beliefs and Sacrificial Rites. London: Longman.
Awolalu, J.O., & P.A. Dopamu. 1979. West African Traditional Religion. Ibadan:
  Onibonoje Press.
Axelson, S. 1970. Culture Confrontation in the Lower Congo: From the Old Congo Kingdom
  to the Congo Independent State with Special Reference to the Swedish Missionaries in the
  1880s and 1890s. Studia Missionalia Upsaliensia 14. Stockholm: Gummessons.
——. 1983/84. ‘Kindoki—ett språk på gott och ont’, Religion och Bibel 42/43, 23–30.
Ayoade, J.A.A. 1979. ‘The Concept of Inner Essence in Yoruba Traditional Medicine’,
  in Z.A. Ademuwagun et al. (eds.), African Therapeutic Systems, 49–55. Waltham,
  MA: Crossroads Press.
Balina, A., A. Mayala & J.M. Mabula. 1971. Sukuma Expression of Traditional Religion
  in Life. Tabora: Kipalapala Seminary.
Baltus, N. 1898. ‘La Maladie du sommeil’, Missions en Chine et au Congo (113), 77.
Barbee, E.L. 1986. ‘Biomedical Resistance to Ethno-medicine in Botswana’, Social
  Science and Medicine 22 (1), 75–81.
Barber, K. 1981. ‘How Man Makes God in West Africa: Yoruba Attitudes Towards
  the Orisa’, Africa 51 (3), 724–745.
218                                   references

Barnard, A. 1979. ‘Nharo Bushman Medicine and Medicine Men’, Africa 49 (1),
——. 1988. ‘Structure and Fluidity in Khoisan Religious Ideas’, Journal of Religion
  in Africa XVIII (3), 216–236.
Barnes-Dean, V.L. 1986. Lugbara Illness Belief and Social Change’, Africa 56,
Barthelemy, J. 1905. ‘Kamoga (Bukumbi)’, Chroniques des Pères Blancs (114), 285–288.
Bascom, W. 1969a. The Yoruba of Southwestern Nigeria. New York: Holt, Rinehart &
——. 1969b. Ifa Divination: Communication between Gods and Men in West Africa. Bloomington
  & London: Indiana University Press.
Batukezanga, Z. 1981. Guérir le malade et la maladie: L’Infirme et son univers en Afrique
  Noire. Kinshasa: Presses Kikesa.
Baumann, O. 1894. Durch Masailand zur Nilquelle: Reisen und Forschungen der Massai-
  Expedition des deutschen Antisklaverei-Komite in den Jahren 1891–1893. Berlin: D. Reimer.
Baxter, P.T.W. 1972. ‘Absence Makes the Heart Grow Fonder: Some Suggestions
  Why Witchcraft Accusations Are Rare among East African Pastoralists’, in M.
  Gluckman (ed.), The Allocation of Responsibility, 163–191. Manchester: Manchester
  University Press.
Beckwith, C., & T. Ole Saitoti. 1980. Maasai. New York: Elm Tree Books.
Beidelman, T.O. 1970. ‘Towards More Open Theoretical Interpretations’, in M.
  Douglas (ed.), Witchcraft Confessions and Accusations, 351–356. ASA Monographs 9.
  London: Tavistock.
Beier, U. 1980. Yoruba Myths. Compiled and introduced by U. Beier. Cambridge:
  Cambridge University Press.
——. 1999. Auf dem Auge Gottes wäscht kein Grass: Zur Religion, Kunst und Politik der
  Yoruba und Igbo in Westafrika. Wuppertal: Peter Hammer.
Benson, J.S. 1974. ‘A Study of the Religious Beliefs and Practices of the Maasai
  Tribe and the Implications on the Work of the Evangelical Lutheran Church in
  Tanzania’, STM dissertation, Northwestern Lutheran Theological Seminary, St.
  Paul, Minnesota.
Berg-Schlosser, D. 1984. Tradition and Change in Kenya: A Comparative Analysis of Seven
  Major Ethnic Groups. Paderborn: Ferdinand Schöningh.
Bernardi, B. 1959. The Mugwe: A Failing Prophet. London, New York & Toronto:
  Oxford University Press for the International African Institute.
Bernaut, F., & J. Tonda. 2000. ‘Dynamiques de l’invisible en Afrique’, Politique
  africaine (79), 5–16.
Berntsen, J.L. 1973. ‘Maasai and Iloikop: Ritual Experts and their Followers’, MA
  thesis, University of Wisconsin, Madison.
——. 1979. ‘Pastoralism, Raiding, and Prophets: Maasailand in the Nineteenth
  Century’, Ph.D. thesis, Department of History, University of Wisconsin, Madison.
Berthold, H. 1927. ‘Der Massai-Doktor’, Der Kolonialfreund (5), 5–7.
Binsbergen, W. van, & M. Schoffeleers. 1985. ‘Introduction: Theoretical Explorations
  in African Religion’, in W. van Binsbergen & M. Schoffeleers (eds.), Theoretical
  Explorations in African Religion, 1–49. London: KPI.
Bjerke, S. 1989. ‘Witchcraft as Explanation: The Case of the Zinza’, in A. Jacobson-
  Widding & D. Westerlund (eds.), Culture, Experience and Pluralism: African Ideas of
  Illness and Healing, 219–233. Uppsala Studies in Cultural Anthropology 13. Stockholm:
  Almqvist & Wiksell International.
Bleek, D.F. 1927. ‘Buschmänner von Angola’, Archiv für Anthropologie XXI (1–2), n.s.,
——. 1928. The Naron: A Bushman Tribe of the Central Kalahari. Cambridge: Cambridge
  University Press.
——. 1929. ‘Bushman Folklore’, Africa 2 (3), 302–313.
                                      references                                       219

Bleeker, S. 1963. The Masai: Herders of East Africa. London: Dennis Dobson.
Blohm, W. 1933. Die Nyamwezi: Gesellschaft und Weltbild, II. Hamburg: Friederichsen,
  de Gruyter.
Blokland, H. 2000. ‘Kings, Spirits & Brides in Unyamwezi, Tanzania’, in R. van
  Dijk, R. Reis & M. Spierenburg (eds.), The Quest for Fruition through Ngoma: Political
  Aspects of Healing in Southern Africa. Cape Town: David Philip.
Blumer, L. 1927. ‘Streiflichter auf Religion und religiöses Leben der Arushaleute’,
  Evangelisches Lutherisches Missionsblatt 82, 75–82.
Boon, J.A. 1982. Other Tribes, Other Scribes: Symbolic Anthropology in the Comparative Study
  of Cultures, Histories, Religions, and Texts. Cambridge: Cambridge University Press.
Bösch, P.F. 1930. Les Banyamwezi: Peuple de l’Afrique Orientale. Münster: Aschendorffsche
Boucher, A. 1928. Au Congo français: Les Missions catholiques. Paris: Tequi.
Bouquet, A. 1969. Feticheurs et médicines traditionelles du Congo (Brazzaville). Paris:
Brain, J.L. 1973. ‘Ancestors as Elders in Africa: Further Thoughts’, Africa 43 (2),
Braito, R., & T. Asuni. 1979. ‘Traditional Healers in Abeokuta: Recruitment,
  Professional Affiliation, and Types of Patients Treated’, in Z.A. Ademuwagun
  et al. (eds.), African Therapeutic Systems, 187–190. Waltham, MA: Crossroads Press.
Brandström, P. 1986. Who is a Sukuma and Who is a Nyamwezi? Ethnic Identity in West-
  Central Tanzania. Working Papers in African Studies 27. Uppsala: African Studies
  Programme, Department of Cultural Anthropology, University of Uppsala.
Brandström, P. 1990. ‘Boundless Universe: The Culture of Expansion among the
  Sukuma-Nyamwezi of Tanzania’, Ph.D. thesis, Department of Cultural Anthropology,
  University of Uppsala.
Brard, P. 1897. ‘Rapport du P. Brard sur les tribus insulaires du Nyanza Méridional’,
  Chroniques des Pères Blancs (73), 152–157.
Braun, W., & R.T. McCutcheon (eds.). 2000. Guide to the Study of Religion. London
  & New York: Cassell.
Brenner, L. 2000. ‘Histories of Religion in Africa’, Journal of Religion in Africa 30 (2),
Brenzinger, M., B. Heine & I. Heine. 1995. The Mukogodo Maasai: An Ethnobotanical
  Survey. Köln: Rüdiger Köppe Verlag.
Buakasa, G. 1968. ‘Notes sur le Kindoki chez les Kongo’, Cahiers des Religions Africaines
  2 (3), 153–169.
Buakasa, T.K.M. 1973. L’Impensé du discours: ‘Kondoki’ et ‘nkisi’ en pays Kongo du Zaïre.
  Kinshasa: Presses Universitaires du Zaïre.
Buchta, H. 1932. ‘Aus meiner Arbeit an den Masai’, Evangelisch-Lutherisches Missionsblatt
  87 (9), 266.
——. 1939. ‘Meine Krankenarbeit an den Masai’, Evangelisch-Lutherisches Missionsblatt
  94 (9), 228–231.
Buckley, A.D. 1976. ‘The Secret: An Idea in Yoruba Medicinal Thought’, in J.B.
  Loudon (ed.), Social Anthropology and Medicine, 396–421. ASA Monograph 13.
  London, New York & San Francisco: Academic Press.
——. 1985a. Yoruba Medicine. Oxford: Clarendon Press.
——. 1985b. ‘The God of Smallpox: Aspects of Yoruba Religious Knowledge’,
  Africa 55 (2), 187–200.
Burns, Fr. 1908. ‘The Bamasai: A British East African Tribe’, Mill Hill Missionaries,
  Publication (9), 168–171.
Butaye, P. 1899. ‘Les Fétiches et les maléfices’, Missions Belges de la Compagnie de
  Jésus, 309–315.
Buxton, J. 1973. Religion and Healing in Mandari. Oxford: Clarendon Press.
Cambier, M.E. 1890. Untitled letter. Missions en Chine et au Congo (23), 367.
220                                   references

Chavunduka, G.L. 1978. Traditional Healers and the Shona Patient. Zambesiana 3.
  Gwelo: Mambo Press.
——. 1986. ‘Development of African Traditional Medicine: The Case of Zimbabwe’,
  in U. Maclean & C. Fyfe (eds.), African Medicine in the Modern World, 59–72.
  Seminar Proceedings 27. Edinburgh: Centre of African Studies, University of
Chavunduka, G.L., & M. Last. 1986. ‘Conclusion: African Medical Professions
  Today’, in M. Last & G.L. Chavunduka (eds.), The Professionalisation of African
  Medicine, 262–264. Manchester: Manchester University Press in association with
  the International African Institute.
Ciekawy, D., & P. Geschiere. 1998. ‘Containing Witchcraft: Conflicting Scenarios
  in Postcolonial Africa’, African Studies Review 41 (3), 1–14.
Circulaire 1926. ‘Circulaire au sujet du Nkassa (poison d’épreuve)’. Unpublished
  document from the Vicariat Apostolique de Brazaville in the archive of the Holy
  Ghost Fathers in Chevilly.
Colson, E. 2000. ‘The Father as Witch’, Africa 70 (3), 333–358.
Comaroff, J. 1982. ‘Medicine: Symbol and Ideology’, in P. Wright & A. Treacher
  (eds.), The Problem of Medical Knowledge: Examining the Social Construction of Medicine,
  50–67. Edinburgh: Edinburgh University Press.
Comaroff, J., & J. Comaroff. 1993. ‘Introduction’, in J. & J. Comaroff (eds.), Modernity
  and Its Malcontents, xi-xxxvii. Chicago & London: University of Chicago Press.
Comstock, G.L. 1979. ‘The Yoruba and Religious Change’, Journal of Religion in
  Africa 10 (1), 1–11.
Cory, H. 1949. ‘The Ingredients of Magic Medicine’, Africa 19 (1), 13–32.
——. 1951. The Ntemi: Traditional Rites in Connection with the Burial, Election, Enthronement
  and Magic Powers of a Sukuma Chief in Tanganyika. London: Macmillan.
——. 1960. ‘Religious Beliefs and Practices of the Sukuma/Nyamwezi Tribal Group’,
  Tanganyika Notes and Records (54), 14–26.
——. n.d.a. ‘Basukuma’. Typescript in the Cory manuscript collection, no. 44, E.A.
  section of the University Library of Dar es Salaam.
——. n.d.b. ‘Bufumu wa Ngoko’. Typescript in the Cory manuscript collection, no.
  59, E.A. section of the University Library of Dar es Salaam.
Cox, J.L. 2001. ‘Missionaries, the Phenomenology of Religion and “Re-Presenting”
  Nineteenth-Century African Religion: A Case Study of Peter McKenzie’s Hail
  Orisha! ’, Journal of Religion in Africa 31 (3), 336–353.
Crowther, S.A. 1855. ‘Journal of Samuel Crowther for the quarter ending Sept.
  25, 1855’. Manuscript in the archive of the White Fathers in Rome.
——. 1892. Experiences with Heathens and Mohammedans in West Africa. London: Society
  for Promoting Christian Knowledge.
Dahlin, O. 2002. Zvinorwadza: Being a Patient in the Religious and Medical Plurality of the
  Mberengwa District, Zimbabwe. European University Studies, Series XXIII: Theology,
  748. Frankfurt am Main: Peter Lang.
Dallas, D. 1931. ‘The Sacred Tree of Ol Donyesha’, Man 31 (42–43), 39–41.
Dalmalm, Å. 1985. L’Église à l’épreuve de la tradition: La Communauté évangélique du Zaïre
  et le kindoki. Paris: Editaf.
Decken, C.C. von der. 1871. Reisen in Ost-Afrika in den Jahren 1862 bis 1865. Leipzig
  & Heidelberg: Wintersche Verlagshandlung.
Denis, D.C. 1979. ‘La Mort—sa conception—les cérémonies chez les Fon et les
  Yoruba de Quidah’, Africa: Revista do Centro de Estudos Africanos da USP 2, 13–44.
Dennett, R.E. 1910 [1968]. Nigerian Studies or the Religious and Political System of the
  Yoruba. New impression 1968. London: Frank Cass.
Dewilde, J. 1894. Untitled letter. Missions en Chine et au Congo (60), 379.
Dimomfu, L. 1984. L’Art de guérir chez les Kongo du Zaïre: Discours magique ou science
  medicale? Les Cahiers du CEDAF 3. Brussels: ASDOC-Studies.
                                         references                                          221

Donovan, V. 1982. Christianity Rediscovered. 2nd edn. Maryknoll, NY: Orbis Books.
Dopamu, P.A. 1977. ‘The Practice of Magic and Medicine in Yoruba Traditional
   Religion’. Ph.D. thesis, Department of Religious Studies, University of Ibadan.
Dornan, S.S. 1925. Pygmies and Bushmen of the Kalahari. London: Seeley.
Douglas, M. 1970. ‘Introduction: Thirty Years after Witchcraft, Oracles and Magic’, in
   M. Douglas (ed.), Witchcraft Confessions and Accusations, xiii–xxxvi. ASA Monographs
   9. London: Tavistock.
——. 1999. ‘Sorcery Accusations Unleashed: The Lele Revisited’, Africa 69 (2),
Drennan, M.R. 1937. ‘Finger Mutilation in the Bushman’, Bantu Studies XI, 247–249.
Drobec, E. 1953. ‘Die Krankenbehandlung bei den Pygmäen und Buschmännern’,
   Zeitschrift für Ethnologie 78 (1), 124–135.
Droogers, A. 1985. ‘From Waste-Making to Recycling: A Plea for an Eclectic Use
   of Models in the Study of Religious Change’, in W. van Binsbergen & M.
   Schoffeleers (eds.), Theoretical Explorations in African Religion, 101–137. London: KPI.
Eades, J.S. 1980. The Yoruba Today. Cambridge: Cambridge University Press.
Estermann, C. 1949. ‘Études et scènes africaines: Les Bochimans du sud de l’Angola’,
   Annales Spiritaines 59 (2), 29–31.
Evans-Pritchard, E.E. 1937. Witchcraft, Oracles and Magic among the Azande. Oxford:
   Clarendon Press.
——. 1956. Nuer Religion. New York & Oxford: Oxford University Press.
Extrait 1895. ‘Extrait d’une lettre du Frère Piessens’, Missions en Chine et au Congo
   (72), 576.
Fabian, J. 1985. ‘Religious Pluralism: An Ethnographic Approach’, in W. van
   Binsbergen & M. Schoffeleers (eds.), Theoretical Explorations in African Religion, 138–163.
   London: KPI.
Farrow, S.S. 1926. Faith, Fancies and Fetich or Yoruba Paganism: Being Some Account of
   the Religious Beliefs of the West African Negroes, Particularly of the Yoruba Tribes of Southern
   Nigeria. London: Society for Promoting Christian Knowledge.
Feierman, S. 1985. ‘Struggles for Control: The Social Roots of Health and Healing
   in Modern Africa’, African Studies Review 28 (2/3), 73–147.
——. 1995. ‘Healing as Social Criticism in the Time of Colonial Conquest’, African
   Studies 54 (1), 73–88.
Fischer, G.A. 1882/83. ‘Bericht über die im Auftrage der Geographischen Gesellschaft
   in Hamburg unternommene Reise in das Massai-Land’, Mitteilungen der Geographischen
   Gesellschaft in Hamburg 1882–83, Vol. 1. Hamburg: L. Friederischen & Co.
Fleisch, P. 1936. Hundert Jahre lutherischer Mission. Leipzig: Verlag der Evang.-Luth.
Flood. G. 1999. Beyond Phenomenology: Rethinking the Study of Religion. London & New
   York: Cassell.
Fokken, H. 1914. Spruchweisheit der Masai. Leipzig: Verlag der Evang.-Luth. Mission.
——. 1917. ‘Gottesanschauungen und religiöse Überlieferungen der Masai’, Archiv
   für Anthropologie 43 (15), n.s., 237–252.
Forde, D. 1951. The Yoruba-Speaking Peoples of South-Western Nigeria. Ethnographic Survey
   of Africa, Western Africa, Part IV. London: International African Institute.
Foster, G.M. 1976. ‘Disease Etiologies in Non-Western Medical Systems’, American
   Anthropologist 78 (4), 773–782.
Fourie, C. 1994. Living Legends of a Dying Culture: Bushmen Myths, Legends and Fables.
   Edited and illustrated by C. Fourie. Hartbeespoort: Ekogilde.
Fratkin, E. 1991. ‘The Loibon as Sorcerer: A Samburu Loibon among the Ariaal
   Rendille’, Africa 61 (3), 318–333.
Fuchs, H. 1910. Sagen, Mythen und Sitten der Masai: Nach der Masaisprache und dem
   Englischen. Jena: Hermann Costenoble.
Galaty, J.G. 1977. ‘In the Pastoral Image: The Dialectic of Maasai Identity’, Ph.D.
   thesis, Department of Anthropology, University of Chicago.
222                                     references

Gall, S. 2001. The Bushmen of Southern Africa: Slaughter of the Innocent. London: Chatto
  & Windus.
Gass, M. 1973 [1919]. Croyances, magie et superstitions des Bagwe ou Basukuma établis au
  sud du lac Victoria-Nyanza. Città del Vaticano: Annali del Pontificio Museo Missionario
  Etnologico XXXVII.
Geschiere, P. 1996. ‘Sorcellerie et politique: Les pièges du rapport élite-village’,
  Politique africaine (63), 82–93.
—— 1997. The Modernity of Witchcraft: Politics and the Occult in Postcolonial Africa (Sorcellerie
  et politique en Afrique —la viande des autres). Translated by P. Geschiere & J. Roitman.
  Charlottesville & London: University Press of Virginia.
——. 2000. ‘Sorcellerie et modernité: retour sur une étrange complicité’, Politique
  africaine (79), 17–32.
Geschiere, P., & C. Fisiy. 1994. ‘Domesticating Personal Violence: Witchcraft, Courts
  and Confessions in Cameroon’, Africa 64 (3), 323–341.
Gillies, E. 1976. ‘Causal Criteria in African Classifications of Disease’, in J.B. Loudon
  (ed.), Social Anthropology and Medicine, 358–395. ASA Monograph 13. London, New
  York & San Francisco: Academic Press.
Good, C.M. 1987. Ethnomedical Systems in Africa: Patterns of Traditional Medicine in Rural
  and Urban Africa. New York & London: Guilford Press.
Guenther, M.G. 1975. ‘The Trance Dancer as an Agent of Social Change among
  the Farm Bushmen of the Ghanzi District’, Botswana Notes & Records 7, 161–166.
——. 1975/76. ‘The San Trance Dance: Ritual and Revitalization among the Farm
  Bushmen of the Ghanzi District, Republic of Botswana’, Journal der Südwestafrikanische
  Wissenschaftliche Gesellschaft 30, 45–53.
——. 1979. ‘Bushman Religion and the (Non)sense of Anthropological Theory of
  Religion’, Sociologus 29 (2), 102–132.
——. 1980. ‘From “Brutal Savages” to “Harmless People”: Notes on the Changing
  Western Image of the Bushmen’, Paideuma (Mitteilungen zur Kulturkunde) 26, 123–140.
——. 1981. ‘Men and Women in Nharo Belief and Ritual’, Namibiana 3 (2), 17–24.
——. 1986. The Nharo Bushmen of Botswana: Tradition and Change. Quellen zur Khoisan-
  Forschung 3. Hamburg: Helmut Buske Verlag.
——. 1992. ‘“Not a Bushman Thing”: Witchcraft among the Bushmen and Hunter-
  Gatherers’, Anthropos 87 (1/3), 83–107.
——. 1999. Tricksters and Trancers: Bushman Religion and Society. Bloomington &
  Indianapolis: Indiana University Press.
Gulliver, P.H. 1951. A Preliminary Survey of the Turkana: A Report Compiled for the
  Government of Kenya. Communications from the School of African Studies, new
  series no. 26, University of Cape Town.
——. 1963. Social Control in an African Society: A Study of the Arusha; Agricultural Masai
  of Northern Tanganyika. London: Routledge & Kegan Paul.
——. 1969. ‘The Conservative Commitment in Northern Tanzania: The Arusha
  and Maasai’, in P.H. Gulliver (ed.), Tradition and Transition in East Africa: Studies of
  the Tribal Element in the Modern Era, 223–242. London: Routledge & Kegan Paul.
Gusinde, M. 1952. ‘Bericht über meine südafrikanische Forschungsreise 1950/51’,
  Anthropos 47 (3–4), 388–404.
——. 1963/64–1964/65. ‘Aus der Geistesleben der Hukwe beim Caprivizipfel’,
  Journal der Südwestafrikanische Wissenschaftliche Gesellschaft 18–19, 35–42.
——. 1965. ‘Die Religionsform der !kung-Buschleute in Süd-Afrika’, Mitteilungen der
  Anthropologischen Gesellschaft in Wien XCV, 36–41.
——. Ms. Diaries in the archive of the Anthropos Institute. St. Augustin, Germany.
Hagenbucher-Sacripanti, F. 1973. Les Fondements spirituels du pouvoir de Loango, République
  populaire du Congo. Paris: ORSTOM.
Hahn, R.A. 1995. Sickness and Healing: An Anthropological Perspective. New Haven &
  London: Yale University Press.
                                      references                                        223

Hallgren, R. 1988. The Good Things in Life: A Study of the Traditional Religious Culture
  of the Yoruba People. Lund Studies in African and Asian Religions 2. Löberöd: Plus
Hatfield, C.R. Jr. 1968. ‘The Nfumo in Tradition and Change: A Study of the
  Position of Religious Practitioners among the Sukuma of Tanzania, E.A.’, Ph.D.
  thesis, Catholic University of America, Washington DC.
Hauge, H.-E. 1979. Maasai Religion and Folklore. Nairobi: City Printing Works.
——. Turkana Religion and Folklore, ed. by D. Westerlund. Studies published by the
  Institute of Comparative Religion at the University of Stockholm 8. Stockholm:
  Department of Comparative Religion, University of Stockholm.
Heinz, H.-J. 1975. ‘Elements of !ko Bushmen Religious Beliefs’, Anthropos 70 (1/2),
Hendriks, R.P. 1952a. ‘Vie sociale et culturelle des Basukuma’. Manuscript in the
  archive of the White Fathers in Rome.
——. 1952b. ‘Cult of Ancestors’. Manuscript in the archive of the White Fathers
  in Rome.
——. 1962. ‘Spirits of the Ancestors (Basukuma)’. Manuscript in the archive of the
  White Fathers in Rome.
Henschel, J. 1983. Alles begann in Bagamoyo: 100 Jahre Kirche in Ostafrika. Aachen:
  Misson Aktuell.
Hersak, D. 2001. ‘There Are Many Kongo Worlds: Particularities of Magico-
  Religious Beliefs among the Vili and Yombe of Congo-Brazzaville’, Africa 71 (4),
Hirschberg, W. 1975. ‘Khoisan Sprechende Völker Südafrikas’, in H. Baumann
  (ed.), Die Völker Afrikas und ihre traditionellen Kulturen, Part I, 383–405. Studien zur
  Kulturkunde 34. Wiesbaden: Franz Steiner.
Hirth, M. 1894. ‘Diaries, Mission du Bukumbi (Kamoga-Laganeda)’. Manuscript in
  the archive of the White Fathers in Rome.
——. 1899. ‘Diaries, Mission du Bukumbi (Kamoga-Laganeda)’. Manuscript in the
  archive of the White Fathers in Rome.
Ho, K.-J. et al. 1971. ‘The Masai of East Africa: Some Unique Biological Char-
  acteristics’, Archives of Pathology 91 (May), 387–410.
Hohenberger, J. 1958. Semitisches und hamitisches Sprachgut im Masai, mit vergleichendem
  Wörterbuch: Eine sprachvergleichende Untersuchung unter Berücksichtigung von rund 50 semi-
  tischen, hamitischen, nilo-hamitischen und anderen afrikanischen Sprachen. Sachsenmühle,
  Fränkische Schweiz: Selbstverlag des Verfassers.
Hollis, A.C. 1905. The Masai: Their Language and Folklore. Oxford: Clarendon Press.
Holm, E. 1965. Tier und Gott: Mythik, Mantik und Magie der südafrikanischen Urjäger.
  Basel & Stuttgart: Schwabe.
Horton, R. 1967. ‘African Traditional Thought and Western Science’, Africa 37 (1,
  2), 50–71, 155–187.
——. 1971. ‘On Taking the Enemy’s Measure’, in The Traditional Background to
  Medical Practice in Nigeria, 38–45. Occasional Publication 25. Ibadan: Institute of
  African Studies in collaboration with the University College Hospital, University
  of Ibadan.
——. 1982. ‘Tradition and Modernity Revisited’, in M. Hollis & S. Lukes (eds.),
  Rationality and Relativism, 201–260. Oxford: Basil Blackwell.
——. 1984. ‘Judaeo-Christian Spectacles: Boon or Bane to the Study of African
  Religions?’, Cahiers d’Études Africaines 24 (4), 391–436.
Hountondji, P.J. 1980. Sur la ‘philosophie africaine’: Critique de l’ethnophilosophie. Paris:
  François Maspero.
Hurskainen, A. 1984. Cattle and Culture: The Structure of a Pastoral Parakuyo Society.
  Studia Orientalia 56. Helsinki: The Finnish Oriental Society.
224                                     references

——. 1989. ‘The Epidemiological Aspect of Spirit Possession among the Maasai of
   Tanzania’, in A. Jacobson-Widding & D. Westerlund (eds.), Culture, Experience and
   Pluralism: African Ideas of Illness and Healing, 139–150. Uppsala Studies in Cultural
   Anthropology 13. Stockholm: Almqvist & Wiksell International.
Idowu, E.B. 1963. Olódùmarè: God in Yoruba Belief. New York: Frederick A. Praeger.
——. 1973. African Traditional Religion: A Definition. London: SCM Press.
Isola, A. 1977. ‘Yoruba Beliefs about Sango as a Deity’, Orita 11 (2), 105–120.
Itandala, A.B. 1983. ‘A History of the Babinza of Usukuma, Tanzania, to 1890’.
   Ph.D. thesis, Department of History, Dalhousie University.
Jacobs, A.H. 1965. ‘The Traditional Political Organization of the Pastoral Masai’.
   D.Phil. thesis, Nuffield College, Oxford University.
Jacobson-Widding, A. 1979. Red —White —Black as a Mode of Thought: A Study of
   Triadic Classification by Colours in the Ritual Symbolism and Cognitive Thought of the Peoples
   of the Lower Congo. Uppsala Studies in Cultural Anthropology 1. Stockholm: Almqvist
   & Wiksell International.
——. 1983. ‘Hälsa och sjukdom i Centralafrika: Aspekter på social och psykologisk
   diagnostik’, Socialmedicinsk tidskrift 60 (7), 380–394.
Janzen, J.M. 1972. ‘Laman’s Kongo Ethnography: Observations on Sources,
   Methodology, and Theory’, Africa 42 (4), 316–328.
——. 1978. The Quest for Therapy: Medical Pluralism in Lower Zaire. Comparative Studies
   in Health Systems and Medical Care 1. Berkeley, Los Angeles & London: University
   of California Press.
——. 1981. ‘The Need for a Taxonomy of Health in the Study of African Thera-
   peutics’, Social Science and Medicine 15B (3), 185–194.
Janzen, J.M., & S. Feierman. 1979. ‘Introduction’, Social Science and Medicine 13B
   (4), 239–243.
Janzen, J.M., & G. Prins. 1981. ‘Introduction’, Social Science and Medicine 15B (3),
Johnson, J. 1899. Yoruba Heathenism. Exeter & London: James Townsend & Son.
Johnson, S. 1921. The History of the Yorubas: From the Earliest Times to the Beginning of
   the British Protectorate, ed. by O. Johnson. London: Routledge and Kegan Paul.
Johnston, H.H. 1886. The Kilima-Njaro Expedition: A Record of Scientific Exploration in
   Eastern Equatorial Africa. London: Kegan Paul, Trench and Co.
——. 1915. ‘Masai’, Encyclopaedia of Religion and Ethics, ed. J. Hastings, Vol. 8.
   Edinburgh: T. & T. Clark; New York: Charles Scribner’s Sons.
Juma, W. 1960. ‘The Sukuma Societies for Young Men and Women’, Tanzania
   Notes and Records (54).
Katz, R. 1976. ‘Education for Transcendence: !Kia-Healing with the Kalahari
   !Kung’, in R.B. Lee & I. De Vore (eds.), Kalahari Hunter-Gatherers: Studies of the
   !Kung San and Their Neighbours, 281–301. Cambridge, MA, & London: Harvard
   University Press.
——. 1982. Boiling Energy: Community Healing among the Kalahari Kung. Cambridge, MA,
   & London: Harvard University Press.
Kipury, N. 1983. Oral Literature of the Maasai. Nairobi: Heinemann Educational Books.
Kimerei, G. 1973. ‘Reconciliation among the Maasae in Comparison with Recon-
   ciliation in the Old Testament’, STM thesis, Dubuque: Wartburg Theological
Kleinman, A. 1980. Patients and Healers in the Context of Culture: An Exploration of the
   Borderland between Anthropology, Medicine, and Psychiatry. Berkeley, Los Angeles &
   London: University of California Press.
Koenig, O. 1956. The Masai Story. London: Michael Joseph.
Köhler, O. 1971. ‘Die “Krankheit” im Denken der Kxoe-Buschmänner’, in V. Six
   et al. (eds.), Afrikanische Sprachen und Kulturen, 317–325. Hamburger Beiträge zur
   Afrika-Kunde 40. Hamburg: Deutsches Institut für Afrika-Forschung.
                                      references                                       225

——. 1978. ‘Tierzauber und Krankheit bei den Kxoe-Buschmännern’, Afrika und
  Überzee LXI (1), 35–58.
——. 1978/79. ‘Mythus, Glaube und Magie bei den Kxoe-Buschmännern’, Journal
  der Südwestafrikanische Wissenschaftliche Gesellschaft 33, 9–49.
Kopytoff, I. 1971. ‘Ancestors as Elders in Africa’, Africa 41 (2), 129–142.
Krapf, L. 1857. ‘Kurze Beschreibung der Masai- und Wakuafi-Stämme im südöstlichen
  Afrika’, Das Ausland: Eine Wochenschrift für Kunde des geistigen und sittlichen Lebens der
  Völker 30 (19), 437–466.
Kronenberg, A. 1979. ‘Die Niloten’, in H. Baumann (ed.), Die Völker Afrikas und ihre
  traditionellen Kulturen, Part II, Ost-, West- und Nordafrika, 157–189. Wiesbaden: Franz
Lacombe, B. 1986. ‘Diseases and Care in Congo in 1983: Image Given by a Survey’,
  in U. Maclean & C. Fyfe (eds.), African Medicine in the Modern World, 189–197.
  Seminar Proceedings 27. Edinburgh: Centre of African Studies, University of
Lagercrantz, S. 1983/84. ‘Karl Laman och hans minkisi’, Religion och Bibel 42/43,
Laman, K.E. 1916. ‘Förfädernas kult hos bakongo’, Svensk Missionstidskrift 4 (5),
——. 1917. ‘Gudstron hos bakongo’, Svensk Missionstidskrift 5 (6), 271–294.
——. 1928. ‘Andetro och Nkisi-kult hos bakongo’, Svensk Missionstidskrift 16 (1, 2 & 5),
  17–30, 75–81, 237–244.
——. 1962. The Kongo III. Studia Ethnographica Upsaliensia 12. Uppsala.
Landei, I.A. Ole. 1980. ‘Christian Nurturing and Maasai Nurturing’, Dissertation
  for Advanced Diploma in Religious Education, Westhill College, Birmingham.
——. 1982. ‘Moral Customs and Moral Education among the Maasai of Tanzania
  and Kenya’, MA dissertation, University of Hull.
——. n.d. Ms. ‘Maasai Culture and Religion’, unpublished book manuscript.
Larsen, Stephan. 1983. A Writer and his Gods: A Study of the Importance of Yoruba Myths
  and Religious Ideas to the Writing of Wole Soyinka. Stockholm: University of Stockholm,
  Department of the History of Literature.
Last, M. 1981. ‘The Importance of Knowing about not Knowing’, Social Science and
  Medicine 15B (3), 387–392.
——. 1986. ‘Introduction: The Professionalisation of African Medicine; Ambiguities
  and Definitions’, in M. Last & G.L. Chavunduka (eds.), The Professionalisation of
  African Medicine, 1–19. Manchester: Manchester University Press in association
  with the International African Institute.
Laube, R. 1986. Maasai-Identität und sozialer Wandel bei den Maasai. Social Strategies,
  Monographs on Sociology and Social Policy 20. Basel: Social Strategies Publishers
  Co-operative Society.
Lawson, E.T. 1984. Religions of Africa: Traditions in Transformation. The Religious
  Traditions of the World Series. San Francisco: Harper & Row.
Leakey, L.S.B. 1930. ‘Some Notes on the Masai of Kenya Colony’, Journal of the
  Royal Anthropological Institute 60, 185–209.
Lebzelter, V. 1928. ‘Die religiösen Vorstellungen der //Khun-Buschmänner, der
  Buschmänner der Etoshapfanne und des Ovambo-Landes und der Ovambo-
  Bantu’, in W. Koppers (ed.), Festschrift, Publication d’Hommage offerte au P.W. Schmidt.
——. 1934. Eingeborenenkulturen in Südwest- und Südafrika: Wissenschaftliche Ergebnisse einer
  Forschungsreise nach Süd- und Südwestafrika in den Jahren 1926–1928, II. Leipzig.
Lee, A.A. 1976. ‘Ngoja and Six Theories of Witchcraft Eradication’, Ufahamu 6 (3),
Lee, R.B. 1967. ‘Trance Cure of the !Kung Bushmen’, Natural History (Nov.), 31–37.
226                                       references

——. 1968. ‘The Sociology of !Kung Bushman Trance Performances’, in R. Prince
  (ed.), Trance and Possession States, 35–54. Montreal: R.M. Bucke Memorial Society.
——. 1984. The Dobe !Kung. Case Studies in Cultural Anthropology. New York:
  CBS College Publishing.
Leighton, A.H., et al. 1963. Psychiatric Disorder among the Yoruba: A Report from the
  Cornell-Aro Mental Health Research Project in the Western Region, Nigeria. Ithaca, NY:
  Cornell University Press.
Le Roy, A. 1892. ‘Les Massais (Afrique orientale)’, Annales Apostoliques 7 (28), 135–147.
——. 1928. Au Kilima-ndjaro: Histoire de la fondation d’une mission catholique en Afrique
  Orientale. Paris: L’Oeuvre d’Auteuil.
Le Scao, J.-L. 1908. ‘Au Loango (Afrique Équatoriale Française)’, Les Missions
  Catholiques (2039), 331–335.
——. 1915. ‘Au Royaume de Satan: Les mangeurs de vie’, Les Missions Catholiques
  (2413), 429–431.
Lewis-Williams, J.D. 1981. Believing and Seeing: Symbolic Meanings in Southern San Rock
  Paintings. London: Academic Press.
Livinhac, L. 1879. ‘Idées religieuses’, Chronique des Pères Blancs (4), 171–173.
——. 1888. ‘Diaries, Mission du Bukumbi (Kamoga-Laganeda)’. Manuscript in the
  archive of the White Fathers in Rome.
Lloyd, L.C. (ed.). 1911. Specimens of Bushman Folklore (collected by the late W.H.I. Bleek
  and L.C. Lloyd). London: George Allen.
Lloyd, P.C. 1960. ‘Sacred Kingship and Government among the Yoruba’, Africa 30
  (3), 222–236.
Lowery-Palmer, A.L. 1980. ‘Yoruba World View and Patient Compliance’, Ph.D.
  Dissertation, Department of Anthropology, University of California, Riverside.
Lucas, A.O., & R.G. Hendrickse. 1971. ‘Yoruba Ideas of Disease Revealed by
  Hospital Patients’. Paper presented at a special seminar on the Traditional
  Background to Medical Practice in Nigeria (April 20–23, 1966). Occasional
  Publication 25. University of Ibadan, Institute of African Studies in collaboration
  with the University College Hospital.
Lucas, J.O. 1948. The Religion of the Yoruba: Being an Account of the Religious Beliefs and
  Practices of the Yoruba Peoples of Southern Nigeria, Especially in Relation to the Religion of
  Ancient Egypt. Lagos: CMS Bookshop.
Lübbert, A. 1901. ‘Über die Heilmethoden und Heilmittel der Eingeborenen in
  Deutsch-Südwestafrika’, Mitteilungen aus den deutschen Schutzgebieten XIV, 78–90.
MacGaffey, W. 1980. ‘African Religions: Types and Generalizations’, in I. Karp &
  C. Bird (eds.), Explorations in African Systems of Thought, 301–328. Bloomington:
  Indiana University Press.
——. 1981. ‘African Ideology and Belief: A Survey’, African Studies Review 24 (2/3),
——. 1983. Modern Kongo Prophets: Religion in a Plural Society. Bloomington: Indiana
  University Press.
——. 1986. Religion and Society in Central Africa: The Bakongo of Lower Zaire. Chicago
  & London: University of Chicago Press.
——. 1990. Art and Healing of the Bakongo Commented by Themselves: Minkisi from the
  Laman Collection. Kikongo texts translated and edited by W. MacGaffey. Stockholm: Almqvist
  & Wiksell International.
——. 1997. Preface to the English Language Edition, in P. Geschiere, The Modernity
  of Witchcraft: Politics and the Occult in Postcolonial Africa (Sorcellerie et politique en Afrique—
  la viande des autres), vii–ix. Translated by P. Geschiere & J. Roitman. Charlottesville
  & London: University Press of Virginia.
——. 2000. Kongo Political Culture: The Conceptual Challenge of the Particular. Bloomington
  & Indianapolis: Indiana University Press.
                                      references                                      227

McKay, D.H. 1950. ‘Venereal Disease in Masai: A Field Survey—June and July
 1950’, East African Medical Journal 27 (Nov.), 451–457.
McKenzie, P.R. 1976. ‘Yoruba Òrìsà Cults: Some Marginal Notes Concerning Their
 Cosmology and Concepts of Deity’, Journal of Religion in Africa 13 (3), 189–207.
——. McKenzie, P.R. 1997. Hail Orisha! A Phenomenology of a West African Religion in
 the Mid-Nineteenth Century. Leiden: Brill.
——. 2002. ‘Hail Orisha! Defended: A Response to James Cox’, Journal of Religion
 in Africa 32 (1), 110–119.
Maclean, U. 1966. ‘Hospitals or Healers? An Attitude Survey in Ibadan, Human
 Organization 25 (2), 131–139.
——. 1971. Magical Medicine: A Nigerian Case-Study. London: Allen Lane, Penguin
——. 1976. ‘Some Aspects of Sickness Behaviour among the Yoruba’, in J.B. Loudon
 (ed.), Social Anthropology and Medicine, 285–217. ASA Monograph 13. London, New
 York & San Francisco: Academic Press.
Magesa, L. 1997. African Religion: The Moral Traditions of Abundant Life. Maryknoll,
 NY: Orbis Books.
Mahaniah, K. 1973. ‘Purification and Millenarianism in Zaire: A Case Study in
 the Socio-Medical and Magico-Religious Practice of the Kongo of Zaire’, Cahiers
 des Religions Africaines 14 (7), 227–244.
——. 1979. ‘The Therapeutic Functions of Authority among the Kongo of Zaire’,
 Cahiers des Religions Africaines 13 (26), 209–229.
——. 1980. ‘La Structure multidimensionelle de guérison à Kinshasa, capitale du
 Zaire’, Zaire —Afrique 20 (150), 591–607.
——. 1982. La Maladie et la guérison en milieu Kongo: Essai sur kimfumu, kingunza et kitobe.
 Kinshasa: Edicva.
Maingard, J.F. 1937. ‘Some Notes on Health and Disease among the Bushmen of
 the Southern Kalahari’, Bantu Studies XI, 285–295.
Malcolm, D.W. 1953. Sukumaland: An African People and their Country: A Study of Land
 Use in Tanganyika. London, New York & Toronto: Oxford University Press.
Marari, L.P. Ole 1980. ‘The Relationship between the Maasai and the Biblical
 Concept of Creation’, B.Div. dissertation, Lutheran Theological College, Makumira.
Margetts, E.L. 1965. ‘Traditional Yoruba Healers in Nigeria’, Man 65 ( July-August),
Marichelle, C. 1910. ‘Tablettes d’un Congolais’, Les Missions Catholiques (2137),
——. 1898. ‘Une École rurale au Congo Français’, Les Missions Catholiques (1591),
Marshall, L. 1962. ‘!Kung Bushman Religious Beliefs’, Africa 32 (3), 221–252.
——. 1969. ‘The Medicine Dance of the !Kung Bushmen’, Africa 39 (4), 347–381.
——. 1976. The !Kung of Nyae Nyae. Cambridge, MA, & London: Harvard University
——. 1986. ‘Some Bushman Star Lore’, in R. Vossen & K. Keuthmann (eds.),
 Contemporary Studies on Khoisan: In Honour of Oswin Köhler on the Occasion of his 75th
 Birthday, 169–173. Quellen zur Khoisan-Forschung 5:2. Hamburg: Helmut Buske
Massek, A.O., & J.O. Sidai. 1974. Wisdom of Maasai. Nairobi: Transafrica.
Merker, M. 1903. ‘Religion und Tradition der Masai’, Zeitschrift für Ethnologie 35 (5),
——. 1904. Die Masai: Ethnographische Monographie eines ostafrikanischen Semitenvolkes.
 Berlin: Dietrich Reimer.
Meyer, B. 1998. ‘The Power of Money: Politics, Occult Forces, and Pentecostalism
 in Ghana’, African Studies Review 41 (3), 15–38.
228                                  references

Middleton, J., & E.H. Winter 1963. ‘Introduction’, in J. Middleton & E.H. Winter
  (eds.), Witchcraft and Sorcery in East Africa, 1–26. London: Routledge & Kegan Paul.
Millroth, B. 1965. Lyuba: Traditional Religion of the Sukuma. Studia Ethnographica
  Upsaliensia 22. Uppsala.
Mogg, J.W. n.d. Bushmen in the Kalahari Desert. SPG World Wide Series 3. London.
Mol, F. 1978. A Dictionary of the Maasai Language and Folklore, English-Maasai. Nairobi:
  Marketing and Publishing.
Morris, B. 1987. Anthropological Studies of Religion: An Introductory Text. Cambridge:
  Cambridge University Press.
Morton-Williams, P. 1960. ‘Yoruba Responses to the Fear of Death’, Africa 30 (1),
—— 1964. ‘An Outline of the Cosmology and Cult Organization of the Oyo
  Yoruba’, Africa 34 (3), 243–261.
Multhaupt, T. 1987. ‘Sozialanthropologische Theorien über Hexerei und Zauberei
  in Afrika’, Anthropos 82 (4/6), 445–456.
—— 1990. Hexerei und Antihexerei in Afrika. München: Trickster.
Munson, R. 1981. ‘Why Medicine Cannot Be a Science’, Medicine and Philosophy 6
  (2), 183–208.
Nadel, S.F. 1952. ‘Witchcraft in Four African Societies: An Essay in Comparison’,
  American Anthropologist 54 (1), 18–29.
Nathalie, P. 1884. ‘Diairies, Mission du Bukumbi (Kamoga-Laganeda)’. Manuscript
  in the archive of the White Fathers in Rome.
Ndoponoi, M. Ole 1986. ‘Initiation among the Maasai and the Christian Faith in
  an Encounter’, B.Div. dissertation, Lutheran Theological College, Makumira.
Ng’weshemi, A.M. 1990. ‘Mulungu: God in Sukuma Belief and Its Implications for
  the Church’, B.Div. dissertation, Lutheran Theological College, Makumira.
Nolte, K. 1886. ‘Krankheiten und Heilmittel der Nama und Buschmänner’, Deutsche
  Kolonialzeitung 3, 629–632.
Nüssler, F. 1931. ‘Krankendienst an den Masai’, Evangelisch-Lutherisches Missionsblatt
  86 (6), 172.
Ogunsakin-Fabarebo, S.I. 1998. ‘Contemporary Theories on Magic: Magun’s Disparate
  Characteristics’, Orita 30 (1–2), 9–20.
Ojo, G.J.A. 1966. Yoruba Culture: A Geographical Analysis. London: University of Ife
  & University of London Press.
Oladapo, T.M. 1984. ‘Traditional Healing among the Ife: A Yoruba Group in
  Nigeria’, Ph.D. manuscript submitted to the Department of Religious Studies,
  University of Ife, Ile-Ife.
Olsson, T. 1975. ‘Religious Documents of the Maasai’, Part I: ‘Prayers’, unpub-
  lished book manuscript.
——. 1982a. ‘Religious Documents of the Maasai’, Part II: ‘Myth’, unpublished
  book manuscript.
——. 1982b. ‘Religious Documents of the Maasai’, Part III: ‘The Oral Expression
  of Religion’, unpublished book manuscript.
——. 1989. ‘Philosophy of Medicine among the Maasai’, in A. Jacobson-Widding
  & D. Westerlund (eds.), Culture, Experience and Pluralism: African Ideas of Illness and
  Healing, 235–246. Uppsala Studies in Cultural Anthropology 13. Stockholm:
  Almqvist & Wiksell International.
Olukoju, E.O. 1997. ‘Esu in the Belief of the Yoruba’, Orita 29 (1–2), 1–14.
Olukunle, O. 1979. ‘Witchcraft: A Study in Yoruba Belief and Metaphysics’, Ph.D.
  thesis, Department of Religious Studies, University of Ibadan.
Olupona, Jacob K. 1991. Kingship, Religion, and Rituals in a Nigerian Community: A
  Phenomenological Study of Ondo Yoruba Festivals. Stockholm Studies in Comparative
  Religion 28. Stockholm: Almqvist & Wiksell International.
                                      references                                        229

Orr, J.B., & J.L. Gilks. 1931. Studies of Nutrition: The Physique and Health of Two African
  Tribes. Medical Research Council: Special Report Series, no. 155. London: His
  Majesty’s Stationery Office.
Osunwole, S.A. 1986. ‘Food Taboos in Nigerian Culture with Special Reference to
  the Yoruba’. Paper presented at seminar on Nigerian Food and Culture, Institute
  of African Studies, University of Ibadan, 26–28 November 1986.
——. 1989. ‘Oral Tradition in Traditional Medicine with Particular Reference to
  the Yoruba’. Paper presented at Second Traditional Medicine Fair, National Arts
  Theatre, Iganmu, Lagos, 9–16 April 1989.
——. n.d. ‘Witchcraft and Sorcery: A Study in Yoruba Beliefs and Medicine’.
  Manuscript in the Department of Religious Studies, University of Ibadan.
Oyebola, D.D.O. 1981. ‘Professional Associations, Ethics and Discipline among
  Yoruba Traditional Healers of Nigeria’, Social Science and Medicine 15 B (2), 87–92.
Pambou, P.-A. 1979–1980. ‘Aspects familiaux et magico-religieux fondamentaux des
  maladies et therapies traditionelles en pays Loango (Congo)’, Dissertation, Licence
  de Sciences Sociales, Académie de Lyon, Université de Saint-Etienne.
——. 1980–1981. ‘Dimension politico-religieuse et therapeutique du phenomène
  social du nkisi et du nkisi si chez les Loango (Congo—Cabinda—Bas Zaïre)’,
  Dissertation, Maîtrise de Sciences Sociales, Académie de Lyon, Université de
Parratt, J.K. 1974. ‘Olorun and the High God Pattern’, The Ghana Bulletin of Theology
  3 (8), 1–10.
Parrinder, G. 1974. African Traditional Religion. 3rd edn. London: Sheldon Press.
Passarge, S. 1907. Die Buschmänner der Kalahari. Berlin: Dietrich Reimer (Ernst Vohsen).
Pedersen, G.D. 1977. ‘Traditional Sukuma Curers: The Bafumu’, Tanzania Notes and
  Records (81/82), 62–65.
Pemberton, J., & F.S. Afolayan 1996. Yoruba Sacred Kingship: ‘A Power Like That of
  the Gods’. Washington & London: Smithsonian Institution Press.
Perlitz, M. 1973. Mungu ist auch in der Steppe: Christen im ostafrikanischen Grasland.
  Erlanger Hefte aus der Weltmission. Erlangen: Ev.-Luth. Mission.
Peron, X. 1995. L’Occidentalisation des Maasaï du Kenya: Privatisation foncière et destruc-
  turation sociale chez les Maasaï du Kenya, II. Paris: Éditions L’Harmattan.
Peterson, D.A. 1985. ‘Spirit Possession among the Maasai in Tanzania’, Africa
  Theological Journal 14 (3), 174–178.
Platvoet, J. 1996. ‘From Object to Subject: A History’, in J. Platvoet, J. Cox &
  J. Olupona (eds.), The Study of Religions in Africa: Past, Present and Prospects, 105–138.
  Religions of Africa 1. Cambridge: Roots and Branches.
——. 2002. Review of Mathias Guenther, Tricksters and Trancers: Bushman Religion
  and Society, Journal of Religion in Africa 32 (4), 502–506.
Pool, R. 1994. ‘On the Creation and Dissolution of Ethnomedical Systems in the
  Medical Ethnography of Africa’, Africa 64 (1), 1–20.
Pouwels, R.L. 1987. Horn and Crescent: Cultural Change and Traditional Islam on the East
  African Coast, 800–1900. African Studies Series 53. Cambridge: Cambridge University
Prince, R. 1964. ‘Indigenous Yoruba Psychiatry’, in A. Kiev (ed.), Magic, Faith, and
  Healing: Studies in Primitive Psychiatry Today, 84–120. London: Free Press of Glencoe.
——. 1979. ‘Symbols and Psychotherapy: The Example of Yoruba Sacrificial Ritual’,
  in Z.A. Ademuwagun et al. (eds.), African Therapeutic Systems, 114–119. Waltham,
  MA: Crossroads Press.
Proyart, L.-B. 1819 [1776]. Histoire de Loango, Kakongo, et autres royaumes d’Afrique. Paris.
Rapports 1910–1911. ‘Rapports Annuels des Pères Blancs’. Archive of the White
  Fathers in Rome.
230                                     references

Ranger, T. 1987. ‘Concluding Summary: Religion, Development and Identity’, in
   K.H. Petersen (ed.), Religion, Development and African Identity, 147–163. Seminar
   Proceedings 17. Uppsala: Scandinavian Institute of African Studies.
Reid, M.B. 1969. ‘Persistence and Change in the Health Concepts and Practices
   of the Sukuma of Tanzania, East Africa’, Ph.D. thesis, Catholic University of
   America, Washington DC.
Reusch, R. 1946. ‘Der Missionsdienst unter den Masai sprechenden Stämmen’.
   Unpublished paper in the archive of the Leipzig Mission, Leipzig, Germany.
——. 1953/54. ‘“Eure Wege sind nicht meine Wege”: Aus der Masai-Mission in
   Ostafrika’, Lutherisches Missionsjahrbuch 1953/54, 87–95.
Rinsum, H.J. van. 2001. Slaves of Definition: In Quest of the Unbeliever and the Ignoramus.
   Maastricht: Shaker Publications.
Rivers, W.H.R. 1924. Medicine, Magic and Religion. London: Kegan Paul.
Romanucci-Ross, L., D.E. Moerman & L.R. Tancredi. 1983. Preface: ‘The Cultural
   Context of Medicine and the Biohuman Paradigm’, in L. Romanucci-Ross, D.E.
   Moerman & L.R. Tancredi et al., The Anthropology of Medicine: From Culture to Method,
   vii–xiii. South Hadley, MA: Bergin & Garvey Publishers.
Sadin, F. 1910. ‘Essai d’étude sur la religion des Bakongo’, Missions Belges de la
   Compagnie de Jésus, 131–140.
Saibull, S. Ole, & R. Carr. 1981. Herd and Spear: The Maasai of East Africa. London:
   Collins and Harvill Press.
Salvadori, C., & A. Fedders. 1973. Maasai. London: Collins.
Sankan, S.S. Ole. 1971. The Maasai. Nairobi, Dar es Salaam & Kampala: East
   African Literature Bureau.
Schans, L. van de. 1955. ‘Les Basukuma: Organisation de la tribu, du pays, lois et
   coutumes’. Manuscript in the archive of the White Fathers in Rome.
Schapera, I. 1930. The Khoisan Peoples of South Africa: Bushmen and Hottentots. London:
   Routledge & Kegan Paul.
Schatzberg, M.G. 2000. ‘La Sorcellerie comme mode de causalité politique’, Politique
   africaine (79), 33–47.
Schebesta, P.P. 1923/24. ‘Die religiösen Anschauungen Südafrikas’, Anthropos 18–19
   (1/3), 114–124.
Schmidt, S. 1973. ‘Die Mantis religiosa in den Glaubensvorstellungen der Khoesan-
   Völker’, Zeitschrift für Ethnologie 98 (1), 102–123.
Schmidt, W. 1929. ‘Zur Erforschung der alten Buschmann-Religion’, Africa 11 (3),
——. 1933. ‘Die Religionen der Völker Afrikas’, Der Ursprung der Gottesidee: Eine his-
   torisch-kritische und positive Studie, Vol. IV. Münster: Aschendorff.
——. 1940. ‘Die afrikanischen Hirtenvölker: Hamiten und Hamitoiden’, Der Ursprung
   der Gottesidee: Eine historisch-kritische und positive Studie, Vol. VII. Münster: Aschendorff.
Schönenberger, P. 1961. ‘Names for “God” Known and Used by the Wanyamwezi’,
   Anthropos 56, 947–949.
Schulz-Weidner, W. 1979. ‘Die Ostatlantische Provinz’, in H. Baumann (ed.), Die
   Völker Afrikas und ihre traditionellen Kulturen, Part II, Ost-, West- und Nordafrika.
   Wiesbaden: Franz Steiner.
Seyffert, C. 1913. ‘Totengebräuche und Todesvorstellungen bei den zentralafrikanischen
   Pygmäen, den Buschmännern und Hottentotten’, Archiv für Anthropologie, n.s., 12,
Shostak, M. 1981. Nisa: The Life and Words of a !Kung Woman. Cambridge, MA:
   Harvard University Press.
Sicard, G.M. 1999. A Tale of a Maasai Girl. Lewes: Book Guild.
Silberbauer, G.B. 1972. ‘The G/wi Bushmen’, in M.G. Bicchieri (ed.), Hunters and
   Gatherers Today: A Socio-Economic Study of Eleven Such Cultures in the Twentieth Century,
   271–326. New York: Holt, Rinehart & Winston.
                                       references                                        231

Simpson, G.E. 1980. Yoruba Religion and Medicine in Ibadan. Ibadan: Ibadan University
Sindiga, I. 1995. ‘Maasai Traditional Medicine’, in I. Sindiga, C. Nyaigotti-Chacha
  & M.P. Kanunah (eds.), Traditional Medicine in Africa, 94–107. Nairobi: East African
  Educational Publishers.
Smith, A., et al. 2000. The Bushmen of Southern Africa: A Foraging Society in Transition.
  Cape Town: David Philip & Athens, Ohio: Ohio University Press.
Söderberg, B., & R. Widman 1973. ‘The “Rediscovered” Material of Laman’s
  Kongo Ethnography’, Africa 43 (4), 351.
Spencer, P. 1988. The Maasai of Matapato: A Study of Rituals of Rebellion. Bloomington
  & Indianapolis: Indiana University Press in association with the International
  African Institute.
——. 1991. ‘The Loonkidongi Prophets and the Maasai: Protection Racket or
  Incipient State?’, Africa 61 (3), 334–342.
Stadler, J. 1996. ‘Witches and Witch-Hunters: Witchcraft, Generational Relations
  and the Life Cycle in a Lowveld Village’, African Studies 55 (1), 87–110.
Steeves, G.W. 1990. ‘Contextualizing the Atonement for the Sukuma Tribe of
  Tanzania’. Abstract of MA thesis, Graduate School, Abilene Christian University.
Stow, G.W. 1910. The Native Races of South Africa: A History of the Intrusion of the
  Hottentots and Bantu into the Hunting Grounds of the Bushmen, the Aborigines of the Country.
  London: Swan Sonnenschein & New York: Macmillan.
Struyf, I. 1933. ‘Le “Kindoki”’, Revue Missionaire des Jesuites Belges 7 (Nov.), 400–402.
Sullivan, S. 2001. ‘Difference, Identity, and Access to Official Discourses: Hai//om,
  “Bushmen,” and a Recent Namibian Ethnography’, Anthropos 96, 179–192.
Swantz, M.-L. 1989. ‘Manipulation of Multiple Health Systems in the Costal Regions
  of Tanzania’, in A. Jacobson-Widding & D. Westerlund (eds), Culture, Experience
  and Pluralism: African Ideas of Illness and Healing, 277–287. Uppsala Studies in Cultural
  Anthropology 13. Stockholm: Almqvist & Wiksell International.
Table n.d.a. ‘Table d’enquêtes sur les moeurs et coutumes indigènes: Vicariats de
  “Mwanza” et de “Maswa”; Tribu des Basukuma’. Manuscript in the archive of
  the White Fathers in Rome.
Table n.d.b. ‘Table d’enquêtes sur les moeurs et coutumes indigènes: Les Yorubas;
  Prefecture Apostolique d’Oyo’. Manuscript in the archive of the White Fathers
  in Rome.
Tanaka, J. 1980. San, Hunter-Gatherers of the Kalahari: A Study in Ecological Anthropology.
  Tokyo: University of Tokyo.
Tanner, R.E.S. 1955. ‘Hysteria in Sukuma Medical Practice’, Africa 25 (3), 274–279.
——. 1956a. ‘An Introduction to the Northern Basukuma’s Idea of the Supreme
  Being’, Anthropological Quarterly 29 (2), 45–56.
——. 1956b. ‘The Sorcerer in Northern Sukumaland, Tanganyika’, Southwestern
  Journal of Anthropology 12, 437–443.
——. 1957. ‘Sukuma Leachcraft: An Analysis of their Medical and Surgical System’,
  East African Medical Journal 36 (2), 119–122.
——. 1958. ‘Ancestor Propitiation Ceremonies in Sukumaland, Tanganyika’, Africa
  28 (3), 225–231.
——. 1959a. ‘The Spirits of the Dead: An Introduction to the Ancestor Worship
  of the Sukuma of Tanganyika’, Anthropological Quarterly 32 (2), 108–124.
——. 1967. Transition in African Beliefs: Traditional Religion and Christian Change; a Study
  in Sukumaland, Tanzania, East Africa. Maryknoll, NY: Maryknoll Publications.
——. 1969. ‘The Theory and Practice of Sukuma Spirit Mediumship’, in J. Beattie
  & J. Middleton (eds.), Spirit Mediumship and Society in Africa, 273–288. London:
  Routledge & Kegan Paul.
——. 1970. The Witch Murders in Sukumaland: A Sociological Commentary. Crime in East
  Africa 4. Uppsala: Scandinavian Institute of African Studies.
232                                   references

Tanner, R., & C. Mitchell. 2001. Religion and the Environment. Basingstoke: Palgrave
Tcherkézoff, S. 1985. ‘The Expulsion of Illness or the Domestication of the Dead:
  A Case Study of the Nyamwezi of Tanzania’, History and Anthropology 2 (1), 59–92.
Thiel, J.F. 1974. ‘Sind die nkisi der Völker des Kwango-Kasai-Zwischengebietes
  Ahnengeister?’, Zeitschrift der Deutschen Morgenländischen Gesellschaft, suppl. 2, 637–643.
Thomas, E.M. 1958. The Harmless People. New York: Vintage Books.
Thomson, J. 1885. Through Masailand: A Journey of Exploration among the Snowclad Volcanic
  Mountains and Strange Tribes of Eastern Equatorial Africa. New and revised ed. London:
  Sampson Low, Marston, Searle and Rivington.
Thurner, I. 1983. Die transzendenten und mythischen Wesen der San (Buschmänner): Eine
  Religionsethnologische Analyse historischer Quellen. Acta Ethnologica et Linguistica 56.
  Series Africana 17. Vienna: Föhrenau.
Turner, V.W. 1968. The Drums of Affliction: A Study of Religious Processes among the
  Ndembu of Zambia. Oxford: Clarendon Press.
Vansina, J. 1975. ‘Südkongo’, in H. Baumann (ed.), Die Völker Afrikas und ihre tradi-
  tionellen Kulturen, Vol.1, Allgemeiner Teil und südliches Afrika. Wiesbaden: Franz Steiner.
Vedder, H. 1912. ‘Die Buschmänner’, Allgemeine Missionszeitschrift 39, 403–416.
——. 1937. ‘Die Buschmänner Südwestafrikas und ihre Weltanschaung’, South African
  Journal of Science 34, 416–436.
Verger, P. 1963. ‘Trance States in Orisha Worship’, Odu: A Journal of Yoruba Edo
  and Related Studies (9), 13–20.
——. 1966. ‘The Yoruba High God: A Review of the Sources’, Odu: University of
  Ife Journal of African Studies 2 (2), 19–40.
——. 1969. ‘Trance and Convention in Nago-Yoruba Spirit Mediumship’, in
  J. Beattie & J. Middleton (eds.), Spirit Mediumship and Society in Africa, 50–65.
  London: Routledge & Kegan Paul.
——. 1971. ‘Tranquilizers and Stimulants in Yoruba Herbal Treatment’. Paper
  presented at a special seminar on the Traditional Background to Medical Practice
  in Nigeria (April 20–23, 1966). Occasional Publication 25. University of Ibadan,
  Institute of African Studies in collaboration with the University College Hospital.
Vinnicombe, P. 1976. People of the Eland: Rock Paintings of the Drakensberg Bushmen as
  a Reflection of their Life and Thought. Pietermaritzburg: University of Natal Press.
Voshaar, J. 1979. ‘Tracing God’s Walking Stick in Maa (a Study of Maasai Society,
  Culture and Religion, a Missionary’s Approach)’, Ph.D. thesis, Catholic University,
——. 1998. Maasai: Between the Oreteti-Tree and the Tree of the Cross. The Church and
  Theology in Context Series 35. Amsterdam & New York: Rodopi.
Waller, R.D. 1995. ‘Kidingoi’s Kin: Prophecy and Power in Maasailand’, in D.M.
  Anderson & D.H. Johnson (eds.), Revealing Prophets: Prophecy in Eastern African History,
  28–64. London: James Currey.
Wanitzek, U. 1986. Kindschaftsrecht in Tansania unter besonderer Berücksichtigung des Rechts
  der Sukuma. Hohenschäftlarn bei München: Klaus Renner Verlag.
Weishaupt, M. 1930. ‘Das Volk der Masai’, Evangelisch-Lutherisches Missionsblatt 85
  (8), 224–225.
Welch, E.A. 1974. ‘Life and Literature of the Sukuma in Tanzania, East Africa’,
  Ph.D. thesis, Howard University.
Werner, H. 1906. ‘Anthropologische, ethnologische und ethnografische Beobachtungen
  über die Heikum- und Kungbuschleute nebst einem Anhang über die Sprachen
  dieser Buschmannstämme’, Zeitschrift für Ethnologie 38 (3), 241–268.
Westerlund, D. 1985. African Religion in African Scholarship: A Preliminary Study of the
  Religious and Political Background. Studies published by the Institute of Comparative
  Religion at the University of Stockholm 7. Stockholm: Almqvist & Wiksell
——. 1986a. Religionshistoriskt och etnografiskt material i europeiska missionsarkiv. Working
                                      references                                       233

  Papers in African Studies 17. Uppsala: African Studies Programme, Department
  of Cultural Anthropology, University of Uppsala.
——. 1986b. ‘Introduction’, in H.-E. Hauge, Turkana Religion and Folklore, 13–21.
  Studies published by the Institute of Comparative Religion at the University of
  Stockholm 8. Stockholm: Department of Comparative Religion.
——. 1989. ‘Pluralism and Change: A Comparative and Historical Approach to
  African Disease Etiologies’, in A. Jacobson-Widding & D. Westerlund (eds.), Culture,
  Experience and Pluralism: African Ideas of Illness and Healing, 177–218. Uppsala Studies
  in Cultural Anthropology 13. Stockholm: Almqvist & Wiksell International.
——. 1991. ‘“Insiders” and “Outsiders” in the Study of African Religions: Notes
  on Some Problems of Theory and Method’, in J.K. Olupona (ed.), African Traditional
  Religions in Contemporary Society, 15–24. New York: Paragon.
——. 1993. ‘The Study of African Religions in Retrospect: From “Westernization”
  to “Africanization”’, in J.K. Olupona & S.S. Nyang (eds.), Religious Plurality in
  Africa: Essays in Honour of John S. Mbiti, 43–66. Berlin & New York: Mouton de
Whyte, S.R. 1989. ‘Anthropological Approaches to African Misfortune: From Religion
  to Medicine’, in A. Jacobson-Widding & D. Westerlund (eds.), Culture, Experience
  and Pluralism: African Ideas of Illness and Healing, 289–301. Uppsala Studies in Cultural
  Anthropology 13. Stockholm: Almqvist & Wiksell International.
——. 1997. Questioning Misfortune: The Pragmatics of Uncertainty in Eastern Uganda.
  Cambridge Studies in Medical Anthropology 4. Cambridge: Cambridge University
Widenmann, Dr. 1895. ‘Beschneidung bei den Masai’, in R. Virchow (ed.), Verhandlungen
  der Berliner Gesellschaft für Anthropologie, Ethnologie und Urgeschichte, 302–303. Berlin:
  A. Asher & Co.
Widlok, T. 2001. ‘Living on Ethnography and Comparison: What Difference Do
  Hai//om “Bushmen” Make to Anthropology (and Vice Versa)?’, Anthropos 96,
Widman, R. 1979. Trosföreställningar i Nedre Zaïre från 1880-talet. Skrifter utgivna av
  Religionshistoriska institutionen i Uppsala (hum. fak.) genom C.-M. Edsman 17.
  Stockholm: Gummessons.
Wijsen, F.J.S. 1993. ‘There Is Only One God’: A Social-Scientific and Theological Study of
  Popular Religion and Evangelization in Sukumaland, Northwest Tanzania. Kampen: Kok.
Wijsen, F., & R. Tanner. 2000. Seeking a Good Life: Religion and Society in Usukuma,
  Tanzania 1945–1995. Nairobi: Paulines Publications Africa.
——. 2002. ‘I Am Just a Sukuma’: Globalization and Identity Construction in Northwest
  Tanzania. Amsterdam & New York: Rodopi.
Wilhelm, J.H. 1954. ‘Die !Kung-Buschleute (mit einer Einführung und besonderen
  Anmerkungen von F.R. Lehmann)’, Jahrbuch des Museums für Völkerkunde zu Leipzig
  XII. Leipzig: Otto Harrassowitz.
Wing, J. van. 1959. Études Bakongo: Sociologie —religion et magie. 2nd edn. Brussels:
  Desclee de Brouwer.
Wolff, N.H. 1979. ‘Concepts of Causation and Treatment in the Yoruba Medical
  System: The Special Case of Barrenness’, in Z.A. Ademuwagun et al. (eds.), African
  Therapeutic Systems, 125–131. Waltham, MA: Crossroads Press.
Woodburn, J. 1982. ‘Social Dimensions of Death in Four African Hunting and
  Gathering Societies’, in M. Bloch & J. Parry (eds.), Death and the Regeneration of
  Life. Cambridge: Cambridge University Press.
Yamba, B. 1997. ‘Cosmologies in Turmoil: Witchfinding and AIDS in Chiawa,
  Zambia’, Africa 67 (2), 200–223.
Yoder, P.S. 1981. ‘Knowledge of Illness and Medicine among Cokwe of Zaire’,
  Social Science and Medicine 15 B (3), 237–245.
Zuesse, E.M. 1985 [1979]. Ritual Cosmos: the Sanctification of Life in African Religions.
  Athens, OH, & London: Ohio University Press.

afterlife, 50–51, 57, 72, 82, 87–88,        cosmology, 4, 86, 103, 121–122
   103, 106, 176                            curses, 7, 79, 107, 150, 152 n. 21,
agriculturalist peoples, 2–3, 14, 29, 33,     156–158, 160, 163, 167, 172, 176,
   36, 41, 70, 82, 85, 87, 150, 189,          186–187, 197
   195–196, 198
ancestors, 6–7, 36, 41, 50, 53, 65, 73,     deities/divinities, 41, 43–50, 52–53,
   82–83, 85, 87–97, 99–101, 103–108,          57–62, 75, 121–124, 126–127, 128
   119–121, 123, 127, 141, 143–148,            nn. 46, 51, 129–148, 177, 179
   154, 165, 169, 171, 180, 189–190,           n. 104, 182, 185, 187, 215 n. 41
   192, 201, 205, 211                       disease, 5–10, 12–16, 19–20, 23 n. 80,
Angola, 26, 35–36, 50                          28, 29 n. 18, 41, 43, 54, 57, 59–60,
anthropology, medical, 4–5, 8                  61, 63, 65, 73–80, 82–83, 91–93,
Arusha, 14–15, 28–29, 31, 70, 73–74,           95, 97, 99, 100 n. 80, 101, 104,
   77, 82–83, 157, 160, 195, 203               105–107, 110, 113, 118–119, 123,
                                               125–128, 130–134, 137–141, 145,
babalawo, 131–132, 137, 139–140,               147, 149, 150, 152–154, 156–160,
   142–143, 181, 214                           163, 167, 171, 176, 189–190, 193,
Bantu, 12, 25–26, 31 n. 26, 32, 35,            195–198, 203, 208, 210–211, 214
   36, 43 n. 10, 54, 73, 80, 104, 119       divination, 3–4, 92, 94–96, 115 n. 62,
   n. 78, 152–153, 162, 189, 195, 198          132–134, 211, 212 n. 29
biomedicine, 6, 8 n. 28, 80–81, 118,        doctors, 115, 128, 156, 210, 212, 214
   162, 184, 193
blessings, 7, 67, 76, 95, 107, 127, 139,    education, 21, 42, 163, 165, 187, 193, 205
   156–157, 158, 161, 176, 186              elders, 7, 30–31, 34, 37, 66 n. 10, 67,
Botswana, 10, 12, 25–28, 209 n. 6              76–79, 89, 107, 150, 156–158, 160,
burial, 50, 72–73, 83, 106–107, 130,           162, 172 n. 48, 197, 205
   135, 144                                 eradication (of witchery) movements, 3,
Bushmen, 12, 25, 42–43, 149, 151,              175, 192, 199–202
   152–153, 195–196, 209                    ethnicity, 3, 5, 32, 81, 154–155, 158,
                                               180, 192, 202
cattle, 29–31, 33, 36, 69, 77–78, 89,       etiologies (of disease), 2 n. 6, 4–6, 8–9,
   162                                         13, 16 n. 61, 104, 193, 202, 208
Chagga, 31, 73, 80, 163                     Evans-Pritchard, E.E., 4 n. 12, 7
chiefs and chieftainship, 30, 34, 37,          n. 26, 166
   39, 86 n. 9, 87, 89, 93, 107, 145,       evil eyes, 150, 197
   150, 167, 169, 171–172, 199–201,
   205                                      God, 17, 43–52, 55, 57–62, 65–70,
Christians and Christianity, 6, 13           72–79, 82, 85, 88–89, 92, 95,
   n. 46, 15, 17, 19, 21, 23, 32–33,         97–100, 104, 106, 109, 115,
   35, 40, 42–44, 46, 48, 51, 69 n. 33,      118–126, 132–133, 135, 137,
   71–72, 73 n. 69, 77, 81–82, 86, 88,       141–143, 148, 150, 154–158, 161,
   89 n. 29, 98 n. 66, 104, 115–116,         169, 177, 181, 185, 198, 204, 213
   118–119, 122–125, 134, 136, 142,
   147, 171, 176, 187, 191–194, 198,        healers, 52, 54–62, 67 n. 17, 72, 74,
   200–202                                    79–82, 95–96, 110, 113, 116, 132, 140,
Church Missionary Society, 20–21, 24,         143, 150–156, 158–159, 162, 167, 169,
   38 n. 54, 122–123, 136, 141, 187           174, 180, 182, 184–187, 194–195,
circumcision, 31, 67, 143, 197, 215           198, 200, 203, 208, 212, 214–215
236                                    index

healing and therapy, 4, 8, 15, 54 n. 102,     191–192, 194–198, 203–204, 207,
  59, 61, 81, 96, 104, 109, 115–116,          210–211, 211 nn. 12, 14, 16, 214
  117, 136, 138–140, 150, 153, 156,         Mbiti, John S., 1, 7
  158–159, 179, 186, 194–196, 204,          medicine, 4, 6, 15, 21–23, 53, 63, 74,
  210, 212–213, 215                           81, 96, 99, 109–110, 132, 138, 151,
heavenly beings, 41, 43, 48, 54, 61, 63       155, 159, 163, 165–166, 181,
herbalists and herbalism, 23, 109, 115,       183–184, 189–190, 202, 205, 209,
  181, 183, 194, 202, 211–215                 210, 212–215
hunting-gathering peoples, 2–3, 12, 25,     missions and missionaries, 9–11,
  27–28, 29 n. 17, 50–51, 56, 149–150,        13–21, 24, 35, 38, 86, 103–104,
  191, 194, 196–197, 203–204                  106, 116, 118, 120, 122–123, 160,
                                              166, 171, 173–175, 187, 191–192,
Idowu, E. Bolaji, 1, 22, 121–122, 124,        200
    125 n. 139, 126, 127 n. 43, 128         Muslims and Islam, 23, 32–33, 40, 65,
    n. 46, 130 nn. 64, 66, 131 n. 73,         81, 101, 123–125, 142, 191–193,
    132 n. 81, 133–134, 135 n. 109,           200, 202
    136 n. 115, 137 nn. 126–127, 138        myth, 3, 38, 42, 45–46, 49–50, 52, 66,
    n. 134, 139 n. 139, 141 n. 156, 142       94, 104, 122, 130
    n. 157
iloibonok, 30, 72, 74, 80, 82, 154–163,     Namibia, 11–12, 26–28, 50
    186 n. 156, 195, 197, 203, 210          nganga/banganga, 109–110, 115–118, 167,
initiation, 43 n. 10, 67, 91, 131–134,         172–174, 202, 213
    139, 145                                Nharo, 10–12, 26–28, 42–44, 46–62,
                                               150–154, 195–197, 204, 209
Jesuits, 18–19, 35, 116                     Nigeria, 21–23, 38, 121, 125, 129,
                                               131, 215
Kenya, 15, 28, 194                          nkisi, 19, 103–105, 108–120, 174, 192,
kindoki, 19, 108, 116, 118, 166–167,           199, 213
   172–176, 199, 201, 207, 213              num, 55–56, 59, 63, 96 n. 56, 150,
kinship, 27, 30, 34, 39 n. 55, 54, 94,         152–153, 204
   105 n. 14, 191, 205                      Nyae Nyae, 11–12, 41, 43, 47, 51–52,
Ko, 12, 26, 49, 51, 61, 152, 209               61
Kongo, 3, 17–20, 26 n. 12, 35–37,           Nyamwezi, 17, 32, 86 n. 8, 88 n. 16,
   103–110, 113, 114 n. 57, 115 n. 62,         90 n. 30, 92 n. 40, 94 n. 50, 97, 99
   116–120, 123, 165–167, 172, 174,            n. 75, 101 n. 87, 166 n. 8, 168, 170
   175 n. 72, 176 n. 78, 190–191, 192          n. 31, 171 n. 37, 211–212
   n. 22, 196, 198, 199 n. 42, 201,
   204–205, 207, 213–214                    offerings, 47, 53, 67, 70, 73, 83,
Kung, 10–12, 13, 26 n. 12, 27–28,              94–95, 107, 117, 130, 139, 141, 143
   41, 43, 44 n. 17, 46, 47, 49–51,         oloirirua/iloiriruani, 71, 79–81
   52 n. 88, 53, 54 n. 104, 55–56,          orisha, 121–122, 123 n. 13, 126–128,
   58–63, 149–153, 196, 204                    131 n. 78, 132–133, 140–142,
Kxoe, 12–13, 26, 47 n. 36, 48 n. 52,           145–147, 177, 187, 193, 202
   58 n. 131, 61, 152
                                            pastoralist peoples, 3, 14–15, 29,
Laman, Karl, 18 n. 66, 20, 35–36,             31–33, 65–67, 70, 72, 74, 82,
  103, 105–120, 172, 173, 174                 82 n. 121, 85, 149–150, 154, 160,
                                              190, 191, 194–195, 197, 203, 210
Maa language, 15, 28, 29 n. 17, 31          polygamy and polygyny, 27, 30, 170,
 n. 26                                        180–181, 196–197
Maasai, 3, 13–15, 26, 28–29, 31–32,         prayer, 44 n. 17, 47, 50, 53, 67–68,
 65–73, 75–78, 80–82, 85–86, 98,              70, 73, 79, 82–83, 89 n. 29, 94–95,
 149–150, 154, 155 n. 41, 156–159,            97–99, 117, 120, 124–125, 127–128,
 161–163, 172 n. 44, 176, 186 n. 156,         130, 132–134, 157, 186
                                     index                                  237

priests and priesthood, 37 n. 48, 42,        141, 165–174, 179–181, 187,
  65–66, 96, 120–121, 127, 129–130,          190–192, 196, 198–199, 200 n. 46,
  134–135, 144, 146–147, 192                 202 n. 55, 204, 207, 211–214

ritual, 4, 30, 34, 43, 54 n. 102, 67,     taboos, 8, 90, 101, 109, 114–115, 130,
   71, 73, 76, 83, 85–86, 89–91, 93,         131, 135–137, 139, 145
   96–97, 121, 128–129, 131, 134–135,     Tanzania, 14–17, 28–29, 31–34, 71,
   139–140, 144–146, 154–158, 168,           80, 82, 96, 159, 163, 194, 197
   189, 192–193, 197, 201, 215            trance dances, 43, 53–56, 58–60,
                                             62–63, 150, 154, 195–196
sacrifices, 47, 70–71, 75, 79, 82,         trees, 36, 41, 59, 70, 76–77, 79, 90,
   89–91, 99, 101, 114, 117, 120, 124,       113–114, 132, 159, 174, 212
   126–127, 130, 133, 134–135, 141,       tsso, 55, 150, 152–153, 204
   143–144, 146, 180
San, 3, 10–12, 25–28, 33, 41–46, 47       van Wing, Joseph, 18, 105–111, 113,
   n. 36, 48–51, 53–54, 56, 59 n. 137,      114–117, 119–120, 166, 172–175,
   61–63, 66–67, 75, 85–86, 96,             213
   116–117, 149–154, 163, 191–192,
   194–197, 203–204, 207–209              White Fathers, 9, 15–17, 21, 86–87,
Schmidt, Father Wilhelm, 11, 13–14,         170
   44–47, 49–50, 51 n. 81, 53 n. 96,      witchery, 3, 7, 20, 86, 91–93, 101,
   63 n. 163, 65 n. 5, 69 n. 31,            105, 108, 122, 135, 141, 145–146,
   70 n. 36, 71 n. 43, 72 n. 64, 78,        149, 150–153, 159–160, 163, 165,
   79 n. 101, 149, 154 n. 31                166–167, 169–171, 174, 176–181,
sin, 61, 76, 77, 93, 94 n. 47, 125          185–192, 194–208, 212
smallpox, 93, 101, 128–131, 134, 144,     women, 27, 31, 36, 37 n. 48, 39, 54,
   147, 179, 184, 215                       62, 67, 70, 77, 81–83, 90, 95–96,
soul/spirit, 51, 56–58, 60, 62, 72, 87,     108, 110, 113, 144–146, 153, 156,
   92 n. 40, 105, 113, 117, 120, 141,       162, 171, 175, 177, 179–180,
   142–143, 150, 173–174, 178               185–186, 188, 196–197, 199
spirit possession, 71, 80–81, 96, 101,    Yoruba, 1 n. 4, 3, 20–23, 25 n. 6, 26
   116–117, 128, 140, 155, 162–163,         n. 12, 38–40, 121–129, 131–132,
   197–198                                  134, 136, 138, 140–142, 144, 145 n.
spirits, 2, 6–9, 41, 43, 48, 50–53,         179, 147–148, 165, 167, 176–177,
   56–63, 65, 71, 73, 79–82, 82 n. 121,     179–181, 182 n. 124, 183, 184,
   85, 88–95, 98–101, 103–106,              186–187, 190–192, 193 n. 23, 196,
   108–111, 114, 116–123, 128,              198–199, 202, 204–205, 207,
   139–143, 145–148, 150, 172, 174,         208 n. 85, 214–215
   189, 192–193, 198, 208, 213
Sukuma, 3, 4 n. 12, 15–17, 26             Zaire, 19–20, 35–36, 189, 213
   n. 12, 32–34, 85–98, 100, 119, 123,
                STUDIES OF RELIGION
                     IN AFRICA

 1. MOBLEY, H.W. The Ghanaian’s Image of the Missionary. An Analysis of the
    Published Critiques of Christian Missionaries by Ghanaians, 1897-1965.
    1970. ISBN 90 04 01185 4
 2. POBEE, J.S. (ed.). Religion in a Pluralistic Society. Essays Presented to
    Professor C.G. Baëta in Celebration of his Retirement from the Service
    of the University of Ghana, September 1971, by Friends and Colleagues
    Scattered over the Globe. 1976. ISBN 90 04 04556 2
 3. TASIE, G.O.M. Christian Missionary Enterprise in the Niger Delta, 1864-1918.
    1978. ISBN 90 04 05243 7
 4. REECK,D. Deep Mende. Religious Interactions in a Changing African
    Rural Society. 1978. ISBN 90 04 04769 7
 5. BUTSELAAR, J. VAN. Africains, missionnaires et colonialistes. Les origines de
    l’Église Presbytérienne de Mozambique (Mission Suisse), 1880-1896.
    1984. ISBN 90 04 07481 3
 6. OMENKA, N.I. The School in the Service of Evangelization. The Catholic
    Educational Impact in Eastern Nigeria 1886-1950. 1989.
    ISBN 90 04 08932 3
 7. JE DREJ, M.C. & SHAW, R. (eds.). Dreaming, Religion and Society in Africa.
    1992. ISBN 90 04 08936 5
 8. GARVEY, B. Bembaland Church. Religious and Social Change in South
    Central Africa, 1891-1964. 1994. ISBN 90 04 09957 3
 9. OOSTHUIZEN, G.C., KITSHOFF, M.C. & DUBE, S.W.D. (eds.).
    Afro-Christianity at the Grassroots. Its Dynamics and Strategies. Fore-
    word by Archbishop Desmond Tutu. 1994.
    ISBN 90 04 10035 0
10. SHANK, D.A. Prophet Harris, the ‘Black Elijah’ of West Africa. Abridged by
    Jocelyn Murray. 1994. ISBN 90 04 09980 8
11. HINFELAAR, H.F. Bemba-speaking Women of Zambia in a Century of
    Religious Change (1892-1992). 1994. ISBN 90 04 10149 7
12. GIFFORD, P. (ed.). The Christian Churches and the Democratisation of Africa.
    1995. ISBN 90 04 10324 4
13. JE DREJ, M.C. Ingessana. The Religious Institutions of a People of the
    Sudan-Ethiopia Borderland. 1995. ISBN 90 04 10361 9
14. FIEDLER, K. Christianity and African Culture. Conservative German
    Protestant Missionaries in Tanzania, 1900-1940. 1996.
    ISBN 90 04 10497 6
15. OBENG, P. Asante Catholicims. Religious and Cultural Reproduction
    Among the Akan of Ghana. 1996. ISBN 90 04 10631 6
16. FARGHER, B.L. The Origins of the New Churches Movement in Southern
    Ethiopia, 1927-1944. 1996. ISBN 90 04 10661 8
17. TAYLOR, W.H. Mission te Educate. A History of the Educational Work of
    the Scottish Presbyterian Mission in East Nigeria, 1846-1960. 1996.
    ISBN 90 04 10713 4
18. RUEL, M. Belief, Ritual and the Securing of Life. Reflexive Essays on a Bantu
    Religion. 1996. ISBN 90 04 10640 5
19. McKENZIE, P. Hail Orisha! A Phenomenology of a West African
    Religion in the Mid-Nineteenth Century. 1997.
    ISBN 90 04 10942 0
20. MIDDLETON, K. Ancestors, Power and History in Madagascar. 1999.
    ISBN 90 04 11289 8
21. LUDWIG, F. Church and State in Tanzania. Aspects of a Changing
    Relationship, 1961-1994. 1999. 90 04 11506 4
22. BURKE, J.F. These Catholic Sisters are all Mamas! Towards the Incultura-
    tion of the Sisterhood in Africa, an Ethnographic Study. 2001.
    ISBN 90 04 11930 2
23. MAXWELL, D., with I. LAWRIE (eds.) Christianity and the African Imagi-
    nation. Essays in Honour of Adrian Hastings. 2001.
    ISBN 90 04 11668 0
24. GUNNER, E. The Man of Heaven and the Beautiful Ones of God. 2003. In
    preparation. ISBN 90 04 12542 6
25. PEMBERTON, C. Circle Thinking. African Women Theologians in
    Dialogue with the West. 2003. ISBN 90 04 12441 1
26. WEISS, B. (ed.). Producing African Futures. Ritual and Reproduction in a
    Neoliberal Age. 2004. ISBN 90 04 13860 9
27. ASAMOAH-GYADU, J.K. African Charismatics. Current Developments
    within Independent Indigenous Pentecostalism in Ghana. 2004.
    ISBN 90 04 14089 1
28. WESTERLUND, D. African Indigenous Religions and Disease Causation. From
    Spriritual Beings to Living Humans. 2006. ISBN 90 04 14433 1
29. FAULKNER, M.R.J. Overtly Muslim, Covertly Boni. Competing Calls of
    Religious Allegiance on the Kenyan Coast. 2006. ISBN 90 04 14753 5