Full text of "The Bantu Speaking Tribes Of South Africa"

THE BOOK WAS
DRENCHED

. 60398 >m

a: ~~ 73 ^

m -< CO

Z^ ^>

' pUP 43 30-1-71 5,000

OSMANIA UNIVERSITY LIBRARY

Call No. S~l 3 &% 3 Accession No. ^ O 5"
Author

This book shojrfd be returned on or berore the date last marked below.

THE BANTU-SPEAKING TRIBES OF SOUTH

AFRICA

PLATE 1

[F. Etlmbtrgcr

Girl Initiates (Bale) of the Tlokwa Tribe, Bcchuanalancl Protectorate
(Their bodies are smeared with white clay)

[frtntispitc

THE

BANTU-SPEAKING TRIBES

or SOUTH AFRICA

An Ethnographical Survey

Edited for the (South African} Inter-University
Committee for African Studies

by
I. SCHAPERA

Contributors :

RAYMOND A. DART MONICA HUNTER

CLEMENT M. DOKE

W. M. EISELEN

A. J. H. GOODWIN

ELLEN P. HELLMANN

A. WINIFRED HOERNLE

PERCIVAL R. KIRBY
EILEEN JENSEN KRIGE
G. P. LESTRADE
J. S. MARAIS
I. SCHAPERA

N. J. VAN WARMELO

LONDON

ROUTLEDGE & KEGAN PAUL LTD

BROADWAY 1IOI/SK: 68-74 CARTER LANE, E.G. 4

First published 1937

Second impression 1946

Third impression 1950

Fourth impression 1953

Fifth impression 1956

Sixth impression 1959

Seventh impression 1962

Printed in Great Britain by
Lowe and Brydone (Printers) Limited, London. N.W.io

CONTENTS

PREFACE xiii

CHAPTER . PAGE

I. RACIAL ORIGINS. By RAYMOND A. DART, M.Sc., M.D., Ch.M.,
F.R.S.S.Af., Professor of Anatomy in the University of the

Witwatersrand, Johannesburg i

Introduction

Historical Background

Racial Features

Cranial Characters : Asiatics in Africa ....
The Racial Basis of the Bantu : A Bush and Negro Matrix
Racial Minglings of Bush and Bantu : The Contribution of the Brown

Race (Hamites)

Subdivision of the Southern Bantu

The Contribution of the Nordic Race ....

7
10

M
'9

H. HABITAT. By A. J. H. GOODWIN, M.A., F.R.S.S.Af., Senior
Lefturer in Ethnology and Archaeology in the University of

Cape Town . . . 33

General Morphology 33

Climate 34

Environmental Regions 35

The Tropical Belt 35

Eastern Cattle Lands 36

The Western Drought Lands 3*

III. GROUPING AND ETHNIC HISTORY. By N. J. VAN WARMELO, M.A.,

Ph.D., Ethnologist, Native Affairs Department, Union of South

Africa 43

Introduction 43

Nguni Group 45

Cape Tribes Proper 4*

Fingo and other Recent Immigrants into the Cape ... 47

Natal Nguni .48

Swazi 5<>

Transvaal Ndebele 53

Recent Nguni Offshoots 54

Shangana-Tonga Group 55

Sotho Group 57

Southern Sotho 5*

Western Sotho &>

Transvaal (Eastern) Sotho 6l

Venda Group *3

Lemba *5

IV. SOCIAL ORGANIZATION. By A. WINIFRED HOERNLE, B.A., Senior

Le&urer in Social Anthropology in the University of the Wit-
watersrand, Johannesburg 6?

Introduction *7

Common Characteristics 68

The Tribe <58

The Household 69

The Sub-Distrift and Distrift 69

The Kin (i) Relatives by Blood 70

vi CONTENTS

CHAPTER PAGE

(ii) Relatives by Marriage 73

Other Groupings 74

Nguni Group 74

(a) Southern Nguni

Um^i 74

Ibandla and ikurutta 77

hiduko or Clan 80

Age and Sex Groupings 81

(J) Northern Nguni

Umu^i 82

IsiBongo 83

Age Groupings 84

Nguni Offshoots 86

Sotho Group

Kin by Blood and by Marriage 86

The Village 88

The Ward 88

Larger Tribal Divisions 89

Households 91

Totem Groups ......... 91

Age Sets 93

Shangana-Tonga and Venda 93

V. INDIVIDUAL DEVELOPMENT. By EILEEN JENSEN KRIGE, M.A. . 95

Child Life 95

Education and Initiation 98

General Education . . 98

Initiation " Schools " 99

Puberty Ceremonies 100

Circumcision 100

The Boxwera 103

The Komana 104

Byalc. or Girls* " Circumcision " School 104

Methods of Teaching in the "Schools" . .' . . .106

Sex Life and Marriage 107

Sex Life ,07

Marriage in

VI. DOMESTIC AND COMMUNAL LIFE. By G. P. LESTBADE, M.A.,

Professor of Bantu Languages in the University of Cape Town 119

The Daily Round 120

The Village I2O

Early Morning 121

The Morning Meal 122

The Later Morning and Early Afternoon . . . .122

The Late Afternoon 125

The Evening Meal , 2 6

The Evening I27

Seasonal Changes in Daily Life 128

Occasional Interruptions of Daily Routine 128

VH. WORK AND WEALTH. By I. SCHAPERA, M.A., Ph.D., F.R.S.S.AF.,
Professor of Social Anthropology in the University of Cape
Town ; and A. J. H. GOODWIN, M.A., F.R.S.S.Af. . . .131

Commodities and their Production nl

Foodstuffs ,j,

Horticulture j l ^

Animal Husbandry .' .' \ \yj

Hunting ,J,

Clothing and Ornaments I4a

CONTENTS vii

CHAPTER PAGE

Dwellings and Household Goods 144

Calendar of Work X4 8

Division of Labour 149

Exchange 153

Property and Inheritance 156

Land 156

Livestock 157

Produce 159

Huts, Household Utensils, and Personal Belongings . . .160

Family Obligations 161

Inheritance 162

Special Gifts and Services 164

Obligations towards Relatives 164

The Position of the Chief 166

The Breakdown of Tribal Economy 170

POLITICAL INSTITUTIONS. By I. SCHAPERA, MA., Ph.D.,

F.R.S.S.AF. 173

The Chieftainship 174

Succession and Minority 174

Prerogatives and Wealth 176

Duties and Obligations 177

Relatives of the Chief
Tribal Councils .
Local Administration .
Citizenship and Status .
Inter-tribal Relations

179
181
185
187
191

IX. LAW AND JUSTICE. By I. SCHAPERA, M.A., Ph.D., F.R.S.S.Af. . 197

Introduction 197

Nature and Sources of Bantu Law 197

Civil and Criminal Law 198

Contracts 200

Civil Wrongs 204

Wrongs against Family Rights 204

Wrongs against Property 206

Defamation 207

Penal Offences 208

Crimes against the Person 208

Crimes against Tribal Authorities 211

Sorcery and other Unnatural Offences 211

Procedure 212

Courts and their Jurisdiction 212

Trials 214

Factors affecting Liability 217

X. MAGIC AND MEDICINE. By A. WINIFRED HOERNLE, B.A. . 221

The inyanga . . 226

Protecting a Village
Heaven Doctors .
Agricultural Ceremonies

Methods of Divination
Ordeals
The Witch and the Sorcerer

240
241

XI. RELIGIOUS BELIEFS AND PRACTICES. By W. M. EISELEN, M.A.,
Ph.D., sometime Professor of Anthropology in the University
of Stellenbosch, Chief Inspe&or of Native Education in the
Transvaal ; and I. SCHAPERA, MA., Ph.D., F.R.S.S.AF. . . 247

The Spirits of the Dead 247

Manifestations of the Spirits 251

viii CONTENTS

The Worship of the Ancestors 254

Family and Tribal Rites .... ... 259

High Gods and Other Deities 262

TheSkyGod 262

Rain-making Rites 265

Lesser Deities 269

Religion and Morality 270

XII. THE MUSICAL PRACTICES OF THE NATIVE RACES OF SOUTH AFRICA.
By PEBCIVAL R. KIRBY, M.A., D.Litt., F.R.C.M., Professor
of Music in the University of the Witwatersrand, Johannesburg
Introduction .......... 271

Instrumental Music 271

Bushman ... . ..... 271

Hottentot 272

Bantu
Vocal Music

Bushmen

Hottentot

Bantu
Glossary of Technical Musical Terms used

274
281
281
282
284
288

XIII. TRADITIONAL LITERATURE. By G. P. LESTRADE, M.A. . .291

Introduction 291

Kinds of Literature 292

Myths, Legends, Fables, and Tales 292

Proverbs and Riddles 293

Songs 294

Praise-poems 295

Origin and Propagation 297

General Characteristics . . . . ' . . . .301

Language, Style, and Technique 302

Language 302

Style 3<>5

Technique 306

XIV. LANGUAGE. By CLEMENT M. DOKE, M.A., D.Litt., Professor of
Bantu Languages in the University of the Witwatersrand,

Johannesburg 309

Main Characteristics 309

General Classification 312

South African Bantu Languages .. 313

Special Features 314

Differences 316

Similarities 318

Grammatical Forms 319

The Substantive 319

The Noun . . . 319

The Pronoun 321

The Qualificative 322

The Predicative 324

Verbal Derivatives 327

Compound Tenses 328

The Copulative 329

The Adverb 330

Theldeophone 330

CONTENTS

CHAPTER PAGE

XV. THE IMPOSITION AND NATURE OF EUROPEAN CONTROL. By J. S.
MARAIS, M.A., D.Phil., Senior Leclurer in History in the

University of Cape Town 333

Cape Colony 33)

Natal 342

The Boer Republics 345

The Transkei 350

The Union 353

XVI. CULTURAL CHANGES IN TRIBAL LIFE. B\ I. SCHAPERA, M.A., Ph.D.,

F.R.S.S.Af. 357

Mechanisms of Culture Contact . 357

Types of Contaft . -357

European Aims and Policies . 359

Native Reactions . 361

Economic Changes . 363

Religion and Magic . 368

Government and Law . 372

Education and Health . 376

Social and Domestic Life . 380

The Trend of Cultural Development . 385

XVII. THE BANTU ON EUROPEAN-OWNED FARMS. By MONICA HUNTER,

M.A., Ph.D 389

The Household 391

Community Life 394

Individual Development 399

XVIII. THE NATIVE IN THE TOWNS. By ELLEN P. HELLMANN, M.A. . 405

Material Culture 406

Income and Expenditure ........ 407

Individual Development 4 12

Birth 4>3

Education and Puberty Rites 4' 5

Marriage 419

Death 4*3

Religion and Magic 4*4

Detribalization 4*9

BIBLIOGRAPHY 435

INDEX OF TRIBAL NAMES 445

INDEX OF AUTHORS CITED 44 8

GENERAL SUBJECT INDEX 45

LIST OF ILLUSTRATIONS

PLATE. PAGE

I. Girl Initiates (Bati) of the Tldkwa Tribe, Bechuanaland Protectorate.

(V. Ellenberger) Frontispiece

II. Physical Types 16

(a) Bhaca man, Transkei. (N. J. van Warmelo.)
(V) Swazi man. (N. J. van Warmelo.)

III. Physical Types 17

(a) Kwena man, Bechuanaland Protectorate. (N. J. van Warmelo.)
() Lemba man, Zoutpansberg, Transvaal. (N. J. van Warmelo.)

IV. Physical Types 56

(a) Venda woman. (N. J. van Warmelo.)

(4) Transvaal Ndebele woman. (N. J. van Warmelo.)

V. Territorial Organization 72

(a) Swazi kraals. (S.A.R. and H.)

(J) Portion of Mochudi Village, Kxatla Reserve, B.P. (I. Schapera.)

VI. Native Homesteads 73

(a) Natal Nguni. (S.A.R. and H.)

(4) Transvaal Sotho. (N. J. van Warmelo.)

VII. "Daughters of Africa" 9<*

Kxatla children, Mochudi, B.P. (I. Schapera.)

VIII. INITIATION CEREMONIES 97

(a) Vhusha rite of the Venda. (N. J. van Warmelo.)
() Kxatla maxwane (male novices). 0- Reyneke.)

IX. Initiation Ceremonies 112

Dance of the abakweta (Cape Nguni). (S. A.R. and H.)

X. Zulu Wedding Dance. (S.A.R. and H.) 113

XI. Household Occupations 120

(a) Drawing water. (I. Schapera.)

() Smearing the Courtyard Floor. (I. Schapera.)

XII. Domestic Scenes 121

(a) Children at Play. (I. Schapera.)

(6) An informal beer drink. (I. Schapera.)

XIII. Agricultural Life 136

(a) Kxatla Fields. (I. Schapera.)
() Threshing Corn. (I. Schapera.)

XIV. Pastoral Life M4

(<t) Milking. (I. Schapera.)

(6) Herdboys riding oxen. (I. Schapera.)

XV. Arts and Crafts 145

(a) Setting a snare for birds. (I. Schapera.)
() Potter at work. (I. Schapera.)

XVI. Interior of a Zulu Hut. (S.A.R. and H.) 160

XVII. Political Life 184

(a) The Great Tribal kxotla, Mochudi, B.P. (I. Schapera.)
(4) Graves of the Kwena chiefs, B.P. (I. Schapera.)

xii LIST OF ILLUSTRATIONS

PLATE PAGE

XVIII. Political Life 1X5

(a) Morning gossip at a local kxotla. (I. Schapera.)
(&) Hearing a lawsuit. (I. Schapera.)

XIX. Magical Practices 232

% (a) Divining the cause of a patient's illness. (N. J. van Warmelo.)

(6) Rain-maker preparing his medicines. (I. Schapera.)

XX Religious Ceremonies 256

Venda thevhula (first-fruits) sacrifices. (N. J. van Warmelo.)

XXI. Music and Dancing 280

Scenes from the domba initiation dance of the Venda. (N. J. van Warmelo.)

XXII. Changing Life in the Reserves 358

(a) Village Church Service. (J. Reyneke.)
(*) Solving the Water Problem. (I. Schapera.)

XXIII. Urban Native Life 406

(a) Scene in a Johannesburg Slum Yard. (. P. Hellmann.)
() Scene in Langa Native Township. (A. J. H. Goodwin.)

XXIV. Urban Native Life jo 7

" War " dances in a Johannesburg Mine Compound. (H. V. Meyerowitz.)
Map 454

PREFACE

T^HIS book is the outcome of a resolution adopted by the (South
African) Inter-University Committee for African Studies in
July, 1934, to " sponsor the preparation and publication of a handbook
of South African tribes ". The need for such a book has long been
felt, not only by teachers and students of anthropology in the South
African Universities, but also by others interested in the racial problems
of the country. There exist some large monographs about individual
groups or tribes, and many useful short accounts, of a more general
nature, published either as articles in scientific journals or as seftions
of works dealing with the continent as a whole. But there has hitherto
been no single comparative survey sufficiently detailed, and sufficiently
catholic in scope and content, to form a satisfactory manual of South
African ethnography. It is our hope that this work will be found to
fulfil reasonably adequately the purpose for which it was written.
Increasing specialization in study has made large-scale collaboration
not only desirable but essential, and although the book may in con-
sequence perhaps have suffered some loss of unity it has certainly
gained in authoritativeness. The contributors have all had considerable
first-hand experience of field investigation among the peoples dealt
with, and almost all of them are or have been engaged in teaching
anthropology, Bantu languages, or allied subje&s in South Africa.

The present state of South African ethnography has to some'extent
diftated the limitations in the range of this book. Certain areas and
ethnic stocks have already been dealt with sufficiently fully in standard
works to make it unnecessary to include them here as well. The
information relating to the Bushmen and Hottentots is summarized
in Schapera's Khoisan Peoples (1930) ; the Bergdama of South- West
Africa are treated exhaustively in Vedder's great work, Die Bergdama
(1923) ; while the small handbook, The Native Tribes of South-West
Africa, issued in 1928 by the South-West Africa Administration, covers
the Ambo and Herero tribes of that Territory, as well as its Bushman,
Hottentot, and Bergdama inhabitants. The present work is therefore
restricted in the main to the Bantu-speaking peoples of the Union of
South Africa and the adjoining British Protectorates, although
reference has occasionally been made by some of the writers to other

ziii

xiv PREFACE

groups for comparative purposes. Our aim all along has been
essentially to summarize existing knowledge rather than to present
merely the results of individual investigations ; but most of the
writers have utilized in their contributions original work whose final
results have not yet been published.

The greater part of the book is devoted to an account of the Bantu
as they were before afFefted by the intrusion of Western Civilization.
The four final Chapters will, however, indicate how considerably
their traditional life has already been altered by this overwhelming
new influence a theme more fully treated in another colleftive work,
Western Civilisation and the Natives of South Africa (edited by I.
Schapera. Routledge, 1934). It has rightly become the fashion in
modern ethnography to study " the changing Native ", and not to
concentrate merely upon his traditional culture. But the understanding
of present-day Native life must rest largely upon a knowledge of
the former culture, and the time is rapidly approaching when such
knowledge will no longer be obtainable in the field. It is highly
desirable, therefore, that more intensive fieldwork should be done in
this country before too much of the old culture has been obliterated ;
and if this book succeeds in stimulating any of its readers to inquire
more fully into some of the topics or peoples dealt with, it will, for
that reason alone, have been worth compiling.

Some reference should be made to the forms used here in writing
the names of Bantu tribes and languages. These forms have been
chosen in conformity with the principles laid down by the International
Institute of African Languages and Cultures, and adopted by the Inter-
University Committee for African Studies. All names are used in
their root form, without prefixes, as nouns and as adjeftives, in the
singular and in the plurad, e.g. a Zulu, two Zulu, Zulu customs, Zulu
(the language), not an umZulu, two amaZulu, isiZulu customs, and
isiZulu respectively. The spelling of the forms is that used in the
languages themselves, e.g. Sotho, not Suto ; but diacritic signs have
been omitted in the commoner names, e.g. Venda, not Venda. The
designations applied to the larger cultural groups are those now in
common use in South Africa, e.g. Nguni, not Zulu-Xhosa ; Sotho, not
Sotho-Tswana. The only one calling for special comment is that of
Shangana-Tonga, for the group hitherto known as Thonga (more
correftly Tonga). Since there are at least three other tribal groups in
Africa bearing the name Tonga, the form Shangana-Tonga has been
chosen as an identifying designation, in conformity with the principle

PREFACE xv

that when two or more tribes have the same name they should be
distinguished from each other by some appropriate geographical or
other label. 1

The task of editing the book was originally entrusted by the
Committee to Professor W. M. Eiselen, of the University of Stellen-
bosch, and Professor I. Schapera, of the University of Cape Town.
Soon after its commencement, however, Professor Eiselen was
appointed Chief Inspeftor of Native Education for the Transvaal
Province, and owing to the pressure of his new duties he felt obliged
to relinquish any further share in the work. He was also unfortunately
unable to contribute the various Chapters originally allotted to him,
apart from his seftion of the Chapter on " Religious Beliefs and
Praftices ". The committee thereupon invited Professor Schapera to
carry on as sole Editor.

The Editor wishes here to thank his collaborators for the generous
assistance they have given him, not only in the preparation of their
own articles, but also in advice regarding the book as a whole. He
would also mention Professor R. F. A. Hoernle and Mr. J. D. Rheinallt
Jones who, although not contributors to the book, have been of much
help in various matters relating to its preparation. He is further
indebted to Mr. V. Ellenberger for the photograph forming the
Frontispiece; to Dr. N. J. van Warmelo for Plates II, III, IV, VM,
Villa, XIXcz, XX, and XXI, and for the map showing the location
of the principal tribes dealt with ; to the Publicity Department of the
South African Railways and Harbours Administration for Plates V<z,
Via, IX, X, and XVI ; to the Rev. J. Reyneke for Plates VIII* and
XXIIa ; to Mrs. E. P. Hellmann for Plate XXIIIa ; to Mr. A.- J. H.
Goodwin for Plate XXIIK; and to Mr. H. V. Meyerowitz for
Plate XXIV.

1 Since the above was written, and after most of the proofs had already been corrected,
the Inter-University Committee for African Studies, at a meeting held in November, 1936,
adopted a recommendation that the form Shangana-Tonga should be employed when writing
of Africa generally, but that when reference is clearly being restricted to South Africa the
shorter form Tonga is sufficiently accurate to stand alone. The occurrence of both forms in
the text is due to the fact that the recommendation came too late to allow of uniform changes
being made throughout the proofs.

CHAPTER I

RACIAL ORIGINS

By RAYMOND A. DART

INTRODUCTION

ANTHROPOLOGICALLY, Africa can be divided into three parts. Two
curves drawn convexly to the north-east across the continent, one
running from Senegal through Khartoum to Mombasa and the other
from Walfish Bay through Johannesburg to East London, separate
it into northern (or north-eastern), central, and southern (or south-
western) belts, predominantly populated to-day by the Brown
(Hamitic or Mediterranean), Negro, and Bush races respectively. The
central or Negro belt can be further subdivided. A line running
parallel to the Equator from the Bight of Benin to Victoria Nyanza
separates it into a northern or Sudanic and a southern or Bantu moiety.
This major division is based on the languages spoken by their
respective peoples. Two other lines, one running north and south
through the Great Lakes to Lake Nyasa and the other running east
and west through the middle of Lake Nyasa (separating the water-
sheds of the Congo and Zambezi) divide the Bantu territory into
eastern, western, and southern regions. The Bantu of this southern
region, embracing Portuguese East Africa, Southern Rhodesia, Angola,
the Union of South Africa, and the British Proteftorates, form the
particular subjeft of this chapter.

For the past 300 years the southern portion of Africa has been
populated by three main groups of people, European, Bantu, and
Bush. Not one of these is a pure race. Each presents a rich racial
pattern which, however intricate to-day, grows ever more complex,
not merely by the fusion of European with European and of Bantu
with Bantu, but also by the past admixture of European with Bantu
and Bush, and the further intermingling of these hybrids with Indians,
Malays, and other Orientals.

Miscegenation is no new feature in the racial history of Africa.
The fossil finds of human bones tell a story of ancient peoples in

2 RAYMOND A. DART

Africa fusing with one another and with aliens. Some ten to twenty
thousand years ago Africa was inhabited by the Boskop race, so
variable in itself as to be suspefted of resulting from the mixing of
as yet unknown, previous races. Already four distinft types of this
race are known : the type represented by the original. Boskop
Skull, the Matjies River type, the Zitzikama type, and the Springbok
Flats type. On the arrival from the North of the Bush race, an ethnic
group comprehending already like other races several fundamental
types, hybridization took place with the Boskop peoples, producing
various combinations and permutations of features within both groups,
so that even at a very early date in South African history the physical
anthropology of the country was a complex ravel. This complexity
was increased by contafts with peoples alien to Africa a faft exempli-
fied by the discovery at Outeniqua (Eastern Cape Province) of ancient
skulls with a Mongoloid (Asiatic) facies in the midst of a Bush-
Boskopoid population. 1 Again, at Mumbwa in Northern Rhodesia,
in the lowest bone-bearing stratum, dating back some 6,000 years,
a European (Nordic) type of skull was found associated with other
bones more Boskopoid than Bush. 2 Cave-deposits prove that, long
before the coming of the Bantu, Southern Africa had attracted aliens.
The Bush people have low-vaulted skulls, yet before the advent of
the Bantu occasional high-vaulted skulls were to be found.

Following the immigrations of the Bantu further hybridization
took place between them and the miscegenated indigenes. In the
prehistoric Bantu era, i.e. before European occupation, we again find
alien types. At Cathkin Peak (Natal) a Bantu skull, instinft with
Mongoloid features, can be assigned to the sixteenth or early
seventeenth century. A European type of skull has been found by
Mrs. Martin in her excavations of pit circles at Penhalonga (Southern
Rhodesia), and another by Dr. Laidler at East London. Shrubsall
described a predynastic Egyptian type, belonging to the Brown
(Hamitic) race, also from the Cape Coast.

The arrival of the Bantu was not only accompanied by extensive
racial mixing but resulted in unleashing very diversified types of
physical features. To-day all types of breadth, length, and height
combinations are to be found in skulls called Bantu ; all types of
cranial form, a wide variation of stature, skin colour, and face type

* L. H. Wells and J. H. Gear (1931), Cave-dwellers of the Outeniqua Mountains," S.
Afr. J. Set., xxviii, 444-469-

R. A. Dart and N. del Grande (1931), " The Ancient Iron-smelting Cavern at Mumbwa,"
Trans. R. Soc. S. Afr., xix, 379-4*7-

RACIAL ORIGINS 3

are there, and to such extent that the variation between individual
members of a tribe is greater than between the averages of one tribe
and its neighbour.

Hitherto physical anthropologists in Africa have been occupied
primarily with the reporting of data. There has been little or no
time for interpreting physical features and their origin. But the
ever-increasing compilation of data makes it desirable for us to
attempt some preliminary correlation of the facts as seen in mass.
Our obje&s then will be to show how history verifies the story of
the bones ; to discuss the outstanding features of the Bantu physical
tangle ; to hint, by comparison and contaft with other types, at the
origin of the Bantu-speaking peoples ; and, incidentally, to fill out
episodes in Bantu history which can now be revealed only by their
physical features.

HISTORICAL BACKGROUND

The sites of early human habitation, in caves, and out in the open,
south of the Zambesi throw no light on the physical or cultural
evolution of the Bantu. They are aliens who entered from the
North by force of iron arms, the secret of whose large-scale manufacture
they had mastered. But history gives some indication of their
movements and the peoples with whom they trafficked and inter-
mingled. Before the coming of the Europeans South Africa was a
dependency of the Orient.

When the Portuguese rounded the Cape they found " Strand-
looper " (beachcomber) and other " Hottentot " communities of
Bush peoples possessing undisturbed the whole southern end of the
continent. From Sofala northwards, however, the coast was inhabited
by Negroes like those of Angola and West Africa ; but the shipping
ports (Kilimane on the mouth of the Zambesi, Mozambique, Melinda,
Kilwa, Zanzibar, Mombasa, Malindi, Lamu, and Mogadishu) were
held and colonized by semi-dependent Moslem chiefs. The Arabs
had penetrated deeply into the interior, settling as far inland as Sena
along the Zambesi. They knew the Great Lakes and had carried
their trade in Negro slaves and elephant ivory across Lake Tanganyika
into the Congo. The European explorers of Central Africa followed
the " roads " occupied by Arab traders.

To the Arabs the whole of this territory down to Sofala was " the
land of the Zing " (or Negro) who were " Kafirs " (or infidels).
Arab geographers, like Masudi (circa A.D. 912), Idris (circa A.D. 1154),

4 RAYMOND A. DART

and Yakut (1179-1229), provide evidence that from the time the
Saxons invaded England down to the Portuguese circumnavigation
of the Cape, the approximate meeting-ground of " Wak-Wak " and
" Zing " (or Bush and Bantu peoples) was in the vicinity of Sofala
or Louren$o Marques.

The Moslem movement, which thrust the Arabs southwards and
westwards through Egypt and over Northern Africa and finally
across the Straits of Gibraltar into Spain (A.D. 711), caused the south-
easterly emigration of the two Arabian chiefs of Oman, the brothers
Suleiman and Said, to the " land of the Zing " (circa A.D. 695). During
the reign of Abdul Malik they escaped, taking their families and a
number of their tribes with them. According to East Coast tradition
the same Abdul Malik built no less than thirty-two towns along the
coast between Mogadishu and the Comoro Islands with the aid of
Syrian craftsmen and placed a Governor in each of them. Arabs did
not enjoy undisputed command of the coast. The Kilwa Chronicles
record the founding of that centre (A.D. 975) by Persians, and in
particular by Ali, the son of Sultan Hassan of Shiraz. By this time
(tenth century), as Masudi reports, the Shirazis (Persians) and the
Adzis (Omanis) had all the coastal trade in their hands. Hassan
and his family of six sons left their native land in seven ships of which
only one, that under Ali (whose mother was an Abyssinian slave),
reached Kilwa in Tanganyika. There he reigned for forty years,
founding the state which predominated along the coast for the following
three centuries, until the coming of the Portuguese.

The Bantu slave-trade provided an unlimited source of man power
and frightfulness in the Moslem wars. Abdul Abbas (brother of the
first Abbasid Caliph) had 400 Negroes in his army, when he massacred
11,000 people in Iraq (A.D. 749). Ultimately, these aliens imported
into Mesopotamia became so numerous and restive that their revolt
under the " Lord of the Blacks " (A.D. 850) was felt all over Arabia.
Their bid for freedom in A.D. 869 under the Persian, Alid Messiah,
resulted in the sack of Basra (A.D. 871) and a decade of pillage at
the hands of an armed force said to number 300,000. The rebels
were finally defeated (A.D. 883). Thus Moslems from the littoral
of the Indian Ocean were freely trading, raiding, and bastardizing
with the Bantu for nearly 800 years before the coming of Europeans
to the East Coast. 1

In the century before the Moslem surges, which had these indireft

1 Vid* W. H. Ingrams (1931), Zanzibar : iu History and its PtopU (London, Witherby).

RACIAL ORIGINS 5

but far-reaching repercussions on the Bantu, the African coast was
dominated by the King of Aksum in Abyssinia. Cosmas Indico-
pleustes (circa A.D. 547) says : " Beyond Barbaria there stretches
an ocean that has the name of Zingion, bordering on the same sea
is the land called Sasos which possesses abundant gold mines " ;
he proceeds to tell how the king sent expeditions to the country which
took six months to go and return, how the people bartered oxen
for gold with the natives and how " the winter of those regions
coincides with the summer amongst us ". This is the first distinctive
record of a black man's territory (Zingion) fringing the Indian Ocean,
and of gold from Africa south of the equator ; it doubtless refers
to Rhodesia. 1 At this time the territory embracing the Rhodesian
gold country is " Sasos ", that is something different from the Bantu
" Zingion ". Presumably we have a record here of how domestic
animals were being introduced into " Hottentot " Southern Africa
nearly 1,400 years ago and before the coming of the Bantu to Rhodesia.
Abyssinia gained her pre-eminent maritime position by destroying
the Himyaritic and Sabsean empire in Arabia (circa fourth century).
The importance of Sabaea (or Sheba) in relationship to South African
prehistory and the movements and physical constitution of the Bantu
people is demonstrated by the records in the Periplus of the Erythraan
Sea and by the study of Bantu linguistics. After Hippalus (A.D. 45)
observed the change of the monsoon in the Indian ocean and made
it public property, a fleet of 120 vessels sailed annually in Roman
times from Myo-Hormos, in Egypt on the Red Sea, to the Malabar
Coast, or Ceylon, at the summer solstice and returned in December
or January. So many Roman and Greek merchants were trading
in the Indian Ocean (or Erythraean Sea) about this time that one
of them wrote the Periplus a kind of travellers' guide. He described
(circa A.D. 60) the ports of Pano (Las Binna), Opone (Ras Hafun),
Serapion (Mogadishu), Nicon (Brawa), and the Pyralae Islands
(Patta, Manda, and Lamu) on his way down the Ausanitic Coast
of the Continent of Azania (Africa) to the Island of Menouthias
(Zanzibar or Pemba) and Rhapta (a port then lying on the mouth
of the Rufiki or the Pangani River). The African export trade he
discussed was in ivory, rhinoceros horn, tortoise-shell, palm oil, and
slaves (required in Egypt and other parts of the Roman Empire) ;
to Africa came lances, hatchets, daggers, awls, glass, wine, and wheat.

* Cf. F. P. Mennell (1902), " The Zimbabwe Ruins near Victoria, Southern Rhodesia/*
Proc. Rhod. Set. Ass.) iii, 81-4.

6 RAYMOND A. DART

Each port was in the hands of a separate chief ; even the most southerly
station mentioned (Rhapta) was administered by Arabians, who were
subject to the people of Muza (Mocha). The coast itself was governed
by a so-called Mapharitic chief, Charibael (Kariba-Il), identified
historically from South Arabian inscriptions as a Sabaean king, who
lived at Saphar (Zafar) and ruled over the Homerites and Sabarites
(Himyarites and Sabaeans) circa A.D. 40-70. By means of embassies
and gifts he held diplomatic intercourse with the Roman Emperors
Claudius and Nero. The Romans sent expeditions against the Sabaeans
from time to time; but they had found wars costly experiments
against this nation, which had been predominant in Southern Arabia
from time immemorial. The Sabaeans rendered tribute only to the
most powerful of the Assyrian kings (seventh century B.C.) and every
biblical student knows of the visit of their queen (Balkis of Sheba)
to Solomon's Court (tenth century B.C.).

This powerful colonizing sea-power extracted its wealth by its
commanding position relative to the maritime trade between Egypt
and the Indian Ocean, whose dimensions in ancient times is best
attested by the fact that Seti I (1380 B.C.) connected the Nile with
the Red Sea by means of a canal. 1 It fell into disrepair, and Necho
(630-527 B.C.) made a spirited attempt at the sacrifice of 12,000 men
to re-establish it. His unsuccessful work was completed by Darius
(520 B.C.) when he attained his desire of commanding the Indian
Ocean trade by subjecting Egypt to Persia. This canal of Darius
was maintained, with periods of intermission, down through Roman
and Moslem times until its closure by the Moslem Mansur (circa
A.D. 770). Any weakness or strife on the part of Egypt, Assyria,
Babylonia, Persia, Greece, and Rome, as successive masters in the
Near East, ensured the liberty and fostered the expansion of the
Sabaean people, who held Egypt's eastern gateway. They were the
Phoenicians of the Indian Ocean and doubtless supplied Necho with
his circumnavigators of Africa, to whose maritime traffic they held
the keys both in the Red Sea and in the historically recorded trading
stations along the East African coast.

We do not know when the Bantu gained sufficient numerical
strength, and political organization, agricultural knowledge, mining
skill, and implement-manufacturing ability to render themselves
independent of the coastal imports of lances, hatchets, awls, and
daggers from coastal traders in the first five centuries of the Christian

1 Vidt Encyclopedia Bntannica, M th ed., art. "Suez Canal".

RACIAL ORIGINS 7

Era, and to avalanche themselves down the eastern side of Africa,
in the way the Goths and Huns hurled themselves across South-
western Europe. Something corresponding with the convulsive
movements of the Gauls into Italy (390 B.C.), of the Teutons into
Gaul (105 and 102 B.C.), and of the Vandals into Italy, Spain, and
finally into Africa (A.D. 429), seems to have taken place in the region
of the Great Lakes and headwaters of the Nile, whence the black
flood had previously constantly threatened and sometimes over-run
Egypt. After destroying the Sabaean power, the Abyssinians were
able to maintain their own independence and to arrest the north-
easterly flow of Bantu ; but their relaxed grip on the Sabaean East
Coast ports and the gold-producing South African belt left the Bush
natives of that territory an easy prey to the southward-advancing
Bantu from the end of the sixth century onwards. We shall probably
not be far from the truth if we place the first great southern migration
of the Bantu at about this period. The physical affinities between
the Nilotic Negroes and our Southern Bantu attest their common
origin.

RACIAL FEATURES

A distribution map of the African races would have presented a
very different picture 2,000 years ago. Few, if any, Negroes occupied
the eastern half of the continent. There, the Bush race were in
direct contact with the Brown race, whose cradle is placed by Sergi
and Elliot Smith in the north-eastern part of Africa. All three races
evolved in Africa and must have been, at one time, sufficiently isolated
from one another to develop their distinctive racial characteristics
and genetic stability. When crossed, their offspring are fertile;
but the anatomical features distin&ive of each type can be traced
in relative purity, or in various degrees of admixture, in their descen-
dants. Let us then consider how these types arose and what features
distinguish them.

The Sahara has been for millenia a great barrier between a free
intermingling of the Brown and Negro races. It has been natural
for anthropologists to imagine that this barren waste, by its isolating
effect, favoured the emergence of these two distinctive types of man-
kind. But the Kalahari desert also extended in a north-easterly dire&ion
across the Viftoria Falls and the two Rhodesias in early Pleistocene
times. It virtually severed the south-eastern from the west-central
part of the continent. This old-time Kalahari barrier must have

8 RAYMOND A. DART

fluctuated with the Ice Ages, as did the Sahara, and at one time probably
separated the Bushmen from the Negro to the west and from the
Brown race to the north. At any rate the cradle of the Negroes,
as nearly as we can determine, was in the basin of the Congo water-
shed. Thence they first spread out in a north-westerly direction
towards the Niger Valley and the Sahara ; and, subsequent to certain
racial hybridization, which we will discuss later, moved in the north-
easterly direction towards the Great Lakes and the sources of the Nile.

Every intelligent person is familiar to-day with the obtrusive
physical characteristics of the typical Negro : the tall and ereft stature ;
well set-up, athletic body of sthenic type, coal-black in colour, and
virtually hairless, save for the tightly-curled woolly mop protecting
the oval head ; the long, thin legs terminating in broad, flat feet.
The skull is infantile in form, being long and relatively narrow, of
moderate height and ovoid in contour, as seen from above. The
eyebrow ridges are negligible or absent ; the forehead is moderately
wide and following the infantile form of the bone curves convexly,
gently, and regularly to the crown of the head, whence it falls away
with equal smoothness of curvature to the rounded occiput. The
form of the orbits is almost square and their margins are strongly
built. The cheeks are wide and, with their supports (zygomatic
arches), are bowed laterally in keeping with the powerful jaw muscula-
ture. The eyes are widely set apart by the broad and flattened nasal
bones. The nose is flat and the nostrils widely expanded. The face
is large and prominent. The jaws are prognathic, accommodating
large, well-formed teeth with a dense ivory covering, which slightly
continue the prognathism and are covered by full, fleshy lips, whose
prominent mucosa is reddish-black or purple in colour. Even the
whites of the eyes are darkened by the strong pigmentation of the
body, but contrast sharply with the deep black of the eyelids and
iris. The chin is not prominent owing to the fleshiness of the lips
and the fullness of the cheeks. The face as a whole is rhomboidal
in form. The skeleton is massive, but the constituent bones are
more slender than those of Europeans owing to their more compact
or ivory-like texture. Their resilience has given rise to an erroneous
impression that the Negro skull is thicker than that of the European.
The air-spaces in the skull are more numerous and larger than those
of the European and are largely responsible for the calibre, resonance
and carrying-quality of the Negro voice.

These characteristics contrast sharply with those of the Brown race,

RACIAL ORIGINS 9

which has spread out from North-East Africa through Southern
Europe to the British Isles and through Southern Asia to Malaya.
They are short and slender, verging on the effeminate in bony structure
and bodily physique. Their brunette bodies have scanty facial hair,
except for a chin-tuft ; but their heads are covered with wavy, black
locks. The iris is brown or black, owing to a considerable retention
of body pigment. The skull is long and relatively narrow, of moderate
height, and pentagonoid (coffin-shaped), when viewed from above ;
the eyebrow ridges are poorly-developed, but the narrow forehead
rises ereft and full above them and the occiput is bulged out behind
into a marked prominence. The orbits are horizontal, ellipsoid, and
with thin margins. The cheeks are narrow, their bony supports
flattened laterally. The nose is small and only moderately developed
but is a definitely elevated ridge. The small jaws do not project
and the chin is pointed. The face as a whole is straight, short, and
narrow and of ovoid form. The teeth are moderate in size. 1

The Bush race resemble the Negro in their broad, flat, and flaring
nostrils and are even more specialized in respeft to their loss of hair ;
its extremely rolled character and sparsity give origin to its peppercorn
distribution. But here the resemblances end. Their bodies are
shorter and more slender than those of the Brown race ; while they
are darker in colour than the Brown, they are not black like the
Negro. The Negro skull we described as infantile in form; that
of the Bushman is foetal. It is narrow relative to its length, but
extremely low, and even more markedly pentagonal than the Brown
skull when viewed from above. Eyebrow ridges are entirely absent
and the forehead, instead of sloping backwards or rising vertically,
bulges anteriorly as it rises towards the crown of the head and then
falls away abruptly, to the prominent occiput posteriorly. 2 The
horizontal orbits are small and circular and have thin margins. The
face is short and straight, and also pentagonoid in form, being widest
opposite the cheek-bones and falling to a point in the chin region.
It is flat and recessed in comparison with the Negro face, for the
jaws project very little, while the bony parts below the lateral parts
of the eyes bulge beyond the nose ; this sub-ocular facial bellying
emphasizes the width in this region. The jaws and the teeth are
small and delicately built and the musculature corresponds with their

1 Cf. G. Elliot Smith (1923), 7% Ancient Egyptians and the Origin of Civilisation (London,
Harper).

s D. Slome (1927), " Curvature of the Bushman Calvarium," Am*r. J. Pliys. Anthrop.,
x,

10 RAYMOND A. DART

form. The facial parts resemble those of the Pygmy and Tasmanian
in being of Australoid or foetal type. The Bush retain primitive
features in the lower jaw 1 which also resembles most closely that
of the Brown race^ the vertebral column, 3 the stru&ure of the feet, 4
and the genitalia. 5

Thus, while the Bush are in many points distinft and more primitive,
the degree of affiliation between them and the Brown race is as great
as or even greater than between either of these races and the Negro.
The physical contrasts are of such a chara&er, that the Negro may
be regarded as a western specialized stock separating very early
from the Sapient stock and pursuing an independent line of evolution
towards tallness, ere&ness, muscularity, infantile head-form, and facial
prognathism, while retaining the primitive, coal-black skin and nasal
conformation. The common ancestral stock of the Bush and Brown
races meantime lost some of its colour ; from this stock arose the
Bushman in the sluggish and unstimulating South and the Brown
Race in the vigorous and favourable North. The Bushman, in
specializing, surpassed the Negro in general loss of body hair and
kinkiness of what remained, but retained the foetal skull, skeletal,
and facial parts which the Negro lost; the Brown Race retained
the primitive, wavy, black hair (chara&eristic also of the Australoid
Race), but lost more colour, expanded their brain-cases and, growing
taller, improved their ereft attitude and bodily agility. Still more
adventurous and lighter-hued offspring from the Sapient stock
reached North- West Africa, Europe, and Asia; there they gave
rise to the Nordic (including the Ainos of Japan and the non-
Mongoloid Eskimo peoples), Armenoid, and Mongoloid stocks
respectively.

CRANIAL CHARACTERS : ASIATICS IN AFRICA

The famous Swede Retzius first drew attention (1840) to the
possibility of distinguishing between people of different racial origin,
by comparing the relation of the width to the length of the head

R. B. Thomson (1913), " The Vertebral Column of the Bushman Race," Trans. R. Soc.
S. Afr., Hi, 365-378 ; G. D. Laing and H. S. Gear (1929). " The Strandlooper Skulls found
at Zitzikama/'S. Afr. J. Set., xxvi, 575-601.

L. H. Wells (1931), " The foot of the South African Native," Amer. J. Pbys. Anthrop.,

M. R. Drennan (1933)* A Short Court* on Physical Anthropology (Capetown : Mercantile
Press) ; ]. Drury and M. R. Drennan (1926-7), " The Pudendal Parts of the South African
Bush Race," Ml. J. S. Afr., xxii, 113-117.

RACIAL ORIGINS II

and expressing it mathematically. It is a cardinal faft that the Negro
shares with the Bush, Brown, and Nordic races the possession of
a long-headed (dolichocephalic) skull, i.e. one in which the breadth
is actually or even less than three-quarters of its length. This is
the primitive form of the human skull. It is found in the man-like
ape Australopithecus (South Africa), and in all known forms of
primitive mankind, such as Pithecanthropus (Java), Sinanthropus
(China), Eoanthropus (England), Neanderthal Man (Southern Europe
and Palestine), Rhodesian Man (Africa), Boskop Man (Africa), and
the living Australoid peoples (Veddas of Ceylon and Australian
aborigines). In this human ocean of dolichocephaly, stretching from
the Cape of Good Hope to the Behring Straits and from Norway
to Australia, there sprang up like an erratic current in Asia the broad-
headed (brachycephalic) type of mankind. This particular human
stream is represented to-day by two main swirls : the Armenoid Race,
which flooded from Turkestan across Mesopotamia and Egypt and
through the Mediterranean area and across Russia and Central Europe
to the British Isles (circa 300 B.C. onwards), and the Mongol Race,
which percolated into the East Indies and Indian Ocean littoral and
rushed like a torrent through the two Americas from Alaska to
Tierra-del-Fuego at an unknown period. 1

A preliminary application of these generalizations, rendered possible
by the collective labours of anthropologists during the last century,
will help to clear the ground in our attempt to interpret the physical
characteristics of the Bantu South of the Zambesi. In any con-
siderable collection of skulls from all available tribes, 2 there occurs
spasmodically, in about i per cent, of the number, a thoroughbred
brachycephal or two Asiatic fish out of their home water. They
appear also in any considerable group of Bush skulls. 3 This Asiatic
influence is not confined to the Bush and Bantu of South Africa.
Amongst 475 Pygmies in Central Africa Schebesta and Lebzelter
found 26 per cent brachycephalic. 4 Figures published by Leys and
Joyce 6 demonstrate fluctuating proportions of the same element in
the various tribes of Kenya, Uganda, Tanganyika, and the Congo.

1 G. Elliot Smith (1934), Human History (London : Cape).

* Such as that in the Anatomy Department of the University of the Witwatersrand.

* J. H. Gear (1929), " Cranial Form in the Native Races of South Africa," S. Afr. J. Set.,
xxvi, 684-697.

4 P. Schebesta and V. Lebzelter (1933), Anthropology of the Central African Pygmies in tnt
Belgian Congo (Prague : Czech Academy of Sciences and Arts).

* N. M. Leys and T. A. Joyce (1913), " Notes on a series of Physical Measurements from
East Africa," /. R. Anthrop. Inst., xliii, 195-267.

12 RAYMOND A. DART

Amongst the Masai, they appear as infrequently as amongst the
South African Bantu; but amongst the Kachamega, Kavirondo,
Kikuyu, Kamba, and Swahili, they account for 10 per cent and upwards
of the population. More than a quarter of the Lamu natives on
the coast and of the Nyema in Belgian Congo, just west of Lake
Tanganyika, are brachycephalic. The Comoro Islanders are so
infiltrated with Asiatic influence that over half their number are
brachycephalic ; they thus resemble the Arabs of Yemen and behind
Aden (Sheker tribe of Bedawi) and the coastal population of Biloch
and Cutch in India more than they resemble their Bantu brethren on
the mainland in cranial characteristics.

Similar nests of brachycephafy stretch across the Congo. Fifteen
per cent of the Tetela near the headwaters of the Lomani tributary
to the Congo River share this feature, and even on the Atlantic
bea-board along the Gaboon Coast near Fernand Vaz more than
5 per cent are to be discovered 1 ; Keith found nearly 35 per cent
amongst the Bushongo, Soko, Sango, and miscellaneous tribes of
Central and Eastern Congo (measured by Torday), and nearly
7 per cent amongst the tribes in the south-eastern (Oban) district
of Nigeria (measured by Talbot) 2 ; about 4 per cent Brachycephals
occur too amongst the Hausa of Nigeria. 3 These fluctuating figures
show how the brachycephalic Asiatic aliens were scattered in seemingly
haphazard fashion across the Continent. More detailed observation
will probably show that the isolated centres where they occur most
frequently represent links in the chain of ancient trans-continental
trade-routes, which supplied the Orient with marketable wealth from
" the land of the Zing ". Onward from the days when the old
Dutch colonists dete&ed the Mongolian appearance of the Eastern
Bush peoples and called them Chinese Hottentots, many writers have
encountered Mongoloid features in the Bush and Pygmy peoples
and in the Negroes of South Africa, the Sudan, and the West Coast. 4
These provide no ethnographical difficulty, for Malayan canoes have
been dredged out of the sea as far south as Port Elizabeth (Algoa
Bay), and the Indonesian type of outrigger canoe is common at
Mozambique, Port Amelia, the Comoro Islands, and other places on

1 R. C. Benington (1911-11), "A Study of the Negro Skull with special reference to
the Congo and Gaboon Crania," Biomttrika y viii, 292-339.

1 A. Keith (1911), "On Certain Physical Chapters of the Negroes of the Congo Free
State and Nigeria," /. R. Anthrop. 7/wr., xli, 40-71.

* A. N. Tremearne (1911), " Notes on some Nigerian Tribal Marks/' /. R. Anthrop. /**;.,
xli, 162-178.

* C. G. and B. Z. Seligman (1932), Pagan Triots of the Nilotic Sudan (London, Routledge).

RACIAL ORIGINS 13

the coast. The Malayo-Polynesian infiltration of Madagascar ethnically
is well-established.

Idris (twelfth century) records the traffic of Orientals thither and
to Sofala, probably from Java and Cambodia. An Arab ambassador
took a Negro slave to the Chinese court in the Sung dynasty (976 A.D.)
from which time onwards Chinese familiarity with African increased.
Masudi (tenth century) commented upon the Chinese African trade,
which was so great in Alberuni's time (circa 1030 A.D.) as to be affecting
the size of the ports in India (e.g. Somanath in the Gujerat Peninsula).
This African trade for ivory, rhinoceros horn, gold, ambergris, tortoise-
shell, animal skins, amber, sandal-wood, and particularly slaves was
never forgotten by the Chinese. It was re-established during the
succeeding Mongolian (1280-1360) and Ming (1358-1644) dynasties
as far south as Madagascar, thus extending over" six centuries. The
Oriental commerce was responsible for spreading the Negro Race
through the East Indies and into Melanesia in the same way as the
slave-trade also dispersed the Negro Race through the West Indies
and America. It made the Bantu language Swahili the lingua franca
of the Indian Ocean.

The Malayan and Chinese people merely followed in the wake
of the Hindus who " not only made trade settlements on the coast,
dating from the seventh century B.C. but apparently penetrated inland
towards the Great Lakes "- 1 The sacred Sanscrit poems or Puranas
describe the great Krishna (Nile) River, running, through Cush-
Ndipa the great lake in Chandristhan, the country of the Mpon
geographical information which gave the correft position in relation to
Zanzibar Island, and was used by Speke in his quest for the sources
of the Nile. The Indians knew Babylon as Baberu, and there is a
story dating from the third century B.C. about Indian mariners there.
After his invasion of India (326 B.C.), Alexander the Great confidently
and successfully sent a large portion of his Macedonian army back
to Mesopotamia across the Indian Ocean in a fleet under his general
Nearchus. Nebuchadnezzar used beams of Indian teak in the temple
of the Moon at Ur and in his own palace. Even before the Chaldean
Empire period the Oriental shipping trade was so vital to the Assyrians
that Sennacherib (circa 694 B.C.) recorded monumentally his destroying
the nests of pirates in the Persian Gulf and his compelling them to
settle at Gerrha. 2 It was these Assyrians who also compelled the

1 Ingrains, op. cit.

F& de L. . O'Leary (1927), Arabia before Muhammad (London, Kegal Paul).

14 RAYMOND A. DART

Sabaeans to pay tribute. The widely diffused interest in, and intimate
contaft with, East Africa at that remote period is shown not only
by Necho's circumnavigation of Africa and by the accurate geographical
knowledge contained in the sacred books of the Hindus, but also
by Darius the Persian monarch's (512-510 B.C.) sending thfe Greek
mercenary Skylax to explore the route down the Indus across the
Indian Ocean and up the Red Sea to Arsinoe near the modern Suez
Cana!, and his execution of Sataspes for failing to circumnavigate
Africa. These further historical fafts explain the maritime influences
which caused so many brachycephalic Asiatics to become dispersed
through the African Continent ; and reveal why, amongst the Ambo
and Herero tribes in South- West Africa, individuals are found to-day
indistinguishable facially from Indians and Persians.

The Asiatic stigmata in the South African native population are
not confined to the Mongoloid traits found in the Hottentots and
other Bush peoples and occurring amongst the " Nyambaan ",
Shangana-Tonga, Xhosa, Sotho, and other groups of Bantu and
most frequently of all amongst the Zulu of the Natal coast (vide
Table VI). Armenoid features are more widely dispersed than
Mongoloid. They are common amongst the Zulu, but even more
pronounced in the Venda of North-Eastern Transvaal, the Shona
tribes of Southern Rhodesia and the Rotse of the Zambesi head-
waters, as well as amongst the Herero and Ambo ; they are expressed
in the s>called " Semitic" traits* of hooked noses and hirsute faces
and bodies. These physical traits, referred to loosely and popularly
as " Semitic " or " Arabic ", are derived not from the primitive
Brown indigenes of the Mesopotamian and Arabian area, but from
the Armenoids, who from their ancient homeland of Turkestan
deluged the " fertile crescent " in successive viftorious waves, from
Dynastic Egyptian times onwards. This formed the avenue of their
approach, by land and sea, to Southern Africa during the historical
period.

THE RACIAL BASIS OF THE BANTU : A BUSH AND NEGRO MATRIX

Retzius's anthropological weapon, the cephalic index, enables us
to separate the Asiatic brachymorphs from the African (and European
and Asiatic) dolichomorphs. As the three African human stirps
(Negro, Bush, and Brown) are long-headed, we need further mathe-
matical weapons to isolate them from one another. Confronting

RACIAL ORIGINS 15

similar problems in the analysis of the Brown (or Mediterranean)
Race, the veteran Italian anthropologist G. Sergi discarded mathe-
matics and classified skulls according to their shape (pentagonoid,
ovoid, ellipsoid, etc.), as viewed from above. Finding these varied
types of skulls recurring constantly throughout Northern Africa and
Southern Europe, Sergi concluded that they were all relatively insignifi-
cant variants of one human racial type, which he called the Mediter-
ranean Race. 1 The conclusion that one predominant race characterized
the area proved correct ; indeed, Elliot Smith showed that this race
extended beyond Africa and Southern Europe into Arabia, Persia,
India, and South-Eastern Asia. 2 Sergi's guiding hypothesis, that
different forms of skull belong to one homogeneous race, is erroneous ;
but it called attention to a much-needed study of head form which
is critical for unravelling the anthropological tangle in Africa as well
as in Europe.

Another shortcoming in Sergi's system (as in that of Retzius)
was that by classifying skulls according to length and breadth alone
it neglected the factor of height. This deficiency in the system has
been neatly overcome by Sergi's fellow-countryman, Frassetto. 8 He
has shown that the dolichomorph and brachymorph groups can each
be divided into three main head-form sub-groups; according as
the parietal bone (which chiefly is responsible for the shape of the
skull) happens to be foetal, infantile, or adult in form. He has thus
reduced Sergi's innumerable categories of skull form to six, based
on biological considerations of bone growth. These six sub-groups
can then be split up according to height factors. They are classified
as hypsicephalic, orthocephalic and chamaecephalic ; according as
their cranial vaults are high, moderately elevated, or low respectively.
This system provides an apparatus for separating human skulls into
eighteen different categories ; it is a mechanism of sufficient isolating
capacity to be of great assistance in African anthropology.

Tables I and III demonstrate ]. H. Gear's application in 1929 of
Sergi's system (as modified by Frassetto) in examining collections
of Bush and Bantu skulls respectively. 4 Previously no very clear
idea existed as to what was and what was not a Bush skull. Even
to-day most diverse and contradictory statements are current in the

1 G. Sergi (1911), L'Uomo (Turin, Bocca).
1 G. Elliot Smith (1934), op. cit.

* F. Frassetto (1909-1918), Lt^ionl antropologia (Bologna : Coop-Tipografica Libraria
giani).
H. Gear, op. cit.

RAYMOND A. DART

literature. Some writers affirm that the Bushmen and Pygmies are
Negroes ; others separate all three from one another and proceed
even further to manufa&ure Hottentot, Strandlooper, and Korana,
and other branches of the African human stock; alternatively,
suggestions have been put forward that Hottentots are hybrids
between the Bush and the Bantu.

Australoid and Boskopoid types do crop up in the living population,
especially amongst the Bush peoples in the south-western ethnical
territory of Africa. They may prove one day to be more prevalent
than has been recognized hitherto ; but they are so few numerically
and so sporadic in occurrence as to constitute anthropological
curiosities. They are atavistic links with Africa's past, heirlooms of

TABLE I
BUSH GROUP DISPERSAL (AFTER GEAR)

DOLICHOMORPHS

BRACHYMORPHS

Penta-
gonoid

Ovoid

Ellip-
soid

Totals

Eury-
penta-
gonoid

Sphe-
noid

Sphse-
roid

Totals

Perccn
tage

Hypsi- .

Ortho- .
Chamae- .

2
19

2
20
*9

3'9
39-2

5<V9

Totals

100-0

Percentage

68-6

*5'S

3'9

2'O

o-o

lOO'O

antiquity preserved in the southern recess of the continent. They
do not belong to the Brown or Negro races, and they do not represent
the fundamental type to which the greater number of individuals
in all branches of the Bush family belong.

In his 51 skulls (comprising 26 Bushmen, 9 Hottentots, 7 Koranas,
5 Strandloopers, and 4 Griquas) Gear found nearly 70 per cent penta-
gonoid in form. Of the twenty-six known Bushmen, 70 per cent
belonged to this pentagonoid group. Further, 57 per cent of the
skulls were low-vaulted (chamaecephalic), and, of these, 76 per cent
were pentagonoid. He therefore concluded that the fundamental
type of skull, common to all Bush peoples of South Africa, is the
chamapentagonoid. In keeping with its baby-like form, its dolicho-
cephaly and chamaepentagonoidy the skull has a small endocranial
capacity, i.e. it is microcephalic.

RACIAL ORIGINS 17

The most highly populated compartment in Table I is that occupied
by these twenty-two chamapentagonoids, which represent the Bush
race in its greatest purity. The low vault, typical of the Bushman,
appears in addition in seven other skulls (six ovoids and one ellipsoid).
The pentagonoid form occurs also in thirteen other skulls, whose
vaults are of greater height than is typical in the Bush race. We will
consider these (and the thirteen ovoids and two ellipsoids) at a later
stage; and merely note, in passing, the single eurypentagonoid
a brachymorphic, Asiatic intruder.

Recently Dr. Galloway and Mr. Wells have gathered together
the information from every available published source where
measurements have been taken of Hottentot, Korana, and Strand-
looper skulls. Clumped together these 239 individuals are distributed
as follows :

TABLE II
DISTRIBUTION OF ALL REPORTED S.A. BUSH PEOPLES COMBINED

Dolichocephalic

Mesaticeph.

Brachyceph.

Totals

Percentage

Hypsicephalic .
Ortho- .
Chamae -

f
61
68

8
50
3

U
118

107

5'8
49-4
44-8

Totals .

*39

lOO'O

Percentage

56-1

40-2

3'7

lOO'O

The relative percentages of the different types of skulls are somewhat
modified by the inclusion of all the variants occurring in this " Bush-
Hottentot" series (such as the Boskopoid and Australoid types),
but the general deduction to be drawn is that further analysis only
corroborates the general truth of Gear's original postulate and demon-
strates that all physical variations from the fcetalized, dolichocephalic,
chamaepentagonoid skull in the Bush peoples are to be regarded as
due to hybridization with other racial types.

Qear also analysed (vide Table III) 94 Bantu skulls (comprising
15 Zulu, 15 Sotho, 5 Xhosa, 3 Fingo, 3 Kalanga, 2 Tswana, 2 " Nyam-
baan ", 2 Mpondo, 2 Ambo, and one each of the Swazi, Rotse, Herero,
Rolong, Hlubi, Chopi, " Mozambique ", Thembu, Ndebele, and Shona
peoples, and twenty-six others of unknown tribe but probably
Transvaal Sotho). In this unselected polyglot he found one foreign

RAYMOND A. DART

(Sphaeroid) Asiatic brachymorph. There were also sixteen Penta-
gonoids, six of which were Chamaepentagonoids indistinguishable
from the typical Bushmen just described. The strong Bush influence
in the Bantu group was further exhibited by twenty-two other skulls
having a low vault (28*7 per cent in all). The best-populated single
compartment in the Bantu group is that occupied by the twenty-five
Ortko-ovoids, which represent the Negro element in its greatest purity.
They aftually account for less than 27 per cent of the group. On
the other hand 45*7 per cent show the ovoid form of skull, which
is the most frequent of the three dolichomorph types of head-form
present in the Bantu. The Southern Bantu are therefore only about
one-third pure-blooded Negro and are more than one-quarter pure-
blooded Bush in constitution (vide also Tables V and Vt).

TABLE III
BANTU GROUP DISPERSAL (AFTER GEAR)

DOLICHOMORPHS

BRACHYMORPHS

Percen-
tage

Penta-
gonoid

Ovoid

Ellip-
soid

Totals

Eury-
penta-
gonoid

Sphe-
noid

Spha-
roid

Total

Ortho-
Chamae-

o
10
6

8
5

4
19
II

54
*7

o
55
7

12-8
58-5
28-7

Totals

lOO'O

Percentage

17-0

45-7

36-1

o-o

0*0

i*i

100*0

To discover the other racial elements that have gone to the manu-
fafture of the Bantu we require to travel farther afield than the Zambesi
River ; but the nature of our problem will be greatly simplified if
we consider these two Tables (I and III) a little farther. Both the
Bush and the Bantu groups are admittedly very mixed, but the Bush
group is more homogeneous (or racially purer) than the Bantu group.
If we consider the feature hypsicephaly, for example, the Bush group
displays only two skulls (4-0 per cent), while the Bantu reveals twelve
(12-8 per cent) of this elevated type. These two high-vaulted skulls
in the Bush group are Ovoids; and, as the preponderant number
of hypsicephalic skulls in the Bantu group are Ovoids, it is apparent

RACIAL ORIGINS 19

a fortiori that the hypsicephaly found amongst Bush peoples has
been handed on to them from the Bantu.

Similarly only two (3-9 per cent) of the skulls in the Bush group
present ellipsoidy, while amongst the Bantu no less than thirty-four
(36-2 per cent) of the skulls possess this feature. We are entitled
to assume that the ellipsoidal impurity amongst the Bush peoples
was a second gift handed on to them through the Bantu.

On the other hand, if we study the phenomenon of orthocephaly
amongst the two principal types of skulls (Pentagonoids and Ovoids)
common to the Bush and Bantu groups a contrary state of affairs
exists. In the Bush group there are virtually three Ortho-pentagonoids
to every one Ortho-ovoid ; in the Bantu there are only two Ortho-
pentagonoids to every five Ortho-ovoids. In other words, while
the Bantu were transmitting their hypsi-ellipsoid impurities into the
Bush race, the latter was responding by enriching the Bantu agglomerate
with onho-pentagonoid impurities which the Bush already possessed
in high degree.

RACIAL MINGLINGS OF BUSH AND BANTU :
THE CONTRIBUTION OF THE BROWN RACE (HAMITFS)

If we can ascertain how the Bantu became infiltrated with hypsi-
ellipsoidal and the Bush with ortho-pentagonoidal impurities, we will
have gone far towards understanding the curious distribution of
cranial types in South Africa and the mechanism whereby it was
produced.

In this quest we are materially assisted by a splendid study of
105 Abyssinian skulls from the Tigre Province, carried out by Sergio
Sergi. 1 He used his famous father's method in such a manner that
we are able to reclassify them in Frassetto's fashion and to compare
direftly the racial constitution of these Abyssinians near the Red Sea
in the north-east of Africa with our Bantu and Bush peoples south
of the Zambesi (vide Table IV).

Various fafts immediately leap to light. In the first place the
Abyssinians are an even more mixed group of people than the Bantu.
They embrace more Asiatics (three Eurypentagonoids and three
Sphenoids) ; and history is corroborated that Abyssinians and other
related North Africans simultaneously received in greater amount

1 S. Sergi (1912), Crania haoessenica. Contribute all' antropologia dell' Africa oriental* (Rome,
Loescher).

RAYMOND A. DART

or a&ually assisted in mediating the Asiatic influences, which are
more sparsely represented amongst the Bush and Bantu races farther
south. But a more remarkable faft is that the most densely populated
compartment in the Abyssinian group is formed by the thirty-one
(Negro) Ortho-ovoids ; they account for an even higher proportion
(i.e. virtually 30 per cent) of the Abyssinians than of the Bantu. It
is not surprising that Sergi also found that 30 per cent of the Tigre
skulls were prognathous like the typical Negro. While the ovoid
quality only affects 36-2 per cent of the skulls (instead of 45-7 per
cent as amongst the Bantu) the Abyssinians are virtually as pure-
blooded Negro as the South African Bantu, and far more considerably
impregnated with Negro blood than are the Bush peoples of South
Africa.

TABLE IV
ABYSSINIAN (TICRE) GROUP DISPERSAL (AFTFR SERGI AND FRASSETTO)

DOLICHOMORPHS

BRACHYMORPHS

Penta-
gonoid

Ovoid

Ellip-
soid

Totals

Eury-
penta-
gonoid

Sphe-
noid

Sphae-
roid

Totals

I ercen-
tage

Hypsi- . .
Ortho- .
Chamae- . .

4
31
3

M
68
16

3
o

2
I
O

16-2
68-6
15-2

Totals .
Percentage

37
35-2

3
36-2

22-8

3
*'.9

3
2-9

o-o

105
100-0

lOO'O

Another impressive feature about this analysis is that although
the Abyssinians show only about half as much chamaecephaly (15*2 per
cent)' as the Bantu (28*7 per cent) and about a quarter of that found
amongst the Bush (57-0 per cent), it is none the less present ; moreover,
50 per cent of the chamaecephalic individuals are Chamse-pentagonoids
like the typical Bushman. Hence the Tigri Abyssinians include relatively
more pure Chamce-pentagonoids of thoroughbred Bush type than do the
undetected Bantu south of the Zambesi. Both the Bush and the Bantu
are flat-nosed (platyrrhine), which accounts for the fact that Sergi
found an even higher percentage (46 per cent) of platyrrhiny than
of prognathism amongst these Abyssinians.

This Bush element in Abyssinia assists our understanding of the

RACIAL ORIGINS 21

physical make-up of the Bantu. The whole of the territory from
Abyssinia to the Zambesi abounds in Bushman relics in the form
of prehistoric rock-paintings and engravings ; it has yielded Bush
remains l ; it harbours to-day at least two tribes (Hadza and Sanda),
who preserve Bush physical features and speak languages belonging
to the Bush-Hottentot linguistic family. 2 Tanganyika Bantu show
22 per cent, or nearly as much, chamaecephaly as do the Bantu south
of the Zambesi. 3 More than a quarter of a large series of skulls from
south-east Kenya were typical dolichochamae-cephalic Bush skulls
and hence the " Hottentot " affinities of the group are strongly
evident. 4 Anatomical, linguistic, and archaeological fafts here all unite
in demonstrating that the territory of the long-headed, low-vaulted,
pentagonal^ microcephalic Bush race extended at one time from the Cape
of Good Hope to the Red Sea.

That this state of affairs preceded the Negro intrusion into the
eastern half of the continent is corroborated by our tables. While
the Abyssinians are more strongly " negrolized " if we may coin
that term than the Bush peoples, they are more closely akin to
our Bush than to our Bantu group in the relatively large number of
Ortho-pentagonoids present. Indeed, the most outstanding point of
contrast between the Southern Bantu and the Abyssinians (vide
Tables II and III) is the sudden relative increase in the number of
Pentagonoids and more particularly of the Ortho-pentagonoids in
Abyssinia. The Ortho-pentagonoids account for about a fifth of the
Tigre population ; they constitute over a quarter of our Bush group.
This type of skull, which occurs nearly twice as frequently amongst
the Abyssinians and more than twice as frequently amongst the
Bush as amongst the Bantu, is characteristic of the Brown (Mediter-
ranean or Hamitic) Race.

The racial features of the Brown Race have already been outlined.
They were the indigenes of the Fertile Crescent, formed by the
Nile Valley, Palestine, and Mesopotamia ; in their midst civilization
emerged. Owing to the bone-preserving qualities of the hot, dry,
Egyptian sands and the Egyptians 9 profound interest in preserving

1 A. Galloway (1933), " The Nebarara Skull," S. Afr. J. Sci., xxx, 585-596.

* D. F. Bleek (1931), " Traces of Former Bushmen Occupation in Tanganyika Territory,"
S. Afr. /. Sci., xxviii, 423-9 ; L. F. Maingard (1934), " The Linguistic Approach to South
African Prehistory and Ethnology," S. Afr. J. Set., xxxi, 117-143-

* v. Schweinitz and R. Virchow (1893), " Kopfmessungen an Ost-Afrikanern," Z. Ethn.,

XXV,

(1931), " A Study of the Negro Skull, with special reference to the Crania from
Kenya Colony," Biomttrika, xxiii, 271-314.

22 RAYMOND A. DART

the dead by mummification, so many thousands of Egyptian skeletons
have been found and measured, that we can follow racial movements
in Egypt from before the dawn of history and study the effefts of
those movements in modifying the skeletal features of her inhabitants
with more exaftitude and over a longer period than in any other
part of the world.

Despite these movements the vast majority of the Egyptians, at
all periods in their eventful history, are dolichocephalic and ortho-
cephalic. These long-heads of moderate cranial height are the typical
Brown people, and they are relatively far more numerous in Egypt
than in Abyssinia. In very ancient Egypt they accounted for approxi-
mately 70 per cent of the population but even in those days 12-20 per
cent of the Egyptians were chamaecephalic like the Bushmen. Miss
Smith has described nine adult " Pygmy " skulls amongst a collection
of Egyptian crania of the Third Dynasty, 1 which testify to the con-
siderable amount of the little people present there in a relatively
pure state 5,000 years ago.

These fafts colleftively demonstrate the intimacy of the fusion
between the Brown and Bush races east of the Great Rift Valley in
Africa prior to the inroads of the Negro upon the eastern sea-board.
The former prevalence and continuous distribution of the Bush Race
over the eastern half of Africa explains their accessibility to the Ancient
Egyptians, and the deep, recurrent interest of the ancient civilized
world in the dwarf people ; it shows why the Bush-Hottentot languages
are so intimately related to the Hamitic group of languages ; it
explains why the Bushmen of South Africa and the Pygmies of
Central Africa and of Southern Asia, despite alien racial admixture,
retain a virtually identical width of skull base 2 ; it helps us to under-
stand how and why by land and sea these primitives of nature could
have become scattered from the Equatorial forests of Africa to the
recesses of Papua and it assists in elucidating the fantastic part these
abbreviated editions of humanity have played in the mythology and
folk-lore of all civilized and semi-civilized peoples.

SUBDIVISION OF THE SOUTHERN BANTU

At this point it is appropriate to remark that much detailed investiga-
tion of the tribes will have to be carried out before we can know the

1 H. D. Smith (1911-12), "A Study of Pygmy Crania based on Skulls found in Egypt/'
Biomttrika, viii, 262-6.
* Schebesta and Lebzelter, op. tit.

RACIAL ORIGINS 23

proximity of physical relationship between tribe and tribe and the
order in which the various tribes entered the territory. Historically
we have little to help us and the position can only be remedied by
the slow process of correlating the fafts of language, tradition, and
culture with physical strufture. The information in the former three,
being more voluminous and accessible, naturally constitutes our
preliminary basis of classification. Linguistically the Southern Bantu
can be subdivided into three main zones :

1. Western, spreading north and south of the Okavango River
and including the Ambo, Herero, and Mbundu (as far north as
Benguella on the Angola Coast) ;

2. South-Central, comprising the Shona tribes of Southern
Rhodesia ;

3. South-Eastern, situated principally in the Union of South
Africa and Portuguese East Africa but also extending into Bechuana-
land and the headwaters of the Zambesi. It includes :

(a) the Nguni peoples (such as the Xhosa, Zulu, and Swazi) on the
south-eastern sea-board, all of whose dialefb are affefted by Bush
clicks ;

(K) the Shangana-Tonga (Thonga) peoples of Portuguese East
Africa, who have no clicks ;

(c) the Sotho peoples, such as the Southern Sotho of Basutoland,
Northern Sotho of the Transvaal, and Tswana of Bechuanaland ; and

(d) the Venda of North-Eastern Transvaal.

It might be anticipated that the Nguni peoples, who have absorbed
Bush " clicks " into their dialets, would also betray more Bush
physical features than do other tribes ; this, in point of faft, is true.
An interesting investigation of certain tribes in the south-eastern
zone by the blood-grouping method has been carried out by Elsdon-
Dew. He finds the blood-group differences so significant that he
can locate three different tribal groups in this zone alone. The first
is formed by the Xhosa and Mpondo, who are the most nearly related
to the "Hottentot"; the second, constituted by the Shangana,
Zulu, Southern Sotho, and Tswana, is intermediate in type between'
the first and the third racial group, which is composed of the Northern
Sotho, the Swazi, " Nyambaan " and Chopi (of Portuguese East
Africa). Of these three groups the last-named, situated most to the
north, approximates most nearly to the postulated "primitive"
type of human blood-grouping. All three groups are more " primitive "

24 RAYMOND A. DART

in their chara&ers than the Bush or the " Hottentot " groups examined
by Elsdon-Dew. These data corroborate our view that the Negro
forms an early off-shoot from the Sapient human stock ; they accord
also with the idea that the first Bantu intruders into South Africa

TABLE v

Vault Characters in

Percentages

CHAMAE-

Below
68.

ORTHO-.

68-73

HVPSI-.

Above
73

No. in
Series

Tribal Name*

Grouping of tribes by cross compari-
parison of cephalic characters

9'5

66-7

23-8

SHANGANA-TONGA

Group A High dolichocephaly
S High hypsicephaly

12-5

50-0

37' J

Bhaca.

20*0

70-0

IO'O

Hlubi ^v^v S

24-0

44-0

32-0

TRANSVAALJX^

SOTHOM ^>

^^^

M-.

fi-8

24-1

S \

-GrouJ> B Low dolichocephaly
+*~^^ High hypsicephaly

26-5

59-8

13-7

1 02

Mpondo^ <p ^--''"'y^

26-7

46-6

26-7

zuiu--"^r \

63-6

9. t

f"?v/tim C* /C<mili ^% A Kut- u!t

\V1 -y increasing Bush admix-

28-0

62-0 .

10-0

"NYAMBAAN^l / tUTc).

35-9

40*0

59-2
60*0

4'9
0*0

103

5birrA5ocAA \

t'* w roupU (otmilar to if but witti
increasing Bush admix-

44- i

48-1

7'7

104

A^a-^/\

ture).

44'4

50-9

4'7

106

^art^ana-/-

roup (Similar to A in dolicho-

48-6

42-6

8-8

TSWANA^

cephaly but differing from
both A and B in loss of

hypsicephaly probably
due -to profound Bush

admixture).

33'9

54-7

n-4

760

Total

'CAPITALS = over 90 per cent, dolichocephalic.
Italics = over 80 per cent dolichocephalic.

Lower case = 70-76 per cent dolichocephalic.

presumably found farthest south to-day were exposed for a
longer time to the effects of admixture with the hamiticized Bush race.
We can confidently expeft that with the extension of such observations
to other branches of the south-eastern zone, and especially to tribes

RACIAL ORIGINS 25

in the south-central and western linguistic zones, very valuable
light will be thrown on the more detailed physical inter-relations of
these highly mingled peoples.

Statistics concerning the cranial chara&ers of over 750 Southern
Bantu have been collected by de Saxe, in the course of an investigation
which is under way. His objeft is to analyse the somatometric
chara&ers presented, by comparing at least a hundred living adult
males of each tribe. He has kindly allowed us to assemble the incom-
pleted data in a tentative tribal grouping (vide Table V). In every
tribe hitherto examined 70 per cent or upwards of the population is
dolichocephalic. The tribes have been arranged on the table in a
gradually increasing order of Bush admixture : as shown by the
ascending chamaecephaly and descending orthocephaly. This arrange-
ment assists in demonstrating that some tribes resemble one another
in cranial features more closely than do others ; and thus makes it
possible to group them. The grouping suggested is not final because,
until adequate numbers have been collefted for all tribes; the groups
are of unequal classificatory value. Moreover the colleftion of data
from other tribes outside the south-eastern zone, to which these
belong, may seriously modify any general opinion now expressed.
At the same time it is of general interest that, already, five distinft
groups appear to be emerging in the south-eastern zone and that
they not only confirm but expand Elsdon-Dew's triple grouping,
though in some points there is slight disagreement.

Thus Elsdon-Dew's affiliation of the Shangana-Tonga with the
Northern Sotho (in Group A) and close dolichocephalic relationship
of these to the Chopi and " Nyambaan " (Group C) is corroborated ;
but our few Swazi appear more akin to the Zulu and Bhaca (Group B),
while the Tswana are more akin to Group C (and A) than to the
Zulu. The Mpondo approximate the Hlubi and Fingo (Group D)
rather than with the Xhosa, Southern Sotho, and Shangana (Group E) ;
but Group D is related to Group B rather than to Group A, while
Group E stands apart. The extension of such cross-analyses is greatly
needed, if we are to discover the more detailed physical relationships
of the tribal groups.

Little detail has been gathered concerning the tribes in the south-
central zone of Southern Rhodesia, and we are in little better case
in respeft of the western zone of South- West Africa. But an analysis
of fifty-four skulls recently made available to us by the kindness of
Dr. L. Gill, Curator of the South African Museum in Capetown,

26 RAYMOND A. DART

has furnished data about forty-eight Ambo and six Herero. They
too are over 90 per cent dolichocephalic; and, taken together,
are 27-8 per cent chamaecephalic, 57-4 per cent orthocephalic, and
14*8 per cent hypsicephalic. They therefore approximate very closely
in physical characteristics the Chopi and " Nyambaan " (Group C
of our table).

This cross-analysis of cranial characteristics has been introduced
to illustrate an avenue of approach to physical anthropological studies
capable of indefinite extension and thitherto relatively unexploited.
Our next table (VI) represents the results of a distributional study
of facial features amongst 977 natives of various tribes, including
the foregoing 760 adult mine natives and a series of 217 facial masks
taken in the field by Professor Cipriani, of Florence, or by various
members of the Anatomy Department in the University of the Wit-
watersrand over the last decade. The analysis was performed by Messrs.
Wells and de Saxe. It provides further information concerning our
provisional groups. The tribes are arranged in order of increasing
percentages of Bush features ; the only serious reversal in order, as
compared with the previous table, occurs in respeft of the Shangana,
who, despite their chamaecephaly, only display 17-1 per cent of
people with Bushman fades. The Zulu, " Nyambaan ", and Sotho
stand out with their relatively high percentage of Mongoloids. The
unsatisfactory term " Caucasoid " has been used in this table to
embrace all facial types other than the Negro, Bush, and Mongoloid ;
the preponderant number of these present the facies characteristic
of the Brown race but there is a considerable percentage of Armenoid
appearance, and even Nordic types are scattered through the groups.
In all the large groups, save the strongly Bushmanized Tswana, die
" Caucasoid " elements tend to account for nearly one-quarter of
the population. Throughout the series the genuine Negro facies
only accounts for 35-63 per cent, or an average of only half of the
population.

In view of these fafts it is not surprising that the mean stature
measurements recorded by different observers of large series of
natives is approximately 5 ft. 6 in. for males and 5 ft. i J in. for females.
Most male measurements fall between 5 ft. 5 in. and 5 ft. 8 in., the
least measurement being 4 ft. 8 in. and the greatest 6 ft. 4} in. It
is obvious that, apart from a consideration of the racial type of
individuals, reliance upon stature as a racial indicator is bound to
prove sterile.

RACIAL ORIGINS 27

A similar range of variation from individual to individual has been
found by Miss Orford and de Saxe in skin coloration, the different
nuances of colour in our records of approximately 1,000 individuals
extending from Nos. 14-35 on von Luschan's scale. The shades
lighter than 23, however, are very rare ; three-fourths of the subjects
fall within the 26-30 range, the order of frequency being 28, 29,
27, 30, and 26 ; the most frequent shade, 28, accounts for one-fourth
of all individuals examined. Seiner found that the Bushmen ranged

TABLE VI

Distribution of facial features in percentages

Cauca-

Mongo-

No. of

Tribal Name*

Negro

Bush

SOld

loid

Cases

53'5

if'S

29-0

2'2

" NYAMBAAN "

54'7

17-1

26-5

1-7

117

Shangana

59-1

18-2

22-7

SHANGANA-TGNCA

43-2

19-3

22'O

5-5

109

Zulu

Sfo

20*0

25-0

Hlubi

60- 9

23-4

1 5-7

6 4

Swazi

n-o

24-2

22-8

CHOPI

ffi

26-0

17-3

1-6

i*7

Xhosa

63-2

26-3

lo-j

Bhaca

46.2

27-2

24-6

2'O

Sotko

44-8

27-6

105

Mpondo

35-7

28-6

35*7

Fingo

43'*

44-6

I2'2

~""

TSWANA

51-2

25-0

22-3

"'5

977

Total

'CAPITALS = over 90 per cent dolichocephalic.
Italics = over 80 per cent dolichocephalic.
Lower case = 70-76 per cent dolichocephalic.

from 1 6-1 8 ; nine-tenths of the subjefts falling between 22 and 25 ; the
most frequent shade, 23, charafterized one-third of all those examined.

As might be anticipated from these data, the tribes more infiltrated
with Bush admixture, such as the Tswana, Fingo, Bhaca, Mpondo,
and even the Sotho, Swazi, and Shangana, are on the whole lighter-
skinned (average 28) than the Xhosa, Zulu, and Chopi (average 29),
and particularly the " Nyambaan " and Shangana-Tonga (average 30).

Physical analysis of the Bantu is thus proceeding, but the data
which will accumulate during the coming decade may profoundly
modify our tentative scheme of grouping. All methods of grouping

28 RAYMOND A. DART

linguistic, cultural, or physical go to prove that the Southern
Bantu did not cross the Zambesi in a single migratory movement
but by a series of inroads separated by centuries. Whether there
were three, five, or more such major movements these records,
colleftively or co-ordinately read, will finally reveal.

THE CONTRIBUTION OF THE NORDIC RACE

We have seen how the ortho-pentagonoid Brown (Hamitic)
impurity in the Bush stock became transferred to the Bantu; it
now remains to discuss how hypsi-ellipsoid impurities impregnated
the Negro stock and became transferred through the Bantu to the
Bush peoples. If our conception be correct that the Nguni peoples,
who retain Bush " clicks " in their diale&s, were some of the earliest
Bantu intruders into Southern Africa, it becomes very interesting
that tribes in this group display a progressive loss of hypsicephaly.

Hypsicephaly is widespread throughout Negro Africa. It is clear
from the averages recorded by de Quatrefages and Hamy that approxi-
mately 25 per cent of the skulls recorded by them from the Congo,
Sudan, Senegambia, and Upper and Lower Guinea were hypsicephalic,
the remainder being chiefly orthocephalic. 1 The percentage varies
from distrift to distrift. If we go up the western half of the con-
tinent, we find the Tetela with 36-5 per cent of hypsicephaly, the
Loango Natives with 69 per cent, and the Gaboon Natives with
52 per cent. Chamaecephaly is encountered very infrequently (none
in 26 Loango; two in 92 Gaboon; three in 74 Tetela). 2 The
hypsicephaly occurs independently of brachycephaly, in fact it is
more frequent where there is least brachycephaly.

Obviously, the hypsicephaly of the Southern Bantu is only part
of a continent-wide phenomenon and the explanation of its presence
must be sufficiently wide to embrace its occurrence in the north as
well as the south. If we return to our early tables we will see that,
while there is slightly more hypsicephaly amongst the Abyssinians
(16-2 per cent) than amongst the Bantu south of the Zambesi (12*8 per
cent in 128 skulls, 14-4 per cent, in 750 living males), there is consider-
ably more ellipsoidy amongst the Bantu (36-2 per cent) than amongst
the Abyssinians (22*8 per cent). Moreover the greater degree of
Abyssinian hypsicephaly is largely accounted for by the presence

1 A. de Quatrefages and E. T. Hamy (1881), Crania ethnica. Let cranes Jet races kumtunes,
a vols. (Paris, BailHere).
* Cf. R. C. Benington, op. ait.

RACIAL ORIGINS 29

in the group of two hypsicephalic Brachycephals, such as are not
found in the Bantu group. It is obvious that the Abyssinians did
not have a sufficiently preponderating amount of hypsi-ellipsoidy
to have been responsible for communicating it to the Bantu ; it is
equally apparent that it could not have come from any other North
African group of peoples ; the Brown Race charafteristic of those
parts has only moderately elevated (orthocephalic) heads, such as
the Negro himself charafteristically possesses.

Hypsicephaly is alien to all three of the races predominating in the
three transcontinental anthropological territorial belts of Africa, and yet
it has percolated through the whole continent, but with decreasing
intensity as we pass from north to south. Only two hypsicephalic
(O voids) skulls were found in the Bush group (Table I) ; 12 (8 O voids
and 4 Ellipsoids) in the Bantu group ; 17 (9 Pentagonoids, 40 voids,
2 Ellipsoids, and 2 Sphenoids) in the Abyssinian group. Now one
definite possible source of hypsicephaly is Asia the short brachy-
cephalic skull must be elevated, if it is to contain a human brain.
We cannot doubt that the Asiatic source is responsible for the hypsi-
cephaly of brachycephalic skulls in Africa ; but nobody would have
the temerity to suggest that brachycephalic admixture could transform
long-headed skulls of moderate height into high-vaulted skulls without
affefting their length. Unless hypsi-ellipsoidy has appeared as a
spontaneous variation amongst the Negro, who passed it on to the
Abyssinian and the Bush peoples, we are led to the conclusion that
the hypsi-ellipsoidy came to the Bantu as to the Abyssinians from
some dolichomorphic people other than the Brown Race. The only
possible known source is the Nordic Race.

The so-called " Nilotes and Nilo-Hamites " along the Upper Nile
Valley and in Uganda were thought by Seligman to be distinguished
from adjacent Negroid tribes by virtue of their dolichocephaly. 1
Aftually amongst the seventy Nuer and Dinka measured by Dr. Pirrie,
72*9 per cent are dolicho-, 24-3 per cent mesati-, and 2-8 per cent
brachycephalic. In our Departmental series of 128 skulls taken at
random from the different tribes of Bantu in the Union of South
Africa 81-2 per cent are dolicho-, 18-0 per cent mesati-, and 0-8 per
cent brachycephalic. If dolichocephaly were the test, then the Southern
Bantu are more " Hamitic " than the " Nilotes and Nilo-Hamites ".
The proximity of relationship between the two groups is brought

1 C. G. Seligman (1913), "The Hamitic Problem in the Anglo-Egyptian Sudan," /. R.
Anthrop. Irut., xliii, {19-705.

30 RAYMOND A. DART

out by the similarity in the above figures ; it is emphasized by com-
paring the altitudinal (height) characters in the two groups. The
Nuer and Dinka are 13-0 per cent hypsi-, 53-6 per cent ortho-, and
34-4 per cent chamaecephalic, and our Southern Bantu collection
is 10-9 per cent hypsi-, 58-6 per cent ortho-, and 30^5 per cent
chamaecephalic. Comparing the Nuer and Dinka measurements
with the list of tribal measurements tabulated above we can see that
the nearest congeners of the Nilotes and Nilo-Hamites in South
Africa are the group formed by the Sotho (of Basutoland), Xhosa,
and Shangana, the chief distinction being that the more northerly
situated Nilotes and Nilo-Hamites are more hypsicephalic.

Various sources of evidence such as the series of seventy-two
African skulls analysed by G. Sergi 1 and the figures published by
Shrubsall for Uganda Natives, 2 etc. could be brought forward to
prove the statement that, in general, hypsicephaly increases as we
proceed from the south-east to the north-west of the African con-
tinent. It would lead us too far from the purposes of this chapter
to discuss how and when these cranial characters of hypsicephaly
and ellipsoidy became transferred from the Nordic Race to the Negro
Race. It is enough to point out that the various features discussed
are found amongst the Bantu and that by studying the relative pro-
portions in which they are distributed, we can conclude that the
Southern Bantu are more closely related to the Eastern Bantu and
to the Nilotes than they are to the Western Bantu. The Western
Bantu did not suffer Brown and Bush admixture to the same extent
as did the Eastern and Southern Bantu.

The dolichomorphic hypsicephaly, which is more pronounced
amongst the Western than the Eastern and Southern Bantu, has
apparently spread across Africa from the Sudan by two different
channels, one eastern (or Nilotic) from the Nile headwaters through
erstwhile " Bushman " Tanganyika and then radiating fan- wise across
the southern terrain from the Indian to the Atlantic Ocean, the other
western (or Nigerian) through the Equatorial forest and the Cameroons
to the uttermost parts of the Congo and Angola. There finally the
two streams converged.

These broad but fundamental features of racial distribution in
Africa are explained satisfactorily if we assume that the Sahara was
a more effective barrier against human mingling prior to the last

1 G. Sergi, L'Uomo, op. cit.

1 In H. H. Johnston (1902), The Uganda Protectorate, 2 vols. (London, Hutchinson).

RACIAL ORIGINS 31

Ice Age than it is to-day. It separated the Nordic from the Negro
Race in the west and the Nordic from the Brown Race in the east.
With the coming of the Ice Age climatic conditions far more con-
genial than those of recent times spread across that desert. The
adventurous Nordic people, scattered hitherto along the Mediterranean
littoral, migrated freely over the Sahara and intermingled to the east
with the north-westerly moving Brown Race and to the south with
the northerly expanding Negro Race. In the Sudan the Negro-Nordic
admixture produced the hypsicephalized negroid stock, whose mobile
history we have been reconstru&ing, whilst depifting the course of
their transcontinental pilgrimage into their present homeland and
its implications for the physical strufhire of the Bantu south of the
Zambesi. We look to the future labours of the growing army of
anthropologists throughout Africa for the gradual elucidation of the
details in this prehistoric epic.

The background of our pi&ure has been limned ; various high-
lights have been tentatively painted in ; their gaunt relief will be
subdued by the labours of those artists whose detailed work is essential
to the masterpieces of the future. Their books are works of the
time to come and will unquestionably appear when an army, equal
to that which has 'laboured in Europe over the last 400 years since
Vesalius, has been organized in Africa to colleft information about
the Bantu similar to that which has been garnered over these centuries
concerning Europeans.

CHAPTER II

HABITAT

By A. J. H. GOODWIN

GENERAL MORPHOLOGY

AFRICA south of the Zambesi consists of a great plateau, somewhat
higher on the eastern side, and falling gradually towards the west.
The height averages between three and four thousand feet, and the
plateau is markedly cut by the valleys of the three great South African
rivers, the Zambesi, the Limpopo, and the Orange, with its Molopo
affluents. Balancing these valleys are three regions which exceed
the 4,000 ft. level : the plateau country between the Zambesi and
Limpopo rivers, which constitutes the greater part of Southern
Rhodesia, the Highveld region stretching south-west of the Limpopo,
and the highlands behind Walvis Bay.

The plateau formation has given rise to the flat-topped and
" spitzkop " hills so typical of the Karroo country and other parts.
The main plateau falls to the coastal plain in a series of terraces,
each clearly delimitated by a mountainous rim. In the south these
terraces are most marked, and form the Great and the Little Karroo,
bounded by the Nieuwveld, the Zwartberg, and the coastal range
which falls to the true coastal belt. This belt is everywhere narrow,
save where it. widens out to form the thornveld which is Portuguese
East Africa. The western coastal belt forms the dry Namib desert
strip, but the eastern edge of the sub-continent falls less steeply to
the sea and is well watered. It is formed of a wide sloping plain,
deeply cut by innumerable streams and their tributaries. The valleys
so made are well wooded and the palm belt and other equatorial
phenomena extend much farther down this coast than on the remainder
of the continent.

A striking feature of the country is its watershed. In the northern
part of Southern Africa the Zambesi cuts far across the continent,
flowing from west to east. Balancing it on the west coast are a few
short rivers, such as the Kunene, which prove a valuable asset to

34 A. J. H. GOODWIN

that region. In the southern part of the Union the position is reversed.
Here the Orange with its tributaries almost transgresses the continent
from east to west, flowing from the Basutoland highlands to the
Atlantic. In contrast the eastern side is here well watered by short,
constant streams which run together to form considerable rivers.
The Orange is mainly a flood river with high evaporation, and dies
to a stream in the winter months, even ceasing altogether in dry years.
Save for the upper Zambesi, which has nursed the Rotse Empire,
rivers have not played a great part in the material culture of the
South African Bantu. Rivers are seldom navigable for great distances,
as the shorter rivers are too swift in their flow to the sea from the
plateau country, while the longer rivers lack water for much of the
year. What little canoe movement has existed on the middle Zambesi
and on the Ngami complex of streams and marshes has been to a
great extent confined. But if rivers have had no great part in the
building of South African cultures, their contents have proved vastly
important. No great riparian tribes exist, no Canoe peoples have
integrated the country with trade routes, but the great mass of the
country depends very dire&ly upon the rainfall and the water which
the rivers provide.

CLIMATE

The bulk of South Africa lies on the track of the south-east trade
winds throughout the year. These winds bring regular rains to
the east coast during the summer months (December-February),
rise sharply over the mountains and feed the residue of their burden
to the eastern edge of the plateau. As the winds pass onwards towards
the west coast they colled more moisture than they deposit, and
rainfall decreases while evaporation increases. The western side,
especially the Namib or coastal desert strip, is thus dry, save for the
high-lying Damara-Nama plateau. In winter (June-August) the
track of these trade winds moves northward towards the Equator.
They still feed parts of the east coast, but the rains fail to pass the
mountain rim. Synchronous with this northward swing of the south-
east winds is a parallel movement of the westerly winds, which now
bring rains to die southern coast from the Cape of Good Hope to
.Port Elizabeth.

The plateau system with its heavy eastern mountain rim, afted
upon by the rain-bearing south-east wind, has thus given us a climate
reflefted on the eastern side by abundant rains, well-wooded country,

HABITAT 35

and the cattle regions east of 25 E. On the western side the desert
reflefts the increased evaporation. To these climatic faftors are also
due the typical mountain forms varying from softly rounded grassy
hills to arid, jagged crags as we pass across South Africa from east
to west.

In all Africa the distribution of the Bantu-speaking peoples coincides
to some extent with the December-February rainfall area. -This
becomes most striking to the south, where Bantu and non-Bantu
tribes have distributed themselves in direft relation to the summer
and winter rain areas respeftively. The Bantu have confined them-
selves to the region of high summer rainfall, while the Khoisan peoples
have had perforce to share the winter rainfall region of the south-west
and the arid marshes south and west of the Kalahari. Only the Hcrero
and Ambo tribes slightly transgress the lines plotted by the limit
of the 10 in. summer rains in South Africa.

ENVIRONMENTAL REGIONS
The Tropical Belt.

The northernmost parts of Southern Africa lie within the Tropics,
but the temperature is everywhere modified by low humidity and
high altitude. Hot days give place to cold nights, and rain also brings
a sudden drop in temperature. The tropical area is best delimitated
by the distribution of the Baobab (Adansonia digitata) and the Mopane
(Copaifera mopane). The former is found throughout tropical Africa
and Madagascar, from Senegal to the Transvaal. It supplies the
natives with bark for cordage and" for string-making. The Mopane
is more confined, forming the typical forest tree of the plateau country
between the Zambesi and Limpopo rivers, with a westward extension
to the Etosha Pan and Kunene region.

A variety of other tropical trees, including the Rhodesian teak and
other water-loving timber trees, are found along river banks in the
tropical region. The Hyphane palm occurs in suitable environments,
and this general moist-tropical vegetation has an extension down the
eastern side of the continent as far as Port Shepstone on the Natal
coast. The dry-loving tropical Mopipi (Boscia rehmanniana) is a
striking tree in the more arid parts. Its soft wood is a common material
used in Native wood-carvings in Bechuanaland.

The conditions in Portuguese East Africa are rendered somewhat
peculiar by the wide, low-lying coastal belt, which here coincides

36 A. J. H. GOODWIN

roughly with the political region. This thornveld is largely a game
country, with relatively low fertility, and a rainfall which is adversely
affe&ed by the presence of Madagascar to the east.

The tropics have their own diseases. Malaria is prevalent every-
where north of 25 S. and extends even farther down the east coast.
Man and the anopheles mosquito are both essential hosts in the life
cycle of the malaria germ, and though the Natives are often carriers,
they themselves are largely immune from the worst effefts of the
fever. The complication of malaria, known as blackwater fever, has
the same distribution, but is rare among Natives.

In contrast with the rest of Africa the sleeping sickness of this
region seems to be carried by Glossina morsitans, the same tsetse
which carries the cattle nagana. The nagana comes farther south
than sleeping sickness which extends only as far as Lake Ngami
in the centre, and to the northern parts of Zululand on the east.
In both these diseases only those tropical distrifts lying below the
4,000 ft. level are affefted. The movements of the fly are further
limited by its dislike of strong light, and by its need for the humidity
which river banks afford. The Natives have thus found it possible
to avoid a heavy incidence of nagana among their cattle by confining
their animals to the high grasslands where light and altitude aid
them. The tsetse, however, still takes a terrible toll of cattle in suitable
parts of the country. Many tribes have in consequence been confined
almost entirely to agriculture.

Termites are largely confined to the high sub-tropical grasslands.
Their depredations affeft standing timber most severely. All wood-
work, even such intricate architectural forms as the Ambo village,
has to be renewed every few years. This deslru&ion and replacement
continues throughout Bechuanaland, Northern Transvaal, and
Southern Rhodesia.

Eastern Cattle Lands.

East of longitude 25 E. are the great cattle lands of South Africa.
All this region lying south of the tropics is good grazing country,
and has proved a suitable field for mixed agriculturists. South of
the extended tropical region on the east coast, the eastward slope
has been cut to form rounded, well-grassed hills. Deep, hot valleys
yield fertile soil and sub-tropical vegetation. While the rainfall is
greatest in the summer months, the winter is by no means dry, and
the chief natural forest areas thus lie along the coast. Even to-day,

HABITAT 37

between George and Port Elizabeth, where the summer and winter
rainfall areas overlap, the forests are composed of an extraordinary
variety of excellent timber trees. The Bantu have made horrid inroads
into the forest country within their areas, preferring the deep leaf-
mould for agriculture, and destroying the timber trees. The Nguni
tribes inhabiting these eastern slopes are mixed hoe-culturists and
cattle keepers. The grassy uplands are used for grazing, while the
watered valleys provide agricultural land. During the winter months
the cattle are taken to the shelter and warmth of the forested valleys,
where grass is abundant.

The whole region, and especially the moist grasslands, is, however,
subjeft to tick-borne diseases, such as tick fever, gall-sickness, and
especially East Coast fever, which seriously afFeft the cattle. Bilharzia
is carried by snails inhabiting the eastward-flowing riyers only.
A much more widespread pest, however, is the locust, of which
two varieties, red and brown, are always to be found throughout
Southern Africa. At irregular intervals 'either species may increase
amazingly, owing to the coincidence of suitable climatic conditions
and the absence of natural enemies. These swarms drive before the
prevailing winds, and wherever they go, by a few hours' feeding,
devastate the countryside, affe&ing grazing and agriculture alike.

The East Coast region rises on the west to the Highveld and the
high Rhodesian grasslands, save where these are cut by the Bushveld
region of the Limpopo and by the tropical belt of the Zambesi valley.
The remainder of the cattle country thus lies on the plateau, and
begins at the marches of the tropical savanna region, with high grass-
lands dotted with trees and shrubs which form the major part of
Southern Rhodesia. This plateau is cut to the north by the Zambesi,
and is defined to the south by the deep Limpopo valley. The rainfall,
and with it the size of the timber trees, decreases fairly rapidly from
Mashonaland on the east to Matabeleland on the west, so that the
grain country of the east here shades to cattle country on the west.

South of Rhodesia lies the strip of Lowveld which follows the
Limpopo and Komati rivers, and then virtually continues south-
south-west from the Limpopo sources to form a marginal region
between the Highveld and the eastern Kalahari. This Lowveld belt
has been a migrational route for various tribes, many of whom were
scattered direftly or indire&ly by the turmoil of Zululand in the
early years of last century. It is thus a region where offshoots of
Sotho and Nguni tribes live side by side.

38 A. J. H. GOODWIN

On the southern edge of this Lowveld region is the Zoutpansberg
where the Venda people have settled between the Shangana-Tonga
tribes of the east and the Sotho tribes of the west. The Venda country
is mountainous and rugged, and shows a varied climate. To the north
it is dry, but the southern and eastern parts are well watered, and
proved excellent forest lands before the general denudation of the
countryside. Sneezewood, ironwood, yellow wood, red almond,
and a variety of other timber trees were represented, together with
trees suitable for tanning leather, for small timber, and other Native
uses. The mountains provide good thatching grasses, and grazing.
So fertile is the red soil that two crops may be raised in a single year.
Like the varied flora, the fauna is typical of the usual game country.

From here the Highveld continues southward to include the Trans-
vaal and the mountain regions of Basutoland. It is a rich cattle region,
with reasonable rainfall, but high evaporation. The country is thus
purely grassland with few trees, bave in the more sheltered valleys.
The rate of soil erosion is everywhere high, and is rendered worse
by cattle-paths, annual grass fires, and the inability of the natives
to appreciate its inevitable results. The Sotho have been forced
towards the mountain masses of Basutoland by the white settlers of
the Free State, and here depend largely upon their cattle. The high
altitude and deficient soil of the eastern part of the Highveld have
forced the Natives to turn to sheep-rearing. The agricultural tradition
is still strong, and a considerable amount of maize and Kaffir-corn
is grown, more especially towards the north-west, which with the
coming of the European became the great " Maize triangle ". Wheat
has taken some hold on the eastern side. The whole region is over-
stocked with cattle, and the results are seen in the continual exodus
of the menfolk to the Union in order to obtain a living from mining
or domestic work under the European.

The Karroo region, south and south-west of the Highveld, has
never been inhabited by the Bantu, and little need be said of this
arid sheep-raising land, save to note that the plateau formation of
South Africa here gives rise to the flat-topped " tafelberg " and
conical " spitzkop " hills, scattered over a wide area, and giving no
enclosed valleys.

The Western Drought Lands.

The Kalahari region of central South Africa is an important
ecological area. It is a region of generally low rainfall, and precipitation

HABITAT 39

decreases as we move south or west. Evaporation is high, and
may even exceed the precipitation in parts. The rainfall is short and
heavy, and varies from 13 inches in the north to 8 inches in the
south.

The north-western limit of the Kalahari may be taken as the Omatako
River, where the red Kalahari sands change to the white sands of
the Omaheke. The main desert extends from the upper Molopo River
to Lake Ngami, its eastern edge lies near the main railway line to
the north, the western edge coincides roughly with the South- West
African border. It is an arid country, not true desert, scored by old
river beds, and lying between 3,000 and 4,000 feet above sea-level.
Undrained depressions or pans scattered throughout the region supply
brackish water after rains. The soil consists of red desertic sand, or
white calcareous tufa formed by the alternating solution and repre-
cipitation of salts in the wet and dry seasons. The extension of the
Kalahari sands as a geological phenomenon suggests that the desertic
region was possibly far more widespread at one time. Here and
there long parallel belts of loose sand are found, sometimes forming
barkhans along the river beds.

The extreme north is regularly flooded in the summer months,
sometimes heavily. The Okavango marshes are 100 miles long,
choked with reeds and laced with channels of stagnant water.
In the rainy season they drain into Lake Ngami and even into the
Makarikari depression. This latter is a great salt-pan in which water
colle&s in the wet season, but soon evaporates to leave a 4 in. crust
of salt. After rains there is profuse vegetation, Acacia giraffe* and
A. detinens thrive along the river banks, while to the north we reach
the mopane, baobab, and commiphora regions.

Even in the driest parts there is a profusion of bulbs and leguminous
plants, while the tsama melon (Citrullus vulgaris) supplies liquid to
men and beasts. The north becomes tropical and less varied. Tsetse
fly, white ants, ticks, and mosquitos abound. Days are hot, but
owing to the altitude and the dry climate, nights are cold, and rain
cools the air suddenly. Winds rise almost daily with the increasing
heat, falling off with the cool of evening. The conditions are thus
not suited to the Bantu cattle-keepers and hoe-culturists, and their
homes are distributed in strift relation to available water.

The northern and eastern parts of the Kalahari are inhabited by
the Tswana tribes. Unlike the Nguni, whose environment allows
them to divide up into small family settlements, the Tswana are

40 A. J. H. GOODWIN

forced to adopt a small town system, the town lying within easy
distance of adequate water, or having its own wells. The town is
loosely surrounded by scattered agricultural lands, while cattle are
often grazed some distance away. During the winter months cattle-
posts are built along distant streams, and the cattle migrate annually
with whole families, or with the younger men of the village. Recently,
with the imposition of peace on these tribes, the tendency has been
to constitute the cattle-posts as permanent homes, and a more scattered
economic strufture is appearing.

The north-western parts of the Kalahari form an interesting complex
in themselves. To the extreme north are the Kunene and Okavango
rivers, lined with teak and other forest trees. The upper Kunene
and upper Okavango bound the country of the Ambo peoples.
Ovamboland proper consists of a great level expanse of white sand,
covered sparsely with grass, acacia, and mopane, known as the
Omaheke country. In natural clearings are palms and baobab trees,
wild figs, and morula trees, from which last a wine is made. The
Ambo tribes are dependent upon short and heavy rains which end
in July, by when the whole region is inundated to form a vast morass
dotted with little islands. During the remainder of the year the
country dries up, and what was impassable flood now becomes a
plain, here and there pitted with water-holes and wells. In order
to combat the rising lime brought to the surface by the high evaporation,
the Ambo build up beds of earth .and cow-dung above the level of
the plain. These are planted with a quick-growing millet which
can take advantage of the rains. Largely owing to the environment,
few cattle are kept, and the breed seems smaller and stockier than
that kept by the Herero.

The Kunene as it passes through Ovamboland banks itself up above
plain-level with the Omaheke sand, then falls steeply to the coastal
belt of desert sand through which it seldom passes.

The Okavango or Kubango river, studded with wooded islands
and riddled with malaria, flows eastward from the white sands of
the Omaheke to the red sands of the Kalahari. With the Chobe or
Kwango river the Okavango spreads like a trellis over Lake Ngami,
which drains northwards to the Zambesi. This Ngami region, though
of tremendous interest anthropologically, is the home of such a
welter of tribes and mixed races and cultures that little direft com-
parison between peoples and environment is possible. Tswana
cattle-keepers, Masarwa hunters, Herero refugees from South-West

HABITAT 41

Africa, and Mpukushu tribes all form a varied population, largely
the scrapings of the Zambesi and the Kalahari.

West of Ovamboland lie the highlands of the Kaokoveld, sparsely
populated by the Tjimba. This is typical game country, and a great
variety of animals is to be found in the Etosha-Kaokoveld reserve.
The bulk of South- West Africa is made up of a southward extension
of the Kaokoveld, which forms a wide plateau at a height of some
6,000 feet, here and there enclosing valleys of typical Karroo country.
In the north is the Otavi limestone area, containing underground
water here and there revealed in a precipitous lake. At Tsumeb
is a great copper deposit originally worked by both Bergdama and
Herero smiths.

The greater part of this high country was the old Herero and
Mbandieru region, a poor cattle country which they used to the limit
with their heavy cattle in spite of the general dryness of the climate.
Their movements southwards were checked by the Nama Hottentots.
The Herero were the richest of a rtumber of related cattle-keeping
tribes, many of which after coming across the Kunene were forced
back into Angola by the pressure of the Nama peoples from the
south. Since the long Herero war with the Germans (1904-7) the
Herero have lost much of their stock, and their movements have
been rigidly confined to a small area.

The remainder of South Africa seems never to have had a great
Bantu population. Here and there small groups have pushed their
way, but with little permanent success. The limit of the Bantu coincides
roughly with the limit of the 10 in. summer rains, and they have
been confined there by the Namib strip to the west, the Hottentots
to the south, together with Kalahari and Karroo thirst-lands. On
the eastern side of the continent they could more easily have advanced,
and once the winter rain area had been reached, these people would
undoubtedly have spread and multiplied. It so happened that they
came into violent contacts with the Hottentot tribes in the region
of the Kei and adjacent rivers, and were held back until the European
colonizers advanced from Capetown along the south coast.

CHAPTER III
GROUPING AND ETHNIC HISTORY

By N. J. VAN WARMELO

INTRODUCTION

INFORMATION concerning the history of the South African Bantu
tribes comes from two sources, European and Native. The personal
observations of the earliest European explorers and navigators, and
of shiprecked seamen, such as Perestrello (1554), the pilot of the
Santo Alberto (1593), and the men of the Stavenisse (1686), establish
beyond doubt not only that the Bantu were already dwelling in
South-Eastern Africa at that time, but also that in part at least they
were the identical tribes found in the same localities at the present
day. Much beyond that these early sources do not go. On the other
hand the European writers who witnessed the subsequent phases of
Native history in this country can only tell us about events of com-
paratively recent date, much too recent in any case to be of use in
an attempt to classify the tribes. Of course, the evidence of the
few men who, like Isaacs for instance, were actually on the scene of
Shaka's rise to power in Zululand, is of the greatest value ; but it
does not supply the wealth of detail we require regarding the course
of the speftacular events which changed the face of South-Eastern
Africa. Above all, these sources tell us little or nothing about the
distribution and names of tribes, their customs, and languages, as
they were before the storm broke that all but swept away many
of them.

The Native sources of history consist of the traditions, legends,
and tales handed down orally from generation to generation in each
tribe. Of these only that small portion is available which has been
recorded in print, mostly by Europeans, and then only too often
in a European language, without the original vernacular version.
This latter circumstance has reduced the usefulness of many of the
Native records in question to an extent unsuspected by the collectors.
It is not out of place here to stress that a scrupulous regard for the

44 N. J. VAN WARMELO

aftual letter and phraseology of Native tradition is essential in the
colleftion and editing of such material. But even were everything
still extant to be rescued from oblivion, it is well to remember that
this also would represent but a fraftion of the wealth of traditional
lore which existed up to say six score years ago, before wars and
famines laid waste to Bantu culture, and caused the premature death
of the aged repositories of tradition. And finally we must bear in
mind that, at the very best, Native tribal tradition is weak in chronology
and scanty in regard for truth. As a rule, three hundred years is the
limit of possibly reliable tradition. Beyond that, legend and fairy
tales begin to luxuriate.

Apart from the sources of information just mentioned, we are
guided, in making our grouping, by what we know of the ethnological
and linguistic charafteristics of the tribes under review. 1 The grouping
given below is intended to have some praftical use : it should serve
to indicate to the investigator where he is likely to encounter similarities
and affinities as he moves from tribe to tribe. Of what nature these
similarities and affinities are in each case, and what therefore are
the charafteristics of each group and subgroup, will appear more
fully in the succeeding chapters of this book. It will further be seen
that the grouping largely coincides with geographical distribution,
and also that the groups are almost a pifture of the language-groups
of South Africa. With primitive folk it cannot be otherwise. Roads,
travel, and communications being virtually non-existent, or at best
feebly developed, a geographical unit inevitably becomes the seat of
a particular dialeft and form of culture, provided it be left undisturbed.
It is the migrations, and the enclaves of not yet assimilated con-
querors or conquered which these leave in their wake, that therefore
make the exceptions and provide the surprises.

The Bantu peoples dealt with in this book fall into the following
major groups, to be reviewed hereL in the order given : Nguni,
Shangana-Tonga, Sotho, Venda, and Lemba. The Venda group is
small in comparison with the rest, although of paramount ethnological
interest ; and the same applies to the Lemba group, which numerically
is most insignificant.

1 Besides the existing literature and museum collections, I have drawn upon my own
observations. Readers familiar with the information already available will easily see for
themselves in how far my judgment has been guided by this fatar.

GROUPING AND ETHNIC HISTORY 45

NGUNI GROUP

The tribes belonging to this group live, with few exceptions,
below the high plateau of the interior, between the escarpment of
the Drakensberg and the sea, and stretch, in a long broad belt of
hundreds of tribes, from Swaziland right through Natal far down
into the Cape Province.

The Nguni are markedly a " cattle people " ; and the presence
of " click " sounds in their language seems to be due, almost
undoubtedly, to contaft with that purely pastoral people, the
Hottentots. The problem is how, where, and when such contaft
was effefted. The existence of this problem is by itself sufficient
to cast serious doubt on the speculations of writers about early Nguni
history, for they do not Account for what we aftually find. The
presence of the clicks in all the Nguni dialefts,- even those of the
Transvaal Ndebele, who have been. living in that province for at
least three to four centuries, seems incomprehensible except on the
assumption of a focus point of Nguni development far in the South,
where contaft with the Hottentots was possible. All this is not in
accord with the theories hitherto put forward as to the way in which
the Nguni came down from the North and occupied their present
home. The accepted chronology tentative of course also does not
appear to meet the case. There is a third grave difficulty : the Lala
enclave which used to occupy approximately the present Southern
Natal. The Lala were largely wiped out a hundred years ago, but
enough remnants are left which may be studied. Not very much of
true Lala custom and speech has survived to be recorded, but even
this has not yet been done, and so we know almost nothing about
them. It is claimed for them that they were of Shona origin, and
some features of their language certainly are reminiscent of Shona
or Tonga; but beyond that nothing definite can really be said.
In addition, an almost impenetrable veil was drawn over the past
a century ago. In the Cape Colony destru&ive frontier wars were
waged, while in Natal it seems that hardly a tribe was fortunate
enough to be left undisturbed during Shaka's reign. Whole tribes
vanished, and everywhere traditions, culture, and material possessions
were lost.

Small wonder, then, that the early history of the Nguni group
as a whole is still a field for inquiry, but hardly a subjeft to be dogmatic
about. The massive volumes of Bryant (Olden Times in Zululand

46 N. J. VAN WARMELO

and Natal) and Soga (The South-Eastern Bantu) have at least achieved
this also that, by coming to conclusions which can by no manner of
means be reconciled, they have shown onlookers that progress is
not being made as merrily as some would have us believe. In short
as regards origins we are up against a wall we are not as yet prepared
to scale. The theories hitherto put forward are not worth repeating
here : they appear to be fanciful, and hardly meet the case. Those
further details of history which do require mention are given below
for each separate sub-group.

In conclusion we should note that, had the events of the past
century, due to Shaka's founding of the " Zulu " empire, not taken
place, praftically the whole Nguni group would have borne a different
appearance. One entire sub-group, for instance, that of the " Fingo
and others ", owes its existence to the tremendous happenings which
accompanied the birth of that empire ; and several other most impor-
tant Nguni offshoots, like the Rhodesian Ndebele, also came into
existence in the same fashion, In the storm centre, where these distur-
bances took their rise, things of course took such a turn that research
largely consists in the fitting together of fragments.

Cape Tribes Proper^

These are the southernmost Nguni tribes. They had already
been in occupation of the present Transkei and Ciskei for centuries
by the time die influx took place. of the immigrant Fingo and other
refugee tribes mentioned under the next sub-group. The Cape tribes
speak the same language with but small variations, but, while a common
descent is traceable for a number of tribes, this is not the case with
most of the others, between whom no sort of genealogical relationship,
as far as tradition goes, can be said to exist.

An example of a group of related tribes is found in the Xhosa,
represented by the Gcaleka of Chief Zwelidumile Sigcau, the Ngqika of
Archie Sandile, and smaller tribes like the Ndlambe, Dushane, Qayi,
Ntinde, and Gqunukhwebe. The last-named are said to be of mixed
Hottentot descent. Tradition regarding the genealogical ramifications
of Xhosa royal lineage and its offshoots has been well preserved, and
is of value as being virtually a synopsis of the relations of the Xhosa
tribes to one another. This Xhosa tradition, according to Soga,
says that some centuries ago the tribe dwelt along the upper reaches
of the St. Johns River, far to the north-east of their present home.
History before that is pure conjefture; while their subsequent

GROUPING AND ETHNIC HISTORY 47

history, which is well described by Soga, is too recent to interest
us here.

Farther east and north-east of the related Xhosa tribes are the
Thembu and their offshoots, such as the Hala, Jumba, and N'dungwana.
With them may be classed the Bomvana, Qwathi, Nqabe, and
Mpondomise, which although not related are nearest to them in
culture. All these tribes also have traditions to the effeft that they
immigrated from the dire&ion of Natal. Otherwise their recorded
history is meagre, containing few indications of value.

Still farther east are the numerous tribes and septs of the Mpondo,
ruled over by several independent chiefs. The Mpondo are again
distinft, in some respe&s, from the tribes mentioned above. They are
divided into a great many clans, of which the largest though not
necessarily the most important in rank are, amongst others, the
Bhala, Kwalo, Gingqi, Kwetshube, Nyauza, Khonjwayo, Nci, and
Ngutyana. There are, further, amongst them some tribes not of true
Mpondo descent, but now looked upon as being praftically their
equals. Besides these we find many descendants of the refugees from
Natal, whom Chief Faku took under his protection. While some
of these managed to maintain their identity intaft, others simply
merged in the Mpondo population. There are for instance se&ions
of Tolo, Tshangase, Tshwawu, Zizi, Hlubi, Bhele, Ngwane, and
similar names commonly encountered amongst the immigrant Fingo.
The earliest known history of the Mpondo also shows them already
settled, several centuries ago, not very far to the North-East of
their present country. Of their relationship to other tribes nothing
definite is known.

Fingo and Other Recent Immigrants into the Cape

When Shaka embarked upon his career of empire building in
Natal, numerous tribes were dislodged either directly or indirectly
as a result of the ensuing state of war. Hence early in the nineteenth
century many thousands of refugees from Natal began to cross over
the Umzimkulu, seeking a new home among the Cape tribes, especially
the Xhosa, and among the European colonists. They came both as
solid tribes, and in large and small bodies of homogeneous or of
composite character. Their numbers were further augmented from
another dire&ion by the fugitive Hlubi of Mpangazitha, who had
been driven out of Basutoland, whither he had first fled, by Mathiwane
and his Ngwane, who were themselves also fugitives from Natal.

48 N. j. VAN WARMELO

The latter Chief presently came down into the Colony himself,
where in August, 1828, his tribe was broken up in an encounter with
European forces.

Many of these refugee tribes returned to Natal when peace was
there restored. But many others remained behind. The real " Fingo "
(Mfengu) were subsequently led out of Xhosaland, but when later
on part of the Western Transkei became vacant they were settled there
and still form the bulk of the local population. There are to-day
many thousands of completely detribalized Fingo. Of those still
carrying tribal names, whether they recognize chiefs or not, the
majority belong to the big Hlubi, Zizi, and Bhele tribes. Smaller
units include the Kunene, Maduna, Gubevu, Tolo, Miya, Khuze,
Mbuthweni and Zotsho. The question as to how far the Fingo still
form a cultural unit, and still know and submit to their hereditary
heads, has not of late been studied. It is not possible, therefore, to
give any indications on this point. Many of the Fingo of course live
in areas of the Transkei where they fall under the control of non-
Fingo chiefs.

The same applies to a large proportion of those immigrants into
the Eastern Transkei who cannot striftly speaking be called Fingo.
To these belong the Bhaca and numerous se&ions of the Wushe,
and clans related to the Bhaca. Others again are independent, such
as the two important Bhaca tribes in Mount Frere under Chiefs
Sikhanda and Mncisana. Two others are in Southern Natal, in
Bulwer and Ixopo districts. In the same sub-group as the Bhaca
may be put the Nhlangwini tribes of Griqualand East and Southern
Natal (Harding and Ixopo distrifts), and a number of small clans
said to be related to them, together with the Xesibe (chiefly in Mount
Ayliff) and the Xolo and Nzimakwe of Port Shepstone. Of some of
these tribes we know more or less how they came to be where they
are now, the exploits of the Bhaca Chief Ncaphayi especially being
fairly well recorded. But of what went before, and of the original
state of the people, as of their present customs and culture, we can
say little or nothing.

Natal Ngu.nl.

Were it not that from about the year 1815 onwards great changes
had occurred in Natal, a very different grouping of its inhabitants
would have been both possible and necessary. As it is, no other
grouping than that outlined below can or need now be formulated.

GROUPING AND ETHNIC HISTORY 49

This is because early in the nineteenth century Shaka, Chief of the small
Zulu clan, began attacking and establishing his supremacy over his
neighbours right and left. Even when due allowance is made for the
inevitable exaggeration in Native accounts of fights, raids, and
massacres, all on an unprecedented scale, tremendous changes
undoubtedly took place in the province. Many large tribes and bodies
of people left the country in all direftions, never to return. Many
more small tribes were wiped out completely. It also appears
that at one time few, if any, people were suffered to live south
of the Tugela River, although as the power of the kings weakened,
many refugees returned and settled in their old homes, but only
to the extent permitted by the recent occupation of the country
by Europeans.

In consequence of these changes, the only grouping now prafticable
is to include all the Natal tribes into one common category. Before
1815 one would have classified them into separate groups : (<z) the
true Nguni or Ntungwa, with perhaps a subdivision for (K) the Mbo ;
and (c) the Lala tribes. But to-day this is impossible, for the rule
of the Zulu kings not only introduced their own Zulu culture every-
where, but also promoted the growth of standard language and
custom. This accounts for the large measure of uniformity found
throughout Natal to-day. Variations in cultural detail do still exist,
but they cannot be defined except in extremes, and they form such
an intricate pattern that dividing lines cannot be drawn. The mixture
of tribes under European rule helped further to stamp out differences.
In addition, the clans originally forming the groups just mentioned
are now scattered haphazardly all over the whole country. One can
say neither which tribes are of Lala stock, nor which were originally
Mbo. We must therefore consider the Natal Nguni as a whole. The
group is named after the province, but it must be noted that provincial
boundaries do not constitute the boundaries of the group. Some
of the Cape tribes previously mentioned live in Natal, and on the
other hand pra&ically all the Natives of the South-Eastern Transvaal
belong to the Natal Nguni.

The tribes of the Natal Nguni are mostly known by the family
or clan names (qitongo) of their Chiefs. These tribal names therefore
do not refleft the aftual composition of the tribes. Though a tribe be
known for instance as the Mthethwa, there may be twenty, fifty or a
hundred different clan names, i^ibongo, represented within that tribe.
The name Zulu is another instance, for it is, striftly speaking, only

SO N. J. VAN WARMELO

the name of the royal clan. To call all the Natives of Natal, or of
Zululand for that matter, " Zulu " is incorreft. It is of course con-
venient, but nevertheless, in all cases where the slightest modicum
of precision is required in stating exaftly what one means, the term
" Zulu " proves to be a most ambiguous and exasperating instrument.
One cannot discuss the " Zulus " except in the vaguest terms. The
use of " isiZulu " for the language, however, seems justified.

The Natal Nguni are divided into more than two hundred inde-
pendent tribes, of which we need name only a few of the most
powerful. To group them in any way whatever according to their
affinities is, as has been said, impossible. In the Southern and South-
Western distrifts of Natal are a number of Nyuswa tribes, the Dumisa,
Cele, Khuze, several tribes of Dlamini, the Nxamalala and many
others. They are flanked, farther north, by the tribes living under the
Drakensberg, the Ngwane and Ngwe (or Ngweni) and others. More
towards the centre are the Mabaso and the powerful and still very
raw tribes of Msinga distrift, the Bomvini, Cunu, and Qanyini.
Proceeding farther towards the coast, we find numerous seftions of
Mkhize and Mafunze, the Gcumisa, two large Zondi tribes, the
Makhanya, Ximba, several tribes of Thulini all near to the coast
from Umzinto to Stanger, the Qadini in Ndwedwe and adjacent
distrifts, the Khabeleni, the Ntuli, Butlielezi, Qwabe, and Mpukunyoni.
Farther to the west, in Msinga and neighbouring distrifts, are the
Sithole tribes, the Thenjini or Thembu, and, in the northernmost
districts, the Hlutyini or Hlubi, Nkosi, Khumalo, and Mdlalose.
Turning thence towards the east we encounter amongst others the
Mbatheni, Ntombela, Gazini, Mthethwa, and in Nongoma and there-
abouts the Zulu. To the extreme north-east are the tribes of Ubombo
and Ingwavuma divisions : the Nxumalo, Zikhali, Myeni, Nibele,
and others, and the Mngomezulu, Nyawo, and Mathenjwa, whose
next-door neighbours are the southernmost Shangana-Tonga, viz.
the Tembe.

Swaji

This sub-group is of recent origin. Before Shaka's time the present
Swaziland was partly occupied by various Sotho tribes about which
we know virtually nothing, but who are best represented to-day in
the Pai and Pulana referred to in the Sotho group. The southern
part of Swaziland was occupied by clans of Nguni origin, mostly
of the variety characterized by the teke^a way of speaking, and by

GROUPING AND ETHNIC HISTORY 51

the customs commonly associated therewith. Commencing with the
increasing power of the Ngwane Chief Sobhuza (ca. 1820), the
" Swazi " people gradually began to come into being, especially
through the conquests of Sobhuza's descendant Mswazi (ca. 1840-
1875), after whom they are named. The latter subje&ed the Sotho
clans of Swaziland, or drove them out, and by extensive raiding
increased his wealth and power. The Swazi, having never been
subdued by force of arms, remained a nation in spite of the imposition
of European control, and are intensely proud and conscious of the
far. The circumstance, however, that the many thousands of Swazi
living outside the borders of Swaziland were freed from the conttol
of the Paramount Chief has weakened tribalism among them. Never-
theless several Swazi tribes near the border may be considered almost
integral parts of the Swazi nation.

In Swaziland itself the old Sotho population seems to have dis-
appeared, except in name. The descendants of the Sotho clans are
colleftively known as abeSuthu or amaKhandzambili. Another non-
Swazi element are the clans from Zululand, no distinction being
made between true Ntungwa and Mbo stock. Thirdly there are the
Swazi proper. To outline areas or to define groups in which these
elements of the Swazi culture-unit preponderate is impossible. The
difficulty lies both in the lack of attention paid to this problem, and
in the comparative uselessness of tribal names as a guide. According
to Nguni custom the tribes are known by the clan name (isibongo)
of their Chief. But since in a tribe any number of clans may be
represented, the Chief's isibongo is but a weak clue to the origin and
composition of his following. There is for instance a considerable
preponderance of Chiefs with the isibongo Nkosi, but this is explained
by the circumstances that Nkosi is the clan-name of the paramount
house. One infers from this that a large proportion of the followers
of these men are not true Swazi, for the latter would be ruled by their
own hereditary Chiefs who have other clan names. On the other
hand it is likely that, where non-Swazi are in control of tribes, these
latter are tribes also largely of non-Swazi origin. All we can do
for the present is to give some indications based on the clan-names
of the ruling families. About the elements that form the tribes
mentioned below it is not possible to say anything more definite.

Taking first the Swazi proper, there are fifty-odd tribes ruled over
by Chiefs with the royal isibongo Nkosi. Their following, as noted
above, is likely to be partly non-Swazi, but on the other hand the

52 N. J. VAN WARMELO

influence of the Chief and his entourage is not negligible. Other
true Swazi tribes are likely to be those under, e.g., Malangatonke
and Siboshwa (both Fakudze), and Chiefs with such vybongo as
Hlophe, Hlatywako, Mamba, Katse, Madvosela, Motsa, Mndzebele,
Shiba, Shongwe, and Tsabetse.

Amonst the clans of " Zulu " origin one finds such vpbongo as
Biyela, Mkhatshwa, Mtsetfwa (Mthethwa), and Zwane.

The original Sotho population of Swaziland is represented by
the " Khandzambili " clans, some of which, however, are said to be
not of Sotho origin. It is natural that when the Sotho turned Swazi
in language and custom, their clan names should also have assumed a
Swazi garb, and that they should -now all be provided with the
ipnanatelo (a sort of complement to the isitongo) 1 demanded by
Nguni usage. Amongst the Khandzambili there are such clan names
as Bhembe, Gama, Gamedze, Magagula, Maseko, Nkambule, and
Sukati. These are not tribal names such as commonly found amongst
the Sotho, but clan names. It is accordingly difficult, if not impossible,
to identify them with Sotho tribes, unless a study of the family names
of the Eastern Sotho is made. But in the case of Maseko at least there
is no shadow of doubt, for one of the three branches of the Pulana Sotho
is named baxaMaSexo. As previously stated, there is reason to believe
that the Pulana are the modern representatives of the old Sotho
population of Swaziland, and Maseko is doubtlessly the Swazi version
of Maiexo.

The Swazi living outside Swaziland, especially far afield, have
followed their own independent lines of development, while those
in the Highveld distrifts of Piet Retief, Ermelo, Wakkerstroom, and
as far west as Witbank and Rayton, have-been influenced by contaft
with the Zulu-speaking people of these areas. Of the Swazi tribes in the
Transvaal may be mentioned the Nkosi of Chiefs Mhola Dlamini,
Dantyi Nkosi, and others, the Shongwe and the Khumalo, all in
Barberton and Nelspruit ; the Hlatywako of Mhlaba and some smaller
tribes (Dlamini, Magagula, Sukazi, Shongwe, etc.), all in Piet Retief,
and a few more in Paulpietersburg distrift ; and finally those small
stray groups living under independent headmen in Sekiikuniland,
Nelspruit, and Pilgrimsrest. The history of these Transvaal Swazi
is not of much importance except to themselves. They are in part
small se&ions who lived on the periphery of happenings in Swaziland,
and were only too glad to be left unmolested in return for submission

1 See below, p. 83 f.

GROUPING AND ETHNIC HISTORY 53

to Mswazi ; and in part they are people left behind by raiding columns
sent abroad by this enterprising Chief. While some of these outposts
subsequently had to be abandoned in the face of Pedi aggression
others managed to cling on and remained where we find them to-day.

Transvaal Ndebele.

These Ndebele tribes must not be confused with the Ndebele
now living in Matebeleland, Southern Rhodesia. The latter left the
confines of Zululand only a century ago under the leadership of
Mzilikazi, whereas the former had by that time already been settled
in their present territory for at least several centuries. Living as they
did surrounded by various Sotho tribes, they could not avoid being
influenced considerably by Sotho culture and language. Some of them
have in fa& become almost pure Sotho in everything but name.

The Transvaal Ndebele fall into two se&ions, Southern and Northern.
They are divided by a considerable stretch of country, which has
only of late been bridged by movements of small groups. They are
also distinft in point of language, for each of the two dialefts is
chara&erized by a number of features peculiar to it alone. The Ndebele
dialefts of the South are better preserved than those of the North,
which have been largely superseded by Sotho. The same may be
said of custom and possibly also physical chara&eristics.

The Southern group to-day comprises a single senior tribe, the
Manala, and a junior tribe, the Ndzundza, which was broken up fifty
years ago and is now represented by over half a dozen seftions. These
two tribes have clung to Ndebele custom and language with extra-
ordinary tenacity. This is not the case with the Hwaduba of Haman-
skraal distrift, who, though said to be descended from the same
source, are now to all intents and purposes a Sotho tribe. The tribes
of the Southern group trace their descent from the tribe of one Msi,
who long ago lived near Pretoria, where the Manala still are. A rather
vague tradition, which may, however, be accepted states that before
Msi's time they had come from the direftion of Natal. The same origin
may be postulated for the Northern tribes.

The Northern group is composed of the Ndebele of Langa (commonly
termed, in Sotho pronunciation, the Laka), represented by several
seftions, mostly in Potgietersrust ; and of the Maune or Letwaba,
likewise represented by several seftions. The Seleka living on the
Bechuanaland border are also said to be of Ndebele stock. This group
finally includes the Moletlane or Sebitiela, who, according to some

54 N. J. VAN WARMELO

traditions, trace their origin to the parent tribe of the Southern
Ndebele. They appear to have trekked away to the North very many
years ago, and settled near the great bend of the Olifant River. An
earlier offshoot of the Moletlane are the Mokopane just outside
Potgietersrust. There is also a recent offshoot under Johannes Kekana,
now settled very near the ancient home of his tribe in Hamanskraal
distrift.

Recent Nguni Offshoots.

Early last century three bodies of Zulu-speaking people fled from
Natal to escape impending destruftion at the hands of Shaka. They
marched north and north-west, living by rapine and plunder.
After many vicissitudes, each of these bodies eventually found a
permanent home, where they perpetuated their race, customs, and
language amongst the alien Bantu tribes which they subjefted to
their rule.

The body led by Mzilikazi trekked over the Highveld of the
Transvaal and settled for a while near Pretoria, and later near Zeerust,
harrying the Sotho tribes and raiding as far south as Basutoland.
When he turned his attention to the emigrant colonists the latter
retaliated and, the country becoming too hot for him, he led his
people over the Limpopo into the present Matebeleland. The Ndebele
kingdom founded there collapsed before the advance of European
colonization, but the language and culture of the invaders has survived
to this day, though in diluted form.

The " Ngoni " or Shangana led by Soshangane, alias Manukuza,
proceeded north and after a sojourn near Chipinga settled in Gasaland,
not far from Delagoa Bay. They were probably not very numerous,
but they were powerful enough to subjeft most of the Tonga com-
pletely and to introduce their own language and customs to such an
extent that the latter should now be termed Shangana-Tonga. When
their empire was also overthrown by Europeans, the Portuguese in
this case, a great many, including the heir apparent, sought safety
in the Eastern Transvaal, where they are still found, mostly in Pilgrims-
rest, but also in the Sibasa distrift, near the Zoutpansberg mountains.

The third body of Zulu-speaking emigrants was led by Zwang-
endaba. At the end of a comparatively brief rest in Gasaland, differences
with Soshangane's people compelled them to seek a wider field farther
north. After an amazing march of some thousands of miles, along
a route which they strewed with the wreckage of raiding and bloodshed,

GROUPING AND ETHNIC HISTORY 55

they eventually settled in the neighbourhood of Lake Nyasa. Their
descendants, who still speak a Zulu of sorts, may be found dwelling
on both sides of the Lake, in groups now widely separated from
one another.

SHANGANA-TONGA GROUP

The Shangana-Tonga form a very large group, the northern
limits of which have never yet been accurately defined. To the South
and South-East they have long been in contaft with the Nguni, but
with the Sotho and Venda farther north contaft was made only
in recent times, by emigrant Shangana-Tonga.

The whole of the original Shangana-Tonga group lives in Portuguese
East Africa. The list and grouping of clans given of them by Junod
(in his Life of a South African Tribe) is not altogether satisfaftory,
but one must add that this is due in part at least to the complexity of
the situation created by the Ngoni conquerors, who purposely dis-
membered tribes and scattered portions of them far and wide. Generally
speaking, however, it may be said that the original Shangana-Tonga
group fell into tkree tolerably well-defined seftions : Southern,
Central, and Northern. To the first belong the clans of Maputa,
Tembe, Mpfumo, and others, classed together by Junod under the
name of Ronga. To the Central Shangana-Tonga belong the clans
of Khosa, Nkuna, Mavunda, Valoyi, Maluleke, Nhlanganu, and others,
classified by Junod into the sub-groups of Nwalungu, Bila, Hlanganu,
and Djonga. To the Northern seftion belong the Hlengwe, Tswa,
and others, extending far to the North and North-East.

About the early history of the Shangana-Tonga group remarkably
little has been put on record. According to Junod the traditions as
to the direftion of immigration differ from clan to clan, some having
come from the North, some from the West, and some from the South.
The inference is that the immigration of the Shangana-Toriga into
their present territory took place in remote times. This is supported
by evidence recorded in early Portuguese writings : Junod cites the
account of Perestrello (1554) as proof that " four or five hundred
years ago at least, the chiefs Tembe, Mpfumu, Manhisa, Libombo, all
of whom still have descendants, were already in the country round
Delagoa Bay ", 1 From the evidence of Shangana-Tonga culture and
language it is difficult even to surmise the nature of their early history,
except to say that they probably supplanted an earlier and different

1 H. A. Junod, 1927, i, 27.

56 N. J. VAN WARMELO

population, of which the Chopi and Tonga of Inhambane are a
survival, that they enjoyed a long period of undisturbed develop-
ment in their present environment, and that they may be, in origin,
the closest relatives of the Nguni. But this is not quite certain, for
the Shangana-Tonga were at one time a purely agricultural people,
possessing no cattle ; and they did not take part in that phase of Nguni
development in which the languages of the latter acquired the click
sounds.

After having lived undisturbed probably for centuries, the Shangana-
Tonga, like the other South African Bantu tribes, were suddenly hit
by the storm that arose when the rise of Zulu power caused many
Nguni tribes of those parts to seek safety in flight. To the Shangana-
Tonga territory came the conqueror Soshangane, with his following
of Zulu-speaking tribesmen. Establishing himself in Gasaland, he
proceeded to extend his rule over the Tonga tribes, while raiding,
exa&ing tribute, and drafting their young men into his ranks in the
process. After his death, his sons Mzila and Mawewe fought over the
chieftainship, the former though the junior remaining vi&orious.
His son again, Ngungunyane, was the last of the independent Ngoni
kings, for his rule was overthrown by the Portuguese. From the first
days of Soshangane's invasion to the end of Ngungunyane's reign,
there was continual fighting and general insecurity. Many of the
Tonga therefore sought safety elsewhere. It thus came about that
several great emigrations took place, with a steady trickle of migration
going on all the time in between. The emigrant Shangana-Tonga
took the only way open to them, to the West, over the Lebombo
hills into the dry low country of the Eastern Transvaal, a country
they had hitherto avoided, but which, being unpopulated, offered no
obstacle to their progress. Proceeding westwards from wherever
they happened to be living in the present Portuguese territory, they
settled all over the North-Eastern Transvaal. It is thus natural that
the bulk of the Shangana-Tonga in the Northern distrifts should
belong to the corresponding Northern seftions of the Shangana-
Tonga group, while those in the Eastern Transvaal are members of
the Central Shangana-Tonga sub-group. The refugees came over
the border in small parties, and settled down wherever they could.
Very often they became the subjefts of Sotho and Venda chiefs,
and though the tendency to reassemble and live together was there,
they usually failed to muster sufficient strength to form tribes of any
importance.

GROUPING AND ETHNIC HISTORY 57

As representative of the Southern group of Shangana-Tonga there
is but one tribe living within the borders of the Union, viz. Chief
Mhlupheki's tribe of Tembe or (in Zulu) Mabhudu, round the west
of Kosi Bay in Ingwavuma distrift, Zululand.

The next sub-group are the Nhlanganu of the Lowveld of Pilgrims-
rest distrift, composed of the tribes of Sobyana, Ndjondjela and some
others, whose followings are, however, much mixed with Shangana.
The Shangana themselves form another sub-group. The name
Shangana actually denotes only those Ngoni who came from Zululand
with Soshangane (whence their name), and their descendants, who
are, of course, largely of mixed blood ; but it also embraces a number
of Tonga who have adopted their masters' language and customs.
To name these folk as members of the Nguni group is therefore only
partly justified, and it is right that they should be mentioned here
also. The Shangana tribes in the Transvaal are under Thulilamahashe,
Gija, Bantom, and other Chiefs of more obscure descent, though with
a considerable following, in Barberton distrift. Another sub-group
is composed of the tribes living about midway, in the distrifts of
Leydsdorp, Tzaneen, and Duivelskloof. By far the largest is the
Nkuna tribe of Chief Muhlava, the others being most insignificant.

The fifth sub-group is that of the Northern Shangana-Tonga.
It is composed of several fair-sized tribes under the Chiefs Mhinga,
Sikundu, Sigalo, and Nwamita, as well as literally scores of small units
under so-called " independent headmen ", many of whom have a very
small following indeed. The existence of all these tiny groups is good
proof that these Shangana-Tonga aftually did immigrate into the
Transvaal over a long period and in the haphazard and leisurely way
we have described. Their individual history has not yet been put oh
record.

SOTHO GROUP

The term Sotho is used for the whole of this group because all
its members, with the exception of some Tswana, call themselves
baSotho. These people differ from their Nguni and Shangana-Tonga
neighbours in some important respects, especially in regard to language
and social organization. The group falls into three distinft sub-groups :
the Southern Sotho of Basutoland and adjoining territories ; the
Tswana or Western Sotho ; and the Eastern Sotho of the Transvaal.
All these sub-groups may be further subdivided into smaller culture

58 N. J. VAN WARMELO

areas, the transitions between which are so gradual that their boundaries
are often difficult to define. The Southern Sotho sub-group as a whole
is easy to define, on account of its isolation from the others ; but a
satisfaftory dividing line between the Western and Eastern Sotho
is not so easily found. Moreover the Eastern sub-group contains
more diversity than any other, and the label, while convenient, is
not too precise.

The early history of the Sotho tribes is still much enveloped in
the haze of conjefture. That they all in the first instance came from
the North is obvious, and the traditions of a number of tribes contain
indications to that effeft. But by what route they came, how long
ago, and whether even the majority of the Sotho tribes are descended
from one parent tribe is all most uncertain and doubtful. The Tswana,
however, do seem to be related and to have come South in several
migrations, but of many other Sotho tribes one cannot say as much,
and it is hardly worth while hazarding a guess. It is not possible
to conneft up all the Sotho tribes with one another, even in the most
fanciful genealogies. This is because there are unmistakably foreign
elements amongst the Sotho, as for instance the Koni, whose name and
Eastern habitat have always been accepted as evidence of Nguni
origin.

Southern Sotho.

Prior to the troubled times attendant upon the rise of Shaka in
Zululand, the inhabitable western part of Basutoland and the adjoining
country were occupied by divers tribes, such as the Fokeng, Tlokwa,
Taung, Kwena, Kxwakxwa, Kxolokwe, Sia, and numerous others.
These were all Southern Sotho, though they differed from one another
in lesser points of culture and dialeft, about which very little is known.
The history of these tribes is admirably described by Ellenberger
(History of the Basuto). He distinguishes between several different
stocks from which the present-day Southern Sotho are descended.
There were the people who had come from the West, and were
therefore related to or descended from the Tswana, such as the
Phuthing, Kxolokwe, Sia, and Tlokwa, and, as later arrivals, the
Lihoya (Dighoya) and Taung ; and there were others who had come
from the East, viz. the Phuthi, who probably formed part of the old
Sotho population of Swaziland.

All these tribes lived peacefully and comparatively undisturbed
until 1822, when the first fugitive Nguni hordes, fleeing from Natal,

GROUPING AND ETHNIC HISTORY 59

broke over the Drakensberg into their territory, and a new era was
ushered in. As elsewhere, tribes were ousted from their homes and
began to wander about. Political chaos and famine reigned. The Tlokwa
under Mantatise set out on a career of rapine and conquest, but ended
up by sustaining great losses themselves. Sebetwane and his followers
trekked northwards through Bechuanaland and the Kalahari until
they reached the upper waters of the Zambesi. Here they founded
the so-called Kololo kingdom of Barotseland, in which Southern
Sotho language and custom still survive to a considerable extent.
Meanwhile in Basutoland itself Moshesh followed another course.
With great political wisdom he accepted all stray people who came to
him for protection, warded off the attacks of Mzilikazi's Ndebele,
built up a great tribe, and, extending his rule, founded what we know
to-day as the " Basuto Nation ". At one time probably all the Southern
Sotho were under his control. This ceased when the boundaries of
Basutoland were defined. Since that day the Sotho in the Orange Free
State have lived a detribalized existence, with the exception of the
two tribes in the Witzieshoek reserve. The few tribes resident in
the Transkei remained under chiefs.

The policy of Moshesh and his successors was to put their kinsmen
in charge of areas, as governors, all over the country. This favoured
the general trend of development towards uniformity in custom and
language. The process is not yet complete, but its end is in sight. To
name, to define, and to classify the tribes constituting the Southern
Sotho sub-group cannot therefore be done satisfaftorily. There are
a great many chiefs of varying rank, many of whom are kin of the
Paramount Chief, the lineal successor of Moshesh. Others are
descended from the royal houses of formerly independent tribes.
The following of all chiefs is mixed up to some extent, but that of a
great number is so mixed that it is hardly possible to indicate which
elements preponderate. Chiefs with large tribes of this description
are, for instance, the Paramount Chief himself and Seeiso, both in
Maseru district, and Tsepo Nkuebe and Solomon, both of Quthing
district.

A large number of Kwena are found under Chiefs Mopeli of
Buthabuthe, Motsoene of Leribe, Sekhonyana and Khoabane of Maseru,
and many other minor chiefs, as well as under Jeremiah Moshesh and
Khorong Lebenya in the Transkei. There we also find the large
Hlakwana tribe under Motheo Sibi. The Fokeng live mostly in the
North, in Berea and Mafeteng districts. The Kxwakxwa (Khoakhoa)

60 N. J. VAN WARMELO

under Matumane and the Kxolokwe (Kholokoe) under Qobela live
in Buthabuthe distrift. Tribes of Sia and Taung are found especially
in Mafeteng. The very mountainous parts of Eastern Basutoland have
been partly occupied in more recent years by a mixed population
drawn from all tribes ; but the distri&s in question (Qacha's Nek
and Mokhotlong) also harbour the Tlokwa of Mosuwe. Another
branch of Tldkwa lives in Mount Fletcher distrift under Chief Scanlen
Lehana. The Phuthi are represented by the large tribe of Chief
Bereng in Mohales Hoek, by some other small ones in Quthing, and
by some groups living in the Transkei, where they are subject to chiefs
not of their own tribe.

In the Northern distri&s, finally, are a number of tribes known as
Tebele, under such chiefs as Boswane (Boshoane) in Leribe and Berea,
Madihotetso in Leribe, and others. They are people of Nguni extraction
who settled in these parts very long ago, and have in part adopted the
customs and language of their Sotho neighbours. Little more is
known about them, their history also being practically unrecorded.

Western Sotho.

This sub-group includes all the Tswana tribes, together with a
number of tribes in the Western Transvaal which are themselves
uncertain whether they should be termed Tswana or not. The Tswana
proper, living all along the fringe of the Kalahari, comprise the Thla-
ping ; the Thlaro ; the Rolong, represented by four sedtions, viz.
the Ratlou, Rathsidi, Seleka (at Thabanchu in the Orange Free State),
and Rapulane, the largest being the Rathsidi Rolong of Chief Monthsiwa
in Mafeking distrift; the Huruthse in Zeerust and Rustenburg
distri&s; some small seftions like the Kubung, Noxeng, and
Kolobeng; and the powerful tribes farther north in Bechuanaland
Protectorate, viz. the Kwena at Molepolole, Ngwaketse at Kanye,
Ngwato at Serowe, and an offshoot of the latter, the Tawana, now
living at Lake Ngami. The Ngwato rule over a number of small
tribes, whom they found already in occupation of the country, and
to whom they are not related, such as the Kaa and Phaleng at Shoshong,
the Khuruthse, the Matswapong, and some Shona (who are not of
Tswana stock).

Amongst those Western Sotho tribes whose Tswana identity is
uncertain, but whose affinities all point to the West, may be mentioned
the Kxatla, represented by the baxaKxafela at Mochudi and Pilansberg
in the West, the Mmakau and Mosethla in the centre (Pretoria),

GROUPING AND ETHNIC HISTORY 6l

the MothSa in the East (Hamanskraal), and some smaller sections ;
the Fokeng or Kwena of Rustenburg, and several other Kwena tribes
in the same distrift (Mmanamela, Modimosana, Mmatau, etc.) ; the
Lete (baxaMmalete) at Ramoutsa in the Proteftorate and in Zeerust
distrift ; the Phalane or Tlase ; the Phiring ; the Taung ; the Thlako ;
the P6; the Tlokwa (mostly in Rustenburg, but with an outlier
at Gaberones) ; and the Hlalerwa. The Hwaduba in Hamanskraal
distrift belong in origin to the Transvaal Ndebele, but have so far
conformed to the ways of their neighbours that they must also be
mentioned here.

The history of all these tribes is imperfectly recorded and scant
in volume. It appears that many centuries ago the first Tswana came
southwards along the edge of the Kalahari. At first they arrived in
small numbers and, finding the Bushmen in occupation, mingled
with and became almost merged in them, the present-day maSarwa
being their descendants. Other early Tswana immigrants degenerated
into the Kxalaxadi of to-day. Of the Dighoya, probably Tswana,
and still visited by Arbousset and Daumas in 1834, no trace remains.
The advent of the bulk of the Tswana, from whom the present-day
tribes are descended, must also date back to many centuries ago.
There are traditions according to which the Ngwaketse, Kwena, and
Ngwato are descended from one original tribe, and other traditions
conneft this latter with the Huruthse and Rolong. The Huruthse,
it may be added, take precedence amongst all these tribes. There is,
of course, some truth in these traditions, which are, however, on the
whole only of the vaguest description. The same applies to the history
of the tribes farther east, for though of late many items of interest
regarding the Kwena tribes of the Rustenburg area have been published
in the vernacular, there is little or nothing of importance available
about the Kxatla and other tribes mentioned above.

Transvaal (Eastern) Sotho.

This sub-group consists of a large main body and several smaller
members which, being numerically so weak, have not hitherto received
much attention. The bulk consists of the tribes of the centre, viz.
those of Sekukuniland, Pokwani, and neighbouring districts. These
are the Pedi and those other tribes, either loosely called Pedi or
speaking the Pedi language, which have been under Pedi control and
influence for a long time, such as the Tau, Kwena (Mongatane and
K6pa), Ntwane, Koni (both those offshoots of Matlala's who migrated

62 N. J. VAN WARMELO

hither from Pietersberg, and those numerous other small groups with
the totems tlou, phiri, phuthi, nare, kwena, nkwe, tau and thSwene, which
are of quite different origin), and the Roka from across the Olifants
River. Farther north, in Pietersburg district, are the tribes of
Mphathlele, Thswene, Mathabatha, Matlala and Dikxale, all of them
Koni from the East, who scaled the mountains round Haenertsburg
and settled on the plains of Pietersburg. There are further the Molepo,
the Tlokwa and some Birwa (from South- Western Rhodesia), and
the big tribes of MoletSe (Kwena) and Xananwa of Blauwberg.

Smaller subseftions of the Transvaal Sotho, showing various
peculiar characteristics, are found in the extreme North-East and East
of the Sotho area. In the North-East are the Phalaborwa, the tribes
of Masisimala, of Mamidja and of Sekororo. The latter are, according
to tradition, of Shona origin, and with them therefore the related
Letswalo, who now live in the Woodbush. Finally there are the
Kxaxa, Mmamabolo's people, and the half-dozen tribes of Lobedu
who have the wild boar (kolobe) as totem and are more closely related
to the Venda than any other Sotho tribes. In the extreme East is a
subdivision formed by the Kutswe, Pai, and Pulana tribes, all except
the first being represented by numerous small independent sections.
They live in or just below the Drakensberg escarpment in
Pilgrimsrest distrift.

The tribes colleftively termed Pedi have had a common history.
The Maroteng, a small tribe of Kxatla origin, made its appearance
in Sekukuniland and gradually subjefted all the tribes living there.
Zulu and Swazi raids almost entirely overthrew this kingdom, but
Pedi power rose to fresh strength during the course of last century.
The history and origins of the tribes thus subjefted by the Pedi,
largely Tau and Koni, has only just been touched upon but not yet
"adequately investigated. Of the Tau it is said that they were originally
Swazi, while the very name of the Koni indicates that they are of
Ngoni origin, but beyond that we know little or nothing. With the
northernmost tribes, matters are equally complicated, for while the
Koni of Matlala and related tribes come from the East, other tribes
appear to have immigrated from the South, others from the South-
West, and the origin of still others remains quite obscure. None of
these tribes seems to have come direftly from the North, except the
Letswalo, who claim a Shona origin. The Phalaborwa again are
most certainly a people from rhe East. It is therefore a highly debatable
question how many of these tribes aftually are of Sotho stock. On

GROUPING AND ETHNIC HISTORY 63

the other hand, there is little doubt that the small extreme Eastern
group, where the dialefts for instance present a striking departure
from normal Sotho, are nevertheless true Sotho, and furthermore
that they are the survivors of the old Sotho population of Swaziland,
to which reference has already been made.

VENDA GROUP

This group is distinguished by a language peculiar to themselves,
though reminiscent of both Sotho and Karanga, and by a culture
sufficiently charafteristic to separate them clearly from other Southern
Bantu. The bulk of the Venda appear to have dwelt in the mountains
of the Zoutpansberg from the earliest times, as they do to-day. They
are a timid, secretive people, and it seems that in their secluded retreat
they have been the guardians of much that is archaic, both in language
and culture. They were shielded from foreign influences by isolation.
The whole country to the south-east of the Zoutpansberg was devoid
of population until occupied by the Shangana-Tonga immigrants
less than a century ago, while the vast area north of the mountains
remains virtually uninhabited to this day. The only line of contaft
with other people, therefore, was to the South and South-West.

To the South the Venda appear to have extended their domain
for a considerable distance. There used to be colonies of them as far
as the Woodbush, amongst Mmamabolo's tribe. In a sense the latter,
together with the other kolobe (wild boar) tribes of that area, may be
regarded as related to the Venda, but the exaft nature of that relation-
ship cannot yet be defined. These southernmost Venda of Groot
Spelonken, Tzaneen, and Pietersburg distrift, who were living far
from the Zoutpansberg mountains, could not retreat thither when
the Shangana-Tonga immigration commenced ; and in the end their
culture and language were swamped, so that little remains of either.
Most Venda Chiefs living along the southern boundaries of Louis
Trichardt and Sibasa distrifts also count numerous Shangana-Tonga
amongst their following; and since, as a rule, the Venda are no
match for the enterprising Shangana-Tonga, the result here also has
been a serious decay of Venda culture. In the south-west, too, the
Venda have been affefted in both language and culture by prolonged
and fairly close contafts with the northernmost Sotho tribes, mostly
Tlokwa.

We may therefore divide the Venda into three sub-groups:

64 N. J. VAN WARMELO

Western, which has been subjeft to Sotho influences; Southern,
which has historical associations with the North-Eastern Sotho tribes,
and more recently has been in intensive contaft with Shangana-
Tonga immigrants, with whom otherwise they have little or nothing
in common ; and Eastern, which has been fortunate enough to escape
being influenced in any way whatever, except of course by the thin
trickle of immigrants and traders from the North and East, an influence
so weak that it escapes dete&ion and is lost to tradition. This last
sub-group has therefore been able to keep traditional Venda culture
almost intaft.

Amongst the Venda, far more than anywhere else in South Africa,
the Chief and his clan are something apart from the rest of the tribe.
The royal clans of the Venda are, with few exceptions, genealogically
related to one another, since most of them claim descent from a some-
what legendary chief, Thoho-ya-Ndou, about whom there are many
stories. His ancestors according to tradition crossed the Limpopo
from Rhodesia, and took possession of the country. They and their
descendants are really the Venda people. They found others in
occupation already, for instance the big tribe of Lwamondo, and the
Ngona, who survive in little more than name. Mphaphuli's people
again are said to have some sort of Nguni origin, hailing from some-
where in the South or South-East. Only the smallest fragment of the
history of all these different tribes has been colle&ed. Such early
history as still lingers refers almost exclusively to the affairs of the
royal house, for commoners do not count. Indeed the tribal divisions
are to such a small extent also cultural divisions that one might truly
say that the Venda have been cut up into tribes for the benefit of
the ruling houses. The relations between tribes are governed by
the feuds or friendships existing between their respe&ive royal houses,
and a tribe consists merely of all those who happen to own allegiance
to a particular dynasty. The Venda tribes, unlike many of the Sotho
tribes, do not differ amongst themselves. There is nothing distinctive
about a tribe as such, and in consequence even tribal names are lacking.
The adherents of a certain Chief are simply known as his people.
The Chief's, name, however, is one which is automatically assumed
by each successor to office, so that the name of the tribe does not
change.

The Western Venda sub-group is composed of the tribes of Chief
Mphefu and his relatives, Chiefs Sinthumule and Kutame, and the
smaller tribe of Musekwa. While Mphefu and about a third of his

GROUPING AND ETHNIC HISTORY 65

people live in a reserve in the Nzhelele valley, the remainder live
scattered about on European-owned lands, both to the west and south-
west of the reserve, where they are in conta& with the neighbouring
Sotho, arid in and along the mountains to the South and South-East
where there is contaft with the Shangana-Tonga. Mphefu also has
more than half a dozen Sotho headmen with their followings subjeft
to him.

The Eastern sub-group, who have been least subjeft to foreign
influences and are no doubt the purest Venda to-day, include the
tribes of Chiefs Tshivhase, Mphaphuli, Lwamondo, Rambuda,
Ne-Thengwe, and Khakhu, all of whom have reserves, and the
smaller tribes of Chiefs Madzivhandila, Mugivhi, and Ne-Tsian$a,
who have no land of their own, and have accordingly forsaken Venda
custom to some extent.

The Southern Venda, as explained above, were formerly linked
up with the tribes farther south, and are now without exception in
close contaft with Shangana-Tonga or Sotho or both, the very tribes
themselves being partly composed of these elements. This sub-group
is composed of the tribes of the following Chiefs, none of whom is
really of much importance : Rasengane, Nngwekhulu, Tshimbupfe,
Masia, and Mashau. The followings of men like Nthabalala, Masakona,
M olema, Mashamba, and Magoro are so mixed that occasionally one
is in doubt whether the name of Venda may still be applied to them.

LEMBA

The Lemba are a small people numbering probably not^ more than
some hundreds of adult males in the Union (chiefly in Zoutpansberg
district) ; while in Southern Rhodesia, according to Mr. F. Posselt,
there may be about 1,500 males, mostly in Belingwe distrift, but also
in Marandellas and Viftoria. In the Union they have no Chief of their
own, but live scattered about amongst the Venda, to whose Chiefs
they are subjeft.

About the history of the Lemba nothing definite is known. They
are strongly suspefted of being Semitic in origin. The reasons for
this belief are, amongst others, that they speak, not Venda as one would
expeft, but a dialeft of Karanga, obviously acquired during a sojourn
farther north ; and that their features, though of course dark, are
distinftly non-Negro. They eat no pork, nor any animal which has
not been kosher-killed by a slitting of the throat; and they do not

66 GROUPING AND ETHNIC HISTORY

intermarry with those not of their race, all such being called
which is identical with washenfr a word used along the east coast
for " wild folk, pagans ". They live by barter only, and used to be
the best craftsmen in metal work and pottery. It is also said that they
were the first to introduce circumcision : at all events, Lemba men
certainly do often take a leading part in these rites.

CHAPTER IV

SOCIAL ORGANIZATION

By A. WINIFRED HOERNLE

INTRODUCTION

THE Bantu peoples have been divided, in the previous chapter, into
five groups of tribes, on the basis of data partly geographical, partly
linguistic, partly historical, and partly cultural. Each of these groups
has sub-groups within it. We shall, therefore, expeft to find the
greatest degree of cultural similarity between tribes belonging to
the same sub-group ; minor cultural differences between one sub-
group and another ; and major cultural differences between the basic
groups. At the same time, even the basic groups have certain types of
institution in common, though these appear in each group in
distinftive forms.

In the present chapter, we shall study those institutions which
together constitute what is known as " social organization " or
" social strufture ". By this is meant the more or less permanent
framework of relationships between the members of a community
which manifests itself in an ordered group-life, with reciprocal rights
and duties, privileges and obligations, of members, determining
behaviour-patterns 1 for each individual member towards other
members, and moulding the feelings, thoughts, and conduft of members
according to these patterns, so that it is only in and through them
that the individual can achieve his personal self-realization and
participate in the satisfaftions offered by the life of his community.

This framework of social organization, or social structure, is
permanent compared with the stream of human lives that, in the
succession of generations, flows through it. Each new generation
finds it there, as an inheritance from untold generations of its pre-
decessors. Yet the framework of social structure is not absolutely
rigid or fixed for all time, but is itself a thing of growth, capable of

1 Tliis term is especially applied to rules of behaviour between kinsmen of different degrees.

68 A. WINIFRED HOERNLE

variation in detail, sometimes giving birth to new developments, at
other times undergoing disintegration and decay.

We shall deal, first with the general, or fundamental, organizing
principles common to the social strufture of all Bantu groups, and
then go on to study the distinftive modifications with which these
general principles are worked out in the social organizations of
different groups. One of these groups, the Nguni, with its sub-
divisions, we shall study in most detail, to have a pattern of reference,
as it were, for the points in which the social organization of other
groups differs from theirs.

The organizing principles underlying the institutions of the
Southern Bantu are to-day, as is shown in later chapters, being
subjefted to considerable strain under the impaft of Western Civiliza-
tion, but they still, to a large extent, control the communal life of the
Bantu. Hence, this chapter will be written in the present tense.

COMMON CHARACTERISTICS
The Trite.

The most characteristic all-inclusive grouping among the Bantu
is undoubtedly the tribe or chiefdom, though organizations on a national
scale are by no means unknown among them. Some of these tribes
number no more than a few hundred, the majority perhaps a few
thousand, while a few are somewhat larger. In Zululand and the
Transkei, some tribes have claimed a membership of more than
ten thousand at various periods of their history, while the more
important Tswana tribes have on the average between ten and twenty
thousand members each.

Each tribe is, in the main, a body of kinsmen, all believing in their
descent from a common far-off ancestor from whom the chief can
claim most direct descent, according to the system of reckoning
descent recognized by the people. But probably it would have been,
even in the very early days, difficult to find a tribe consisting exclusively
of people recognized as kinsmen. Even in the small original tribe of
Zulu, with its 2,000 tribesmen and its few dependent clans, we hear
in the days of Senzangakona (born c. 1757) of at least two groups
of non-related people having been absorbed, the Nzuza of Sotho
extraction and the Ntuli of the Bhele tribe. 1

Membership of a tribe is determined, therefore, more by allegiance

Cf. A. T. Bryant, 1929, 57, 58.

SOCIAL ORGANIZATION 69

to a chief than by birth, and the unity of the tribe depends funda-
mentally on the common loyalty of the tribesmen to their chief.
Chieftainship in consequence is a very important institution in the
whole political life of the people. A popular chief gradually enlarges
his tribe by accession of refugees from other chiefs, while an unpopular
one loses his adherents and becomes " chief of the pumpkins ", as
a picturesque Native saying puts it.

The Household.

Within the tribe, the outstanding social unit is the household, a group
consisting typically of a man with his wife or wives and dependent
children, together with any other relatives or unrelated dependants
who may be attached to him, but composed frequently also of other
combinations of close relatives.

In the Nguni and Tonga tribes these household groups are also
the local units. They are scattered over the tribal territory in fairly
small settlements or kraals (Zulu, umu^t; Xhosa, um{i) y distributed
irregularly at some little distance apart, and generally situated near
a stream or on the slope of a hill. Next to the kraal are the gardens
cultivated by its inhabitants, while their cattle graze on the common
pastures in the vicinity. Within this household, the individual family
or rather a mother and her children stands out definitely as a group
apart, inhabiting its own home (Zulu, indlu). For the Bantu universally
allow polygyny in their social system, and a fair proportion of elderly
men have several wives, as well as other dependants living in their
household with them. Nowhere are these wives equal. " Bantu
social structure knows no equals. . . . The first born of the same
parents is always the superior of those born after him/ 9 1 and this
superiority is extended to his descendants with varying consistency
among different tribes. " Children of the several wives of a polygamist
take their status from their mother. No co-wife of a polygamist is the
equal of a co-wife, but their rank is not determined in the same way
everywhere." 2 The different systems will be described presently.

The Sub-Distria and District.

A number of these kraals, grouped together on the same hill-ridge
or in the same river valley, form a distinft social unit which we may
call a sub-districi (Zulu, isigodi; Xhosa, ibandla), with its own

1 N. J. van Warmelo, 1931, n. * van Warmelo, loc. cit.

70 A. WINIFRED HOERNLE

recognized administrative and judicial system under a headman
(Zulu, unum^ana). Varying numbers of these sub-distrifts are grouped
together into larger districts (Zulu, isifunda ; Xhosa, greater ibandla),
and all come under the jurisdiction of the Chief, who himself controls
such a distrifl and appoints or confirms subordinates in all the others.
In the Sotho tribes, the people tend to colleft together in villages
including a number of different household groups. In Basutoland
and the Transvaal these villages are numerous and as a rule fairly small,
including, say, from ten to fifty independent households. In Bechuana-
land, on the other hand, the members of each tribe live for the most
part in one large central town and several much smaller outlying
villages. 1 The central towns, where the chiefs of the tribes have their
headquarters, are often of considerable size, their population running
into several thousands. Owing to this mode of settlement, which is
determined largely by the scarcity of surface water and the consequent
necessity of congregating together at suitable spots, the cultivated
lands of each town are generally some distance away, extending as broad
belts for many miles across country, while the cattle are kept at special
grazing posts, often a day's journey away by foot Whatever the form
of settlement, however, each household group is always clearly marked
off from the rest.

The Kin (i) Relatives by Blood.

Beyond the intimate circle of the household there is the wider
circle of the kin, relatives either by birth (blood) or by marriage. A
child begins with relatives by descent only ; and these are always
bilateral, relatives through the father, and relatives through the mother.
It is only upon the child's own marriage that he acquires relatives-
in-law of his own. The attitude of his parents to their relatives by
blood and by marriage undoubtedly forms an important basis for
the behaviour patterns built up between the children and their kin.
Kinship bonds play a part of paramount importance in the lives of the
people. None of the tribes has any great degree of economic division
of labour. Each household is to a very great extent a replica of every
other, so that economic interdependence is not one of the main
principles of linking different households together. Instead, kinship
bonds, bonds of common descent, ramify through the country and link
together those, wherever they may be, who claim common descent
from an ancestor, be that on the paternal or the maternal side. This

1 I. Schapera, "The Old Bantu Culture," in Schapcra, 1934, 7.

SOCIAL ORGANIZATION 71

fundamental system of grouping is the same for all the tribes and so is
perhaps best described here.

The type of behaviour expected of the members of a family towards
one another is much more clearly defined and more similar throughout
Bantu society than is the case in modern European society. In all
tribes, the newly-founded family normally takes up its residence
among the relatives of the husband, and the chief control over the
children belongs by custom and in law to the father and his kin. Disci-
pline and authority, economic duties, legal affiliations, and controls,
therefore, play a great part in determining the behaviour of children
to their father and his kin, while affeftion and love find less restraint
in the relation of the children to their mother and her kin.

The father is undoubted head of his family and has complete
authority over his children as long as they remain in his household,
and even afterwards in lesser degree. If head of the household, he
controls the land and the animals to a very large extent, and he direfts
the lives of all his subordinates. He is responsible to the outside world
for all the aftions of the members of his household, prote&ing them
when they are in trouble and answering for their misdeeds. He is
their representative at the tribal courts and is the intermediary between
them and the world of ancestral spirits. Respeft, deference, and even
awe, are noticeable in the behaviour of his children towards him.
The same characteristics may be found, also, in the behaviour of the
wife towards her husband, however much this may be mitigated by
true devotion between them.

Between the mother and her children there commonly grow up
in Bantu society the deepest bonds of affeftion known to these people.
Obedience and respeft for the mother are demanded, but pure un-
selfish love really dominates her attitude to her children. She is
devoted to their welfare, watching over them and protecting them in
all aspefts of their lives. Not so much obedience, therefore, but an
ever-present expectancy of consideration dominates the attitude of
the children to their mother.

Among the children a striCt hierarchy prevails, based on the
seniority which serves as a fundamental principle of behaviour in
Bantu society. The elder brother always takes precedence between
brothers, and so, too, between sisters the privilege of age is maintained.
Between brothers and sisters the sex differentiation often dominates
the behaviour. Sisterhood and brotherhood most often overrule age
differences, and there is a prescribed type of behaviour for a brother

72 A. WINIFRED HOERNLIJ

towards his sister and vice versa. Ties of affeHon link these people
together through life, though in some tribes the marriage arrangements
may bring about strain through a son's marriage depending on the
lolola (or toxaJi) cattle obtained for his sister. 1

Outside this intimate circle of the immediate family, the same
principles of kinship and seniority hold sway. The father forms one
of a close-knit group with his brothers. The latter are everywhere
grouped under a kinship term which we may translate " father " ;
and these " fathers " are distinguished as " great " or " little " fathers,
according as they are older or younger than the child's own father.
To them, as a group, the same general deference and obedience are
due as to the father, mitigated somewhat for the younger ones, and
modified naturally by the experiences of life according to temperament
and general closeness of contaft.

If all these men are a child's " fathers ", their children are all his
" brothers and sisters ". The child behaves towards them on the same
general principles that he has learnt to pra&ise towards his blood
brothers and sisters in the intimate circle of family and household
life, subjeft to the principle of seniority. This is sometimes counted
on the basis of physiological age, sometimes on the basis of the status
of the parents, irrespective of the age of the children themselves.
The same general system applies to the children of a number of sisters,
and justifies the statement that " parallel cousins ", i.e. children of
brothers on the one hand and children of sisters on the other, behave
towards each other like blood brothers and sisters.

The behaviour to a father's sister is definitely controlled by her
membership of the same family as the father. If he is " father ", she
is " female father ", with the same background and traditions as he.
She knows the peculiar ways and traditions of the paternal home far
better than the mother is thought to do, so that on important occasions
she may be called in to see that family ways are maintained. The most
formalized relationship between such women and their brother's
children is probably to be found among the Tonga and Venda.

A maternal uncle similarly plays his pan in the lives of his sister's
children, on the basis of the behaviour controlling his relation to his
sister. If she is " mother ", he is " male mother " (malume) in most
of the tribes, a relative of the mother's family of the same generation
as she, and sharing in the whole gentle protective attitude of the mother
to her children. His role is more formalized in the Sotho, Venda,

1 See below, pp. 113 f.

PLATE V

(A) Mochucli village. Kxatla Re-serve, B.1*.
TERRITORIAF. ORGAN I/A FION

[/. Schaficta

[face p. 72

PLATE VI

(a) Natal Nguui

l&A.K. tuul H.

(6) Transvaal Sotho

(Kxaxa of Mphathlela's Location, Pictersburg)
NATIVE HOMESTEADS

L,V. J. van Warmth

SOCIAL ORGANIZATION 73

and Tonga groups than among the Nguni, for in these groups the
families of brother and sister are drawn together in marriage, while in
the Nguni tribes no marriage is allowed with a blood relative less than
at least four generations removed, be it on the father's side or on the
mother's. This difference in the attitude to " cross-cousins ", i.e.
the children of a brother and a sister, ramifies in very important ways
throughout the whole of social behaviour.

Grandparents are inclined to spoil their grandchildren every-
where, and the Bantu are no exception to this rule. But the patrilineal
principles which play so big a part among these people, giving the
father's relatives the greater responsibility in the bringing-up and
control of the children, tend to make the paternal grandparents strifter
and more critical of the behaviour of their grandchildren than is
necessary in the case of the maternal grandparents.

Here, then, is a large body of kin drawn intimately into contaft
with the lives of each generation. Looking out of the world from his
own home, the Bantu child knows where he may seek hospitality and
succour of every kind ; where, also, he may of right be called upon
to render assistance in case of need. The barriers of reserve shutting
off human beings from one another are largely down so far as these
classes of relatives are concerned, so that for economic assistance,
for friendly counsel, in time of sorrow and in time of joy, these are the
natural categories of people to turn to, the core of people with whom
one is close-knit from birth in a web of reciprocal rights and duties.

The Kin (if) Relatives by Marriage.

Marriage being a contraft between two families as much as between
the two individuals chiefly concerned, the behaviour of a large number
of people to one another is changed by each marriage that takes place.
In principle, the behaviour of people related to one another by marriage
is somewhat restrained and striftly regulated. Relatives by marriage,
when they meet at each others' homes, behave most formally and
corre&ly. Meat and beer are prepared as for honoured guests, but
often there is no intimacy and ease of real human affe&ion. This formal
behaviour applies especially to relatives older than the husband or
wife, contaft with juniors being in all cases much less strained. Correft
behaviour is especially demanded of a man towards his mother-in-
law and of a woman to her father-in-law and all men of his status.
In the Nguni tribes, the behaviour especially of wives to their elder
relatives-in-law is so striftly regulated that there is a special term,

74 A. WINIFRED HOERNLfe

hlonipha, for this behaviour. Hlonipha afFefts all aspefts of a married
woman's life, regulating her movements about the household into
which she is married, her liberty in eating, even her vocabulary for
everyday things.

Other Groupings.

Other forms of organization, universal among these tribes, are
more formal descent-groups, such as lineage, clan, kxoro, and also
groups based on age and, in a less organized manner, on sex. It is on
one or other of these types of grouping that the military organization
rests ; but, as there are differences from group to group, these types
of social strufture are best dealt with separately for the different
groups. Social groups for fellow-craftsmen are so casual that they are
hardly to be called formally organized groups.

NGUNI GROUP

The Nguni must be clearly distinguished from all other groups of
tribes in their social organization, in that they rigidly prohibit marriage,
or sexual relations of any kind, with people related through any of
the four grandparents. This fundamental rule has most important
repercussions through the social strufture. In these tribes also we
find the most complicated organizations of the household. The social
system of the Zululand tribes was subjected to a severe strain through
the wars of Shaka and his immediate successors, and many of the
fundamental institutions were developed there in somewhat novel
ways, whereas the Southern Nguni, in spite of the disturbances which
eruptions of people from the North brought in their train, and in
spite of the many wars "waged between Bantu and European, have
preserved their social system in a less altered form. It will, therefore,
be as well to describe the Southern Nguni system first, and then indicate
the differences which we find farther north, without, however,
implying that the Northern Nguni system is a development from
precisely the situation which we find among the Southern Nguni.

(a) Southern Nguni
Urny.

Among all the Southern Nguni tribes, we find that each household
(urny) is -an independent territorial settlement with its own name.

SOCIAL ORGANIZATION 75

In each the cattle kraal forms the pivot of the arrangement of the huts.
The principal hut of the household faces the entrance to the cattle
kraal ; and the space between them, the ikundla (courtyard), is essen-
tially private to the owners of the w/wp, those there by right of birth.
In a semi-circle to either side of this chief hut are ranged the other
huts of the umfi. Formerly the um^i tended to be larger than it is
to-day. It consisted then of a patriarch with his wives and dependent
children, his married sons with their families, and possibly also un-
related married dependants (Xhosa, v(induna). An unify then, might
contain twenty to forty adult men related in the male line, each with
his family, whereas to-day " an average um^i contains four to five
adults, and four children." * The basic characteristic of this urny
is the overwhelming stressing of the bonds linking the adult males and
all the children, through the emphasis put on descent from a patrilineal
ancestor. Though from one point of view we have in this settlement
a number of " families " represented by the living-huts (i^inJlu) of
the married women, from another point of view we have a group
of close-knit relatives guarding their lineage traditions against " out-
siders ", i.e. the women brought by marriage into the lineage.

Polygyny is an ideal which every well-established man strives to
attain, and the more important households reveal a complex grouping
of wives. The first wife married is always the principal wife (inkosika^i) y
except in the case of a chief, whose principal wife is often chosen
only after he is well-established in the chieftainship, and whose
khay. cattle are collected from all the principal households of the tribe.
Among the other wives, one other stands out in the typical polygynous
household. She is the " right-hand wife " (urnfcqi wasekunene), so
called from the position of her huts in the household, the Xhosa
counting right and left from the point of view of a person who looks
out from the hut door towards the cattle kraal. 2 Any other wives
married are affiliated to the principal wives as subordinates and are
called Qadi. Each family, or " house " (indlu), whether principal or
supporting, forms an independent unit with its own property, except
in the case of a " seed-bearer ", who may be married if the wife of a
principal house proves to be barren. The house of a woman so married
has no independent property, and any children she bears are regarded

1 Cf. M. Hunter (1936), if, and also for whole detail of description.

The Zulu distinguish right and left from the point of view of one looking up into the
iMifi from the cattle-kraal gate. Thus, what is the " right-hand " se&ion for the Xhosa is
the '* left-hand " section for the Zulu.

7 6 A. WINIFRED HOERNLE

as the children of the woman for whom she is " seed-bearer ' \ l An
important chief may have yet another type of wife. He marries a
woman, the ixiba or isipnJa wife, and places her, together with
dependants, within the principal ump (or an um^i called by the same
name) of his father, or grandfather, or other important relative, and
thus keeps " alive " a household with which much of tribal history is
closely linked.

Among the Xhosa tribes, the widows of a man usually remain
at their late husband's ump. Any children they bear at his kraal
after his death have the same status as children begotten by him. Many
widows, however, establish a separate urny with their grown-up
sons and each becomes the principal woman in her um^ while yet
others return to their childhood homes (kayo) where they join the
ranks of the numerous group of aS&zp, i.e. husbandless women
(whether unmarried, divorced, or widowed) who have borne children.
Such women, living with their own relatives, have a much easier
time than do women living with their husbands among relatives by
marriage. Among the Mpondo and related tribes, the widow, if still
of child-bearing age, is taken by the male next-of-kin to her husband
(father and ascending kin excepted). Any children born of this
union are regarded as children of the original husband. This is the
custom of ukungena.

The two principal wives, together with their supporting wives,
may each have quite a large seftion of the a/rap within which to
some extent life runs separately ; or, when there is a grown-up son
in the right-hand house, or this se&ion of the household is that of a
chief, a separate ump may be founded for it at some distance from the
principal um{i, or even in some different part of the country. Each
um^i has its own name, and each important family hut within it may
have its own name. The principal son of the family of the um{i has
the right to use this name, and many names of umji of important
chiefs have become the distinguishing names of groups of people,
since the adherents of a headman or chief may call themselves by the
name of the headquarters of their leader. This is true also of the
Northern Nguni tribes.

Within the amp, great or small, the principles of conduit already
described are constantly at work the principle of seniority ; the
principle of die categories of blood kin, with distinftive patterns

1 W. T. Browidee, " The Transkeian Territories of South Africa," /. Afr. Soc., xxiv (1914-5),
116.

SOCIAL ORGANIZATION 77

of behaviour towards each category ; the principle of sex differences ;
the principle of distinction between kin by blood and kin by marriage.

Unless she is the inkosika^i of an umji, a woman has a large duty of
hlonipha to the family into which she has married. She must not
intrude upon the men's side of the hut ; she must not cross the court-
yard past this side of the hut ; she must not unnecessarily cross the
courtyard itself or enter the cattle kraal. She must not mention the
personal names of the elder relatives of her husband or of her husband
himself; nor even use words containing the principal syllables of
their names. She must defer to them in every way. Should she meet
them in the pathway, she steps aside ; she does not eat freely with them
nor appear in working undress before them until she is thoroughly
established in the umg. by bearing children for them. Until special
ceremonies are performed to release her from the restriftion, she does
not drink milk nor eat meat at her husband's home. These eating
restrictions may be abandoned after the performance of an appropriate
ceremony, but those of speech and behaviour remain until a woman is
head of her own umji.

Co-wives of a polygynist may have a strong fellow-feeling for one
another. They are subjeft to the same restrictions ; they share in the
maintenance of the domestic life and care of the children throughout
the UM(I. Yet, beneath this co-operation, there is the still more closely-
knit little group within each indlu, the group of a mother and her
children ; and the supreme aim of a woman's life is to further the
interests of her own children first and foremost. There is, therefore,
an underlying tension and strain between the families of an um%i,
which often reveals itself in mutual accusations of sorcery and witch-
craft.

Whilst the ump is thus to a large extent a little world of its own,
with the head (unumjana) controlling the conduft and fortunes of its
members, it is also the smallest link in the chain of other types of
organization which play important roles in the life of the community.

Ibandla and Ihundla.

The group of people living in an ump of some size constitutes what
is known as the itandla of the ump head. The adult men, especially,
form a distinft social unit for many community purposes. In economic
undertakings, on social occasions, for legal obligations, in the army,
this little group constitutes the smallest unit of organization and is
treated as a whole.

78 A. WINIFRED HOERNL&

A number of households, generally closely related to one another
and living in a conveniently marked-off area, such as a valley or ridge
of a hill, constitutes the next group in the territorial, social, and political
hierarchy. This little group of households, but chiefly the adult
men thereof, constitutes a " great ibandla ", that of a petty chief
perhaps the head of the principal clan represented in the cluster of
households ; perhaps a friend or dependant of the Paramount Chief,
or some remote relative of his who has been given this small district
to control. The umji of the petty chief is called the ikundla of the
distrift. It is the social, administrative, judicial, and military centre
of the little area.

Many of these little districts are organized into a few much larger
distri&s, known also as umhlaba or ibandla. 1 The headquarters of a
chief, either a member of the ruling clan of the Paramount Chief,
or the head of a clan which has khon^ i.e. submitted to the supremacy
of the Paramount Chief in return for domicile in his area, is the
organizing centre of this area, and the people in it constitute the
ibandla of the chief. At the head of the whole hierarchy comes the
Paramount Chief, with all his tribesmen organized into amabandla 2
under him.

The amabandla of the Paramount Chief are organized in the follow-
ing way. The tribal area is divided out among different " houses "
of the Chief, who has several household settlements in different parts
of the country. The principal son of each house (or his descendant)
claims all the people in his distrift as his ibandla, and these may be
organized into several minor amabandla under sub-chiefs and head-
men. Dues in meat, beer, and labour, which are owed to the Para-
mount Chief, are paid not to his far distant headquarters, but to the
member of his family representing him in the area. Headship of an
ibandla is inherited, but the area controlled by the head of an ibandla
is not a fixed unit. If people do not like the head of their ibandla,
they may take their dues and their cases to the next headman, and if
this becomes a custom his area will be enlarged, that of his rival
decreased. In course of time, there may be a reorganization of the
headship, since each succeeding chief has sons to place, and often
puts a member of his own immediate family in the area controlled
by descendants of his brothers or other remote relatives. These
latter then sink to the rank of sub-chiefs, and the more immediate

1 The term umhlaba is sometimes used as referring to the territorial area ; whereas ibandla
refers to the group of human beings.
Plural of ibandla.

SOCIAL ORGANIZATION 79

representative of the Paramount Chief controls the area, if his per-
sonality enables him to grip the affeHon of his people. These
amabandla all have recognized names, sometimes the name of the
head of the clan under whom the area was first established, sometimes
the name of the urny. of the chief of the royal clan who first administered
the area. 1 In this latter case, in some of the tribes, e.g. Gcaleka,
Ngqika, Bomvana, the ibandla name may be the name of a favourite
ox in the chief's herd, and hence this name and the name of the group
is called the inkabi (ox). 2 Among the Mpondo, this custom is unknown :
ibandla and house names of chiefs are either descriptive or the names
of the first founders.

The subordination of all distrift chiefs to the Paramount was
formerly marked by the obligation of the district chiefs to pay the
Paramount death dues and fines for murder and witchcraft, and to
attend military reviews and first-fruit ceremonies. The distrifts
were in the past the administrative areas ; they were to a large extent
settled by a body of related people ; and they played a large part in
the social life of the people. Even to-day, at a feast, meat and beer
are laula (called out and distributed) by amabandla (or inkabi)
so many pots of beer to this ibandla and such and such joints to
that. At small feasts, meat and beer are laula by the minor amabandla
of private individuals and chiefs' sons who have no civil authority;
at bigger gatherings, they are laula by the amabandla of distrift
chiefs. 3

Among the Mpondo, the army was formerly organized on an
ibandla basis. Each petty headman collected his followers and made
his way to his immediate superior, who in his turn led his ibandla
to his superior, until finally in a few divisions, four or five, the army
gathered at the tribal headquarters. 4

The whole system of territorial settlement, only partly based on
kinship ties, is definitely co-ordinated through the dominant position
of one clan, members of which keep a grip on the different areas,
and render loyal support to the head of their clan through the all-
pervasive principles of rank, based on a hierarchy of seniority. From
time to time, however, a dominating personality may break through
this seniority principle and, wresting the leading role from a senior
house, may arrogate it to himself and his own " house ".

1 For this whole account, cf. Hunter, 1036, 378 ff.
Cf. P. A. W. Cook, 1931, 19 ff.
Cf. Hunter, 1936, 364 ff. ; and Cook, loc. cit.
4 Hunter, 1936, 400 ff.

8o A. WINIFRED HOERNLE

Isiduko or Clan.

The clan is a kinship group. It consists of a group of people claiming
descent, and taking its name, from a common ancestor in the male line.
Within the clan is a number of lineages (Xhosa, usapo or intsapo),
the members of which generally have a common grandfather in the
male line. Lineages and clans are named after a common ancestor, but
in addition there are isifongo, praise-names, conne&ed with each clan
and each lineage, which are used on all ceremonial occasions and when
it is desired to honour the members of the group. 1 These isi&ongo
names are often names of still more remote ancestors and are valuable
as revealing links of connexion between otherwise separate clans.

In any distrift a large number of people may be members of one clan,
that of the chief or headman, but the two types of grouping, territorial
and kinship, never coincide absolutely, and members of one clan may
be found even in several different tribes.

The lineage and clan are extremely important groups, the lineage
being, next to the household, the most intimate social group and
principal ritual group within the tribe. Among the Mpondo, the lineage
is the group within which alone milk is drunk, though the Xhosa
drink milk in any lineage of their clan. In all the tribes, marriage and
the custom of ukumetsha (premarital sexual relations) are strictly
prohibited within the clan, though lineages are constantly breaking
away to form new clans and then marriage is allowed. The clans have
a regular order of precedence, that of the Paramount Chief being most
direftly descended from the common ancestor of the tribal nucleus,
though there are always also clans of alien origin whose affiliation is
purely political, and whose position in the hierarchy a matter of
historical tradition.

The lineage undoubtedly is the most important purely kinship group
among most of these tribes, the group which exercises most social
control over the people and within which there is the most intimate
social contaft and the most stable system of reciprocal rights and duties.
Among the Xhosa alone would the clan seem to be a close-knit body
within which milk customs, military organization, and also ritual
customs prevail. While the Mpondo army was organized on the
ibandla basis, the Xhosa army was organized on a clan basis, 2 each
clan constituting a seftion of the force. Even to-day it is by no means
unknown for representatives of a Xhosa clan, wherever they may. be

1 Hunter, 1936, 52 f. * J. H. Soga, 1932, 68.

SOCIAL ORGANIZATION 8l

domiciled, to combine in sending a sacrificial offering of oxen to the
remote headwaters of that river in the Transkei which the clan members
claim as the first headquarters of their ancestors. The Paramount
Chief is, in all the tribes, the leading representative of the dominant
clan in the tribe, the " royal " clan, to which most of the important
chiefs under him also belong. The royal clan of both the Gcaleka
and the Ngqika tribes is the Tshawe ; of the Thembu, the Mtande ;
of the Mpondo, the Nyausa. Taking the Gcaleka tribe as an example,
we find that there are some twenty-five clans all tracing their descent
from the common ancestor, Xhosa, and strictly ranked according
to their direftness of descent. These clans have hived off from time
to time from the common matrix, as it were, sometimes because of
personal feuds and rivalries, sometimes because of intermarriage be-
tween the different lineages or " houses ", after the common descent
was at least four generations removed from the common ancestor.
These royal clans are grouped together as the inTshinga division
of the Gcaleka people. Another group of clans belonged originally
to Sotho or other non-Xhosa tribes, but to-day are subordinate to the
Gcaleka chiefs. They form the division called the iQauka, and though
they are the commoners among the clans they have to-day the con-
trolling influence in tribal affairs. 1

Age and Sex Groupings.

Though the hierarchy of age is as important among the Southern
Nguni as in any of the other groups, and though in many of these tribes
(e.g. Xhosa and Bomvana) circumcision and group initiation cere-
monies remain, while they have long since died out in other tribes,
(e.g. the Mpondo and all Northern Nguni tribes), yet no tribe-wide
organization on the basis of age exists among them. Within the house-
hold, rigid precedence is given to seniority, as we have seen, and so
also within the ibandla both on social and military occasions, but the
age grouping is subordinate to some other (clan or territorial) grouping.
It is among the Northern Nguni and tribes of other groups that we
find more institutionalized groupings on the basis of age. Groupings
on a sex basis form naturally on all social occasions, and sex differentia-
tion plays its part through the whole social system, but, as far as we
know, there exists in the Southern Nguni tribes no formalized
organization on a sex basis.

1 Soga, 1932, 22.

82 A. WINIFRED HOERNL&

() Northern Nguni

It has been shown in the previous chapter that probably three
different migrations of tribes are represented in the Zululand-Natal area,
and it is quite possible that there were originally many differences of
custom and perhaps also of social organization. Whether such exist
to-day we do not at present know, since we have no scientific study
of the Southern Natal tribal groups ; and the changes brought about
in the social system of all these tribes by the rise of the Zulu power have
been so marked, that for the present it is justifiable to give an account
of the system imposed on this area by the Zulu. For, even in spite of
vast changes brought about by white dominance and administration,
this social system shows a remarkable tenacity and to a large extent
still controls the lives of the tribes, not only in these special areas,
but also in other parts of Africa where offshoots of these people have
settled down (e.g. Ndebele, Ngoni).

Umuji.

As with the Southern Nguni, so too with the Northern : the house-
hold (Zulu, umuii\ or kraal, is the primary territorial unit. The
organization of this umu^i may, among chiefs and headmen, be even
more complicated than that of the Southern Nguni, but the funda-
mental strufture is the same. The core of the umuy, is a body of patri-
lineal relatives whose wives are all from clans not related otherwise
to them. The individual family, in its own indlu, or home, is a strongly
integrated group, owning its own cattle, its own lands, and its own
granaries, and forming through life the group with the most deep-
set emotional ties and common domestic experiences.

These families have different status within the umu^i. To-day,
the original outer fence encircling the whole umuy is gone praftically
everywhere, but typically there was such an encircling wall of palisades,
reeds, or bushes. With or without fence, the umiqi is still organized
into its different divisions all circling round the cattle kraal (isitaya).
Important polygynous households are divided into three seftions in
the place of the Southern Nguni two. 1 At the top of the umwp, opposite
the great gateway of the outer fence (where this exists), is the seftion
of the inkosikap, the mistress of the umup. This seftion is known from
its principal hut as the indlunkulu y but it may also contain the hut
of a seed-bearer (umloGokap), the huts of married sons, and unmarried

1 A. T. Bryant, 1923, 47-51.

SOCIAL ORGANIZATION 83

grown-up children, as well as the homes of servants or dependants in
the background. And the whole may well be screened off by individual
fences, secluding the living huts from the general gaze of members
of the umuii To the left l of this indlunkulu seftion is the ikhohlo,
or left-hand branch, of the polygynous family, with its principal
wife. She is in important families the first wife married in the youth
of the kraal head, but is relegated to second position in the kraal when,
later in life, he marries the daughter of some important man and places
her in the indlunkulu. The ikhohlo wife may also have several subordi-
nate wives (a/wa6/6/), as well as married sons and dependants, in
her seftion of the umuii. Usually, at the time when the indlunkulu
is founded, the headman also organizes the third seftion of his family,
the inQadi. One of his wives is appointed head of this seftion, with
others subordinate to her. This seftion is most definitely an appendage
of the indlunkulu^ and is called on to supply an heir if none is born
to the inkosika^i or to the wife specifically appointed as her seed-
bearer, whereas the ikhohlo seftion, once the indlunkulu. is founded,
becomes to a very large extent loosened from the other seftions of the
umu%i y and tends to move off to found a separate umu^ as soon as a
married son is able to take over the responsibility for his mother and
her dependants. In addition to these three important seftions of his
polygynous family, a chief may keep " alive " kraals of his father and
grandfather by placing in them i(v(inda in the same way as is done by
the Southern Nguni. 2 Within the umw(i there obtain the same rigid
distinftions as among the Southern Nguni between relatives by birth
and wives ; the same general kinship system with its principles of
seniority and deference to those of higher rank and status ; and the
same rigid system ofhlonipha for married women.

Isi&ongo.

The kin also play the same part as they do farther south, and the
exogamous patrilineal clan with eponymous name controls the more
intimate social life of the people, and definitely constitutes a group
within which sexual relations of any kind are prohibited. The clan
is called isfoongo, a word referring more particularly to the name
of the group, while the terms ufilo&o, umquba, or w*{a/o, indicate the
group itself. Each clan, in addition to its own eponymous isfoongo,
has an address name, called the isithakaielo, which is the name of a

1 Scil. " left " as one looks up from the great gate of the cattle-kraal ; cf. above, p. 75, note.
Bryant, 1923, 47-51.

84 A. WINIFRED HOERNLfc

famous ancestor in the ancestral clan from which many have sprung. 1
Many different clans have Mnguni for their isithakaielo, which is the
justification for calling the whole group "Nguni", since all the
tribes, or at least very many of them, seem to claim a far-off origin
from one unknown ancestor, Mnguni.

To-day, the government recognizes over 200 tribes (isvpve) in
Zululand and Natal. These are composed of a central core of representa-
tives of the chief's clan, together with seftions of a number of other
clans which are often more numerously represented in another tribe
where one of their clansmen is chief. This situation is the aftermath
of the rise of the Zulu political domination and its destru&ion by the
Europeans. Groups of people were scattered hither and thither and
now have settled down in smaller or larger tribes which are only
partly bodies of kinsmen.

Within the tribal area there is the small distrift area (isigodi) with
its headman, or induna, whose umu?i is the administrative centre ;
and the larger district (uifundd)^ of which there are a few in each
tribe, ruled over by the head of the clan dominant in that particular
area and subordinate to the Chief of the Paramount's clan.

The Zulu kings, from Shaka to Cetshwayo, maintained control
of their conquered domain by planting royal kraals at different spots
in the conquered territory, maintaining there a military force, and
putting in residence some close relative of their own who might
be trusted to act as their " eyes and ears " and represent their interests,
though the drudgery of administration was done by specially appointed
officials (isfinJuna). Uncles, brother, aunts, mothers, and wives might
be so placed in these royal kraals, called amakhanda? when they were
specifically the headquarters of certain regiments. More remote tribes
were simply decimated and their wealth carried off, or they were laid
under tribute to be paid from time to time as the di&ator ordered.

Age Grouping.

The hierarchy of age is more organized here than farther south.
In the large utnuy. of a polygynous man, the children of similar age will
constantly be found together and form, together with their age mates
of the little sub-distrift around their home, the intanga, the age set
the members of which spend so much of their childhood days together.

1 Bryant, 1920, 681 ff.

Plural of ikkanda. In the amakhanda the royal family occupied the upper part of the
kraal (in quarters called the isigodlo), while the regiments occupied the two wings of die ikhanda.

SOCIAL ORGANIZATION 85

From the age of about eight, in the heyday of Zulu power, these
young boys might act as mat-and-food carriers for their elders when
going to regimental meetings. Between the age of 18-20 they were
regularly enrolled by the king into a regiment, or fautko, with its own
name, its own regimental officers from a senior regiment, and its own
military headquarters, which might be a newly established ikhanda,
or a seftion of an already existing ikhanda^ to reinforce the numbers
of an older fautho. These amadut/io were formed very often in
the days of Shaka and Dingane, less frequently later ; but they are still
being formed to-day, though the system of military kraals is gone.
When the regiments are collefted together, they constitute the army
(imp'i) of the people. In the military drill the soldiers frequently march
many abreast, and such lines are called vpcheme, each consisting of
from three to eight or more men. Yet another division recognized
is the iviyo or isigada, a company of about 500 men. For mobiliza-
tion purposes the soldiers are assigned to one of a few royal kraals
within easy distance of each other. A divisional commander is at the
head of all the men so assigned to him. Sometimes the men are
organized in their regiments, sometimes the divisional group of each
regiment forms under its divisional commander. 1 During the Zulu
domination, these regiments were undoubtedly mainly a fighting
force. The younger regiments especially were a restless, unruly element
in the state. They did, however, also guard the king's cattle ; they
built their ihhanda and kept it in order ; and they might be called
on at any time to hoe the chiefs gardens ; to lay out a new head-
quarter utnu(i ; or to do any other work that the king might order.
From the regiments he drew his messengers, his administrative officers
for the distrifts, and his councillors. All these officials were drawn
from many different tribes in Zululand, as they were singled out
for bravery, or for some piece of work for the king which drew his
attention to them.

The independent Zulu kingdom really came to an end when
Cetshwayo was defeated by the British in 1879. To-day, the descend-
ants of Cetshwayo are still to a certain extent recognized as the supreme

1 In 1914, at the ihlambo of Chief Solomon Dinuzulu, 10,000 warriors were present at his
chief kraal, Mahashini, organized in the following regiments in order of rank, the highest
coming first: imbhokocrebomvu ; ufelaphakati ; udakwa; ucijimpi; uvukayibambe ;
inqab'ukucetshwa ; upondolwendlovu ; intabayezulu. These regiments belong to five
divisions, linked with five royal kraals : (a) gqikazi, linked with the gqikazi kraal to which
belonged both Cetshwayo's and Dinuzulu's mothers ; (6) mpisendlini Mahashini, Dinuzulu's
kraal ; (c) kubhaza, affiliated to Cetshwayo's Usutu kraal ; GO nodwengu, Mpande's chief
kraal; and () Undi, kraal built by Cetshwayo.

86 A. WINIFRED HOERNLt

representatives of all the Zululand tribes, but there is a tendency for
these tribes to split by grouping themselves round two different lines
of descent of the Zulu royal clan, 1 the descendants of Senzangakona
and those of his brother Sojiyisa. The one seftion is called the Usutu
seftion, after the name of one of Cetshwayo's umuy.\ the other,
the Mandlakazi seftion, after an umup of Sojiyisa. 2

Nguni Offshoots

Most offshoots of the Nguni group, such as the Southern Transvaal
Ndebele, the Swazi, the Rhodesian Ndebele, and the Ngoni of Nyasa-
land and Tanganyika, have preserved the same fundamental social
system. But there are one or two new principles which must be noted.
Among the Southern Transvaal Ndebele, each isfoongo, or clan, has a
special species of animal linked with it known as its {//a, or taboo,
which may not be named or eaten or used by the members of that
clan. This feature has undoubtedly been copied from the surrounding
Sotho tribes. 3 The Rhodesian Ndebele and the Nyasaland Ngoni
have introduced a new hierarchy of rank resulting from their conquest
of indigenous tribes in their new homes. Both groups have attempted
to maintain their purity of blood to some extent, and both have
maintained an amazing pride in what they consider their stronger social
system and their tradition of military prowess and might. The Rhode-
sian Ndebele recognize three social grades, the pure Ndebele, a middle
rank with Ndebele fathers and Shona mothers, and finally the
commoners, the Shona rank and file, merged in their social system.
The situation is somewhat similar among the Ngoni.

SOTHO GROUP
Kin by blood and ly marriage.

The Sotho tribes all allow certain types of kin to marry. This
different attitude to marriage between kinsfolk is reflefted right through
the social system, and makes for important differences in the social
groups, even where these are based on the same principles as among

i T, rS enzan g a k na Mpandc Cetshwayo Dinuzulu Solomon boy (Usutu kraal)
Jama ~~~LSojiyisa Mapita Zibebu Bokwe (Mandlakazi kraal)

* The Mandlakazi section sent no official representative to the ihlambo ceremony after the
death of Solomon Dinuzulu, but on the occasion of the visit of the Prince of Wales, the Maqdla-
kazi regiments fell in with the Usutu regiments and formed one great ukumbi, or crescent of
men, with them.

* N. J. van Warmelo, 1930, 15.

SOCIAL ORGANIZATION 87

the Nguni. There is the same basic type of classification of kin into
relatives on the father's side and relatives on the mother's side.
Seniority is stressed, and patrilineal filiation is dominant in the social
system. Relatives by marriage are, however, not striftly marked off
from relatives by blood, but, on the contrary, are to a large extent
the same people. There are not the same rigid categories of relatives,
but a more complex range of possibilities of behaviour towards the
same persons ; e.g. the maternal uncle may also be the father-in-law,
and so, too, the paternal uncle may be the father-in-law. Such com-
binations are quite impossible among the Nguni.

All Sotho tribes show a strong preference for a man's marriage
with the mother's brother's daughter, though they also encourage
marriage between the children of two brothers and allow marriage
with the father's sister's daughter and with the mother's sister's
daughter. This last type of marriage is, however, distinctly less
frequent than the other types and in certain tribes is forbidden. 1
Where the bride comes to her husband's home, not as a stranger,
a non-relative, but as a close kinswoman, known probably from child-
hood at least to the parents of the husband, if not to the husband him-
self, there is bound to be a less strained system of behaviour within
the household. True, the relationship between mother-in-law and
son-in-law is carefully controlled and regulated ; so, too, is the
behaviour of a bride to her father-in-law. But we find little of the
rigidly enforced hlonipha system of the Nguni tribes. In the Sotho
tribes, most of the restriftions on behaviour and in address can be
terminated by the performance of a ritual ceremony once the couple
has settled down.

A brother and sister are obviously very closely linked together
in the Sotho tribes when the son of one marries die daughter of the
other. They are intimately linked, however, in still other ways.
A father frequently pairs off his sons and daughters, as far as may
be possible, telling the son that he is to use the boxadi cattle 2 of such
and such a daughter in order to obtain a wife for himself. The children
of the sister, then, have definite claims on their mother's brother,
since he has benefited quite considerably from the boxadi cattle handed
over by their father. They have an honoured position in his home, 3

1 W. Eisclen, 19283, 81.

1 Boxadi is the term for the cattle given by a man to his wife's father when he marries her ;
cf. below, Chap. V.

* The children of a woman for whom no boxadl has been transferred are brought up at
their maternal uncle's or maternal grandfather's home, but they are there as dependants and
their position is much lowlier.

88 A. WINIFRED HOERNLt

and he may be called on to help in providing loxadi for his sister's
son. In these tribes, also, the maternal ancestors are thought to inter-
vene of right in the lives of their daughter's descendants, and the
maternal uncle may be called on to sacrifice for his sister's children
when they are ill. He receives a beast when his sister's daughter is
married, in return for the services he has performed for her. Between
his children and his sister's there is a joking pattern of behaviour,
which indicates immediately that the behaviour is exposed to stresses
and strains. When two families are so intimately linked and anything
goes wrong with one, the other is bound also to be upset. The sister's
children, in this situation, have the whip-hand, since the brother
established his home with cattle she brought into the family, so that on
occasion the maternal uncle's children may feel themselves awkwardly
placed in relation to their cross-cousins.

The Village.

In all three Sotho sub-groups, we find true village life in place of
the independent households of the Nguni tribes. In Basutoland, these
villages (metse, sing, motse) are small, rarely having more than 250
inhabitants and mostly not more than a score. 1 The inhabitants of a
motse are mostly relatives and chiefly in the paternal line, though not
necessarily so. Each close group of relatives builds its huts round the
lesaka, or cattle kraal, so that in a large village the individual groups
can be distinguished by the little paths separating the different hamlets
from one another, while the charafter of the whole as a unity is revealed
in the common khotla (Tswana, kxotla ; North Sotho, kxoro) which
is, as a rule, to one side of the village and usually has a tree for shade.
This khotla is the common gathering place of the men, and the centre
of village administration and control. The Northern Sotho system
is essentially similar, but in the Tswana tribes much greater develop-
ment is seen.

The Ward.

In these Bechuanaland tribes, the people are divided into wards,
consisting of a number of families united under the leadership of a
kxosana (headman), whose position is hereditary in the male line.
Most of the wards are named after some distinguished ancestor of
their headman. The constituent families are in most cases direftly

1 Cf. F. Laydevant, 1931, 216.

SOCIAL ORGANIZATION 89

related either by birth or by marriage to the headman, although this
is not necessarily the case. Intermarriage between members of the
same ward is permitted. Children normally belong to the ward of their
father, but in exceptional circumstances can become members of some
other ward, e.g. that of the mother or wife where this is different from
their own. The ward is essentially a localized administrative unit.
The central feature of each is the cattle kraal (lesaka\ and adjoining
it the kxotla^ or court, where lawsuits are heard and other local business
is dealt with under the supervision of the headman, assisted by the more
important heads of families. The headman is responsible to the
chief for all that goes on in his ward, and the headmen of all the
wards together constitute an advisory council to the chief, being
consulted by him in all cases of emergency. 1 Members of the same
ward form a distinft seftion within the mephato (age-regiments)
under the headship and immediate authority of the one most senior
in status. In tribal gatherings the people are sometimes ordered by the
chief to sit according to their wards ; and one or more whole wards
may be allotted to a son of the Chief as his particular adherents. 2
Among the Kxatla there are some sixty-eight of these wards in the
tribe, with a population ranging from a maximum of 1,500 to a
minimum of forty and an average of 280. Among the Ngwato there
are no less than 300 of these wards with a somewhat larger average
of population in each. A ward may constitute either a quite separate
village, or more generally part of a village (whereas among the Southern
and Northern Sotho the village and the ward are usually one), but
in the latter case it is always socially and administratively distinft
from the other wards in the village. 3

Larger Tribal Divisions.

Among the Kxatla and the Ngwato, the wards are grouped together
into larger divisions of the tribe. In the Kxatla tribe, " there are five
of these major divisions, named respectively BaKxosing, BaMorema,
BaThsukudu, BaMabodisa, BaManamakxoti. These divisions rank
in precedence according to the order in which they have just been
listed. The first division, the BaKxosing, embraces all the wards derived

1 On this whole section, see I. Schapera, " The Old Bantu Culture," in Schapera, 1934,
18 ff. ; and Schapera, 1935, 203 ff., from which this account is taken.

1 Schapera, 1935, 213.

* A ward is called kx6r6 or bcotla by the Kxatla ; rarely mots*. The Ngwato call a ward
most usually motse and less frequently kxotla. The Pedi use kx6r6 for the men's gathering
place in the village (mout), and maintain there a central fire from which women are excluded.

9 A. WINIFRED HOERNLE

from the chiefs of the ruling dynasty for wards often come into
being through separation from some parent group. The next two
are more remotely conne&ed, but are also regarded as of true Kxatla
stock. The last two are made up for the most part of alien groups
absorbed into the tribe in the early days of its history. Each division
also contains some foreign groups more recently incorporated, the
general rule being that where a group of strangers is accepted into
the tribe, they are allowed to form a separate ward of their own,
with their leader as headman, and are then placed by the chief within
one of the five main divisions. Within each division, the wards are
graded in rank according to their seniority of descent or historical
status. This order of precedence was formerly striftly observed in
connexion with such communal ceremonies as the eating of first-
fruits, the initiation of boys and girls into adult membership of the
tribe, and the rites at the establishment of a new town or village ;
it is still of some social importance." l " Among the BaNgwato there
are only four major divisions in the tribe, and each division contains
wards of dihcosana (patrilineal relatives, however remote, of the
reigning dynasty) ; wards ofbathlanka (commoners of long-established
Ngwato stock) ; wards of bafhaladi (comparatively recent accretions
to the tribe) ; and even some wards of malata (subjeft peoples, such
as the MaSarwa and MaKxalaxadi). Nor are the divisions ranked in
any order of precedence, although within each division it is possible
to rank the wards of dikxosana according to seniority of descent,
and the wards of batUanka according to seniority of status. Another
notable point of difference is that, whereas among the BaKxatla the
Chief of the tribe always belongs to the BaKxosing division and is
the headman of its senior ward, among the BaNgwato the chief
has no ward of his own, while different Chiefs have belonged to
different divisions." 2 These major divisions do not seem to exist
among the Southern and Northern Sotho and have not been reported
for any other Western Sotho tribe.

The single Sotho ward certainly corresponds most closely with the
ibandla of the Southern Nguni, but the interconnexion of these
organizations in the tribal whole is very different in the two systems.
However, the inTshinga and iQauka divisions of clans among the
Xhosa show the same tendency to separate the lineal descendants of
the tribal ancestor from the descendants of ancestors taken into the
political system for a variety of reasons.

1 Schapcra, 193$, 205. * Schapera, 1935, 206.

SOCIAL ORGANIZATION 9 1

Households:

Within the ward we find the households, each with its own home-
stead or lapa. The members of a family husband, wife, and dependent
children live together in the same lapa, 1 but where there is more than
one wife each is entitled to her own lapa. All the huts of the same lapa
are enclosed within a courtyard by a low rectangular wall. The
malapa of close family connections, such as father and married sons,
may be found joined on to one another by connecting courtyard walls,
while separate family units will have a space between their walls and
those of their neighbours. In the old days when polygyny was
commonly practised, the general rule was that the wife first married
ranked as the senior or great wife. The other wives were of less
importance, but also ranked according to their order of marriage. The
only exceptions to this rule of precedence were : (a) when a man,
after marrying one or more wives, married the girl to whom he was
first betrothed, as in cases of infant betrothal ; or (K) when he married
the daughter of his maternal uncle. This new wife then took precedence
over all the others, and acquired the corresponding privileges. The
most outstanding expression of this difference in status between a man's
wives is found among the Pedi, where the mother of the heir to the
chieftainship must be a woman chosen by the tribe, whose loxadi
is collected from the tribe, and who, moreover, kindles with her
husband, when she comes to his mosaic, or capital, a new ritual fire,
from brands of which new fires must be kindled in all homesteads
of the tribe. This woman then becomes the wife of the country and
an intimate relationship exists between her son and the tribe from his
birth. 2

Totem Groups.

In all the Sotho tribes we find " a wider system of grouping which
cuts across the limits of the tribes. The members of such a wider group
(for which there is no special native term) all regard themselves as
intimately bound up, in some mystical way, with a certain species of
animal or natural object, known as their seano (object of reverence),
serito (honour), seila (taboo), or seboko (praise-name). The name of
this animal or object is used as a ceremonial or laudatory form of

1 Other close relatives may also be found living in their own lapa, e.g. a widow and her
children; cf. Schapera, 1935, 220.
1 W. Eiselen, 1931, 38.

9 2 A. WINIFRED HOERNLE

address, just as is the isiOongo of the Nguni tribes. There are special
myths telling how each group originally obtained 'the seano y and all
the members of a group have to observe various taboos and other
usages in connection with the animal or objeft which they revere ".*
Formerly, " if it was an animal, no one of that group might eat its
flesh, use its skin, or even touch it, lest some serious misfortune befall
him. Still less would he dare kill it, unless it was harmful and did mani-
fest damage, and in that case he had to be ceremonially purified after-
wards." * There is no bar to marriage of people having the same
seanSy but in cases of mixed marriage the children take the seano of
their father. There also does not appear to be any kind of social
solidarity among the members of such a group. Frequently they may
be scattered over many different tribes, as in the case of those whose
seano is the crocodile (kwena) ; they are found all over the Sotho
area. On the other hand, a single tribe can also include members of
many different seano groups. In a single Ngwato ward, the following
seanS are to be found: kwena (crocodile); moyo (heart); tlou
(elephant) ; phuti (duiker) ; nare (buffalo) ; kxabo (ape) ; tau or
sebata (lion) ; kubu (hippopotamus) ; kolobe (boar). 8 Sometimes a
group, or part of a group, would for some historical or political
motive discard its existing seano and adopt a new one, with a corre-
sponding change in the taboos it had to observe.

It was groups of people with different seano or seboko who settled
in Basutoland and were gathered together in a confederacy of tribes
by Moshesh in the early nineteenth century. 4 Moshesh belonged to a
branch of the Kwena, which is the dominant group in Basutoland
to-day. Gradually his family has asserted its authority over the other
tribes in the country. He himself parcelled out large se&ions of the
country among the four sons of his principal wife. Their descendants
are the big chiefs in Basutoland to-day. But the Paramount Chief, the
principal heir to the chieftainship of Moshesh, is gradually gathering
the administration into closer touch with his own immediate family,
by rt placing" his own sons over the heads of the descendants of
other branches of Moshesh's family. The status of the different tribes is
thus being diminished, and the whole population welded together into
a " nation ", rather than a confederacy, with administrative authority

* Schapcra, " The Old Bantu Culture," in Schapera, ^1934, 18, 19.

Schapera,

, up. cit, 19. Schapcra, 1935, 214, note.

i of Kxatla, Pedi, Phuti, Kholokoe, Sia, Tlokoa, Tlou, Kwena, Tswene, Hlokoana,
Khoakhoa, as well as groups of people from Zululand and the Transkei. Cf. Laydevant,
1931, 207 ff. ; Ellcnbergcr and MacGregor, 1912 ; and J. C. MacGregor, Basuto Tra&tion*
(Capetown, 1905).

SOCIAL ORGANIZATION 93

chiefly in the hands of the descendants of one family, the senior branch
of the family of Moshesh.

Age Sets.

In the Sotho tribes, the members of an age set (mophato, pi.
mephato ; also called thaka by North and South Sotho) used all to
undergo circumcision and a period of group-training of a very different
nature from that of the Nguni. 1 These age sets are ranged in a
hierarchy of men, and still to-day play a considerable part in the lives
of the people, especially among the Northern and Western Sotho,
where an age group may be called upon to perform many different
kinds of communal duties for the tribe or chief. In Basutoland, it is
rather specific villages, or distrifts, which are called upon for this work
for the chief of their area, who is entitled to use free labour for
ploughing, cultivating and reaping, not only a tribal land but also on
land allotted to his principal wife.

SHANGANA-TONGA AND VENDA

There is no space to go into similar detail concerning the social
organization of the other groups dealt with in this book. Funda-
mentally, the organization is of the same type, with the underlying
kinship stru&ure, the hierarchy of age, the dominance of one lineage
in the political stru&ure.

The Shangana-Tonga organization, since the overlordship of
Soshangane and his descendants came to an end, is once more of
the original simple tribal type. 2 Within the tribe, the patrilineal
lineage is the dominant kinship organization of the same eponymous
type as among the Nguni, and there is nothing corresponding to the
totemic grouping found among the Sotho. Within the kinship system,
attention must be drawn to a type of marriage among kin which links
the Shangana-Tonga with the Ndau of Portuguese East Africa and
the tribes of the Shona group. In all these tribes, marriage of cousins,
parallel or cross, is forbidden, but a man who has transferred cattle
for a wife has certain claims on her family which he may go far in
exafting. Should his wife prove barren or otherwise fail to fulfil
her marital obligations, the man can marry another woman from
his wife's family, her younger sister, her brother's daughter, or, in

1 See below, Chap. V, pp. 99 f.

1 Cf. H. A. Junod, 1927; A. A. Jacques, 1928, 3270348.

94 SOCIAL ORGANIZATION

the last resort, her brother's wife the woman who has been
established by means of his cattle ! In the Tonga tribes, further, the
most extreme type of behaviour is allowed between a man and his
sister's son, i.e. between the malume and the ntukulu. The ntukulu
has a preferential position in his malume s home and is permitted very
great liberty with his malume's property and his malume' s wife. The
malume^ on the other hand, must always be solicitous for his sister's
children, sacrifice for them if need be, and maintain contaft with them
at all times. We find, then, that the mother's family plays a much
greater part in the lives of her children in these tribes than is the case
with the Sotho or the Nguni.

The Venda * tribes praftise cross-cousin marriage, and the father's
sister is a person of dominant importance in the lives of her nephews
and nieces. The sister is closely linked to her brother right through
life, and, in the case of the chief, the father's sister (makhadi?) and the
sister play a part in the administrative and religious life of the people
not paralleled in the other groups with which we are dealing.

Here also we find the mutupo grouping, which is very similar to the
group called by its seano or seboko name among the Sotho. It is a
patrilineal system of grouping, usually named after some animal with
which there is a ritual connexion, preventing the members of the
mutupo from killing or eating that animal. The mutupo members
may observe various food- and other taboos, and generally there is
a number of praise phrases (tshikhodo) by which members can be
honoured in address. They are not exogamous and are not limited
in membership to any one tribe. In other respe&s the social organiza-
tion is very similar to that of the Northern Sotho.

1 Cf. H. A. Stayt, 1931, chaps, xi and xv; and-G. P. Lestrade, 1930, 306-322.

CHAPTER V

INDIVIDUAL DEVELOPMENT
By EILEEN JENSEN KRIGE

CHILD LIFE

IN Bantu society, where the status of a man is measured largely by
the size of his kraal and that of a woman by the number of children
she has borne, the birth of a child is hailed with great joy as an event
of importance to the whole village. For a long time after birth the
young child is believed to be weak, peculiarly liable to be injured
by harmful magical forces in the outside world. It is therefore confined
to its hut until the umbilical cord drops off and then is allowed out
only after it has been specially strengthened. This treatment varies
from tribe to tribe. Among the Zulu, Xhosa, and other South-Eastern
tribes, the child is held in the smoke of burning animal charms, given
medicine to drink, and scarified and rubbed with the charred remains
of the charms. The amulets worn by infants, the rites performed from
time to time during babyhood, the belief of the Shangana-Tonga
that " the child grows by medicine ", all lend support to Junod's
theory that the Bantu look upon babyhood as a marginal period
a period of ill-health ended at weaning, among the Shangana-Tonga,
by a rite similar to that performed after a period of illness or mourning. 1

The Bantu child is not weaned until two or three years old. Should
the mother become pregnant before this, she would " cut the road "
of the first child, making him thin or paralysed, an a&ion that is
severely censured. Sexual intercourse will, however, have been
resumed long before weaning, very often, as among Tonga, Lobedu,
and others, accompanied by a rite to ensure the child against harm.

The name, that of an ancestor or one given to commemorate some
event at the time of birth, is received usually when the child is taken
out of the hut for the first time. Later in life it will be given other
names, but the birth-name is the one always used by the parents and
others of their generation. A very common rite found among

* H. A. Junod, 1927, i, 58-
95

96 EILEEN JENSEN KRIGE

Shangana-Tonga, Venda, and Sotho, the significance of which is
difficult to discover, is showing the baby to the moon. The child
may simply be taken outside with the words " There is your rakhadi
(father's sister) ". Sometimes there is a special rite by means of which
the child " enters the law " or the parents " take possession of it "- 1

After weaning, the child becomes one of a group of many mis-
chievous toddlers to be seen about every Bantu village. 2 It is never
lonely nor does it lack playmates, for the village includes a number of
adult women with children more or less the same age. Toddlers are
looked after by their mothers and elder sisters, who take great pride
in teaching them the correft way of greeting their elders, of receiving
gifts, of dancing to the clapping of hands, and countless other things.
When naughty, they are frightened by tales of monsters who, in
Bantu folklore, carry off disobedient children. Their greatest danger
is that of falling into the open fires that burn in the huts ; but for the
rest they live a care-free life, with no clothes to keep clean, and few
of the dangerous instruments of civilization that take their toll of
child life. Parents are indulgent without unduly pampering their
children.

At an early age children learn not to sit with or eat with people
older than themselves. They spend most of their time with those of
their own age, play together or work together, and are recognized
by their elders as a group, from which colle&ive responsibility in
herding and other occupations is expefted. At first no wider than the
immediate family circle, this group, as the child grows and comes into
contaft with an ever-wider circle of people, includes first other
children in the neighbourhood, and finally, at circumcision or enrol-
ment into a regiment, embraces all of the same age within the tribe.
The educational value of these age sets is very great : not only are
selfishness, bad temper, and other faults more effectively checked by
the group than they could ever be by parents, but the younger children
are striftly controlled by the group just older than themselves.

The irresponsible period of early childhood does not last very
long, for at the age of five to six boys begin to herd lambs and calves,
and girls to nurse their brothers and sisters or help their mothers to
draw water, fetch firewood, and gather spinach. Loiter the girl begins
to grind, cook, and hoe with a light hoe, and by the time she is twelve
is capable of doing all the housework expected of a woman. The boys,

1 Junod, 1027, i, f6; D. Kidd, 1906, 12 f.

* Among the Shangana-Tonga and other tribes children are sent to their mother's people
for a year or two after weaning.

PLATE VII

" Daughters of Africa "

Kxatla children, B.P.
(Notice the method of carrying the babies)

PLATE VIII

f.V.,7. van Warm fin

(a) Vhushu rite of the Yenda

[J.Keynrke

(b) Kxatla maxwane (Male novices)

about to be whipped in the **0//a

(This ceremony is held annually until it is decided to send the boys through the

boxwera, or initiation proper)
INITIATION CEREMONIES

INDIVIDUAL DEVELOPMENT 97

after many a beating for letting the goats stray into the gardens, will,
at the age of ten or twelve, begin to herd cattle with the bigger boys,
an important step in their development which, in Zulu society, did
not take place in the old days till their " ears had been opened " in
the ear-piercing ceremony. 1

While herding all day, the boys acquire a wealth of veldlore. They
learn to know the names of all the edible birds and plants, they learn
to make traps and organize their own hunts and fights, becoming
adepts at killing birds on the wing by hurling knobkerries at them.

There is great similarity in the games played by children in different
South African tribes. Everywhere one finds boys and girls playing
with dolls, clay oxen, and string figures ; there are dancing and singing
games, tests of knowledge and also mixed games, played by both sexes
together. Many of the boys 9 games, such as stabbing at a tuber rolled
down a slope, give them an excellent training in exaftness of aim,
which will stand them in good stead later on when they hunt ; while
the miniature villages in which Sotho and Venda boys and girls play
at being grown-up, cooking, grinding, and hunting, just as they have
seen their elders do, afford a valuable training in the duties of later life.

The status of the child in later life will depend upon the accidents
of his birth his relationship to the chief, the importance of the clan
of which he is a member, and the position of his mother in the village
(whether she is a chief wife or a minor wife). During childhood, how-
ever, he is thought to be of very little importance. But, though he
has to show great deference towards everyone older than himself,
his individuality and his rights over any possessions that he may have,
such as goats, are always respefted. There are also occasions on
which, by virtue of his sexual immaturity and ritual purity, he plays
a leading part in ceremonial for the welfare of the whole tribe. Among
the Venda, Northern Sotho, and Eastern Tswana, it is small boys and
girls who must sprinkle the country with medicine to make the rain
fall 2 and, when birds become troublesome and have to be upa'd
or scared away by magic, children again play a leading part. 3 Among
the Zulu only boys under puberty may eat the remains of the black
bull strangled at the first-fruit ceremonies, and for the annual thevhula
ceremony of the Venda all preparations must be made by young and
innocent girls. 4 Children in Bantu society, therefore, besides being

1 Mahlobo and Krige, 1934, 163.

* W. Eiselen, 1928, 3*7-39* H. A. Stayt, 1931, 311 ; E. J. Krigc^ 1931, 223 ; I. Schapera,
1910, 211 ff.
1 Krige, 1931, 228-9. N. J. van Warmelo, 1932, 155.

98 EILEEN JENSEN KRIGE

essential for the prestige and social position of both men and women,
are, as cattle-herds, nurse-maids, and general helpers in the village, of
considerable importance in the tribal economy, while at certain periods
their co-operation is essential in ritual for the general welfare of the
tribe.

EDUCATION AND INITIATION
General Education.

Bantu education, the preparation of the child for the work of life,
differs from that of the European, not only because the life led by
adults in these two societies is so different, but also because their
theory of the sort of preparation that is necessary is not the same.
European education is conceived largely as the handing down to the
next generation of a system of knowledge, essential for overcoming
the difficulties and solving the problems of adult life. In Bantu society
success and welfare are not bound up with knowledge alone, and
difficulties are overcome not so much by the application of science,
as by use of magic and appeal to ancestors. For the natural forces to
work, for rain to fall and crops to mature, it is essential that correft
ritual be observed, taboos remain unbroken, and custom be faithfully
adhered to. What a man knows is therefore less important than what
he does, how he lives and behaves ; success and welfare are closely
related to morality and any change from traditional ways is looked
upon with suspicion.

The moulding of the individual to the social norm is achieved
chiefly through ritual and ceremonial. By aftive participation in ritual
he is made to feel the importance of those things that are of
value to society, and the correft emotional attitude to cattle
and crops comes, not so much from any conscious teaching, as
from imitation and participation in rites in which these values are
emphasized.

Economic education is carried on within the family with a strift
dichotomy of sex. The Bantu child learns by doing and imitating.
A boy watches his father cut wooden pillows, helps him make baskets
and put up a door in the hut, till later on he is himself able to perform
these tasks. A girl helps her mother 'weave mats, fetches and carries
for her while she is making pots, till she becomes quite proficient
herself.

Training in social behaviour comes partly from the older members

INDIVIDUAL DEVELOPMENT 99

of the family, partly from one's own age set, while the educational
value of play and games is as great in Bantu society as in our own.
The riddles and folk-tales recounted on dark nights not only give
full play to imagination, but inculcate the values and moral ideas
obtaining in their society.

Initiation " Schools ".

Formal education is given in the initiation schools, which play
a very important part in the life of every individual. The Bantu
tend to conceive of the development of the individual as a series of
stages clearly marked off from one another. Puberty marks an
important stage, marriage another. To pass successfully from one
such stage to the next, it is considered necessary to secure the aid
of forces that can influence one's life for good or bad. It is usual,
therefore, at such times to have ceremonies by means of which a break
with the faults and weaknesses of the previous stage is efFe&ed, the
initiate is strengthened by magic and appeal to the ancestors, and is
instrufted in the duties and privileges he is about to assume. These
rites emphasize the change that is taking place ; and the emotions they
evoke, together with the impressionable age of the scholars, help to
give a lasting effeft to the precepts inculcated.

There are not many schools before puberty. The Zulu ear-piercing
ceremony takes place at the age of seven or eight, and the Venda
thondo is said to be a school in which boys from the age of about
eight may, together with others much older than themselves, receive
instruftion every evening till they reach puberty. 1 It is wrapped in
secrecy, is attended mainly by children of high rank, and an important
part of the instruction is training in stealth and individual daring,
so useful for military purposes. The young men serving as warriors
at the chief's kraal spend much of their time in the thondo^ which as a
permanent institution is the nearest approach to the European con-
ception of a school. It has, however, this advantage over our schools,
that it is an institution to which adults also belong and plays an
important part in Venda life. There is also a thondo for girls, in which
women and girls of rank are taught Venda etiquette. 2

1 According to van Warmelo (1932, 109), its main function seems to be that of a meeting-
place for the garrison of warriors who regularly spend the night in the Chief's kraal, in addition
to which it provides education for the young.

1 van Warmelo, 1932, in.

100 EILEEN JENSEN KRIGE

Puberty Ceremonies.

Puberty, the great dividing line between childhood and the physical
maturity of adult life, is, among many tribes, notably those of the
South-East, the occasion of ceremonial in which the young boy or
girl is secluded in a hut for a while, given strengthening medicines
or committed to the care of the ancestors, and taught a number of
things. 1 The ceremonial is accompanied by much singing and dancing
by the age-mates of the initiate. 2 The Khoba or Vhusha ceremony for
girls in the Northern Transvaal is more severe. The girl is beaten
and made to repeat formulae (many of which concern sexual funftions)
while standing in water, and there is in the final dancing also a good
deal of mummery. At the Vhusha the Venda girl is warned not to
become deflowered before marriage, being shown how to have
intercourse without this occurring. 3 The educational importance of
these puberty ceremonies, however, lies not so much in the teachings,
as in the nature of the whole ceremonial, which is meant to impress
upon the individual the faft that he is now no longer a child.

Circumcision.

Circumcision is widespread in South Africa. 4 Among the tribes
practising it, it is regarded as an essential step in the life of every man.
The uncircumcised is considered a boy all his life, may take no part

1 Individual puberty ceremonies for girls at or about the time of their first menstruation
are found among all Nguni and Shangana-Tonaa, as well as among Venda and Northern Sotho.
The Tswana and Southern Sptho, on the other hand, have only the bale or byaU national schools
for girls, corresponding to circumcision of boys. In the North Transvaal puberty ceremonies
and ByaU exist side by side. Puberty ceremonies for boys are not as widespread as for girls,
being found, as far as published sources indicate, only among Venda, Natal Nguni, and, in
a vestigial form, among Shangana-Tonga.

* For a description of such puberty ceremonies, see E. ]. Krige, 1936, chap, v ; and
M. Hunter, 1936, chap. iii.

1 Cf. Stayt, 1931, 108. Cf. also Junod, 1927, i, 177-9, and van Warmelo, 1931, 37-51,
for descriptions of these ceremonies.

4 Once practised by all the South African Bantu, circumcision has died out in several Nguni
tribes. The Zulu discarded it in or just before Shaka's days, the Natal tribes no longer practise
it, the Swazi dropped it in the days of Malunge, regent during Mbandzeni's minority (Engel-
brecht, 1930, ai), the Mpondo abandoned it in 1867 under Faku's rule, thoudi a small seftion,
die AmaNqanda, which at the time of Faku's prohibition lived in Tembuland, as well as a
few Mpondomise and Fingo groups in Pondoland, still retain it (Hunter, 1936, i6j). Among
the Ronga seftion of the Shangana-Tonga circumcision fell into disuse more than a hundred
years ago Qunod, 1917, i, 72), though it is still practised extensively by the Shangana-Tonga
of the Transvaal. In Bechuanaland it has been abandoned by all tribes except the Tlokwa
and MmaLete (ed.). While dying out in some tribes, circumcision has been gaining ground
in others, and in recent years has been adopted* by the Venda, who did not circumcise.
The present distribution of circumcision cannot be fully plotted out, owing to complete lack
of information on many tribes. We do, however, know from published sources that it is practised
by Xhosa, Thembu, Fingo, and Bomvana ; some of the Tswana tribes ; the Southern Sotho :
the Pedi, Masemola, Lemba, Ndebele, Matlala of North-West Transvaal, Mmamabolo of
Woodbush, Transvaal Shangana-Tonga, Venda and Lobedu, Xananwa, Letswalo, and Khaha.
In Portuguese East Africa it is practised by Chopi and some Shangana-Tonga.

INDIVIDUAL DEVELOPMENT IOI

in the councils and deliberations of men, is looked down upon by
women and may not marry. Among the Sotho this school has a
national character not found in other tribes : it is held by the chief
every three or four years for the whole tribe, the boys are specially
taught to honour the chief and the tribe and all are enrolled into a
regiment at the end of the school. 1

Preparations for the school vary : Tswana novices are for a long
time previously subje&ed to taboos and severe beatings for the faults
of their childhood ; Pedi boys work for the chief for a few years
before the school. The night before the formal opening is usually
spent at the chief's kraal, where the first separation rites take place,
such as shaving and being beaten with medicated sticks. Next day
the operation takes place in a secluded valley in order of rank and
amidst loud singing to drown the cries of the boys. 2 Medicines are
applied to the wounds and it would seem that considerable skill is
shown by many operators. 3 The foreskins may be thrown into
running water, placed in the monumental stone filled with ashes at
the end of the school (Matlala), hidden by each boy in an antheap
(Xhosa), or even eaten in medicated meat (South Sotho).

For three months the initiates will now live in a lodge, which has
been specially prepared. Here they are completely cut off from
ordinary life, observe many taboos, use special terms for many obje&s
and are in the direft charge of those who went through the previous
school. The men who live in the lodge must refrain from sexual
intercourse, lest the wounds of the boys do not heal, and any man who
wishes to enter has to know the correct passwords. The uncircumcised
may not come near, on pain of death.

One of the main obje&s of this school is to train the boys in courage
and endurance. They are beaten on the slightest pretext, have to
sleep on their backs on the bare ground without covering, and are
severely punished if unable to repeat the formulae and songs that are
so great a feature of the school. They receive a good training in
hunting and in some tribes wooden figures of animals are shown the
initiates. Not only do many of the songs contain reference to sexual

1 For detailed descriptions of circumcision schools of these tribes, see C. Hoffmann, 1914-
15, 81-112; P. Ramseyer, 1928, 40-70; J. T. Brown, 1921, 419-427; W. C. Willoughby,
1909, 228-24; > C. A. Wheelwright, 1905 251-5 ; N. Roberts and C. A. T. Winter, 1915,
561-578 ; Junod, 1927, i, ?i~94 ; W. Eiselen, 1929, passim.

1 Among die Nguni, the boys are expected to show their manliness by uttering no cries.

1 For observations of a European doclor on the cleanliness and good results obtained, see
G. A. Turner, " Circumcision amongst South African Natives," ML J. S. Afr., x (1914-15),

102 EILEEN JENSEN KRIGE

functions, but during the whole period language that at other times
would be considered obscene is used to the women who bring the food.

After the wounds have healed a great change occurs. Whereas
the initiates have hitherto always sat facing the west, they are now
made to face east : they turn from the darkness of childhood to the
light of manhood. This is sometimes the occasion of a feast, and
among the Lobedu it is now that the boys are given strengthening
medicine. What this consists of, we do not know, but among the
Southern Sotho and the Transvaal Ndebele it contains human remains. 1
A pole, probably of phallic significance, is now brought into the lodge.
The boys have to greet it respeftfully and from here instruftion is
given. At the end of the school the hair is shaved, new clothes and
a new name given, and the initiates are received back into society
amidst great feasting. The lodge is burnt. The period of circumcision
is a critical one for the whole tribe : there must be no quarrels and
jealousies at home and dancing is taboo.

Circumcision among the Xhosa and other Transkei tribes has
neither the secrecy nor the formality of the Sotho schools. It is local
in character, only from two to twenty boys forming each school,
and it is not connefted with the formation of regiments. 2 In fact,
circumcision here differs little from the ordinary ear-piercing
ceremonies of the Zulu. The seclusion, strift only until the wounds
have healed, seems to have no other significance than the medical
need for keeping the boys from " uncleanness ", sexual and other,
that is believed to affect wounds. There is said to be no formal
teaching except of dancing, no hardships are undergone, and, once
their wounds have healed, the boys spend their time day after day
dancing and love-making at feasts held by each father in turn, till the
school is concluded by a lefture on how they should behave in
future. 3 Basutoland schools fall midway between those of the
Transvaal tribes and those of the Nguni ; the boys are in the charge
of only one or two teachers, there is more formality and secrecy than
among the Xhosa, while there are in addition elements, e.g. the
eating of medicated meat and the manner in which boys express their
desire to undergo initiation, 4 that appear to be borrowed from puberty
and Sutwa ceremonies in Natal.

1 Cf. H. Dieterlen, Sesuto-English Dictionary, s.v. sehocre; and H. C. M. Fouric, 1921, 134.
1 Xhosa regiments are clan-, not age-groups.

* For descriptions of these ceremonies, see P. A. W. Cook, 1931, 71-64; A. Schweiger,
1-6$.
H. E. Mabille, 1905-6, 247-8.

INDIVIDUAL DEVELOPMENT 103

The Boxwera.

Among the Northern Sotho and Tswana circumcision is supple-
mented by the boxwera, a school taking place sometimes a year later,
but often, as among Pedi and Letswalo, only four or five years later. 1
During the period between the two schools the boys find themselves
without status : they are looked down upon even more than the
uncircumcised, and should they beget a child it would be killed.
This school and the girl's byak, which occur at the same time and in
many respefts supplement one another, last a whole year. 2 The lodge
of the boys, like that of the girls, is in the chief's kraal, their seclusion
being by no means as strift as during circumcision, though they may
not be seen without the reed mask and costume chara&eristic of the
boxwera. Many months are spent in the preparation of these elaborate
costumes. When addressed by the uncircumcised, the boys have to
reply in the whistling language. 3 Apart from the work done for the
chief, cutting poles, softening skins, hunting, the boys are taught
" laws ", mainly rules of conduft in married life, the necessity for
providing for one's mother in her old age, and in dealings with men,
keeping peace and order. An important feature is the songs and dancing
in very heavy, clumsy costumes which cause real suffering to the
wearer, constituting a great test of endurance. The boys are now
initiated into the secret of the wonderful bird or beast, the ruling
spirit of the by ale. This consists of a conical frame covered with
strings of bark, which serves to cover the man or men afting the
part of this mysterious creature, the whole being surmounted by a
head. It is said to come from a deep pool, it whistles out its commands
on a peculiar flute and appears in the village only on moonlight
nights. During the period that follows its appearance, the whole
tribe is subjeft to strift rules : it is taboo to make a noise in the villages,
it is taboo to dance in the day or get drunk or beat one's wife. The
school ends in a feast for both boys and girls, after which the lodge

1 Pedi, Khaha, Mmamabolo of Woodbush, Letswalo, Lobedu, Matlala, and probably many
other North Transvaal tribes hold this school. For accounts of its nature, which has nowhere
been fully studied, see Junod, 1929, 131-147 ; C. Hoffmann, 1914-15, 100-111 ; W. Eiselen,
1929, 40 f. ; G. Beyer, 1926, 258-9. Note that the term boxwira is used by Tswana and others
who have only one school for the circumcision school. (Note by editor : Some Tswana tribes
have two schools, the first being termed " the white boxwtra " and the second " the black
boxwlra ").

* Among Khaha, Lobedu, and other North Transvaal tribes Box w era and byale occur together ;
but it is not clear if this is also true of the Pedi and others.

Bantu languages being tone-bound, it is possible to convey by whistling almost as much
as one can by the words themselves.

104 EILEEN JENSEN KRIGE

and costumes are burnt. Among the Pedi, Letswalo and Matlala, the
boys receive cuts from chin to ear, a rite that in other tribes belongs
to the Komana school. It would seem that this is a merging of the two
schools.

The Komana.

The komana is a rain rite as well as a school, which, among the
Lobedu, used to be held twice a year for about ten days. 1 It was
opened in the usual way by running the gauntlet, after which each
boy received four circular cuts from mouth to ear on either cheek,
a painful and lengthy process. The main objeft of this school appears
to be the introduction of the boys into the deepest secrets of the tribe.
They are, among the Lobedu, shown the sacred drums, among which
is the rain drum, and learn for the first time that the badaja or spirits
of ancestors, who are supposed to come back to earth after certain
ceremonies, throwing stones when offended, to frighten the uninitiated,
and speaking in the whistling language, are in reality old and trusted
men who aft the part. Many special songs are sung, but there are no
thrashings or hardships of any kind.

Byale or Girls 9 " Circumcision " School.

While the tyale appears to be a single culture complex in South
Africa, the ceremonies held in Basutoland and Bechuanaland differ
from those in the Northern Transvaal in the complete absence of the
mysterious beast which controls the whole school in the north. 2 The
opening rites in the Northern Transvaal consist of beating the girls
with medicated rods and smearing those holding official positions
(the chief's daughters and some others) with the stomach contents
of an ox that has been slaughtered for the occasion (Lobedu). The
initiates wear skin skirts and bands across the chest of woven bark
and grass, 3 which are believed to ensure fruitfulness, and they may
not smear their bodies with fat. The day is spent working, and
besides being made to perform various exercises, such as hopping
on one leg and picking up stones, they are subje&ed to other hardships.

1 There is no adequate description of this schopl, found among the Lobedu and other North
Transvaal tribes. It is mentioned by Junod, 1929, 144, and more fully described by . J.
Krige, 1931, 216-220.

( For 'descriptions of this school, see : Mabille, 1905-6, 250-1 ; Ellenberger and MacGregor,
1912, 288-9 ! T. Brown, 1926, 89. For ceremonies in the North Transvaal, see : G. Thilenius
1915, 72-5 ; H. M. Franz, 1929, 1-5 ; Stayt, 1931, 138-141 ; van Warmelo, 1932, 79-103.

In Basutoland they wear masks as well.

INDIVIDUAL DEVELOPMENT 105

They are severely beaten, their meals of dry porridge must, like those
in the boys' circumcision school, be eaten very quickly, and unpleasant
tasks, such as that of eating fresh cow-dung, are set. They must be
very humble and respe&ful to all and have to use the special terms
characteristic of the school whenever they speak.

The nights are spent elongating the labia minora, for which medicine
made of parts of a bat is effeflive, and in singing and dancing. Some
of the songs among the Lobedu are in praise of various important
sibs, others have reference to sexual fun&ions, while the most im-
portant song of all, that in honour of the Bird of the School, contains
advice (in metaphorical terms) not to be deflowered. Sometimes the
Bird terrifies the girls by coming into the courtyard and stamping out
their fire. When mosabetho songs are sung, the men who have been
through all three schools come and " choose " the women they like,
with whom they have sexual intercourse after the dance, and towards the
end of the school the iyale girls are allowed to join in, though, should
anyone be impregnated, the usual fine will have to be paid.

The " showing of the dihSma ", or mummeries, plays an important
part in the school. One of the Xananwa dikoma is called Phudi. It is
the model of a goat with eyes of a certain red fruit used in rain medicine,
and is a sign of fruitfulness. The girls are asked what it is and beaten
if they do not know. Another, Mapono, is a woman disguised in a
dress of reeds with a headpiece of horns. She comes dancing in and
chases the girls, who flee in terror. Another kSma, Motlankana, is
a girl dressed as a man, with imitation penis, who, after inviting every
girl to accompany " him ", seizes one and simulates sexual intercourse
while a song is sung. Some time during the school the operation,
usually a cut on the clitoris, 1 is performed and the ceremony is ended
in a dance and feast. The significance of the by ale has not been properly
studied, but there can be no doubt that it is closely associated with
fertility and is, like the boxwera, a preparation for marriage.

Venda boys and girls are prepared for marriage in the domba^ which
appears to be a fertility rite, the central feature of which is the python
dance. The python is closely associated with rain in Venda thought :
when killed it must be thrown in the river lest rain be withheld, while
python skin renders a barren woman fertile. 3 In the domba, boys and
girls are, by means of formulae and the use of mummery, taught about

1 This is true for Lobedu and Venda. Xananwa, Tlokwa, and Moletse make a diagonal
cut below the urethra (Franz, op. tit.), while the Mmamabolo deflower the girls with a cow horn.
* Stayt, 1931, 111-124; van Warmelo, 1931, 52-79.
Stayt, 1931, 309-310.

106 EILEEN JENSEN KRIGE

childbirth, marriage, and sexual life, but so indireft is this teaching,
and so generally unintelligible to the European who approaches
the matter without a full knowledge of its significance, its context
and setting and the whole configuration of values and institutions
in v which it fulfils its true moral funftion, that it is described,' except
by a few investigators, as merely obscene.

Methods of Teaching in the " Schools "

Every Bantu child, in addition to the education received in his own
home, which enables him to satisfy his economic needs, thus passes
through at least one and often more than one school. 1 If we are to
understand the value of these schools in tribal life we must not judge
them by the knowledge inculcated, so much as by the manner in which
they fit the individual (or are thought to fit him) for the life he is to
lead as an adult. The fun&ion of each school is different, but, broadly
speaking, we can say that every school is a preparation for a new
step in the life of the individual, in which an attempt is made to
imbue him with the qualities and the knowledge necessary for that
stage. The qualities believed by the Bantu to be most important for
success in adult life are ability to bear children and strength, courage,
and endurance in the work of life ; hence these things are emphasized
in the schools. In addition they learn to honour the chief and tribal
custom, respeft those older than themselves, value those things which
are of value to the society, and observe tribal taboos, especially those
conne&ed with sex life.

The methods employed vary. Fertility and other qualities of adult-
hood can be obtained only by means of rites, medicine, and invocations
to the ancestors for their blessings, and we must never lose sight of
the fad that Bantu initiation schools are primarily rites, the inculcation
of knowledge and precepts being of secondary importance. A common
method employed, in the teaching of precepts and of showing initiates
their new duties is the use of songs and mummery, a method that
has a great appeal to primitive and illiterate people. To train them
in courage and endurance they are subje&ed to hardships and tests.
Teaching in Bantu schools is seldom direft and usually symbolical :
boys in the circumcision school are taught to be truthful, not by
exhortations, but by having their tongues clamped with a piece of

1 In the North Transvaal and in Zululand, girls have two schools, boys three ; in Basuto-
land and Bechuanaland, girls and boys have one, while there are tribes that have two schools
for boys and one for girls. The Venda are so school-ridden that, in addition to three of their
own for both boys and girls, they have two foreign adoptions, one for girls and one for boys.

INDIVIDUAL DEVELOPMENT 107

stick, and the formulae that they learn are couched in language that
only the initiated can understand. 1 Some of the formulae teach pre-
cepts useful in everyday life, others contain sexual teachings. These
formulae, often described as licentious and valueless, may appear to
the observer or reader, unaware of their setting and significance, as
crude ; but to the native they do not appear so. Out of their context
they might convey to the native mind some obscenity, corresponding
somewhat to the translation into licentious language by indiscriminate
use in ordinary social relations of the scientific terminology of a
professor in physiology. But even if that is so, they are taught in
an atmosphere of such awe and respeft for the school, and all it stands
for, that their psychological effeft cannot be otherwise than to
emphasize the importance of the things that are taught. One cannot
respeft the obscene, nor can precepts approved of and inculcated by
society as a whole be thought obscene by the individuals composing
that society. Even when judged from the European standpoint,
Bantu schools do achieve success in inculcating obedience, discipline
and general good behaviour, qualities that make many a European
employer prefer " boys " that have been through the schools to
others.

The growing popularity of initiation schools in the Transvaal
is not unconnected with the faft that they are a valuable source of
income to the chief, who not only receives fees from initiates, but
also enjoys their services before, during, and after the school. The
secrecy and mystery surrounding the schools is an attradtion to the
young. The schools of the Sotho have many elements in common
with the secret societies farther north, and appear to have been
influenced by them.

SEX LIFE AND MARRIAGE
Sex Life.

Bantu initiation schools and many other customs can be better
understood when looked at in the light of the Bantu attitude to sex.

1 Many of the formulae are so archaic as to be almost meaningless, others may not be grasped
at first, but are understood better after the individual has been through a number of schools
in other capacities, such as shepherd, older teacher, and so on. It is often stated that much
of what happens in the schools is meaningless to the Natives themselves. Care must, however,
be taken not to be too hasty in our judgments. We know very little of the schools and most
of what we do know has come at second-hand from individuals who have been through the
schools only as initiates. Much of what we teach in our own schools is neither understood
nor remembered by the pupils. Yet it would be misleading to conclude that our schools as
institutions have no value. Further, it is the ritual, rather than the teachings, that is of primary
importance in Bantu schools.

io8 EILEEN JENSEN KRICE

The Southern Bantu have not that horror of sexual intercourse that is
associated with the Christian idea that sex and sin are almost
synonymous. They look upon the union of the sexes as something
natural and good. Further, the value of the sex aft in perpetuating
the race, and its close association with human fertility and thus with
social welfare, make it an integral part of a number of rites. There
are times when sexual relations are essential when moving to a new
village (among the Shangana-Tonga), when a widow (Shangana-
Tonga) or a homicide (Zulu) has to be purified and there are
occasions when the sex aft is striftly taboo after a death, after
childbirth. 1

The sex life and funftions of the individual are believed to influence
nature : a menstruating woman is a danger, not only to any man who
has relations with her, but also to the cattle who would become ill
if she were to go into their kraal and, among the south-eastern tribes,
to drink their milk. Any irregularity in child-bearing, such as the birth
of twins, an abortion, or a miscarriage, is thought by many tribes
to have an effeft on the rain. If a Pedi woman has an abortion, it is
considered enough to desiccate the country with heat, and a rite will
have to be performed before rain will fall, 2 while among the Lobedu
abortions and twins have to be buried in wet soil, lest they cause
drought. It is clear that here the fertility and sexual funftions of
human beings, and the rain, so essential for the fertility of the fields,
are thought to be so closely connefted that irregularities in the one
easily affeft the other.

If the sex aft is a powerful means for good, it is at the same time
considered to be full of dangers. It brings in its train a certain
" uncleanness " or defilement which may prevent the healing of
wounds and the recovery of sick people. Hence, among the Zulu,
people who have had sexual conneftion, together with menstruating
or pregnant women and mourners, are debarred from coming near
anyone who is ill. Similarly, during circumcision ceremonies or in
dme of war (Shangana-Tonga) the sexual aft is taboo. It is also
possible for a man, through intercourse with a woman, to leave in
her a mysterious power of conveying disease to the next person who
has conneftion with her ; hence the dangers of adultery.

In all the Southern Bantu tribes the birth of a child out of wedlock

1 According to Junod, 1910, 146, it is only the sex relations of the married that are important
in ritual. Among some Shangana-Tonga, when the sex aft is taboo, e.g. during the mourning
period, the gangisa custom (sex play of the unmarried) is not taboo.

1 Junod, 1910, 140.

INDIVIDUAL DEVELOPMENT 109

is frowned upon as an offence, and sexual morality consists in avoiding
this and in observing sexual taboos of the type mentioned above.
Bantu children, even before puberty, indulge in play at sexual inter-
course ; but this is either connived at or looked upon with amusement
and toleration by adults because it can have no social consequences.
After puberty, however, care must be taken that no children are born,
and we find some form of social control, which varies from tribe
to tribe. 1 Premarital sex intercourse is a recognized institution and
among Venda and Zulu definite instru&ion is given at puberty on
how to have intercourse without becoming deflowered. 2 Only
external intercourse is, however, allowed, to facilitate which the
Zulu pluck the pubic hair. The Swazi guard against possible con-
sequences by placing a piece of soft goatskin over the female organ. 3

If a Shangana-Tonga girl has no lovers people laugh at her, saying
she must be malformed, while a boy who lacks -lovers is subjefted
to a certain rite to render him more successful. 4 In this tribe, as among
Mpondo and Venda, both boys and girls may have many lovers.
The Zulu, however, are strifter. Here the praftice of klobonga is
allowed only between couples that are qomad (privately betrothed)
and permission has first to be obtained from the group of elder girls,
who control the girl's aftions very striftly. 5 A girl may thus hlobonga
only with one man, but the man may have several lovers, since he is
not confined to one wife. The Mpondo boy must send certain presents
to the parents of his lover (ior. to the mother, goats and later an
umryobo beast to the father), 6 while the lover of a Venda girl works for
Jier father, often side by side with her husband-to-be. 7 Ceremonies
and occasions more particularly associated with these pra&ices are
girls' puberty ceremonies (among south-eastern tribes), marriage
ceremonies, and the domba and byale schools.

Girls are periodically examined by their mothers to see if they are
intaft 8 and there are effeftive sanftions to secure virginity at marriage.

1 Premarital sex intercourse is a recognized institution among Venda, Lobedu, Zulu, Shangana-
Tonga, Swazi, Mpondo, Fingo, Xhosa, Pedi, and Mmamabolo of Woodbush. There is no
definite information for other tribes.

* Stayt, 1931, 108; Kohler, 1933, 34. * Engelbrecht, 1930, 21.

* Junod, 1910, 146. * Krige, 1936, 105-6; cf. Kohler, 1933, 31 f.
' Hunter, 1936, 181. Though such unions sometimes lead to marriage, this is not generally

approved of, parents preferring their daughters to marry someone she does not know so well.
(Hunter.)

f Stayt, 1931, in. Venda informants state that it is considered wrong for a man to dhavula
with his fiancee. If, as among the Lobedu, a taboo sets in between couples the moment they
are betrothed, this is quite comprehensible.

* See Kohler, 1933, 39-40, for a discussion on whether it is possible for these women to
establish the virginity or otherwise of a girl.

EILEEN JENSEN KRIGE

Among Zulu and Mpondo, defloration brings defilement on the whole
age-group of girls, who will maltreat her for the shame she has brought
upon them, while the Kxatla songs of mockery were said to be a
most powerful deterrent in the old days. 1 Furthermore, there is
always a distinction made in marriage ceremonies between brides
who are virgin and those who are not. Boys are discouraged both by
public opinion and the fine imposed as damages.

Premarital intercourse is thus a socially recognized and socially
controlled institution among the Southern Bantu, who regard r as
striftly moral so long as it conforms to the rules. These praftices form
part of a system and philosophy of life and for this reason, while we
may disapprove of them we cannot, with justice, label them as
immoral. 2

Within marriage, too, there is considerable sexual freedom, though
the attitude towards adultery is not the same in all tribes. It is strictest
among Shangana-Tonga and Zulu, where adultery is a ground for
divorce. 3 Among the Cape tribes, the Swazi, Tswana, Pedi, 4 and
others, adultery is not a ground for divorce and, though legally
punishable, is very common indeed. The Sotho of Woodbush dis-
approve of adultery in theory, though in praftice and in private it is
looked upon as something desirable. 5 Boys are, moreover, taught
in the circumcision school here and in neighbouring tribes to over-
look adultery on the part of their wives and to refrain from gossip
about the love affairs of others. Women are allowed considerable
freedom as long as they aft with discretion and it is not unknown for
husbands to indicate to their wives men who may be suitable as lovers.
Even here, however, adultery is punishable by a fine.

Men are, subjeft to the fine payable for adultery, allowed greater
freedom than women, and there is in many tribes a recognized class
of women with whom married men may have relations with impunity.
This includes amadika^i (divorced women and unmarried mothers,
for the seduftion of whom no fine is payable unless pregnancy results) 6
among Cape tribes ; " wives " of women among Lobedu ; and

1 Schapera, 1933, 67.

* Kohlier, 1933, 34, calls external intercourse a " perversion of the natural funftion " and
draws attention to " nervous disorders " which it brings in its train. This would be an interest-

ing subjeft for investigation.

Junod, 1927, i, 195. Yet adultery is
4 Among Pedi adultery is common and ir

common among the Shangana-Tonga.

, I in certain circumstances considered even justifiable,

e.g. if the husband remains away from home after his wife's child has been weaned.

Hoffmann, 1933-34, 229.

Whitfield, 1929, 413. Among Thembu no fine is payable in the case of a second pregnancy
of an unmarried woman, but custom in this case varies.

INDIVIDUAL DEVELOPMENT III

widows, who among the Xhosa are expefted to have intercourse with
strangers to raise up seed to their dead husbands, 1 and among Pedi
and other Transvaal tribes are looked upon as " every man's wives ". 2

Marriage.

When the individual has passed through whatever schools and
ceremonies obtain in his tribe for initiation into adulthood, he is
ready for marriage. To regard Bantu marriage as the culmination
of sex love would be giving this institution a wrong perspective. In
marriage it is the social, legal, and economic aspe&s that to the Southern
Bantu are the most important sex relations may, as we have seen,
be had outside marriage. 3 Further, marriage is primarily an affair
between groups, involving the two families concerned even more than
the individuals. The personal predileftions of the couple do not
carry nearly the same weight as the good name of the family of the
girl, her ability to bear children, work well, and get on amicably with
her mother-in-law, for whom she will at first have to work.

The choice of a partner in marriage is thus not surrounded by the
romantic conceptions of many Europeans. Further, if the individual
concerned belongs to the Sotho group, he will have known all his
life whom he is destined to marry, for here a man is expe&ed to marry
his mother's brother's daughter, if one of suitable age is available. 4
Marriage with the father's sister's daughter is permitted but not very
common, as well as marriage with the father's brother's daughter. 5
Individual choice is further limited among the central tribes by infant
betrothal. 6 The Nguni group and the Shangana-Tonga, on the other
hand, do not allow marriage with relatives, and here there is therefore
greater freedom of choice, though parents nevertheless arrange many
marriages. Some girls, among Venda, Lobedu and other Northern
Transvaal tribes, become the " wives " of women, for here any
woman with the necessary lobola may marry a girl. She will point
out some man with whom the girl may have children, allow her

1 Whitfield, loc. cit. * Eiselen, 19280, 49.

1 The relative importance of social and economic faftors and purely romantic love in European
marriage varies greatly. Very often social factors unconsciously determine the choice of the
individual.

4 Marriage with the mother's brother's daughter is found among Venda, Sotho, and Transvaal
Ndebele.

6 Whereas the Huruthse prefer cross-cousin marriage to that with die father's brother's
daughter, the Pedi prefer the latter form.

Very common among the Pedi (Harries, 1929, 3), and found also among Venda, Lobedu,
Tswana (Brown, 1926, 159) and Swazi (Engelbrecht, 1930, 21). In such cases lobola is usually
handed over while the girl is still young, so that she can be nurtured on its milk. Cattle may
even be handed over before the child is born.

EILEEN JENSEN KRIGE

own husband to have access to her, or leave her free to choose her
own lovers. 1 If a young man dies without issue among Venda,
Lobedu and Swazi, a woman may be married for him and his house
will continue. 2 This is, however, rare. Nevertheless, in spite of all
these customs, it does happen that a young man when attracted to a
girl asks his father to arrange a marriage, which may be done in the
usual way : and if things go badly, recourse can be had to elopement.
If a girl really dislikes a man she is seldom forced to marry him.

Marriage negotiations may be opened either by the parents of the
boy or by the parents of the girl, in addition to which there are one
or two other forms. While by far the most widespread method is for
the boy's parents to take the first step, both methods may be found
side by side, and among the Xhosa it was, in the old days, more
usual for the girl's people to open negotiations ; this is still the most
usual method among the Zulu in cases where the girl is of high rank. 3
The Zulu and other tribes of Natal have what is known as the ukulaleka
or ukuganisela custom in which the girl, at the instigation of her lover,
her father, or without the knowledge of either, presents herself at
the kraal of the man she wants to marry. His people will then have
to report her arrival to her father and open marriage negotiations.
For a man to refuse such a girl is almost unheard of in Zululand ; so
much disgrace would be attached to it that the girl would immediately
balekela someone else. This custom leaves a girl considerable freedom,
for if her father wishes to marry her to someone she dislikes she can
always run away to her lover.

The Nguni peoples have the ukuThwala, a kind of abduftion or
forced ukubaleka in which the boy and his friends carry off the girl
(sometimes even at the instigation of her own father). Among the
Zulu this takes place when an engaged girl, or her father, breaks her
contraft, and a penalty over and above lobola will be demanded only
if the boy had no right to thwala. It is not at all clear from the evidence
whether the ukuThwala obliges the girl's parents to accept the position
or whether they may still refuse to consent. 4 Abduftion or elopement
among the central tribes and Tonga has not the recognition that

1 A woman who has no sons herself has a claim on her brother's daughter who would have
married such a son. In Basutoland a widow with cattle could marry a wife and get some
male relative or friend to beget children for her. Further, if the first wife of a chief had only
daughters, a wife could be married for the eldest and handed over to the next male relative
to raise seed. /The eldest son of this union was the heir (Whitfield, 1929, 175).

* Whitfield, 1929, 213 ; Engelbrecht, 1930, 8.

* Among the Mpondo there is some stigma attached to the first step being taken. by the
girl's people unless she is of high rank (Hunter, 1936, 187).

4 See Whitfield, 1929, 79-8;.

INDIVIDUAL DEVELOPMENT 113

ukuThwala has. The girl will run away with her lover to some
indulgent relative and the boy's people have to report the matter
to her father who, among Venda and Lobedu, usually accepts the
position but exacts lobola. Such an aftion is considered a disgrace.

Once negotiations have been opened, the all-important question
of lobola arises. No custom has been so widely discussed or given rise
to so many misconceptions as this one of lobola, or the handing over
of cattle in marriage. Because it involves material possessions and is
often accompanied by considerable bargaining, it was looked upon
by Europeans at first as a sale. Apart from the fact that the words for
selling and for paying lobola are not the same, the husband has no
rights of property or possession over the woman, nor can he dispose
of her or ill-treat her. Lobola is not a dowry, for " the cattle are where
the girl is not " and though she may, when in trouble, claim help
from anyone who holds her lobola, she is not able to use it and does
not necessarily benefit by it. Only through lobola can a man claim
the children of a woman ; once this has been paid, all her children
are his, too, whether he is their real father or not. There is some truth
in the idea that lobola is a compensation to the group that has lost a
member, to restore the disturbed equilibrium. There is, however,
no direct relation between the amount of lobola and the value of the
girl and where, as has been the case in times of famine, such things
as stones and mice are handed over, it can hardly be looked upon as
compensation. It is rather an attempt to soften the blow and try to
obtain the friendship of the girl's family. That lobola is not mere
compensation is shown also by the fact that there is a lifelong bond
between a man and his parents-in-law. He is expe&ed to assist them
whenever they need his help 1 and for this reason is called by the
Zulu " the handle of a hoe ", being, like this most important implement
in society, a friend and helper. Furthermore, the. gifts handed over
by the girl's father may sometimes equal or even exceed the lobola
paid (Cape Nguni). Lobola has many aspects ; but its importance
lies in the fact that lobola is marriage and to the Southern Bantu
marriage without lobola is inconceivable.

The amount handed over varies, according to the times and the
wealth of the boy's family, though everywhere it is considered an
honour to give a high lobola. The nearest approach to a fixed amount
is found among the Lobedu, where there are seven named beasts
(in addition to the virginity beast), but even here there can be variations

1 See Stayt, 1931, 144-

114 EILEEN JENSEN KRICE

between a minimum of two cows and calves and the option of giving
three additional beasts. 1 In most tribes a special beast, often a cow
and calf, is handed over to the girl's mother in recognition of her
virginity. 2

Among many tribes the handing over of lobola is a life-long process.
The Swazi man need not pay a single head of cattle till late in married
life, and often instalments are handed over after the birth of each
child. 3 The essential feature of lobola among the Xhosa, according to
Soga, " is that there is no finality in it. Cattle pass from the husband
to his wife's parents or their family throughout the lifetime of the
husband and beyond it." 4 The ukuteleka custom is much in vogue
among the Cape tribes, whereby the father of a girl from time to
time impounds her until the husband hands over an additional amount
of lobola*

The Shangana-Tonga, Pedi and most of the central tribes prefer
the full amount to be handed over before marriage ; but in practice
this is seldom done. The Huruthse feel that lobola should be handed
over before the birth of the first child, but it'is often left till the initiation
of the eldest, when it is handed over in a lump sum 6 (which these
tribes prefer). The Venda complete payment of lobola usually only
after the marriage of the wife's first daughter. Marriage can be arranged
without lobola if a man be poor, on the understanding that the lobola
for the first daughter will be given to the wife's father, a custom known
as ethula among Zulu and Cape tribes, and found also in other tribes,
such as the Kxatla and Pedi.

The person primarily responsible for payment of lobola (in the first
marriage of a man) is his father, who usually pairs off brother and
sister, the girl's lobola being used to acquire a wife for the boy. Other
relatives may, however, also be liable for contributions ; and among the
Sotho, as among the Swazi, 7 the mother's brother is held responsible
for at least one head of cattle, which the Huruthse can legally force
him to pay.

In the protrafted ceremonial that is necessary for Bantu marriage
the most important aftors are the two families rather than the

1 J. D. Krige, 1934, 1 35-14 9.

This beast, payable also by a seducer, is the ingquthu of Zulu and Transkei tribes, Lobedu
tsotd* ya musJune, Huruthse serufd, Venda /u/{<z</{f , Swazi umsula nycmbeti.

Engelbrecht, 1930, 13. Soga, 1932, 266.

The Venda have a similar custom. See Lestrade, 19306, 198.

Lestrade, 1926, 939-940.

' Engelbrecht, 1930, 14. The malumt is on the other hand entitled to one or more of his
niece's lobola^ which he holds for sacrifices when things go wrong with her or her children.

INDIVIDUAL DEVELOPMENT 115

individuals. The father of the girl will neither accept nor rejeft a
suitor without the consent and support of other members of the
family or lineage, who will be called in to discuss the matter. Hence
it is always said of a girl " She has been married to such and such
people " or " over at such and such a kraal ". In a matter so delicate
as marriage negotiations the Bantu consider it essential to have
" go-betweens ", through whom all transaftions are made, and who
will be better able to withstand any rebuffs. There may be a go-
between on each side, as among the Lobedu, or the go-betweens may
be chosen by one of the parties, bridegroom's among Tswana, Zulu
and Xhosa, bride's among the Venda.

The moment negotiations have been opened, there is a barrier
between the two families, which must be bridged by the tireless
efforts of the go-betweens. The distrust is shown in a number of ways
and continues right until the marriage feasts are held. The girl's
family will be rude and insulting when the first instalment of lobola
arrives and very often a mock-fight takes place, the girl's male relatives
trying to prevent the drivers of the lobola cattle from entering the
kraal. They are made to feel that the girl is a loved and valued member
of her family, who are in no way anxious to give her up. This hostility
is ceremonially expressed also in the mutual vituperation between the
two families when 'the marriage nears completion (Zulu, Shangana-
Tonga). Simultaneous with these expressions of hostility are efforts
at friendliness. When lobola is handed over the girl's father will
slaughter a goat or beast, part of which may be sent to the boy's
people. Meanwhile ceremonial visits between the boy and girl will
be taking place ; among the Tonga these visits are from both sides,
among the Lobedu only on the part of the boy. Such visits help to
promote feelings of friendship.

When finally, after years of negotiation, the bride sets out for her
new home, 1 a beast will, in most tribes, be slaughtered to secure for
her the blessings of her ancestors. 2 Among the Shangana-Tonga,
this is the marriage feast at which boy and girl are publicly recognized
as man and wife. The Xhosa do not hold a feast at the girl's kraal,
both families contributing towards that held at the groom's ; other

1 Among Venda, Pedi, and some Tswana a man lives with and must work for his father-
in-law for a year or more before being allowed to take home his wife.

1 Among the Tswana and Southern Sotho the sheet of fat round the entrails is rubbed with
medicine and placed round the girl's neck, probably to secure fertility. Among the Huruthse
this ceremony seems to be associated with payment of lobola^ for if a child whose mother's
lobola has not yet been paid has to enter the initiation school it is performed on the child to
give it the necessary status.

n6 EILEEN JENSEN KRICE

tribes have feasts at both kraals. The girl must show extreme relu&ance
to leave home. Among Sotho tribes the bridal party sits down
at frequent intervals, refusing to go further till persuaded by a present ;
among the Zulu the bride makes various attempts at flight during
the festivities at the boy's kraal.

While showing unwillingness to join the boy's group, the girl
does not leave home empty-handed. She takes with her presents
for various members of the boy's family, her own household utensils
and, among Nguni tribes, also a number of beasts, one of which is
slaughtered at the wedding feast. 1 The exchange of gifts between the
two families does not stop at marriage. The Zulu woman once or
twice a year carries beer to the kraal of her son-in-law and a Lobedu
wife never returns empty-handed after a visit to her former home.

For the bridegroom, marriage is a transition to full manhood ;
for the woman it is this and more. She has to be aggregated into a
new family and for her the proceedings are marked by separation
from her own home and seclusion at the new one until she is accepted
into the family. 2 One of the first aggregation rites for the girl takes
place, among the Zulu, when she has to touch the gall of the beast
slaughtered at the marriage by her husband's people. But the process
of incorporation is a slow one, marked by very humble and submissive
behaviour on the part of the bride and the observation of a number
of taboos. The entry of bride and groom into their new life is marked
sometimes by mutual inoculation or " mixing " ceremonies (Venda
and Lobedu), sometimes by sitting together on one mat or being
smeared with fat (Shangana-Tonga and Southern Sotho).

No marriage is complete without children. Should the wife be
barren or die childless, a sister or other younger relative would have
to take her place. 3 Any children born will be considered children of
the barren sister, and the seed-raiser has no hut or status of her own.
No lobola is due for her, 4 though a gift of from one to three head of
cattle is usually given to thank the parents. If, however, the deceased

1 The Xhosa girl brings from one to three beasts, and often many more if her father is rich.

* She is, however, never fully incorporated. When she is ill, sacrifice must be made to
her own ancestors. For a comparison of the position of married women in the village among
Nguni and Sotho respectively, cf. above, Chapter IV.

1 The Huruthse go so far as to send a woman who has already been married, if necessary ;
while Zulu often send a younger sister of the bride as her inhlan^i to remain with her in her
new home and take her place, if need be*

4 This is mentioned for the Southern Sotho (Whitfield, 1929, 175), Swazi (Engelbrecht,
1930, 6), Venda (Lestrade, 1930, 202), Huruthse (Lestrade, 1926, 941-2), and is common all
over the Union. Among Pedi a " nominal charge " never exceeding three head of cattle,
is always made (Whitfield, 1929, 208).

INDIVIDUAL DEVELOPMENT "7

has borne children, her husband would have to pay lobola in the
usual way on marrying her sister. Among some tribes, e.g. Shangana-
Tonga, 1 Ndebele, 2 Swazi, 3 marriage with the wife's sister is at all times,
not only on the death or sterility of the first wife, considered very
appropriate, and it is usual for a Swazi man when visiting his wife's
people to engage one of her sisters to spend the night with him. If
a child is born he would either give lobola for the girl or, if on friendly
terms with her people, he might be allowed to give a beast and take
the child, leaving the girl to be married to someone else.

A form of marriage related to the sororate is that with the wife's
brother's daughter. Among the Venda, a man may, failing his wife's
sister, be given one of her brother's children. 4 This brother, having
used his sister's lobola, is responsible for satisfying the just claims of
her husband. Among the Swazi, too, a man is entitled to his wife's
brother's daughter, 5 while Shangana-Tonga and Lobedu consider
this kind of marriage appropriate at any time. The Lobedu kinship
term for a girl's father's sister's husband is " husband ".

Should a woman's husband die there will never be the bogey of
destitution before her. She will be able to remain in the village of
her husband's people and be cared for by them. The levirate obtains
everywhere except among the Xhosa, where the brothers of the
deceased are not allowed to have sexual connexion with the widow,
though even here she is expefted to " raise seed " to the deceased by
intercourse with strangers. 6 Though in certain circumstances an
elder brother may inherit the wife, 7 it is, on the whole, the junior
levirate that obtains, and the union is always recognized by the
slaughter of a beast, accompanied by some sort of purification. In
cases where a son succeeds his father as village-head, with responsibility
for all kraal inmates it is, among Sotho and Shangana-Tonga, possible
for him to have sexual relations with a young wife of his father, never,
of course, his own mother. Other tribes look upon such a union with
horror. The widows are always allowed to choose which of the
eligible male relatives they would like as seed-raiser, and, in general,

1 Junod, 1927, i, 247. * Eiselen, 1928*:, 417-

Engelbrecht, 1930, 6. * Stayt, 1931, 152.

*. Engelbrecht, 1930, 2. _ , v .,

The levirate is nevertheless present in all neighbouring tribes Mpondo, Thembu, Xesibe,
Bhaca, Fingo, Hlubi. .

7 Among the Venda either an older brother or a younger may " raise seed , but among
the Huruthse, if an elder brother inherits a woman, the resulting children are regarded as his
own and inherit from him. Here it has, however, been known for a paternal uncle to raise
up seed in the ordinary way (Lestrade, 1926, 242). It would seem that among the Swazi
the eldest brother of deceased has a prior claim to the chief wife (Engelbrecht, 1930, 9).

Ii8 EILEEN JENSEN KRIGE

a woman who has borne her husband a number of children would not
be prevented from returning to her own home as long as her children
remained in their father's kraal. The socially approved aftion, is
however, to remain where she is and raise seed to her husband.

After marriage the individual is recognized as a full member of the
community with all the privileges and duties that this entails. As the
years go by, his status grows ever greater in this society in which
respeft for age is so marked a charaftertistic. Mature age is the Bantu
ideal : the old man is honoured for his wisdom and experience while
the old woman enjoys respeft and freedom from many taboos of her
younger days.

CHAPTER VI

DOMESTIC AND COMMUNAL LIFE
By G. P. LESTRADE

MOST of the aspe&s from which Bantu tribal life may be regarded
social, political and economic organization, law, religion, art, and
ceremonial have received fairly adequate and connected treatment
in the scientific literature concerning these peoples. The more prosaic
features of domestic and communal life have, however, been curiously
neglected. Either anthropologists have felt that these sides of Bantu
life are of minor importance compared to the other aspects with which
they have dealt ; or, perhaps, they have considered that daily routine
and even special occasions are made up of a series of incidents forming
part of one or other set of cultural activities which they have described,
so that such routine and such occasions need no independent descrip-
tion of their own. The result of this attitude has been that these
subjects have had but scant and sporadic treatment in the scientific
literature, and that it is chiefly in popularized, imaginary and often
rather fanciful portrayals of tribal life that we have any detailed and
connected descriptions of the daily routine of a typical individual,
family, clan, or tribe, or of the communal activities in which groups of
these from time to time indulge.

Since the conditions under which the various Bantu-speaking
peoples live vary considerably, and since their social and economic
systems show many important differences, it is impossible to give
one generalized picture that will be applicable to all, nor would a
detailed description of the life of one particular tribe in this regard
serve as a guide to the rest. Under the circumstances, it has been
thought wise to adopt here a method midway between the two
extremes, and to give a fairly generalized description of the domestic
and communal life of a single group of tribes. The group chosen
is that with which the writer happens to be best acquainted, namely
the Northern Sotho. Such a description will, it must be repeated, not
suit other tribes, and will not, in its details, be entirely applicable to
every tribe even in the group chosen. But in its main features it may

119

120 G. P. LESTRADE

be taken as applicable not only to other Sotho tribes but also to the
adjacent Venda and Shangana-Tonga of the Transvaal ; and will,
it is thought, form not too inaccurate a guide to the life in these respefts
of the other Bantu-speaking peoples of South Africa.

It must be stressed beforehand that daily routine and the nature of
communal a&ivities are influenced to a very important extent by a
number of fa&ors in tribal organization. Chief among these are
habitat, domestic and communal organization, the division of labour
among sexes and according to age, and the distribution of special
offices and funftions among individuals. As all these are described
in detail in other chapters, it will be unnecessary to do more than
merely mention them here. Further, no attempt will be made here to
deal with the life of children, since this also is dealt with elsewhere.
Our main concern will be with adult life as lived from day to day,
and as interspersed with special occasions, in the tribal village.

THE DAILY ROUND
The Village.

Let us imagine, as the scene upon which the events we are going
to describe take place, a typical village of this area : a cluster of huts,
a dozen or so in number, with an open square leading into the settle-
ment, which also has a council-place, a cattle-kraal, and smaller
stockaded enclosures. Each hut is surrounded with a courtyard, and
has various appurtenances, such as storage-places, drying-places,
cooking-places, etc. The various courtyards are separated from each
other by palisading, winding alleyways connefting courtyard with
courtyard. The whole forms a compact and close unit of habitation.
The lands for cultivation, the pasture-grounds for the stock, the
watering-places, the drinking-water supply, the washing-places, the
spots where wood is gathered all these are, however, a considerable
distance from the village which uses them. The surrounding villages
are also not near-by reckoned by European standards : the average
distance between villages is at least a mile or so. The chief's village is,
it may be, a good many miles off. A village such as we have been
describing would be occupied by a man with his wives and their
children, together with perhaps an unmarried younger brother of
the man and, it may be, one of his father's widows, while the wives
of young married sons would help to swell the number. The huts
would be allotted for occupation by the inhabitants of the village on

PLATE XI

(a) Drawing water (Kxatla-

f/.

. . L (?) Smearing a courtyard floor (Kxatla)

(llic designs on the floor arc traced out with a mixture of fresh cowdunic and

earth)
HOUSEHOLD OCCUPATIONS

PLATE XII

() Children at play (Kxatla

(b) An informal beer drink (Kxatla)
DOMESTIC SCENES

. Schape,a

DOMESTIC AND COMMUNAL LIFE 121

the principle that each wife, with her children, is entitled to a hut
of her own, while the other huts are allotted according to the sex
and marital condition of the remaining inhabitants. The head of the
village occupies each of his wives' huts in turn, young unmarried
men and elder boys share a hut or huts as do young unmarried women
and elder girls, and young married couples are allotted a hut to them-
selves. Let us follow such a little community through one of its
typical days.

Early Morning.

The Bantu are early risers, however late a night they may have had.
They get up as soon as the sun shows above the horizon, or even
earlier on the slightest provocation, when, in their own phrase, the
horns of the cattle are just visible. Since they sleep completely naked
in their blankets, since full bathing, when done, is done at the bathing-
places some distance off and later in the day, 1 and since other ablutions
are performed in public view, their first aftion after waking is to dress,
after which the huts, which have been shut tightly all night, are opened
to let in some light and air. The women usually get up first, and wash
before going about their morning work. They then see to it that there
is water for the menfolk to wash, and proceed to their other work.
If the water supply is some distance off, washing-water may have
been fetched the previous evening, as well as wood for making the
fires. If the watering-place is comparatively near-by, the women
go out to fetch water and gather wood. In the meantime the men
have got up, though, especially if they have been several times during
the night to have a look at the cattle-kraal, they may sleep a little
longer. Youths and boys, who have to take the stock to pasture, are
up betimes, and may break their fast with some scraps of food left
over from the previous evening's meal before doing the morning
milking. If there is nothing left over, they may either take some
uncooked meal and some cooking-utensils with them to prepare their
morning meal in the veld, or they may eleft to live on such small
game as they can trap or kill there, or they may even go hungry till
they return home ravenous for the evening meal. Those who have
to be off on a journey are also away betimes, with or without some
scrap of food. By this time the men are also up. Early morning washing

1 Except by young people who have recently gone through the initiation ceremonies;
these, for some period after they have returned home, keep up the habit, acquired during the
initiation-period, 'of bathing in the early morning.

122 G. P. LESTRADE

is not so obligatory with them as it is for women and recently-initiated
youths, and many eleft to remain unwashed till just before the morning
meal. They usually sit for a little while outside in the sun, and issue
orders to the younger men about what is to be done during the day.
The latter are also soon off to their various tasks, such as building or
repairing a hut, or any other of the manual work which falls to their
share. The men then repair to the council-place, where, unless other-
wise engaged, they usually spend most of their day. The younger
women and girls who may not have gone with the other women
to fetch water and to gather wood now set about other tasks in the
home, such as sweeping and cleaning, washing cooking-utensils,
stamping maize, and looking after the children.

The Morning Meal.

As the middle of the morning is reached, the women who have been
out return with water and wood, and begin to prepare the morning
meal, the first of the two meals of the day. This usually consists only
of porridge, though occasionally some relishes are also prepared with
it, such as monkey-nut sauce or a sort of puree of wild spinach. When
this is ready, usually between ten and eleven o'clock, it is ladled out
into the wooden food-bowls, and allowed to cool and harden slightly.
Those who are out in the veld or on a journey of some little distance
do not return home for the morning meal, nor do the men and youths
who are at the council-place or engaged in manual labour respectively.
The women and children eat their share at home in the courtyards,
usually under the eaves of the huts. The share of the others is brought
to them in their bowls by the younger women, usually together with
water for washing the hands. Nothing, except perhaps a calabash of
water at the end, is drunk with the meal. But the men at the council-
place may already have drunk a little beer supplied to them by the
head of the village. The men usually eat in age-set groups, sharing
their food with one another. The young men and youths at work
eat where they happen to be, the food being brought to them there
by their people. After the morning meal is consumed, the bowls
are fetched back by the younger women, and the work of the day
goes on.

The Later Morning and Early Afternoon.

It is here that a brief reference to the activities of the men at the
council-place is best made. Most of the time is spent in talk, with or

DOMESTIC AND COMMUNAL LIFE 123

without a background of work to accompany it. The talk may be
merely desultory, in which case most of the men engage in some seated
handicraft as well braying skins, carving objefts out of wood or
horn, making wire-work, weaving baskets, etc. The talk may also
be important, in which case accompanying work tends to stop. Tribal
and village news is exchanged, affairs are discussed, messengers come
and go, visitors are announced and received, petty disputes are talked
over and settled, lines of common aftion are determined upon. At
ordinary times there is a pleasantly informal air about the proceedings
at the council-place. The men sit about in scattered fashion, on the
ground, up against a tree, on logs of wood, or, if they have them, on
little stools, which they carry about with them or leave where they
usually sit. They may even lie about, and, if there is no important
business going on, if they have had a late or interrupted night, and if
the day is warm, they may indulge in a quiet nap while the others go
on talking.

But this informality changes when something occurs which demands
more formal attention. Then everybody sits politely upright in a
body, listening attentively to what is being said and taking an aftive
part in the proceedings. It may be, for instance, that an important
visitor arrives, or that a matter of some moment is before them. Then
the loose desultory talk stops, side-pursuits are suspended, everybody
is quiet, paying close attention to what is happening, and speaking each
in his appointed turn as occasion requires. The formal procedure in
connexion with the trial of petty cases will be described elsewhere :
this is perhaps the best place to indicate briefly the pi&uresque
ceremonial, common among the people we are describing, in the
formal reception of an important visitor other than the chief of the
tribe. The visitor, accompanied by his own entourage of spokesmen,
first comes to the outskirts of the village, where he is met by a
messenger, to whom his spokesmen announce him. The messenger
departs to the council-place, and soon returns inviting the visitor and
his men to accompany him there. They are seated on the outskirts
of the group. After a little preliminary silence, the senior man in
the council-place asks after the visitor's health. The word is passed
from mouth to mouth through a number of intermediaries, and finally
reaches the visitor's entourage, who communicate it to him. He
answers in the same way, but in the reverse order, speaking first to his
own spokesmen, who pass the word to the hosts, till it finally reaches
the first speaker. Several such formal questions and answers pass

124 G. p - LESTRADE

between the parties for a while. After a time the proceedings become
less formal and cumbersome, the intermediary spokesmen gradually
fall out, and the visitor talks direftly with his chief host. When the
time comes for the visitor to leave, formality is resumed, ending
in formal permission to depart, and the visitor and his men being
formally conducted away from the council-place by the messenger.
If beer is available, some will be brought and consumed, the hosts
drinking first to show that it is not poisoned. If there is no beer, the
hosts will apologize for its absence, usually in the hyperbolic phrase
that the whole village is dying of hunger. If the visitor intends to
make a somewhat longer stay, food will be cooked for him and his
spokesmen and brought together with that of the other men. For a
very important visito^ a goat or sheep will be killed, or, on a first-
class occasion, a beast.

While the older men are at the council-place, the younger men and
older boys, or such of them as are not out with the cattle, will be busy
at their manual labour. The proceedings here are of course quite
informal, many jokes are interchanged and much talk and shouting
goes on. The exchanging of rude jokes with the younger women who
pass by, on the part of the young men who may stand to them in
cross-cousin relationship, forms a rather noticeable feature in the
course of the day's work. Occasionally, one or more of the older men
from the council-place will come to inspeft the progress of the work,
and issue further orders as to how it is' to be carried on.

The women, in the meantime, busy themselves about their manifold
household duties. One or two will remain at home, usually the older
ones, to look after the home and the small children, while perhaps
one or two others will busy themselves making pots or stamping
maize or preparing beer. Others, accompanied by young girls, and with
their small babies on their backs, will set out to the lands, where they
remain till the late afternoon weeding, hoeing, sowing, reaping, etc.
When the work in the fields is done for the day, they will go to the
water-supply to fetch water, and will gather some wood, all in pre-
paration t for the evening meal. The work on the lands is often
performed communally. In theory, every woman has her own piece
of land, which she is supposed to look after herself, and for which
she is responsible. In prafHce, women often club together over this
work. One woman may invite a number of others to help her work
her piece of land, on the understanding that she will help them with
theirs when the time comes. As a reward for this co-operation,

DOMESTIC AND COMMUNAL LIFE 125

she usually provides them with some beer and, it may be, some
food, which is consumed when the hard work is over.

It is toward the commencement of the late afternoon that the usual
time for bathing occurs. Older and younger people, men and women,
have their own appointed bathing-places. The women usually bathe
on their way home from the fields, and the others who have remained
at home, men and women, go out to bathe when the day's work
proper is done.

The Late Afternoon.

The later afternoon ushers in the less strenuous part of the day.
Important matters at the council-place are settled by then, and it may
be that someone, perhaps the head of the village, provides some beer
for the men to drink. Beer-drinking is carried on with a fair amount
of ceremonial. The pot is brought by a young woman, who places
it before the men. Several calabashes are brought at the same time.
One of the men, usually the right-hand man of the village-head, takes
charge, ladling out the beer with one calabash and pouring it into the
others, which are handed round in turns. The man whose wife has
provided the beer usually drinks first, to show that it is not poisoned.
Others drink after him, the few calabashes being used by all the men
in turn. Older and more important men are given the first share,
others wait till later. If there is nothing to do at the council-place,
the men will wander off, some going to visit friends, exchanging
gossip and snuff and possibly getting a calabash of beer in the process :
or they may go off to a regular beer-party somewhere in the neighbour-
hood, perhaps at the chief's kraal, or at some other village where some
festive occasion is being celebrated. As the afternoon wears on and
the men are more and more full of beer, voices grow higher and
tempers may rise, vociferous disputes and wranglings and sometimes
a&ual fighting may occur. In a plentiful season, no man reaches the
end of the afternoon perfectly sober.

The women also, on their return from the fields, or after their day's
work at home, indulge in a certain amount of relaxation till the rime
comes for preparing the evening meal. They go out visiting,
exchanging gossip, and, in the case of the older women, getting
some snuff, and, if it is a fairly plentiful season, usually managing to
obtain some beer as well. Women do not sit in on men's beer-parties,
but have some beer provided for them separately, which is drunk
with considerably less ceremonial. In the course of such visits, a

126 G. P. LESTRADE

certain number of business transa&ions, such as the purchase and sale
or exchange of household utensils, take place, and a good deal of
good- or ill-natured wrangling takes place over such and other items.
As the afternoon wears on towards dusk, the youths and boys come
home with the stock, which has fed and been watered. The men will
by this time be drifting home gradually, and, as soon as the stock
has returned, they will go to inspeft it, after which the evening milking
takes place. If a goat, sheep or beast has been slaughtered, its meat
will have been cut up by this time, and the various portions will be
allotted to those entitled to them. There is a regular system about
the division of the various parts of the carcass, certain houses and
certain persons being entitled to certain fixed portions. The same holds
of any animal that may have been killed in the chase and which is
brought home to be divided.

The Evening Meal.

The younger women will by this time have got the fires going and
the cooking-pots on. The women, old and young, now busy them-
selves about the preparation of the evening meal. Its basis, like
that of the morning meal, is of coarse porridge ; but it is a more
substantial meal. Meat, or in its absence some relishing sauce,
forms a substantial accompaniment to the porridge. The porridge is
once again ladled into the individual wooden food-bowls, the sauces
and meat are served up in cbmfnon dishes made of clay. When the
food is ready, the members of the family sit down to the evening meal.

The evening meal is essentially a family affair. For the morning
meal, as has been stated, the members are dispersed and served where-
ever they happen to be. But everybody is at home for the evening
meal, including of course any guests that may have arrived. The meal
is eaten outside, usually under the eaves of the huts, but sometimes
elsewhere. Among the tribes under strong Venda influence, important
persons, and nearly all the men, eat inside the huts. The older men,
the younger men, the women, and the children, constitute little groups,
each eating apart from the others, with a common meat-dish and
individual porridge-bowls. Eating is done with the fingers, the hands
being usually washed before the meal is taken. The children often get
only the scraps of meat left over by the grown-ups, but are allowed
literally to lick clean the grown-ups' platters. Since the evening meal
is the only substantial one of the day, the quantities of food consumed
at a sitting are very great according to European standards of capacity.

DOMESTIC AND COMMUNAL LIFE 127

The youths and boys especially, who have been out on scant rations all
day, put away enormous amounts of the thick heavy porridge.

The Evening.

After the evening meal the whole family relaxes, and family life
appears in its most pronounced and cosiest form. People sit round the
fires, the men and older boys in one group, the women and children
in another. It is then that tales are told. The men will relate such of
the day's happenings as they are disposed to communicate, or will tell
hunting-stories or tales of long-ago events ; or they may indulge in
some of the typical men's games, such as the moruba game, 1 or they
may exchange proverbs, or, in lighter mood, riddles, and generally
joke and pass the time away. The children crowd round the older
women, for this is when grandmothers' tales are told tales of ogres
and witches, animal-stories, and the many other tales referred to in
the chapter on tribal literature. The children indulge freely in riddle-
games, and in other games of various kinds. For the young people,
the evening is the time of love-making, to which the older people,
still under the benign effefts of the afternoon's beer and the soporific
influence of the heavy evening meal, turn a judiciously blind eye.
Assignations are made and trysts are kept, especially at times when
there is dancing of aa evening, as often happens on moonlight nights.
Dances are the only things, as a rule, that entice people away from their
own homes in the evenings ; but dancing of this kind, as distinft
from ritual dancing, which falls into another category, is for the
younger people, the elders contenting themselves with looking on
and consuming such beer as may be provided.

When the evening draws on toward night, the inhabitants of the
village retire to rest one by one as sleep overtakes them. Some of
the men may go out for a last reassuring look at the cattle-kraal,
the women will look to see that the fires are dead or at least so far
died down that there is no risk of anything catching fire during the
night. The courtyard entrances are barred, and then everybody
retires to his or her hut. Sleeping-mats are spread out, some improvised
headrests are put ready, and, dressed only in their sleeping-blankets
or skins, young and old retire to rest. The doors are tightly closed,
and windows there are none. Everything must be shut to keep off
not only possible evildoers but, more dangerous still, the evil spirits

1 Described for the Shangana-Tonga people by Junod (1927, i, 345 ff-) under the name
tshula. The name mancata is often applied to this game in general anthropological literature.

128 G. P. LESTRADE

that go about during the night and get into the huts through any
available openings to attack the defenceless inhabitants. During the
night, especially if the cattle are uneasy, some of the men will be up
several times and go to the cattle-kraal to inspect. For the rest, every-
thing is peaceful until dawn, when the next day's round commences.

SEASONAL CHANGES IN DAILY LIFE

As may be imagined, seasonal changes play a considerable part
in modifying ordinary routine. During the ploughing- and reaping-
seasons, for instance, when every able-bodied woman and, especially
nowadays, every available able-bodied man, is at work in the fields
for a considerable part of the day, the villages stand practically empty
except of children and very old people for a great deal of the time ;
and daily routine is changed accordingly. In the middle of winter,
when pasture for the cattle is scarce, the flocks with their herdsmen
have to be taken far afield to graze, often so far that it is impossible
for them to go out and come back in the same day, so that the cattle
are kept for some period at the cattle-posts, the herd-boys staying
with them, with a consequent change in the habits of the village. At
another time, the chief may call out most of the young men to perform
their share of labour for him, and the younger women may be similarly
called upon to till the chief's fields ; and again such a process results
in a considerable disturbance of ordinary daily routine. It would be
impossible and unnecessary to enumerate and describe such disturb-
ances of daily routine in detail ; but their possibility must be clearly
kept in mind.

OCCASIONAL INTERRUPTIONS OF DAILY ROUTINE

But there are various occasional disturbances of daily routine
which we shall have to note, since it is on these occasions that we
find the communal, as distinct from the domestic, side of village
life in its most pronounced and striking form. We refer here to
festive occasions, such as a wedding, or a birth-ceremony, or the return
of the initiates after their period of seclusion in the tribal schools, or,
in fact, any occasion of the kind which serves as a reason for inter-
rupting the normal flow of daily domestic life.

Let us imagine ourselves on the eve of the wedding of some
important man's daughter. We will remember that, for a considerable

DOMESTIC AND COMMUNAL LIFE 129

period, the even tenor of life in the villages has been slightly disturbed
by the going to and fro of the messengers who have arranged the
details of the marriage, which, incidentally may have formed one of
the staple topics of conversation round about for some time. But now
the arrangements are made, and the day for the a&ual wedding-
ceremony is near. Messengers are sent out to the neighbouring villages
to acquaint them formally with the faft that the wedding is to take
place, and to invite them formally to be guests thereat. No invitations
are sent to those blood-relations who will attend as of right, nor are
invitations sent to those whose presence is not desired, although the
absence of such an invitation seldom succeeds in keeping them away.
For some days beforehand, an air of excitement prevails in the villages
affected. Many preparations are made : the men decide which of the
cattle are to be slaughtered for the occasion of the feast, the women
are busy making beer, stamping maize, and preparing the ingredients
for the dishes that will accompany the porridge and meat.

On the day itself, excitement rises to a very high pitch. Little if
any work is done. All activities that can wait at all are completely
suspended, and even essential services, such as the pasturing and
watering of the cattle, are scamped as much as possible. In the very
early morning, life goes on very much as on ordinary days, except
for the prevailing air of excitement. But soon the nature of the day
makes itself manifest in the unusual nature of people's activities. The
men proceed with the killing of the animals destined for the feast :
there is much shouting at every step in the business, much earnest
confabulation about how the pieces are to be cut and distributed. The
women have got the beer in readiness, and a portion of it is consumed
already in the morning, while the morning meal is usually but a
scanty one in anticipation of the feast to follow.

The aftual wedding ceremony now begins to take place. This is
described elsewhere in this book : but some indication ought to be
given of the demeanour of hosts and guests, of young and old, of
men and women. It is the duty of the hosts to provide the most lavish
hospitality they can, aided by their relations and intimate friends :
and the guests are by no means modest in their demands, or shy about
accepting their rightful share of the hospitality, or about voicing their
complaints when it is not lavish enough to their taste especially when
they have already had enough beer to loosen their tongues. The Bantu
soon grow noisy in crowds, especially when there is some beer going,
and the din at a wedding feast is enough to make the uninitiated think

130 G. P. LESTRADE

that some violent free fight is going on, when in faft people are only
amusing themselves talking and exchanging jokes and a few personal
pleasantries. The younger people, who are not given any beer, are
usually fairly quiet, although many a conversation has to be shouted
to be mutually audible. The older people let themselves go fairly
thoroughly, particularly the older women, who will from time to
time break out into high-pitched ululations, trilling with their tongues ;
or, if they are very excited, doing mimic dances in pas seul, running
about, accosting people in an otherwise quite unseemly manner, and
generally behaving quite differently from the way they do on ordinary
occasions. The men are a little more restrained in their enjoyment :
but, as the beer begins to do its work, they too become noisy. One
may start a mimic dance, imitating the killing of an enemy in battle
or of an animal in the chase ; another may burst forth into a praise-
poem, reciting his own praise-poem or that of the bridegroom's or
bride's clan ; yet another may start a song, or, in quieter mood, begin
telling some interminable tale to which few listen. The children lark
about, playing games and generally getting in everybody's way,
while dogs and pigs run about squealing and adding to the hubbub.
As food is brought in the course of the feast, there is no general return
home for the evening meal as on ordinary days, and the meal, being
one of the central points of the day, is usually got ready and eaten
some time before the usual hour. Men, women, and children all eat
in separate groups, and are further divided according to age-sets,
each person eating in company with his or her coevals, separately
from the others.

Towards evening, a dance usually ends up the proceedings : there
is more beer, more excitement, more shouting, more din generally,
and merriment is kept up till late into the night. A considerable
amount of sexual licence prevails, especially among the younger
people. Impromptu dancing- and singing-competitions are held,
" choosing-games " are played, and these usually result in a clandestine
meeting later outside. When the night is so far advanced that even the
most indefatigable reveller is tired out, people start going home as
best they can. Soon the village sinks into the calm of exhaustion, to
return next day to its ordinary humdrum routine. But the events of
such a day form an ever-fruitful topic of conversation for a long time
afterwards.

CHAPTER VII

WORK AND WEALTH
By I. SCHAPERA AND A. J. H. GOODWIN

COMMODITIES AND THEIR PRODUCTION
Foodstuffs*

THE Bantu live partly upon the natural resources of their environment.
Wild animals provide them with meat, and wild plants and fruits
with vegetable food. But they control the bulk of their own food-
supply by cultivating plants and breeding livestock. Their principal
crops are maize, introduced originally from America via Europe,
and millet or " Kafir corn " (sorghum vulgare), indigenous to the
country. Kafir corn is still the staple food in Bechuanaland, but almost
everywhere else is now less extensively used than maize ; except for
making beer. Sweet cane (eleusine) and several species of melons,
pumpkins, peas, beans, ground nuts and sweet potatoes are also grown,
but in lesser quantity. The principal domestic animals used as food
are cattle, goats, sheep, and fowls. The cattle originally kept seem
to have been of two main stocks, the one small, long-horned and
straight-backed, the other large and humped ; but they have become
considerably mixed with breeds introduced by the Europeans. The
goats and sheep are short-legged, long-haired animals of hardy
but inferior stock. The fowls are mostly of the common small hybrid
variety.

The principal daily dish in most tribes is of maize or Kafir corn.
Green mealies (maize) are sometimes boiled or roasted on the cob ;
but more generally the shelled grains are either roasted dry to pop-
corn or boiled in water until soft, or stamped and then boiled. Kafir
corn is generally stamped or ground, and then stirred into boiling
water to make a stiff porridge. Fire, it may be noted here, is every-
where made by friftion : a thin, hard stick is twirled rapidly between

1 The sections marked with an asterisk were drafted by A. J. H. G., and extensively revised
by I. S., who is also alone responsible for the rest of the chapter. Apart from museum specimens
in Capetown, studied for the sections on technology, the following secondary sources have
been principally used : Junod, 1927 (Shangana-Tonga) ; Krige, 1936 (Zulu) ; Hunter, 1936,
and Soga, 1932 (Cape Nguni) ; Stayt, 1931 (Venda). Quotations, abstracts, and specific
citations are acknowledged more fully below. All references to the Tswana, unless otherwise
indicated, are based upon observations in the field by I. S.

132 I. SCHAPERA AND A. J. H. GOODWIN

the hands into a notch cut in a softer stick, and the resulting spark
caught up in dry grass. The subsidiary crops are sometimes eaten
separately, but serve mainly as relishes with the staple foods. They too
can all be prepared in several different ways to lend considerable
variety to the daily menu. Wild fruits and berries, a welcome addi-
tion in season, are generally eaten raw. Wild herbs are boiled in a
little water to form a vegetable stew, often eaten mixed with meal.
They are regarded as essentially women's food, eaten by men only
when away from home hunting or travelling. In times of crop failure,
however, they may become indispensable food for all. Locusts, when
procurable in large quantities, are dried and ground to meal ; while
caterpillars and flying ants are often made into sauce for flavouring
food.

The milk of cattle and goats is everywhere an important item of
food, especially among the Nguni, where it rivals cereals as the staple
dish. It is generally allowed to ferment and thicken in a calabash or
skin bag. It is then either eaten separately, or mixed with some other
dish like boiled ground maize. Among the Nguni no person may take
it as food at the home of people not related to him ; while women may
not eat it at all when menstruating, immediately after childbirth,
newly bereaved, or in some other way ritually impure. Sweet milk is
often given to children, but seldom drunk by adults ; although when
cooked it is freely used in the preparation of some vegetable dishes.

The only other beverage, apart from water, is beer. Several varieties
are made. The most popular, known as utskwala (Nguni) or iyalwa
(Sotho), is brewed from Kafir corn. It is less alcoholic than light
European beers, and must be taken in large quantities to produce any
marked effect. It contains a fair amount of solid matter, and so is
nourishing as well as stimulating. Children, however, are not as a
rule allowed to drink it. A much milder beer, known as leting (South
Sotho) or mahewu (Nguni), can also be made from Kafir corn, although
maize is more generally preferred. The Venda and Shangana-Tonga
make other drinks, often highly intoxicating, from the fruits of the
morula tree (ScUrocarya caffrd), the leaves of the prickly pear (Opuntia
vulgaris), or the sap of the fan palm ; and some Tswana tribes brew
a potent honey mead. These beverages, however, are not so commonly
used, and on the whole are not so well liked, as the ordinary grain
beer. 1

1 For a detailed account of Native beverages, their alcoholic content and nutritive value,
cf. Turner, 1909, 31-54.

WORK AND WEALTH 133

Beer is regarded by the Bantu as an essential food. In times of
plenty it is not only freely consumed, but often is the principal or
even sole food for many men for days on end. But it also plays a very
important part in their social life. " The whole social system of the
people is inextricably linked up with this popular beverage, which
is the first essential in all festivities, the one incentive to labour, the
first thought in dispensing hospitality, the favourite tribute of subjeQs
to their chief, and almost the only votive offering dedicated to their
spirits." l Among the Lobedu, a Northern Sotho tribe, " beer is a
common means of exchange or payment for services rendered, and
is in evidence in a number of transactions that in other tribes are
conducted by means of mealies and goats " ; and " it is in evidence
in almost all ritual and ceremonial as a celebration of important
occasions, binding together different groups or individuals, (and)
effefting a reconciliation where things go wrong ", 2 To a considerable
extent this applies also to all other tribes ; and it may be added that
in most of them the harvest thanksgiving takes the form of a national
beer-drink, preceded by an offering of beer to the ancestors of the
Chief.

The Bantu are much less fastidious than Europeans about the
kinds of wild animal they will eat, although carnivora, dogs, monkeys,
crocodiles, snakes, as well as such birds as the owl, crow, and vulture,
are generally taboo. The Sotho and Venda also prohibit people from
eating their own totem animals, a taboo not found in other tribes,
although certain clans among the Swazi and Transvaal Ndebele
have analogous taboos. The Shangana-Tonga eat fish, but most other
tribes, even those living along the coast, refuse to do so, regarding them
as unclean food akin to snakes. Hunting, however, is carried on too
sporadically to be a regular source of meat. The domestic animals,
again, are slaughtered as a rule only on special occasions, although all
dying of disease or other causes are eaten. Meat accordingly figures
so seldom in the everyday diet as to be considered a distinft luxury,
eagerly sought for; and whenever available is consumed in large
quantities at a time. It can be prepared in several different ways, the
most common being to boil it in water or to roast it on wooden spits
over an open fire. Women may not eat the flesh of a domestic animal
which has aborted or died in calving, lest they become similarly
affii&ed; nor may they eat eggs, which are held to make them
lascivious. Among the Cape Nguni, too, men will not normally eat

1 Stayt, 1931, 48. E. J. Krigc, 1932, 347, 357-

134 I- SCHAPERA AND A. J. H. GOODWIN

poultry, regarded as essentially women's food ; while almost every-
where animals killed by lightning are taboo.

Meat and most vegetable foods are generally eaten salted. The
salt is either collected from natural deposits along the coast or in
salt pans, or, as among the Venda and Shangana-Tonga, is recovered
by evaporating solutions of salted earth. Water, used for drinking,
cooking, and washing is obtained from streams, springs, and other
natural sources. No attempt is made to conserve it, nor are any
precautions taken against pollution, cattle being allowed access to
the same sources of supply.

The principal narcotic is tobacco, formerly cultivated extensively.
It is taken as snuff, the ground tobacco leaves being mixed with the
ashes of some special tree, such as the prickly aloe, to make it more
pungent. It is freely used by both men and women. Some people
also smoke hemp, which, however, is not nearly so widely cultivated
or used.

* Horticulture.

Among food-producing a&ivities cultivation of the soil is on the
whole most important. The fields are in most tribes situated within
walking distance of the homesteads, from which they are visited
daily when necessary. But among the Tswana they are frequently
a good distance away, owing to the concentration of the people in
large village settlements ; and most families therefore have special
homesteads alongside their fields, where they go to live during the
agricultural season. Fields are scattered about somewhat irregularly
wherever suitable soil is available. The Bantu realize that soils vary
in fertility, and in selefting sites for cultivation are guided by such
signs as the kind of grass or bush growing on the land. Fields also
vary considerably in shape and size, but are on the whole fairly small,
seldom exceeding two or three acres in extent. 1 Isolated fields are
prote&ed by fences of dead thorn bush or some similar material ;
but where the fields of several people lie side by side they are not
usually enclosed, although uncultivated strips or other conventional
boundary marks clearly distinguish them.

The sajne field is cultivated for several successive seasons until
it shows signs of exhaustion. It is then left lying fallow, while new
land is cleared and worked. The Bantu thus praftise the shifting
cultivation characteristic of primitive agriculture in many parts of

1 Stayt, 1931, 34; Hunter, 1936, 72 f.

WORK AND WEALTH 135

the world. Among the Mpondo, and probably in most other tribes also,
" it is customary for each um^i (household) to turn over some new
ground each year, either by enlarging an existing field, or taking new
ground on one side, and leaving a part of the old fallow, or beginning
a new field altogether. New fields when begun are quite small, but room
is allowed for expansion, and the area is increased each year." l The
bigger trees on virgin land are burned down, the smaller ones, and
bushes cut with axes, and the grass and weeds uprooted with the hoe.
The loose foliage is then piled up in heaps and burned, the ashes
providing the only fertilizer the ground ever receives, apart from the
manure incidentally deposited by cattle grazing on the stalks after
reaping.

Planting generally commences as soon as the first rains have fallen.
The seed is scattered over the surface by hand, or spat out from the
mouth, and then progressively covered in with an iron-bladed hoe
or, as among the Cape Nguni, with a simple wooden spade or pointed
stick. The Shangana-Tonga, however, first hoe the ground and then
make shallow holes into which the seed is carefully placed, a few
grains at a time. 2 All the different crops are grown together in the
same field, their seeds being mixed and simultaneously sown. Since
most people cultivate more than one field, planting is a fairly long task.
As the young crops shoot up, they are thinned out by hand, while the
accompanying weeds are removed with the hoe. This is an arduous
and lengthy task, as weeding is generally done twice. Pumpkins,
melons, and sweet cane ripen first, and, together with green mealies
(maize), are harvested and consumed as they become available. No
one, however, may eat of these first fruits until the Chief has cere-
monially done so. Kafir corn and mealies take much longer to ripen.
From the time they start seeding, the people spend all day at the
fields, often perched on specially-erefted platforms, and endeavour
by throwing stones, shouting, and other methods to scare away the
granivorous birds flocking there in great numbers. The agricultural
season ends with the final harvest, when the ears of Kafir corn and
maize are broken off by hand, loaded in large baskets, and stacked in
temporary granaries until thoroughly dry. The stalks are left standing
as food for the cattle. Peas, beans, and ground nuts are generally
gathered in at about the same time.

Some ears of maize and Kafir corn are set aside as seed for the next
planting. The remainder are threshed. Kafir corn is threshed with

1 Hunter, 1936, 72 f. * Junod, 1927, ii, 23.

136 I. SCHAPERA AND A. J. H. GOODWIN

heavy wooden flails on a specially-prepared floor of hard, beaten
earth. Maize may be threshed in the same manner, but more often
is husked and shelled by hand. All grain not immediately required is
stored away. The Nguni and Venda generally bury it in pits dug in
the cattle-kraal. The pits are covered with flat stones, sealed with
cow-dung, and hidden by loose manure or earth. Supplies are taken
out as required, and the pit resealed. 1 Sometimes, however, they store
it in a wicker cylinder, plastered with mud, and perhaps raised on a
wooden platform. This form of granary is found in most other tribes
also ; but the Sotho and Shangana-Tonga more generally use large
baskets of coiled weave, or a clay drum raised on clay legs reinforced
by poles or stones and covered with a small thatched roof. 2

An essential element of agricultural technique is the employment
of magic, held to be indispensable to the success of every enterprise.
Certain " medicines " are buried in the corners and centre of the
field, or burned so that the smoke blows over it; in order to ensure
an abundant crop ; others are burned, buried, or otherwise used to
keep away worms, birds, hail, and other pests ; and still others are
used to prevent the diminution of the crops by witchcraft or theft
once they are on the threshing-floor or in the granaries. There are
also many taboos of various kinds to be observed by the cultivators
lest the growing crops meet with disaster. These medicines and rites
are independently used by every cultivator. 3

Other rites are performed by the Chief on behalf of the tribe as a
whole. The Bantu, as Miss Hunter says, have no scientific control
over the supply of water necessary for growth no irrigation is
used 4 ; but to satisfy this most pressing of their agricultural needs
they have developed " an elaborate magical technique ", in the form
of numerous rainmaking ceremonies performed either by the Chief
himself or by his professional rainmakers. In the well-watered parts of
South Africa these ceremonies are generally carried out only in times of
drought ; but in the more arid regions of the west, inhabited by
Tswana and Northern Sotho tribes, they take place every year. In
these tribes, too, various important taboos (e.g. on the cutting of
certain trees) must be observed by the whole community during the
early part of the rainy season, lest drought, hail, or some similar
disaster destroy the young crops. These taboos are proclaimed afresh

1 Hunter, 1936, 86; Krige, 1936, 44; Stayt, 1931, 36 f.

* Junod, 1927, ii, 27 ; Schapera and Goodwin, field data.

* Cf. Junod, 1927, ii> 28 & ; Krige, 1936, 192 ff. ; Hunter, 1936, 76 ff.
4 Hunter, 1936, 79.

PLATE XIII

/j" Kxatht fields

hmi; Horns can In* sfvii in ih- lorr^rouud in ihr light

Threshing corn (Kx.nl.i 1
(The woman in the foreground, right, has a baby on her back)
AGRICULTURAL LIFE

I face p. 136

WORK AND WEALTH 137

every year by the Chief, who also formally lifts them in due course.
In most tribes, again, the Chief distributes among his people seed
specially " doftored " with certain ingredients, including very often
portions of flesh from the body of a human being killed for the purpose.
This they mix with their own seeds to promote the fertility of the
crops. It may finally be noted that elaborate public ceremonies are
almost everywhere held before sowing, at the time of the first fruits,
and after harvest, when the Chief's ancestral spirits are either invoked
for a bountiful season or thanked for having provided it. 1

* Animal Husbandry.

Agriculture may be the principal source of subsistence, but the
Bantu themselves attach more importance to their cattle. Animal
husbandry is perhaps most strongly developed among the Nguni,
who inhabit the most favoured parts of the country ; whereas among
the Shangana-Tonga and Venda disease, warfare, and other faftors have
greatly reduced the size of the herds, and the presence of the tsetse-
fly in many parts is an insuperable obstacle to stock-farming. But
cattle are kept wherever it is at all possible. They are not merely
a source of food, in the form of milk and occasionally meat. Their
skins provide material for clothing, shields, bags, and other useful
objects ; their horns are made into receptacles; and their dung is used
both as fuel and as the cement plastered on walls and floors. The
oxen in many tribes serve as beasts of burden and as a means of
transport. Often, too, they are trained specially to race without
riders, contests between them being one of the most favoured sports ;
and such racing cattle are sometimes commemorated in tradition
long after their death.

Cattle are further " the principal medium of exchange, and the
medium in which court fines are levied. . . . (They) are the means
of keeping on good terms with the ancestral spirits, and so of securing
health and prosperity, because the maintenance of good relations with
the ancestor spirits depends upon making the proper ritual killings of
cattle at various stages in the life of the individual, and in sickness. . . .
Cattle are also the means of obtaining sexual satisfaftion, since a legal
marriage cannot take place without the passage of cattle, the right to
limited sexual relations is legalized by the passage of a beast, and the
fines for illegal relations are levied in cattle. The possession of cattle
gives social importance, for they are the means of securing many

1 For descriptions of these ceremonies, see below, Chap. XI.

138 I. SCHAPERA AND A. J. H. GOODWIN

wives and adherents, and of dispensing hospitality and showing
generosity, on which virtues status largely depends. Also the posses-
sion of cattle in itself gives weight and dignity to the owner "- 1

Cattle therefore loom largely in a man's thoughts. They are his
principal form of wealth, his most treasured possession ; and anything
concerning them and their welfare focuses his attention. The Bantu
languages abound in terms minutely differentiating cattle according
to sex, age, colour, and shape of horns, and reflecting the intense
interest taken by the people in their beasts. A man often knows his
cattle by name ; and his bull, the pride of his herd, is hailed in laudatory
phrases as it comes out of the kraal. The slaughter of an animal is
normally reserved for ceremonial occasions, whose importance is
heightened by the magnitude of the sacrifice. Some Cape Nguni
tribes, especially the Bomvana, even have sacred herds of cattle,
held inalienably by the Chief in trust for the tribe, and sacrificed to
his ancestors in times of national importance. Cattle raids are among
the most frequent causes of warfare, and cattle theft one of the most
serious offences. Kxomo modimo wa moxae, modimo wa nko e meetse ;
hcomo kotlanya dithSaba, o bolaile banna ba le bantsi, says a Tswana
song : " cow god of the home, god with the moist nose ; cow that
makes the tribes fight, you have killed many men " an apt summary
of both the religious and political importance of cattle.

The cattle-kraal, a circular enclosure of stout poles in which the
cattle are kept at night, is not only the central feature of most village
settlements, but also the centre of village life. Inside it the cattle
are killed for sacrifices ; here men meet for secret discussions ; women
may not enter unless they belong to the family, and even then not
when they are menstruating ; and here, too, the family head is often
buried, wrapped in the skin of a newly-slaughtered ox. Adjoining
it is the meeting-place of the men, where they sit daily to eat and gossip
round the fire and sometimes to feast, where justice is administered,
and where strangers arejreceived. Women may not come there except
when bringing food, and must in other ways hlonipha (respect) it.

But despite this intense pre-occupation with cattle, the Bantu
pay little atention to quality. For most social purposes one full-
grown beast is as good as another ; and if anything preference is shown
for animals of a particular colour or with horns of a certain shape rather
than for well-developed oxen or good milch cows. The cows give
little milk, and then only after calving, so that this important source

1 Hunter, 1936, 69.

WORK AND WEALTH 139

of food fluctuates considerably in yield. And since it is every man's
ambition to have as many cattle as possible, no matter how poor
they may be, there is little attempt to prevent the propagation of
inferior stock. Most bulls are castrated, only one or two being kept
in each herd for breeding purposes ; but they are not castrated early
enough, and as bulls always run freely with the herd, mating is indis-
criminate, although the widespread custom of lending out cattle
mitigates to some extent the dangers of constant inbreeding. 1

No fodder of any kind is grown for the cattle. They feed solely
upon the available pastures, except after harvest, when they graze
on the stalks in the fields ; but the pra&ice is widespread of burning
parts of the veld to provide early spring grazing. All land not aftually
under cultivation is common pasturage. " There are no clearly
defined areas to which particular individuals have prescriptive grazing
rights." 2 But among the Mpondo, and probably other Nguni tribes
also, people living in the same locality regard the area surrounding their
villages as their special grazing ground, and try to exclude others, even
to the extent of driving away their herds ; while among the Tswana
each ward similarly has its own special area, where no other cattle
may graze without permission.

In the rainy season cattle graze near to the villages, but during the
dry winter grass must often be sought many miles from home. On
tHe well-watered eastern side of South Africa the uplands form
summer grazing, while in winter the cattle are taken down to the deep,
sheltered valleys, where grass is lush and abundant. In the more
arid regions to the west, where as among the Tswana the people are
settled in large villages, the cattle are kept all the year round at special
grazing posts, often a day's journey or more from home. These
cattleposts are during the dry season located close to river beds, where
pits are dug to water the cattle ; in the rainy season they are moved
away to fresh pastures. No cows at all are kept at the villages. Milk is
sent there from time to time in the milksacks, but appears less fre-
quently in the daily diet than it does farther east.

By day the cattle are let out to graze in the vicinity of their kraal,
to which they return at sunset. In autumn, when there is plenty of
grass and water after the summer rains, they may be unattended. But
during the rest of the year they are always herded, especially during
the dry season, when they must be taken to water. At such times,
in the more arid regions, they are often watered only every two

1 Stayt, 1931, 38. Hunter, 1936, uy

140 I. SCHAPERA AND A. J. H. GOODWIN

or three days. The cows are in some tribes milked in the morning
before they go out and in the evening after they come back ; but in
others they are brought back in the forenoon to be milked for the
only time, and then taken out again to graze. The calf sucks first,
but is soon driven off; the milker, sitting on his haunches to one side
of the cow, then fills his wooden pail, after which the calf is allowed
to suck again. The calves are always separately herded near the kraal,
the very young ones being kept in it all day.

The Bantu know from practical experience where to find good
grazing and water for their herds, how to proteft them from wild
animals, and what remedies to apply for such common ailments as
liver disease, ophthalmia, and protrafted delivery. But, as Stayt
remarks, they really understand little about cattle diseases, "and
epidemics are quite beyond their control." l In these and such other
aspe&s of stock-raising where experience alone cannot afford. them
safe guidance, they fall back once again upon magic. They bury
" do&ored " sticks in the kraal to proteft the cattle from witchcraft
and disease ; they sprinkle their cows and inoculate their bulls with
certain " medicines " to stimulate fertility ; they burn other " medi-
cines " to promote the general well-being of the cattle ; and as an
additional safeguard forbid newly-widowed and other ritually impure
persons to approach the animals closely. Women generally are pro-
hibited from handling the cattle in any way, or even from walking
through a herd, particularly when menstruating, newly pregnant,
or in some other way " impure ". 2 But there do not appear to be any
tribal ceremonies relating to cattle similar to these noticed above in
regard to horticulture.

Goats and sheep are kraaled and herded separately from the cattle.
Sheep are not numerous, but even the poorest people have goats,
which among them fulfil the same uses as cattle. They provide milk
and meat ; their skins are made into clothing ; and they also serve as
mediums of exchange, lobola payments, and sacrifices. "The
distinguishing terms applied to cattle are applied also to goats, and
they are supposed to suffer from contaft with umla^a (ritual impurity of
women). But emotions are not aroused by goats as by cattle. No one
ever makes praises of goats." 3 Fowls, the only domesticated bird,
are very common. They are occasionally fed with grain, but otherwise

1 Stayt, 1931, 40.

1 Schapera, 19346, 561 ff., gives a detailed account of Kxatla herding rites.

1 Hunter, 1936, 71.

WORK AND WEALTH 141

left to fend for themselves. They generally roost on trees or on
specially-erefted perches in the household enclosure. They are often
killed for eating or sacrificed to the spirits, and are also used as mediums
of exchange. Dogs, the only other indigenous domestic animal,
are used principally in hunting. They are never eaten or sacrificed,
and are often half-starved and brutally treated ; but good hunting
dogs are much valued and sometimes well cared for.

^Hunting.

The meat obtained by the slaughter or death of domestic animals
is supplemented by the spoils of the chase. In Bechuanaland and other
areas where game is abundant this may be an important source of
food. But hunting is nowhere a regular occupation, except with a
few people. It is carried out fairly sporadically, as occasion and
opportunity permit, especially during the months when other food is
growing scarce. Game is therefore " merely a welcome addition
to the daily diet ; compared to the importance of vegetable produce
it cannot be described as in any way a staple food ", 1 The skins and
hair of the animals killed also provide material for clothing and certain
forms of decoration, their horns are made into receptacles and musical
instruments of various kinds, their teeth and claws are used as orna-
ments, and their fat mixed with the " medicines " for many forms of
magic.

The principal weapons employed in hunting, as in warfare, are the
spear, the axe, and the club. The two former have wooden shafts
and iron heads ; the latter is a short stick knobbed at one end. The
Venda alone make much use of the bow and arrow. The Southern
Tswana have borrowed this weapon slightly from the Bushmen;
but it is found nowhere else among the South African Bantu. 2 For
proteftion against the attacks of wild animals and other enemies,
all the tribes use shields of untanned hide. Among the Sotho the shape
varies considerably, from a wide hour-glass to a circle ; the Nguni,
and the tribes they have considerably affe&ed, prefer a plain oval.

People may hunt wherever they like within the tribal territory,
no private rights to hunting land being recognized. Hunting by
pursuit is generally carried on colleftively, and big communal hunts
are often organized by Chiefs and headmen. The men sometimes form
a huge circle round a spot where game is known to be abundant,

1 A. I. Richards, 1932, 106.

Cf. L. F. Maingard, " History and Distribution of the Bow and Arrow in South Africa/'
S. Afr /. Sci., xxix (1932), 711-723.

142 I. SCHAPERA AND A. J. H. GOODWIN

and close in gradually, shouting and sending their dogs ahead of
them to draw the animals, which are then speared when attempting
to break through. Sometimes again, as when hunting big game,
they patiently stalk an animal, or pursue it with their dogs, until it is
cornered. One or two men then rush in to hamstring it with their axes
so that it cannot move, when it is killed by spear. This method of
hunting is often exceedingly dangerous, especially against the elephant,
rhinoceros, buffalo, or eland; and men skilled in it are renowned
for their courage and dexterity. Single hunters similarly pursue smaller
buck with their dogs until the exhausted animal is cornered. Another
common method of hunting by large parties is to construft a long
fence with occasional openings at which pits are made. The game is
driven towards the fence, and in attempting to escape through the
opening falls into the pits. Such pits, their bottoms lined with upright
stakes, occasionally poisoned, upon which the animals fall and are
impaled, are also dug for big game along river banks or on the way
to waterholes. Smaller game and game birds are trapped in snares of
various kinds, the commonest being a running noose and slip-knot,
or are thrown down with the club.

As in agriculture and animal husbandry, various forms of magic
are employed to ensure success. Hunters of big game are ceremonially
purified before setting out and on their return ; they are inoculated
in the wrists with medicines to ensure steadiness of hand and aim ;
they carry various charms, and must observe certain taboos while out
in the veld. Women may never accompany a hunting party, nor should
any reference be made to matters of sex ; the animal pursued must
not be mentioned by its real name, a conventional substitute being
employed ; and careful attention is paid to omens, of which there are
many. Dogs and weapons, similarly, are washed with medicines or
doftored in some other way to make them more efficient. Without
these precautions,- the people maintain, the hunters will fail in their
quest, and may even be seriously injured.

^Clothing and Ornaments.

The old Bantu dress has been widely replaced, either completely
or partially, by clothing of European material and pattern. But it can
still be seen in some less influenced tribes and localities, where it is
made principally of skin. The type of clothing varies somewhat from
group to group, and within the group from person to person according
to sex, age, and rank. Generally speaking, however, women wear long

WORK AND WEALTH 143

aprons or skirts, hanging down before and behind from the waist.
A much smaller apron is sometimes worn underneath over the pudenda.
Nguni women further have a bodice across the breasts and under
the armpits, but in other groups the upper part of the body is left
uncovered. In cold weather cloaks of soft skin or fur, often carefully
made and decorated, are added to the ordinary dress. Babies are
carried on the back in cradleskins slung round the waist and knotted
over the breast. They go about completely naked except possibly
for a bead or string girdle and an occasional necklace or charm. Young
girls wear short fringes of skin, beadwork, or vegetable fibre to cover
the pudenda. Older girls wear similar fringes or aprons reaching to
the knee, and, among the Sotho, a longer apron at the back. Among the
Venda, however, they wear a loinskin brought between the legs and
tucked behind into the waist girdle to hang down in a flap. Girls
passing through the initiation ceremonies wear elaborate dresses of
vegetable fibre ; and in such tribes as the Zulu there are also distinftive
forms of dress for betrothed girls and pregnant women.

Men among the Cape Nguni and Shangana-Tonga wear nothing
but a penis sheath of calabash or palm-leaf covering the glans ; but
among the Natal and other Northern Nguni they also wear skin aprons
or tail and fringe sporrans both before and behind. Elderly men have
a hard ring of polished beeswax built into the hair of the head. The
Sotho and Venda wear a loinskin, one end of which is drawn between
the legs and tucked in or knotted at the back or side. They have
neither the penis sheath nor the headring. In cold weather the men don
skin cloaks like those of the women, but only members of the royal
family may use leopard or lion skins. The bigger boys wear the same
dress as the men, the smaller boys tiny skin flaps shielding the genitals.
Boys undergoing the initiation ceremonies are often completely
covered in dresses of vegetable fibre. Both men and women further
use flat sandals of oxhide, attached to the foot by a thong, and caps
of skin or fur. Among the Sotho conical hats of woven reed are often
seen.

The ornaments used by women consist mainly of beads and wire.
The wire, made in varying thicknesses of iron or copper, is generally
twisted into bracelets worn round the neck, arms, and ankles, often
in great numbers. Certain kinds of copper bracelet and ornament are
among the Venda and Sotho worn only by women of rank. Beads are
worked into headbands, necklaces, breast ornaments, waistbands, and
bracelets of varying size and colour. Except for ostrich eggshell

144 I- SCHAPERA AND A. J. H. GOODWIN

beads, locally made by some tribes or, as in Bechuanaland, bought from
Bushmen, beads have always been imported ; and no ceramic beads
of Native make are known, save for a few crude earthenware specimens,
unbaked and ill-glazed. Among the Venda, Northern Sotho, and some
other tribes certain kinds of beads are greatly prized and handed
down as heirlooms. Other ornaments are made of shell or bone, ivory
bracelets being in most tribes a prerogative of rank.

Men generally content themselves with wire bracelets worn round
the wrists and ankles ; but among the Nguni, young men, especially
when courting, often decorate themselves with beadwork. The Zulu
have developed a special " language of the beads ", different combina-
tions of colour and pattern having conventional meanings attached to
them. 1 Warriors, magicians and youths in ceremonial dress also wear
wild animal tails and strips of skin tied round the arms, loins, legs,
and ankles in various forms and sizes, necklaces of teeth, and head-
dresses of ostrich and other feathers. In some groups of tribes (e.g.
Northern Sotho and Cape Nguni), bead necklaces of a special type are
a symbol of Chieftainship.

^Dwellings and Household Goods.

Huts are used principally as bedrooms and stores, most a&ivities
taking place in the open, except in wet weather. They vary consider-
ably in stru&ure and shape. Two main varieties may, however, be
distinguished : the beehive hut of the Nguni, and the cone-and-
cylinder hut of the Sotho, Venda, and Shangana-Tonga. The former
has a framework of long saplings, planted in a circle in the ground,
and bent inwards and tied together to form a hemisphere. Shorter
saplings are tied horizontally across the frame at frequent intervals
to strengthen it. The whole is then completely covered with a strong
thatch of grass. The roof is supported from the inside by one or more
upright poles ; and a semicircular opening, generally so low that people
must crawl in to enter, is left as a door. The Cape Nguni plaster the
interior with mud and cowdung, but this does not appear farther
north. The Zulu, however, sometimes spread rough mats of grass
on the outside over the thatch.

The other type of hut has a cylindrical wall. This generally consists
of vertical poles bound together with wickerwork and plastered both
inside and out with mud and cowdung, although sometimes, as

1 Cf. F. Mayr, " Language of Colours among the Zulu expressed by their Beadwork Orna-
ments," Annals, Natal Museum (1907), i, 159-165.

PLATE XIV

[I. Schaleia

(b) Hcrdboys riding oxen (Kxatla)
PASTORAL LIFE

\fateti. 744

PLATE XV

Setting a snare for birds (Kxatb)

(b) Potter at work (Kxatla)
ARTS AND CRAFTS

L/.

WORK AND WEALTH 145

among the Tswana, it is made entirely of earth. An opening big enough
for people to walk through comfortably serves as a doorway. The
roof is of saplings bound together with wickerwork to form a conical
frame, which is then thatched. It is separately made, and when finished
is put on the wall, which it often overlaps considerably. Its point is
often, but not always, supported from the inside by a central pole ;
and sometimes its edge also rests on an outer circle of vertical poles,
between which and the wall a small circular veranda is thus created.
The floor in both types of hut is made of beaten earth smoothed over
at frequent intervals with coatings of mud and cowdung. A shallow
depression serves as a fire-hearth for cooking and warmth in wet
weather. There are no windows. Air gets in and smoke escapes
through the thatch and the door.

The huts belonging to one household are clearly marked off from
all others. Among the Nguni and Shangana-Tonga, where the house-
hold is also a separate local settlement, the huts are ranged in strift
order of seniority round the cattle-kraal. Sometimes, but not always,
the whole settlement is enclosed by a wooden palisade. Among the
Tswana, on the other hand, representing the extreme trend of Sotho
development, great numbers of households cluster together to form
large villages. The huts of each household are located irregularly
within a low, rectangular courtyard, formed by a fence of wickerwork
or reed mats or by an earthen wall or wooden palisade. The cattle-
kraal, owing to the peculiar conditions of local settlement, is always
separately built. The floor of the courtyard is made and smeared like
that of the huts ; and the walls of both are often ornamented with
broad geometric patterns of charcoal, kaolin, ochre, and other clays.

Huts and homesteads are fairly often renewed. The ravages of white
ants necessitate frequent rebuilding ; while in many tribes, especially
Nguni and Shangana-Tonga, huts are burned or pulled down after
the death of an occupant, and on the death of the household-head
the whole homestead is destroyed and a new one built elsewhere.
The site is often sele&ed beforehand by divination, and is always care-
fully " doftored " against witchcraft and other evils. Husband and
wife must have ritual sexual intercourse in a new hut before it can be
occupied. Among the Sotho the Chief must annually renew the
boundary charms protecting the village and its inhabitants.

Household utensils, all made from materials at hand, are fairly
numerous. Big-bellied, wide-mouthed pots for fetching and storing
water, for brewing and drinking beer, and for cooking, are made of

146 I. SCHAPERA AND A. J. H. GOODWIN

clay. The base of the pot is moulded in a sherd ; the sides are built
up with clay rings or strips, pressed to shape, and the whole smoothed
off. In the larger pots the lower half is moulded separately from the
mouth ; the two halves are then brought together, and the join made
good. The pots are finally placed in a shallow pit, surrounded with
piles of wood or cakes of cowdung, and fired. To proteft them from
spoiling in the process, various taboos must be observed, the most
widespread being that no ritually impure person may be present at the
firing. Both the shape of the pot and the choice of applied designs
vary somewhat from group to group. The Sotho prefer a single band
of recurring triangles, black on the red of the pot, painted on and
burned in. The Nguni prick up the wet clay with a pointed stick
to give the effeft of patterns worked in raised keloids. The Sotho
further make clay images of animals and human beings for use in
the initiation ceremonies ; while children everywhere make and play
with unbaked toy cattle of clay.

Headrests, bowls, platters, spoons, grain stampers and stamping-
blocks, milk pails, and porridge stirrers, as well as sticks, clubs, weapon
shafts, pipes, snuff-boxes, divining bowls (Venda), certain musical
instruments, poles for hut-building, and palisades for cattle kraals
and fences, are all made of wood. The Nguni peoples are on the whole
poorer in wooden utensils than the Sotho and Venda, some of whose
products show a high artistic finish and may be elaborately carved
and decorated. The common household utensils are laboriously
carved from seftions of tree trunks ; the rest are shaped from branches
according to need. An axe is employed for chopping, an adze for
trimming, a gouge for hollowing and scraping, and a knife or more
usually a spear for whittling. Boring and the application of design
are generally done with a hot wire.

Drinking vessels, ladles, snuff-boxes, resonators for some musical
instruments, and receptacles for ointments, beer, and milk are made
from calabashes, specially grown for the purpose. Basket containers
of various shapes and sizes, beer strainers, sleeping and eating mats,
trays, brooms, and string are made of reeds, grass, and other vegetable
fabrics. String is usually twined from the inner bark of such trees
as the Acacia, which is stripped, chewed and rolled on the thigh.
A bundle of reeds tied together makes a simple broom. Basketwork,
generally of the coiled or spiral variety, is adapted to many purposes.
The vast grain basket of the Sotho is of a composite grass strand,
sewn together with string. The small milkpail of the Zulu is made of a

WORK AND WEALTH 147

similar base, so tightly sewn together with rushes or grass into a solid
receptacle that liquid poured into it quickly expands the strands and
makes the basket water-tight. The conical baskets widely used for
carrying grain and other commodities are made with a light alternating
basketry stitch. Beer strainers, for separating the crudely-ground
grain from the thinner liquid, are roughly shaped conical baskets of
twilled flat reed, or a bias check or multiple plait of grass.

The horns of animals are used as holders for medicine, or converted
into musical instruments, pipes, and snuff-boxes; their bones and
tusks are made into ornaments of various kinds, awls for basketwork,
snuff-spoons, and divining tablets ; their teeth and claws are used for
necklaces and other ornaments ; and their hair is used as the base for
wirework or made into switches and ornamental headdresses. Their
skins are made into clothing, shields, bags, and bellows. Skins used
for clothing are pegged out on the ground, cleaned and allowed to dry ;
then carefully scraped with an iron blade, rubbed witlj some sort
of fat, beaten, and finally worked in the hands until quite soft and
pliable. They are then cut to the necessary shape, or sewn together
with animal sinew to make a cloak. They are not often tanned, although
use and the deliberate application of ochres and fats give the impres-
sion of a heavily-loaded leather. The bag carried by every man
consists of a whole game-skin worked to a soft leather and tanned or
dyed in a bark infusion.

The adzes, axes, gouges, and knives used in making wooden and
other utensils are of iron, as are hoes, spears, wire bracelets, and
many other objefts. Most tribes can mine in a simple fashion, but the
Cape Nguni and Shangana-Tonga are largely dependent upon others
for their supply of the metal, local deposits being rare. The Bantu
are unable to melt iron ore, but can obtain sufficient heat with their
double-bag bellows to reduce it to a " bloom " in an enclosed char-
coal furnace, either of termite-heap or built of clay. The bloom is then
wrought to shape with a stone hammer. All tools and weapons are
tanged, the tang being heated and forced into the wooden handle.
Copper is also worked in the Northern and Western Transvaal, where
it is found in fair quantities. The ore is melted in the furnace, cast
into rod-like ingots, and drawn through the graded holes of a simple
drawplate to make wire. This is bound round wildebeest or cattle
hair for bangles and other ornaments, or used as a lashing on spears
and clubs. Gold is worked in limited areas only, of which Mapungubwe
in the North-Western Transvaal may be taken as typical. Here, in

148 I. SCHAPERA AND A. J. H. GOODWIN

addition to wirework, gold plates and tacks are extensively used for
ornamenting obje&s.

CALENDAR OF WORK

The aftivities outlined above are not all carried on simultaneously
and continuously throughout the year. The Bantu, like most primitive
peoples, depend essentially upon the rotation of the seasons for their
food supply, and their aftivities are accordingly determined. This is
most clearly seen among the Tswana, whose life falls every year into
two clearly-distinguished phases. Their villages are fully inhabited
during the winter months only, when there is much bustle and activity.
As soon as the first rains fall, about November, the people go out to
their fields, remaining there right through the agricultural season
until the harvest is reaped in June. During this time village life is
almost at a standstill. In the other tribes, whose fields are within easy
distance of their homes, there is not the same sharp division ; but here
too agricultural work determines the time for most other occupations.

The periods of most intense agricultural aftivity are at planting
(late November, December, and early January in Bechuanaland),
weeding (January, February, and March), scaring away the birds (May),
reaping and threshing (June and early July). Most other occupations
are carried on at suitable moments in between. 1 The clearing of new
fields is generally done during the two months preceding planting.
Basketwork and woodwork are done mostly in the months between
harvest and planting, but also between weeding and reaping ; pot-
making is confined mainly to the warm, damp period immediately
before and during the early rains ; while hut-building is in Bechuana-
land done only after the harvest, but in other tribes between weeding
and harvesting as well. Cattle-herding, too, as we have seen, is affected
by the seasons, the cattle having separate winter and summer grazing ;
while hunting, especially of big game, is carried on most aftively
in the dry winter months, when the movements of the animals are
restricted by the scarcity of surface water.

Running parallel to this seasonal a&ivity is the food consumption.
The first fruits of the new year become available in the Matter half
of the rainy season. For the next three months or so the people revel
in the green mealies, pumpkins, melons, and sweet cane supplementing
the staple cereals, while milk is abundant. With the gathering-in of
the harvest, grain, and therefore beer is plentiful ; and as the people

1 Cf. Hunter, 1936, nof., for Mpondo seasonal calendar.

WORK AND WEALTH 149

now have more leisure, this is the great festive season. 1 Weddings,
initiations and other ceremonies of a similar nature are most often
celebrated ; tribal meetings, lawsuits, and other public business are
most aftively pursued ; and there is much informal visiting and
entertaining. The green foods, however, are more and more replaced
by the standard diet of boiled or roasted mealie grains and Kafir
corn porridge ; while the milk supply gradually diminishes as grazing
becomes poorer and the cattle are taken further away. The spring and
early summer months see the people reduced almost entirely to the
standard cereal foods and such wild plants as can be gathered after
the first rains have fallen. What Bryant says in this conneftion of the
Zulu applies equally well to most other tribes, as far as vegetable
produce is concerned. " Thriftlessness was, and is, one of the
characteristic defefts in the Zulu nature. He has inherited nothing of
the saving instinft. No sooner are the fruits of the new season mature
and permitted for general consumption than he forthwith initiates a
wholesale attack upon them. This habit so materially reduces the
amount left over for harvesting that, after a very few months, his
total store of food is at an end. In perhaps eight families out of ten
there is a normal annual recurrence of severe dearth throughout the
spring or early summer months. During the whole of this period,
members of all such families, children as well as adults, have to be
usually content with but one full meal a day, generally taken in the
evening time. Very often I have known whole districts of children
who got not even that." 2

DIVISION OF LABOUR

In many forms of production the only division of labour is between
the sexes. Everywhere among the Bantu different occupations are
traditionally assigned to men and to women ; and despite occasional
variations, the lines of division are fundamentally the same throughout.
In agriculture, the men clear the new fields of bush and grass, except
among the Shangana-Tonga, where this is women's work. 3 The
a&ual tilling of the soil, from the initial planting to the final reaping
and threshing, is very largely done by women and the older girls,
although younger children of both sexes assist in driving away birds,
and men occasionally take part in planting, weeding and reaping.
The women also look after the fowls. But all work connefted with
cattle, goats and sheep herding, milking, making thick milk, washing

1 Hunter, 1936, na. Bryant, 1908, 9. * Junod, 1927, ,

15 I. SCHAPERA AND A. J. H. GOODWIN

the milking utensils, and slaughtering is essentially within the
province of the men, women, as we have seen, being prohibited by
taboo from handling these animals. Adult men really do not often
herd or milk, unless poverty compels them to adopt this as a means
of livelihood. The younger boys go out with the small stock and
calves, while the older boys and youths look after the cattle ; but the
men closely supervise their a&ivities. Hunting is done by the men
and youths. Women gather edible wild plants, but men may also
gather wild fruits and berries when out in the veld.

Both sexes take part in hut-building, each having special tasks
to perform. Among the Nguni and Sotho the men cut wood and do
all timberwork, while the women cut grass, thatch, and make the
floor. Where necessary, women also plaster the wall of the hut or,
as among the Tswana, themselves make the earthen wall. Among
the Venda and Shangana-Tonga the men not only do all timberwork
but also thatch. The women merely make the floor and plaster the
wall. Cattle-kraals, grain pits, fences and wooden palisades are all
made by the men, but the earthen walls surrounding many Sotho
courtyards arc built by the women, who also do all the decorating.

Housework is almost entirely done by women and girls. They
stamp and grind corn, prepare food and make beer, wash the cooking
and eating utensils, smear the walls and floors of the huts and court-
yards, clean the huts and keep them in good repair. Men, however,
occasionally cook meat, particularly at feasts or when by themselves,
while boys at the cattleposts prepare and cook their own food. The
women also fetch water and colleft firewood, and do most other
carrying work, such as transporting grain home from the fields, or
bringing in the poles cut by the men for building.

The occupations just dealt with are in no way specialized. Every
man is expefted to be able to herd cattle, hunt, and do all the other
work normally performed by men ; and so too every woman is
expefted to be able to till the soil, cook, make beer, and do all the
other work normally done by women. The necessary knowledge is
acquired mainly through increasing participation in the work of the
household. Children are required and taught from a fairly early
age to be of assistance to their parents. Girls are made to help in fetching
water and firewood, stamping or grinding corn, cooking, and smearing
and cleaning the huts. They start by imitating these activities in their
play, and are gradually drawn into aftual domestic work under the
instruction of their mothers and older sisters. As they grow older,

WORK AND WEALTH 151

they begin to take part in agriculture. Young boys are put to herding
small stock and calves. Older boys herd and milk the cattle, learn to
handle weapons, hunt for themselves or accompany the men, help
build the cattle-kraals, and do other work of a similar nature. At the
initiation ceremonies marking the transition from childhood to man-
hood they are emphatically reminded that cattle-herding and warfare
are the two dominant spheres of masculine aftivity, while girls are
similarly exhorted in regard to agriculture and housework.

To a considerable extent, too, all such work is carried on by every
household for itself. Each has its own fields, cultivated by its women,
and generally also its own livestock, looked after by its men and boys ;
every man hunts, and every woman gathers edible plants. It is thus
able to produce dire&ly the food it consumes. Every household
further builds its own huts as required, and all the necessary cooking,
cleaning and other housework is done by its women and girls.

Household tasks are as often done colleftively as individually.
Even in such simple aftivities as fetching water, collefting firewood
and stamping corn, the women and girls prefer to work in company,
the presence of others affording a welcome relief from the monotony
or burden of the task. The boys herding cattle spend much of their
time playing together while their animals graze ; and hunting, as we
have seen, is undertaken for the most part by groups of men. The
major a&ivities, like clearing a new garden, planting, weeding, reaping
and threshing, building the framework or wall of a hut, thatching,
putting on the roof, fencing a field or cutting timber, are almost
invariably done by several people working in co-operation to save
time and energy. Each married woman, e.g. has her own fields, but
usually the fields of the same household are worked colleftively,
each being planted, weeded and reaped in turn by all the women
and girls. 1

Sometimes the members of a household are sufficient for the
purpose. More generally the assistance of outsiders is invited. Often
enough help is given by near relatives, who are indeed expefted to
assist. But in all the tribes there is also found the institution of the
work-party (Nguni, ilima ; Sotho, letsema). 2 Anybody with a big
task on hand with which he and his household alone cannot cope, or
which he wishes to complete reasonably soon, will invite his neigh-
bours and friends to help him. He brews a large quantity of beer,
or slaughters an animal, and makes it known that with this he will

1 Cf. Hunter, 1936, 87. Schapera, 1935 MS. ; Hunter, 1936, 88

15* I. SCHAPERA AND A. J. H. GOODWIN

entertain all those coming to work with him on a certain day. Any-
body wishing to do so can take part and receive his share in the feast.
Payment is sometimes also made in milk, porridge, salt or tobacco.
The work is lightened by rhythmical songs of various kinds and by
the presence of so many other people, while the feast awaiting them
is a stimulus to eager a&ivity. But it is not only this material reward
that makes people lend a hand. Poor people have no other method
of getting sufficient labour to carry out big tasks and, as they are
therefore all dependent upon one another for assistance, it is good
policy to help others and thus ensure their willing co-operation when
required for one's own work. And people notoriously slack in
attending work-parties often find their neighbours equally reluctant
to help them in turn.

It is only in regard to the making of household utensils and imple-
ments that some degree of specialization is found. Here there is also
a primary division of labour between the sexes, determined mainly
by the material employed. All work in wood, leather, bone and metals
is done by men ; potmaking and beadwork by women ; and basket-
work, except among the Shangana-Tonga, where it is restricted to
men, is carried on by both sexes, although each makes baskets of a
different type. But many of these crafts are known to only a small
number of people of the sex concerned. The outstanding examples of
such specialization are metalwork (men), and pottery (women),
the craft in both cases being largely confined to certain families,
within which it is handed down from parent to child. This is true
also of basketwork among the Shangana-Tonga, woodwork among
the Venda, kaross-making among the Tswana, and certain forms of
basketwork done by men among the Mpondo. 1 But woodwork,
leatherwork, beadwork and basketwork are on the whole more
generally diffused. Most if not all people are able to make at least
certain utensils, and thus supply their own needs. In each case, however,
a few people are also noted for the superiority of their produts,
since there is considerable variation in skill which here is of greater
significance than in food-producing or building ; or else make certain
objefts which others cannot. Magic, seen above to be an essential
part of produftive technique, is also very largely a highly-specialized
art. A few forms of magic connefted with agriculture, animal
husbandry and hunting are commonly known and praftised ; but
most of the rites and their associated medicines are known only to

1 Junod, 1927, ii, 125 f.; Stayt, 1931, H; Hunter, 1936, 99; Schapera, field data.

WORK AND WEALTH 153

the professional magicians, who may be either men or women. Among
the Shangana-Tonga certain men are also professional hunters of
big game, 1 but specialization of this kind appears to be on the whole
uncommon.

No specialist, however, ever devotes himself exclusively to one
particular occupation as a means of livelihood. They all herd and
hunt, cultivate their fields, build their huts, and carry out the ordinary
domestic work like the rest of the tribe ; the special craft they also
pradtise is merely a subsidiary source of income. This is as true of
the smith and the magician as of the potter or worker in wood. Nor
is there any special craft which cannot also be learned by outsiders.
Any person wishing and able to do so may become a magician or
smith, a kaross-maker, woodworker, or potter, although the first two
occupations require an apprenticeship for which a fee must be paid
to a praftising professional. Perhaps the only major instance where
any particular occupation is confined to a special caste is with the
Lemba, until quite recently the only coppersmiths and potters of the
Venda and neighbouring tribes. 2 Certain forms of magic, notably
rainmaking, are however often traditionally confined to members of
special clans, apart from the chief.

EXCHANGE

The general self-sufficiency of the household in regard to food,
shelter and many other produ&s, the faft that everybody is a herds-
man or cultivator, builder or housewife, and the relatively slight
development of specialization in other occupations, are refle&ed in
the marked absence of systematic trade. There is little produftion
for exchange, except with pots, baskets, iron goods and similar
utensils and implements ; and even here " there are no emporia where
one may regularly acquire these articles, no markets, periodical or
otherwise, at which they are offered for sale, no merchants going about
buying them up and selling them again ". 3

A certain amount of irregular trade is nevertheless carried on. A
man requiring metal goods, pots, baskets, wooden utensils, skin
cloaks or similar objefts which he does not himself make will procure
them from an expert craftsman. He goes direftty to the latter, and
either buys the objeft he wants, if it is already available, or, as is
frequently necessary, orders it to be made. Such articles may also

1 Junod, 1927, ii, 59 ff. Stayt, 1931, 18, 52, 58 f.

1 Lcstrade, 1934, 440-

154,1. SCHAPERA AND A. J. H. GOODWIN

be acquired by exchange from other people possessing more than they
need at the moment. Livestock, too, are fairly often obtained in this
way ; while in times of food shortage grain is sought from more
fortunate neighbours. There is no standardized medium of exchange.
Grain, meat, cattle, small stock, fowls, hoes, and spears are all
exchanged for one another and for other objefts, or used as payment
to magicians and herdsmen. There are, however, certain stabilized
relative values. Pots and baskets are almost universally exchanged
for their content in grain. Among the Tswana, again, two goats are
given for a sheep or a bag of corn, and ten goats or five bags of corn
for a heifer ; among the Mpondo ten spears are given for a young
beast and ten goats for a full-grown beast ; while among the Venda
two hoes are given for a goat. 1

Barter of this description, always on a small scale, takes place not
only within the tribe, but also between members of different tribes.
In Bechuanaland, e.g., people are in times of famine often compelled
to purchase corn in some neighbouring tribe less sorely afflifted.
The Shangana-Tonga, who do very little work in iron, obtain most of
their metal goods from the Venda; the Tswana buy skins from
Bushmen and Kxalaxadi for spears, knives, tobacco and dogs ; the
Mpondo get copper rings from the tribes farther east, in exchange for
corn ; and the Southern Sotho formerly traded " otter skins, panther
skins, ostrich feathers and the wings of cranes " to the Zulu, as
ornaments for warriors, in exchange for cattle, hoes, spearheads,
necklaces and copper rings. 2 Permission to trade must in such cases
be obtained from the Chief, who usually expefts and receives a present
in return.

Commodities are not only used in this vay to acquire goods which
people cannot produce for themselves. They may also be given in
exchange for labour. We have already seen that people helping in
work parties are paid with beer, meat or other foodstuffs. So, too,
the magician is paid for his special services, generally in livestock,
sometimes in grain or metal goods. It is fairly common also for a
man with no sons or young male relatives to look after his cattle to
employ some other man or boy to do so. The herdsman is paid a
heifer, which with its offspring then belongs to him. This form
of wage-labour must be distinguished from the widespread custom
of uhusisa (Nguni) or xo fisa (Sotho), by which the wealthy man

1 Stayt, 1931, 75; Schapera, 1935 MS.; Hunter, 1936,134.

1 Casalis, iStfi, 169 (S. Sotho); Junod, 1927, ii, 138 (Shangana-Tonga); Schapera
1935 MS. (Tswana); Hunter, 1936, 134 (Mpondo).

WORK AND WEALTH 155

places one or more of his cattle in the keeping of another. The herds-
man is entitled to use their milk for his own purposes, and may be
given some of the meat when an animal dies, but he may not sell or
slaughter them. It is also usual, if the cattle flourish under his care,
to reward him with a heifer from time to time.

Serving another man like this is perhaps the principal way in which
a poor man can acquire cattle. " Cattle raiding afforded an opportunity
for enterprising men to get rich quickly, but the principal way in
which a poor man acquired wealth was to serve a chief, or wealthy
man, and receive in return the loan or gift of cattle ... A fundamental
idea of (Bantu) social economics is that it is in no way degrading to
ask a gift of another, that to dispense gifts is the mark of a chief,
and that he who is given gifts becomes the giver's ' man '. Every
chief and every wealthy man has . . . men come to ask gifts, and
prepared in return to perform services. The gift may be of stock or of
produce ; most frequently it is a loan of cattle. . . . The services usually
performed are building huts, cutting bush for the kraal, assisting m
cultivation, going messages, a&ing as mouthpiece of his ' chief '
in public (e.g. arranging a marriage agreement for him), and praising
(ukubonga} his ' chief '." l And the man lending out his .cattle not only
has the task of herding them simplified. It serves partly to insure
against total loss from disease, witchcraft or some other agency which
might annihilate them should they all be concentrated in one kraal ;
and is also a means of disguising the full extent of his riches, and so
of escaping the jealousy and evil designs of less fortunate neighbours.
Above all, perhaps, it is a means of acquiring prestige : the greater
the number of retainers thus attached to a man, the higher his status
and the more considerable his influence in the tribe generally.

Barter and payment for special services are only two of the
mechanisms by which goods are circulated and thus made more widely
accessible. Lobola payments, the fines and compensations levied at
the courts, the various gifts made by relatives to one another, and the
various forms of tribute paid to the Chief, all serve the same end.
They will be dealt with below in discussing the special obligations
arising from kinship and citizenship. But it may also be noted that a
good deal of miscellaneous borrowing goes on in regard to such
objefts as household utensils, food and even livestock; and is a
frequent means of making good any immediate deficiency in a
particular commodity.

1 Hunter, 1936, 135.

IS 6 I. SCHAPERA AND A. J. H. GOODWIN

PROPERTY AND INHERITANCE

The food, livestock, dwellings, clothing, household utensils and
other objects acquired by direct production, by barter, or in return
for serving others, are used by each household to satisfy its own
needs. But the rights exercised over these goods by members of a
household are defined in various ways by tribal law and custom.
Possession of goods entails certain obligations, not only within the
household itself, but towards other people also. These rights and
duties vary according to the nature of the commodity.

Land.

The livelihood of the Bantu is intimately bound up with their
system of land tenure. They erect their dwellings on the land, cultivate
it, graze their livestock upon it, and hunt over its surface. They use
its water for domestic purposes and for their herds .and flocks ; they
eat the wild fruits and other foods it produces, and make medicines
from its vegetation ; they convert its wood into huts, palisades and
various utensils, and its reeds and grass into basket-work, thatch
and string ; and they extract from it metals, clay for their pots, and
earth for the floors and walls of their huts. In the regulation of the
land and its resources provision is made for each method of
exploitation.

All the land occupied by the tribe is vested in the Chief and
administered by him as head of the tribe. This he does through his
sub-Chiefs and headmen, who regulate the distribution and use of
land in their respective areas. The land is not his personal possession
with which he can deal as he pleases. " With the exception of that
portion of land which is reserved for the Chief and his family, on
more or less the same basis as for anyone else, none of the land belongs
to the Chief, nor can he dispose of it except gratuitously and to
members of his own tribe." l The natural resources of the land
earth, water, wood, grass, clay, edible plants and fruits are all
common property, never reserved for the use of any particular
persons. 2 So, too, anybody may graze his cattle and hunt wherever
he pleases ; although, as we have seen, people living in the same
area often tend to assert exclusive claims over its grazing, while in
Basutoland the Chiefs reserve special areas for winter grazing. It

1 Lestrade, 1934, 430.

1 Little information is available regarding rights to mineral deposits. Among the Tswana
it is said that their exploitation was striUy controlled by the Chief; but no further detail;
could be obtained, as Native mining has long since disappeared.

WORK AND WEALTH 157

is only in regard to land for residence and cultivation that private
rights are universally recognized. Every household-head has an
exclusive right to land for building his home and for cultivation.
Generally he can take up such land for himself within the area con-
trolled by his sub-Chief or headman, provided that he does not
encroach upon land already occupied or cultivated by others. Failing
this, it is the duty of his headman to provide him gratuitously with
as much land as he needs.

Once a man has taken up or been granted such land, it remains in
his possession as long as he lives there. He has a prescriptive right
over his arable land, whether it is still uncleared, being cultivated,
or lying fallow. Other people can graze their cattle on the stalks
once his crops have been reaped, and any woman can gather firewood
or wild plants from it ; but no one else can cultivate it without his
permission, and on his death it is normally inherited by his children.
He also has the right, subjeft to the approval of his headman, to give
away part of it to a relative or friend, or to lend it to someone else.
But he can never sell it or dispose of it in any other way in return for
material consideration. Should he finally abandon the spot, his land
reverts to the tribe as a whole and can subsequently be assigned to
someone else. The only other way in which he can lose his right to
the land is by confiscation, if he is found guilty of some serious crime.

The household-head in turn must divide his land among his
dependants according to their needs. Each of his wives is entitled
to at least one field for her own special use. No other wife can cultivate
this field, nor can her husband alienate it from her house without
her permission, and on her death her own children have a preferential
right to it. The household-head must similarly find arable land for
the wives of other married men living under his immediate jurisdiction.
Often enough he also has a common household field, not assigned to
any special wife, but worked by them all for the household as a whole. 1

Livestock.

Cattle and other domestic animals are all private property ; and
have their ears slit, lopped or perforated in different ways as dis-
tinftive marks of ownership. The Chief has no say in their distribution,
save in regard to cattle looted in war ; nor has any man the right to
demand cattle from him or other political authorities by virtue merely

1 On Bantu land tenure, cf. especially Schapera, 1935 MS.; Harries, 1929, chap, in;
Hunter, 1936, 112 ff.

IS 8 I. SCHAPERA AND A. J. H. GOODWIN

of tribal membership, although, as we have just seen, he has such a
claim to land. Moreover, cattle once acquired can not only be given
away outright, but can also be sold, slaughtered or otherwise disposed
of by the owner. 1 A man may, however, be made to surrender one
or more of his cattle as damages for a wrong he has done, or ate punish-
ment for an offence he has committed. In the former instance, the
cattle will belong to the victim of the wrong ; in the latter, they go
to the Chief or other political authority. In exceptional instances the
Chief may even confiscate all the cattle of a man who has committed
some very grave offence. The Chief, too, is entitled on certain
occasions to receive tribute in cattle from his subje&s, and a man's
relatives also have certain claims upon his cattle. These special claims
will be dealt with more fully below.

Men, women and children may all possess cattle in their own right.
A man may acquire cattle through natural increase of cattle already
owned, through inheritance, as lobola for his daughter or sister, by
gift from a relative or friend or the Chief, by service, by barter, and
through some other more or less fortuitous circumstance, e.g. as
damages. A married man must set aside some of his cattle for the
house of each wife. The cattle thus allocated, together with some of
those received as lobola for the daughters of any house, belong to
that particular house. No house has any claim to cattle belonging to
another, except by way of loan ; the husband cannot use such cattle
except for the benefit of that house, nor can he alienate them without
the consent of its wife and eldest son. These cattle provide food for
the inmates of that house ; they and their increase are used to lobola
wives for its sons ; and when the man dies they pass to its heirs.
The cattle a man does not specifically allocate in this way he may do
with as he pleases. He may lend them out to other people to herd
as mafisa y or he may sell or slaughter them. He may and should consult
the other members of his family before thus disposing of them, but
is not bound to follow their advice. If, however, he squanders them
recklessly, his heirs may appeal to the Chief, who will then take him
to task. In some tribes (e.g. Tswana), a man should in any case during
his lifetime set aside one or more heifers or other livestock for each
of his sons if he can afford to do so. He must also both among the
Sotho and the Nguni give each daughter one or more beasts at
marriage. He should further, out of such cattle, make the necessary
sacrifices to his ancestors on behalf of his dependants.

1 Lestrad* 1934, 4 33-

WORK AND WEALTH 159

Cattle belonging to an unmarried man are controlled by his father
or guardian. A son has no right to give away, sell, lend or otherwise
dispose of them without the consent of his father. The latter, on
the other hand, may use the cattle as he thinks fit on behalf of his son.
But should he dispose of them for his own ends he is expefted to
replace them. Once the son is married, and sets up his own household,
he can dispose of his cattle as he pleases, subject to the obligations
noted above.

A woman may acquire cattle in her own right by gift from her
father or husband, as her share of the lobola given for her daughter,
and as payment for specialized work (e.g. magic). These cattle run
with her husband's but he cannot treat them as his own. They also
do not form part of his estate, being separately inherited by the
woman's own children. On the other hand, she normally cannot
dispose of her cattle without her husband's consent and through
his agency, although if he objects unreasonably she can appeal to her
own relatives or to the headman. A divorced wife takes her own cattle
back to her parental home. Her husband, however, is generally
entitled to keep back any that he has himself given her, while if she
has any children the cattle must remain behind to be inherited by them.

Produce.

The milk yielded by cattle belongs to the house to which they are
assigned, and is reserved for the use of its inmates. Similarly, every
woman normally has her own granary, in which she keeps all the corn
reaped from her field, except for the small tribute which must in many
tribes be paid to the Chief. From her corn she prepares daily the food
for her children and other dependants, as well as for her husband.
She must also send to her co-wives part of whatever dish she has
prepared, so that in this way the produce of the whole household is
shared by all its members. The husband does not generally interfere
with his wife's use of her corn ; but she cannot sell it, give it away, or
grind it in large quantities to make beer without his consent. The corn
from any common household field is direftly controlled by the
husband. It is used mainly for brewing beer for work-parties or feasts, for
purchasing cattle and other commodities, and for entertaining visitors.

Meat obtained from the slaughter of any domestic animal is shared
by all members of the household. Special portions are allotted to
the wives, older sons and older daughters. Some it it must also be
given to such near relatives as paternal and maternal uncles and aunts ;

160 I. SCHAPERA AND A. J. H. GOODWIN

and in every tribe well-defined rules lay down which particular
portions must be given to each category. Game is generally divided
among all the people taking part in a hunt. The first man to strike
an animal claims the body and the skin. Others wounding it are
entitled to the legs or other portions, according to the order in which
they helped kill it l ; and some of the flesh is shared with the rest of
the party. The Chief also is entitled to hunting tribute in various
forms, the nature of which will be dealt with below.

Huts, Household Utensils, and Personal Belongings.

Huts, although built by the whole household, often with the
assistance of outsiders, are generally referred to by the names of the
wives occupying them. Each married woman is entitled to her own
hut. . The allocation of huts is controlled by the husband ; but once
a wife has been provided with her hut, she is regarded as its mistress,
and supervises everything connected with it. But her husband has
the right to enter it whenever he pleases 4 and if she is divorced, it
remains behind with him.

Household utensils are for the most part controlled by the wife.
She generally brings with her at marriage such things as pots, baskets,
hoes, and sleeping-mats, the gift of her parents. Others, she makes
for herself, or purchases with corn, fowls or beer, or is given by her
husband. But among the Mpondo it is the duty of her parents to
replace her utensils when they wear out or break. 2 She may do as she
likes with these objefts, without consulting her husband; but if
he finds her dissipating them recklessly, he may intervene to stop
her. If divorced, she takes back all that she has brought from her
own home. The rest may also be awarded her, or divided between
her and her husband, as the court decides.

Personal effefts are the private property of the people for whom
they are made or acquired. In the case of women and children they
consist mainly in clothing and ornaments. Men also have sticks and
weapons. Skin clothes are made by each man for himself, his wife and
children, except among such tribes as the Mpondo, where a woman's
clothes, even after marriage, are provided by her father or brother. 3
In all tribes, too, certain forms of clothing, like karosses, must often
be bought from specialists. Ornaments are made by women for them-
selves, their lovers, husbands and children, but are as often purchased.

1 Cf. Junod, 1927, ii, 56 ; Krige, 1936, 206.

Hunter, 1936, 128. Hunter, 1936, 128.

3
J3
N

WORK AND WEALTH l6l

Men generally cut their own sticks, but purchase or inherit their
weapons. The proceeds obtained by people from the sale of the
obje&s they make are also regarded as their exclusive property.
Adults may do as they please with their personal belongings, but
children may not dispose of them without the consent of their parents.

Family Obligations.

We may now conveniently review the principal obligations between
members of the same family in regard to property and services. As
already shown, most produ&ive aftivities are carried on within the
household. Both husband and wife must perform on behalf of the
household the various tasks imposed by tribal life. It is the special
duty of the wife towards her husband and children to till her fields,
cook food, and carry out her other domestic tasks. It is the special
duty of the husband towards his wife and children to provide them
with huts in which to stay, land to cultivate, and cattle to milk. He
must further see that they are properly clothed and have the other
necessities of life. The property each acquires must be administered
for the benefit of the others as well. A father must help his sons with
cattle to lobola wives ; must see that his daughters receive domestic
utensils to take to their new home ; and must allocate cattle to his
wives, so that on his death there is provision for the sons of each
house. Children, again, must help in all the work of the household
according to their age and sex. Sons when grown up must support their
parents, and should place at the disposal of their fathers anything
they obtain by hunting, purchase, or service for others.

Brothers and sisters must similarly help one another. Most tribes
among the Sotho and Nguni have the custom of " linking " a sister
to a brother. 1 The latter is then responsible all his life for the special
needs of this sister. He must provide her with clothes when necessary,
give her meat whenever he slaughters, and, as among the Mpondo,
supply her " with some of the stock necessary for initiation rites
before her marriage, as well as with a wedding outfit, and with clothing
and utensils after marriage ". 2 In return he receives the lobola cattle
given for her, while after his own marriage she must come to assist
his wife in any important domestic work. Brothers are ampng the
Tswana, and probably elsewhere, linked together in the same way.
An elder brother must support and proteft his younger " linked "

1 Cf. especially Schapera, 1935 MS.; Hunter, 1936, 123 f.
1 Hunter, 1936, 129.

162 I. SCHAPERA AND A. J. H. GOODWIN

brother, help him with lobola cattle, and should give him other cattle
from time to time. The younger brother, in turn, must always help
his senior in work, and, if their father is dead, should place at his
disposal whatever he earns. Each must give the other meat whenever
he slaughters.

Inheritance.

Inheritance among the Bantu is governed by various traditional
rules coming into force at the death of a person. A man may before
his death indicate what he wishes done with some of his personal
belongings ; but his other property must be inherited in the customary
manner and cannot be dealt with otherwise. The principal heir is
generally the eldest son of the great wife. But the father, given good
cause, may disinherit him by publicly disowning him before the Chief.
The heir will then be the eldest remaining son.

When a married man dies, leaving a wife and children of both sexes,
his eldest son, as general heir, inherits all the livestock which at the
death of his father have not been allotted or donated to any special
house or person. He also inherits any livestock specifically assigned
to his mother's house. He must out of these provide for his younger
brothers, in particular by helping them to lobola wives for themselves.
Daughters normally receive no cattle at all, nor does the widow;
but the heir must maintain and support them while they are living with
him. Should he fail to do so, they can complain to the Chief. The
estate of a polygamist is similarly divided. The eldest son in each house
inherits all livestock assigned to that house. The eldest son of the
great house further inherits such property as has not been assigned
to any house ; but must see that each house has sufficient stock to
provide wives for its sons. The fields, the huts, and the household
utensils remain with the widow or widows, who use them as long as
alive. But in many tribes, especially Nguni and Shangana-Tonga,
the homestead is destroyed on the death of the household-head.
The heir of each house then builds his own home, where he is joined
by his mother and her other children. These tribes also burn or bury
most of a dead man's personal belongings. Any weapons and other
objefts that may be preserved go to the principal heir, who must
share them with his younger brothers and with certain other relatives.
Deceased's maternal uncle, e.g., is among the Sotho entitled to some
of his weapons, as well as to a bull from his herd.
If a married man dies leaving no sons, his estate comes under the

WORK AND WEALTH * 6 3

control of his nearest male relative, generally his younger brother.
This man must maintain and support the widow and any daughters
from the property in his charge. He is expefted by cohabiting with
the widow, or by arranging for this to be done by someone else, to
see that she duly bears a son, who will ultimately inherit the dead
man's property. In a polygamous household, if there is no son by
the great wife, the general heir will be the eldest son of her affiliated
wife, and failing one the eldest son of the house next in rank. Failing
a son in any house, the nearest male relative of the dead man becomes
the general heir. But here again he is expefted to " raise seed ", either
in person or by proxy, to the various houses in which sons are desired.

If no heir is found by any such praftice, a man's property is inherited
by his father, if still living ; if not, by the eldest brother in the same
house. Failing a brother in the same house, the heir is the eldest
brother in the great house. Failing a brother there, the eldest brother
of the house next in rank inherits. In the entire failure of brothers,
the heir is the father's eldest brother in the same house, or his senior
male descendant. Failing one, the estate passes to the father's eldest
brother in the great house, or his senior male descendant ; and so on,
descending from house to house according to their rank, until a
male heir has been found to inherit the property and look after the
widow and daughters. In the entire absence of a male heir, the estate
falls to the Chief. 1

When a married woman dies childless or without sons, her husband,
if still living, may bring in another woman, generally her younger
sister or some other near relative, to " raise seed " for her. This woman
will then use the cattle, fields and other property assigned to deceased's
house, and they will be inherited in due course by her own children.
But if the dead woman leaves any children, all property belonging
to her in her own right, including livestock, are among the Mpondo
inherited by her youngest son, who also inherits her fields. 2 Her hut
and certain personal belongings are however destroyed or buried
with her. Among the Tswana, on the other hand, the youngest son
inherits the huts as well as the fields and livestock ; but the corn and
household utensils go to the eldest daughter, who must out of them
provide for her younger sisters and for the maternal relatives of her
mother. 3 Accurate information regarding other tribes is not available.

The estate of an unmarried man is administered by his father, if

For general discussion of inheritance, cf. WhitfieW, 1929, chap. yi.
Hunter, 1936, 119 f. ' Schapera, 1935 MS.

164 I. SCHAPERA AND A. J. H. GOODWIN

still alive. After the latter's death it goes to the heir of the house to
which the deceased son belonged. The property of an unmarried
woman goes to her mother, passing after the latter's death to the
other daughters. Her cattle, however, are inherited in the same way
as those of an unmarried son. But if she, although unmarried, has
children of her own, they inherit all her property.

SPECIAL GIFTS AND SERVICES

It has been noted above that a man's property and labour are not
only utilized to satisfy his own wants and those of his household, but
must on occasion be placed at the disposal of others. Among obliga-
tions of this kind hospitality plays a large part. Every visitor should
be offered some food, great stress being laid upon the observance of
this rule. Moeng xoroxa dijo di a tonale, say the Tswana : " stranger
arrive, so that there may be food." At feasts there are always a number
of uninvited guests, but certain pots of beer and portions of meat are
invariably set aside for them. And whenever a man slaughters he
should reserve some of the meat for the neighbours attracted to his
home. A selfish or niggardly man is despised ; a generous man is
warmly praised, and is generally a welcome guest in turn when he
visits others. We have also seen that people are often asked to assist
their neighbours at work. No one is compelled to do so ; but the
tradition of reciprocal help, without which many common tasks
could not be satisfactorily accomplished, is generally a sufficient
incentive, quite apart from the feast rewarding those taking part.

Obligations towards Relatives.

There are, however, in all the Bantu tribes more special obligations.
Relatives by blood or by marriage are entitled of right not only to
hospitality, but also to gifts of various kinds and to labour assistance.
They should visit one another frequently, and must be specially enter-
tained whenever they do so. They must be invited to all feasts, where
they frequently sit in small groups according to their status, each being
given special portions of meat, beer, and whatever else is available.
And, as we have seen, whenever a man slaughters he must send special
portions of the meat to the relatives living near him. Those out of
reach should from time to time be given an ox or goat instead. They
are expe&ed to reciprocate in kind on some future occasion. A man
given meat, e.g., will somewhat later repay it with several pots of
beer, or with a substantial bowl of porridge. So, too, whenever a

WORK AND WEALTH 165

man has any big task on hand, he may instead of organizing a work-
party call upon his relatives for assistance ; and they are expefted to
provide it. When a woman is confined after childbirth, her own female
relatives as well as those of her husband help to fetch water, colled
firewood, cook and do all the other household work to which she
cannot attend. They help her in the same way to prepare for a feast.

A few more specific instances of kinship obligations may briefly
be reviewed. 1 A man's paternal relatives are preferential partners in most
co-operative aftivities. His senior uncles and aunts should contribute
towards the lobola he gives for his wife, having in return the right
to receive part of the lobola obtained for his sister or daughter. They
must also contribute towards the gifts of cattle, food, clothing and
ornaments made to boys and girls on passing through the initiation
ceremonies. If a man is unable to pay a fine imposed upon him at
court, they should assist him to do so. If childless, or having children
of one sex only, they may be given a nephew to help herd their cattle,
or a niece to help in the household work ; and they may even com-
pletely adopt such children.

Similar obligations are created by marriage. The lobola transaction
itself, with all the other accompanying gifts, is one of the principal
means of circulating goods. Moreover, once a marriage has been
established relatives-in-law must help one another at work in the same
way as relatives by blood. A married woman assists and is assisted
by her husband's female relatives in the various major domestic
and agricultural tasks, they share their food, and generally fun&ion
as interdependent units of one social group. Her own relatives make
her and her family gifts of food, clothing, household' utensils and
similar objefts, help them at work, and frequently adopt their children
for a while. Her husband is expefted to reciprocate in kind. He should
help his wife's people at work, not only when requested but also of his
own accord, make them occasional gifts of meat and other kinds of
food, and invite them to all feasts in which he plays a leading part.

A child for whose mother lobola has been paid is always treated with
great respeft and has many privileges at her parental home. His
maternal uncle has many special duties to perform towards him,
especially among the Sotho and Shangana-Tonga. The uncle must
slaughter for the feast at the birth of a nephew or niece, provide them
with new outfits of clothing when they are initiated, help them with
food and clothing when they get married, and contribute towards
1 Cf. especially Schapera, 1935 MS.; Hunter, 1936, 47 ff., 121 fF.

166 I. SCHAPERA AND A. J. H. GOODWIN

the lobola paid by his nephew. The nephew has the right, among the
Tswana, to slaughter a beast at his uncle's kraal for food, or to take
a heifer for breeding purposes ; or he may give his uncle an old
bull, a damaged kaross,or some other useless objeft, claiming in return
a young bull or heifer, or a new kaross. The uncle, in return, always
receives a substantial portion of the lotola given for his niece ; he
receives the first animal earned by his nephew, part of every animal
he slaughters or kills in the chase, and any objeft he picks up in the
veld ; for every animal his nephew kills or takes from his kraal he
can later exercise a similar privilege; and he inherits part of his
nephew's estate. 1 Cross-cousins among the Tswana similarly can
help themselves freely to one another's personal belongings, the only
relatives to whom this privilege is accorded.

These forms of service and gift to relatives are not compulsory,
in the sense that people can be sued at court for failing to provide them.
But it will be said that they do not show their relatives due honour
and respeft, and considerable ill-feeling may result, together with
refusal on the part of the slighted relative to work with or take any
interest in the welfare of the culprit. The main incentive to conformity
with these obligations is reciprocity. In the relative absence of
industrial specialization and consequent economic interdependence,
kinship serves to establish greater social cohesion within the com-
munity, and to integrate its activities into a wider co-operation than
obtains within the restricted limits of the household. The so-called
" communal system " of the Bantu is largely the manifestation of this
close bond of solidarity and reciprocity arising out of kinship and
affe&ing well-nigh every aspeft of daily life.

The Position of the Chief.

Apart from being obliged to help his relatives in this way, a man
must pay tribute, both in labour and in kind, to his Chief and other
political superiors. 2 But whereas relatives cannot in court enforce
fulfilment of their special rights and privileges, the Chief has the
power to punish anybody failing to render him the customary tribute.
The nature of this tribute varies somewhat from group to group.
Generally speaking, however, the Chief has the right to send any
member of the tribe where and on what errand he likes. Moreover,
he can call upon them collectively to perform various tasks for him

Junod, 1927, i, 267 ff. ; Schapera, 1935 MS.
* Eiselen, 1932,1, 297-306; Schapera, 1935 MS.; Hunter, 1936, 384 ff.

WORK AND WEALTH 167

and his household, as well as on behalf of the tribe at large. Among
the Sotho and Northern Nguni work of this kind is generally done for
him by the regiments, whom he summons for the purpose whenever
he wishes. Among the Tswana, e.g., the men's regiments not only
constitute the tribal army, but in time of peace are often called upon
to round up stray cattle, to destroy beasts of prey, to hunt for the
Chief, to clear new fields for his wives, or to build his huts and cattle-
kraals ; while the women's regiments help build and thatch the huts,
fetch earth and smear their walls and floors whenever required, fetch
water and colleft firewood for the Chief's wives, and cultivate their
fields. 1 In other tribes similar work is generally done locally, through
the sub-Chiefs and headmen.

Although the Chief thus has a right to the services of every member
of the tribe, he also has many servants direftly attached to his house-
hold and performing its work. Among the Nguni and in many other
tribes young men come from all parts of the tribe to serve at the
Chief's court, where they perform such menial tasks as fetching wood
and water, making fires and herding cattle. Among the Tswana he
has also a large number of serfs, recruited from inferior communities
like the MaSarwa (Bushmen) and the Kxalaxadi. Many of them live
in the veld, hunting for the Chief; others stay at his cattleposts
herding his cattle; others are brought into the town, where they
live in his homestead and do all the routine domestic work. Most of
them are attached to him in person, and are taken over by his successor.
Others are allotted to his different houses, and are then inherited in
the house to which they are attached. They are not free to leave or to
transfer their allegiance to some other master, and if they desert can
be followed up and brought back. Their spouses, if they marry outside
the Chief's household, also become his servants. He can, on the other
hand, give or lend them to someone else, whose servants they then
become ; and it is usual that when his daughter marries he gives her
some of these serfs to help her in her new home. They are regarded
now as her sole property, and will be inherited by her children. Other
servants are drawn from the ordinary members of the tribe. Each
house of the Chief is put under the care of a certain ward or certain
families. These retainers remain attached to the same house, their
children continuing to work for it and its descendants. They live at
their own homes, and have the same rights as other members of the
tribe ; but they also have special obligations towards the house to

1 Schapera, 1935 MS.

168 I. SCHAPERA AND A. J. H. GOODWIN

which they are allotted. Some of the men are put in charge of its
cattle, while some of the women and girls help in the daily domestic
work ; and the group as a whole cultivates the fields of that house,
builds its huts, and performs all similar work. 1

The Chief is also entitled to various forms of tribute in kind. There
are in most tribes one or more large public fields specially cultivated
for him every year by local divisions of the tribe. He provides the
seed ; but the people do all the work, from clearing the field to
harvesting and threshing the produce, all of which is then given to
him. Among the Sotho and Shangana-Tonga he further receives a
basketful of corn from every woman .after she has reaped her own
harvest. The presentation of this corn is an occasion of great ceremony,
constituting part of the harvest thanksgiving. The Chief must also
be given the breast-portion of every big game animal killed in the
chase, ostrich feathers, the skins and claws of all carnivora, especially
of the lion and the leopard, and the hides and tusks of the elephant,
hippopotamus and rhinoceros. The skins of all animals killed in any
communal hunt organized by him are also his property.

But his most important source of wealth is cattle. He possesses
by far the largest herds in the tribe. All cattle looted in war are brought
to him. Some of them he distributes among the successful raiders
or among men who have otherwise distinguished themselves ; but
he always keeps a generous portion for himself. Stray cattle must be
brought to him by their finders ; and if not claimed within a certain
period, are held to belong to him. He receives gifts of cattle from the
head of every family on his installation as Chief; and among the
Tswana similar gifts are made to him by the leading members of the
tribe after his initiation. The lobola paid for his great wife is frequently
made up of contributions from the tribe ; while the lobola he receives
for his daughters forms a substantial addition to his wealth, since the
amount paid for a Chief's daughter is much higher than that for a
commoner's daughter. Among the Sotho, again, the father of every
boy going through the initiation ceremonies must pay the Chief
a fee of one beast. He is entitled to all the fines imposed in his court for
homicide and serious bodily assault, and to a portion of the damages
paid in civil cases brought to his court for settlement. So, too, when-
ever an important tribesman dies, one of his cattle must be sent to
the Chief as formal notification of the loss. He can even confiscate the
entire property of tribesmen found guilty of witchcraft or some other

1 Schapera, 1935 MS.

WORK AND WEALTH 169

grave crime. Finally, people anxious to keep in his favour find it
expedient to make him occasional gifts of cattle ; and he sometimes
makes special tours of his territory to colleft such gifts.

But all this accumulation of wealth by the Chief, all this labour
rendered to him, must be utilized by him not only for his own benefit,
not only for the maintenance of his large household, but also on behalf
of the tribe as a whole. One quality always required of a Chief is
generosity. He must provide liberal hospitality in beer and meat
for people visiting him or assisting at his court, or summoned to
work or to fight for him. He rewards with gifts of cattle and some-
times wives the services of his councillors, warriors, and retainers.
Many of his cattle are distributed for herding among his retainers,
who live upon the milk ; while in times of famine he must provide
the people with corn from his granaries. The annual tribute of corn
he receives he must use to make the beer given to the people at the
harvest thanksgiving ; every man bringing him the skin of a lion or
leopard must be rewarded with the gift of a heifer ; the meat of the
game killed at a tribal hunt he distributes among those taking part ;
the breast-portions of big game animals he cooks for the men at his
court, or sends to paupers or invalids. The Chief is thus looked upon
as the source of wealth, of reward, and of sustenance in times of
trouble ; and this constitutes a powerful sanction for his authority.
A Chief too niggardly to provide lavishly for his subjefts readily
becomes unpopular, and may even find them tending to desert him
for some more liberal rival, serious an offence as this is.

The Chief, moreover, must perform certain special duties for
the material welfare of his people. As we have already seen, he controls
the use and distribution of the tribal land ; he also organizes big
tribal hunts, especially when beasts of prey have been ravaging the
cattle; he organizes raids on neighbouring tribes to obtain more
cattle, and must also see to the defence of his own people's herds
from attack by others ; he is responsible almost everywhere for the
rainmaking ceremonies upon which the growth of the crops is held
to depend, while in many tribes he also distributes charmed seed
among his subjefts. Very often he also direftly organizes their
agricultural aftivities. Among the Tswana, Northern Sotho and
many others, he summons his people at the beginning of a new agri-
cultural season to hoe and sow his various public fields. Until they
have done so, they may not under penalty of punishment do the
same work on their own fields. Again, when the first fruits of the

170 I. SCHAPERA AND A. J. H. GOODWIN

new season are ripe, no one may eat of them until the Chief has
ceremonially done so. So, too, no one may harvest or thresh his crops
until this has been done with the produce of the Chiefs fields. In
this way the Chief" ensured that these operations would be performed
at a time which the experienced elders of the tribe knew to be most
favourable "- 1

Sub-Chiefs, and headmen also are entitled to exaft free labour from
their people for all public purposes, and to send them about freely
on official business. But they cannot claim tribal labour for their own
private purposes, although if popular may willingly be given assistance.
They may occasionally also be given presents of beer, meat, and similar
commodities. But there is no recognized form of tribute to be paid
to them other than the fines levied in their courts. In some tribes,
however, a Chief may permit vassal-Chiefs under his rule to claim
from their own people the various forms of hunting tribute to which
he himself is entitled, and also to have fields cultivated for them.
But they are allowed this by grace, and not by right ; and he may
whenever he wishes withdraw the privilege.

THE BREAKDOWN OF TRIBAL ECONOMY

The coming of the European has brought to the Bantu new forms
of organization largely incompatible with the traditional ways of
earning a living, and fresh needs which cannot be met by the traditional
methods of production and exchange. These innovations and their
effefts are dealt with more fully in a subsequent chapter, 2 but may
conveniently be outlined here to bring out the contrast between
Bantu and European systems of economy. The Native, while
attempting to cling to his own scale of values, finds it impossible not
to admit the competing claims of a system in which exchange is
carried out and values adjusted by means of prices. This has afFefted
domestic industries : the potter is no longer so aftive where the trader
has tin-ware for sale; the Native adzes and axes are everywhere
giving way to the European article. The traditional arrangements for
storing foodstuffs come to be upset by the pra&ice of sale to the
Europeans immediately after the harvest for cash or chits giving the
right to purchase goods. Imitation of the Europeans gives rise to
wants unknown before, for clothing and utensils of all kinds, and for

1 Rtpon of Native Economic Commission, 1930-2 (U.G., 22, 1932), 56.
1 See below*

WORK AND WEALTH 171

school education, which in the Native teachers creates a new class
of specialists who, on account of their training, often play a larger
part in public life than their age or social origins might seem to
warrant.

At the same time, the spread of European rule is depriving the
Chief of many of his fun&ions, limiting his rights of dispensing
justice, and, by striftly limiting the amount of land at the tribe's
disposal, is even preventing him from fulfilling his primary obligations
as father of his people. This again lessens the respeft in which he is
held, and makes difficult the full exercise of his claims over the labour
and property of his subjefts. The Government, moreover, demands
tax payments in money, thereby increasing the breach in the old
system of produftion and exchange. The need for earning tax money,
together with their relu&ance to sell cattle, drives many men out to
work for wages in mines or industries or on farms away from the
tribal home. Their absence, often prolonged, means changes in the
organization of family work; their additional earnings mean a
different scale of family incomes. The communal solidarity which is
the key to the traditional Bantu methods of making and sharing
wealth is lessened ; ties of relationship or neighbourhood, of respeft
for the elders and loyalty to the Chief, become less binding.

So a new class arises, who, in the words of one Tswana Chief, are
like bats : neither birds nor mice, neither Natives nor Europeans.
This description is not unjust. For the Bantu have certainly not yet
learned to make the most of the European method of catering for
wants by means of constantly adjusted price-changes. The Native
is not a close bargainer, and it will be long before his introduction to
money is complete. None the less, it is a dying system which we have
been describing above, a form of organization already well on the
way to being transformed in many parts of South Africa.

CHAPTER VIII
POLITICAL INSTITUTIONS

By I. SCHAPERA

THE basic unit in Bantu political life is the tribe (Nguni, isipve;
Shangana-Tonga, ri$aka y nyimta; Sotho, sethfaba, morafe; Venda,
lushaka). This may be defined, for all practical purposes, as the body
of people organized under the rule of an independent Chief. Each
tribe has its own name, occupies its own territory, manages its own
affairs, and afts as a single united body in war. But it is primarily
through their allegiance to the same Chief that the members of a tribe
are conscious of their unity ; and indeed the tribe is most often named
after the Chief himself or one of his ancestors. So, too, the Chief
himself is spoken of as Chief over his people, and not as Chief of the
territory they inhabit : e.g. Griffith Lerotholi is not " Paramount
Chief of Basutoland ", but morena e moholo oa BaSotho, " the great
Chief of the (Southern) Sotho."

The nuclear stock of a tribe is generally composed of people all
claiming descent from the same line of ancestors as the Chief. But even
the smallest tribe contains many alien families or groups, while in the
larger tribes only a small proportion of the people may belong to the
original stock. Usually as a Chief gains in power, prestige, and wealth
his tribe will be enlarged by the accession of refugees from other
Chiefs ; while, on the other hand, the unpopularity of any Chief will
gradually lessen the number of his adherents. Disputes over succession
to the Chieftainship, or quarrels arising from other causes among
members of the royal family, may lead to a split, the malcontents
either seeking refuge with some other Chief or setting up as an
independent tribe of their own. Then again, conquest by war may
lead to the incorporation of several different tribes under the rule
of one Chief, or to the disintegration of others, whose members scatter
all over the country seeking refuge where they can. The tribe is thus
not a closed group, like the clan. It is an association into which people
may be born, or which they may voluntarily join, or into which they
may be absorbed by conquest ; and which they may, for one reason
or another, leave again.

173

174 I- SCHAPERA

From time to time there have arisen in South Africa large Bantu
States, in which many different tribes were amalgamated into a single
political unit. The Zulu under Shaka, the Rhodesian Ndebele under
Mzilikazi, the Shangana under Soshangane, the South Sotho under
Moshesh, the Swazi under Sobhuza and Mswazi, the Pedi. under
Thulare, and the Ngwato under Kxama and his predecessors were
all powerful combinations in which one Chief had brought under his
rule many of his neighbours. 1 The Nguni rulers introduced certain
new features of social organization, to which reference will be made
below ; the others simply extended the traditional system of adminis-
tration to embrace the conquered tribes, allowing them a large measure
of autonomy under their own hereditary leaders but often appointing
sub-Chiefs to watch over them. Most of these large States have since
broken up, mainly as a result of wars with the Europeans, and have
reverted to the original system of small tribes. But in Basutoland,
Swaziland, and Northern Bechuanaland (Ngwato) the composite
" nations " still flourish. Here, however, allowing for the difference
in scale, the administrative system has on the whole remained funda-
mentally the same as that of the smaller tribes.

At the head of, the whole tribe is the Chief (Nguni, inkosi ;
Shangana-Tonga, host; Venda, khosi; Sotho, morena, kxosi). He
is assisted in the execution of his duties by various forms of council.
Local divisions within the tribe, such as distrifb, sub-distri&s, villages,
and wards, are in turn administered by their respe&ive heads, assisted
by small local councils. Each petty local authority is responsible in
the first place to the head of the larger local unit within which he
dwells. The latter, again, is responsible dire&ly or through some
senior local authority to the Chief. There exist also such institutions as
the army, age-sets, civil and ritual assemblies, through which the
tribe as a whole is on occasion marshalled direftly before the Chief.
In effeft, therefore, the government of the tribe is ultimately concen-
trated in the hands of the Chief; but the existing social and territorial
organization is used to delegate matters of more purely local concern
to subordinate authorities.

THE CHIEFTAINSHIP
Succession and Minority.

Chieftainship is hereditary. Many instances are known where it
has been usurped or acquired in some other way by trickery or force ;

1 For die history of these tribes, see above, chap. iii.

POLITICAL INSTITUTIONS 175

but as a rule the Chief succeeds automatically to his office by right
of birth. Kxosi ke kxosi ka a tsetswc, say the Tswana : (< a Chief is
Chief because he is born to it." But the rules of succession vary some-
what from group to group. Among the Nguni, Sotho, and Venda
Chieftainship normally passes from father to son. The rightful heir
is the eldest son of the Chief's " great wife ", the woman for whose
lobola cattle have been contributed by members of the tribe. 1 Among
the Shangana-Tonga a Chief is usually succeeded by his brothers in
turn, and only when the last brother has died does the succession revert
to the sons of the eldest. 2 Failing a direft heir, the Chief is everywhere
succeeded by the man next in order of seniority. Frequently, however,
the succession is disputed by rival claimants, even when the real
heir is well known ; and there will be strife and tribal disruption.
Women do not normally succeed, although they may at times a&
as regents. But among the Lobedu and certain other small North
Sotho tribes the Chief must always be a woman. She is, however,
for social purposes regarded as a male, marries " wives " (who then
cohabit with her male relatives), and is succeeded by the eldest daughter
of the " great wife ". 3 This form of Chieftainship is, as far as the
Southern Bantu in general are concerned, decidedly exceptional.

The heir is normally installed as Chief soon after the death of his
predecessor. Occasionally the leading councillors of the tribe, if
they consider him unsuitable for the Chieftainship, will intrigue
against him and bring about the succession of a more satisfactory
though junior relative. But as a rule the claims of legitimacy are more
than sufficient to counterbalance personal disqualifications. The new
Chief is first " doctored " with various medicinal charms to give him
the power of commanding the obedience and respect of his people.
He is then formally invested with the insignia of Chieftainship, the
occasion often being one of elaborate public ceremonial ; is entrusted
with the sacred tribal regalia, and enters into the inheritance of his
predecessor's household and property.

If the heir is still too young to take up office, the dudes of Chieftain-
ship are entrusted to a regent. The regent is generally, among the
Sotho and Nguni, the late Chief's senior surviving brother or an
older son by a junior wife. He normally afts alone, but sometimes,
especially among the Swazi, the heir's mother may have as much or

1 On the "great wife", aee above, chap, iv, pp. 75, 83, 91.

1 Junod, 1927, I, 4io-4.

Cf. A. A. Jacques, 1934, 37T-3 8 *-

i?6 I. SCHAPERA

even greater authority in the direction of tribal affairs, and the male
regents merely a&s as her mouthpiece. The regent while in office has
all the powers, rights, privileges, and duties of the Chief. He also
receives and keeps as his own the various forms of tribute customarily
paid to the Chief. He is expe&ed to hand over the Chieftainship when
the heir comes of age, i.e. when he has passed through the ceremonies
admitting him into adult membership of the tribe. But tribal histories
show that a regent has often enough been tempted by the fruits of
office to retain the Chieftainship permanently. If he has been a success-
ful and popular ruler, he is generally assured of support, with the
result that there may be a split in the tribe. But if on the other hand he
has abused his position, he will be forced, perhaps even sooner than
really necessary, to make way for the heir; and should he resist
the penalty may be assassination. 1

Prerogatives and Wealth.

The Chief, as head of the tribe, occupies a position of outstanding
privilege and authority. He is the symbol of tribal unity, the central
figure round whom the tribal life revolves. He is at once ruler, judge,
maker and guardian of the law, repository of wealth, dispenser of
gifts, leader in war, priest and magician of his people. Among the
Venda and some Northern Sotho, as among Shaka's Zulu, he is
elevated to an almost godlike eminence: his person is sacred, his
subjefts bow before him in humble adoration and obeisance, and his
smallest gesture is greeted with a chorus of fulsome adulation. 2 Among
the Venda, indeed, " not only is the Chief thus regarded as semi-
divine during the greater part of his life : towards the end of it, or
sometimes long before, he actually confers godhead upon himself,
when after abjuring all contaft with women, and putting away his
wives, he performs a solemn solitary dance (u pembela) which makes
him in very truth a god (MuJiimu)." 3 In other tribes, such as the
Tswana, he is far more approachable ; but even here he is greatly
revered and always treated with a good deal of outward respeft.
His deeds are extolled in magnificent eulogies recited on all public
occasions by special bards. He himself is often addressed not merely
by his official title but in words or phrases of extreme adulation,
while among the Tswana he is correftly known by the personification

1 Cf. Schapera, 1935 MS. (Tswana); Lestrade, 19300, 319 (Venda).

1 Cf. Stayt, 1931, 201-3; Lestrade, 19300, 311-12; Krige, 1936, 237-9.

* Le^irade, loc. cit.

POLITICAL INSTITUTIONS 177

of the tribal name. 1 His initiation, installation, and marriage are
occasions of great public festivity. There are certain insignia of rank,
such as leopard skins, ivory armlets, or bead necklaces of a particular
kind, which only he is allowed to wear ; and he alone, in the Nguni
tribes, is entitled to the royal salute bayede.

The Chief's household is usually far larger and more elaborate
than those of ordinary tribesmen. He always has many wives, both
at his own residence and (as among the Nguni and Shangana-Tonga)
at military kraals or other places of importance throughout the tribe ;
while numerous officials, retainers and other adherents cluster round
him and his court. He has the first choice of land for building upon,
for cultivation and for grazing his cattle. He and his family take
precedence in the tribe in matters of ritual, e.g. at the first-fruits and
initiation ceremonies. This precedence is a prerogative so jealously
guarded that anyone daring to infringe it is severely punished as a
potential rival. The Chief alone has the right to convene tribal
meetings, arrange tribal ceremonies, and impose the supreme penalties
of death and banishment. He receives various forms of tribute from
his people, both in kind and in labour. 2 He further has the right in
general to obedience from his subje&s in all matters of public interest,
as well as in minor matters of more personal concern. Failure to
comply with his orders is a penal offence. All other offences against
him are generally punished more severely than similar offences against
ordinary tribesmen ; while disloyalty or revolt against his authority
is a major crime, punished as a rule by death and the confiscation of the
culprit's property.

Duties and Obligations.

But the Chief's life is not merely one of immense privilege. He has
many duties to perform for the tribe, duties which if faithfully carried
out may impose an enormous burden upon his time and energy. He
must in the first place watch over the interests of his subjefts and
keep himself informed of tribal affairs generally. He is said to be the
father or herdsman of his people, and as such has to look after them,
treat them well and justly, and see that no harm or misfortune befalls
them. He must give ear to all his subjefts, irrespeftive of rank ; and
much of his time is spent daily in his official courtyard (Sotho,

1 The Chief of the Ngwato, e.g., is addressed and referred to as MoNgwato, and the chief
of the Kxatla as Mokxatla.

Cf. above, chap, vii, pp. 166 ff.

178 I. SCHAPERA

kxotla\ Venda, khoro\ Nguni, inkundla\ Shangana-Tonga, hubo\
listening to news, petitions, and grievances from all parts of the tribe.
His accumulated wealth, as shown in the last chapter, must be used for
the tribal benefit as well as for his own maintenance ; and much of
his popularity depends upon his reputation for hospitality and
generosity.

The Chief is the executive head of the tribe. Nothing of any
importance can be done without his knowledge and authority. But
in administering tribal affairs he must always consult with his councils,
both private and public ; and it is one of his main duties to summon
and preside over meetings of these councils as occasion arises. With
them he must decide upon questions of peace and war, and see to the
proteftion or relief of his people in case of war, pestilence, famine, or
some other great calamity. He must see that the local divisions of the
tribe are effe&ively governed by their sub-Chiefs or headmen, and
take any a&ion that may be necessary to ensure this. He controls the
distribution and use of the tribal land, of which he is often figuratively
termed the owner ; and he must in many tribes regulate the sowing
and harvesting of crops, the organization of tribal hunts, trade relations
with strangers, and other economic aftivities. He is the head of the
tribal army, and as such organizes military expeditions, often accom-
panying them himself, performs the necessary war magic for the
success of his troops, and disposes of the prisoners and loot. He is
also the representative and spokesman of his tribe in all its external
relations, and so must receive and entertain other Chiefs or their
messengers, and communicate in the same way with his neighbours
when necessary. All strangers visiting the tribe must be reported
to him, while none of his own people may go away without his know-
ledge and permission.

The Chief is further responsible for maintaining law and order
throughout the tribe. He must proteft the rights of his subjefts,
provide justice for the injured and oppressed, and punish wrongdoers.
He is the supreme judge, whose decision is final. Appeals lie to him
from the verdi&s of the lower courts ; while certain serious offences,
such as treason, homicide, assault, rape and sorcery must be dealt
with only by him and his court. He may further initiate and, after
discussion with his councillors and the tribe as a whole, promulgate
new laws and regulations for the better conduft of tribal affairs ;
and similarly abolish old usages which the tribe has outgrown.
Legislation of this sort does not seem to have played a conspicuous

POLITICAL INSTITUTIONS 179

part in the old tribal life, the Chief being more concerned to maintain
the existing laws than to alter it ; but in more recent rimes decrees
have often been found necessary to meet the new conditions produced
by contaft with the Europeans.

As religious head of the tribe, the Chief always plays an extremely
important part in its ritual life. He must arrange for the celebration
of all the great ceremonies, such as rainmaking and other agricultural
rites, tribal purifications, initiations, and the charming of the army,
upon which the welfare of his people is held to depend. He is the
custodian of various sacred obje&s symbolizing the unity and
prosperity of his tribe, such as the inkatha or sacred coil of the Zulu
and the mhamba of the Thonga, both containing among other
ingredients " dirt " from the bodies of past and present Chiefs.
He further possesses horns or vessels filled with a potent magical
paste giving him power over his enemies, and must use them for
" doftoring " the tribe in time of war or other emergency. Even if
he himself has not learned to perform the necessary fites, he has
official magicians to do so ; and they always work at his request and
under his direft supervision.

In many of these rites the Chief is the link between his people and
the ancestral spirits governing their welfare. His dead ancestors are
held to afford supernatural protection and assistance to the people
they once ruled; and on all important occasions he will sacrifice
and pray to them on behalf of the tribe as a whole. The role he thus
plays as tribal priest a role which only he, as ruling Chief, can
fill helps to explain the great reverence in which he is always held
by his people. He himself becomes a tribal god after his death. He is
buried in some traditionally sacred spot ; apd, especially if he has
been renowned or well-beloved, his descendants will come there with
their people to pray to him for rain and harvest, peace and prosperity. 1

Relatives of the Chief.

The Chief's paternal relatives (Sotho, diheosana, borrangwanakxosi ;
Venda, magota ; Swazi, bantanenkosi, etc.) share to a varying extent
in the rights and privileges of his exalted position. Generally speaking,
the more closely a man is related to the Chief, the more powerful
a person he is in tribal affairs generally. If in addition he is personally
capable, his authority will be correspondingly greater. If on the other

1 See below, chap, xi, for a fuller account of the Chief's religious position.

l8o I. SCHAPERA

hand he is incompetent, a waster or a drunkard, his influence will
be correspondingly less. But owing to his inherited status he will
nevertheless continue to command some respeft from his social
inferiors. The Chief's brothers and uncles, in particular, have the
right to be consulted by him on all important tribal matters, and often
exert considerable influence and authority as his private advisers.
One of them generally afts for the Chief during his absence or illness ;
and in certain tribes (e.g. Venda, Ngwato) this man's position is
constitutionally demarcated. 1 Others are appointed by the Chief as
sub-Chiefs over outlying distrifts or vassal-tribes, while among the
Sotho they may further have individual jurisdiction as leaders of
mephato (age-regiments). Without their aftive support he can do
little ; but if they are solidly behind him his hold over the tribe will
be greatly strengthened.

On the other hand, it is usually these men who form the Chief's
most dangerous rivals, especially when he is weak or tyrannous.
They are jealous of their rights, and adtively resent any failure on
his part to co-operate with them in ruling the tribe. Even if they do
not dispute the succession with him before he is installed, they may
subsequently intrigue against him, relying for support upon the
prestige of their birth and such personal popularity as they possess.
Dynastic troubles of this sort are of frequent occurrence in Bantu
society ; and although they may not always lead to civil war or to the
secession of an ambitious brother or son of the Chief with his
following, they split the tribe into fa&ions whose continual disagree-
ments greatly disturb the general peace and prosperity. The success
of a Chief's reign is therefore often determined by his treatment of
his close relatives and by their own attitude towards him.

The Chief's mother, especially when her lobola cattle have been
contributed by the whole tribe, is its most important woman. She is
spoken of as the " mother " of all his people, and is treated by them
with the utmost respeft, so much so that among the Sotho and others
her home is a recognized asylum to which vitims of persecution can
flee. She sometimes afts as regent during her son's minority, or if not
must be consulted by the regent in all matters affefting the royal
family and estate ; while after his accession the Chief frequently
seeks her advice in political matters. Among the Swazi she occupies
a position of unique importance. She is, next to the Chief himself,
the leading member of the tribal government, and afts as the principal

Stayt, 1931, 197-8; Schapera, 1935, MS.

POLITICAL INSTITUTIONS 181

regent during his minority. She has her own kraal, which is the ritual
capital of the tribe ; the annual first-fruits festival is held there, as
well as meetings of the tribal councils, and a regiment is always
stationed there to proteft her and to work for her. She exercises
separate judicial powers ; she is the custodian of the sacred rain
medicines, which her son uses in collaboration with her > he invariably
consults her before taking aftion in any important matter, and should
follow her advice ; and she can take an aftive part in all council
meetings. 1 Among the Venda the Chief's senior paternal aunt
(makhadii), or failing her his senior sister, has a somewhat analogous
position. She must be consulted on all important tribal affairs, exercises
a very great influence over the Chief's personal and political conduft,
and is treated with most of the respeft and formality accorded to him. 2
The sisters and daughters of the Chief, as leaders of the female
age-regiments, may among the Tswana also command special
authority over the women of the tribe, apart from the general respeft
they enjoy by right of birth. It is fairly common in all tribes for the
Chief to marry them to sub-Chiefs or headmen with large followings,
and so consolidate more strongly his hold over his people ; or to
other Chiefs with whom he is anxious to be allied. The Venda and a
few neighbouring North Sotho tribes appear to be exceptional in
that among them a Chief may appoint one (or several) of his sisters
as headman over some village or even distrift, the succession to this
office then tending in some cases to be confined to females. 3

TRIBAL COUNCILS

In administering the affairs of the tribe, the Chief is assisted in the
first place by a small number of confidential advisers, with whom he
consults informally and secretly. Most of them are his own close
relatives, such as his senior uncles and brothers ; a few others may
be influential sub-Chiefs or headmen ; and there may also be one or
two devoted personal friends and adherents. A Chief when he succeeds
will generally retain the advisers of his father, but as times goes on may
discard some of them and seleft his own. The office is not necessarily
hereditary. These men help the Chief to formulate policy, and discuss
with him every measure which will subsequently have to be referred

1 B. A. Marwick, Th* Natlvts of Swaziland, MS. 1934.

Stayt, 1931, 196-7; Lestrade, 19300, 314.

Lestrade, 1930*1, 314-15; Stayt, 1931, 315; Jacques, 1934, 177-3*2 passim.

1 82 I. SCHAPERA

to the great tribal council. They should keep him informed of public
opinion and of what is happening in the tribe generally, and must
remind him of matters needing consideration and attention. He is
not bound to follow their advice, and may even ignore it if he wishes ;
but he will not as a rule deliberately oppose their united opinion.
They must further keep check over his own behaviour, and are
expefted to warn and even reprimand him if he goes wrong.

One of these advisers is the special official commonly known as
the Chief's induna (North Nguni). 1 He is not as a rule a member of
the royal family, but a commoner of conspicuous loyalty, ability, and
trustworthiness. The office often descends from father to son, but not
necessarily, as the Chief has the right of selection and appointment.
But the induna of one Chief will generally continue to serve in the
same capacity under the latter's successor. He is the Chiefs right-
hand man, afting as intermediary between him and the council and
tribe, and as his mouthpiece on all formal and many informal occasions.
All matters referred to the Chief for discussion or a&ion must go to
him in the first instance. He receives all cases that come to the Chief's
court, makes the necessary arrangements for their hearing, and may
himself aft as judge in the absence of the Chief; he assists the Chief in
leading the army and in carrying out the great tribal ceremonies ;
he is responsible for the fulfilment of all the Chief's orders; and he also
supervises the running of the Chief's household. Important Chiefs
ruling over large tribes may have several such officials, sharing between
them the various functions just mentioned. 2

The Chief must on occasion consult a much wider, more formal
council. The exaft strufture of this great tribal council (Nguni,
itanJla, inkundla; Venda, kAoro; North Sotho, kxoro\ Tswana,
lekxotla ; etc.) varies somewhat from group to group. It is, however,
made up as a rule of the Chief's private advisers, some of his more
distant relatives, all sub-Chiefs and headmen of local divisions and
portions of the tribe, and various influential commoners appointed
because of their ability. Some of these councillors are always in
residence at the Chief's kraal or village, or come to live there for a
while, helping him to deal with the cases brought to his court. But
on important occasions the whole body is specially summoned together.

1 The corresponding terms elsewhere are : umphakathi omkhulu (Xhosa), ndjuna (Shangana-
Tonga), mukhoma (Venda), monna-kx6r6 (N. Sotho), ntona or kola ya kxosi (Tswana).

Among the Venda, e.g. the mfhasi (" master of the ground M ) is the legal adviser and
Assessor judge ; mukhoma (" leader ") runs the Chief's household j and ntgota is the Chief's
head councillor and ceremonial assistant (Stayt, 1931, 198-9).

POLITICAL INSTITUTIONS 183

It deals with all matters of tribal policy ; and no aftion can be taken
nor can any new law come into force until it has been approved of
here. The decision previously arrived at by the Chief and his private
advisers is put before the members of the council, who then freely
discuss and criticize it. They may accept, modify, or rejeft it ; and
whatever decision they reach is generally carried into effeft. The
Chief is in strift theory able to override their unanimous opposition,
but in praftice he rarely ventures to do so. Their co-operation is
essential for the successful government of the tribe ; and should any
Chief aft contrary to their collective advice the result will probably
be a split. It is this council also which exercises the greatest check
upon his behaviour, inasmuch as the members seldom hesitate to take
him severely to task for his shortcomings.

Among the Nguni, Shangana-Tonga and Venda this council
is in effeft the governing body of the tribe. It represents the people
and their opinion, and is the link between them and the Chief. After
the meeting each local official returns to his own distrift, where he
informs his assembled followers of the decision arrived at, which is
now binding upon the tribe as a whole. But among the Sotho all
matters of public concern are dealt with finally before a general
assembly of the initiated men. This assembly (pitsS) must be summoned
by the Chief whenever occasion arises. It is usually, but not invariably,
preceded by a meeting of the great council. The Chief, who presides,
puts the issue before the people. Anybody present can then take part
in the ensuing debate, although preference is given to influential or
elderly men. In theory great freedom of speech is permitted. In
praftice it is not always exercised, except under provocation or a
sense of grievance, owing to the fear, by no means unfounded, of
subsequent reprisals from the Chief. But if the occasion calls for it,
he will be contradifted, criticized, reprimanded, and even violently
attacked with little hesitation. Should most of the speakers express
different views from those favoured by the Chief and his advisers,
the latter will attempt to argue them round ; but no Chief would dare
oppose public opinion as here revealed unless he is looking for trouble.
Normally, however, the voices of the councillors, if they support him,
will carry sufficient weight with the general assembly for the decision
previously arrived at to be approved, and thus enable the Chief to
carry his policy into effeft. 1 Similar tribal assemblies are held among
the Nguni and Shangana-Tonga before war and during the annual

1 Schapera, 1935 MS.

184 I. SCHAPERA

firstfruits ceremonies ; but they are of comparatively rare occurrence
otherwise, and do not play nearly as conspicuous a part in the system
of government.

The existence of these councils greatly limits the Chief's aftual
exercise of his power. Political life is so organized that effeftive
government can result only from harmonious co-operation between
him and his people. Kxosi ke kxosi ka morafe, say the Tswana :
" a Chief is Chief by grace of his tribe." Despite the faft that ultimate
control over almost every aspeft of tribal life is concentrated in his
hands, and the very considerable authority he consequently has,
he is very seldom absolute ruler and autocratic despot. In order to
get anything done, he must first gain the goodwill and support of his
advisers and councillors, who play a considerable part in restraining
his more arbitrary impulses. Any attempt to aft independently of
them is not only regarded as unconstitutional, but will also generally
fail. A good deal, of course, depends upon the personal character of
the Chief. A forceful and energetic man can succeed in dominating
his subjefls and ruling to all intents and purposes as a di&ator. The
great Ngwato Chief Kxama (Khama) was an outstanding man of this
type. Cases are even known where such Chiefs have been able for
many years to exercise the cruellest of tyrannies. The most notorious
of all was Shaka, although he was by no means unique. But the
average Bantu Chief depends for much of his power upon the assistance
of his councillors, and his behaviour is accordingly controlled. Under
a weak and incompetent Chief their influence is very pronounced,
fa&ions arise, and " kingmakers " are by no means unknown.

Moreover, the Chief himself is not above the law. Should his
aftions run counter to accepted standards of what is right and proper,
he is severely criticized by his councillors and the people at large.
Should he injure any of his subjefts, he is expefted to make the cus-
tomary reparation ; and in some tribes (e.g. Swazi, Tswana) may even
be tried and punished by his own council. If he persists in flagrantly
misruling, his people will begin to desert him ; or a civil war may be
instigated, in the hope that he will be overthrown and one of his more
popular relatives take his place ; or, as a final resort, he will be poisoned
or assassinated at the first suitable opportunity. 1 Instances abound
in which despotic or hopelessly unsatisfactory Chiefs have thus been
put out of the way by their despairing or infuriated subje&s, as happened

1 Cf. Lestrade, 19300, 312-13, for a concise statement of the limitations upon the absolute
autocracy of the Venda Chiefs.

PLATE XVII

s/?> The great tribal Avil/d, Moc-ImeH, B.P,

(I.

I/. Stkupnm

(b) Graves of the k\\ena duels ^.Sechele an<l Sebcle)
in the tribal cat tic-kraal, Molepolole, B.P.

POLITICAL LIFE

PLATE XVIII

[i.

(a) Morning gossip at a kxotla (Mochudi, IIP.)

(b) Hearing a lawsuit (Kxatla)
POLITICAL LIFE

f /. &hafiira

POLITICAL INSTITUTIONS 185

to Shaka himself. So great, however, is the reverence attached to the
Chief by virtue of his birth and ritual position that his people will put
up with much from him that would never be tolerated in one of
lesser rank ; and it is only under extreme provocation that drastic
action will be taken against him.

LOCAL ADMINISTRATION

Apart from the central Government, composed of the Chief and
the councils mentioned above, there exists in every tribe a system
of local government. The territory occupied by the tribe is divided into
local units, varying in size and importance, each of which comes under
the jurisdiction of a recognized authority. The structure of these
units has already been dealt with in a previous chapter, 1 so that we need
do no more here than briefly outline the manner in which they are
respectively administered.

Although every household or kraal is to some extent a distinct
legal and administrative group, the smallest effective political unit
is the sub-district or ward, under the control of a local headman.
The headman is sometimes a member of the Chief's family, sometimes
a commoner. He is either formally appointed by the Chief, or at least
confirmed, the latter being the more usual since the office tends to
be hereditary. The headman is responsible to the Chief for the peace,
order, and good government of his area, in which he is the Chief's
representative. He must help his people in their troubles, and sponsor
them before the Chief; he regulates their occupation and use of land,
and controls the right of strangers to settle there ; he prays and
sacrifices on their behalf to his ancestors, and performs other cere-
monies for their well-being and protection. He deals with cases which
the kraalheads have not been able to settle, or have referred to him
in the first instance ; and he has the power to impose fines and other
forms of punishment. But he must himself refer all cases of serious
difficulty to his superior authority, and there is also an appeal from his
verdict. He must further see that his people pay the customary forms
of tribute and carry out the commands of die Chief, conveyed to
him by special messengers or at council meetings. He must himself
visit the Chief from time to time, not only to assist at the council
but also to report upon the affairs of his area. 2

1 Cf. above, chap. iv.

* Schapera, 1935, MS. (Tswana) ; Lestrade, 19300, 308 f. (Vcnda).

186 I. SCHAPERA

His own privileges are relatively few. His home is built in the
most favoured part of his area, and he has the pick of the land for
cultivation and grazing. He can claim free labour from his people for
such public services as building and repairing his official courtyard,
taking messages or executing the verdiCb of his court ; bat he does
not as a rule receive any private tribute, either in labour or in kind,
except 'with the special permission of the Chief. By virtue of his
office, however, he is as a rule greatly respefted, not only by his own
people but in the tribe as a whole ; and if he has a large following he
may play an important part in tribal politics generally.

The headman is assisted by a small council (ibandla, etc.) embracing
his own senior male relatives and the more important kraalheads,
together with any other elderly men of repute in his area. They sit
with him in judgment over the cases that come to his court; he
discusses with them all other matters relating to the public life of
the group, and consults them on tribal questions generally in connection
with meetings of the Chief's council. They also keep check upon his
own behaviour. They can, if need be, reprimand him severely;
but more generally, if he is negligent, incompetent, or unduly severe,
they report him to the Chief, who may then fine him or even depose
him in favour of a more suitable man.

Large districts (or, among the Tswana, outlying towns) are in the
same way administered by what may be termed sub-Chiefs. These
men are sometimes close relatives of the Chief, or important
commoners specially placed there by him as his representatives
and deputies ; sometimes they are the headmen of the senior local
clans or wards ; and sometimes they are the hereditary rulers of the
vassal-tribes inhabiting that district. In the larger tribes there are often
several grades of sub-Chiefs, their importance varying with the size
of the area and the number of people under their jurisdiction. An
important sub-Chief may thus have jurisdiction over several minor
sub-Chiefs, who in turn each have separate jurisdiction over several
headmen. The functions of the sub-Chiefs are on the whole similar
to those of the headmen of sub-distriCts or wards ; but they also aft
as courts of appeal from the verdiCts of the latter, and have an over-
riding authority in all other matters. They are assisted by councillors
embracing the headmen of their district as well as their own private
advisers. They are themselves direCHy responsible to the Chief. He
issues orders to them, which they must carry out under penalty of
punishment ; he can receive appeals from their courts ; he can alter

POLITICAL INSTITUTIONS 187

their boundaries at any time ; and he can, if their conduCt of affairs
is unsatisfactory, install someone else in their place.

The system of local administration may thus be regarded as one of
ever-widening jurisdiction extending upwards from the household
or kraal. In many respeCts the kraalhead, the headman of a sub-distriCt
or ward, the sub-Chief of a district or town, and the tribal Chief himself
have identical or similar functions. The main difference lies in the
size and composition of the group over which each of them has
jurisdiction, and in the corresponding range of his power. But within
the limits of his jurisdiction he has undivided power: judicial,
administrative, economic, and on occasion also military and religious.
One point of special interest in all these groupings is the extent to
which the personal character of the head plays its part. The more
popular a man is, and the higher his prestige, the greater the number
of followers he can retain and attraCt. His power is based upon their
support ; it is therefore necessary for him, if only to satisfy his ambition
and pride, to treat them generously and justly. This is as true of the
kraalhead as of the tribal Chief; and its practical recognition has
contributed greatly towards ensuring what, in spite of many lapses,
is on the whole an equitable system of administration.

CITIZENSHIP AND STATUS

A distindion is made in every tribe between its own members and
aliens. Membership of the tribe, as previously noted, is defined not
in terms of birth, but of allegiance to the Chief: it is possible for
people not born into a tribe to become subjeCts of its Chief either by
conquest or by placing themselves voluntarily under his rule. Every
member of the tribe must conform to its laws, and must also carry out
certain special obligations, such as working for the Chief, paying him
tribute, fighting in his army, and attending tribal meetings. He is
entitled in return to land on which to ereCt his home, cultivate his
crops, graze his cattle, and to all other facilities for earning a liveli-
hood ; and to protection by the tribal authorities against undue inter-
ference with his person, his activities, and his property.

The extent to which members of the tribe enjoy the advantages and
protection of the law varies according to their rank, sex, and age.
Hereditary rank in the Bantu social system is largely confined to the
Chiefs, sub-Chiefs, and headmen, with their respective families. The
special rights and duties of these political authorities have already

l88 I. SCHAFERA

been discussed. Commoners can acquire individual prestige or
influence through knowledge of the law, conspicuous loyalty to the
Chief, ability in debate, the praftice of magic, bravery and skill in
war, wealth, or some similar faftor ; but distinctive qualities of this
kind do not necessarily lead to enhanced social status, unless such a
man is made a headman or tribal councillor. Social distinftions are
often also made according to tribal origin in matters of ritual and
public life. Incorporated foreigners or their immediate descendants
seldom command the same influence or receive the same consideration
as tribesmen of long-established stock; and among the latter a
similar range of differentiation exists according to seniority of descent.

Apart from the political authorities, all members of the tribe, both
male and female, can generally be grouped, as far as personal status
is concerned, into : (a) heads of households, and () other inmates of
households. The head of the household occupies his position by
hereditary right : he is not appointed or even formally confirmed by
higher authorities, save in cases of disputed succession. He not only
controls the social, economic, and religious life of his dependants,
as elsewhere described, but is also their legal and administrative head.
He is entitled to their respeft, obedience, and service ; keeps order
over them, and deals with any disputes or quarrels arising among them ;
is responsible for their conduct in matters afiefting outsiders, watches
their interests at the tribal courts, and can be held liable for their
misdeeds and debts. He is assisted in the performance of his duties
and the exercise of his rights by the older men of his household, with
whom he informally discusses matters of immediate domestic concern.
But his authority is moral and persuasive, rather than compulsory.
He can inflift corporal punishment upon his own wife, children, or
servants, but in regard to other members of the household he has no
means of enforcing his will except through general domestic discipline,
and must refer all important issues to his local headman, to whom
there is also an appeal from his decisions. The continuity of the house-
hold is secured by a recognized system of devolution, the eldest son
of the principal wife being the general heir and legitimate successor
of his father. He not only takes the latter's place in respeft of the use
and enjoyment of property, but also undertakes his duties and respon-
sibilities, and must shoulder any outstanding liabilities.

Children claim descent from their father (i.e. the husband of their
mother) ; and succession to office and inheritance of most forms of
property go normally from father to son. Legitimacy is determined as

POLITICAL INSTITUTIONS 189

follows. The child of an unmarried woman belongs to her own family,
and is subject to her father or guardian. Its own father has no claim
to it unless he makes a special payment in cattle. Even then the child
is subordinate to his legitimate children in social and economic rights
and privileges, nor can it inherit except in default of male issue on the
part of his wives. The child of a married woman, born during the
subsistence of her marriage, belongs to her husband, whether he is
actually its father or not. From this it follows that a child born of
adulterous intercourse does not belong to its natural father, even
although he may have paid damages to the woman's husband in
respeft of the conception of the child. The child born to a widow at
her late husband's home is regarded as his ; it bears his name, and
has the same rights as the children born to her during her husband's
lifetime. This is particularly so where the widow has been taken over
by her husband's brother or some other approved relative under the
" seed-raising " custom. The child born to a divorced woman by
any man to whom she is not married belongs to her former husband
if his lobola cattle have not been returned, and to her own people if
they have been. If a widow or divorced woman remarries at her own
home, and a new lobola is paid, any child subsequently born to her
belongs to her new husband.

The status of a child is sometimes altered by adoption. This is most
commonly practised among the Sotho, where a married couple with
no children of their own, or children of one sex only, will adopt the
child of a close relative or friend. An adopted child comes under the
authority of the foster-parents, and has all the rights and duties of
their own legitimate offspring. Such adoption must be distinguished
from another widespread custom by which a child is sometimes sent
to live for a while at the home of its mother's people or other relatives.
Its father is expected to contribute towards its keep while it is there,
which he generally does by paying an ox or more as " maintenance
fee ". This fee must be paid before he can regain control of the child.

Unmarried people are all regarded as minors, even if they have
passed through the initiation ceremonies. They are under the particular
control of their own parents, especially of the father, and are also
subject, in household affairs generally, to the authority of the house-
hold-head. If they are still minors on the death of their father, one
of the paternal uncles or some other close relative takes his place as
their guardian. He has full control over their persons and property, but
must account to the Chief for his stewardship. Should he abuse his

190 I. SCHAPERA

trust, either by ill-using the children or by squandering their property,
the Chief appoints another guardian from among the members of the
family. There must always be a male guardian, even if the mother
is still alive. The Chief himself afts as " upper guardian " of all widows
and orphans, and should see to it that they are properly cared for.

Minor children cannot leave home or marry without the consent
of their father or guardian. They may not sue or be sued unless
assisted and represented by him. They may own property, but cannot
dispose of it without his consent. They must give him all their earnings,
refer to him all their business transactions, and generally obey him
in all matters. In return they have the right to maintenance, proteftion,
and good treatment ; to be informed of all household affairs ; to be
instru&ed and properly initiated into tribal life ; and to be provided,
when the time arrives, with a wife or husband, as the case may be.

A man when he marries and sets up his own household passes for
all praftical purposes out of the tutelage of his father ; but even so is
expefted to render him a good deal of obedience and service, to show
him every respeft, and to support him in his old age. On the other
hand, so long as he lives in the same household, he is subjeft to the
same authority as its other inmates in matters pertaining to the
household generally. But he controls his own wife and children, and
must be consulted by the household-head in all dealings with them ;
and is capable of suing or being sued in respeft of his own private
dealings or transa&ions. He should, however, refer all such dealings
to his household-head, as by doing so he is prote&ed ; and is bound to
inform him of all his movements and responsibilities. The same rights
and duties apply to other married male inmates of the household.

A woman when she marries passes from the legal control of her
parents into that of her husband, who now becomes her guardian,
and as such responsible for her aftions. She cannot sue or be sued
except through him, while if she does wrong he is generally held
liable and must pay any fines or damages incurred. It is his duty to
proteft her and to treat her kindly and considerately; to cohabit
with her regularly, subjeft however to the special taboos on sexual
intercourse between married people; to provide her with a hut,
food, clothing, and maintenance generally, and to .assign fields, cattle,
and other property to her house for her use. If ill-treated, she can go
back to her own home, and her husband cannot reclaim her until he
has paid one or more cattle to her people as damages. The wife, in
return, must work for her husband and be faithful to him, bear and

POLITICAL INSTITUTIONS 191

nourish children for him, work her fields and prepare the food, and
generally occupy herself with the many domestic duties which family
life entails. She can acquire property in her own right, but must
consult her husband in all her business dealings. She must be in all
respefts subservient to his will, and must live wherever he chooses to
build his home.

In public life generally women do not have the same rights as the
men. They are never consulted on questions of tribal policy, and all
the political offices are kept in the hands of the men. Among the
Venda and Swazi, it is true, the Chief's female relatives may play an
important part in the tribal government ; but such instances are on
the whole exceptional, and in no way enhance the political status of
women generally. Moreover, certain spots in all the village settle-
ments, such as the public courtyard, and under various conditions the
cattle-kraal, are reserved for men, and normally inaccessible to women.
Finally, preference is everywhere given to sons, and a woman bearing
daughters only is often regarded with contempt. She may even have
to request her parents to provide a " substitute wife " to bear a son for
her house.

Among the Tswana the most subordinate position in the tribe is
occupied by the MaSarwa (Bushmen) and Kxalaxadi. The former
especially constitute a distinft servile class. They are attached to the
Chief and other wealthy or prominent members of the tribe, for whom
they must perform all kinds of menial services. They can be transferred
altogether or lent by one man to another, irrespe&ive of their own
wishes ; if ill-treated they have no other remedy than to run away if
they can, but are often followed up and brought back forcibly;
if living in the veld, they can have their children taken away from them
and brought to their master's home for domestic service ; they have
no right of a&ion against their master for any wrongs he infli&s
upon them, such as seducing their women ; and they have no share
in the tribal government, being deprived even of membership of the
regiments. 1

INTER-TRIBAL RELATIONS

The boundaries between tribes are as a rule roughly defined by
natural barriers or objefts, such as rivers, mountains, or watersheds.
If relations are friendly, a good deal of visiting takes place between them.

1 Cf. E. S. B. Tagart, Report on the Conditions existing among the Masarwa of the Beckuanaland
Protectorate (Pretoria, 1932), 4-10 ; and London Missionary Society, The MaSarwa : Report
of an Inquiry (Tigerkloof, 1935).

IQ2 I. SCHAPERA

Intermarriage is also fairly common, although generally people prefer
to marry within their own tribe. Traders and magicians go from one
to the other seeking business ; the courts of the one are open to the
members of another with grievances against some local tribesman ;
or refugees may flee from one to the other. All visitors of this kind
must be reported to the Chief or local sub-Chief, and their presence
explained. If they have come to trade, they must first obtain his
permission to do so, generally by making him a payment " to open
the way " for them. If they have come as refugees, they must offer
him allegiance (Nguni, ukuhhonia) and secure his goodwill by a similar
payment, generally of livestock, in return for which they are given
permission to settle. Temporary visitors mostly reside with a relative,
friend, or acquaintance, who, in accordance with the widespread
tradition of hospitality, must entertain them and give them presents.
They are under the proteftion of the local sub-Chief or headman.
If anyone injures them, they are entitled to justice and reparation. But
they in turn are expected to behave with great circumspeftion and to
obey the laws of the tribe ; and should they commit any offence they
will be punished and ordered to leave the country.

Diplomatic relations are maintained between neighbouring Chiefs
by means of recognized court messengers. Each Chief informs his
neighbours of all important events in his tribe, such as the holding of
an initiation school, the outbreak of some pestilence, or the death of
his predecessor ; and invites them to his own installation or marriage.
If there is trouble between a Chief and members of his tribe, it is often
the praftice for one of the parties to call in some neighbouring Chief
to try and reconcile them. Chiefs descended from the same ancestor,
but now ruling over different tribes, frequently continue to recognize
in some way their relative status by birth. Thus among the Tswana
no Chief would in the olden days celebrate the firstfruits festival,
or hold an initiation school, until he had received permission to do so
from the Chief of the Huruthse, who was regarded as senior to all the
rest in line of birth. 1 So, too, among the Xhosa there exists a so-called
" paramountcy ", under which the Chiefs of junior tribes recognize
the Chief of the tribe from which they have separated as their superior.
Cases are sometimes referred by them to him for settlement, or carried
to him on appeal from their verdi&s ; and they pay him a certain small
nominal tribute. 2

1 J. Mackenzie, Ten Years North of the Orange River (Edinburgh, 1871), 356.

* Dugmore, in MacLean, 1858, 30; Lestrade, MS., Notes on Xhosa Political Organisation.

POLITICAL INSTITUTIONS 193

Formal requests may also pass from one Chief to another for rain, or
for a military or matrimonial alliance, or for the extradition of fugitives
from justice. Matrimonial alliances between Chiefs are often contrafted
for political purposes, the two tribes henceforth regarding themselves
as related and therefore often lending each other assistance in war.
Where a direft request is made for military assistance, some cattle or a
young girl of the Chief's family generally accompany the request as a
gift to the Chief approached. A request for rain is also paid for in cattle.
Fugitives are as a rule returned only if the two Chiefs are on very
friendly terms, for every Chief gladly welcomes new accessions to
his tribe. If the man is a criminal fleeing from justice he may be sent
back ; but if he has fled because of ill-treatment or other injustice, and
does not wish to go back, he will be protected by the Chief to whom
he has fled. His own Chief cannot follow him up and take him by
force, but is expe&ed to send messengers to ask for his return.

Hostile relations between neighbouring tribes arise most frequently
out of disputes at the boundaries over land or water rights ; cattle
thefts ; interference with visiting subjefts ; or the refusal to hand over
fugitives. Messages will pass between the Chiefs, claiming or refusing
satisfaction, until, unless one of them gives way, armed conflift is
inevitable. Wars were fairly common in the olden days. They were
generally short, sudden affairs. The usual mode of attack was to
surround an enemy village by stealth during the night, rush it at dawn,
kill as many men as possible, capture the cattle, women, and children,
burn the huts, and then beat a hasty retreat home. The captive women,
children, and cattle were taken to the Chief, who, after keeping for
himself what he wanted, distributed the rest among his warriors and
councillors. Such raids and counter-raids would continue until one
of the Chiefs sent formally to beg peace from his more powerful
neighbour. This granted, the weaker tribe would have to pay its
conquerors a regular tribute of cattle and other valuables, in token
of submission and allegiance. It was, however, often allowed to
retain a large measure of autonomy under its own Chief. It might
after a time refuse to continue paying tribute, in which case an attempt
would be made to exaft it by force. If the attack was beaten off, the
people would once again regard themselves as completely independent.
But if defeated, they and their territory were often wholly absorbed
into the tribe of their conquerors. The vanquished Chief might in
such cases be retained as sub-Chief over his own followers and distrift,
although more frequently one or more brothers or sons of the victorious

194 I. SCHAPERA

Chief would be placed in charge with a fairly considerable following of
his own.

Shaka and his imitators not only conducted their wars on a far
larger scale and in much more ruthless manner, but pursued a some-
what different policy in regard to the tribes they conquered. Shaka's
own policy, where he did not completely exterminate or expel a tribe
from its ancestral lands, was to obliterate every vestige of tribal entity
by merging under his direft rule, as integral parts of his tribe, all who
survived his conquests. In this way the Zulu nation was built up
out of the remnants of more than one hundred tribes. 1 His generals
who fled with their armies to carve out their own kingdoms established
still another form of organization, in which the Nguni nucleus became
the ruling aristocracy, while the indigenous tribes were reduced to a
servile condition. This is best seen among the Ndebele of Southern
Rhodesia, where, under Mzilikazi and his successor Lobengula, the
people were divided into three distinft social classes : ate^ansij the
descendants of the original Nguni nucleus ; abenhla, the descendants
of the Sotho and other peoples absorbed into the tribe during its stay
in the Western Transvaal ; and finally the amahole, Shona captives
taken in raids in Southern Rhodesia. The latter were regarded as a slave
class, with whom neither of the others should intermarry. 2

Some reference is necessary here to the system of military organiza-
tion. As a rule the army consists of all the able-bodied men in the tiibe.
But differences exist in regard to the manner in which they are
organized. The Zulu army, with its system of age-regiments dis-
tributed over the tribal territory in special military kraals or barracks,
has already been described in a previous chapter. 3 The kindred Swazi
and Rhodesian Ndebele both have somewhat similar systems of
organization. Among the Cape Nguni, on the other hand, the army
is not organized on an age basis. The members of each clan (isiduko)
together form a single unit within the tribal army, under the command
of their hereditary headman or under an appointed fighting captain.
These clans, among the Xhosa, are grouped into two large divisions,
the first embracing all the clans of royal descent, and the other all
clans of common or alien descent. Each division has a separate com-
mander, but the two are united under the tribal Chief as supreme
commander. 4

1 Cf. Bryant, 1920, passim.

H. M. Hole, "The Rise of the Matabele," Proc. Rkod. Set. Ass. t xii (1912-13), 144.

' Cf. above, chap, iv, p. 8y.

Soga, 1932, 63 ff.

POLITICAL INSTITUTIONS 195

In the Sotho tribes, e.g. the Tswana, age once more forms the basis
of military organization. Each regiment (mophato, age-set) consists
of men simultaneously initiated into adult membership of the tribe.
These regiments are formed at intervals of several years apart. In
time of war each fights as a separate unit in the army, under the com-
mand of a son or brother of the ruling Chief, initiated at the same time
as the other members. The whole army is led by the Chief, or in
his absence by the commander of the senior regiment in order of
formation, assisted by the Chief's induna. There is no formal retire-
ment of any mophato from aftive service, but as a rule the army in
time of war is made up of those regiments still strong in numbers and
whose members are not on the average too old to be aftive fighters.
In time of peace these regiments, as among the Zulu, Swazi, and
Rhodesian Ndebele, have to perform various kinds of labour service
for the Chief. 1

1 Schapera, 1935, MS.

CHAPTER IX
LAW AND JUSTICE

By I. SCHAPERA

INTRODUCTION

Nature and Sources of Bantu Law.

BANTU law in its traditional manifestations is not emoodied in any
formal codes. There are