African political systems

AFRICAN POLITICAL SYSTEMS

AFRICAN POLITICAI
SYSTEMS

Edited by

M. FORTES, M.A., Ph.D.
and

E. E. EVANS-PRITCHARD, M.A., Ph.D.

Published for the

INTERNATIONAL AFRICAN INSTITUTE
by the

OXFORD UNIVERSITY PRESS

LONDON

NEW YORK

TORONTO

Oxford University Press, Amen House, London E.C.4

QW|GOW NEW YORK TORONTO- MELBOURNE WELLINGTON
IfSlblBAY CALCUTTA MADRAS KARACHI KUALA LUMPUR
CAPE TOWN IBADAN NAIROBI ACCRA

First Edition 1940

Reprinted 1941 , 1948 , 1950 , 1955 , 1958 and 1961

^rnima m ureat jantatn
by Jarrold & Sons, Ltd, Norwich

To

C. G. SELIGMAN, M.D., F.R.S.

In token of respect and admiration
for his great contributions to the

study of African ethnology

EDITORS* NOTE

T HIS book 1$ both an experiment in collaborative research
and an attempt to bring into focus one of the major problems
of African sociology. Many dogmatic opinions are held on the
subject of African political organization and are even made use of
in administrative practice; but no one has yet examined this aspect
of African society on a broad, comparative basis. This book will,
we hope, prove the need for and indicate some of the possibilities
of such an investigation. Many of the problems it brings into the
foreground can only be solved by further research; but the oppor¬
tunity for such research is rapidly passing and if it is not grasped
now may be lost for ever.

We regard this book as the first stage of a wider inquiry into the
nature and development of African political systems. In addition
to further research into native political systems, such an inquiry
would include the study of the development of these systems
under the influence of European rule. This problem is not only
sociologically important, it is of pressing importance to the
peoples of Africa and to those who are responsible for governing
them.

We hope this book will be of interest and of use to those who
have the task of administering African peoples. The anthropo¬
logist’s duty is to present the facts and theory of native social
organization as he sees them. It is no light duty; for a thorough
training and laborious field investigations are indispensable to its
proper performance. Whether or not an anthropologist’s findings
can be utilized in the practical tasks of administration must be
left to the decision of administrators themselves.

We are grateful to the International African Institute for
sponsoring the publication of this book. Though several of the
contributors carried out their field research as Fellows of the
Institute, it is in no way responsible for any points of view or
opinions expressed in the book. We have to thank Professor
Radcliffe-Brown for much help and advice, as well as for the
Preface which he has kindly contributed. Our greatest debt is to

v “* AFRICAN POLITICAL SYSTEMS

our fellow workers. Without their collaboration the book would
have been impossible. They will agree, we are sure, that it is an
apposite contribution to social science at the present time.

M. F.

E, E. E.-P.

Oxford,

March, 1940.

NOTE TO THIRD IMPRESSION

No changes have been made in any of the papers included in
this book for the third impression. The Table of Contents has,
however, been brought up to date so as to enable students who
may wish to do so to get into touch with the contributors. Since
the original publication of this book in 1940 several of the con¬
tributors have published more detailed studies of some of the
topics they have dealt with here. These studies are listed at the
end of the present edition for convenience of reference.

September, 1947.

M. F.

E. E. E.-P.

CONTENTS

EDITORS’ NOTE.p. vii

PREFACE. Professor A. R. Radcliffe-Brown, M.A. (Cantab.), Emeritus

Professor of Social Anthropology in the University of Oxford . xi

INTRODUCTION. Dr. M. Fortes and Professor E. E. Evans-Pntchard i

Aims of this Book. A Representative Sample of African. Societies.

Political Philosophy and Comparative Politics. The Two Types
of Political System Studied. Kinship in Political Organization.

The Influence of Demography. The Influence of Mode of Liveli¬
hood. Composite Political Systems and the Conquest-Theory.

The Territorial Aspect. The Balance of Forces in the Political
System. The Incidence and Function of Organized Force. Differ¬
ences in Response to European Rule. The .Mystical Values Asso¬
ciated with Political Office. The Problem of the Limits of the
Political Group.

THE KINGDOM OF THE ZULU OF SOUTH AFRICA. Max

Gluckman. B.A. (Hons.) ( Witwaterstand). D.Phil. (Oxon).
Professor of Social Anthropology in the University of Manchester 25

Historical Introduction. The Zulu King and the State. Status
and Political Power. The Tribes within the Nation. Sanctions on
Authority and' the Stability of the State. The People and their
Leaders. The Period of European Rule. Conclusion.

THE POLITICAL ORGANIZATION OF THE NGWATO OF
BECHUANALAND PROTECTORATE. L Schapera, M.A.

(Cape town), Ph.D., D.Sc.(London), F.R.S.S.Af. Professor of

Social Anthropology in the University of Cape toton ... 5 &

Ethnic Composition and Territorial Constitution. The Adminis¬
trative System. Powers and Authority of the Chief. Rights and
Responsibilities of Chieftainship.

THE POLITICAL SYSTEM OF THE BEMBA TRIBE—NORTH¬
EASTERN RHODESIA. Audrey 1 . Richards , M.A.(Cantah),

Ph.D.(London). Reader in Social Anthropology at the London^

School of Economics and Political Science, University of London 83

Bantu Political Organization—Some General Features. ^ The
Bemba Tribe: Tribal Composition—Social Grouping—Kinship
—Local Grouping—Rank—Other Principles of Social Grouping
—Economic Background —White Administration. Bases of
Authority: The Dogma of Descent—Legal Rules of Descent and
Succession. Functions and Prerogatives of Authority: The Head-
man—the Chief. The Machinery of Government: Administrative
—Military—Judicial—Advisory. . The Integration of the Tribe.
Post-European Changes: New Authorities introduced —Effects
of the 1929 Ordinances.

THE KINGDOM OF ANKOLE IN UGANDA K. Oherg, A.M.,

Ph.D.(Chicago), Escola Livre de Sociologia e Politica, Sdo Paulo,

Brazil . . - • • • • • • Iil

Traditional and Historical Background. Political Status, the
King and the Royal Kraal. Tribute. The Cult of Bagyendanwa.
Succession. Conclusion.

X

CONTENTS

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA.

S. F. Nadel, D.Phil.{Vienna), Ph.D(London). Reader in Anthro¬
pology, King's College, Durham University . . . . 165

Introduction. Demography. Economic System. Settlement.

Political Organization: The Position of the Chief—The ‘Offices
of State 3 " —Succession to Chieftainship—Administration of the
“•Colonies’—Taxation—Jurisdiction—Territorial Rights. _ The
Claim to Autonomy. Social Stratification. Integrative Mechanisms:
Economic Co-operation and Community Life—Tradition and
Mythology—Religion—Conclusions. The Evolution of the Kede
State.

THE POLITICAL ORGANIZATION OF THE BANTU OF

KAVIRONDO. Gunther Wagner, Ph.D.(Hamburg). . . . 197

Introductory. Definition of the Political Unit. The Internal
Political Structure: Enactment of Laws—The Continuity of Law
and Custom—The Perpetuation of Relationships over Periods
during which they are Inoperative—The Transmission of Law
and Custom to Succeeding Generations—The Restoration of
Breaches of the Law'. External Political Structure. The Nature
of Political Authority: The Privileges of Primogeniture—Wealth
—The Quality of being an Omugasa —Reputation as a Warrior—

The Possession of Magico-Religious Virtues—Age.

THE POLITICAL SYSTEM OF THE "FALLENSI OF THE
NORTHERN TERRITORIES OF THE GOLD COAST.'

M. Fortes, M.A.{Cape town), Ph.D.{London). Reader in Social
Anthropology in the University of Oxford . . . 239

The Country and the People. Character of the Political System.

Warfare. The Network of Clanship and the Fundamental Cleavage
of Tale Society. Limiting Factors: Kinship, Local Contiguity and
the Economic System. Authority and Responsibility in the
Lineage System. Tale Religion. Chiefship and Tmdaanaship.

The Complementary Functions of Chiefs and Tmdaanas. Tm¬
daanas and the Wider Community. The Secular Authority of
Chiefs and Tmdaanas .

THE NUER OF THE SOUTHERN SUDAN., E, E. Evans-Pritchard,
M.A.{Oxon.), Ph.D.{London). Professor of Social Anthropology
in the University of Oxford. Fellozv of All Souls

Distribution. Tribal System. Lineage System. Age-set System.

Feuds and other Disputes. Summary.

INDEX.

272

297

LIST OF MAPS

The Distribution of the Peoples Dealt with in this Book
The Country of the Kede .......

Sketch Map of Kavirondo . .

Sketch Map of Taleland.

The Nuer and Neighbouring Peoples . . .

164

196

238

280

PRE'FACE

By Professor A. R. Radcliffe-Brown

Tunc et amicitiam coepemnt jungere habentes
Finitima inter se, nec laedere, nec violare.

Non tamen omnimodis poterat concordia gigni; ^

Sed bona, magnaque pars servabant foedera casti:.

Aut genus humanum jam turn foret omne peremptum,

Nec potuisset adhuc perducere saecla propago.

Lucretius.

T HE comparative study of political institutions, with special
reference to the simpler societies, is an important branch of
social anthropology which has not yet received the attention it

deserves. The publication of this volume affords an opportunity
for a brief statement of the nature of that study as it is conceived

by the Editors and myself.

The task of social anthropology, as a natural science of human

society, is the systematic investigation of the nature of social
institutions. The method of natural science rests always on the
comparison of observed phenomena, and the aim of such com¬
parison is by a careful examination of diversities to discover
underlying uniformities. Applied to human societies the compara¬
tive method used as an instrument for inductive inference will
enable us to discover the universal, essential, characters which
belong to all human societies, past, present, and future. The
progressive achievement of knowledge of this kind must be the
aim of all who believe that a veritable science of human society is
possible and desirable.

But we cannot hope to pass directly from empirical observations

to a knowledge of general sociological laws or principles. The
attempt to proceed by this apparently easy method was what
Bacon so rightly denounced a § leading only to a false appearance
of knowledge . 1 The immense diversity of forms of human society
must first be reduced to order by some sort of classification. By
comparing societies one with another we have to discriminate and
define different types. Thus the Australian aborigines were

1 Novum Organum, I, civ.

AFRICAN POLITICAL SYSTEMS

xii

divided into some hundreds of separate tribes, each with Its owe
language, organization, customs, and beliefs; but an examination
of a sufficient sample shows that beneath the specific diversities
there are such general similarities that we can constitute and
describe In general terms an Australian type . 1 The type Is of
■course an abstraction just as ‘carnivore’ ox ‘ungulate’ Is an abstrac¬
tion, but it is an abstraction only a little way removed from the
concrete reality. When a number of such types have been
adequately defined they in turn can be compared one with another
and a further step in abstraction can be made. By such a process,
obviously requiring the labour of many students over many years,
we may reach classifications and abstract concepts more precisely
defined and more exactly representing empirical reality than the
concepts indicated by such phrases as ‘primitive society 5 , ‘feudal
society’, ‘capitalist society’, that occur so abundantly in
contemporary writing.

In attempting to classify human societies, difficulties are met
with of a kind that do not exist in other sciences, such as zoology
or chemistry. Two societies or two types may resemble each other
in one aspect of the total social system and differ in another. It
is therefore necessary to compare societies with reference to some
particular aspect or part of the whole social system, with reference,
for example, to the economic system or the political system or the
kinship system. Thus the present volume presents materials for
the comparison of certain African societies with reference to their
political organization alone.

This, of course, involves making an abstraction of a different
kind. For in any social system the political institutions, the
economic institutions, the kinship organization, and the ritual
life are intimately related and interdependent. In science there
are right and wrong ways of making abstractions; the right ways
are profitable in that they lead us to important additions to our
knowledge ; the wrong ways are not merely unprofitable, but are
sometimes obstructive. If we are to study political institutions in
abstraction from other features of social systems we need to make
sure that our definition of ‘political’ is such as to mark off a class
of phenomena which can profitably be made the subject of
separate theoretical treatment

The successful use of the comparative method depends, not
1 Radcliffe-Brown, Social Organization of Australian Tribes.

Xlll

■ PREFACE

only upon the quantity and quality of the factual material at our
disposal, but also upon the apparatus of concepts and hypotheses
which guide our investigations. The difficulty in science is not
in finding answers to questions once they have been propounded,
but in fin din g out what questions to ask. In a natural science ot
society the comparative method takes the place of the experiments
method in other sciences and what Claude-Bemard said of the
latter is equally true of the former. ‘The experimental method
cannot give new and fruitful ideas to men who have none; it can
serve only to guide the ideas of men who have them, to direct their
ideas and to develop them so as to get the best possible results.
As only what has been sown in the ground will ever grow m it, so
nothing will be developed by the experimental method except the
ideas submitted to it. The method itself gives birth to nothing.
Certain philosophers have made the mistake of according too
much power to method along these lines. _ ,

The factual material available for a comparative study ot t e
political institutions of the simpler societies is inadequate both
in quantity and quality. It is to be hoped that the publication of
the essays contained in this volume may stimulate other anthro¬
pologists to give us similar descriptive studies. The quality ot
descriptive data, their value for comparative study, depends to a
considerable 7 extent on how the observer understands the
theoretical problems for the solution of which the data he collects
are relevant. In science, observation and the selection of what to
record need to be guided by theory. In the study of the simpler
societies the anthropologist finds that the concepts and theories
of political philosophers or economists are unserviceable or
insufficient. They have been elaborated in reference to societies of
a limited number of types. In their place, the social anthro¬
pologist has to make for himself theories and concepts 'wffich wi
be universally applicable to all human societies, and, guided by
these, carry out his work of observation and comparison.

In some regions of Africa, it is easy to define what may be
called the ‘political society 5 . This is so for the Ngwato, e
Bemba, and Ankole, where we find a tribe or kingdom ruled over
by a chief or king. But the difficulty that is presented ui other
regions is well illustrated by the discussion m Dr Wagners
essay on the Bantu Kavirondo tribes.’ Something of the same

ip. IQQ ff-

xiv • AFRICAN POLITICAL SYSTEMS

difficulty occurs also with the Tallensi and Nuer described in this
volume and with many other societies in various parts of the
world. It would doubtless be possible to find a definition of the
word ‘state’ such that we could say that certain African societies,
such as Ankole or the Bemba, are states, while others are stateless
societies. This does not help us, however, to solve our problems.

Every human society has some sort of territorial structure.
We can find clearly-defined local communities the smallest of
which are linked together in a larger s'ociety, of which they are
segments. This territorial structure provides the framework, not
only for the political organization, whatever it may be, but for
other forms of social organization also, such as the economic,
for example. The system of local aggregation and segregation,
as such, has nothing specifically political about it; it is the basis
of all social life. To try to distinguish, as Maine and Morgan did,
between societies based on kinship (or, more strictly, on lineage)
and societies based on occupation of a common territory or
• locality, and to regard the former as more ‘primitive’ than the
latter, leads only to confusion.

In studying political organization, we have to deal with the
maintenance or establishment of social order, within a territorial
framework, by the organized exercise of coercive authority
through the use, or the possibility of use, of physical force. In
well-organized states, the police and the army are the instruments
by which coercion is exercised. Within the state, the social order,
whatever it may be, is maintained by the punishment of those
who offend against the laws and by the armed suppression of
revolt. Externally the state stands ready to use armed force
against other states, either to maintain the existing order or to
create a new one.

In dealing with political systems, therefore, we are dealing with
law, on the one hand, and with war, on the other. But there are
certain institutions, such as regulated vengeance, which come
between the two. Let us first consider law, and within the field
of law the machinery of repressive justice. Within a locally-
defined community, an individual may commit some act or adopt
some mode of behaviour which constitutes some sort of attack
upon or offence against the community itself as a whole, and
thereupon the offending person may be put to death or excluded
from the community or in some way made to suffer. In simple

XV

PREFACE

societies the actions which are thus repressed, and which are
therefore, in those societies, crimes or public delicts, are mos
commonly various forms of sacrilege, incest—which is_ itsell
generally conceived as a kind of sacrilege—'witchcraft m the
sense of the exercise of evil magic against members of the same
community, and sometimes the crime of being a bad lot, i.e.
habitually failing to observe the customs of the community.

Dr. Wagner, in his essay on the Bantu Kavirondo, describes
how offenders could be expelled from their group or could be pu
to death by what he speaks of as lynching, and writes 1 . uc group
action in the face of threatened danger, taken spontaneously, i.e.
without a hearing of the case and often on the spur of the moment,
is clearly not the same as institutionalized jurisdiction of the tnba
society through recognized judicial authorities. ut 14
highly likely that if such actions could have been carefully observ
it would have been found that they were directed by leaders who
had some .measure of recognized authority., In the tngo e o ^
Kamba and Kikuyu and in the injoget of the Kipsigis and Nand
where individuals who had offended against the community were
put to death or otherwise punished, this was done by an orderly

procedure directed by men in authority.® , - ,

F My own view is that in collective -actions of this kind, in which
it may be said that the community judges and the commum y
inflicts punishment, we may see the embryonic form of cnram
law. That there is often no trial results from the fact tha
offence is often patent, well known to all the community. Other¬
wise the relatives and friends of the accused would come to his
defence and the procedure would be checked. If there is doubt
then, in Africa, recourse may be had to some form o
oath. It would be a serious mistake, I believe, to accept D.
Wagner’s view and regard actions of this kind as f^menul^
the same sort of thing as actions of retaliation by a person who has
suffered injury in his rights against the person ^sponsible for the
injury. The punitive action is to be regarded as the direct expres

sion of public sentiment. , ,

Within small communities there may be little or no need fo
penal sanctions. Good behaviour may be to a great extent the

AFRICAN POLITICAL SYSTEMS

result of habit, of the conditioning of the individual by his early
upbringing. In addition, there are two other kinds of sanctions.
There is first the sanction of moral coercion as distinguished from
physical coercion; the individual who does wrong is subjected to

open expressions of reprobation or ridicule by his fellows and thus
is shamed. What is effective here is the direct expression of public
sentiment. When a person whose behaviour is unsatisfactory is
subjected to some sort of boycott we have a condition intermediate
between the moral and satirical sanctions and the penal sanctions
proper.

Secondly, there are the various kinds of ritual or supernatural
sanction. The most direct of these is constituted by the unques¬
tioned belief that certain actions bring misfortune upon the person
who is guilty of them. For us as Christians the expected misfor¬
tune is eternal torment in the fires of Hell; for an African it is
most commonly sickness or death. In any particular instance, the
mode of behaviour which is a failure to observe ritual obligations
may or may not be also subject to a moral sanction; it may be
reprehensible or it may be merely foolish; in the former case it is
a sin, In the latter an unlucky act or failure to act. In other words,
in the case of sin there is a moral sanction of reprobation added
to the belief that the sin will lead to misfortune for the sinner.

When a person has committeed a ritual offence, his own concern
if it is a matter of luck, or both that and public sentiment if it is
one of sin, will induce him to perform some ritual act of expiation
or purification by which the effects of his misdeed are believed
to be obviated. In some societies the sinner,must perform a
penance, which may be regarded as a self-inflicted punishment.

But in some instances it may be believed that the effects of the
sin will fall not only upon the sinner, but upon the whole com¬
munity, or that the whole community is polluted by the sin; and
the offending person may be put to death or driven out of the com¬
munity as a collective act of expiation. Here we come back again
to the penal sanction. Thus in Ashanti crimes such as incest or
murder or sacrilege are sins—are conceived as offences against
the gods—which bring misfortune upon the whole country, so that
the criminal must be put to death in order that the misfortune
may be avoided.

The kinds of belief which underlie the ritual or supernatural
sanctions may provide a basis for what may be called indirect

PREFACE xvu

penal sanctions. Thus in some African tribes we find a regular
practice of imprecation against wrongdoers. A person who has
committed an offence, whether it is or is not known who he is,
may be officially cursed by the elders or by persons having
authority and power to act in this way. The curse * normally
accompanied by some ritual or magical act through which it s
effective. It is believed that the guilty person will fall sick and die
unless the curse is removed.* Again, in many African societies a
person who is accused or suspected of witchcraft or some other
offence may be compelled to take an oath or submit to an ordeal,
the belief being that if he is guilty he will fall sick and die. _

Thus the rudiments of what in the more complex societies is
the organized institution of criminal justice are to be found in
these recognized procedures by which action is taken y or on
behalf of the body of members of the community, either directly
or by appeal to ritual or supernatural means, to inflict punishment
on an offender or to exclude him from the community. In African
societies the decision to apply a penal sanction may rest with
people in general, with the elders, as in a gerontocracy, with a
limited number of judges or leaders, or with a single chief or king.

There is another side of law, in which'we axe concerned with
conflicts between persons or-groups, or with injuries in ic e y
one person or group upon another, and with action by or o
of the^ community l msolv. the conflict or to necrrre that fac¬
tion is given for the injury. In this field of law also we find a
minimum of organization in some of the simpler societies, the
effective force which controls or limits conflict, or which comp
the wrong doer to give satisfaction to the person he has injured, is
^ public opinion or, as it may perhaps better be calledpublic
sentiment. A person who has suffered wrong or m 3 ur y ^
of another and cannot in any other way obtain faction mg
take forcible action. If public sentiment is on his.side,the conffict
may be resolved in a way that is regarded as just, and so^satisfies
the^community. There is often some conventional recogm
form of procedure by which an injured person may g

public sentiment on his side. Knowledge that such action of seff-
redress is possible is itself often sufficient to restram those who
might otherwise commit injurious acts or to mduce them

i For an example, see Peristiany, Social Institutions of the Kipsigis, PP- 87-8

l 88, 192.

xviii AFRICAN POLITICAL SYSTEMS ,

offer satisfaction when they have been at fault. Public sentiment
may be strong enough to compel the parties to a conflict to settle
the matter by negotiation either directly or by means of a go-
between.

A step towards the establishment of a judicial system is taken in
some societies by the recognition of certain persons as having
authority to act as arbitrators, or to give judgement on the rights
and wrongs of a dispute submitted to them, and suggest a settle¬
ment, though they have no power of physical coercion by which
to enforce that judgement. The authority of the judge or judges
may be conceived in different ways. He or they may be thought of
as the representatives of the community, giving voice to the public
sentiment; or as persons whose wisdom enables them to settle
disputes; or as having special knowledge of right custom; or,
again, as having qualities that may be called ‘religipus/ similar to
those of the priest or the medicine-man; and they may even be
thought to be divinely inspired. Thus the court, if we may call
it so, even where it has no coercive power, always does have
authority.

Recourse may sometimes be had to ritual or supernatural
sanctions in cases of disputed rights. If evidence is so conflicting
that the judge or judges find it impossible to come to a decision
they may resort to the application of an ordeal or oath. If a
person refuses to abide by a decision of the court they may, by
imprecation or the threat thereof, compel him to do so.

In a fully developed court of civil justice, the judge has power to
enforce his judgement by some form of penal sanction. The chief
of the Ngwato tribe, for example, has that power.

In seeking to define the political structure in a simple society,
we have to look for a territorial community which is united by the
rule of law. By that is meant a community throughout which
public sentiment is concerned either with the application of direct
or indirect penal sanctions to any of its own members who offend
in certain ways, or with the settlement of disputes and the provi¬
sion of just satisfaction for injuries within the community itself.
Thus, for the Nuer, Dr. Evans-Pritchard has indicated that
one character by which the political unit—the tribe—is to be
defined is that it is the largest community which considers that
disputes between its members should be settled by arbitration

But we have to recognize that in some societies such a political

PREFACE

xix

community is indeterminate. Thus amongst the Australian
aborigines the independent, autonomous, or, if you will, the
sovereign, group is a local horde or clan which rarely includes
more than ioo members and often as few as thirty. Within this
group, order is maintained by the authority of the old men. But
for the celebration of religious rites a number of such hordes come
together in one camp. In the community so assembled there is
some sort of recognized machinery for dealing with injuries
inflicted by one person or group on another. To give an example:
if a man has had his wife stolen from him and the thief, from
another horde, is present in the assembled camp, the injured man
will make known his wrong by raising a clamour in the recognized,
appropriate way. The public sentiment of the whole assembly,
being appealed to, may compel the offender to submit to having a
spear thrust into his thigh by the injured husband.

The point to be noted is that such assemblies for religious or
ceremonial purposes consist on different occasions of different
collections of hordes. Each assembly constitutes for the time
being a political society. If there is a feud between two of the
constituent hordes, it must either be settled and peace made or it
must be kept in abeyance during the meeting, to break out again
later on. Thus on different occasions a horde belongs temporarily
to different larger temporary political groups. But there is no
definite permanent group of this kind of which a horde^ can be
said to be a part. Conditions similar to this are found in some
parts of Africa—for example, among the Tallensi. 1

There are exceedingly few human societies known to us in
which there is not some form of warfare, and at least a good half of
the history of political development is in one way or another a
history of wars. The comparative study of war as a social institu¬
tion has not yet been undertaken. Amongst the various different
kinds of warfare that can be distinguished, what we may call wars
of conquest have been important in Africa, as they have been in
Europe. When such a war is successful it establishes one people
as conquerors over another who are thus incorporated into a larger
political society, sometimes in an inferior position as a subject

people. . ,. ,

But the institution of war may take a different form m which

i i n f ra p . 230 ff. Where a political structure of this kind exists, it is generally
either ignored or completely misunderstood by colonial administrators.

AFRICAN POLITICAL SYSTEMS

two communities stand in a permanent relation such that war
between them is always a possibility and does from time to time ■
occur, though neither seeks to conquer the other and absorb it as a
conquered people in a larger political unity. In a political system
of which this is true, the occurrence or the possibility of war
gives us the readiest means of defining the political structure.

But it is very difficult to draw an exact dividing line, valid for all
societies, between war and feud. In a single society, as in some
parts of Australia, different kinds of armed conflict are recognized,
from duels between two groups by appointment as to time and
place, in which each side avoids, if possible, killing any of the
enemy, but seek to inflict non-mortal wounds, to £ wars to end war’
which only occur at relatively infrequent intervals and result in

many deaths. .

. There is one kind of feud which needs to be recognized as being

of importance in any attempt to define political structure in some
simple societies, viz. the institution of regulated retaliation for

homicide. Where that exists, when a man is killed,' his relatives,

or the members of his clan or group, are entitled, or in some

societies obliged, by custom to take the life either of his killer or
of a member of his clan or group. Public sentiment regards such

vengeance as just and proper so long as the law of talion is

observed— that is, that the injury inflicted is equivalent to the

injury* suffered, but not greater.

Feuds, or collective actions using force or the threat of force,
of the -kind to which ’this example belongs cannot be regarded as

the same thing as war. The action is limited to obtaining satisfac¬
tion for a particular injury and is controlled by the general public
sentiment of the community in which it takes place. But, on the
other hand, though the idea of justice is involved, such actions
cannot be properly regarded as falling within the sphere of law.

Thus in simple societies the political structure in one of its
aspects, viz. as grouping together individuals within a territorial
framework, which implies, of course, the separation of group
from group within the total system, has to be described in terms
of war, feud, and the exercise of recognized authority in settling
disputes, finding remedies for injuries, and repressing actions
regarded as injuring not certain individuals, but the community
as a whole.

Amongst some writers on comparative politics, there is a

XXI

PREFACE

tendency to concentrate attention too much on what is called the
6 sovereign state 5 . But states are .merely territorial groups within a
larger political system in which their relations are defined by war
or its possibility, treaties, and international law. A political
system of this kind, such as now exists in Europe, of sovereign
nations linked by international relations, is only one type of
political system. Political theory and political practice (including
colonial administration) have often suffered by reason of this
type of system being set up, consciously or unconsciously, as a
norm.

There is a second aspect of political structure. The social
structure of any society includes some differentiation of social
role between persons and between classes of persons. The role
of an individual is the part, he plays in the total social life-
economic, political, religious, &c. In the simplest societies, there
is little more than the very important * differentiation on the basis
of sex and age and the non-institutionalized recognition of leader¬
ship in ritual, in hunting or fishing, in warfare, and so, on, to which
we may add the specialization of the oldest profession in the
world, that of the medicine man. As we pass from the simpler
to the more complex societies we find increasing differentiation
of individual from individual and usually some more or less
definite division of the community into classes.

As political organization develops there is an increasing differen¬
tiation whereby certain persons—chiefs, kings, judges, military
commanders, &c. —have special ■ roles in the social life. Each
such person may be said to hold or fill an office administrative,
judicial, legislative, military, or other. The holder of an office in
this sense is endowed with authority, and to the office there
attach certain duties and also certain rights and privileges.

In Africa it is often hardly possible to separate, even in thought,
political office from ritual or religious office. Thus in some
African societies it may be said that the king is the executive head,
the legislator, the supreme judge, the commander-in-chief of the
army, the chief priest or supreme ritual head, and even perhaps the
principal capitalist- of the whole community. But it is erroneous
to think of him as combining in himself a number of separate
and distinct offices. There is a single office, that of king, and its
various duties and activities, and its rights, prerogatives, and
privileges, make up a single unified whole.

^ AFRICAN POLITICAL SYSTEMS

Besides the development of political office, though not inde¬
pendent of it, we have to take account of the various forms of
political inequality. The simplest example of this is afforded by
differentiation on the basis of sex and age; men usually take far
more part than women, not only in war, but also in maintaining
internal order, and : older men, as a rule, have more authority
than younger ones. . Gerontocracy—rale by elders—-is a form of
political organization that is found in some parts of Africa. In
some tribes of East Africa it is systematized by means of a definite
structure of age-sets and age-grades. Where a society is under a
chief or king, we may find an element of gerontocracy combined
with the monarchical principle.

The Banyankole described in this book are an example of a
division into politically unequal classes. The political power rests
with the pastoral Bahima, who thus constitute a ruling class,

while the agricultural Bairu are in an inferior position. In this
instance and in a number of others there is good reason to believe
that this differentiation into politically superior and inferior classes
is the result of conquest, but it is going far beyond the evidence to
assume that political inequality has in all instances arisen in this
way. In the Banyankole and similar tribes, the political inequality
is associated with other differences and the' class structure is
maintained by the difference in the mode of life of overlords and
subjects and by the absence of intermarriage.

Thus in the comparative study of pdlitical systems we are
concerned with certain special aspects of a total social structure,
meaning by that term both the grouping together of individuals
into territorial or lineage groups and also the differentiation of
individuals by their social role either as individuals or on the basis
of sex and age or by distinctions of social classes.

Social structure is not to be thought of as static, but as a condi¬
tion of equilibrium that only persists by being continually renewed,
like the chemical-physiological homostasis of a living organism.
Events occur which disturb the equilibrium in some way, and a
social reaction follows which tends to restore it. Sometimes a
system may persist relatively unchanged for some length of time;
after each disturbance the reaction restores it to very much what it
was before. But at other times a disturbance of equilibrium may
be such that it and the reaction which follows result in a modifica¬
tion of the system; a new equilibrium is reached which differs

xxiii

PREFACE

from the one previously existing. With a serious disturbance the
process of readjustment may take a long time.

A political system, as we have seen, involves a set of relations
between territorial groups. How the study of this as an equilibrium
system may be approached in African societies is illustrated m the
last two essays of this book, on the Nuer and the Tallensi. Within
the community, the political constitution must also be studied as
an equilibrium system. Dr. Gluckman’s essay on the Zulu shows
how the former system of a balance between the power of the
chief, on the one side, and public sentiment, on the other, has
been replaced by one in which the chief has to maintain as best he
can some sort of balance between the requirements of the European

rulers and the wishes of his people. . .

No attempt can be made to indicate the great variety o
equilibrium situations that can be studied in the political systems
of African peoples. It must suffice to draw attention to the need of
studying political organizations from this point of view.

In writings on political institutions there is a good deal o
discussion about the nature and the origin of the State which is
usually represented as being an entity over and above the human
individuals who make up a society, having as one of its attributes

something called ‘sovereignty’, and sometimesspokenofas

having a will (law being often defined as the will of the State) or as
issuing commands. The State, in this sense, does not exist in the
phenomenal world; it is a fiction of the philosophers What does
exist is an organization, i.e. a collection of mdividual huma
beings connected by a complex system of relations Within that
organization different individuals have different roles and some
are in possession of special power or authority, as chiefs or elders
capable of giving commands which will be obeyed, as legislators o
judges, and so on. There is no such thing as the power of the
State; there are only, in reality, powers of individuals—kings
prime ministers, magistrates, policemen, party bosses, and
voters. The political organization of a society is that aspect ot t
total organization which is concerned with the control an
regulation of the use of physical force. This, it is su § ge ’
provides, for an objective study of human societies by the method
of natural science, the most satisfactory defimtionofthespec
class of social phenomena to the investigation of which this book
is a contribution.

INTRODUCTION

By M. Fortes and E. E. Evans-Pritchard

1. Aims of this Book

/^\NE object we had in initiating this study was to provide a eon-
\~/venient reference book for anthropologists. We also hope that
it will be a contribution to the discipline of comparative politics.
We feel sure that the first object has been attained, for the societies
described are representative of common types of African political
systems and, taken together, they enable a student to appreciate
the great variety of such types. As the sketch-map on p. 2 shows,
the eight'systems described are widely distributed in the con¬
tinent. Most of the forms described are variants of a pattern of
political organization found among contiguous or neighbouring
societies, so that this book covers, by implication, a very large part
of Africa. We are aware that not every type of political system
found in Africa is represented, but we believe, that all the major
principles of African political organization are brought out in these
essays.

Several contributors have described the changes in the political
systems they investigated which have taken place as a result of
European conquest and rule. If we do not emphasize this side of
the subject it is because all contributors are more interested in
anthropological than in administrative problems. We do not wish
to imply, however, that anthropology is indifferent to practical
affairs. The policy of Indirect Rule is now generally accepted in
British Africa. We would suggest that it can only prove advan¬
tageous in the long run if the principles of African political
systems, such as this book deals with, are understood.

II. A Representative Sample of African Societies

Each essay is a condensation of a detailed study of the political
system of a single people undertaken in recent years by the most
advanced methods of field-work by students trained in anthropo¬
logical theory. A degree of brevity that hardly does justice to
some important topics has been necessary for reasons of space.

the distribution of the peoples dealt with in this book

1. Zulu 4. Banyankole

2. Ngwato s. Kede

3. Bemba 6 . Bantu Kavirondo

7 . Tallensi

8. Nuer

INTRODUCTION

3

Each essay furnishes, nevertheless, a useful standard by which
the political systems of other peoples in the same area may be
classified. No such classification is attempted in this book, but
we recognize that a satisfactory comparative study of African
political institutions can only be undertaken after a classification
of the kind has been made. It would then be possible to study a
whole range of adjacent societies in the light of the Ngwato
system, the Tale system, the Ankole system, the Bemba system,
and so on, and, by analysis, to state the chief characters of series
of political systems found in large areas. An analysis of the
results obtained by these comparative studies in fields where a
whole range of societies display many similar characteristics in
their political systems would be more likely to lead to valid
scientific generalizations than comparison between particular
societies belonging to different areas and political types.

We do not mean to suggest that the political systems of societies
which have a high degree of general cultural resemblance are
necessarily of the same type, though on the whole they tend to be.
However, it is well to bear in mind that within a single linguistic
or cultural area we often find political systems which are very
unlike one another in many important features. Conversely, the
same kind of political structures are found in societies of totally
different culture. This can be seen even in the eight societies in
this book. Also, there may be a totally different cultural content
in social processes with identical functions. The function of
ritual ideology in political organization in Africa clearly illustrates
this. Mystical values are attached to political office among the
Bemba, the Banyankole, the Kede, and the Tallensi, but the
symbols and institutions in which these values are expressed are
very different in all four societies. A comparative study of
political systems has to be on an abstract plane where social pro¬
cesses are stripped of their cultural idiom and are reduced to
functional terms. The structural similarities which disparity of
culture conceals are then laid bare and structural dissimilarities
are revealed behind a screen of cultural uniformity. There is
evidently an intrinsic connexion between a people’s culture and
their social organization, but the nature of this connexion is
a major problem in sociology and we cannot emphasize too
much that these two components of social life must not be
confused.

4

AFRICAN POLITICAL SYSTEMS

We believe that the eight societies described will not only give
the student a bird's-eye view of the basic principles of African

political organization, but will also enable him to draw a few,
perhaps elementary, conclusions of a general and theoretical kind.
It must be emphasized, however, that all the contributors have
aimed primarily at giving a concise descriptive account and have
subordinated theoretical speculations to this end. In so far as they
have allowed themselves to draw theoretical conclusions, these
have been largely determined by the view they have taken of what
constitutes political structure. They do not all take the same
view on this matter. In stating our own views we have found it
best to avoid reference to the writings of political philosophers,
and in doing so we feel sure that we have the support of our
contributors.

IIL Political Philosophy and Comparative Politics

We have not found that the theories of political philosophers
have helped us to understand the societies we have studied and
we consider them of little scientific value; for their conclusions
are seldom formulated in terms of observed behaviour or capable
of being tested by this criterion. Political philosophy has chiefly
concerned itself with how men ought to live and what form of
government they ought to have, rather than with what are their
political habits and institutions.

In so far as political philosophers have attempted to understand
existing institutions instead of trying to justify or undermine them,
they have done so in terms of popular psychology or of history.
They have generally had recourse to hypotheses about earlier
stages of human society presumed to be devoid of political institu¬
tions or to display them in a very rudimentary form and have
attempted to reconstruct the process by which .the political
institutions with which they were familiar in their own societies
might have arisen out of these elementary forms of organization.
Political philosophers in modem times have often sought to
substantiate their theories by appeal to the facts ■ of primitive
societies. They cannot be blamed if, in doing so, they have
been led astray, for little anthropological research has been con¬
ducted into primitive political systems compared with research
into other primitive institutions, customs, and beliefs, and still

INTRODUCTION 5

less have comparative studies of them been made . 1 We do not
consider that the origins of primitive institutions can be discovered
and, therefore, we do not think that it is worth while seeking for
them. We speak for all social anthropologists when we say that
a scientific study of political institutions must be inductive and
comparative and aim solely at establishing and explaining the
unif ormities found among them and their interdependencies with
other features of social organization.

IV. The Two Types of Political System Studied

It will be noted that the political systems described in this book
fall into two main categories. One group, which we refer to as
Group A, consists of those societies which have centralized
authority, administrative machinery, and judicial institutions in
short, a government—and in which cleavages of wealth, privilege,
and status correspond to the distribution of power and authority.
This group comprises the Zulu, the Ngwato, the Bemba, the
Banyankole, and the Kede. The other group, which we refer to
as Group B, consists of those societies which lack centralized
authority, administrative machinery, and constituted judicial
institutions—in short which lack government and in which
there are no sharp divisions of rank, status, or wealth. This group
comprises the Logoli, the Tallensi, and the Nuer. Those who
consider that a state should be defined by the presence of govern¬
mental institutions will regard the first group as primitive states
and the second group as stateless societies.

The kind of information related and the kind of problems
discussed in a description of each society have largely depended
on the category to which it belongs. Those who have studied
societies of Group A are mainly concerned to describe govern¬
mental organization. They therefore give an account of the status
of kings and classes, the roles of administrative officials of one
kind or another, the privileges of rank, the differences in wealth
and power, the regulation of tax and tribute, the territorial divi¬
sions of the state and their relation to its central authority, the

i We would except from this stricture Professor R. H. Lowie, though we do
not altogether accept his methods and conclusions. See his works Primitive
Soaetl (1020) and Origin of the State (1927). We are referring only to smthro-
ooloaists The work of the great legal and constitutional historians like Maine,
ViLogrldo^ and Ed Meyer'falls into another category. All students of political
institutions are indebted to their pioneer researches.

6 AFRICAN POLITICAL SYSTEMS

rights of subjects and the obligations of rulers, and the checks on
authority. Those who studied societies of Group B had no such

matters to discuss and were therefore forced to consider what, in
the absence , of explicit forms of government, could be held' to
constitute the political structure of a people. This problem was

simplest among the: Nuer, who have very distinct territorial divi¬
sions. The difficulty was greater for the Logoli and Tallensi, who
have no clear spatially-defined political units.

F. Kinship in Political Organization

One of the outstanding differences between the two groups is
the part played by the. lineage system in political structure. We
must here distinguish between the set of relationships linking the

individual to other persons and to particular social units through
the transient, bilateral family, which we shall call the kinship
system, and the segmentary system of permanent, unilateral descent
groups, which we call the lineage system. Only the latter estab¬
lishes corporate units with political functions. -In both groups of
societies kinship and domestic ties have an important role in the
lives of individuals, but their relation to the political system is of
a secondary order. In the societies of Group A it is the adminis¬
trative organization, in societies of Group B the segmentary
lineage system, which primarily regulates political relations
between territorial segments.

This is clearest among the Ngwato, whose political system
resembles the pattern with which we are familiar in the modem
nation-state. The political unit is essentially a territorial grouping
wherein the plexus of kinship ties serves merely to cement those
already established by membership of the ward, district, and
nation. In societies of this type the state is never the kinship
system writ large, but is organized on totally different principles.
In societies of Group B kinship ties appear to play a more pro¬
minent role in political organization, owing to the close association
of territorial grouping with lineage grouping, but it is still only
a secondary role.

It seems probable to us that three types of political system can
be distinguished. Firstly, there are those very small societies, none
of which are described in this book, in which even the largest
political unit embraces a group of people all of whom are united
to one another by ties of kinship, so that political relations are

INTRODUCTION 7

coterminous with kinship relations and the political structure and
kinship organization are completely fused. Secondly, there axe
societies in which a lineage structure is the framework of the
political system, there being a precise co-ordination between the
two, so that they are consistent with each other, though each
r emains distinct and autonomous in its own sphere. Third y,
there are societies in which an administrative organization is the
framework of the political structure.

The numerical and territorial range of a political system would
vary according to the type to which it belongs. A kinship system
would seem to be incapable of uniting such large numbers of
persons into a single organization for defence and the settlement
of disputes by arbitration as a lineage system and a lineage system
incapable of uniting such numbers as an administrative system.

VI. The Influence of Demography

It is noteworthy that the political unit in the societies with a
state organization is numerically larger than in those without a state
organization. The largest political groups among the lallensi,
Logoli, and Nuer cannot compete in numbers with the quarter to
half million of the Zulu state (in about 1870), the 101,000 of the
Ngwato state, and the 140,000 of the Bemba state. It is true that
the Kede and their subject population are not so populous, but it
must be remembered that they form part of the vast Nupe state.
It is not suggested that a stateless political unit need be very sma
—Nuer political units comprise as many as 45,000 souls nor tha
a political unit with state organization need be very large, but it is
probably true that there is a limit to the size of a population that
can hold together without some kind of centralized government.

Size of population should not be confused with density ot
population. There may be some relation between the degree ot
political development and the size of population, but it would be
incorrect to suppose that governmental institutions are found m
those societies with greatest density. The opposite seems to e
equally likely, judging by our material. The density of the Zulu
is 3 • 5, of the Ngwato a- 5, of the Bemba 3 -75 per square mile while
that of the Nuer is higher and of the Tallensi and Logoli very
much higher. It might be supposed that the dense permanent settle¬
ments of the Tallensi would necessarily lead to the development

8

AFRICAN POLITICAL SYSTEMS

of a centralized form of government, whereas the wide dis¬
persion of shifting villages among the Bemba would be incom¬
patible with centralized rule. The reverse is actually the case. In
addition to the material collected in this book, evidence from
other African societies could be cited to prove that a large popula¬
tion in a political unit and a high degree of political centralization
do not necessarily go together with great density.

VII. The Influence of Mode of Livelihood

The density and distribution of population in an African society
are clearly related to ecological conditions which also affect the
whole mode of livelihood. It is obvious, however, that mere
differences in modes of livelihood do not determine differences

in political structure. The Tallensi and the Bemba are both
agriculturalists, the Tallensi having fixed and the Bemba shifting
cultivation, but they have very different political systems. The
Nuer and Logoli of Group B and the Zulu and Ngwato of Group A
alike practise mixed agriculture and cattle husbandry. In a general
sense, modes of livelihood, together with environmental condi¬
tions, which always impose effective limits on modes of livelihood,
determine the dominant values of the peoples and strongly influ¬
ence their social organizations, including their political systems.
This is evident in the political divisions of the Nuer, in the dis¬
tribution of Kede settlements and the administrative organization
embracing them, and in the class system of the Banyankole.

Most African societies belong to an economic order very
different from ours. Theirs is mainly a subsistence economy with
a rudimentary differentiation of productive labour and with no
machinery for the accumulation of wealth in the form of com¬
mercial or industrial capital. If wealth is accumulated it takes the
form of consumption goods and amenities or is used for the
support of additional dependants. Hence it tends to be rapidly
dissipated again and does not give rise to permanent class divisions.
Distinctions of rank, status, or occupation operate independently
of differences of wealth.

Economic privileges, such as rights to tax, tribute, and labour,
are both the main reward of political power and an essential means
of maintaining it in the political systems of Group A. But there
are counterbalancing economic obligations no less strongly backed
by institutionalized sanctions. It must not be forgotten, also, that

INTRODUCTION 9

those who derive maximum economic benefit from political office
also have the maximum administrative, judicial, and religious

responsibilities. . r i

Compared with the societies of Group A, distinctions of rank

and status are of minor significance in societies of Group
Political office carries no economic privileges, though t e posses
sion of greater than average wealth may be a criterion ol the
qualities or status required for political leadership; for m these
economically homogeneous, equalitarian, and segmentary societies
the attainment of wealth depends either on exceptional personal
qualities or accomplishments, or on superior status m the lineage
system.

VIII. Composite Political Systems and the Conquest Theory

It might be held that societies like the Logoli, Tallensi, and
Nuer, without central government or administrative machinery,
develop into states like the Ngwato, Zulu, and Banyankole as a
result of conquest. Such a development is suggested for the Zulu
and Banyankole. But the history of all the peoples treated in this
book is not well enough known to enable us to declare with any
degree of certainty what course their political development has
taken. The problem must therefore be stated in a different way.
All the societies of Group A appear to be an amalgam of different
peoples, each aware of its unique origin and history, and all except
the Zulu and Bemba are still to-day culturally heterogeneous
Cultural diversity is most marked among the Banyankole and
Kede, but it is also clear among the Ngwato. We may, there¬
fore, ask to what extent cultural heterogeneity m a society is cor¬
related with an administrative system and central authority, ihe
evidence at our disposal in this book suggests that cultural and
economic heterogeneity is associated with a state-like political
structure. Centralized authority and an administrative organiza¬
tion seem to be necessary to accommodate culturally diverse
groups within a single political system, especially if they have
different modes of livelihood. A class or caste system may result
if there are great cultural and, especially, great economic diver¬
gencies. But centralized forms of government are found also with
peoples of homogeneous culture and little economic differentiation
like the Zulu. It is possible that groups of diverse culture are the
more easily welded into a unitary political system without the

IO

AFRICAN POLITICAL SYSTEMS

emergence of classes the closer they are to one another in culture.
A centralized form of government is not necessary to enable
different groups of closely related culture and pursuing the same
mode of livelihood to amalgamate, nor does it necessarily arise

out of the amalgamation. The Nuer have absorbed large numbers
of conquered Dinka, who are a pastoral people like themselves
with a very similar culture. They have incorporated them by
adoption and other ways into their lineage system; but this has
not resulted in a class or caste structure or in a centralized form
of government. Marked divergencies in culture and economic
pursuits are probably incompatible with a segmentary political
system such as that of the Nuer or the Tallensi. We have not the
data to check this. It is clear, however, that a conquest theory of
the primitive state—assuming that the necessary historical
evidence is available—must take into account not only the mode of
conquest and the conditions of contact, but also the similarities
or divergencies in culture and mode of livelihood of conquerors
and conquered and the political institutions they bring with them
into the new combination.

IX. The Territorial Aspect

The territorial aspect of early forms of political organization
was justly emphasized by Maine in Ancient Law and other
scholars have given much attention to it. In all the societies
described in this book the political system has a territorial frame¬
work, but it has a different function in the two types of political
organization. The difference is due to the dominance of an
administrative and judicial apparatus in one type of system and its
absence in the other. In the societies of Group A the administra¬
tive unit is a territorial unit; political rights and obligations are
territorially delimited. A chief is the administrative and judicial
head of a given territorial division, vested often with final economic
and legal control over all the land within his boundaries. Every¬
body living within these boundaries is his subject, and the right to
live in this area can be acquired only by accepting the obligations
of a subject. The head of the state is a territorial ruler.

In the other group of societies there are no territorial units
defined by an administrative system, but the territorial units are
local communities the extent of which corresponds to the range of
a particular set of lineage ties and the bonds of direct co-operation.

INTRODUCTION 11

Political office does not carry with it juridical- rights over a.par¬
ticular, defined stretch of territory and its inhabitants. Member¬
ship of the local community, and the rights and duties that go with
it, are acquired as a rule through genealogical ties, real or fictional.
The lineage principle takes the place of political allegiance, and th$
interrelations of territorial segments are directly co-ordinated with
the interrelations of lineage segments.

Political relations are not simply a reflexion of territorial rela¬
tions. The political system, in its own right, incorporates terri¬
torial relations and invests them with the particular kind of
political significance they have.

X The Balance of Forces in the Political System

A relatively stable political system in Africa presents a balance
between conflicting tendencies and between divergent interests. In
Group A it is a balance between different parts of the administra¬
tive organization. The forces that maintain the supremacy of the
paramount ruler are opposed by the forces that act as a check on his
powers. Institutions such as the regimental organization of the
Zulu, the genealogical restriction of succession to kingship or
chief ship, the appointment by the king of his kinsmen to regional
chiefships, and the mystical sanctions of his office all reinforce the
power of the central authority. But they are counterbalanced by
other institutions, like the king’s council, sacerdotal officials who
have a decisive voice in the king’s investiture, queen mothers
courts, and so forth, which work for the protection of law^ and
custom and the control of centralized power. The regional
devolution of powers and privileges, necessary on account of
difficulties of coirimunication and transport and of other cultural
deficiencies, imposes severe restrictions on a king’s authority. The
balance between central authority and regional autonomy is a very
important element in the political structure. If a king abuses his
power, subordinate chiefs are liable to secede or to lead a revolt
against him. If a subordinate chief seems to be getting too powerful
and independent, the central authority will be supported by other
subordinate chiefs in suppressing him. A king may try to buttress
his authority by playing off rival subordinate chiefs against one
another.

It would be a mistake to regard the scheme of constitutional
checks and balances and the delegation of power and authority to

i2 AFRICAN POLITICAL SYSTEMS

regional chiefs as nothing more than an administrative device.
A general principle of great importance is contained in these
arrangements, which has the effect of giving every section and
every major interest of the society direct or indirect representation
in the conduct of government. Local chiefs represent the central
authority in relation to their districts, but they also represent the
people under them in relation to the central authority. Councillors
and ritual functionaries represent the community’s interest in the
preservation of law and custom and in the observance of the ritual
measures deemed necessary for its well-being. The voice of such
functionaries and delegates is effective in the conduct of govern¬
ment on account of the general principle that power and authority
are distributed. The king’s power and authority are composite.
Their various components are lodged in different offices. Without
the co-operation of those who hold these offices it is extremely
difficult, if not impossible, for the king to obtain his revenue,
assert his judicial and legislative supremacy, or retain his secular
and ritual prestige. Functionaries vested with essential subsidiary
powers and privileges can often sabotage a ruler’s acts if they
disapprove them.

Looked at from another angle, the government of an African

state consists in a balance between power and authority on the one
side and obligation and responsibility on the other. Every one
who holds political office has- responsibilities for the public weal
corresponding to his rights and privileges. The distribution of
political authority provides a machinery by which the various
agents of government can be held to their responsibilities. A chief
or a king has the right to exact tax, tribute, and labour service
from his subjects; he has the corresponding obligation to dispense
justice to them, to ensure their protection from enemies and to
safeguard their general welfare by ritual acts and observances.
The structure of an African state implies that kings and chiefs
rule by consent. A ruler’s subjects are as fully aware of the
duties he owes to them as they are of the duties they owe to
him, and are able to exert pressure to make him discharge these
duties.

We should emphasize here, that we are talking of constitutional
arrangements, not of how they work in practice. Africans recog¬
nize as clearly as we do that power corrupts and that men are
liable to abuse it. In many ways the kind of constitution we find

INTRODUCTION *3

In societies of Group A Is cumbrous and too loosely jointed to
prevent abuse entirely. The native theory of government is often
contradicted by their practice. Both rulers and subjects, actuated
by their private interests, infringe the rules of the constitution.

Though it usually has a form calculated to hold in check any
tendency towards absolute despotism, no African constitution can
prevent a ruler from sometimes becoming a tyrant. The history
of Shaka is an extreme case, but in this and other instances where
the contradiction between theory and practice is too glaring and
the infringement of constitutional rules becomes too grave, popular
disapproval is sure to follow and may even result in a movement of
secession or revolt led by members of the royal family or subordin¬
ate chiefs. This is what happened to Shaka.

It should be remembered that in these states there is only one
theory of government. In the event of rebellion, the aim, and
result, is only to change the personnel of office and never to
abolish it or to substitute for it some new form of government.
When subordinate chiefs, who are often kinsmen of the king, rebel
against him they do so in defence of the values violated by his
malpractices. They have an Interest greater than any other
section of the people in maintaining the kingship. The Ideal con¬
stitutional pattern remains the valid norm, in spite of breaches of
its rules.

A different kind of balance Is found in societies of Group B. It

is an equilibrium between a number of segments, spatially juxta¬
posed and structurally equivalent, which are defined in local and
lineage, and not In administrative terms. Every segment has the
same interests as other segments of a like order. The set of inter¬
segmentary relations that constitutes the political structure is a
balance of opposed local loyalties and of divergent lineage and
ritual ties. Conflict between the interests of administrative
divisions is common in societies like those of Group A. Sub¬
ordinate chiefs and other political functionaries, whose rivalries
are often personal, or due to their relationship to the king or the
ruling aristocracy, often exploit these divergent local loyalties for
their own ends. But the administrative organization canalizes and
provides checks on such Inter-regional dissensions. In the
societies without an administrative organization, divergence of
Interests between the component segments Is intrinsic to the
political structure. Conflicts between local segments necessarily

i 4 AFRICAN POLITICAL SYSTEMS

mean conflicts between lineage segments, since the two are closely
interlocked; and the stabilizing factor is not a superordinate
juridical or military organization, but is simply the sum total of

inter-segment relations.

XI . The Incidence and Function of Organized Force

In our judgement, the most significant characteristic dis¬
tinguishing the centralized, pyramidal, state-like forms of govern¬
ment of the Ngwato, Bemba, &c., from the segmentary political
systems of the Logoli, the Tallensi, and the Nuer is the incidence
and function of organized force in the system. In the former
group of societies, the principal sanction of a ruler's rights and
prerogatives, and of the authority exercised by his subordinate
chiefs, is the command of organized force. This may enable an
African king to rule oppressively for a time, if he is inclined to do
so, but a good ruler uses the armed forces under his control in the
public interest, as an accepted instrument of government—that
is, for the defence of the society as a whole or of any section of it,
for offence against a common enemy, and as a coercive sanction to
enforce the law or respect for the constitution. The king and his
delegates and advisers use organized force with the consent of
their subjects to keep going a political system which the latter take
for granted as the foundation of their social order.

In societies of Group B there is no association, class, or segment
which has a dominant place in the political structure through the
command of greater organized force than is at the disposal of any
of its congeners. If force is resorted to in a dispute between seg¬
ments it will be met with equal force. If one segment defeats
another it does not attempt to establish political dominance over
it; in the absence of an administrative machinery there is, in fact,
no means by which it could do so. In the language of political
philosophy, there is no individual or group in which sovereignty
can be said to rest. In such a system, stability is maintained by an
equilibrium at every line of cleavage and every point of divergent
interests in the social structure. This balance is sustained by a
distribution of the command of force corresponding to the distri¬
bution of like, but competitive, interests amongst the homologous
segments of the society. Whereas a constituted judicial machinery
is possible and is always found in societies of Group A, since it has

INTRODUCTION *5

the backing of organized force, the jural institutions of the Logoli,
the Tallensi and the Nuer rest on the right of self-help.

XII. Differences in Response to European Rule

The distinctions we have noted between the two categories into
which these eight societies fall, especially in the kind of balance
characteristic of each, are very marked in their adjustment to the
rule of colonial governments. Most of these societies have been
conquered or have submitted to European rule from fear of inva¬
sion. They would not acquiesce in it if the threat of force were
withdrawn; and this fact determines the part now played in their
political life by European administrations.

In the societies of Group A, the paramount ruler is prohibited,

by the constraint of.the colonial government, from using the

organized force at his command on his own responsibility. This
has everywhere resulted in diminishing his authority and generally
in • increasing the power and independence of his subordinates.
He no longer rules in his own right, but as the agent of the colonial
government. The pyramidal structure of the state is now main¬
tained by the latter’s taking his place as paramount. If he capitu¬
lates entirely, he may become- a mere puppet of the colonial
government. He loses the support of his people because the
pattern of reciprocal rights and duties which bound him to them
is destroyed. Alternatively, he may be able to safeguard his
former status, to some extent, by openly or covertly leading the
opposition which his people inevitably feel towards alien rule.
Very often he is in the equivocal position of having to reconcile his
contradictory roles as representative of his people against the
colonial government and of the latter against his people. He
becomes the pivot on which the new system swings precariously.
Indirect Rule may be regarded as a policy designed to stabilize
the new political order, with the native paramount ruler in this
dual role, but eliminating the friction it is liable to give rise to. ^

In the societies of Group B, European rule has had the opposite
effect. The colonial government cannot administer through
aggregates of individuals composing political segments, but has to
employ administrative agents. For this purpose it makes use of
any persons who can be assimilated to the stereotyped notion of an
African chief. These agents for the first time have the backing of
force behind their authority, now, moreover, extending into

16 AFRICAN POLITICAL SYSTEMS

spheres for which there is no precedent,. Direct resort to force'in
the form, of self-help in defence of the rights of individuals or of
groups is no longer permitted; for there is now, for thie-first time,
a paramount authority exacting obedience in virtue of superior
force which enables it to establish courts of justice to replace self-
help. This tends to lead to the whole system of mutually balancing
segments collapsing and a bureaucratic European system taking
its place. An organization more like that of a centralized state
comes into being.

XIII. The Mystical Values Associated with Political Office

The sanction of force is not an innovation in African forms of
government. We have stressed the fact that it is one of the main
pillars of the- indigenous type of state. But the sanction of force
on which a European administration depends lies outside the native
political system. It is not used to maintain the values inherent in
that system. In both societies of Group A and those of Group B
European governments can impose their authority; in neither are
they able to establish moral ties with the subject people. For, as we
have seen, in the original native system force is used by a ruler
with the consent of his subjects in the interest of the social order.

An African ruler is not to his people merely a person who can
enforce his will on them. He is the axis of their political relations,
the symbol of their unity and exclusiveness, and the embodiment
of their essential values. He is more than a secular ruler; in that
capacity the European government can to a great extent replace
him. His credentials are mystical and are derived from antiquity.
Where there are no chiefs, the balanced segments which compose
the political structure are vouched for by tradition and myth and
their interrelations are guided by values expressed in mystical
symbols. Into these sacred precincts the European rulers can
never enter. They have no mythical or ritual warranty for their
authority.

# What is the meaning of this aspect of African political organiza¬
tion? African societies are not models of continuous internal
hamrony. Acts of violence, oppression, revolt, civil war, and so
forth, chequer the history of every African state. In societies like
the Logoli, Tallensi, and Nuer the segmentary nature of the social
structure is often most strikingly brought to light by armed con¬
flict between the segments. But if the social system has reached a

INTRODUCTION

17

sufficient' degree of stability, these internal convulsions do not
necessarily wreck it. In fact, they may be the means of reinforcing
it, as we have seen, against the-abuses and infringements of rulers
actuated by'their private interests. In the segmentary societies,
war is not a matter of one segment enforcing its will on another,
but is the way in which segments protect their particular interests
within a field of common interests and values.

There are, in every African society, innumerable ties which
counteract the tendencies towards political fission arising out of the
tensions and cleavages in the social structure. An administrative
organization backed by coercive sanctions, clanship, lineage and
age-set ties, the fine-spun web of kinship— all these unite people
who have.different or even opposed sectional and private interests.
Often also there are common material interests such, as the need
to share pastures or to trade in a common market-place, or com¬
plementary economic pursuits binding different sections, to one
another. Always there are common ritual values, the ideological
superstructure of political organization.

Members of an African society feel their unity and perceive
their common interests in symbols, and it is their attachment to
these symbols which more than anything else gives their society
cohesion and persistence. In the form of myths, fictions, dogmas,
ritual, sacred places and persons, these symbols represent the
unity and exclusiveness of the groups which respect them. They
are regarded, however, not as mere symbols, but as final values in
themselves.

To explain these symbols sociologically, they have to be trans¬
lated into terms of social function and of the social structure which
they serve to maintain. Africans have no objective knowledge of
the forces determining their, social organization and actuating
their sociaTbehaviour. Yet they would be unable to carry on their
collective file if they,, could not' think and feel about the interests
which actuate them, the institutions by means of which they
organize collective action, and the structure of the groups into
which they are organized. Myths, dogmas, ritual beliefs * and
activities make his social system intellectually tangible and
coherent to an African and enable him to think and feel about it.
Furthermore, these sacred symbols, which reflect the social system,
endow it with mystical values which evoke acceptance of the social
order that goes far beyond the obedience exacted by the secular

18 AFRICAN POLITICAL SYSTEMS

sanction of force. The social system is, as it were, removed to a
mystical plane, where it figures as a system of sacred values beyond
criticism or revision. Hence people will overthrow a bad king,
but the kingship is never questioned; hence the wars or feuds
between segments of a society like the Nuer or the Tallensi are
kept within bounds by mystical sanctions. These values are com¬
mon to the whole society, to rulers and ruled alike and to all the
segments and sections of a society.

The African does not see beyond the symbols; it might well be
held that if he understood their objective meaning, they would
lose the power they have over him. This power lies in their
symbolic content, and in their association with the nodal institu¬
tions of the social structure, such as the kingship. Not every kind
of ritual or any sort of mystical ideas can express the values that
hold a society together and focus the loyalty and devotion of its
members on their rulers. If we study the mystical values bound
up with the kingship in any of the societies of Group A, we find
that they refer to fertility, health, prosperity, peace, justice—to
everything, in short, which gives life and happiness to a people.
The African sees these ritual observances as the supreme safe¬
guard of the basic needs of his existence and of the basic relations
that make up his social order—land, cattle, rain, bodily health,
the family, the clan, the state. The mystical values reflect the
general import of the basic elements of existence: the land as the
source of the whole people’s livelihood, physical health as some¬
thing universally desired, the family as the fundamental pro-
creative unit, and so forth. These are the common interests of
the whole society, as the native sees them. These are the themes
of taboos, observances and ceremonies in which, in societies of
Group A, the whole people has a share through its representatives,
and in societies of Group B all the segments participate, since
they are matters of equal moment to all.

We have stressed the fact that the universal aspect of things
like land or fertility are the subjects of common interest in an
African society; for these matters also have another side to them,
as the private interests of individuals and segments of a society.
The productivity of his own land, the welfare and security of his
own family or his own clan, such matters are of daily, practical
concern to every member of an African society; and over such
matters arise the conflicts between sections and factions of the

society. Thus the basic needs of existence and the basic social
relations are, in their pragmatic and utilitarian aspects, as sources
of immediate satisfactions and strivings, the subjects of private
interests; as common interests, they are non-utilitarian and non-
pragmatic, matters of moral value and ideological significance.
The common interests spring from those very private interests

to which they stand in opposition. . ...

To explain the ritual aspect of African political organization in
terms of magical mentality is not enough; and it does not take us
far to say that land, rain, fertility, &c., are ‘sacralized’ because
they are the most vital needs of the community. Such arguments
do not explain why the great ceremonies in which ritual for the
common good is performed are usually on a public scale. They
do not expl^m why the ritual functions we have been describing
should be bound up, always, with pivotal political offices and
should be part of the political theory of an organized society.

Ag ain, it is not enough to dismiss these ritual functions of chief-
ship, kingship, &c., by calling them sanctions of political authority.
Why then are they regarded as among the most stringent respon¬
sibilities of office? Why are they so often distributed amongst a
number of independent functionaries who are thus enabled to
exercise a balancing constraint on one another? It is clear that
they serve, also, as a sanction against the abuse of political power
and as a means of constraining political functionaries to perform
their administrative obligations as well as their religious duties,

lest the common good suffer injury. . .

When, finally, it is stated as an observable descriptive fact that
we are d oling here with institutions that serve to affirm and pro¬
mote political solidarity we must ask why they do so. Why is an
all-embracing administrative machinery or a wide-flung lmeage
system insufficient by itself to achieve this ?

We cannot attempt to deal at length with all these questions.
We have already given overmuch space to them because we
consider them to be of the utmost importance, both from the
theoretical and the practical point of view. The ‘supernatural
aspects of African government are always puzzling and often
exasperating to the European administrator. But a great deal
more of research is needed before we shall be able to understand
fully. The hypothesis we are. making use of is, we feel, a
st imulating starting-point for further research into these matters.

20

AFRICAN POLITICAL SYSTEMS

That part of it which has already been stated is, perhaps, least

controversial. But it is incomplete.

Any item of social behaviour, and therefore any political rela¬
tion, has a utilitarian or pragmatic content. It means that material
goods change hands, are disbursed or acquired, and that the direct
purposes of individuals are achieved. Items of social behaviour
and therefore political relations have also a moral aspect; that is,
they express rights and duties, privileges and obligations, political
sentiments, social ties and cleavages. We see these two aspects
clearly in such acts as paying tribute to a ruler or handing over
blood-cattle in compensation for murder. In political relations,
consequently, we find two types of interests working conjointly,
material interests and moral interests, though they are not sepa¬
rated in this abstract way in native thought. Natives stress the
material components of a political relation and generally state it
in terms of its utilitarian and pragmatic functions.

A particular right or duty or political sentiment occurs as an
item of behaviour of an individual or a small section of an African
society and is enforceable by secular sanctions brought to bear
on these individuals or small sections. But in a politically
organized community a particular right, duty, or sentiment exists
only as an element in a whole body of common, reciprocal, and
mutually balancing rights, duties, and sentiments, • the body of
moral and legal norms. Upon the regularity and order with which
this whole body of interwoven norms is maintained depends the
stability and continuity of the structure of an African society. On
the average, rights must be respected, duties performed, the senti¬
ments binding the members together upheld or else the social
order would be so insecure that the material needs of existence
could no longer be satisfied. Productive labour would come to
a standstill and the society disintegrate. This is the greatest
common interest in any African society, and it is this interest
which the political system, viewed in its entirety, subserves. This,
too, is the ultimate and, we might say, axiomatic set of premisses of
the social order. If they were continually and arbitrarily violated,
the social system would cease to work.

We can sum up this analysis by saying that the material interests
that actuate individuals or groups in an African society operate
in the frame of a body of interconnected moral and legal norms
the order and stability of which is maintained by the political

INTRODUCTION

21

organization. Africans, as we have pointed out, do not analyse
their social system; they live it. They think and feel about it m
terms of values which reflect, in doctrine and symbol, but do not
explain, the forces that really control their social behaviour. Out¬
standing among these values are the mystical values dramatized
in the great public ceremonies and bound up with their key
political institutions. These, we believe, stand for the greatest
common interest of the widest political community to which a
member of a particular African society belongs—that is, for the
whole body of interconnected rights, duties, and sentiments; for
this is what makes the society a single political community. That
is why these mystical values are always associated with pivotal
political offices and are expressed in both the privileges and the
obligations of political office.

Their mystical form is due to the ultimate and axiomatic
character of the body of moral and legal norms which could not
be kept in being, as a body, by secular sanctions. Periodical
ceremonies are necessary to affirm and consolidate these values
because, in the ordinary course of events, people are preoccupied
with sectional and private interests and are apt to lose sight ot the
common interest and of their political interdependence. Lastly,
their symbolic content reflects the basic needs of existence and
the basic social relations because these are the most concrete and
tangible elements of all social and political relations. The visible
test of how well a given body of rights, duties, and sentiments is
being maintained and is working is to be found m the level ot
security and success with which the basic needs of existence are
satisfied and the basic social relations sustained.

It is an interesting fact that under European rule African kings
retain their ‘ritual functions’ long after most of the secular
authority which these are said to sanction is lost. Nor are the
mystical values of political office entirely obliterated by a change
of religion to Christianity or Islam. As long as the kagskp
endures as the axis of a body of moral and legal norms holding
a people together in a political community, it will, most probably,
continue to be the focus of mystical values. .... .

It is easy to see a connexion between kingship and the interests
and solidarity of the whole community in a state with highly
centralized authority. In societies lacking centralized government
social values cannot be symbolized by a single person, but are

22

AFRICAN POLITICAL SYSTEMS

distributed at cardinal points of the social structure. Here we
find myths, dogmas, ritual ceremonies, mystical powers, &c.,
associated with segments and defining and serving to maintain
the relationship between them. Periodic ceremonies emphasizing
the solidarity of segments, and between segments, as against
sectional interests within* these groups, are the rule among the
Tallensi and Logoli no less than among the Bemba and Kede.
Among the Nuer, the leopard-skin chief, a sacred personage
associated with the fertility of the earth* is the medium through
whom feuds are settled and, hence, inter-segment relations
regulated. The difference between these societies of Group B and
those of Group A lies in the fact that there is no person who
represents the political unity of the people, such unity being
lacking, and there may be no person who represents the unity of
segments of the people. Ritual powers and responsibility are
distributed in conformity with the highly segmentary structure of
the society.

XIV. The Problem of the Limits of the Political Group

We conclude by emphasizing two points of very great im¬
portance which are often overlooked. However one may define
political units or groups, they cannot be treated in isolation, for
they always form part of a larger social system. Thus, to take an
extreme example, the localized lineages of the Tallensi overlap one
another like a series of intersecting circles, so that it is impossible to
state clearly where the lines of political cleavage run. These over¬
lapping fields of political relations stretch almost indefinitely, so
that there is a kind of interlocking even of neighbouring peoples,
and while we can see that this people is distinct from that, it is not
easy to say at what point, culturally or politically, one is justified in
regarding them as distinct units. Among the Nuer, political
demarcation is simpler, but even here there is, between segments
of a political unit, the same kind of structural relationship as there
is between this unit and another unit of the same order. Hence the
designation of autonomous political groups is always to some
extent an arbitrary matter. This is more noticeable among the
societies of Group B, but among those of Group A also there is an
interdependence between the political group described and neigh¬
bouring political groups and a certain overlapping between them.
The Ngwato have a segmentary relationship to other Tswana

INTRODUCTION

23

tribes which in many respects is of the same order as that between
divisions of the Ngwatd themselves. The same is true of the other
societies with centralized governments.

This overlapping and interlocking of societies is largely due to
the fact that the point at which political relations, narrowly defined
in terms of military action and legal sanctions, end is not the point
at which all social relations cease. The social structure of a people
stretches beyond their political system, so defined, for there are
always social relations of one kind or another between peoples of
different autonomous political groups. Clans, age-sets, ritual
associations, relations of affinity and of trade, and social relations
of other kinds unite people of different political units. Common
language or closely related languages, similar customs and beliefs,
and so on, also unite them. Hence a strong feeling of community
may exist between groups which do not acknowledge a single
ruler or unite for specific political purposes. Community of
language and culture, as we have indicated, does not necessarily
give rise to political unity, any more than linguistic and cultural
dissimilarity prevents political unity.

Herein lies a problem of world importance: what is the relation
of political structure to the whole social structure? Everywhere
in Africa social ties of one kind, or another tend to draw together
peoples who are politically separated and political ties appear to be
dominant whenever there is conflict between them and other social
des. The solution of this problem would seem to lie in a more
detailed investigation of the nature of political values and of the
symbols in which they are expressed. Bonds of utilitarian interest
between individuals and between groups are not as strong as the
bonds implied in common attachment to mystical symbols. It is
precisely the greater solidarity, based on these bonds, which
generally gives political groups their dominance over social groups
of other kinds.

THE KINGDOM OF THE ZULU OF SOUTH AFRICA 3
By Max Gluckman

/. Historical Introduction

I DESCRIBE Zulu political organization at two periods of
Zulu, history—under King Mpande and. to-day under
European rule. Zulu history has been well described by Bryant
and Gibson, and I here give only ■ a bare outline which can
be filled in by referring to their books. 1 2 I have used historical
records partly to illustrate the functioning of the organization in
each period and partly to discuss changes in the nature of the
organization.

The Nguni family of Bantu-speaking people who later formed

the Zulu nation migrated into south-eastern Africa about the
middle of the fifteenth century. They were pastoralists practising
a shifting cultivation.. They lived in scattered homesteads occupied
by male agnates and their families; a number of these homesteads
were united under a chief, the heir of their senior line, into a tribe.
Exogamous patrilineal clans (men and women of common descent
bearing a common name) tended to be local units and the cores of
tribes. A tribe was divided into sections under brothers of the
chief and as a result of a quarrel a section might migrate and
establish itself as an independent clan and tribe. There was also
absorption of strangers into a tribe. Cattle raids were frequent,
but there were no wars of conquest. By 1775 the motives for war
changed, possibly owing to pressure of population. Certain tribes
conquered their neighbours and small kingdoms emerged which

1 The information contained in this article was largely collected during
fourteen months* work in Zululand (1936-8), financed by the National Bureau
of Educational and Social Research of the Union of South Africa (Carnegie
Fund). I wish to thank the Bureau for its grant. I have also used many books,
dispatches, and reports about Zululand in the last 100 years. For a biblio¬
graphy of these, and an account of Zulu society, see E. J. Krige’s Social Systems
of the Zulu (Longmans, 1936).

2 A. T. Bryant, Olden Times in Zululand and Natal (Longmans, 1938); J. Y.

Gibson, The Story of the Zulus (Longmans, 1911). The account of the Zulu
nation in this article is reconstructed from histories, contemporaneous records,
and my questionings of old men.

26

AFRICAN POLITICAL SYSTEMS

came into conflict. In this struggle Shaka, head of the Zulu tribe,
was victorious; by his personal character and military strategy, he
made himself, in ten years, master of what is now Zululand and
Natal, 1 and his troops were campaigning far beyond his boun¬
daries. He organized a nation out of all the tribes he had sub¬
jected. His chief interest was in the army and he made whole-time
warriors of his men; he developed the idea of regiments formed of
men of the same age, and quartered them, for most of the year, in
large barracks built in different parts of his country. They trained
there for war, herded the king’s cattle and worked his fields. The
men were forbidden to marry till the king gave them permission,
as a regiment, to marry into a certain age-regiment of girls.
Shaka’s rule was tyrannous and he fought a war every year;
therefore, when in 1828 he was assassinated by his brother,
Dingane, the people gladly accepted Dingane as king.

During Shaka’s life English traders settled at Port Natal on
friendly terms with the Zulu. Later the Boers entered Natal,
defeated the Zulu in 1838, and confined them north of the Tugela
River. Dingane’s rule was also tyrannous and his people began to
turn from him to his brother, Mpande. Dingane plotted to kill
Mpande, who fled with his followers to the Boers in Natal; from
there he attacked and routed Dingane and became king. The Zulu
now entered on a period of comparative peace, for Mpande only
occasionally raided the Swazi and Tembe (Thonga); to south and
west were European states and the strongly entrenched Basuto.
However, during his reign two of his sons fought for his heirship;
Cetshwayo was victorious and he became king when Mpande died
in 1872.

In 1880 the British defeated the Zulu, deposed Cetshwayo and

divided the nation into thirteen kingdoms. Three years later they
tried to reinstate Cetshwayo; for various reasons civil war broke
out between the Usuthu (the Royal) section of the nation and tribes
ruled, under the King, by the Mandlakazi Zulu house, which was
united to the royal house in Mpande’s grandfather. The king died
but his son, Dinuzulu, with Boer help defeated the rebels who
fled to the British. In 1887 the British established a magistracy in

1 An area of some 80,000 square miles, occupied, according to Bryant’s
estimate, by about 100,000 people. I think this figure is too low. It may be
noted that tribes fleeing from Shaka established the Matabele, Shangana, and
Nguni nations.

THE KINGDOM OF THE ZULU

27

Zululand and restored the Mandlakazi to their homes. Dinuzulu
resisted, was defeated and exiled. The Zulu were divided into
many tribes and white rule was firmly established. Dinuzulu was
later appointed chief over a small tribe (the Usuthu), but was
again exiled after the Bambada Rebellion in 1906. He died in
exile and his heir was appointed Usuthu chief; on his death he was
succeeded by his full-brother as regent. The government has

This is a sketch of the territorial organization 01 tfce Zulu nation under King
Mpande. It is presented only as a plan, and not as a map.

The shaded area is the King’s, containing his capital, other royal homesteads,
and military barracks (which are also royal homesteads). Numerals show tribal
areas: there were many more than fourteen. .

In tribe 3, of which Q is the capital, small letters show wards under indunas.
In tribe 1 X, Y, and Z are the homesteads of the chief and two of his impor¬
tant brothers: the men of the tribe are attached to these homesteads to constitute

the military divisions of the tribe. ...

In tribe 11, the dots represent homesteads spread over the comitry. Equlusim
and Ekubuseni are royal homesteads which were heads of national army dm-
sions, though they lay outside the King’s area.

28

AFRICAN POLITICAL SYSTEMS

passed from Britain to Natal and in 1910 to the Union of South
Africa.

II . The Zulu King and the State

Certain kinship groupings persisted through the devastating
wars and the great change in political organization of Shaka’s and
Dingane’s reigns. The clans had disappeared as units, and
members of a single clan might be widely dispersed over Zululand:
they retained their clan-name and their respect for the head of their
senior line. 1 Pockets of clansmen were, however, still to be found
in various parts. The important kinship groups which were the
basis of social organization were still formed by the inhabitants of
separate homesteads. At the head of a homestead was the senior
male by descent of the group. Nearby there might be found
homesteads of men of the same clan and they all acknowledged the
heir of their senior line (the lineage-head) as their head. Some
distance away there would perhaps be clan-kinsmen, living under
a different political authority, but recognized as part of the group
and therefore entitled to take part in affairs affecting it. Among
these local agnatic groups there were often homesteads of other
relatives by marriage or matrilineal relationship: then came a
stretch of country occupied by members of another group, simi¬
larly constituted. Strangers might attach themselves to an
important man, as his servants or dependants, and would be
absorbed with their relatives into his kinship group as ‘quasi-
kinsmen’ ; they retained their clan-name, but could not marry into
their superior’s own lineage, though they could marry into his clan.
The second important change in Zulu family life was caused by
the younger men having to serve at the king’s military barracks,
which kept them from home most of the year. In the homesteads
the older men and the boys herded the cattle and the women
worked the fields. Each homestead had its own fields and cattle¬
fold. A demographic survey would show the homesteads scattered
at some distance apart (a few hundred yards to a mile or two)
along the hills which, intersected by deep bush-filled valleys,
characterized the interior of Zululand. The fields were mostly

1 Men and women with the same clan-name could not marry one another. No
new clans have been formed in the period since the clans ceased to be local, politi-
cal units, as in the past a chief desiring to marry a woman of his clan would split
off her lineage and make of it a separate clan. Dinuzulu attempted unsuccessfully
to form a new clan of a Mandlakazi lineage into which he married.

THE KINGDOM OF THE ZULU

29

along the ridges and the banks of streams; the low valleys, unin¬
habited because of fever, were winter-grazing and hunting'grounds.
The coastal tribes lived, similarly distributed, on the malarial
sandy plain between the hills and the sea. Communication be¬
tween different parts of Zululand was fairly easy; men went from
all parts to the King’s barracks and marriage between members of
widely separated homesteads was common.

The Zulu nation thus ponsisted of members of some hundreds of
clans, united by their allegiance to the king. The people belonged
to the king and he therefore took the fine in cases of assault or
murder. In the earlier period of Nguni history, political alle¬
giance tended to coincide with kinship affiliation. Thus the Zulu
tribe (abakwazulu) consisted originally largely of descendants of
Zulu, a junior son of Malandela, as distinguished from the Qwabe
tribe, the descendants of Qwabe, the senior son of Malandela.
To-day the term abakwazulu still means the members of the Zulu
clan, but it has also the wider meaning of all the people who pay
allegiance to the Zulu king. Collectively, whatever their clan
names, they are politely addressed as ‘Zulu’. Political and kinship
affiliation came to be. distinct also in the smaller political groups
into which the nation was divided. These were composed of
members of many clans, though they might have a core of kins¬
men: members of a single clan might be found in many political
groups. While the kinship basis of political groups disappeared, the
new ones which emerged were described in kinship terms, for any
political officer was spoken of as the father of his people, and his
relationship to them was conceived to be similar to that of a father
and his children. The territory of king or chief may be referred to
as umzi kaMpande (the homestead of Mpande) or umzi kaZibebu
(the homestead of Zibebu), as umzi kabatti is the family homestead
of So-and-So. The children of the king are. not supposed to refer
to him as ‘father’, for, ‘is not the king father of his people, not of
his family only’.

The king also owned the land. All who came to live in Zululand
had to acknowledge his sovereignty. Abakwazulu has too the
meaning of the people of Zululand (Kwazulu) and the Zulu word

izwe means nation, tribe, or country. 1 The same rule applied to

1 It must be noted that the Zulu form of describing the clan is locative:

abakwazulu, ‘the people of the Zulu clan’, literally, ‘the people of the place
of Zulu’, not genitive, abakaZulu ,. ‘the descendants of Zulu’.

3© AFRICAN POLITICAL SYSTEMS

the smaller political groupings and to find out who a man’s chief is,
one asks either ‘Who is your chief?’ or ‘Of whose district are you?’
The relation of the political unit to land may be defined: any one
coming on to land belonging to a political authority became
subject to that authority, and all his subjects were entitled to land
in his area.

The Zulu nation may therefore be defined as a group of people
owing allegiance to a common head (the king) and occupying a
defined territory. They combined under the king to attack or
defend themselves against outside groups. In addition to con¬
trolling relations with other Bantu-speaking peoples and the
Europeans, the king exercised judicial, administrative, and legis¬
lative authority over his people, with power to enforce his decisions.
He performed religious ceremonies and magical acts on behalf of
the nation. All the tribes which made up the nation spoke dialects
of the same language and had a common culture. •

The kings Mpande and Cetshwayo had no subjects directly
under their control. They lived in a tract of land occupied only
by royal homesteads and military barracks 1 : outside of this tract
Zululand was divided into a large number of political groups. The
inhabitants of the largest divisions of the nation I call ‘tribes’, and
their heads I call ‘chiefs’. The tribes were divided into smaller
groups (wards) under relatives of the chief or men of other clans
(indunas), responsible to the chiefs.

The king was approached with ceremonious salutations and
titles of respect which, say the Zulu, increased his prestige. He
was addressed as the nation. What tradition and history was
common to all the Zulu had to be told in the names of the Zulu
kings and it was largely their common sentiment about the king
and his predecessors which united all Zulu as members of the
nation. At the great first-fruits ceremonies and in war-rites, the
king was strengthened and cleansed in the name of the nation. He
possessed certain objects, inherited from his ancestors, and the
welfare of the country was held to depend on them. This cere¬
monial position of the king was backed by his ancestral spirits.
They were supposed to care for the whole of Zululand, and in the
interests of the nation the king had to appeal to them in drought,
war, and at the planting and first-fruits seasons. They were

1 All military barracks were royal homesteads. They were built on the plan
of ordinary homesteads but were very large, housing some thousands of men.

THE KINGDOM OF . THE ZULU 31

praised against the ancestors of other -kings*' The king was in
charge of, and responsible for, all national magic. Shaka expelled
all rainmakers from his kingdom, saying only he could control the
heavens* The king possessed important therapeutic medicines
with which he would treat all his ailing important people. All
skilled leeches had to teach the'king their cures* Finally, when
people died and a person was accused of killing them by sorcery,
no sentence was supposed to be executed unless the king’s witch¬
doctors confirmed the verdict. These religious and magical duties
of the king, in performing which he was assisted by special,
hereditary magicians, were vested in the office of kingship; though a
king might be killed, his successor took over these duties, and the
spirits of tyrants were even supposed to become a source of good
to the people who had slain, them.

The ritual of these national ceremonies was similar to that of
tribal ceremonies of pre-Shaka times, but Shaka militarized them
and the men paraded for them in regiments. The ceremonies were
chiefly designed to, strengthen the Zulu at the expense of other
people, who were symbolically attacked in them. It was this
military orientation of Zulu culture under the king which largely
unified his people. A man was called isihlangu smkosi (war-shield
of the king). The dominant values of Zulu life were those of the
warrior, and they were satisfied in service at the king’s barracks and in
'his wars. To-day old men talking of the kings get excited and joyful,
chanting the king’s songs and dances, and all Zulu tend, in conver¬
sation, to slip into tales of the king’s wars and affairs at his court.

The regiments belonged to the king alone. They lived in
barracks concentrated about the capital; the chiefs had no control
over the regiments and assembled their own people in territorial,
not age, divisions. This organization probably persisted from the
period before Shaka 1 began to form ‘age regiments’^ In those
times the chief of a tribe seems to have assembled his army in
divisions which he constituted by attaching the men of.-certain
areas to certain of his important homesteads. The tribes within
the. Zulu nation were (and to-day still are) organized for fighting
and hunting on this basis. The king alone could summon the age
regiments. The nation also was divided for military purposes in
the same way as a tribe was divided. For the king attached

1 The idea of age regiments was originally developed on the basis of old
age-grades by a chief, Dingiswayo, who was Shaka’s patron.

AFRICAN POLITICAL SYSTEMS

3 *

certain groups of tribes to certain of his royal homesteads. 1 call

each of these divisions and the royal homestead {ikhanda, head)
to which it was attached a ‘head’. Most of the homesteads which
were ‘heads’ were in the king’s area; one or two lay outside it.
Some were also barracks in which were quartered particular regi¬
ments. The division into ‘heads’ was not purely territorial, for
once a. man was attached to a ‘head’ he could not change his
attachment even if he moved into a tribal area attached to a dif¬
ferent ‘head’. His sons inherited his attachment; when they were
ready to be enrolled as soldiers, they went to the ‘head’ to which
their father belonged, and later from all the ‘heads’ the king
assembled-all the young men and formed them into a new regi¬
ment with its own barracks. Therefore each ‘head’ contained
members of all regiments and each regiment contained members of
all the ‘heads’. In a barracks or on parade, the ‘heads’ within a
regiment had set places according to the seniority of establishment
of their respective royal homesteads. The members of a ‘head’
supported the prince of their royal homestead; King Cetshwayo,
therefore, when he succeeded to the throne, strengthened his own
head by attaching to it more tribes. Each regiment had com¬
manders who were usually princes, chiefs, or the brothers of
important chiefs, but were sometimes brave commoners.

This centralization of the regiments in the king’s area gave him a
position in Zulu life entirely different from that of any of his chiefs.
It continually brought the men close to his capital, where they lived
on the bounty of his cattle and grain, supplemented by food sent
from their homes. But though it brought the regiments under the
king’s control, it robbed him of personal followers, since all the men
were attached to some chief. It may be noted that this centralization
seems to have been effected when the Zulu were fighting few wars
but maintaining a large standing army; Zulu prestige was so great
that there was little likelihood of other Bantu raiding them, even
though the Zulu troops were stationed far from the borders. 1

The king was also the supreme court of the nation and appeals
from the chiefs’ courts went to him. He was called on to decide

1 This organization may be contrasted with a lack of similar organization
among the Tswana and Sotho and is perhaps a reflection of the scattered estab¬
lishment of the Zulu. In the large Tswana towns the men could easily be
summoned to the chiefs home. But chiefs of tribes such as the Swazi and
Thonga seem to have kept only one or two regiments near them: th$ concentra¬
tion of the whole Zulu army about the king is unique in southern Africa.

THE KINGDOM OF THE ZULU

33

difficult cases. There were always in residence at his capital some
indunas of cases (izinduna zamacald) who heard these cases and
gave verdicts in the king’s name. Most of these indunas were
chiefs ruling areas of their own; others were sons, brothers, and
uncles of the king, and there were commoners ‘lifted up 5 by the
king for their wisdom and knowledge of law. In all the councils
of men throughout the land, the indunas were supposed to mark
men skilfulin debate and law and their ability might get them into
the king’s council. Two of his indunas were more important than
the others: the one was more specifically commander of the army and
was a chief or prince; the other was called the ‘great indund (prime
minister) and had weightiest voice in discussing affairs of state. He
was always an important chief, never a member of the royal family.

The king was supposed to maintain the customary law. Zulu
have illustrated this to me by quoting a case in which Mpande had
to decide against one of his favourites and then sent men to wipe out
the successful litigant’s family so as to make it impossible for the
decision to be carried out. But he could not decide, against the law f
for his favourite. Nevertheless, the king could in deciding a case
create new law for what he and his council considered good reason.

The king was supposed to follow the advice of his council. If he
did not, it is said that the council could take one of his cattle. The
Zulu believed that the welfare of the country depended on the
king’s having wise and strong councillors ready to criticize the
king. In council the king (or a chief) was supposed to put the
matter under discussion before the council and himself speak last
so that no one would be afraid to express his own opinion. The
king might inform his close councillors of his views and they
could put these to the council; he should not put himself in a posi¬
tion where he would be contradicted. But no councillor should
express a strong opinion; he should introduce his points with some
oblique phrase deferring to the king. The king ended the discus¬
sion and, if he were wise, adopted the views of the majority. The
council could also initiate discussions on matters of tribal or
national interest. It seems that in fact the king did consider his
councillors’ views and did not act autocratically. Sometimes he
would excuse an action by saying that it had been done by his
indunas without his authority, and this does seem to have hap¬
pened; and in dealing with Europeans on occasion the kings said
they were willing to do something,, and then backed out on the

34 AFRICAN POLITICAL SYSTEMS

plea that their indunas had decided against it. The king’s power

and the councillors 5 insistence on their rights and jealousy of one

another might all affect the course of discussion and the decision
on any matter or case.

From his subjects’ point of view, one may say that the main
duty they owed the king was military service, including labour
service. The king was also entitled to certain royal game, though
he had to reward the hunters. In addition, it was customary to

give him gifts of grain, beer, cattle and, some say, girls. As he

also received most of the cattle and women captured in war and

fines for certain offences, he was easily the richest man in the
nation. 1 In return for this, he was expected to feed and help his

people generously. He had to care for his regiments and give them

their shields; in famine he was expected to help all his people and

also at all times those in difficulties. Thus if the king ruled accord¬
ing to tradition, he was generous to his subjects, using his wealth
for them; he gave them justice; he protected their interests; and

through him they hoped to satisfy their ambitions on battlefield
and in forum.

III. Status and Political Power

All the members of Shaka’s family enjoyed a higher status as a

result of his victories. Neither he nor Dingane had any children
and it was the descendants of Mpande who came to form the
royal family, though certain important collateral lines were re¬
garded as princes. Any child of these lines, and the children of
their daughters and adopted women, were referred to as abant-
wana (= children, but is equivalent to princes and princesses).
They formed the superior rank in Zulu society, in status above
even the chiefs; some of them also ruled as chiefs of tribes.
Princes of the Zulu lines, and chiefs of other clan lines who were
princes by royal women, were among the most powerful chiefs in
the land. But the closer a royal prince was (and is) by birth to the
reigning king the higher his social status, though he might exert
less influence in the nation than other princes or even commoners.
To a lesser extent the same rules applied to the reigning families
within the tribes; the close relatives of a chief were the aristocracy
in his tribe.

1 Shaka made all trade with Europeans pass through his hands; and later
only important people were allowed to buy certain goods from traders.

35

_ THE KINGDOM OF THE ZULU

■ Any member of the Zulu royal family had to be greeted cere¬
monially by commoners, including chiefs. Any royal prince
might also be greeted by some of the praise names of the king,
such as ‘source of the country 9 , if the king were not there, though
the royal salute, Bayede , and the names inkosi (king), inganyama
(lion), should be strictly reserved for the king himself. This
status of the princes brought some of them political power. Shaka’s
brothers became chiefs in the areas in which they settled. Mpande
followed the practice of big, polygynous chiefs and settled his sons
in various areas as chiefs there. The king was therefore head by
descent of the powerful aristocratic Zulu lineage which was looked
up to by all Zulu, and his position in the national organization was
strengthened, since tribes scattered through Zululand were ruled
by his close relatives, who were bound to him by strong kinship
ties of mutual assistance and by their common membership of the
royal lineage. Marriage ..between the royal family and families of
chiefs established similar ties. The king would marry off a sister,
a daughter, or even some girl belonging to him, to a chief, and her
son (who ranked as a prince in the nation) should be heir. How¬
ever, the princes might draw to themselves followers beyond those
given them by the king, and as in the past brothers of tribal chiefs
had broken away to establish independent tribes, so the princes
within the nation were a potential threat to toe king, especially if he
misruled.' They were ready to intrigue against him and take ad¬
vantage of the people’s dislike of him. Zulu custom says the king
should not eat with his brothers, lest they poison him. ‘ His rela¬
tives on his mother’s side and by marriage were said to be his
strongest supporters, for their importance in national life came
from their relationship to him, rather than their relationship to the
royal lineage.

Zulu therefore state, on the one hand, that the king rules with
the support of his brothers and uncles, and, on the other hand,
that the king hates his brothers and uncles, who may aspire to the
throne. In practice, it appears that more often the princes and
chiefs competed for importance at court, i.e. they intrigued against
one another, rather than against the king. While Mpande lived his
sons also struggled for power. The most important of these
struggles was for Mpande’s heirship. The rule of succession is that
the heir is bom of the woman whom the king makes his chief wife.
Mpande first appointed Cetshwayo heir, for Cetshwayo was bom

3 6 AFRICAN POLITICAL SYSTEMS

of a wife given him by Dingane. Then he began to favour Mbuyazi,
son of his most beloved wife. Each son had his own following.

Cetshwayo was supported by his most important brothers and the

big chiefs* and he routed Mbuyazi: from that time he began to rale.
Zulu succession has been very chequered: the first two kings were
ousted by a brother* and Cetshwayo fought for the kingdom.
When the British deposed him, his chief wife had had no children,
so he appointed Dinuzulu, then about seventeen, heir. A son was
bom to the chief wife after Cetshwayo S s death and was made a
chief by the British; Dinuzulu objected, as he feared his brother
would come to be a rival.

Royal rank therefore tended to cany political power either in the
form of a personal following or else of great weight in tribal and
national councils. Otherwise high rank in the nation, with
respect, was accorded to all political officers, whether they were
chiefs or councillors of the king, and to his important servants and
brave warriors.

One other principle also gave high rank—namely, kinship
seniority within any kinship group. As stated above, Zulu, the
founder of the Zulu clan, was the junior brother of Qwabe, the
founder of the Qwabe clan. To-day the Qwabe chief is one of the
few chiefs who will not recognize the superiority of the Zulu king:
he claims that he himself is superior by birth. People to whom I
have put his claim consider that it is invalid: Shaka founded the
Zulu nation and therefore his heirs are entitled to rale it. Never¬
theless, they say, the king should ‘respect’ the kinship seniority of
the Qwabe chief. This principle worked through all the clans.
Independently of political power or boundaries, the people con¬
tinued to pay respect to the lineal head of their clan. They might
take inheritance cases to him and assist him with the bride-wealth
for his chief wife, even if they lived under a chief of another clan
line.

IV. The Tribes within the Nation

Zululand was divided into a large number of tribes of varying
sizes. 1 In Zulu theory the chiefs (or their ancestors) of all these
tribes were ‘raised up’ by one or other of the kings. By this the

1 Estimate of population: The nation about 1870 probably numbered a
quarter to half a million; tribes varied from a few hundreds to several thou¬
sands. The later regiments consisted of nearly 8,000 men.

THE KINGDOM OF THE ZULU 37

Zulu mean that they held power subject to the king and that ulti¬
mately, at the time when the Zulu nation was being created and
consolidated, Shaka or his succeeding brothers made their
ancestors into chiefs, or allowed them to continue their rule in a
particular area. The kings either recognized existing chiefs or
sent some man with a following to colonize an uninhabited area.
One of the important ways in which a man obtained political
status was by royal birth, as described in the preceding section.
Other men were the heads of the remnants of tribes which had
been independent before Shaka’s conquest and there were clan-
heads to whom, in the years after the initial wars had scattered
their people, their followers returned. The kings on occasion also
rewarded personal body-servants, brave warriors, and learned
councillors by putting them in charge of districts. But usually the
chiefs were princes or the heads of clans. The lineal heads of
certain clans had no political power; other clans were represented
by chiefs in various parts of Zululand. It was even expedient for
the kings to recognize clan-heads as chiefs since kinship affiliation,
was still a principle uniting people and cognizance had to be taken
of the groups thus formed.

From the earliest times political officers had been succeeded by
their sons and under the kings this rule continued to be recognized.
Zulu still say that an induna or chief had his position because he
was given it by the king; but if he died his heir, unless hopelessly
incompetent, should succeed him. And, failing the heir, the king
(or chief) should appoint a close relative to act as regent and the
position return to the main line if possible. Zulu say the heir has
a right to be appointed, but it depends on the king’s will; yet it is
recognized that if the heir is passed over he and his followers may
cause trouble. For example, I heard an important chief discussing
with two of his brothers the appointment of a successor to a re¬
cently deceased induna of the X-clan. The brothers were

against a descendant being appointed, as they held that the dead
man had intrigued with the local magistrate to be recognized as
independent. The chief replied that there was no proof that his
sons would act in that way because he had; in any case, the area

was thickly populated by the X-- people and he asked his

brother how they would like it if a stranger were put in control of
their own area? He concluded: Tf we do that, we shall have
trouble with the X-people . 5 (I need scarcely note that the

3 8 AFRICAN POLITICAL SYSTEMS

magistrate was not party to this intrigue.) Nevertheless, in a small
indtma's area the king (or chief) had power to appoint a parvenu:
the king would hesitate to interfere directly in the succession to a
large tribal chieftainship, or a chief with a powerful induna , for a
number of the people might remain loyal to the man whom they
considered the rightful heir. There were occasions, however, on
which the king favoured one or other claimant to a chieftainship
to suit his own ends. If the homestead of an important son of the
king or a chief (especially a Zulu clan chief) was built in a com¬
moner induna 9 s area the prince became political head there and the
induna became his induna. However when Mpande wanted to
settle one of his sons as a chief, he asked a chief for land for him.

The chiefs had certain powers delegated to them by the king.
Their most important duties were judicial and administrative.
They tried cases coming to them on appeal from their indunas 9
courts and investigated breaches of the law. Though in theory all
fines for bodily hurts went to the king, in fact the chiefs seem to
have kept these; however, they periodically sent large herds of
cattle as gifts to the king. They were supposed to forward difficult
cases and cases involving important estates to the king. In their
decisions they were bound to follow laws issued by the king and
from them appeal lay to his court. They had power to execute
judgement, but no power of life and death. In practice, some
powerful chiefs were fairly independent and are said to have
executed sorcerers and adulterers. As judicial heads in their
districts, they had to report to the king all grave misdemeanours
and they had to watch over the public weal. An example of how
the king delegated administrative duties to his chiefs is Cetsh-
wayo’s appointing a coastal chief to facilitate the passage of
labourers from Thongaland to Natal through Zululand, which he
had agreed to do for the Natal Government. As the chiefs were
often in attendance on the king, they could not perform these duties
themselves, but delegated them to trusted relatives and indunas.

The king communicated with his chiefs by runners. To imper¬
sonate a king’s messenger was punishable by death. Thus orders
to mobilize at the capital, projected laws and matters of national
import were announced to the people by the king through his
chiefs, though many announcements were made at the first-fruits
ceremony. When necessary, the chiefs passed on these orders to
their indunas in charge of wards and these reported to the heads of

THE KINGDOM OF THE ZULU 39

lineage groups and homesteads. All the people were entitled to
express their opinion on affairs and they did this through the heads
of their kinship groups and then their immediate political officers.

In addition, the chiefs and indunas had administrative duties within
their own districts, including the allocation of land, the main¬
tenance of order, trying of cases, watching over their districts’
welfare, taking ritual steps to protect the crops, looking for sor¬
cerers. Chiefs, like the king, received gifts of com and cattle, but
they levied no regular tribute. They could cal out their subjects
to work their fields, build their homesteads, arrest malefactors, or
hunt. In turn, they were expected to reward these workers with
food and to help their people who were in trouble. Like the king,
too, they were bound to consult and listen to a council composed of
their important men.

Thus authority from the king was exercised through the chiefs,
his representatives in various districts. They ruled through their
brothers and indunas of smaller districts, under whom were the
lineage- and homestead-heads. Zulu political organization may
therefore be seen as delegated authority over smaller and smaller
groups with lessening executive power. From inferior officers
there was an appeal to higher ones; in theory the king’s will was
almost absolute. At the bottom were the heads of kinship groups
who could issue orders and arbitrate in disputes within their
groups, but who could not enforce their decisions, except over
women and minors. On the other hand, as the groups became
smaller the ties of community and kinship grew stronger, and as
force lessened as a sanction other social sanctions increased in
importance. The dependence of men on their senior relatives in
religious and economic matters, as well as in trouble, was strong;
even at the barracks they shared huts with their kinsmen and
relied on them for food and support in quarrels. In kinship
groupings the main integrating activities and social sanctions were
based, on reciprocity and communal living. Some kinship rules
were backed by judicial sanctions, but when these obligations were
enforced at law, force was used on the chief’s judgement, not on the
obligation itself.

I have described the tribes and smaller groups as part of a pyra¬
midal organization with the king at the top in order to bring out
the administrative framework which ran through the social
groupings, but the position of the head of each group in the series

40

AFRICAN POLITICAL SYSTEMS

was different, for he was related to the members of his group by
different ties from those linking them to the head of the larger
group of which it was part. Besides the ties of sentiment, home¬
stead and lineage, heads exercised authority because of their

kinship status and their importance in their inferiors’ social and
economic' life; indunas and their followers shared in common
social, and often economic, activities, as well as political affairs;
tribesmen were attached to their chiefs mainly by political bonds;
and "all Zulu to the king by their military duties. The, average
Zulu’s importance decreased the bigger the group of which he was
a. member. The king’s position in the state was essentially his
establishment in the ‘barrack area’. He symbolized for the Zulu
their identity as a nation as against the Swazi and other Bantu, and
European, Powers. The nation was a federation of tribes whose
separate identities were symbolized by their chiefs. The tribes
were even autonomous within the national organization for on
occasion many tribesmen supported their chiefs in quarrels with
the king, though some were swayed by national loyalties. 1 How¬
ever, it was in the relations between tribes that tribal identities
mainly appeared. There existed between the tribes a strong
hostility which radically affected the course of Zulu history after
the Zulu War of 1880; it was mirrored at court in the competition
of the chiefs for power. For the people of any tribe of some
strength were proud of their traditions and their chiefly line, were
loyal to their chief and quick to resent any attempt by other chiefs
to interfere in their tribal affairs. Occasionally, especially on the
borders of tribes, this hostility broke out in affrays. It appeared
most clearly in the people’s attachment to their own chief as against
other chiefs. Therefore, as will be seen in the next section, the
chiefs tried by ruling well to win adherents from other chiefs.
Nevertheless, the chiefs were often related to one another and on
friendly terms. As part of the administrative machinery they
served together on the king’s council and they might even combine
to constrain the king.

Within a tribe, there was a similar opposition between sections.
The tribes were divided, as described in the paragraphs on the
Army, into sections attached to homesteads of the chief, his

1 This is how Zulu describe it; in fact, they may have been moved by self-
interest or other motives, but their actions axe described in terms of tribal and
national values.

THE KINGDOM OF THE ZULU

41

brothers, and his uncled; the adherents of each of these home¬
steads were very jealous of their ‘prince’s’ prestige and felt a local
loyalty to him as against the adherents of other ‘princes’. Before
and after the death of a chief, these groups vied with one another
to have their ‘prince’ nominated as heir, and were even ready on
occasion, despite their tribal loyalties, to support him against the
heir when he assumed power. Faction fights between these sec¬
tions continue to-day, often flaring up over trivial matters; and
when Government assumed rule in Zululand it inherited a rich
legacy of their feuds and of inter-tribal feuds. Even the members
of wards under commoner indunas often came to blows, for at
weddings and hunts they assembled as members of military sec¬
tions or wards, and if a fight started between two men their fellow
members would support them. Thus in every Zulu political group
there was opposition between its component sections, often mani¬
fested through their leaders, though they co-operated in matters
affecting the welfare of the whole group.

The opposed groups within the nation were united by the
common service of their leaders in the council of the larger group
of which they were part. The administration ran in separate
threads from king to a particular chief, to a particular induna, to a
particular lineage-head; all these threads were woven together in
the council system. Though the group-heads were the main part
of what bureaucracy there was in the simple Zulu social organiza¬
tion, their functions as bureaucrats and as group-heads were not
entirely identical. In previous paragraphs some of their functions
as group-heads have been reviewed. As administrators, they
watched their people’s interests and ruled them according to the
orders of their superiors, and they also used their people’s backing
in their struggles for administrative power, perhaps against the
people’s interests. They and the officers about a court were the
link between a ruler and his subjects, but frequently tended to
become a barrier between them, for they were jealous of their
rights, resented any encroachment on their privileges and some¬
times acted independently of the ruler. The people had to consider
these officials in approaching their rulers, the rulers were largely
compelled to conduct their relations with the people through them.
There was therefore an unstable balance of duties and interests
between the group-heads acting as courtiers and other courtiers,
and the rulers and the people.

42

AFRICAN POLITICAL SYSTEMS

V . Sanctions on Authority and the Stability of the State

The king was bound to consider custom and his council. The
Zulu king rarely called full meetings of the nation for discussion;
he consulted their wishes through the chiefs. The people could
not themselves criticize the king, but he might suffer if he dis¬
regarded their feelings entirely. The king was supposed to be just
and generous and princes and chiefs were educated in, and con¬
scious of, the tradition of good rule. The Zulu point to their his¬
tory and show its lessons. Was not Shaka killed because he
oppressed the people, so that Dingane did not fear to kill him ? In
turn, many people supported Mpande against Dingane. Mpande,
the just and generous king, ruled long I have been told that if a
chief troubled his people, his family and indunas would poison
him, but my informants could not give me a case in which this
was done.

It required a long period of suffering before the people would
turn against their rulers. Kings and chiefs were said to have many
spies, and it was difficult to organize armed resistance to the king,
though Zulu point out that all Shaka’s spies did not save him from
assassination. The king was backed with great force and a rebel¬
lion required that jealous chiefs and princes should unite. An
early European visitor to Shaka records that his policy was to keep
his chiefs at loggerheads with one another, and the Zulu admit
this as a method of rule, pointing out that Government uses it
to-day in dividing up Natal and Zululand into 300 chieftainships.
Outside of the royal family there was no one who could hold
together the nation and this was recognized by the chiefs.

The people depended for leadership against an oppressive
ruler on their nearer political officers. The Zulu had no idea of
any political organization other than hereditary chieftainship and
their stage of social development did not conduce to the estab¬
lishment of new types of regime. Their only reaction to bad rule
was to depose the tyrant and put some one else in his place with
similar powers, though individuals could escape from Zululand to
other nations’ protection; that is, the people could take advantage
of the princes’ and chiefs’ intrigues for power and the latter in
intriguing sought to win the backing of the people. The king’s
policy was therefore to prosecute any one who threatened to be
able to take his place: he had to meet rivals, not revolutionaries.

THE KINGDOM OF THE ZULU 43

The kings killed all brothers whose rivalry they feared. Uncles
(fathers in the kinship system) were less likely to oust the king,
and while the people should not complain against the king to his
brothers they could appeal to one of his uncles. The kings, and
all officers, were always on the watch for these threats to them. As
the medicines of a ruler were believed to make him immune to the
influence of his inferiors, if he felt ill in the presence of some
person he could accuse the latter of sorcery.

The king had to treat all his brothers (and chiefs) carefully, lest
they became centres of disaffection against him. The tension
between the king and his brothers was a check on the king’s rule
because his subjects could shift their allegiance to his brothers. In
addition, because the Zulu were strongly attached to their imme¬
diate political heads, the chiefs, and would even support them
against the king, the chiefs had power to control the actions of the
king. On the other hand, the chiefs remained dependent on the
king. He could enlarge the powers of his favourites or assist the
rivals of a recalcitrant chief.

Within tribes the chiefs held power under similar conditions.
They could use armed force against disobedient or rebellious
subjects though they had to inform the king that they were doing
this. Theye were stronger checks on their rule. Their subjects
could complain to the king if they were misruled. Though a man
could in theory sue the king, he was not likely to do so; a chief
could be brought before the king’s court. Misrule by a chief
would strengthen the hands of his brothers within the tribe and
these brothers, unless the king intervened, might seize power.
A quarrel with an important brother or subject might induce him
to live elsewhere with his personal adherents. While misrule drove
subjects to other chiefs a good and generous rule would attract
followers. The Zulu have it that a chief should be free and
generous with his people and listen to their troubles, then they
will support him in war and ‘not stab him in his hut’. The forces
of fission and integration which marked the early political organi¬
zation were still at work in the Zulu nation and to benefit by them
it behoved a chief to rule wisely and justly in accordance with the
wishes of his people.

The Zulu had loyalties to their various political heads. While
these loyalties did not generally conflict, if king, chief, or induna
abused his power the people would support one of their other

44 ■ AFRICAN POLITICAL SYSTEMS

political heads against him, and in their intrigues for power the
political heads were ready to take advantage of this. Thus the
potential conflict of these loyalties was a strong check on misrule
and gave the people some control over their rulers.

VI. The People and Their Leaders

The working of these forces depended on the fact that political
leadership was personal. In theory, any one could approach his
superiors through their courtiers, though it might take some days.
A chief (and even the king) was supposed to deal with his people
himself and should not altogether delegate this duty. Chiefs and
indunas knew most of their subjects, with their relationships and
ancestry; if a stranger arrived at a capital all details about him
were asked. To a lesser extent this applied to the king. The chief
attended his people’s weddings and sent his condolences, or
visited them, if a relative died. The Zulu sum this up by saying
‘the people respect their chief, but the chief ought to respect his
people’.

This intimacy between the chief and his people, despite the
ceremonial which surrounded him, was largely possible because
there was no class snobbery among the Zulu. The chief was still
regarded and treated as the ‘father of his people’; ‘they are your
father’s people’, he was told; ‘care for them well’. And did not the
chief belong to the tribe, especially if it had subscribed the bride¬
wealth for his mother? There was no insurmountable barrier to
marriage between his and any of his subjects’ families. Though
the courtiers had greater knowledge of affairs than the provincials
had, the Zulu all had the same education and lived in the same
way; and any one could take his part in the chief’s council or
assist in judging a case. Birth, age, courage, and wisdom all
affected the attention a man would get; but every one could speak.
Wealth brought a chief closer to, did not remove him from, his
people. For under the conditions of Zulu life wealth did not give
a chief opportunity to live at a higher level than his inferiors. He
had more wives and bigger homesteads, but he could not surround
himself with luxuries, for there were none. Wealth, in the form of
well-filled granaries and large herds of cattle, gave a man power
only to increase the number of his dependants and to dominate
many inferiors. From the point of view of the chief, it may be said
that he had to be rich in order to support his dependants; and

THE KINGDOM OF THE ZULU

45

besides this there was no use for wealth. 1 On the other hand, the
wealth of a commoner attracted dependants and gave the rich man
political status. Moreover, the kinsman of a wealthy man would
not quarrel readily with him, so that there was little likelihood of
his kinship group breaking up. However, there were few ways in
which a commoner could acquire wealth: he might by magical
practice, or as a reward from king or chief for some deed, or as
booty in war. I have been told that only important men owned
cattle. The rich Zulu loaned out cattle to other people to herd for
him; they could use the milk, and also the meat of animals which
died, and this contract made them dependent on the cattle-owner
because he could inflict great hardship on them by taking away his
cattle. When the chief did this, it gave him a hold over his people
and prevented them from easily changing their allegiance and
going to some other chief. 2 Wealth therefore attracted followers,
and as they increased and had children the wealthy man could
collect about him a substantial group of dependants which was a
political unit. Kinship alone, within a lineage, also created poli¬
tical units; even the head of a homestead had political duties. A
notable feature of Zulu political organization throughout Zulu
history is the creation of new groups as people moved about,
settled and increased, and the heads of all these groups were
minor political officers who might in time achieve prominence.
Since leadership was personal, these groups were not merely
absorbed into existing political groups; their leaders became
officers within the organization.

There was thus a constant creation of new officials which, with
the rise in rank of brave warriors and wise men, permitted of a
high degree of social mobility. Any man, whatever his rank by
birth, could become politically important if he had the ability,
though those already established in high positions watched
jealously over their rights and privileges. Thus it is said that
sometimes if a chief became jealous of an inferior he would kill
him on a trumped-up charge of sorcery, though it seems that more
often these charges were brought by other men in the chief s
court. They were (and are) frequent in court circles.

1 To-day the position is. different. Wealthy men can build European houses

and buy motor-cars, clothes, ploughs, &c. . , . *

2 A certain chief in Zululand to-day has a remarkable hold over his people.
he has 16,000 cattle (out of a tribal total of 54>°°°) loaned out among them.

46

AFRICAN POLITICAL SYSTEMS

The Zulu say that to-day there is more security of life than in
olden times, when a man might be killed for anything. Despite
this, and despite accounts of lawlessness and favouritism, the old
Zulu declare that they got justice from their chiefs and help in
times of trouble. They deny that chiefs could be bribed. Mpande
and Cetshwayo both gave decisions against important chiefs. The
chiefs were undoubtedly cruel and capricious on occasion, but
they were generous, though one old man who had been much at
court said to me sadly: ‘There is no chief who is kind.’ The old
Zulu generally shake their heads over the harsh rule of the past;
and then speak of the glories under it. The life of the subjects
seems to have varied with the character of their chief.

During the time of the kings, the State bulked large in the
people’s lives. In council and on the battle-field only could high
ambitions be satisfied. In the smaller districts the men were
always busy on administrative and judicial affairs when they were
at home, especially after they retired from the active service,
during which they had to spend a large part of their time at the
military barracks. Here they starved for days, then feasted
royally on meat. They lounged with their fellows, hunted,
danced before the king, paraded for the national ceremonies. The
king would sit and talk with his important men, discussing the
law, mighty deeds, and history. Tribal and ward capitals were the
centres of social life in their districts. The evolution of the barrack
system affected Zulu social life considerably: it controlled mar¬
riages, and, though the old sexual labour division values re¬
mained, it was necessary for girls to assist in the work of herding
and milking. The young men were not available for work at home
and food had to be sent to them at the barracks. For the moment
when they would be enrolled as warriors they waited eagerly,
longing to join a regiment.

VII. The Period, of European Rule 1

Between 1887-8 the British Government finally took over
the rule of Zululand, despite Dinuzulu’s armed opposition. In

1 My observations on modem Zulu politics have been made especially in the
districts of Nongoma, Mhlabatini, and Hlabisa, and to a lesser extent in Ngotshe,
Vryheid, Ingwavuma, and Ubombo. Certain observations have also been made
in towns, on travels in more southerly districts, and at gatherings of chiefs and
Zulu with Mshiyeni, the Regent of the Zulu royal house.

THE KINGDOM OF THE ZULU

47

a short time Government rule was confirmed. 1 To-day it is a vital

part of Zulu life: of ten matters I heard discussed one day in a
chief’s council seven were directly concerned with Government.
Fifty years of close contact with Europeans have radically changed
Zulu life along the lines known all over South Africa. 2 The
military organization has been broken and peace established. The
adoption of the plough has put agricultural labour on to the men,
and they go out to work for Europeans in Durban, Johannesburg,
and elsewhere. The development of new activities and needs, the
work of various Government departments, missions, schools,
stores, all daily affect the life of the modern Zulu. Communica¬
tion has become easier, though pressure on the land is greater.
Money is a common standard of value. The ancestral cult and
much old ceremonial have fallen into disuse.

Zululand is divided into a number of magisterial districts, which
are divided into tribes under chiefs, 3 who are granted a limited
judicial authority and who are required to assist the Government
in many administrative matters. 4 Within a district the magistrate
is the superior political and judicial officer. He is the representa¬
tive of Government. His court applies European law and is a
court of first instance and of appeal from chiefs in cases between
Natives decided according to Zulu law. He co-operates with other
Government departments, and with the chiefs and their indunus .
This, according to statute, is the political system: the chiefs are
servants of Government under the magistrate, whom they are
bound to obey. In Zulu life the magistrate and the chief occupy
different, and in many ways opposed, positions.

The modem Zulu political system is ultimately dominated by
the force of Government, represented in the district by the police.
They are few in number, for the area and population they control,

11 lack space to discuss historically the way in which Government rule has
been accepted, but have tried to make this implicit in my account of the system

2 Se e i # Schapera (Editor), Western Civilization and the Natives of South

^^^r^opiJation^gures ’ and maps see N. J. van Warmelo, A Preliminary
Survey of the Bantu Tribes of South Africa (Union Government Printer, Pretoria,
1935). Magisterial districts comprise about 30,000 people; tribes vary trom

tens of to a few thousand taxpayers. . _

* These duties are defined by the Natal Code of Native Law, Proclamation
No. 168 /1932. See W. Stafford, Native Law as Practised in Natal (Witwaters-
rand University Press, Johannesburg, 1935 )*

AFRICAN POLITICAL SYSTEMS

48

but behind them lies the overwhelming military power of the
Europeans. The magistrate is backed by this power and he is

vested with all the authority of the white upper class in the South
African community. In the development of new activities which
has marked the change in Zulu life, Government has played a
leading role. It has established peace, encouraged men to go to
work for Europeans, supported schools, started health, veterinary,
and agricultural services. The magistrate, therefore, not only
applies Government regulations, but he is also the chief head
of the organization which is bringing new enterprise and
some adaptation to new conditions to the Zulu. 1 He has to
do many things which the chief cannot do for lack of power,
organization, and knowledge. People go to the magistracy with
questions and troubles. Thus the magistracy has come to stand
for many of the new values and beliefs which to-day affect Zulu
behaviour.

However* while the Zulu acknowledge and use the magistracy,
their attitude to Government is mainly hostile and suspicious.
They blame it for the new conflicts in their community; they
point to laws which they consider oppressive ; 2 they regard
measures which Government intends in their interests as being
designed to take from them their land and cattle, and cite in
argument the encroaching of whites on Zululand in the past and
what they regard as a series of broken promises to them. More¬
over, many of these measures conflict with their pleasures, beliefs,
and mode of life, as, for example, the forbidding of hemp-smoking
and of sorcery trials, and the dipping of cattle and control of cattle
movements. Therefore while Government requires the chiefs to
support its measures, the people expect their chiefs to oppose
them. And, indeed, the chiefs are usually opposed to them. This
position was clearly emphasized in 1938, when a chief who opposed
the building of cattle paddocks to prevent soil-erosion was praised
by his people, but condemned by officials; a chief who asked for
a cattle paddock was praised by officials, but condemned by his
people. They complained against him to the Zulu king. For

1 1 lack space to discuss the relations of different Government departments
or the role of missionaries, and the reaction of the Zulu to changes, many of
which have not been purposefully made.

2 See D. D. T. Jabavu on ‘Bantu Grievances’ in Western Civilization and
the Natives of South Africa, op. cit. These are outlined from the point of view
of an educated Native.

THE KINGDOM OF THE ZULU 49

the people look to their native leaders to examine Government
projects and ‘stand up for the people’ against them.

The imposition of white rule and the development of new
activities have radically curtailed and altered the chief’s powers.
He is subordinate to Government rule; he cannot compel, though
he levies, labour service; he still owns the land, but it is less and
subject to Government control; he has lost his relatively enormous
wealth and often uses what he has in his own, and not his subjects’
interests; he is surpassed in the new knowledge by many of his
people. The men now have less time to devote to their chief’s
interests. A chief may try to enforce old forms of allegiance which
some subjects will not render and this leads to conflict between
them. If he tries to exploit or oppress a man, the latter can turn
to the magistrate who will protect him. This last important point
needs no elaboration, though it may be noted that as far back as
the civil wars the different factions tried to persuade the British
Government that they were in the right and should be helped.
The chief can compel only that allegiance which Government, in
its desire to rale through the chiefs, will make the people render,
though his disapproval is a serious penalty in public life. Never¬
theless, the chief still occupies a vital position in the people’s life.
Not only does he lead them in their opposition to Government,
but he also has for them a value the magistrate cannot have. The
magistrate cannot cross the barrier between white and black. He
talks with his people and discusses their troubles, but his social
life is with other Europeans in the district. The chiefs social life
is with his people. Though he is their superior, he is equal with
them as against the whites and ‘feels together’ with them. ‘He
has the same skin as we have. When our hearts feel pain, his heart
feels pain. What we find good, he finds good.’ A white man cannot
do this, cannot represent them. The Zulu are ignorant of Euro¬
pean history and it can have no value for them: the chiefs, and
especially the king, symbolize Zulu traditions and values. They
appreciate with their people the value of cattle as ends in them¬
selves and of customs like bride-wealth which are decried by
Europeans. The chief is related to many of them by kinship ties
and any man may become so related by marriage; the social and
endogamous barrier between whites and blacks cannot be satis¬
factorily crossed. The Zulu acknowledge their chiefs’ position
largely through conservatism and partly because Government

50

AFRICAN POLITICAL SYSTEMS

recognizes it. But a chief is usually chief by inheritance: ‘he has

the blood and the prestige of chieftainship and they extend to
his relatives; the magistrate has only the prestige of his office.’
By this contrast Zulu express the chief’s position as it exists
independently of Government’s acknowledgement and rooted in
the values and habits of the people. Chiefs and members of the
royal family are greeted with traditional modes of respect. Their
family history is retailed. Their capitals are centres of social life.
They are given loyalty and tribute.

I have outlined the opposed positions of chief and magistrate:
the balance between them is the dominant characteristic of the
political system. However, it shifts from situation to situation in
Zulu life. A certain minimum of allegiance to both magistrate
and chief is legally enforced by Government; the influence of each
may vary above that minimum with their characters and relations
to each other, or according to the matter considered. A sym¬
pathetic magistrate who understands the Zulu will draw them to
him, especially from a chief who is unsatisfactory; a harsh magis¬
trate keeps people away from him and they go more to their chiefs.
Even more the balance shifts for different individuals in different
situations. A man who considers the chief to be biassed against
him, favours the magistrate as impartial; but for him the chief is
the source of justice when the magistrate enforces an unwelcome
law. The people rally to the chief when they oppose measures
such as the reduction of bride-wealth. If the chief tries to force
labour from people, they compare him unfavourably with the
magistrate who pays for the labour he employs. Though in many
situations it cannot be done, the Zulu constantly compare Native
and European officers and switch their allegiance according to
what is to their own advantage or by what values they are being
guided on different occasions.

It has been necessary for this analysis to emphasize the opposi¬
tion between chief and magistrate. It is strong, and appears in
the jealousy each often has of the other’s power. But in routine
administration the system functions fairly well. Chiefs and
indunas actively assist in the administration of law and the carrying
out of certain activities. The magistrates, keen on their work and
anxious to see their districts progress, may as individuals win
the trust of their people, though it is never complete and the
fundamental attitude to Government remains unchanged. They

THE KINGDOM OF THE ZULU

5i

represent their people to Government, and the administration, in
developing the Native reserves, seems to be coming into conflict
with Parliament in so far as Parliament represents white interests.
But though in general the system works, the opposition between
the two sets of authorities becomes patent over major issues. Then
ultimately the superior power of Government can force a measure
through unless it depends on the willing co-operation of people
and chief. The Zulu now have little hope of resisting Govern¬
ment rule and sullenly accept Government decisions. In the
chiefs’ councils, they vent their opposition in talk.

In evaluating this reaction to modem political institutions it is
necessary to distinguish between two groups of Zulu, the pagan
and the Christian (or schooled). Any schooled Zulu is in general
much readier to accept European innovations than are the pagans.
However, the majority of Christians have the same attitudes as
the pagans, though their complaints against Government and
whites may be differently formulated. Some better educated
Christians measure the chief’s value by the materialistic standard
of the practical work done by Government and hold that the chiefs
are reactionaries opposing progress and they favour a system like
the Transkeian Bunga. They are possibly moved by desire for
power themselves. In general it may be said that most schooled
Zulu regard the magistracy with more favour than do pagans, but
among the best educated Zulu, who come most strongly against
the colour bar, there is a tendency to a violent reaction to their
own people and culture and values away from the Europeans.
Nevertheless, it is through the Christians that the Europeans
introduce most new ideas into Zulu life. This is causing hostility
between pagans and Christians and creating, on the basis of
differences in education, adaptability, enterprise, and values, a
new opposition in the nation. Aside from these Christians, there
are the pagans who attend on whites, seeking some advantage and
trying to profit from the political situation: thus Zulu unity against
the whites is weakened. The people tend not to see a conflict in
their own actions, though they feel and suffer under it, but often
they criticize other Zulu for their allegiance to the whites, saying
that they are selling their people to the white man.

Though all Zulu tend to be united against the Europeans, old
tribal loyalties and oppositions are still at work and faction fights
frequently occur. Tribes are often hostile to one another, but

S* AFRICAN POLITICAL SYSTEMS

they are again beginning to support the Zulu king. He is recog¬
nized legally only as the head of a small tribe in Nongoma district,
though Government recognizes his superior status and through
him speaks to, and hears from, the Zulu people. He has been
used on several occasions to settle disputes in other tribes and
always gets precedence over other chiefs. The present Regent is
Government’s nominee to the Union Native Representative
Council. Government thus recognizes the continued existence of
the Zulu nation. The strength of Zulu nationalism is growing
after a period of weakness. The head of the royal house is again
the king (inkosi); other chiefs are abantwana (princes) or abanum-
zana (big people). He exercises influence, aside from what
Government allows him, in other tribes. Nearly all the tribes of
Zululand and Natal and some in the Transvaal acknowledge him
as their king, though many of them were never ruled by the kings
and fought in various wars for the whites against them. The king’s
present power is partly due to the fact that he symbolizes the great
tradition of the Zulu kings, which gives the Zulu their greatness
as against other peoples, such as the Swazi. Bantu national
loyalties, pride, and antagonisms are still strong despite a growing
sense of black unity. The king’s power is also part of the reaction
against white domination, for the Zulu feel that he has the ear of
Government and therefore power to help them in their present
difficulties, and that he has the courage and strength to oppose
Government. Nevertheless, under Government each chief is
independent. Jealousy and desire for power still divide the chiefs,
but only the Mandlakazi and Qwabe chiefs are jealous of the king,
though he could not get all the tribes to adopt his nominee to
represent the Natal Natives in the Union Senate; but other chiefs,
find that, as representatives' of the king, their position among
their people is stronger than it is as independent Government
duets. As such, their people suspect them of being afraid to
criticize Government. The allegiance they give the king varies
from constant consultation to recognition when he travels. All
Zulu crowd to see him when they can and heap gifts on him.

Within a tribe there remains the divisions into sections under
brothers of the chief or indunas which sometimes leads to figh ting ,
rhe chief must rule according to tradition or the tribe will support
his brothers and weaken his court, though the magistrate is, as
pointed out, the strongest sanction on misrule. If a chief palters

THE KINGDOM OF THE ZULU ‘ 53

to Government, Ms subjects may turn from Mm to a more obdurate
brother, or sometimes if the chiefs say they approve of a measure,
the people may accept it. The cMef has to pick Ms way between
satisfying Government and his people and has to control political
officers over whom he has only slight material sanction, though,
since these officials and the councils of the people are not legally
constituted by Government, he may disregard them.

I am unable, for lack of space, to examine the way in wMch the
political system functions in modem Zulu social and economic
life; or the effect on the political situation of the division of the
white colour-group into Afrikaans-speakers and English-speakers,
and other divisions within it. Briefly, it should be noted that the.
white group itself has contradictory values in approaching the
Natives; though many Europeans are influenced by both sets of
values, the missionaries, various other Europeans, and adminis¬
trators, educationalists, and people in similar positions give more
active expression to the Christian and liberal values. Many of
these Europeans are on very friendly terms with Zulu. They fight
for Zulu interests and the Zulu recognize tMs to some extent,
though they still regard them as whites and therefore suspect.
In economic life the ties between Europeans and Zulu are strong.
TMs may be seen in the traders who have to compete for Zulu
customers and in the various labour employers competing with
each other for the limited supply of labour. They attempt to get
the goodwill of cMefs in their enterprises and at the Rand mines
members of the royal family are employed to control Zulu workers
as well as to attract them there. MeanwMle, the recognition
accorded by these labour employers and traders, and also by
missionaries, to the cMefs adds to their powers in the present
situation, even while the labour flow and Christianity are weaken
ing in other ways the tribal organization.

VIII. Conclusion

Zulu political organization has been twice radically altered. On
both occasions the people quickly acknowledged their new rulers"
power and the new organization functioned fairly well; but the
old organization, which retained its values and significance,
affected the functioning of the new one of which it was made
a part. Meantime, despite the changes brought about by the
centralization of authority and the regimental system under the

54

AFRICAN POLITICAL SYSTEMS

kings, and to-day by the labour flow and the development of new
social groups and values and modes of behaviour, the smaller
social groups have remained relatively constant.

The essence of both the systems described is the opposition of
like groups and the potentially conflicting loyalties of the people to
different authorities. The nation was a stable organization, for this
opposition was principally between the tribes which were united in
the king’s position and his regiments. The circulation of the
rulers wealth was necessary to enable them to maintain their close
relationship with their people. The conflict of loyalties to officials
of different rank, often intriguing against one another, came into
the open as a check on misrule. Therefore, despite the apparent
autocracy of king and chiefs, ultimately sovereignty in the State
resided in the people. However, though a ruler might be deposed,
the office was not affected. In actual administration, the loyalties
of the people and the competition of officers did not often conflict,
since the administrative machinery worked through the heads of
groups of different type: the main opposition was between similar
groups, co-operating as parts of a larger group.

To-day the system is not stable, for not only is Zulu life being
constantly affected and changed by many factors, but also the
different authorities stand for entirely different, even contra¬
dictory, values. The Zulu, with their strong political organiza-
tion, have reacted against white domination through their political
authorities, who were incorporated in Government administra¬
tion. The modem political organization of Zululand is the oppo¬
sition between the two colour-groups represented by certain
authorities. Each group makes use of the leaders of the other
group if it can for its own purposes. The opposition between the
two groups is not well-balanced, for ultimately it is dominated by
the superior force of Government, against which the only reaction
of the Zulu is acceptance or passive disobedience. The threat of this
force is necessary to make the system work, because Zulu values
and interests are so opposed to those of the Europeans that the
Zulu do not recognize a strong moral relationship between them¬
selves and Government, such as existed, and exists, between them¬
selves and their king and chiefs. They usually regard Government
as being out to exploit them, regardless of their interests. Govern¬
ment is now largely a sanction on oppressive chiefs, but the old
checks on the chiefs act to prevent them becoming merely

THE KINGDOM OF THE ZULU

55

subordinate tools of Government. Government, too, does much

work that the chiefs cannot do.

The opposition is synthesized by co-operation In everyday
activities; by the position that an individual white official wins In
the people’s esteem so that he comes even to stand for them
against Government, i.e. he enters the black, as opposed to the
white, colour-group; and by the attempt of the people to exploit
the opposed political authorities to their own advantage. In addi¬
tion, divisions of each large group into political groups and
opposed groups with conflicting ideals and interests act to weaken
each group within itself and to lessen the main opposition.
Members of dissident black groups, or individuals supporting
Government in some matter, may be said to be supporters of the
magistrate against the chief. In these ways social, economic, and
other ties between Zulu and Europeans are bringing the Zulu
more and more to accept white rule.

Within the one political organization there are officials, white
and black, who have entirely different positions in the people’s life
and whose bases of power are different. These officials represent
values which may be contradictory. By their allegiance in dif¬
ferent situations to the officials representing each set of values, the
people are prevented from being faced with a patent conflict of
these values. Nevertheless, as the chief’s material power is puny
compared with Government’s, the position he occupies is largely
a centre for psychological satisfaction only and white domination
is accepted by the Zulu, resignedly hostile . 1

1 Since this essay was written, Dr, Hilda Kuper’s book An African Aristocracy;
Rank Among the Swazi, has appeared (1947). Dr. Kuper’s book gives import¬
ant comparative material for an understanding of the political institutions of
the Nguni-speaking Bantu.

THE POLITICAL ORGANIZATION OF THE NGWATO
OF BECHUANALAND PROTECTORATE 1

By I. SCHAPERA

I. Ethnic Composition and Territorial Constitution

T HE Native inhabitants of Bechuanaland Protectorate, most
of whom belong to the Tswana (Western Sotho) cluster of
Bantu-speaking peoples, are politically divided into eight separate
tribes (merafe, sing, morafe)* Each has its own name, occupies its
own territorial reserve within which no European may own land,
and, subject to the overriding authority of the British Adminis¬
tration, manages its own affairs under the direction of a chief
{kgosi, morena) f who is independent of the rest. The Ngwato
(commonly termed BaMangwato) are the largest and historically
the best-known of these tribes. Their great chief, Kgama III
(c. 1837-1923), attained world-wide prominence as a zealous
convert to Christianity, a fanatical prohibitionist of alcoholic
liquor, and a strong supporter of British imperialism in central
South Africa. He promoted in various ways the economic progress
of his people, and keenly encouraged the spread of education. The
latest census, in May, 1936, showed that no less than 28 per cent,
of the population was literate—a factor considerably affecting the
modem relationship between the chief and his subjects. 2 Kgama’s
innovations were not confined to this partial acceptance of
Western civilization, but included several changes in the legal and
administrative system of the tribe. As a result of both this and the
active intervention of the British Administration, the Ngwato
have departed considerably from their traditional system of
government, and so offer an interesting field for the study of
African political development.

The tribe occupies a reserve 39,000 square miles in extent,

1 This article is based upon field investigations made in 1935 for the Bechu¬
analand Protectorate Administration. For a more detailed account of Tswana
social and political organization in general, the reader is referred to my Handbook
of Tswana Law and Custom (Oxford, 1938), chaps. i—vi, xvi.

2 The corresponding figures for other Tswana tribes are: Kgatla, 28 per
cent.; Ngwaketse, 15 per cent.; Kwena, 9 per cent.; Tawana, 1*9 per cent.

THE NGWATO OF BECHUANALAND PROTECTORATE 57

much less than the territory it claimed when the Protectorate was
established in 1885. Its population, returned in 1936 as 101,481,
is by no means homogeneous. Only one-fifth belongs to the
nuclear community, comprising the ruling dynasty and other
descendants of the people who founded the tribe by separating
early in the eighteenth century from the Kwena. 1 The remainder
are bafaladi (foreigners, ‘refugees’), who became subject to the
Ngwato chiefs at various times through conquest in war, volun¬
tary submission, flight from an invading enemy, or secession from
some other tribe. Most of them retain sufficient corporate life to
be regarded as separate communities or groups of communities
within the tribe. Some, like the Kaa, Phaleng, Pedi, Tswapong,
Kwena, Seleka, Khurutshe, Birwa, and Kgalagadi, are themselves
of Tswana origin, or come from the closely allied Northern Sotho
cluster. Others are linguistically' and culturally distinct. The
Kalaka (who actually outnumber the Ngwato proper), Talaote, and
Nabya belong to the Shona group of Southern Rhodesia,* and the
Rotse,Kuba,and Subiatothe peoples of north-western Rhodesia;
the Herero are refugees from South-West Africa, and the Sarwa
are Bushmen, with a large admixture of Bantu blood and culture.

Practically all the 'Ngwato proper, members of the dominant
community, are concentrated round the chief in the capital town
of Serowe (pop., 25,000). 2 Serowe also contains many groups of
foreigners. But the great majority of these are scattered over the
rest of the reserve, people of the same stock tending to inhabit the
same localities. They live in some 170 villages, ranging in size
from small settlements, of less than 100 people each, to such
relatively large centres as Shoshong, Mmadinare, Bobonong, and
Tonota, with populations of 2,000 or more. For administrative
purposes, the villages some distance away from Serowe were
gradually grouped by Kgama and his successors into districts,
based partly upon geographical convenience and partly upon
ethnic considerations. 8 The district of Shoshong is inhabited

1 The Kwena (living in a reserve immediately south of the Ngwato) are
generally regarded as the parent stock from which the Ngwaketse, Ngwato,
and Tawana are derived.

2 Built in 1902. Before that the tribal head-quarters was located at Palapye,
built in 1889, and before that at Shoshong.

3 The districts already existed, in the sense that people of the same stock
inhabited the same part of the tribal territory; but Kgama gave them concrete
form by appointing men specially to administer the outlying parts of the
Reserve for him (see below, p. 61).

AFRICAN POLITICAL SYSTEMS

58

mainly by Kaa, Phaleng, and Kgalagadi; Matshana and Maga-
lapye by Herero and Kgalagadi; Tswapong by Tswapong, with
smaller groups of Birwa, Rotse, and Seleka; Bobonong by Birwa;
Mmadinare by Talaote, Seleka, and Tswapong; Tonota by
Khurutshe and Kalaka; BoKalaka by Kalaka; and BoTletle and
Matsha by many small groups of Kalaka, Khurutshe, Herero,
Kuba, Nabya, Subia, and Kgalagadi. The Sarwa are not confined
to any particular district, but are scattered widely over the whole
reserve.

The inhabitants of a small village generally belong to the same
tribal community. For administrative purposes they are held to
constitute a single ‘ward* (motse, Village’), under the leadership
and authority of an hereditary headman. The ward is a patrilineal
but non-exogamous body, most of whose members belong to the
family-group of the headman, but normally it also includes several
other families or family-groups 1 attached to him as dependants.
The bigger settlements all contain a number of wards, not neces¬
sarily of the same community. The village in such cases must be
regarded, not as a local unit divided for convenience into smaller
segments, but as a cluster of self-contained social groups inhabit¬
ing one centre. Within it each ward has its own hamlet, clearly
separated from the rest, and its own kgotla (council-place), where
lawsuits and other local business are dealt with. Altogether there
are some 300 wards in the tribe, of which no less than 113 are
located in Serowe. They vary considerably in size, but on the
average contain from 200 to 400 people each.

This grouping into wards, common to all the Tswana, explains
the facility with which immigrants or conquered peoples were
absorbed into the tribe. Single families or family-groups of
strangers were placed by the chief in some existing ward, i,e. under
the immediate control of a particular ward-head. A larger group
would be recognized as a separate ward in itself, with its leader as
headman, or divided into a number of wards, according to its size
and existing kinship or territorial organization. Every person in
the tribe must belong to a ward and, save in exceptional circum¬
stances, he must always live in the same place as his fellow
members.

1 A family-group ( kgotlana ) is a collection of households whose heads are all
descended in the male line from a common grandfather or great-grandfather.
The senior descendant in line of birth is the ‘elder* ( mogolwane ) of the group.

THE NGWATO OF BECHUANALAND PROTECTORATE 59

Among the Ngwato proper, as contrasted with the foreigners,
there are two main categories of wards, distinguished according
to their mode of origin. In the days when polygamy was still
practised, a chief would group his sons by one wife into a single
body, under the authority of the eldest, and attach to them some
people of common birth as servants. The present headmen of the
wards thus created, all descendants of former chiefs, are accord¬
ingly known as dikgosana, ‘royal headmen 5 (sing, kgosana, prince,
chieftain 5 ). 1 Each chief would also allocate the supervision of his
own cattle among several faithful commoners; each of whom, with
his immediate relatives and the cattle herds placed under him, thus
became the nucleus of a new ward. The headman of such a ward
is termed a tnotlhanka, ‘common headman (lit., servant). Many
wards, both Ngwato and foreign, have since their foundation
become subdivided, one or more family-groups seceding because
of internal disputes or some similar factor. Such offshoots might
then be recognized by the chief as new wards, and given land on
which to erect their own village or hamlet.

Within a ward each family-group manages its own affairs, under
the leadership of its elder, and settles by arbitration disputes
involving any of its people. But all members of the ward fall
under the general control of their headman. He allocates land
to them for residence, cultivation, and grazing, can freely com¬
mand their services for all public purposes, is their official repre¬
sentative and spokesman, and supports and protects them in their
dealings with outsiders. He must see that they carry out the
commands of his political superiors, and formerly also collected
the tribute they paid to the chief. He judges cases which the other
elders of his ward have not been able to settle, or which are beyond
their competence to try; and, unlike them, can impose fines and
thrashings as punishments. He also conducts various religious
and magical ceremonies on behalf of the ward as a whole, although
with the spread of Christianity his functions in this direction are
disappearing. His close paternal relatives, the elders of the
remaining family-groups, and any other men of repute and ability

1 Thus, the headmen of the Tshosa, Kgope, and Mauba wards are descended
from Chief Kesitilwe; the headmen of the Maeketso, Seetso, Mokomane,
Sedihelo, Tshweu, and Seiswana wards from his son, Makgasana; the headmen
of the Morwakwena and Rammala wards from the latter’s son, Molete; and
the headmen of the Ramere, Monageng, Mmualefe, Modimoeng, and Ramasuga
wards from Molete’s son, Mathiba.

AFRICAN POLITICAL SYSTEMS

in his ward assist and advise him; and during his absence, or after
his death, the man next to him in line of succession, normally his
eldest son by his first wife, automatically takes his place. 1 Matters
of importance are discussed at a general meeting of the men in
the ward, the opinions they express helping the headman to reach
a decision.

Where a ward has become subdivided, the headman of the
parent group is the first court of appeal from the verdicts of the
others. In the bigger centres outside Serowe, the ward-head senior
to the rest in birth is also the headman of the whole village. As
such, he can hear appeals from the verdicts of his colleagues, and
has an overriding authority in all other matters, e.g. the distribu¬
tion of land and the organization of collective undertakings. He
is also the medium through whom the chief communicates with
the inhabitants of the village.

The hereditary chieftain of each foreign community continues
to rule over his people according to their own laws and customs.
He controls the activities of his dependent ward and village head¬
men, tries cases they are unable to settle or appealed from their
verdicts, and deals with other matters pertaining to the community
as a whole. But he is himself subject to the authority of the chief
to whom he is responsible for the general order, peace, and good
government of his adherents, and to whom also there is an appeal
from his decisions.

Each community living outside tribal head-quarters was
formerly placed under the protection of some prominent Ngwato
motlhcmka (common headman) resident in the chiefs town. This
man, whose responsibilities were hereditary, was expected to keep
in touch with the people and their affairs, visit them periodically
to collect tribute for the chief, and while there try cases brought
to him on appeal. He also informed them of developments at
head-quarters, looked after them whenever they came there, and
transmitted their grievances to the chief. In time it became
evident, with the expansion of the tribe, that many subject com¬
munities lived too far away from head-quarters for this method

T * hereditary principle runs right through the Ngwato political system
Sn^S * at ^ministration of any group is vested not so muchTn
one particular person, as m the whole family of which he is the head- and that

he tair* 61 thiS present on any occasion when action must

people of the ioup. VlrtUe ° f *“* birthright > t0 exercise authority over the other

THE NGWATO OF BECHUANALAND PROTECTORATE 61

of supervision to be sufficiently effective. Kgama therefore devised
a more direct system of administration. He grouped the more
remote communities into the districts to which reference has
already been made, and In each placed a resident governor, usually
a member of his own family, but sometimes a prominent and
reliable common headman. His successors have continued and
extended this policy.

The governor Is accompanied to his district by his Immediate
relatives, who assist and advise him. His main duties are to
communicate the chief’s orders and messages to the people under
his control, hear appeals from the verdicts of their chieftains,
settle disputes between different communities, organize and direct'
local public undertakings, supervise the collection of hut-tax and
tribal levies, and advise the chief on local political and economic
conditions. All matters that he cannot himself settle he must
refer to the chief, to whom there Is also an appeal from his own
decisions. Should he abuse his authority, or otherwise prove
incompetent, he may be recalled by the chief, as has happened
on several occasions within recent years. Some other man is then
sent to take his place. Failing this, his appointment tends to be
permanent, and may even become hereditary, unless the chiel
sees reason to intervene. The ‘protector’ of a foreign community
continues to represent it at Serowe, and is still the medium
through whom its people must approach .the chief when they
come there to appeal against their district governor or for some
other official purpose. But he no longer visits them to claim
tribute, the collection of which was abandoned by Kgama, nor
does he try the cases in which they are involved. He is now little
more than their ‘consular agent’, his administrative duties having
been taken over by the district governor.

All the wards in the tribe, both Ngwato and foreign, are finally
grouped into four parallel ‘sections’ (dikgotla, sing, kgotla), named
respectively, after the leading ward in each, Ditimamodimo,
Basimane, Maaloso, and Maalos&-a-Ngwana. The origin of this
grouping is not clearly known, but it seems to have arisen from
the practice of giving the heir to the chieftainship a large cattle
post and creating a new ward of commoners to look after it. The
Ditimamodimo cattle post, and the ward created for and named
after it, are said to have been established by Chief Molete for his
son, Mathiba; the Basimane by Mathiba for his son, Kgama I;

62

AFRICAN POLITICAL SYSTEMS

the Maaloso by Kgama I for his son, Kgari; and the Maaloso-a-
Ngwana for Kgari’s son, Sekgoma I (father of Kgama III). The
heir, on coming to the chieftainship, relied considerably upon the
assistance and support of the retainers thus attached to him, and
as a rule placed under their supervision all the wards originating
in his reign. .The creation of new sections ceased in the reign of
Sekgoma I, who consolidated the system into its present form.
All wards since created were placed, at the discretion of the chief,
within one or other of the existing sections.

In Serowe each section has its own quarter of the town, within
which its component wards are located. The headman of the
nuclear ward in each is also headman of the whole section.
Associated with him in the administration of its affairs is the
senior kgosana of the section, i.e. the royal headman most closely
related to the chief. 1 These two men act together as the senior
judges of the section. They hear all cases referred to them directly
or on appeal from the other ward-heads of their section in Serowe,
or from a district governor where members of their section are
involved; and until they have done so the case cannot come before
the chief. As heads of the section they also command much
greater influence in the tribe than do the other ward-heads. The
chief frequently consults them on questions of tribal policy, and
may depute them to deal on his behalf with important adminis¬
trative matters. They speak for their people at tribal meetings
where the opinions of each section are separately canvassed, and
can also summon meetings of their own to discuss matters of
sectional or tribal interest. The section to which the chief himself
belongs and, as already indicated, different chiefs have belonged
to different sections—is generally regarded during his reign as the
ruling section of the tribe. He relies more upon its senior headmen
for help in formulating tribal policy, and looks to them particularly
for support in all his plans and undertakings.

1 The descendants of Sekgoma I, who are all more closely related to the chiei
than other royal headmen, are associated with him in the central government
of the tribe, and do not take part in the sectional administration. The next
senior royal headman in each section is regarded as its responsible royal head-
man J Jkf P°sition is held in Ditimamodimo by the headnian of Ditlharapa
ward, m Basimane by the headman of Tshisi ward, in Maalosd by the headman
of Maboledi ward, and in Maaloso-a-Ngwana by the headman of Menyatso
ward, all of whom are descendants of Kgama I.

THE NGWATO OF BECHUANALAND PROTECTORATE 63

II. The Administrative System

The administrative system just described, and the social
organization upon which it is directly based, separate the members
of the tribe into groups distinguished one from another by local
powers and loyalties. In certain respects each section, district,
community, village, ward, and family-group is independent of the
rest, managing its own affairs under the direction of a recognized
head whose authority extends over almost every sphere of public
life. The many communities of which the tribe is composed
frequently also differ from one another in language, custom, and
tradition, and so have not even a common cultural background.
We must now consider how all these groups are welded together
and given a solidarity and cohesion enabling the tribe as a whole
to present a united front to the outside world in defence or
aggression, maintain law and order and adjust disputes between
the members of one group and another, and carry on large
collective undertakings.

One of the mechanisms through which this is achieved is the
administrative hierarchy into which the various forms of local
authority are graded. In Serowe, as we have indicated, the elder

of a family-group is directly subordinate to his ward-head. The
ward-head, in turn, is subordinate, either directly or through the
headman of the ward from which his own is derived, to the head¬
man of the nuclear ward in the same section. 1 The sectional
headman, finally, is subordinate to the chief. In the outlying
districts, the ward-head is, either directly or through his village
headman, subordinate to the hereditary chieftain of his tribal
community. The latter, again, is subordinate to the district
governor, who is finally subordinate to the chief. In each case
the superior authority has powers overriding those of the lesser
authorities in his own group. The latter must obey his commands
and carry out his instructions, must refer to him all cases which
they are unable to settle or with which they are not competent
to deal, and there is an appeal to him from all their judicial and
executive decisions.

The judicial system is fundamentally the same for all courts.
The victim of a civil wrong, such as breach of contract, seduction,

1 In the foreign communities living in Serowe, the local chieftain of each is
intermediate between the ward-head and the sectional headman.

64 AFRICAN POLITICAL SYSTEMS

adultery, trespass, damage to property, theft, or defamation, may

either pass it over or, through the elder of his family-group, try to
arrive at an agreed settlement with the offender. Failing this, he
takes the matter to the court of defendant’s ward-head. Crimes,

such as offences against political authorities acting in their official
capacity, breaches of the laws decreed by the chief, rape, assault,
homicide, and sorcery, can never be compounded, but must always
come to trial. All trials are heard in public, and any member of
the tribe has the right to attend and take part in the proceedings,
no matter in what court they are held. The parties concerned and
their respective witnesses are heard in succession, listened to
intently and uninterruptedly, and closely questioned by the people
present. The judge then throws the matter open for general dis¬
cussion, and the merits of the case are publicly argued by those
wishing to do so. This is one of the principal functions of his
personal advisers. Finally he sums up, in the light of the opinions
thus expressed, and either pronounces his verdict or, if he feels
that the case is too important or difficult, refers it to the court of
his political superior. If either party is dissatisfied with the ver¬
dict, he can likewise appeal against it. The case is then heard
again from the very beginning at the superior court, pending whose
decision action is suspended. A case originating in a family-group
may thus pass through three or four grades of intermediate court
before ultimately reaching the chief.

In effect, therefore, the existing social and territorial organiza-
tion is used to delegate matters of more purely local concern to
the subordinate authorities, but the government of the tribe as a
whole is concentrated in the hands of the chief and his personal
advisers. The chief is the central figure round whom the tribal
life revolves, and through whom the activities of the tribe are
ordered and controlled. He is at once its ruler and judge, maker
and guardian of its law, and director of its economic life, and in
the olden days was also its leader in war and its principal priest
and magician. It is primarily through allegiance to him that the
members of the tribe express their unity. He calls and signs him¬
self 6 Kgosi ya baNgwato’, 'Chief of the Ngwato people’; he is
ceremonially addressed, by the personification of the tribal name,
as MoNgwato ; the tribe itself is named after his ancestor, the
legendary founder of the royal line; and he is its representative
and spokesman in all its external relations. Like his subordinate

THE NGWATO OF BECHU AN ALAND PROTECTORATE 6s

authorities, he is assisted in his work by his close paternal relatives
and other personal advisers. But on occasion he also invokes the
aid of wider councils and other forms of assembly drawn from the
tribe as a whole, and so binding its people together still further. 1
Sometimes he consults the headmen alone, but more frequently he
summons all the men of Serowe, or even of the whole tribe, to a
meeting where public business is discussed, while through the
regimental organization he may bring them together for work or,
in the olden days, for war. These forms of assembly cut across the
parochial loyalties of ward, village, and community, and so are
among the most conspicuous means of uniting the members of
the tribe.

This system of central administration still prevai s, but since
the establishment of the Protectorate it has been somewhat
modified. The European Administration has not only limited the
powers of the chief and other tribal authorities, and' altered the
structure of their courts; it has also introduced its own govern¬
mental institutions. The Ngwato are now ruled by both European
and Native authorities, and the latter occupy the subordinate
position. It will be as well, therefore, to review briefly the part
played by the European Administration in the regulation of tribal
affairs before we proceed to discuss in more detail the past and
present powers and functions of the chief and his councils.

Bechuanaland Protectorate, together with Basutoland and
Swaziland, is under the general legislative and administrative
control of a High Commissioner responsible to the Secretary of
State for Dominion Affairs in Great Britain. The Territory itself
is directly governed by a Resident Commissioner, with head¬
quarters at Mafeking. 2 The Ngwato Reserve, one of the twelve
administrative districts into which it is divided, is under the
immediate jurisdiction of a District Commissioner stationed at
Serowe. He is assisted to maintain law and order, and carry out
his other duties, by a small body of police and a few subordinate
European and Native officials. Some technical officers represent¬
ing the medical, agricultural, and veterinary branches of the

1 For a more detailed sketch of these advisers and councils, see below, pp. 71-2.

2 Mafeking, oddly enough, is located in the Union of South Africa, and not
in Bechuanaland Protectorate itself, whose southern border is twelve miles
north of the town. This anomalous position is a survival of the days (1885—95)
when what is now British Bechuanaland, in the northern Cape Colony, was
included in the Protectorate.

AFRICAN POLITICAL SYSTEMS

Administration also live and work in the Reserve. Its total
European population, comprising not only Government officials,
but the London Missionary Society staff, traders, railway em¬
ployees, and others, was 376 in 1936, men, women, and children
included.

The general policy of the Administration, in dealing with the

Natives, has been ‘to preserve the tribal authority of the chiefs and
the laws and customs of the people, subject to the due exercise of
the power and jurisdiction of the Crown, and subject to the
requirements of peace, order, and good government’. It early
took away from the chiefs the right to make war or enter into
independent political agreements, removed cases of murder and
culpable homicide, as well as all cases involving Europeans, to
the jurisdiction of the European courts, allowed appeals from the
verdicts of the chiefs in very serious cases, defined the boundaries
of the tribal territories, and imposed a regular annual tax upon all
adult male Natives. But for many years thereafter the manner in
which the chief administered the tribe was not a matter of Gov¬
ernment concern, except when it led to open trouble, and in most
cases the tendency was to support him as far as possible in his
dealings with his people. More recently, however, the chief’s
judicial powers were further curtailed. In 1919 provision was
made for hearing appeals from his verdicts in any type of case,
civil or criminal, by the establishment of a combined court presided
over jointly by him and by the local District Commissioner. In
1926 divorce proceedings between Natives married according to
European civil law were also brought under the jurisdiction of the
District Commissioner’s Court, and in 1927 the trial of alleged
sorcerers was removed from the tribal courts when the imputation
or practice of ‘witchcraft’ was made a statutory offence. These
encroachments did not pass unchallenged. Tshekedi, who became
Regent in 1926 for his fraternal nephew Seretse (son of Kgama’s
son, Sekgoma.II), joined the chiefs of several other tribes in pro¬
testing against them, but they were nevertheless made law.

Finally, in 1934, the powers of the chief were for the first time
clearly defined, and the status and powers of the tribal courts
regularized. The Native Administration Proclamation (No. 74 of
I 934 ) specifies the rights, powers, and duties of the chief and
other tribal authorities, makes succession to and tenure of the
chieftainship subject to the approval of the Administration, which

THE NGWATO OF BECHUANALAND PROTECTORATE 67

has the power to pass over an unsuitable heir or suspend an in¬
competent or otherwise unsatisfactory chief, provides machinery
whereby the tribe can depose a chief, makes conspiracy against
the chief a statutory offence, and establishes a formal Tribal
Council to assist him in the execution of his duties. The Native
Tribunals Proclamation (No. 75 of 1934) removes from the
jurisdiction of the tribal courts all cases in which the accused is
charged with treason, sedition, murder or attempted murder,
culpable homicide, rape or attempted rape, assault or intent to do
grievous bodily harm, conspiracy against the chief, and a variety
of statutory offences. All other cases, both civil and criminal, in
which Natives only are concerned, can still be tried by the tribal
courts according to Tswana law and custom. But, in place of the
many grades of court in the tribal system, the Proclamation
recognizes only three whose decisions are legally binding. These
are styled Junior and Senior Native Tribunals and the Chief’s
Tribunal respectively. From the last there is an appeal to the
District Commissioner’s Court and thence, under certain condi¬
tions, to the Special Court of the Protectorate. The Proclamation
further defines the constitution of each tribunal, laying down
that it shall have a limited membership appointed by specified
tribal authorities, provides for the keeping, in all cases tried, of
written records open to inspection by the District Commissioner,
severely curtails the forms of punishment that may be imposed,
and lays down various rules to govern procedure.

These Proclamations, first drafted in 1930, had been frequently
and fully discussed by the Administration with the chiefs and
their tribes. The necessity for them had become more and more
evident. As the old chiefs died, they were succeeded by young men
educated for the most part in schools outside the Protectorate, and
so cut off from adequate first-hand experience of tribal government
and jurisdiction. Instances multiplied of drunkenness and irre^
sponsibility, neglect of duty, misappropriation of hut-tax and
other tribal moneys, and of serious internal disputes, all calling for
stronger Administrative control; while increasing educational,
veterinary, and agricultural development made it still more
desirable to define clearly the relations between the Administra¬
tion and the tribes. Tshekedi opposed the Proclamations from the
beginning, and even after their promulgation did not put them
into force. Finally, in December, 1935, he and Bathoen, Chief of

68

AFRICAN POLITICAL SYSTEMS

the Ngwaketse, took legal action against the High Commissioner.
They claimed that the Proclamations were of no legal force and
effect, because in them the High Commissioner inter alia both
altered certain Native laws and customs, thus exceeding the
powers conferred upon him, and violated treaty rights reserved to
their respective tribes. The case was heard in the Special Court in
July, 1936. Judgement was reserved and delivered in November.
The court found in favour of the High Commissioner, holding
that, while the Proclamations undoubtedly altered existing Native
law and custom, he had acted within his powers; while on the
question of treaty rights a ruling by the Secretary of State, that the
power of the Crown in Bechuanaland Protectorate ‘is not limited
by Treaty or Agreement 5 , was taken as conclusive. The Proclama¬
tions have therefore become binding upon the Ngwato, and
Tshekedi has now begun to apply them. It is too soon yet to
determine in detail how they affect the tribal administration, but
some attempt will be made below to indicate their main tendencies.

Ill . Powers and Authority of the Chief

The general effect of the Proclamations, and of previous
developments, is that although the chief and his councils still
administer the affairs of the tribe, the supreme political and
judicial authority in the Reserve is now the Administration, acting
through the District Commissioner. Nevertheless, the chief, as
head of the tribe, is always treated by his subjects with a good deal
of outward respect. His installation and marriage are occasions of
great public festivity, and his death evokes universal mourning.
His household is usually far larger and more elaborate than those
of ordinary tribesmen. In the olden days he always had many
wives, retainers, and serfs, and even to-day the number of his
direct dependants is very great. He receives various forms of
tribute from his people, imposes levies upon them, and claims free
labour from them for personal as well as tribal purposes. He has
the first choice of land for his home, fields, and cattle-posts; he
and his family formerly took precedence in the tribe in matters
of ritual; and he alone has the right to convene full tribal meetings,
create new regiments, arrange tribal ceremonies, and, in the olden
days, impose the supreme penalties of death and banishment.
Failure to comply with his orders is a penal offence. All other
offences against him are generally punished more severely than if

THE NGWATO OF BECHUANALAND PROTECTORATE 69

committed against ordinary tribesmen. Disloyalty and revolt
against his authority are major crimes, punished as a rule, in the
olden days, by death and the confiscation of the culprit’s property,
and nowadays by banishment or some other penalty inflicted
directly by the Administration.

Despite the curtailment of his traditional rights and powers,
the chief still plays a very prominent part in the government of
the tribe. He decides upon questions of tribal policy, and can
make regulations binding upon his subjects. 1 He determines upon
and arranges for the execution of all important public works. He
supervises the conduct of his subordinate authorities, and in case
of extreme incompetence or abuse of office can replace his district
governor by some other man, or depose the hereditary head of
a ward or community, whose duties then fall to the man next in
line of succession. Much of his time is spent daily in his kgotla
(council-place), where he listens to news, petitions, and com¬
plaints from all over the tribe, and gives orders for whatever
action may be necessary. He must protect the rights of his sub¬
jects, provide justice for the injured and oppressed, and punish
wrongdoers. Before the establishment of the Protectorate, he was
the supreme judge of the tribe. This function has been taken
over by the Administration, but his court is still the highest Native
tribunal, to which a right of appeal lies from the verdicts of the
others. In the olden days, such serious offences as treason,
homicide, assault, rape, and sorcery could be punished only by
him and his court. They have now been removed to the juris¬
diction of the European authorities, but he still adjudicates over
all other breaches of tribal law and has greater punitive powers

1 Legislation of this sort does not seem to have played a conspicuous part in
the old tribal life, the chiefs being more concerned to maintain the existing law
than to alter it. In more recent times, owing to the new conditions created by
contact with Europeans, legislation by the chief has become a frequent occur¬
rence. Kgama introduced so many changes that the people themselves some¬
times distinguish between ‘traditional Native law’ and ‘Kgama’s law’, although
both are equally binding upon them. Among other things, he accepted Chris¬
tianity as the official religion of the tribe; abolished hogadi (bride-wealth), and
discouraged polygamy among the Ngwato proper, although not among his
subject communities; abolished most of the old tribal ceremonies; prohibited
the .sale, manufacture, and drinking of Kafir beer and other intoxicating liquors
throughout his Reserve; regulated the sale of com and breeding-cattle to
European traders; modified the customary rules of inheritance, so that daughters
should also inherit cattle; prohibited the movement of wagons through the
villages on Sunday; and officially protected certain big game animals.

70

AFRICAN POLITICAL 'SYSTEMS

than the judges of the lesser courts. He controls the distribution
and use of the tribal land, organizes large collective hunts, and
regulates trade relations with outsiders and the time for sowing
and harvesting crops.

With the extension of European control, the chief's adminis¬
trative duties have greatly increased. He is responsible to the
Administration for maintaining law and order in the tribe, pre¬
venting crime, and collecting hut-tax and other dues. He must
carry out all orders and instructions issued to him, and render
any assistance required from him, by responsible officers of the
Government; and he is expected to co-operate with the District
Commissioner and other members of the Administration in all
sorts of political, economic, social, and educational schemes and
developments. His formerly undivided control over every aspect
of public life has thus been diffused through various Government
departments with superior authority. He must, further, deal with
the traders, missionaries, would-be concessionaires, and other
Europeans living in his Reserve, visiting him, or writing to him;
and must issue to his subjects receipts for tax payments, permits
for the sale of cattle and com, and passes to leave the Reserve on
v'sits to the Union or in search of work. The complaint some¬
times made against Tshekedi that he is an ‘office chief rather
than a ‘kgotla chief indicates sufficiently the change in adminis¬
trative methods that all this has entailed.

Formerly the chief was also the head of the tribal army. He
organized military expeditions, often accompanying them himself,
performed the necessary war magic, and disposed of the prisoners
and loot. With the abolition of inter-tribal warfare under Euro¬
pean government, all this has disappeared. Formerly he also
organized the great tribal ceremonies upon which the welfare of
his people was held to depend. But Kgama, from the time he was
converted, fought against these ‘heathen’ practices, and after he
became chief deliberately ceased to observe them. Since they
could not be celebrated without his authority and participation,
they have altogether died away, and with them his functions of
tribal priest and magician. The Ngwato are now officially a
Christian tribe, acknowledging the ritual leadership of the local
missionaiy. The latter is therefore to some extent a rival authority,
whose claim to the allegiance of the faithful has at times brought
him into conflict with the chief. The monopoly given by Kgama

THE NGWATO OF BECHUANALAND PROTECTORATE 71

to the London Missionary Society in his Reserve has also been
a source of subsequent trouble with immigrant communities
professing some other variety of Christian faith.

On all questions of tribal policy the chief is expected to consult
with his immediate paternal relatives. The sectional headmen
and other prominent local leaders are also held to be among his
rightful advisers. When any matter of outstanding importance
arises, he further summons a general meeting of headmen, and
so obtains the views of all the important men in the tribe before
taking any action. Except for such meetings, held very infre¬
quently, the chief’s advisers were until recently not organized
into a definite body with limited membership. He consulted them,
severally or collectively, whenever he wished, and varied them
according to the issues involved. He relied only upon the men
whom he could trust, and ignored others, however important
their standing, who were openly hostile to him or whom he
regarded with suspicion.

In November, 1925, however, immediately after the death of
Sekgoma II, the tribe, at the suggestion of the Resident Commis¬
sioner, elected a Council of thirteen to assist the temporary acting
chief (Gorewang, son of Kgama’s brother, Kgamane). The
Council was not at all popular, the people feeling that it was an
Administrative device to undermine the chief’s power by limiting
his freedom of action; and Tshekedi, when he returned from
school in February, 1926, to take over from Gorewang, success¬
fully insisted upon its abolition. The Native Administration
Proclamation has now reintroduced the idea-of a well-defined
Tribal Council. It requires the chief to nominate publicly, and
with the approval of the tribe, the men entitled to advise him as
councillors, and directs him to consult with them in the exercise
of his functions. A councillor’s tenure of office is subject to the
discretion of the Resident Commissioner, and not of the ^ chief,
whose powers are thus limited in another direction. This was
one of the features in the Proclamation against which Tshekedi
most strenuously protested, maintaining that such a limited body
as it proposed to set up was alien to the tribal system; and the
court, in giving judgement, found that it was undoubtedly a
departure from the traditional method of government.

This is not the only change that has taken place. During and
since the time of Kgama, the chief has come to rely upon European

7* AFRICAN POLITICAL SYSTEMS

advisers apart from Government officials. He often consults
the local missionaries on many aspects of tribal life remote from
religion, seeks advice from traders on matters of finance, and
frequently employs the special services of' lawyers in Mafeking
and Cape Town. Tribal policy is therefore no longer determined
merely by the chief and his traditional Native advisers; it is
moulded also in some degree with the aid of various interested
European personalities, not always regarded with favour by the
Administration.

All matters of tribal policy are dealt with finally before a general
assembly of the adult men in the chiefs kgotla (council-place).
Such meetings are very frequently held, at times almost weekly.
Normally only the men present in Serowe attend them, any
decisions reached being communicated to those in other parts
of the Reserve through the district governors and other local
authorities. But on important occasions the people of the districts
are also summoned, and the question at issue is debated by the
tribe as a whole. Among the topics discussed in this way are tribal
disputes, quarrels between the chief and his relatives, the imposi¬
tion of new levies, the undertaking of new public works, the
promulgation of new decrees by the chief, and the relations
between the tribe and the Administration. The decisions made
are generally the same as those previously reached by the chief
and his personal advisers, who as leaders of the tribe can sway
public opinion; but it is not unknown for the tribal assembly to
overrule the wishes of the chief. Since anybody may speak, these
meetings enable him to ascertain the feelings of the people
generally, and provide the latter with an opportunity of stating
their grievances. If the occasion calls for it, he and his advisers
may be taken severely to task, for the people are seldom afraid to
speak openly and frankly. The Administration has contributed
greatly towards the retention and present vigour of these assemblies,
by making a practice of getting its officers to discuss with the tribe
in kgotla developmental schemes, new or projected laws, and other
matters affecting the relations between the two bodies.

In his judicial and administrative capacities, the chief again
relies primarily upon his personal advisers. They help him hear
and judge cases at his court, although, as we have seen, any other
member of the tribe may attend and take part in the proceedings.
They also act as his state messengers on important occasions, he

THE NGWATO OF BECHUANALAND PROTECTORATE 73

selects his district governors from among them, and he may
delegate them to try cases on his behalf, supervise the execution
of his verdicts, and undertake other duties of a similar kind. If
he is ill, or away from head-quarters, his heir, if old enough, or
else some other very close paternal relative, acts as his deputy.
For such minor tasks as carrying ordinary messages, he uses any
tribesman at hand, and he also has a few official policemen of his
own, who see that his decrees are enforced and act as messengers
of his court on most routine occasions. Within recent years he
has also begun to employ paid secretaries and other assistants to
handle his correspondence, collect tax, issue passes and receipts,
and attend to other routine business of the same kind. The chief’s
principal secretary, owing to his access to all confidential docu¬
ments and the close association in which he works with the chief,
has become one of the strategic men in the tribal administration;
and many royal headmen regard with resentment Tshekedi’s
employment of a Kalaka headman in this capacity, whereas
Kgama and Sekgoma II both relied upon very close relatives.

Major enterprises are organized through the system of age-
regiments (mephaio) into which the whole tribe is divided. A
regiment consists of people of the same sex and of about the same
age, and every adult in the tribe must belong to one. The regi¬
ments are formed at intervals of several years apart, when all the
eligible boys or girls, as the case may be, are grouped together into
a single body. In the olden days they simultaneously went through
an elaborate series of initiation ceremonies, but nowadays they are
simply called together and told the name of the new regiment to
which they henceforth belong. Each regiment of men is com¬
manded by a member of the chief’s own family (his brother, son,
or fraternal nephew); while each group of men in it belonging to
the same section, district, community, village, or ward is led by
some similarly close relative of the appropriate headman. The
headman himself leads the men of his group in his own regiment.
The heir to the chieftainship commands his own regiment during
the lifetime of his father, but on succeeding to office ceases to do
so, the effective leadership passing to the member of the royal
family next in rank. The women’s regiments are organized along
similar lines.

The men’s regiments originally constituted the tribal army in
the event of war, and were used at other times as a labour force.

AFRICAN POLITICAL SYSTEMS

74

Their former function has now disappeared, but the latter has been
greatly intensified owing to the new forms of activity resulting
from the introduction of Western civilization. They can be called
up, whenever the chief wishes, for such tasks as making dams,
rounding up stray cattle, building schools and churches, hunting
beasts of prey, cutting down bushes in the chief’s fields, building
his huts and cattle-kraals, making roads and aerodromes, cutting
boundary paths, preparing agricultural showgrounds, rounding up
offenders against the law, and escorting distinguished visitors.
The women’s regiments, again, are employed to put up the walls
and thatch the roofs of the chief’s huts, draw water for any royal
or tribal work, get wood for the chief’s wife, clean the village, fetch
earth and smear the walls and floors of the chief’s homestead, and
weed his wife’s fields. Only the chief can mobilize a whole regi¬
ment for work, but district governors and other headmen may
summon their own followers by regiments to perform purely
local tasks of a similar nature.

Regimental labour is both compulsory and unpaid, and failure
to answer a summons to work can be punished by a fine or thrash¬
ing. Within recent years, with the spread of education, on the one
hand, and the increased burden of work, on the other, complaints
have become common about the brutal methods sometimes use!
for rounding up defaulters and stragglers, and about the hardships
and losses imposed by such calls upon people engaged in work of
their own. These were among the grievances mentioned in a peti¬
tion lodged against Tshekedi in 1930 by eight members of the
tribe, and largely substantiated in the Administrative inquiry that
followed. As a result of such abuses, present also in other tribes,
the Native Administration Proclamation has made it illegal for the
chief to exact free labour from his people except for certain clearly
specified purposes.

IF. Rights and Responsibilities of Chieftainship

The authority of the chief is derived in the first place from his
birthright. The chieftainship is hereditary in the male line, passing
normally from father to son. In the days when polygamy was
practised, the rightful heir was always the eldest son of the
"great’ wife, i.e. of the woman first betrothed to the chief. Failing
a son in her "house’, the eldest son of the wife next in rank suc¬
ceeded. Sometimes, however, there were disputes regarding the

THE NGWATO OF BECHUANALAND PROTECTORATE 75

■lative status of a chief’s wives, with the result that there would be
tore than one claimant to the succession. Sometimes, too, a
;gent acting for a minor chief would attempt to usurp t e c 1
tinship permanently, or an ambitious prince would revolt agamst
n unpopular ruler. But in every instance the rival claimant was
imself of royal birth. No man who was not a senior member of
tie ruling family could ever hope to become recopized as chi .
Kgosi he hgosi ha a tsetszve’, say the Ngwato (A chief is chief
lecause he is born (to it)’). Since, as we have seen, the whole soci
ystem of the tribe emphasizes the principle of hereditary rank,
ince every local authority (except the district governor, a re a
ively recent creation) owes his position to the fact that h
egitimate heir of his predecessor, the chief’s status as head of the
elding family in the tribe is sufficient m itself to secure for hi
the respect and obedience of his people. /O / H .

Under the European Administration, hereditary succession
the chieftainship still prevails, but with certain mcdifications^ It
was early laid down that no chief could exercise J^fficUonover
his tribe unless he had been recognized by the High Cornmi
sioner and confirmed by the Secretary of State ' In -®! Ct ’ the
made no difference to the succession, as among the Ngwato the
rightful heir was always accepted by the Administration as chief.
Under the new Proclamations, however, the Admimstranon as
assumed the right of refusing to recognize or confirm the tor
chief, if he appears, after public inquiry, not to be a fit and prop
person to exercise the functions of the chieftainship . So far no
occasion has occurred for such refusal. But the possffiiffiy it
embodies of passing over the legitimate successor m favour of
a junior member of the royal family introduces a
foreign to Ngwato law. Formerly such an event could only
occur as the result of deliberate usurpation, resultmg probably

m By virtue of his des(lt^t 5 /hiif was formerly the lbAbet ™^
his people and the spirits governing their welfare His dead
ancestors were held to afford supernatural protection and assistance
Tffie people they had once ruled, and on all
he would sacrifice and pray to them on behalf of thetnbeTh
role he thus played as tribal priest, a role which only he, as senior
descendant of the ancestral gods, could fill, he^s to exp am
great reverence in which he was always held by his

“ 6 AFRICAN POLITICAL SYSTEMS

Kgama’s acceptance of Christianity, however, deprived the chief¬
tainship of almost all its ritual significance, and so of a powerful
sanction for its authority. The whole tribe is by no means firmly
attached to Christianity, and many of the old practices are still
carried on more or less surreptitiously, especially among the
subject communities. But the people no longer look to the chief
for spiritual benefits, and certainly do not accord him the same
pious reverence which contemporary observers show that Sekgoma
I received as tribal priest and magician.

The chief’s hold over his people was formerly also strengthened
by marriage. His own wives were drawn mostly from the families
of his close relatives, other influential headmen, and the chiefs of
neighbouring tribes. Since in the Tswana social system a man’s
maternal relatives are expected to be among his strongest sup-
porters, the chief made sure in this way that his sons, particularly
the heir, should always have a powerful backing. At the same time,
the practice of polygamy gave ample scope for intrigue. The rela¬
tives of each wife watched jealously over her interests, and did their
utmost to further the fortunes of her sons. As a result many feuds
arose which still play their part in tribal politics. Kgama’s aban¬
donment of polygamy did away with these sources of possible
conflict but also deprived the chief of the political advantages he
“; ld J b f n by J^iciously selecting his wives. He is, however,

11 able to consolidate his hold over the tribe by marrying off his
sisters and daughters to headmen with large followings. Kgama

STlo TTl 1C ? “ t0 r e ? Ct ty marryil * of ** daughters to
the local chieftains of the Khurutshe, Kaa and Talaote respec¬
tively, and three others to prominent royal headmen, two of whom
were irnpomnt enough to be appointed district go^mors.

,A,! C t 1Cf S *T er is further dependent upon the uses to which

«b p „« r Wd of the ,ribe - he

an , ‘ . fr ® hl f sub J ects in c om, cattle, wild animal skins, ivory
d ostrich feathers, retamed most of the cattle looted in war arid
kept all unclaimed stray cattle and part of the fines impost fc Ss

entire prope^of^ T* ° f He C ° uld also c ^cate the

for an/othefleriousoff^ 11 C ° T nSp '™ g a S ainst ^ or banished

^well^ 1 S^uZ^ S’Ws pipit SpmoS

se™ dir^ffio

THE NGWATO OF BECHUANALAND PROTECTORATE 77

work. Foremost among them were the batlhanka (common head¬
men) whom, as already noted, he put in charge of his cattle and
other servants. The latter were drawn mainly from the ranks of
the Kgalagadi and especially the Sarwa, who occupied the position
of Serfs. At first they merely hunted for him, the skins and other
spoils they gave him forming an important part of .his income;
but under Kgama they were gradually taught to herd cattle also and
to carry out menial household tasks.

Owing to the wealth he thus accumulated, the chief was always
the richest man in the tribe. He was, however, expected to use his
property, not only for his own benefit, but also for the tribe as a
whole. He had to provide beer and meat for people visiting him,
assisting at his kgotla, or summoned to work for him; reward with
gifts of cattle and other valuables the services of his advisers,
headmen, warriors and retainers; and, in times of famine, supply
the tribe with food. ‘Kgosi ke mosadi wa morafe it was said (‘The
chief is the wife of the tribe’, i.e. he provided the people with sus¬
tenance). One quality always required of him was generosity, and
much of his popularity depended upon the manner in which he
displayed it. Kgama is still gratefully remembered as an extremely
liberal chief, who not only imported com for his people in times of
scarcity—on one occasion to the value of between £2,000 and
£3,000—but also bought many wagons, ploughs, guns, and horses,
which he distributed among them, while in several instances he
paid large sums of money to free some of them from debt.

In this connexion, the relationship between the chief and his
batlhanka deserves special mention. These men, as we have seen,
were placed as common headmen in charge of the chief’s cattle-
posts. The cattle entmsted to them were the hereditary property
of the chieftainship, so that the batlhanka were always attached to
the ruling chief himself. Each motlhanka was required to provide
the chief’s household with milk and meat from the cattle under his
care, and to come with his followers to perform such other work
as might be demanded of him. In return, he could use the cattle
as he pleased: he kept the rest of their milk, slaughtered a beast
whenever he wished, paid bogadi (bride-wealth) for his sons out
of them, and exchanged them for other commodities, while on his
death they passed to his children. He was also given the Sarwa
inhabiting the region where the cattle grazed, and kept most of
their hunting tribute for himself. The chief, however, had the

AFRICAN POLITICAL SYSTEMS

ultimate claim not only to these cattle (known as kgamilS , ‘milk-
pail’, cattle), but to everything else acquired by the motlhanka.
The entire property of a motlhanka was regarded as kgamelo ; and
since the chief could withdraw his kgamSlo whenever he wished,
he could at any time ruin the holder.

This system obviously bound the common headmen very closely
to the chief. They were dependent upon him for their entire
subsistence, and therefore of necessity were among his most loyal
adherents. They were looked upon as and proved themselves his
strong supporters against the intrigues of his uncles and brothers,
and, since they could never be potential rivals for the chieftain¬
ship, he came to rely more and more upon them, until in time they
became the most influential group assisting him to govern the
tribe. The chiefs own relatives, it may be added, were never
entrusted with kgamila cattle, although he was expected to
provide them also with cattle and servants of their own.

To-day the chief no longer receives most of the tribute formerly
paid to him, Kgama having abandoned its collection. Cattle¬
raiding, at one time an important source of wealth, disappeared
with the abolition of inter-tribal warfare. A still more drastic
change occurred about 1900, when, as a result of disputes with his
son Sekgoma regarding the ownership of kgamilo cattle, Kgama
declared that henceforth all such cattle would be regarded as the
private property of their holders, and that he as chief renounced
all rights over them. On the other hand, the coming of Western
civilization provided him with new sources of income, such as the
aimual subsidy until very recently paid for mining concessions in
the Reserve, the annual commission paid by the Administration
on the amount of hut-tax collected from his people, and the cash
levies he imposed at various times to finance public undertakings

P l? hls d f bts ; AI1 this mon ey was formerly controlled and
used by him as he pleased. The present tendency, quite recently
initiated by the Administration, is to divert the money and other
mvenue mised from tax, court fines, levies, and simL soumes

S^and ^ ^ 2part fr0m the Chief ’ s P ersonal ^come.

restT/th! 3 aSS1S ,! Sare paid annuaI salarie ® out of the fund, the

a^suktf th° n T tem ? USed /° r s P ecificaII y tribal purposes. As
. , 1 of these imitations of his income, coupled with the rela

srir- d ; f iiving he ™ st -- d ^S

the chief has been deprived to a great extent of his traditional role

THE NGWATO OF BECHU AN ALAND PROTECTORATE 79

t>f repository of wealth apd dispenser of gifts, and so of yet
.another important sanction for his authority.

In the last resort, the power of the chief rests upon his personal
character. Political life is so organized that effective government
can result only from harmonious co-operation between him and his
people. l Kgosi ke kgosi ka morafe\ says the proverb (‘The chief is
chief by grace of the tribe 5 ). Even in the olden days, despite the fact
that control over almost every aspect of tribal life was concen¬
trated in his hands, and that his power was in consequence very
considerable, he was seldom absolute ruler and autocratic despot.
In order to get anything done, he must first gain the support of
his advisers and other headmen, who thus played an important
part in restraining his more arbitrary impulses. Any attempt to
act without them would lead to obstruction, if not to open revolt.
A forceful and energetic man like Kgama could succeed in domina¬
ting his subjects and ruling in effect as a dictator— but at the cost
of some painful disputes with his closest relatives. On the other
hand, a weak chief like his son, Sekgoma II, became the virtual
puppet of certain royal headmen, whose influence over him
created an opposition which led to much subsequent trouble.

Tribal politics is in fact made up to a considerable extent of
quarrels between the chief and his near relatives, and of their
intrigues against one another to command his favour. As we have
seen, they are entitled by custom to advise and assist him in his
conduct of public affairs, and they actively resent any failure on
his part to give them what they regard as their due. Since the
time when Kgama, as a newly converted Christian, incurred the
active hostility of his father (Sekgoma I), the tribe has been rent
with dissensions in the royal family. Kgama, after expelling his
father from the chieftainship, quarrelled successively with his
own brother, Kgamane, his half-brothers, Mphoeng and Rradit-
ladi, and his son, Sekgoma, all of whom were banished from the
Reserve with their followers; while his increasing reliance first
upon his son-in-law, Ratshosa, and then upon the latter’s sons,
John and Simon, alienated other royal headmen. Sekgoma,
restored to favour after twenty years, became Chief in 1923 ; and
soon, instigated mainly by the Ratshosa group, banished Phethu,
son of Mphoeng. Phethu was pardoned shortly afterwards, and
on his return began plotting against the Ratshosas. TshekedPs
accession in February, 1926, gave him his opportunity, and a

80

AFRICAN POLITICAL SYSTEMS

series of steps was taken against the Ratshosas which led to their
attempt, in April, to kill Tshekedi. For this they were imprisoned
by the Administration and banished from the Reserve, but they
have continued through their local partisans to be a disturbing
factor in tribal life. Meanwhile, Tshekedi had trouble with his
half-sisters, whom he also had to banish. This was followed by a
conspiracy to claim the chieftainship for an illegitimate son of
Sekgoma II, by a petition against Tshekedi’s rule organized by
several of the royal headmen, and finally by his open breach with
the Rraditladi family, who are alleged to have been implicated in
most of the preceding intrigues. All these disputes split the tribe
into factions whose continuous agitations against one another
obviously made the Chief’s position very difficult. It is evident
enough that the success of a chief’s reign is determined in no small
measure by his personal relations with his near kinsmen.

Formerly the chief’s power was to some extent limited also by
tribal law. If he committed an offence against one of his subjects
the victim could get some prominent man to intervene; the chief
was then expected to make amends for the wrong he had done
But so great was the reverence attached to him by virtue of his
birth and ntual position that the people would put up with much
from him that would never be tolerated in one of lesser rank • and
often enough, m practice, the victim had no real remedy except
to leave the tnbe and transfer his allegiance to some other chief

KeX prOVOcation ** drastic action would

taken. If the chief flagrantly misruled the tribe, or in other

woffid'Shd l h ° Stility ° f the pe °P Ie ’ ^ fading headmen
would withdraw their support and publicly attack him at tribal

fon GiVen 0 " T S ** * ^ t0 wholesale mi gra-

tion Given sufficient provocation, the people might even begin

to plot against him, in the hope that he would be overthrown md

one of his more popular relatives take his place; or as a last resort

f ^f of .revolt have occurred often enough in the past
history of the tnbe, «d they did not always meet with faflu^

The imposition of European rule deprived the people of‘the

Z:^r A f eS ^ •° mierly P ossessed against oppression md

^ more inlocal

THE NGWATO OF BECHUANALAND PROTECTORATE 81

policy was to rule as much as possible through the chief, the
Administration tended in most cases to uphold his authority,

without inquiring too closely into the merits of the trouble*
Freed, in consequence, from fear of the sanctions formerly
restraining him, he became more arbitrary in action and jealous
of any challenge to his authority. Kgama was a man of exceptional
ability and enterprise, who through sheer force of personality was
able to do much for the material advancement of his people; but
he was also firmly insistent on his rights as chief, and brooked no
opposition. The right of appeal from the chief’s verdicts estab¬
lished in 1919 did little to help, since, by an astonishing arrange¬
ment, the appeal court consisted of the District Commissioner and
the chief himself!

The Native Administration Proclamation has at last provided
more effective machinery for controlling the chief and protecting
the tribe against oppression or maladministration. While making
it a statutory offence for any tribesman to ‘conspire against or
subvert or attempt to subvert’ the authority of the chief, the
Proclamation specifically states that this provision does not apply
to bona fide criticism of his rule. On the other hand, should the
chief (or any other tribal authority) fail to carry out the duties
imposed upon him, he can be tried by the District Commissioner
and, if convicted, fined or imprisoned. Moreover, if he' at any
time ‘neglects or fails to discharge properly his duties as chief, or
becomes physically incapable of carrying them out properly, or
abuses his authority and oppresses his people, or otherwise proves
to be a bad chief’, he can, after having been given an opportunity
of defending himself, be suspended from the exercise of his duties
as chief; and until his suspension has been withdrawn, some one
else, appointed either by the tribe or by the Administration, will
act in his place. If the tribe so desires, but only then, the chief
may even be deposed permanently. If necessary, he may also be
ordered, after suspension or deposal, to leave the Reserve, and
not to enter it again until given permission to do so.

As an institution, the chieftainship is still greatly honoured and
respected, and the people still look primarily to the chief as their
ruler and guide. But his loss of many old ritual and economic
functions, the presence of a rival leader in the form of the mis¬
sionary, and, above all, his subjection to the Administration, have
inevitably deprived him of much of his authority. Moreover,

82

AFRICAN POLITICAL SYSTEMS

educational advancement, and the possibility of escape created by
labour migration, have made the people more openly critical of
his conduct, and they no longer respond so readily to the many
demands he makes upon their services. There is no tendency as
yet to advocate abolition of the chieftainship, but the stricter
measures introduced by the Administration to keep the chief
under control and to protect the tribe from abuse have been
generally welcomed, especially by his more literate subjects, who
apparently hanker .after a ‘constitutional monarchy 5 of the kind
they have learned to know from their school-books.

At the same time, the chief himself is in a by no means happy
position. His people look to him to protect their interests, and
often enough his actions are inspired by genuinely patriotic rather
than purely selfish motives. The Administration, again, on the
one hand holds him responsible for the maintenance of.peace,
order, and good government, and on the other expects him to
see that the laws it imposes and the instructions it issues are duly
carried out. It has already happened that a situation may arise
in which the chief must choose between his duty to the Adminis¬
tration and what he regards as his duty to the tribe. If he attempts
to enforce the wishes of the Administration, he is only increasing
his own difficulties by arousing the hostility of the people; if he
disobeys the Administration, he is liable to punishment and even
suspension. These and some of the other problems nowadays
confronting the chief were recently discussed in an article by
Tshekedi, 1 which shows how far removed the modern system of
administration is from the day, little over forty years ago, when
Kgama was told in London by the Secretary of State that he would
be allowed to continue ruling his people ‘much as at present 5 .

1 ‘Chieftainship under Indirect Rule’, J. R. Afr. Soc., vol. xxv (1936),
pp. 251-61.

THE POLITICAL SYSTEM OF THE BEMBA TRIBE-
NORTH-EASTERN RHODESIA

By Audrey I. Richards

I. Bantu Political Organization—Some General Features

T HE political systems of most of the Bantu peoples known to
us show certain striking similarities, particularly as far as
South and Central Africa are concerned. We are apparently
dealing in each case with a tribal organization that is an outgrowth
of a smaller lineage group, either split off from a parent stem in
search of independence and new territory, or scattered by the
onslaught of an enemy. In South, Central, and to a lesser extent
East Africa most of the ethnic groups now known as tribes have
a surprisingly short history of occupation of their present habitat
—rarely more than 200 years and sometimes as little as 50 to
100. For this reason the original kinship structure of the immi¬
grant people can still be recognized as the framework of their
political system. Authority is almost invariably based on descent,
whether within the family, the village, the district, or the nation,
and the chief of the tribe combines executive, ritual, and judicial
functions according to the pattern of leadership in each constituent
kinship unit. Like the family head, he is a priest of an ancestral
cult, believed in many cases to have a mystic power over the land,
and he invariably claims rights over his people’s labour and
produce. The hierarchy of Bantu society allows only one type of
authority, one basis of power, and one set of attributes in its
leaders in most of the tribes so far described.

Besides this personal relationship established by tradition
between the Bantu subject and his chief, yet another feature of the
political organization depends on facts of kinship, emotional,
legal, and ritual. Political power and prerogatives tend to become
concentrated in the hands of descendants of the original lineage
group, of which the chief is the living representative, and in many
areas tribal cohesion seems very largely to depend on the pre¬
dominance of this ruling line, whether the latter is to be reckoned
as the first Bantu people to occupy the particular territory, or
whether it conquered the earlier inhabitants and subsequently

g 4 AFRICAN POLITICAL SYSTEMS

built up a new state. We find tribes called after the name of the
original leader of the nuclear community 1 (e.g. the Zulu, or the
lineage groups of North Basutoland loosely described as Bamoheng,

or Bamoketela): and others’ called after his clan (e.g. Kwena,
Mangwato). The descendants of the first chief's clan may form
a ruling caste (e.g. Zulu, Swazi, Bemba), and the total number
of clans In the tribe may be arranged in order of precedence based
on the tradition of the original migration into the area, or else the
degree of relationship with the descent group of the chief . 2 The
next of kin of the chief may play a definite part in the political
organization, may claim rights to territorial chieftainships or
villages, membership of tribal councils or smaller advisory bodies
(e.g. the council formed by the chief’s brother, sister,- and near
relatives among the Venda), or they may act as a regency council
at the chief’s death (e.g. Venda, Tswana, Swazi).

These then seem to be common features of Bantu political
organization—the position of the chief as head of a community
held together by bonds, real or fictitious, of kinship and as priest
of an ancestral cult, and a political structure based on the domin¬
ance of a leading family line or clan. It is the differences in the
machinery of government and in the incidence of tribal authority
within this common pattern that make the interest of a comparative
work like the present. In examining a particular case, there are
a number of different conditioning factors which seem to account
for these variations in political structure. Of these the most
obvious appear to be the following: (a) the length of time the
tribe has inhabited its present territory; (b) the type of immigra¬
tion, whether by peaceful penetration, ejection of other units or
their amalgamation ; 3 (c) the emphasis placed on different prin-

1 Schapera uses this term to describe a ruling group which has conquered
and finally amalgamated other peoples often of foreign stock (cf. p. 57).

2 cf. the heirarchy of Ganda clans: the,precedence observed in tasting first-
fruits in order of clan seniority among the Sotho peoples, and the respect
still accorded to the Zulu and Swazi clans that have split off from the original royal
stock, when it became necessary to contract marriages between members of
one house.

3 Compare the differences in size and homogeneity between single tribes
largely^ of one stock with a single paramount chief (Swazi, Bemba); the
congeries of small autonomous tribelets with similar cultural features but no
supreme head (the swamp peoples of north-eastern Rhodesia or the low-veld
tribes of northern Transvaal); or the amalgamation of a number of different
e r“ c £f° u P s * nto one empire by conquest or absorption (the old Luba Empire
of the Congo, or that of the Zulu under Shaka or the Basotho under Moshesh).

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 85

ciples of social grouping such as descent, age, sex differences, or
local ties by which the tribe may be integrated and the incidence
of authority determined; (d) the economic bases of the people’s
activities which affect their degree of dispersal, the form of leader¬
ship required, and the economic values associated with political
prerogatives; (e) the type of foreign rule to which the tribe is
subject and the European elements that are effecting its political
development, i.e. variations in policy from the administrative
system known as Indirect Rule in Tanganyika, Northern Rhodesia,
Uganda, to the more direct government by the white man in South
Africa or the attempts made to create new political institutions
for Natives, such as the Bunga system of the Transkei.

All these factors will be found to account for differences in
political organization among a number of the kindred peoples
known as the Bantu, and I shall try to analyse the Bemba system
along these lines.

IL The Bemba Tribe

(a) Tribal Composition . The Bemba tribe at present inhabits
the Tanganyika plateau of north-eastern Rhodesia, between the
four great lakes—Tanganyika to the north-east, Nyassa to the
east, and Mweru and Bangweolu to the north-west and west
respectively. They number to-day about 140,000, very sparsely
scattered over the country at a density of an average 3*75 per
square mile.

The Bemba trace their origin to the area now known as the
Belgian Congo and declare that they were originally an offshoot
from the great Luba people which inhabits the Kasai district.
The fact that the first ancestor of the Bemba is known as Citi
Muluba (‘Citi the Luban’) substantiates this tradition, together
with the cultural similarities still noted between the two peoples
and the fact that Luban words, no longer understood by the
Bemba commoner, are still used as part of the religious ritual at
the paramount chiefs court. The legends of immigration are
numerous and circumstantial. The first arrivals apparently crossed
the Lualaba River, which forms the western boundary of their
present territory, about the middle of the eighteenth century, and
travelled north and east until they established their first head¬
quarters near Kasama, the present administrative centre of the
Bemba country. From the sociological point of view, their history

36 AFRICAN POLITICAL SYSTEMS

dates from this period, since the composition of the invading
group still determines the title to chieftainships, rank, succession
to various offices, and the order of precedence of a number of the
older clans.

The Bemba are to all intents and purposes a homogeneous
group. They form a quite distinct political unit from the Bisa,
Lala, Lunda, and other neighbouring tribes with similar traditions
of origin and marked cultural and linguistic affinities. The Bemba
declare that their forefathers found the country empty on their
arrival, and, whether this is true or not, there seems to have been
no strong opposition from whatever groups occupied the territory.
The war-like habits of this tribe seem to have developed later,
when they spread into the surrounding districts, pushing back
their neighbours, such as the Lungu to the north-west, the Bisa
and the Lala to the west and south respectively, and the Cewa to
the east. The dominance of the Bemba chiefs was still more effec¬
tively enforced by the import of Arab guns in the nineteenth
century. Where they did not dislodge the occupants of the sur¬
rounding country, their chiefs appointed members of the royal
family or specially faithful subjects to hold the district for him
(ukulashika) and to collect tribute of ivory tusks, grain, iron¬
work, salt, or other goods. The empire of the Bemba extended at
one time right up into the Congo and to the shores of Lake
Tanganyika, and they exerted influence over most of the present
Bisa and Lala country to the south. With the coming of the white
man at the end of the nineteenth century, their authority over the
surrounding tribes collapsed, and though Bemba chiefs still rule
over Bisa villages, e.g. in the Chinsali and Luwingu districts, it is
rather a case of tribal admixture on the borders than a large-scale
incorporation of foreign elements such as has occurred in the his¬
tory of some of the Southern Bantu states. For the purposes of the
present inquiry, we can reckon the Bemba as a homogeneous tribe
with a histoiy of settled occupation of their present territory
lasting about 200 years.

The distinctive marks of tribal membership are the following:
( a ) common name Babemba, still uttered with a good deal of
pride in such phrases as *Fwe Babemba' (‘We, the Bemba , ) J used
to preface bragging references to the exploits of the tribe as com¬
pared to those of surrounding peoples, who are still sometimes
referred to contemptuously as slaves (bashya), (b) The common

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 87

language ( Cibemba ), which forms a distinct dialect in native eyes,
although it does not differ very considerably from the neighbour¬
ing Cibisa or Cilala. (c) The tribal mark, a vertical cut on each
temple about 1 in. long behind the eyes, (d) The common
historical traditions of the people—even young men at the present
day speak with pride of the coming of their fathers from Luba-
land, and take delight in describing the military exploits of their
ancestors and the ferocity of the old chiefs, (e) Their allegiance to
a common paramount chief, the Citimukulu, whose overlordship of
the Bemba territory is unquestioned.

(b) Social Grouping. (1) Kinship. The Bemba are a matrilineal
tribe practising matrilocal marriage. Descent is reckoned through
the mother and a man is legally identified with a group of relatives
composed of his maternal grandmother and her brothers and
sisters, his mother and her brothers and sisters, and his own
brothers and sisters. His membership of this group determines
his succession to different offices and his status in the community,
although in a matrilocal society it only occasionally determines his
residence. He also belongs to a wider descent group, the clan
(umukoa, plur. imikoa) which is also traced in the woman’s line.
Each umukoa is distinguished by the name of an animal, plant, or
natural phenomenon, such as rain. It has a legend of origin
usually describing the split-off of the clan ancestors from the
original lineage group, and an honorific title or form of greeting.
Clans are in effect exogamous, since a man may not marry a
woman he calls ‘mother’, ‘sister’, or ‘daughter’, and these terms
are extended to the limits of clan membership on the maternal
side. Through his clan affiliation, a man traces his descent, rank
—if he belongs to the royal clan—rights to succeed to certain
offices, such as hereditary councillorship (cf. pp. 100, 108), and
claims to his relatives’ help and hospitality.

Some clans have a higher status than others, according to
whether their original ancestors arrived in the country as part of
the following of the first Citimukulu, or alternatively, split off as
a separate descent group later. Thus the crocodile clan (Bena
tjandu) is the umukoa of the first immigrant chief and stands
highest in status (cf. ‘Rank’), while various others, such as the fish
clan, millet clan, &c., are said to be of similar antiquity. The
hereditary councillors described later belong to these clans. AH the
imikoa are paired with opposite clans that perform reciprocal ritual

88

AFRICAN POLITICAL SYSTEMS

duties for each other, but this form of social grouping does not
seem to affect the political organization at all at the present day. 1

Within the clan, smaller lineage groups are recognized. These
have no distinct name; though the Bemba often refer to them as
‘houses* (i amaianda , sing, itjanda) of the same clan. Such a house
consists of the direct descendants of one particular ancestress
traced back to three or four generations—five at the most. Within
this smaller descent group, succession to office is usually limited,
and chieftainships tend to become hereditary within three or four
generations in such lines . 2 Social replacement of one man for
another, either as an heir, an officiant in a religious ceremony, in
fulfilment of a marriage contract (in the case of a woman), or in
compensation for blood guilt in the old days, tends and tended to
take place within the ‘house* and not the clan, though members of
the umukoa do replace each other if there is no one more nearly
related within the irjanda to do so.

It is the smaller descent-group which is important in consider¬
ing the influence of the ancestral spirits (imipashi, sing, umupashi)
over the living, either as affecting the welfare of their descendants
in general or as entering the wombs of pregnant women of that
descent-group to act as guardian spirits to the children as yet
unborn.

Apart from the descent-group that determines his status, there
is the body of kinsmen with whom a Bemba co-operates actively in
daily life. These are the people with whom he may choose to live,
and who gather together at any important event in his life, such as
marriage, the birth of a child, illness, or a death. This group is
known by a distinct term, the ulupwa* It has a bilateral basis, since
it is composed of the near relatives on both sides of the family and
also relatives in law. The balance between the powers of the
maternal and paternal relatives is a very even one in Bemba society,
m spite of the legal emphasis on the matrilineal side, and the ties
uniting the members of the ulupwa are very strong, 3 Though it is

n 5 eCip ^ Cal G 3 an ReIationshi P s among the Bemba of North-Eastern

.-Y? 6 if* ^. kuIa > Bw fy? Cangala, claimed, in the course of a succession
S n , ear y ektlVes ’ ** chadren of his grandmother, Nakasafye,
co ““ dere . d 88 “separate ‘house’, as distinct from the children and
grandchildren of her sister, Mukukamfumu II (cf. chart on p. 102).

(193$ ^ Mother ‘ nght m Central Africa’ in Essays presented to C. G. Seligman

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 89

more usual to live with kinsmen on the matrilineal side, the grand¬
father or the mother’s brother, yet a man may choose to live with
his father’s people by preference, and they play an important part
at all the great ceremonial occasions in his life. The strength of the
bilateral ulupwa is in fact one of the distinguishing features of the
Bemba kinship system as compared with the strongly patrilineal
societies of South Africa to my mind. It affects the political system
in two ways. First, it allows for a much greater variety in the
composition of the village, and more possibilities of change in its
membership; and, secondly, we find in the case of the chiefs
relatives that the ulupwa of a ruler is an important unit in the whole
political machine. A ruler’s sons receive positions and office as
well as his heirs, the maternal nephews; and his father’s relatives
and those related to him by marriage are also favoured, so that his
grip over the country is a strong one.

(2) Local Grouping. The local unit in Bemba society is the
village (umushiy plur. imishi). It contains on an average thirty to
fifty huts, and is a kinship unit first and foremost. A village comes
into being when a middle-aged or elderly man has acquired a big
enough following of relatives to justify his applying to the chief for
permission to set up a community on his own. He usually builds
near other relatives, but land is so plentiful that it is perfectly
possible for him to settle almost where he pleases within the
chief’s domain. The core of the village consists in the first place of
the headman’s own matrilocal family group, i.e. his married
daughters with their husbands and children, and probably
members of his matrilineal descent-group, i.e. his sisters and their
children. Polygamy is rare. A chief will have a number of wives,
say ten to fifteen, but commoners do not often have more than one.

A successful headman will be able to attract more distant rela¬
tives to him, both on the patrilineal and matrilineal side. On his
death he may be succeeded by his heir, and such a local com¬
munity may continue in existence with frequent changes in its
composition, for two, three, or even more generations. Indeed, the
village of the hereditary officials of the paramount chief (bakabilo,
cf. pp. 100, 108) remain permanently fixed in one village.
Thus in every district there are a number of new villages brought
into existence by the chief’s favour (ukupokelafye kuli mfuntu) and
therefore specially dependent on his support. These include com¬
munities newly gathered together by commoner headmen, as

9 o AFRICAN POLITICAL SYSTEMS

described, as well as existing villagers which have been given, with
or without the inhabitants’ goodwill, to a relative of the chief.
Besides these new headmanships, there are those founded in the
chief’s predecessors’ reigns and described as such, and on the
whole less dependent on the present ruler. The proportion of new
to old villages in Citimukulu’s district in 1933 was as follows:

On 160 villages:

per cent.

New villages .. .... .. .. 28

Villages with one previous holder of the headmanship 16

Villages with two previous holders of the headmanship 10
Villages with three or more holders of the headmanship 40
Villages constituted from remnants of two old villages 6

The skill with which he allots headmanships, and the positions in
which he places his own relatives, contribute greatly to a chiefs
power.

In spite of the provisions for inheritance of headmanships, the
Bemba village is an impermanent community from many points of
view. It moves every four or five years, in keeping with the
practice of shifting cultivation, and is liable to disruption at the
death of an important member or at any loss of popularity by the
headman. The plentiful supply of land and the many alternative
possibilities of kinship grouping provide ample opportunities
for a man to change from one village to another if he pleases, and
in any case he is almost bound to live in a series of communities
during his lifetime, e.g. the village of his birth, that to which he
moves when he marries, any other village he may go to when he
acquires the right to move his wife and family from her people’s
care, and lastly, in some cases, a community of which he may
acquire the headmanship through succession to his maternal uncle.
Hence, although a man’s companions and fellow workers are those
of his umuski and he speaks with some affection of the village of his
birth or of his mother’s people (icifulo), yet the bonds of kinship are
much stronger than those of the impermanent local group. A
Bemba is a member of a ulupwa and may move as he pleases to live
with any of the relatives composing it, and he is the subject of a
chief and may obtain permission to live in any part of the latter’s
territory ,but his ties to a given locality are not necessarily strong. 1

1 cf. my Land, Labour and Diet in Northern Rhodesia (1939), chap. vii.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 91

A chiefs village (■ umusumba ) is very much larger than that of
a commoner. Inhabitants of the capital are composed of relatives
of the chief, his followers, and also a number of families which
moved there originally to win royal favour and have become
accustomed to court life. 1 Since a chief’s reputation depends
largely on the size of his capital, and his councillors, courtiers, and
administrative officers were drawn largely from his villagers, the
umusumba is an important unit in the political machine. The late
Nkula’s village had about 400 huts when I visited it in 1931, that
of the Citimukulu 150 in 1938. The capitals of pre-European days
were evidently very much larger. These communities were divided
into sections (ifitente,sing.icitente) and though nowadays there are
nine ifitente at the paramount’s village, there were formerly thirty
to forty, according to native accounts.

The whole Bemba territory is divided into districts {ifyalo,
sing, icalo). The icalo is a geographical unit with a fixed boundary
and a name dating from historical times, e.g. the district of the
Citimukulu is known as Lubemba, the country of the Bemba, and
that of Mwamba, Ituna. These districts are territories originally
allotted to members of the royal family, but once so divided they
have never been sub-divided to provide smaller chieftainships
for a new generation of princes as has happened in some parts of
South Africa.

But the icalo is also a political unit. It is the district ruled over
by a chief with a fixed title—the' name of the first ruler to be
appointed over each particular strip of land, always a close relative
of one of the earlier Citimukulus. There are several types of chief,
the paramount, who has his own icalo, as well as being overlord of
the whole Bemba territory; the territorial chiefs, five or more in
number, who have under them sub-chiefs who may rule over very
small tracts of country or, rather, over a few villages. 2

Each of these chiefs is known by the same title mfumu and each
icalo is a more or less self-contained unit, a replica ofthe social struc¬
ture of the other. Each capital has its own court, however small.
Each chief has rights over the labour of his own villages. They
work for him only and not for the paramount as happens among

1 The phrase ‘umzvino musumba* (‘inhabitant of the capital*) is used to indicate
a ‘chief’s man* or a person of specially polished manners and knowledge of
affairs.

2 Mwamba has as a sub-chief, the Munkonge, and the Nkula has Shimwalule,
Mwaba, Mukuikile, Nkweto, &c.

92

AFRICAN POLITICAL SYSTEMS

the Zulu, Swazi and other tribes with the regiment system. The
icalo is also a ritual unit. At each capital are the sacred relics
(babenye) of the first holders of the chiefly title and their ancestral
spirits are thought to act as tutelary deities of the district, and are
worshipped at the umusumba , at village shrines, and old hut sites
throughout the country, and are also commonly supposed to act
as guardian spirits to children born within the icalo A Naturally the
ritual and political organization of the paramount J s capital is more
elaborate than that of his inferiors, but even the smallest sub-chief
maintains his miniature court and tries to ape the state of those
above him, while the bigger territorial chiefs sometimes rivalled the
power of the Citimukulu in the old days.

The territorial chieftainships are arranged in order of prece¬
dence, according to their nearness to the centre of the country—
Lubemba—and the antiquity of their office. To the most im¬
portant of these chiefdoms—the Mwambaship, the Nkulaship, the
Nkolemfumuship, and the Mpepoship, for instances—the Citimu¬
kulu appoints his nearest relatives, the one succeeding the other in
order of seniority. Thus the present Citimukulu, Kanyanta, has
acted in turn as the Nkolemfumu, and the Mwamba before suc¬
ceeding to the paramountcy (cf. chart on p. 102). On the other
hand, the sub-chieftainships have tended to become concentrated
in local branches of the royal family, and the paramount's strong
grip over the country and his intimate knowledge of affairs at the
courts of his fellow chiefs is certainly weaker in these outlying
districts than in the case of chiefdoms ruled by his close relatives.

To the commoner, membership of an icalo means his allegi¬
ance to the chief of that territory. He will describe himself as an
inhabitant of a district, such as Icinga, i.e. mwine Icinga or,
alternatively, as the subject of its chief, Nkula, i.e. mwine Nkula ,
and both terms are synonymous. He may move from village to
village within the icalo, but he remains his chief's man. The latter,
in his turn,, reckons his assets, not in terms of the size of his
territory or its natural resources, but rather by the number of his
people and in particular the villages he has under his rule.

(3) Rank. Rank is a marked feature of Bemba society. It is
based on kinship, real or fictitious, with the chief. All members of
the royal crocodile clan (Bena rjandu) are entitled to special
respect, precedence on ritual and social occasions, and sometimes

1 Hence the great preponderance of one or two birth-names in.each district.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 93

to claims on the people’s services. The potential heirs of a chief
within his own branch of the family—that is to say, his brothers,
maternal nephews, or maternal grandsons—are treated with
particular deference. The former two categories are described as
chiefs and addressed by the title mfumu , while the latter, only
slightly lower, in status, are referred to by a special name
beshikulu ha mfumu ('grandchildren of the chiefs’) and have their
own ritual and social prerogatives.

Women of the royal line, the mothers, sisters, maternal nieces,
and granddaughters of the chiefs are called banamfumu and are
treated with much the same deference as are the men of the family.
The mother of the paramount is highly honoured, succeeds to a
fixed title—the Candamukulu —takes part in tribal councils, and
has several villages of her own. The sisters of chiefs are privileged
persons, protected and supported by their royal brothers, and
usually granted one or more villages to rule. They are above the
law in matters of sex morality, and a princess is allowed to have as
many lovers as she pleases, provided she produces many children
as potential heirs to the throne.

Not only members of the royal clan, but also persons who
merely belong to the ulupwa of the chief, can claim high rank, i.e.
his relatives on his paternal side, and his own sons. Some fathers
of chiefs were nobodies and were quickly forgotten, but some have
been famous men, honoured by their sons when the latter suc¬
ceeded to the throne. The children of chiefs, though not members
of his clan, and therefore not heirs, are also entitled to special
privileges, and the bana bamfumu ('children of the chief’) form a
class of their own. They are brought up at the court, where they
are treated in many ways more favourably than the heirs them¬
selves and are able to claim headmanships and even chieftainships . 1
Even the half-brothers of chiefs, through other fathers (>bakaulu ),
have rights to special treatment at court.

Added to this, already numerous class of royal personages are
the descendants of close relatives of dead chiefs. Roughly speak¬
ing, any person who can claim to be maternal nephew, grandson, or
son of a chief is succeeded by a man who continues to hold the
same rank by the ukupyanika system described on p. 9^*
then addressed as 'chief’ or ‘son of chief’. The descendants of

1 A few chieftainships are definitely handed on to ‘sons of chiefs’ instead of
to ‘chiefs’, e.g. the Makassaship, the Lucembeship, or the Muhkongeship.

94 AFRICAN POLITICAL SYSTEMS

royal princesses are also entitled to honour, as well as those of
wives of chiefs and even consorts of princesses. It will be seen,
therefore, that the royal rank is a very large one. Any one who can

possibly claim connexion of any sort with any chief, dead or
living, does so, although the perquisites of rank are in most cases
honour only and the possible favours of the chief, rather than any
material assets. Every one outside the royal clan, or ulupwa , is an
umupabij or 4 ordinary person’, and in old days there was a slave
class below—men and women captured in battle or enslaved to
their own people for some crime. These individuals were known
as bashya. The term is now used as an opprobrious epithet
especially for foreigners—often assumed to have been enslaved by
the Bemba formerly. Slavery itself no longer exists.

(4) Other Principles of Social Grouping. Age is not a principle
of social grouping among the Bemba. Precedence is reckoned on
the basis of seniority, as in most Bantu societies, and there are
special terms used to describe the different stages of life, suckling,
infant, child, adolescent, unmarried, married, old, &c. But there
are no regiments based on age, as in South and parts of East
Africa, and the boys initiation ceremonies so often found associated
with such institutions do not exist among this group of the
Central Bantu.

There are no occupational groups, with the exception of certain
specialist fishing communities on the banks of the big rivers, and
in the old days there were specialist hunters of big game. Secret
societies, such as the ubutwa y which is common among neighbour¬
ing tribes over the Congo border, and has been adopted by the
Bisa of the swamps, do not seem to have been introduced among
the Bemba.

To conclude, Bemba society is as yet undifferentiated to any
large extent. The tribe is an outgrowth of a lineage-group which
has occupied its present territory for 200 to 300 years, and has
remained more or less homogeneous. The original kinship struc-
ture is still apparent. All the social groups to which a man
belongs are ultimately based on kinship—whether it is his house-
hold village or descent group, and there are no other forms of
association such as age-sets to cut across this original grouping by
descent. Rank consists of membership of the clan of the first
immigrants to enter the land.

(c) Economic Background. The Bemba are an agricultural people

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 95

like most of the Central Bantu group to which they belong. They
keep no cattle. Tsetse-fly at present prevents their keeping stock
over most of the country, but in any case they seem to have no
pastoral traditions, whatever they may have had formerly. Thus
they have no means of storing wealth as have the Southern Bantu.
Their marriage contracts are fulfilled by service and not by the
passage of cattle. In the old days military glory and the extraction
of tribute from conquered peoples seems to have been the dom¬
inant ambition of the Bemba chiefs, and their wealth consisted
in the size of their following and the amount of service they were
able to command. This fact profoundly influences their position
at the present day (cf. p. 116).

The soil of most of this district is poor and it ha$ not attracted
white settlement. The staple crop is finger millet (eleusine core-
cana), while some kafir com, a little maize, legumes, and pumpkins
are also grown. The people practise shifting cultivation of a
primitive type, and the plentiful supply of land and the lack of
any localized natural resources which might attract the inhabitants
to settle in one area rather than another all affect the political
system. As has been shown, they decrease the strength of local
ties as against political or kinship affiliations, and they account for
the fact that the power to distribute land is not an important
prerogative of leadership in distinction to conditions in most
Southern Bantu tribes.

Hunting and fishing contribute a small share of the fooa-supply
only. Organized marketing does not exist, and under modem
conditions no cash crop has been found for this area. This fact,
together with the absence of opportunities for local employment,
forces the adult male population to look for work outside the tribal
area, with resultant effects, as will be seen, on the political system
of the tribe.

(d) White Administration . The type of white administration
introduced in this area is described on pp. 112-20.

Ill . Bases of Authority

The positions of leadership in Bemba society consist of the
following offices: (a) territorial rulers (chiefs and headmen),
(b) administrative officers and councillors ; (c) priests, guardians of
sacred shrines, and magic specialists with economic functions, {d)
army leaders in the old days. Succession to all these offices is

9 6 AFRICAN POLITICAL SYSTEMS

based on descent in nearly every case. Chieftainships were limited
to one clan, as we have seen; some of the councillorships (i.e. the
bakabilo) are confined to a few of the older clans;.and head-
manships, though they may be won through the chiefs favour,
tend to become hereditary in their turn. All priestly offices are
hereditary without exception, as is natural where an ancestral cult
of this type is practised. Magico-economic specialists, particularly
those in charge of fishing villages, usually acquire their powers by
descent also, as do some of the doctors and diviners ( tjanga ). In
each case the supernatural powers almost invariably correlated
with political authority in this area are conferred by a rite,
of great complexity, in the case of the succession of a chief,
known as ukypyanika. For these reasons it is essential to study
the dogma of descent by which these powers are believed to
be transferred from one generation to another, and the legal
rules of succession by which status and office are passed from
one man to another.

(a) The Dogma of Descent. By dogma of descent I mean, first,
those theories of procreation 1 which express a people’s beliefs as
to the physical contribution of the father and mother to the
formation of the child, and hence the traditional conception of
the physical continuity between one generation and the next; and
next their beliefs as to the influence of the dead members of each
social group over the living, and hence the social identification 2
of a man with the line of his dead ancestors.

Among the Bemba it is believed that a child is made from the
blood of a woman which she is able to transmit to her male and
female children. A man can possess this blood in his veins, but
cannot pass it on to his children, who belong to a different clan.
Physiological paternity is recognized. Children are often described
as being like their fathers, and are expected to give the latter
affection and respect although they have no legal obligations to
them under the matrilineai system. ‘We take our fathers’ presents
because they begot us,’ they say. But it is nevertheless the physical

1 This term was first introduced by Malinowski, who showed how the rules
of matrilineai descent among the Trobriand islanders are buttressed by beliefs
that the father makes no physical contribution to the birth of his child. Similar
material published by Rattray from the Ashanti area shows a belief in a double
contribution of blood from the mother and spirit from the father correlated
with a bilateral emphasis on descent.

2 To use a term employed in a very stimulating manner by Radcliffe Browne.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 97

continuity of the mother's line of ancestors which is the basis of
legal identification with her descent group. 1 A royal princess
might even produce an heir by a slave father in the old days
without lowering her child's prestige. The relationship between
brother and sister, which is a very close one, legally and ritually,
is based on the fact that the two were bom from one womb, and
in the case of the royal family it appears to be equally strong when
the two are children of different fathers. These theories of pro¬
creation account, not only for the matrilineal descent of the
Bemba, on which succession to chieftainship is based, but also
for the rank accorded to the royal princesses as mothers of
chiefs, and the headmanships and other positions of authority
given them.

The Bemba dogma as to the influence of the dead over the living
is also of the utmost importance as a basis for political authority.
The spirit of a dead man (umupashi, plur. imipashi) is thought to
survive as a guardian presence associated with the land or village
site formerly inhabited, and as a spiritual protector of different
individuals bom in the same lineage group and called by the same
name. The imipashi of dead chiefs become tutelary deities of the
land they ruled over, and responsible for its fertility and the
welfare of its inhabitants. They can be approached by the successor
to the chieftainship at various sacred spots in the territory and at
the sacred relic shrines (bahenye) in his own village. A chief is
said to be powerful because he ‘has great imipashiV It is for this
reason he is described as the umwine calo , ‘owner of the land',
and it is important to note that in every case the most important
imipashi and the most sacred relics are those of the first chiefs to
enter the land, or the first occupants of a chieftainship.

This dogma as to the influence of the dead over the living
inhabitants of a district, or the members of a descent group, is
very similar to the general Bantu pattern. But the Bemba belief
in the social identification between the dead man and his appointed
successor seems to me to be particularly complete. It is the basis
of the belief as to the supernatural influence exerted by the chief
in his own person as distinct from his direct approach to the spirits

1 The patrilineal tribes -on the Nyassaland border consider the Bemba theory
of procreation as entirely ridiculous. One Ngoni expressed his contempt, thus:
4 If I have a bag and put money in it, the money belongs to me and not to the
bag. But the Bemba say a man puts semen into a woman and yet the child
belongs to her and not to him!*

98 AFRICAN POLITICAL SYSTEMS

in prayer. When a man or woman dies, his or her social personality
must be immediately perpetuated by a successor who passes
through a special ritual ( ukupyanika) and thus acquires the name
the symbols of succession (a bow for a man and a girdle for the 5
woman), and the umupashi of the dead man. By this social
identification, a man assumes the latter’s position in the kinship-
group, uses the same kinship terms and, in the case of a chief, it
is almost impossible to tell when a man is describing incidents
which took place in his own life or those of an ancestor two or
three generations dead. So important is this social perpetuation
of the dead considered that immediately after a death, before the
successor has finally been appointed, a small boy or girl, usually
a maternal grandchild, is chosen to inherit the name of the deceased
temporarily (ukunzoa menshi, ‘to drink the water’). He or she is
given some small piece of the latter’s property and thereafter
addressed as grandfather or grandmother, or whatever the right
kinship term may be.

In the same way, a chief, once he has succeeded to the name
the spirit, and the sacred relics of his predecessor, has mag ic
influence over the productive capacity of his whole territory. His
ill health or death, his pleasure or displeasure, his blessings or
curses, can affect the prosperity of the people, and even his sex
life reacts on the state of the community. 1 For a chief to break
a sex taboo is an act which may cause calamity to the whole people
and the rites by which he is purified after sexual contacts form
one of the most important elements in the politico-religious
ceremonial requiring the participation of thirty or forty hereditary
officials (bakabilo) m the case of the paramount. Conversely
legitimate sex intercourse, especially as prescribed on certain
ritual occasions, may actually be a health-giving influence. Any

Jf ^ 5?® a certam de S ree of supernatural influence in his own
g as the successor to his predecessor’s umupashi, but a chief
has considerably more. For all these reasons, ritual precauttns
guard the sacred person of a chief. Special 4 oos mustTkeS
to preserve the ritual purity of the ruler’s sacred fire, and his sacred

hereditary councillors

grave with them Thi<? infx +•; ’ ^° r ^ ear the land’ into the

k - 1 ’-*«

It is probable that in the old d*v7th*nTu u-f' Godfrey Wilson. In fact
definition of a ‘divine ldng*. * em ^ a ciuefs would fall under Frazer’s

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 99

food, and to protect his person and that of the sacred relics from
the contagion of illness, death, or sex defilement. 1

The ritual by which a successor to the chieftainship is converted
from an ordinary individual to a ruler with almost divine powers,
has a good deal of political importance. It confers authority on
the priests—in this case hereditary officials (bakabilo) who cany
it out—and gives them, as we shall see, considerable power to
check the chief himself. The complete ritual by which the
umupashi of a dead ruler is liberated to guard the land he governed,
and the new heir is installed, is too complex to describe here and
now. Briefly speaking, it consists of the desiccation of the body
during a period of a year, from one kafir-corn harvest to the
next; its burial in a special grove (with human sacrifices in the
old days); and the building of a shrine on the site of the deserted
capital. To make the new chief, bakabilo must preside at the
installation of a new great wife, arrange for the sexual purification
of the royal pair, and the lighting of their new sacred fire. 2 They
must hand over to the heir the heirlooms (babenye) of which they
have been in charge during the interregnum, and must finally
found a new village and build again the sacred huts in which the
relics are to be kept. Such a ceremonial may take eighteen months
to two years and the participation of all the bakabilo and hereditary
buriers ( bhfmgo ) in the case of the paramount; a lesser time and
very many fewer priestly dignitaries in the case of the territorial
chiefs. The secrecy and awe surrounding these ceremonies is,
I believe, one of the ways by which the people’s reverence for
their chiefs is maintained.

(b) Legal Rules of Descent and Succession. Against this back¬
ground of beliefs as to the continuity between one generation and
another, the nature of descent and succession is defined exactly
by legal rule. Descent in the royal family is reckoned to the time
of first occupation of the country, and twenty-five to thirty
Citimukulus are remembered. In the case of a territorial chief,
the line of ancestors is not so long, and most are described as
having been ‘born in the country’. Most of the names honoured
are those of men but some are those of women, and it seems that

1 cf. my Land, Labour and Diet in Northern Rhodesia (i 939 )» chap. xvii.

2 Hence the importance of the great wife of the chief (umukolo ua cold) in
the political life of the tribe and the belief that her behaviour also influences
the welfare of the land.

100 AFRICAN POLITICAL SYSTEMS

the first ancestress to inhabit a new chiefdom, or one who was a
mother of numerous powerful sons and was thus able to found a
new branch, could claim to be so respected . 1 But it is to the men
holders of titles that most shrines are built.

The hereditary officials (bakabilo) also trace their descent
to the first arrival for the most part, and tell stories which account
for their right to the ritual offices they hold to-day, e.g. the
bafinga who now bury the chief, claim to be the descendants of
those who buried the first Citimukulus when on the march. This
reckoning of descent to a definite epoch in history very clearly
remembered is of service in maintaining the myth of absolute con¬
tinuity of the chiefly lines. In actual fact, the present Citimukulu is
a descendant of one Cileshye, who seized the throne from the
occupier, Cincinta, only four generations back. This branch of
usurpers is able to claim descent from the first Citimukulu all the
same. The first ancestors are remembered very accurately and
their sacred relics kept. The ensuing vagueness in the chain seems
to be of no account.

In most types of succession whether to the name and spirit of a
dead man or to his office, there are usually two or three potential
heirs, and although there are certain rules of priority, it is practi¬
cally never the case that there is one child known as heir to the
chieftainship from birth and brought up as such, as occurs in those
South African tribes in which the eldest son of the great wife must
always succeed. A Bemba chief, or commoner, is succeeded by his
brothers in order of age, next by his sister's children, and, failing
them, by his maternal grandsons. Difficulties arise when there
is a choice between an older classificatory 'brother', not a sibling,
but possibly a mother's sister's son, or an even more distant
'brother' still, and a young man, a maternal nephew who is
the child of the deceased's own sister, with whom, as we have seen,
his ties are very close. Here the principles of primogeniture con¬
flict with that of propinquity of kinship, in the case of a branch of
a family that has been in existence for three or four generations,
and it is probable that in these cases the nearest heir is appointed
unless he is manifestly unsuitable, when the more distant 'brother'

1 e.g. Bwalya Cabala, the first ancestress said to have been fetched from
Lubaland by her brothers when the latter had occupied what is now Bemba-
iand; or the Nakasafye, grandmother of the present Nkuia, who is described
as having started a new line, and was evidently a woman of great character as
well as the mother of many sons.

■ THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA ioi

or 'maternal nephew' is selected. I never heard of a regent being
appointed for a young man as is commonly done in those Bantu
tribes where the heir to the throne is known from his time of
birth.

The situation is more complicated in the case of succession to
chieftainships, since through the custom of inheriting one big
territorial chieftainship after another within the paramount’s
immediate family, a tradition has grownup that, e.g., the holder of
the Mwambaship should always succeed to the Citimukuluship,
whatever the priority of kinship. This claim w T as put forward
in the last succession dispute (1925) and is commonly supported
by Government officials who naturally prefer a fixed system
of succession to the discussion of rival candidates’ rights that
seems to have been the older procedure.- There is also a ten¬
dency becoming more and more evident for certain of these bigger
chieftainships to be confined to sub-branches of the main royal
line, as distinct from sub-chieftainships which are nearly always
given to descendants of local branches of the crocodile clan (e.g. the
Mwabaship). This constant growth and separation of different
sub-lines or houses of the royal clan seems to have been continuous
in the past. The chart of the present central branch of the Bena
rjandu should make the situation clear. It will be seen that the
first and second Citimukulus in this line were siblings, and were
succeeded by another pair of own brothers— Citimukulu III and
IV, the sons of the first ruler’s eldest sister, Candamukulu. The
paramountcy then passed to the line of a younger sister, Bwalya
Cabala, tradition stating that the eldest maternal nephew of Citi¬
mukulu III and IV, then holding office as Mwamba, refused to
succeed to the office for various reasons. The title then passed to
another pair of brothers in succession, Citimukulu VI and VII,
the sons of a younger daughter of Candamukulu—Nakasafye.
Hence the famous dispute of 1925 just referred to, between
Kanyanta, now Citimukulu, and his mother’s mother's .sister’s
grandchild, Bwalya Cangala, then holding the Nkulasbip, and
reckoned as Kanyanta’s classificatory brother. Bwalya claimed
that he was own maternal nephew of the dead chief, Ponde,
and Kanyanta that he came of an older line and that it had now
become established that the Mwambas always succeeded the
Citimukulus. The Government supported the latter claimant,
but there seems to have been very little to choose between the

(Probably some generations missed here)

Chart showing succession in central branch of the royal faMy (taken from a table made by E. B. H. Goodall and kept at the Government Office, Kasama)
Men shown thus: CITIMUKULU Women shown thus: Candamukulu

Note—( i) Two or more titles means the successive holding of two or more chieftaincies. (2) Citumukulu I designates the first Paramount

in this particular lineage group.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 103

legal rights of the two rivals, and the rather complex machinery
of tribal deliberation on such matters (cf. p. 109) was not called
into motion.

The chart also shows clearly the way in which certain chieftain¬
ships have also tended to become fixed in different family lines of
this main branch of the Bena tjandu even during the last four
generations, i.e. the Mwambaship, the Nkolemfumuship, the
Mpepoship in the line of the chieftainess Mukukamfumu, and the
Nkulaship, the Cikwandaship, and the Ceweship in the line of
Nakasafye. The separation between these two branches will
probably become wider and wider with time. In the case of the
outlying sub-chieftainships, this separation off of local family lines
of the royal clan has proceeded even further. To conclude, there
are definite rules of succession according to Bemba kinship but the
type of matrilineal succession usually provides two or three pos¬
sible heirs, and in the case of the bigger chieftainships there are a
number of different factors, such as the paramount’s control over
the more important ifyalo, the traditional order of succession to
different offices, local feeling in the case of sub-chieftainships, and
last, but not by any means a negligible point, the personal qualities
of the candidates themselves.

IV. Functions and Prerogatives of Leadership

The functions of the territorial heads, i.e. chiefs and headmen,
seem to be derived from two sources—the position of the leader
as head of a kinship group and his role as the representative of a
line of dead ancestors in a particular district. In the case of a
headman, these two aspects are indistinguishable, while the latter
predominates where a chief is concerned.

(a) The Headman. Bemba headmen are described as looking
after, keeping, or actually ‘herding the people’ (ukuteka bantu). As
senior kinsman of most of the villagers, a headman is responsible
for the discipline of the children and young people; he hears
cases informally and directs some economic activities. There
are few activities carried out by the whole community in common
except fishing and hunting, but besides organizing these latter
pursuits a good headman initiates each new agricultural process and
. encourages and criticizes the younger men and women. Land is not
often a matter of dispute in this area. The headman does not allot
individual plots, but listens to cases should any arise. He is said to

io 4 AFRICAN POLITICAL SYSTEMS

‘feed his people 5 and actually does so if they are in need, besides
dispensing hospitality to strangers.

The head of the village acts as its ritual head. In the old days, he
put up one village shrine to his own ancestors and one or more
others to the dead chiefs of the land. This is still done in out-of-
the-way parts of the country and in most places, I think, prayers
are offered to these tutelary deities, whether shrines are built to
them or no. The headman, like the chief, also influences the life of
the community through his own person. He must ‘warm the
bush 5 (ukukafye mpanga) by an act of ritual intercourse with his
wife before the huts of a new village are occupied. He blesses
seeds for sowing, axes for tree-cutting, and first-fruits. His fire
stands for the life of the community as a whole and must be
ritually lighted when occasion demands. He presides over the
special divination rites connected with village activities, such as
the founding of a new community or the death of a member, and
blesses new babies or individuals who are sick.

In the political hierarchy, the headman has his definite place.
No Bemba may cultivate land except as a member of a village
group, and the headman is responsible for organizing the supply of
tribute and labour which must be paid to a chief by the community
as a whole. He accompanies his villagers to court when, they have
cases to present and often speaks for them. He transmits the orders
of a chief to his people and nowadays those of the Government.
His prerogatives are few in number. As head of a kinship-group,
he can command personal service from his younger relatives and
should be able to exact one day 5 s work from his people on the first
day of tree-cutting and sowing. He is always given tribute of
beer or meat. But probably, apart from these few economic pri¬
vileges, the Bemba headman values most his position of authority,
his small following, and the favour of his chief.

The sanctions for his authority nowadays are mainly his popu-
brity, together with the strength of kinship feeling, and the belief
of the Bemba that it is dangerous to allow an older relative to die
injured. His supernatural powers were a source of strength in
the old days, but to a very small extent now, and it must be
admitted that the forces which keep a village together are riot
very strong. It is a constant fear to a headman that his people
will melt away.

(b) The Chief . The functions of the different types of chief

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 105

differ only in degree. All are said to look after their people, to
‘work the land 5 , and, with reference to their supernatural powers,
to ‘spit blessings over the land 5 ( ukufunga mate). Their political
duties consist in the administration of their capitals and also of
their territories as a whole. A large umusumba means plenty of
coming and going, enough workers for joint enterprises, a large
panel of advisers for court cases, many messengers to keep in
touch with the surrounding villages—-in short, the possibility of
keeping the tribal machine running. To maintain and even
augment such a community by his popularity and his reputation
for generosity is one of the chief's important political tasks. He
has also to keep contact with the people widely dispersed over his
icalo and to appoint new headmen, amalgamate old villages, and
decide as to the selection of heirs to old titles. On his success in
these last duties the integration of his people as a political unit
largely depends.

As a judicial authority, the chief presides over his court with
advisers selected from his village, and in the old days he alone
could hear charges of witchcraft and, in the case of the greater
territorial chiefs, put the accused to the poison ordeal (; mwafi ). In

the economic sphere, he initiates agricultural activities by per¬
forming the customary ceremofiy before each begins; he makes
big gardens with the aid of tribute labour from which he is able
to fill large granaries and thus find the wherewithal to feed his
following; he controls directly certain fishing and hunting enter¬
prises : and he criticizes and directs the gardening work of his own
villagers. 1

The ritual duties of a chief consist in the observation of the
taboos for the protection of his own person and the safety of the
sacred relics at his disposal, and the carrying out of a number of
rites for the sake of his whole icalo —in the case of the paramount,
for the whole tribe. These last consist of economic rites, tree¬
cutting, sowing, and first-fruit ceremonies, those performed in
case of national calamity, and for success in war in the old days.
He was formerly bound to protect the people from witches and
used to employ a special doctor at his court to destroy, by burning,
the bodies of those found guilty of this offence.

In the old days the chief organized military expeditions,

1 cf. my Land . Labour, and Diet in Northern Rhodesia (1939% chap, xiii, for a
full account of the chief’s economic powers.

io6

AFRICAN POLITICAL SYSTEMS

although he did not necessarily take part in the fighting. As
one chief put it, £ If we were killed, the whole icalo would fall to
pieces 5 . The ruler had certain military captains in his following,
could call up men to fight, direct their operations from afar, and
arrange for the performance of war magic for success before battle
and for purification from the stain of blood after it.

The prerogatives of a chief consist in rights over the labour of
his people, who are required to do a few days 5 tribute labour each
year and to answer sudden calls for help if made; and also claims
to tribute in kind, usually paid in the form of an annual present
of beer and/or grain, and portions of animals killed in the hunt. 1
It is through this tribute that he is able to pay his advisers,
servants, labourers—and soldiers in the old days. Formerly, he
maintained rights to certain monopolies, such as ivory tusks, salt
from the big inland deposits at Mpika, and guns and cloth traded
from Arabs. Slaves or booty captured by the army were brought
to him, and he had a number of his own people enslaved for various
offences. Besides these economic prerogatives, he commanded
great, one might almost say abject deference, and had the satis¬
faction of seeing his following grow, his authority increased, and
his power over life and death over his subjects recognized.

The sanctions for a chief’s authority are numerous, and they
were still greater in the old days. The most important of these
has already been described as the people’s belief in their rulers
descent from a long line of ancestors and the supernatural powers
thought to be so conferred. Besides this, a reputation for generosity
and a system by which advancement could only be attained through
royal favour naturally bound people to him. Much of his power
also rested m the old days on force. A chief practised savage
mutilations on those who offended him, injured his interests,
laughed at him or members of his family, or stole his wives. A
number of these mutilated men and women still survive in Bemba
country to-day. Command over the army and over the supply
of guns also lay m the chief’s hands and there is no doubt that
the greatness of the Bena rjandu rested to a large extent on fear,
The people explain that the royal family were named after the
crocodile because ‘they are like crocodiles that seize hold of the
common people and tear them to bits with their teeth’.

1 AH these dues

are very much harder to exact nowadays (cf. p. 116).

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 107

V. The Machinery of Government

Within each district there are a series of officials, messengers,
&c., who carry out the activities of government and the different
forms of ritual on which the chief's power depends. Some of
these are personal followers of the chief promoted by him for
their special loyalty (e.g. the bafilolo , basano ), while others are
hereditary officials who are more independent of their ruler's favour
(e.g. bafilolo and bafingo). All these different dignitaries can be
classed under various functional heads, i.e.:

(a) Administrative . These include the executive officials in
charge of business in the umusumba and those responsible for
carrying out the chiefs orders in the icalo at large. Within the
capital the most important are the heads of divisions (bafilolo),
who are appointed from among the chiefs personal friends. These
are charged with keeping the peace of the village, organizing the
tribute labour from the capital, allotting land for cultivation, which
is often necessary in the bigger settlements, arranging hospitality
for visitors—an important task at the capital—and acting as a panel
of advisers on all occasions (cf. ‘Judicial’, below). Besides these
elder men, there are at the umusumba a number of courtiers and
in the old days young men (bakalume ba mfumu). Young boys,
often members of the royal clan, were, and still occasionally are,
sent to court to be educated there, and some families remain
as courtiers for several generations apparently. All these act as
messengers, attendants, and in the old days took duty as
executioners.

As regards the country at large, the main difficulty was keeping
in touch with the scattered villages. The Bemba have no general
meeting like the pitso of the Sotho peoples or the libandla
of the Nguni. For the chiefs orders to be conveyed to his
villages, messengers have to go to and fro. Other officers are
required to recruit the tribute labourers and to demand beer
or produce for the chief, and to apprehend criminals. Since
some villages are sixty miles or so from the capital, an enor¬
mous amount of time is spent in coming and going in this way
and even with the introduction of the bicycle a great many
messengers of one sort or another are still required. In the
old days courtiers and younger relatives of the chief acted in
this capacity. Nowadays they have anything from four to

AFRICAN POLITICAL SYSTEMS

xoS

twelve uniformed messengers, kapasus, and for the rest they
go short of service.

(b) Military. There was no general military organization in
this tribe, but attached to each big court were one or two captains
(baskika). Some of these were hereditary, with ritual functions
connected with war magic, and others appointed at the chief’s
will. They now act as specially trusted messengers.

(c) Judicial. There is no fixed composition to a Bemba court,
although its procedure is laid down by custom. At a small chief’s
court, the elderly men of the village attend, while the bafilolo act
as advisers at the big imisumba. Cases go on appeal from sub-chief
to chief, chief to paramount, and in the event of a case of extreme
difficulty presenting itself, the Citimukulu can summon from their
villages some of his hereditary priests or councillors, the bakabilo
(cf. below). Witnesses are brought by each party to a case and
are marshalled by the bafilolo. The senior man present claps as
each point is made to mark the recognition of the court, and the
chief himself finally sums up and gives judgement. The advisers
speak when asked a point of precedent or law, and influence the
chief s final decision by black looks or alternatively enthusiastic
clappings of the hand.

(d) Advisory . There is no council or meeting of all the adult men
of the tribe for special occasions, as among many Southern Bantu.
Sub-chiefs have a panel of village elders and relatives to advise
them, while the biggest territorial chiefs have hereditary officials
who combine political and judicial with ritual functions. In the
case of the paramount, these officials—the bakabilo —number
between thirty-five and forty and form an advisory council
on special matters of State. The bakabilo have been described
as having descent as long as that of the chief himself in many
cases and possess sacred relics in their own rights. The power of
these relics is so strong that the Citimukulu is not permitted to
pass through their villages for fear that one chieftainship should

7? °l h f- Bakahil ° are immune from tribute, wore special
feather head-dresses in the old days, and even now claim special

coMtrv 6 ^? t0 ?t glVe ? to a chief when travellin g about the
country They call themselves Fwe Babemba (‘We, the Babemba’),

may not leave the central territory (Lubemba) for long, must be

twf TIT f-? yaI dktrict ’ and kee P taboos similar to
those of the chiefs. They succeed by a special accession

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 109

ceremony and are burled according to particular rites. They are
divided into groups according to the order of their ancestors
arrival in the country, and each has a special office based on the
privileges of his original ancestor, e.g. the care of the royal drum,
the right to sit on a stool in the chiefs presence, or the duty to
call him in the morning by clapping outside his door.

The main duties of the bakabilo in native eyes are ritual, as has
been described. They are in charge of the ceremonies at the sacred
relic shrines and take possession of the babenye when the chief
dies. They alone can purify the chief from the defilement of sex
intercourse so that he is able to enter his relic shrine and perform
the necessary rites there. They are in complete charge of the
accession ceremonies of the paramount and the bigger territorial
chiefs, and some of their number are described as bafingo, or
hereditary buriers of the chief. Besides this, each individual
mukaUlo has his own small ritual duty or privilege, such as lighting
the sacred fire, or forging the blade of the hoe that is to dig the
foundations of the new capital.

Besides their priestly duties, the bakabilo acted as regents at
the death or absence of the chief, and any question of succession
or other matter of tribal importance is placed before the bakabilo,
and the big ^eremonies I witnessed at the chief’s capital were
all made occasions, of such discussions. The procedure is complex,
but an effective method of deliberation. The paramount sends
two special hereditary messengers, also bakabilo , to place the
matter before the council. The senior members speak and if
a difficulty arises they refer the matter to the head priest of the
land, the Cimba, who sits apart with his own following, and gives
decisions on matters of tribal precedent or suggests rewording
decisions to be carried to the chief. Some of the discussion is
carried on in archaic cibemba.

The importance of the bakabilo's council is the check it holds
over the paramount’s power. These are hereditary officials and
therefore cannot be removed at will. Two or three of the bakabilo
have been chased out of the country in the past for overweening
pride, according to tradition, and the Cimba was removed from
office in 1934, but only after the tribe had suffered for many years
from the results of a species of megalomania to which he seemed
to be subject. Otherwise the bakabilo are immune from the
chief’s anger and exert a salutary influence over him by refusing

no

AFRICAN POLITICAL SYSTEMS

to perform the ritual functions that are necessary to the chief’s
state. 1

Other advisory officials consist of the near relatives of the chief
himself. These do not attend discussions as to succession to
chieftainships, but are constantly informed of the progress of
affairs. The paramount’s mother and the Makassa (the eldest
‘son of the chief’) play an important part in this way. In the past
senior members of the royal family seem to have intervened
occasionally when some chief was behaving too outrageously, as,
for instance, in the case of a sub-chief, Fyanifyani, apparently
attacked by a sort of blood-lust. This man was removed from
his office, according to history.

In brief, the Bemba system of government is not a democratic
one in our sense of the word. The elder commoner has fewer
rights to speak on tribal matters than have the Zulu, Swazi, or
even some of the Sotho peoples. The affairs of the icalo are in
the hands of a body of hereditary councillors whose offices and
most of whose deliberations are secret. But I was impressed by
the sense of tribal welfare which these bokabilo showed, and they
were quite able to discuss and shrewdly adapt some old tribal
precedent to modem conditions. Their strength, as regards tribal
government at the present day, is their esprit de corps and sense
of responsibility; their weakness, the fact that in the eyes of the
people and the Government their function is mainly a ritual one.

VI. The Integration of the Tribe

The integration of the tribe depends chiefly on the sentimenl
of tribal cohesion and loyalty to the paramount, and the means
by which the activities of the different districts are brought under
one control in this widely dispersed group. The dogmas of kin¬
ship have been shown again and again to be the basis of tribal
feeling and of the allegiance given to the territorial and paramount
chiefs. In other Bantu tribes there is some tribe-wide organiza¬
tion such as the Nguni regiment system, that seems to act as an
integrating force. There are also forms of public ceremonial at
which all the adult men of the tribe are gathered, or all the warrior
classes. The first-fruit ceremonies of the Swazi or of the Zulu in

1 Drniiig 1934 1 found the paramount living in grass huts. He was unable
to build his new village because the bokabilo, indignant at his behaviour, refused
to perform the foundation ceremony for the new community.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA m

the old days are an example. The big tribal councils of most of the
African peoples described as being attended by ‘every one 1 and
in reality very large meetings, must also act as occasions when the
loyalty of the tribe is fostered. Among the Bemba much of the
tribal ritual is secret, as has been shown, and the advisory council
is composed of what might be called an aristocratic caste. If the
bakabilo meet in sitting on the open ground in the capital, as I
have seen happen, they use archaic language on purpose, so that
the common people cannot understand. It is no occasion for high-
flown oratory or any of the demagogue’s arts. On the other hand,
the Bemba chiefs were formerly considered very nearly divine,
and the belief in their supernatural powers is still strong enough
to. integrate the tribe. The sacredness of the royal ceremonial
largely depends on its secrecy and the fact that only persons of the
right descent can take their part in the ritual. The ordinary people
do not attend the ceremonies except in the case of some inhabitants
of the capital, but they value their secret nature and speak con¬
temptuously of the Bisa and neighbouring tribes with less complex
rites. The number of the bakabilo, each scattered through the
chief’s icalo and each with his own ritual function, sometimes
secret from his own fellows, also adds to the strength of the whole
ceremonial system. Each is insistent that his part is absolutely
essential to the welfare of the tribe, and his own village is con¬
vinced to that effect, too. Another integrating factor is the belief
in royal descent and presence in the society of such a large
number of men and women who claim chiefly rank. These are dis¬
persed all over the country, generally in charge of villages, and they
naturally support the chiefs from whom they derive their power.

As regards the activities of the different if yah, it has been seen
that these are self-contained units and there is no regular provision
for regular meetings of icalo heads. They are linked by the over¬
lordship of the paramount, who acts as judge of their court of
appeal, and the different tiny states are bound together because of
the close relationship between their different chiefs. Messengers
constantly go from one court to another to inquire after family
matters, the children of one chief are sent to be brought up at the
capital of another, the chiefs themselves take office first in one
icalo and then another, and even the Citimukulu takes no important
step, ritual or political, without consulting his ‘brothers’, the big
territorial heads.

112

AFRICAN POLITICAL SYSTEMS

But here again ritual is one of the big integrating forces. The
Citimukulu can initiate a series of sacrifices (ulupepo lukalamba ),
which start at his relic houses and spread to all the shrines through¬
out the land. The bakabilo are sent from Lubemba to bury any of
the bigger territorial chiefs who die in their distant ifyalo and to
install the new heir. The paramount prays for rain on the rare
occasions when it is required, on behalf of the whole tribe. Thus
for ritual purposes, in spite of the quarrels and jealousies between
different lines of the royal family, the whole Bemba country can
be said to act as a whole and to be conscious of its unity. If the
paramount chief were to turn Christian before the political institu¬
tions of this tribe have been considerably adapted, tribal cohesion
would, I think, be very much weakened, whether temporarily or
permanently.

VII. Post-European Changes

The advent of British rule in Northern Rhodesia changed at
once the position of the Bemba chief and his political machinery,
and it continues to do so in an increasing variety of ways. Some
of these changes are due to the actual introduction of new authori-
ties into the area—whether Government officials, missionaries, or
other Europeans—who have either replaced the old Bemba
officials, divided the spheres of authority with them, or intro¬
duced entirely new conceptions of the functions of government
itself. Others seem to be mainly the result of changed economic
conditions, particularly the introduction of money, the institution
of wage labour, and the provision of opportunities for money¬
making in industrial undertakings outside the territory Such
factors, over which the Administration has often little direct con¬
trol, have inevitably affected the position of the Bemba chief.
1 hey have altered the people’s conception of authority, destroyed
the whole basis of labour on which the powers of the chief
epen e , and the old correlation between political authority,
economic privilege, and military strength.

The total effects of white domination on the Bemba political
organization have not yet proceeded to their full length, but it will
be well to indicate some of the changes produced by the introduc-

0f §0vernment ’ e -g- the alteration in the
balance of the old tribal system, and the resultant weakening of
the personal relationship between subject and chief upon which

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 113

the whole structure of authority depended. To do so, it will be
simplest to try to indicate the position when the Europeans first
took over the administration of the country, and to compare this
with the situation produced by the introduction of a modified
system of Indirect Rule in 1929.

(d) New Authorities Introduced . In 1900 north-eastern Rhodesia
was placed under the control of the British South Africa Company
by an Order-in-Council after its officials had established posts in
or near the Bemba country at Kasama, the present administrative
centre of the northern province of Northern Rhodesia, in 1899 and at
Mirongo, near the present Chinsali, in 1896. This administration
continued until 1924, when the Colonial Office assumed control.
To the Bemba, considered a particularly fierce and warlike tribe
before the arrival of the white man, the superior force of the new
administration must have been immediately apparent. The
officials of the British South Africa Company were better armed.
They at once intervened in a case of disputed succession over the
Mwambaship which had then fallen vacant, and appointed their
own nominee instead of his maternal uncle, who was endeavouring
to take the title by force. One by one the functions of the old
chiefs were taken over by the new authorities. New courts of law
were introduced, and though some Native customary law was
administered by the white officials, yet here as elsewhere customs
considered ‘repugnant to natural justice and morality* were
prohibited, this category was such a large one 1 that to the natives
it must have seemed like the introduction of a new code. Certain
totally new offences were also created, e.g. the killing of elephants
and a number of other forms of game, the digging of game-pits,
and the use of primitive iron-smelteries. The penalties for legal
offences were changed too. For mutilation at the hands of the
chief, enslavement, and compensation paid to the injured party
was substituted imprisonment, beating, fines paid to the Govern¬
ment, and the death sentence. New demands on natives’ goods
and services were also made, such as the hut-tax of 3^
afterwards changed to a poll-tax, which now stands at 7$. 6A,
and the enrolment of natives as carriers, road-builders, &c.,
in Government pay. The B.S.A. administration recognized
the Bemba chiefs and in 1916 defined their authority more

1 e.g. accusations of witchcraft, murders for ritual purposes, the use of the
ordeal in determining guilt, &c.

II4 AFRICAN POLITICAL SYSTEMS

exactly, 1 but they were used in the main as executive officials and
shorn of most of their authority and their privileges, such as the

owning of slaves, the possession of arms, the right of mutilation,
the power of administering the poison test, and the collection of
the ivory tusks.

The sanctions for the power of the new administration were, in
native eyes I think, its military strength' and the fact that it had

overcome the once powerful Bemba chiefs, and later, as new
economic values were acquired, its apparently endless wealth. 2
But when talking with elder natives one is aware how largely the
pattern of fear and personal allegiance accorded to the old chiefs
was transferred with little modification to the new authorities.
The same terms are used for both: there is the same assumption
that the tax, game-laws, and even the paid employment of natives
are all dues demanded by the Government for its own aggrandise¬
ment, as was the tribute of the chiefs in the old days. 3 There is
much the same belief in the ruler's complete omnipotence, and a
similar expectation of sudden arbitrary action, even as I noticed,
when the most good-natured and reasonable officials appeared to
be concerned. It is no exaggeration to state that each Government
station is in effect a native capital or umusumba. Each has its dis¬
trict officer, an authority like the chief with a following to whom
allegiance may mean advance, and who is regarded with mingled
fear and loyalty. Each has its force of messengers and police,
employs its own labour.

The missionary bodies in the country must also be regarded
as new authorities set up in the tribe. The White Fathers entered
the Bemba country just ahead of the B.S.A. administration
and set up their first post near Kasama, in the heart of the Bemba
country. They can still be said to dominate this central district,
although the Church of Scotland Mission and the London
Missionary Society also operate elsewhere. Each different
mission station must also be regarded as an umusumba. Many

1 c ^* R e P°rt the Commission appointed to inquire into the financial and
economic position of Northern Rhodesia (1938), p. 179.

2 natives seemed to me to comment most on the ferocity of the
Government officials (ubukali), and the youriger to speak of the wealth of the
Administration.

3 Jt 1S common for old men and women to refer to their sons who have either
been recruited or else gone voluntarily to the mines as having been ‘seized by
the Government’, and to speak of opportunities arranged for the sale of their
gram as having their food ‘seized’ by the native commissioner.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 115

are big establishments, with enormous and impressive cathedrals.
They own and cultivate ground, attract a following, have villages
on their estates regarded as ‘mission people/ just as the hena
musumba are considered to be the personal following of a chief.
Each society, again, has introduced what is, in native eyes, its
own new code of laws, often differing from those of the Govern¬
ment and those of the chiefs, e.g. most missions prohibit polygamy,
some divorce, others beer-drinking, dancing, or religious cere¬
monial of different kinds. In native eyes at present there are
certain well-known rules binding on the Christian members of
the community, sometimes even bringing them into conflict
with the other authorities of the society, the district official and
the chief, and a new category of offence known as fya busenski
(‘things of heathendom’), or fya kale (‘things of the past’) believed
to be strongly condemned. 1

- Besides their own villagers, the missionary societies exert
authority over Christians scattered in nearly every community in
the territory, and their grip over these distant ‘subjects’ must in
some cases be just as strong as those of the chiefs of old days.
At the Roman Catholic missions at any rate, each baby of Christian
parents is registered and summoned at the right time for instruc¬
tion, however far away he or she lives. Each village is constantly
visited by travelling native teachers and evangelists, and by the
white missionaries themselves.

The sanctions of the missionaries’ authority are many. On the
positive side, their teaching and their way of living command
a new allegiance and a new opportunity for advance and their
personalities very often inspire admiration, affection, and personal
loyalty. On the negative side, there is the introduction of a new
supernatural sanction quite as powerful as those that supported
the chief’s authority, 2 and the threat to withdraw the Christian
members of a community in which the Christian law is being
flouted by a headman or prominent member. This acts as a
powerful deterrent in the case of many Roman Catholic villages,

1 This last is, of course, an injustice to the modem missionary, who is often one
of the first to try to enc9urage interest in and respect for native custom.

2 The fear of hell-fire, and, in the case of the older and less educated natives,
the fear of curses seriously believed to be uttered by missionaries, evidently on
the strength of some such statement as ‘God will purnsh you if you behave like
that’—such a belief being almost inevitable in an area where chiefs were thought
to have power to curse.

IX 6 AFRICAN POLITICAL SYSTEMS

since the break-up of the whole settlement may become Inevitable

In such circumstances.

There are no other organized bodies of Europeans In this area
—no big bodies of farmers or settlers, for instance. But it may
be said that all Europeans are In a sense in a position of authority.
They all have the sanction of wealth and the power of employing
a following. All have the high social status that enables them to
talk to a chief as an equal, or more usually as an inferior, and all
are believed by the natives to be backed, whether right or wrong,
by the administrative officers of the district however unjustifiably.

Besides these new authorities introduced into the political arena,
the power of the Bemba chief is inevitably reduced by his economic
position, which must have grown increasingly bad since the first
days of European occupation. The Bemba rulers were never rich
compared to a number of African potentates. Theirs is a poor
country. They possess no cattle which could be converted to
money under modern conditions; the ivory of their country no
longer belongs to them, nor their mineral rights. Land has no
financial value as yet, and the salaries given to chiefs have always
been low compared to the income, say, of the paramount chief of
the Barotse. 1 Added to this, the service on which the Bemba
chiefs depended is cut down by half or more by the absence of
men at the mines, and what remains is often given unwillingly.
These chiefs never apparently exacted court fines as a regular
thing, after the fashion of the cattle-owning Bantu, and have not
yet put a levy on the earnings of men away at the mines, as has
been done in some parts of South Africa.

In view of these facts, it may be asked how the power of the
Bena tjandu survived at all up till 1929, when a determined effort
was made to restore it. Partly because of their closely knit kinship
structure, but also, I think, because of the strength of the super¬
natural beliefs on which their authority was so largely based.
These were certainly weakened by the introduction of Christianity
and the prevention or discouragement of many tribal rites, but it
is impossible to treat them as mere survivals at the present day.
Even young men are affected by such beliefs. For much the same
reasons, the chiefs’ courts continued to function alongside of the

1 The Barotse chief receives an annual subsidy of £1,700, together with
£850 from the British South Africa Company and £1,500 from the Zambesi
saw-mills, as compared to the £60 a year of the Bemba paramount.

THE BEMBA TRIBE OF NORTH-EASTERN RHODESIA 117

District Commissioner's courts in spite of thirty years or so of
non-recognition. Natives took there the cases they believed the
European magistrates did not understand, i.e. ritual matters,
affairs with their roots in past history, and certain civil actions.
Thus there were still some functions which the people believed
the Bemba chief could perform better than all the new white
authorities in the territory, besides the great historic tradition
behind his authority.

Effects of the 1929 Ordinances. By the Native Authority and
Native Courts Ordinances of 1929 a form of Indirect Rule was
introduced into Northern Rhodesia, the power of the authorities
then instituted being still further extended and more closely
defined by a subsequent Ordinance of 1936. Chiefs, and in some
cases councils of chiefs, were constituted as native authorities by
these measures. They were given jurisdiction over definite terri¬
tories, and encouraged to issue orders on matters of hygiene,
bush-burning, the movement of natives, the constitution of
villages, &c. Native courts were also recognized and given juris¬
diction over all cases except witchcraft, murder, issues Involving
Europeans, &c. The chiefs were given salaries, small in actual
fact and quite inadequate to the needs of an administrator at the
present time, 1 yet in the eyes of natives, who are unaccustomed to
see large sums of money at any one time, they were substantial
marks of Government, favour and often described as such. Clerks
and kapasus (messengers), at exceedingly low salaries, were also
attached to the courts. No financial control was given until
recently (1936), when native treasuries were set up. 2

As far as I could judge, 3 this new policy certainly did much to
restore the personal prestige of the chiefs. The loss of an authority
which was largely based on ritual was compensated for to some
extent by the evident support of the Government, Most Bemba
realize clearly that their chiefs are still merely servants of the
Administration and note the fact that they cannot imprison or try
the most important cases,, and that their judgements are liable to
be reversed, but they often commented to me on wliat seemed to

1 Citimukulu receives £60 per annum, Mwamba £s°> an£ ^ other chiefs
less in proportion. For details showing the inadequacy of these amounts cf.
my ‘Tribal Government in Transition*, Joum. African Society, vol. xxxiv, 1935 *

2 The Bemba Treasury had a balance of £143 in 193$ on a sum of £i, 3°3
to be divided between six districts. See Report cited on p. 114*

3 1 arrived in the country the year after the introduction of these measures.

Il8 AFRICAN POLITICAL SYSTEMS

them evident signs of favour bestowed by the new policy on their
rulers. As we saw, the judicial machinery of the Bemba never
really disappeared with the institution of white courts, but the
legal recognition of the chiefs’ courts by the 1929 Ordinance
certainly increased the latter’s status tremendously. Successive
Government reports have described the new native courts as
functioning well, while few charges of excessive fines or hearing
fees have been made against the chiefs. Such complaints as I heard
seemed to come mainly from educated natives belonging to other
areas, particularly from Nyasaland, who found themselves subject
for the first time to Bemba law. In the legislative field, determined
efforts were made by the district officials to restore the self-
confidence of the native authorities and to encourage them, not
only to resume functions of government they had lost, but also
to take on new tasks, such as the issue of orders as to hygiene,
&c. In fact, as regards the personal position of the Bemba chiefs,
it may be said that there was a gradual increase in status due to
Government support, added to a respect and fear which had never
been entirely lost. Their power is in many ways surprising, in
view of their poverty, their lack of means to enforce any decisions
taken, and the presence in their territories of other authorities in
command of their subjects and in particular of large numbers of
Christians bound by codes that are not recognized in the chiefs’
courts. 1

Apart from the chief’s own position, it is necessary to review
the political system as a whole. How far was the old machinery
of Government re-established by the introduction of the 1929
Ordinances? And, more important still, is such a machinery
adapted to the new needs of the tribe ? The answer to the first
question is a decided, ‘No’. The whole balance of authority has
been altered, partly by the changed economic position of the chief
and partly by lack of Government recognition. We saw that the
executive and judicial officials on whom the chief relied were
kept together by hope of rewards and food and in some cases in
virtue of their religious functions. Nowadays the chief is less able
to feed his councillors, principally owing to the decay of the tribute
labour system. He does not consider himself obliged to distribute

1 Some missionary societies have followed the Government policy and have
invited chiefs to take a part in school education committees, &c. j others have
viewed any increase in the chief’s authority as a retrograde step.

THE BEMBA TRIBE OF NORTH-EASTERN* RHODESIA 119

money in the same way as food, and in any case has not sufficient
to enable him to reward his people adequately. So that at a time
when he is asked to take on new functions of administration his
following is becoming smaller and smaller. 1 have seen Citimukulu
hearing cases alone with his paid clerk and watched the bakabilo
melting away during the midst of the discussion of important
matters owing to want of food. This makes an impossible situa¬
tion which may be changed for the better by the greater measure
of financial control which the Government has recently granted
to the native authorities , 1 though the sums now allotted to native
treasuries are small, and lack of funds has always prevented the
compensation of chiefs for the lack of their tribute labour and
other perquisites that has occurred in Nigeria, Barotseland, and
elsewhere.

Government recognition of the political organization of the
tribe and its purposeful adaptation to modern conditions is also
essential at the present time. Apart from the economic break¬
down at the chiefs’ courts which has just been indicated, much
of the trouble has been due to the fact that no serious investigation
of the judicial, executive, and advisory machinery of Government
was made in the first instance. Chiefs were constituted as authori¬
ties with little study of the way in which their orders were to be
enforced. They were listed as ‘members of court’, but, though
headmen and council were mentioned as eligible to sit on such
courts, the presence of the latter was not apparently compulsory.
The bakabilo’s important advisory functions as a tribal council
and a potential regency council were not recognized until anthropo¬
logical research in the area had been made . 1 The unfortunate
result was that the chief felt free to act without this former check
on his power and openly expressed to me his delight in the fact.
The councillors, on the other hand, felt discouraged, and declared:
‘The Government likes the chiefs. It does not listen to us, the
Babemba? Hence a political system that could never have been
described as democratic now provides less check than ever on the
chief’s authority.

The difficulty from an administrative point of view is evident.
Here is a system of political authority based largely on hereditary
ritual privilege. To abandon the bakabilo council is to do without
a body of men with strong traditions of government and a sense of

1 Twenty-five of the bakabilo are now given £1 a year (see Report cited, p. 144).

120

AFRICAN POLITICAL SYSTEMS

the public weal; to rely on them without at the same time training

them in new functions, and adding to them other elements,
especially for the more educated members of the tribe, is to build
on a foundation that cannot endure, and to deny to the commoner
the experience of administration he will require to have in the
future. Similarly, the co-ordination of the activities of each of the
smaller political units known as ifyalo is at present one which is
based on the intimate relationship between the chiefs ruling each,
and the ritual which unites them. There is no official representa¬
tion of each district on the Paramount Council, since all the baka~
bilo are drawn from his own territory, however much they consider
themselves responsible for the affairs of the ■whole tribe. Ritual
prohibitions still prevent the frequent meetings of the big terri¬
torial chiefs themselves, though they do occasionally come to¬
gether at Government ndabas outside their own capitals. Hence,
if it ever became necessary to provide for closer co-ordination of
the different districts or for local representation, it appears that a
general tribal council of the type of the National Council of the
Basuto or the libandla of the Swazi would have to be constituted
in this tribe.

THE KINGDOM OF ANKOLE IN UGANDA

By K. Oberg

I. Traditional mid Historical Background

A NKOLE is but one of a series of small Native kingdoms
^ stretching from north to south along the western borders

of Uganda Protectorate. Both geographically and anthropolo¬
gically, this is an interesting region. Bounded on the east by the
great barrier of Lake Victoria and on the west by the mountain
mass of Ruwenzori, and a chain of lakes extending from Lake
Albert to Lake Tanganyika, it forms a corridor leading from the
broad grasslands of the Upper Nile to the plateaus of Belgian
Ruanda and Tanganyika Territory. Geographically this corridor
is typical African savanna with its rolling grass-covered hills and
sparse acacia scrub.

Some time in the dim past this region was occupied by Bantu¬
speaking, agricultural Negroes. The rainfall, though scanty, was
sufficient to permit a fairly even distribution of the population over
the country, scattered thinly in the drier plains of the east, but
more densely in the hilly regions of the west. Later in the history
of Africa this same corridor provided a pathway over which waves
of Hamitic or Hamiticized Negro cattle people migrated south¬
ward. These pastoralists, with their vast herds of long-horned
cattle, are believed to have been crowded southward from southern
Abyssinia and many believe them to be of Galla origin. Whatever
be the exact location of their original home or their specific tribal
connexion or the reasons for their migration, there is no doubt
that these people are closely linked to the Hamites in blood and in
certain customs concerning cattle. What is more important,
however, is the fact that whenever these pastoralists settled upon
territory already occupied by the Bantu agriculturists they made
a uniform adjustment, they conquered the agriculturalists, and
established themselves as a ruling class. Thus when the British
took over the management of Uganda, some forty years ago, they
found everywhere in this corridor the pastoralists as rulers and the
agriculturalists as serfs. The pastoralists calling themselves
variously as Bahima or Bafauma and the agriculturalists as Bairn or

1 22

AFRICAN POLITICAL SYSTEMS

Bahera. Although it is said that the Bahima were once united and
held sway over a great empire called Kitara, in historic times this
territory was split up into kingdoms stretching from Bunyoro

through Toro and Ankole to Ruanda in Belgian territory. The
kingdom of Buganda also belongs to this classification, but,
owing to the greater percentage of the agricultural population,
sharp racial differences soon disappeared, the royal family alone
stressing its foreign descent.

This upland corridor, then, is the stage upon which Bahima and
Bairu have for centuries played their political drama. Intriguing
and heroic as the opening scenes of this drama undoubtedly were,
the story as historical fact is lost to us for ever. But to every
Muhima this past is a living, fascinating reality. Whether it be
embodied in song, recitation or fireside tale, the theme is the same.
It is of raiding, of wandering, of battle between clans and kings, of
famine and disease, of sorcery and sacrifice. Through this vast
body of myth and legend, we are able to grasp something of the
epic nature of the struggle which took place on the plains of
western Uganda and which, in one instance, gave rise to the
kingdom of Ankole.

Long ago, these legends say, there were Bahima and Bairu in the
land. The Bahima lived in eastern Ankole with their cattle while
the Bairu tilled the soil in the west. In those days the Bahima had
neither king nor chiefs, but important men in the clans settled
disputes. Among the rich men the following are still remembered:
Nyawera lived in Kashari and belonged to the Abaitera clan;
Rwazigami lived in Rugondo and belonged to the Abasite clan;
Ishemurindwa of the Abaishekatwa clan lived in Masha; Karara
of the Abakoboza clan lived in Ruanda; Rwanyakizha of the Aba-
rami lived in Nshara; while Mariza of the Abararira clan lived in
Bukanga. There were other clans as well, too numerous to mention
here. These Bahima are spoken of to-day as being the first Bahima
of Ankole. There are no stories of how they got there. Other
clans came into Ankole later from the neighbouring countries.

The Bairu lived in Rwanpara, Shema, Buhwezhu, and Igara.
It is not known whether they had clans or whether they were
organized under chiefs. There were no wars between the Bahima
and the Bairu in those days, each living in his own section of the
land and trading beer and millet for milk and butter.

Then very suddenly a strange people appeared. They were

THE KINGDOM OF ANKOLE IN UGANDA 123

called the Abachwezi. From the cycle of songs and legends, these
Abachwezi seem to be the same figures who played such an impor¬
tant role in the past of the Banyoro, Baganda, Batoro, Abakaragwe,
and, at least certain groups, among the Banyanruanda. While
statements as to their origin and disappearance differ, there is at
bottom a fundamental agreement about their character, doings,
and direction of their movements. All legends point out the fact,
for instance, that the Abachwezi came from the north, that they
were not very numerous, that they conquered the people in their
way and then disappeared southward finally vanishing into lakes
or craters.

According to legend, these Abachwezi were wonderful people.
‘They were like the Bahima, but more brilliant. One could not look
them in the face because their eyes were so bright that it hurt one’s
own eyes to look at them. It was like looking at the sun. They
wore bark cloth and went about in cow-hide sandals. Their
women covered their faces in public and were guarded by eunuchs.
The important Abachwezi built large grass houses and had their
kraals near by. They had many cattle and lived on milk, meat, and
beer, especially beer mixed with honey. They were great hunters
and magicians.’ This description of the Abachwezi is the most
matter-of-fact. Most accounts deal with their superhuman feats,
their terrible strength, their power of making themselves invisible,
their wealth in cattle, the beauty of their women, and the ruthless
domination of all whom they conquered.

Most stories agree in stating that Ndahura was the great con¬
queror, the leading Omuchwezi. In Ankole, Ndahura is not as
well known as in Toro and Bunyoro, where he is considered as the
first Abachwezi king. He is there considered the son of a former
king and a woman called Nyinyamwiru (mother of Mwiru) and it
was he who is said to have consolidated the Kitara Empire. After
the conquest he reorganized his kingdom into districts, appointing
his sons and henchmen as rulers. To Ankole he sent Wamara, who
is there considered the first Abachwezi king. The word ‘Wamara’
comes from the verb okumara, to finish or to complete. Many
Banyankole explained that he finished what his father began—
namely, the conquest of an empire. Wamara lived at Bwera, which
was then a part of greater Ankole, and ruled over Ankole, Karagwe,
and a part of southern Buganda. Wamara was said to be kind to his
henchmen and always ready to help them out of difficulties. When

124

AFRICAN POLITICAL SYSTEMS

their cattle died, he gave them cows from his own herd and helped
to pay their marriage-prices. By means of his magical powers, he
could make rain, make barren women bear children, and prevent
disease among cattle. Wamara had many sons and relatives, who
are all described in song and legend. There was Murindwa, the
eldest son of Wamara, who was noted for his wisdom; there was
Mugenyi, who herded his cows at Bwera and is said to have built
the earthworks at Biggo; Kazoba was another son, ‘beautiful as a
berry and as timid as a marsh antelope’ ; Ibona was a diviner;
Riangombe, a hunter; Kagoro, a conqueror of new pasture lands,
foremost in all wars; Mugasha was the maker of water holes for
cattle.

The Abachwezi, then, lived in Ankole, ruling the country,
herding their cattle and performing miracles, but they did not

remain long. Misfortune came to them. First, the people began to
disobey the Abachwezi and even their wives turned against them.

Then the cows began to die and the people to sicken of a new
disease. When the cow Bihogo and the leading bull died, the
Abachwezi became worried and wanted tamiove on. The final

catastrophe was the murder of Murindwa by the members of the
Abasingo clan. He was said to have been thrown into a pit and to
have died there. But before leaving, Kagoro avenged the death of
his brother by killing many of the Abasingos, and Wamara cursed
the clan and to this day the curse is maintained in Ankole. The
Abachwezi then moved on into Karagwe with what cattle they had

left. Mugasha is the tragic hero of this retreat On being forced to

leave Ankole, he became very sad and tried to commit suicide. The
other Abachwezi, however, prevented him from accomplishing
this by magic. Instead of disappearing into Lake Kyaikambara,
like the other Abachwezi, Mugasha went to the Sesse Islands in
Lake Victoria and he is said to be there still.

While the Abachwezi were fleeing from Ankole, Katuku, a
Mwiru headman, overtook them and persuaded Ruhinda, one of
the younger Abachwezi, and his mother, Nzhunwakyi, to return
to Ankde with turn. There was a long debate before Ruhinda
final y assented. He is then said to have been hidden along with the
tilt ^ Bagyendanwa ' h ? Katuku and his follower/for some

S nf Vi* Pe T ° n “. a L gak mled over ^ole, Ruhinda came
the R hldm? ? d , eStaklshed himself as ruler of the Bahima and
the Bairu, and thus he became the founder of the Abahinda

THE KINGDOM OF ANKOLE IN UGANDA 125

dynasty, which rales Ankole to this day. Many Banyankole to-day
firmly believe that the Abachwezi did not die, but disappeared
and that they will return again to rale over them. In the mean¬
time, the spirits of the Abachwezi still rule over the land and a cult
has grown up, the members of which make periodic offerings to the
Abachwezi spirits. This Abachwezi spirit worship is to-day known
as the Emandwa cult. Furthermore, in Ankole there are many
places and relics connected with the Abachwezi. These places
have become sacred to the Banyankole and are avoided by them,
and the relics have become symbols of kingship. So impressed
were the Banyankole with the Abachwezi that when the white men
came to Ankole they believed them to be the Abachwezi returned.
The Europeans were different, more powerful, able to do unac¬
countable things. The Europeans, like the Abachwezi, are able to
travel in the air, to make a fire without leaving ashes, and to
travel over the country with great speed. Another tale has it that
the Europeans are not really the Abachwezi, but their servants
sent to punish the Banyankole for their ill treatment of their former
rulers.

The cycle of Abachwezi legends, then, is the Muhima’s version
of his cultural history, particularly as it relates to the origin of his
political institutions. We cannot, of course, consider this version
as exact history. Yet its sociological significance is far-reaching.
It describes Ankole as first occupied by the agricultural Bairu and
a few pastoral Bahima, living in relative isolation and without a
developed political organization. It describes subsequent Bahima
migrations, a period of struggle, and a final subjugation of the
Bairu by the Bahima and the establishment of a kingdom. But
even more than this, it provides the political structure with a tradi¬
tional legendary background which lends to it a traditional
sanctity and a foundation of absolutism and permanence.

But we do not need to go to native legends in order to account
for the origin of the Banyankole kingdom. Evidence lies before the
student on every hand. Even to-day we can observe the environ¬
mental and social forces which gave rise to the particular com¬
plexion of Banyankole society and its political institutions.

The role of the environmental factors of climate and topography
in bringing the pastoral Bahima and the agricultural Bairu into
contact cannot be underestimated. Ankole, as we have seen, is a
section of a long and narrow belt of savanna country stretching

126

AFRICAN POLITICAL SYSTEMS

along the eastern side of the western Rift Valley from Lake Albert
in the Sudan to Lake Kivu in Ruanda. It is fairly well established
that Hamiticized negro peoples migrated southward along this
route and that in time this belt of grassland was filled with cattle-
keepers. Similarly, climatic conditions made possible the settle¬
ment of this area by relatively dense agricultural populations.

But contact due to environmental conditions alone does not
account for permanent subjugation, the payment of tribute, and a
state structure. Isolation, segregation, and extermination were
alternative w T ays of adjustment. Isolation was not possible, due to
the nature of the country and the density of the population. All
the Bairn could not move into the relatively small hilly areas. On
the other hand, as we have shown, legends relate a stage of segre¬
gation before the pressure of Bahima population brought about a
general settlement of Ankole by the pastoralists. Extermination
was perhaps possible, but the Bahima chose to dominate the Bairn
because it paid to dominate. Although the agricultural technique
of the Bairn did not produce a great surplus, it could produce,
under pressure, enough beer and millet to make domination profit¬
able. In this connexion, we must always remember that the
Bairu had to supply a population only one-tenth its own size. Had
the numbers been reversed, exploitation would not, perhaps, have
been successful. On the other hand, agricultural production was
not such that it could have supported the Bahima entirely. The
Bahima, then as now, lived upon their cattle and forced their serfs
to give them as much beer, millet, and labour as possible without
destroying their source of supply. In this connexion, it might be
illuminating to contrast the situation in which the Masai found
themselves. As they swept down in the extensive plains of the
eastern Rift Valley region, they found there only a few wandering
Wanderobo hunters who neither interfered with the pastoral
habits of the Masai nor offered possibilities for economic exploita-
tion. Exchange relationships were established, but exploitation
leading to political domination through a state organization did
not arise.

Bahima domination of the Bairu arose not only because these
racially and economically different people were brought into con¬
tact m large numbers by environmental conditions and because it
was economically profitable, but also because the Bahima were
able to dominate. Bahima herdsmen, accustomed to protecting

127

THE KINGDOM OF ANKOLE IN UGANDA

their herds from animal and human enemies, were individually

superior fighting men. Constant raiding and counter-raiding

developed a military discipline which could be expanded and put
to political uses. The organization of the Bahima kraal was a
larger collective enterprise than the Bairn homestead. The
unilateral ekyika or lineage offered wider political and military
co-operation than the relatively smaller Bairn oruganda or ex¬
tended family. Thus, even without a further development of the
political organization, the Bahima had the advantage in fighting
experience and co-operation.

Once the Bahima of Ankole had conquered the Bairu and had
imposed their will through a state organization, they were faced by
a new situation, they had to defend their country, their cattle and
the Bairu subjects from external attack. Defence forces and
counter-raiding were not a guarantee of security. The most
satisfactory method of preventing aggression lay in the permanent
subjugation of raiders. Conquest of other cattle people, less
strongly* organized, became a necessary feature of state defence.
Here again we can contrast the situation in which the Bahima
found themselves with that met by the Masai. The Masai were
pre-eminently cattle-raiders, making sudden attacks upon the
villages and homesteads of their settled neighbours, taking what
cattle and goods they could find, and then retreating to their
plains. They did not invade the territory of their neighbours, for
they did not require expansion of their pasture lands, nor was the
land of their neighbours, like that of the Kikuyu and the Kavir-
ondo, ideal cattle country. Furthermore, the Masai were not
subjected to permanent pressure by the surrounding tribes. They
were primarily attackers and not attacked.

While conquest of surrounding cattle people was imposed by
the needs of defence, it had its aspect of economic profit. It paid
to dominate these weaker groups, for tribute in cattle could be
extracted from them. Conquered cattle people came under the
rule of a king’s representative, who undertook the collection of
tribute and its presentation to the king. An interesting feature of
these conquered cattle people was that, being Bahima, they soon
amalgamated with their conquerors. The Bahima of Empororo,
who were formerly independent, were conquered and for a time
paid regular tribute, but, with the increase of pressure from
Ruanda, they fought along with their conquerors and were

128 AFRICAN POLITICAL SYSTEMS

subsequently given equal political and legal status with the Bahima
of Ankole,

Exactly how the Bahima of Ankole were organized into a State
and how they created the political society of the Banyankole
kingdom we cannot now tell. But the elements of the situation
as we have enumerated them are observable. There were external
conditions to which the Bahima adjusted themselves and by the
detailed analysis of the processes of adjustment these conditions
became evident. This adjustment process corresponds to the
functions of the Bahima State. First among these functions was
the domination of the Baira, expressed by inferior legal status and
the obligation of tribute payment; and along with inferior legal
status went inferior social status amounting essentially to a caste
difference. Secondly, the State defended the territory and the
people of Ankole from external raiders and conquerors. Thirdly,
the State embarked upon a programme of conquest which was
limited only by similar ventures on the part of neighbouring
kingdoms.

II. Political Status

What is political action but the creation and destruction of forms
of social organization through the exercise of organized power ?
No sooner were the ethnically different Bahima and Bairn brought
into contact by the environmental and social forces already des¬
cribed than they were forced to define not only their relationships
to one another, but also to modify the relationships binding the
members of each group to one another.

No longer were the Bahima cattle men free agents, united in
extended families and loosely knit lineages and clans; they were
now also members of a political group. If the Bahima were going
to further their interests as Bahima, they had to organize and act
together as Bahima. At bottom this new relationship was based
on Bahimaship—upon race and cattle-ownership. But this special
political bond had to be created, had to be consciously entered
into. It involved leadership, co-operation, submission to authority.
It gave rise to kingship and the dynastic principle, the organiza¬
tion of military forces and chieftainship. In short, it welded the
Bahima into a State, the nucleus of the Banyankole kingdom.

This new political relationship was established through
okutoizha, or clientship. A Muhima cattle-owner would go before

THE KINGDOM OF ANKOLB IN UGANDA 139

the Mugabe, or king, and swear to follow him in war and would
undertake to give the Mugabe a number of cattle periodically to
keep this relationship alive. On the other hand, clientship could
be broken by the omuioizha , or client, refusing to pay homage.
This was a perfectly recognized way of breaking off the relation¬
ship, and it was only when a number of Bahima banded together
in order to defy the king more effectively that this act was con¬
sidered rebellion. Even then, if the rebellious people resumed
homage payments, they would be pardoned by the king. If, how¬
ever, a Muhima induced others to pay homage to himself and
raided cattle without giving the Mugabe a share, he could be
accused of treason and the Mugabe would move his forces to
suppress him.

Clientship involved a number of obligations on the part of the
client towards the Mugabe. Foremost among these duties was.
military service. Every Muhima, even if he were not a member of
a military band, had to go to war when called upon. Any cattle
which a Muhima acquired through a private raid were claimed by
the Mugabe and a part of them had to be handed over to him.
Clientship obliged every Muhima to make periodic .visits to the
Mugabe’s arurembo (kraal), with homage payments. He was also
obliged to assent to the giving of cattle to the Mugabe’s collector
when the king was in need of cattle. Whenever a Muhima died,
his heir had to report to the Mugabe and renew the bond of
clientship by giving a ‘cow of burial’.

In return for military service and the payment of homage, the
client received protection. First, the Mugabe undertook to shield
the cattle of his client from cattle-raiders and to retaliate when his
client had suffered from raids. If a client had lost all of his cattle
through raid or disease, the Mugabe was obliged to help the man
start a new herd. Secondly, the Mugabe maintained peace
between his clients. No client was permitted to raid or steal the
cattle of another client or to do harm to his person or dependants.
If breaches of the peace occurred, the transgressor was accused
and tried before the Mugabe. In cases of murder, the Mugabe
granted the kinsmen the right of blood revenge. Finally, the
Mugabe was instrumental in enabling his clients to enlarge their
herds and pasturage by raids and conquest.

To sum up, then, the Bahima State consisted of the cattle-
owning freemen and their leader, the Mugabe. The specific tie

13°

AFRICAN POLITICAL SYSTEMS

which bound the herdsmen to their leader was in the nature of
clientage, obutoizha. The Bahima alone were politically organized
in that they combined to create and maintain the particular system
of differential relationships which existed in the kingdom. The
fundamental force which the State exercised in maintaining these
relationships was military power. Now let us examine the ties
which bound the subject peoples to the king.

To the Bahima, the word ‘Bairn’ signifies serfdom, a legal status
inferior to that existing between themselves. The index of Bairu
status was race and the dependence upon agriculture for a liveli¬
hood. The difference in status is perhaps best expressed by a
statement of the limitations of Bairu rights.

The Bairu were not permitted to own productive cows. For
services rendered to the Bahima, they were sometimes given
barren cows and bull calves. These cattle the Bairu either kept
for making marriage payments or slaughtered for food. If a
Mwiru did have productive cows in his possession, any Muhima
could take them away from him. There is a story among the
Bairu that long ago they owned cattle, but that these cattle were
taken from them by the invading Bahima. Some veterinary
officers in western Uganda believe that this is true and that the
cattle which the Bairu owned were of a different breed from the
present-day Ankole longhorn. This belief they base upon the
existence of shorter homed stock upon the fringes of Bahima
country, as, for instance, the cattle of the Bakiga of Kigezi.

The social distinction between the Bahima and the Bairu was
maintained by a strict prohibition of marriage. No Mwiru could
marry a Muhima woman. The Bahima, when questioned upon
this matter, laugh and say that such a marriage is quite unthink¬
able. Not only is the idea of such a marriage repugnant to the
Bahima, but the validation impossible, as the Bairu, in former
times, did not possess the cattle necessary for the bride-price.
Bahima men did not marry Bairu women, for it was illegal to give
the Bairu cattle, which alone legitimized marriage and offspring.
On the other hand, however, Bahima men took concubines from
among Bairu girls. These women had no status as married women
and were usually described as servant girls. Bairu concubines were
especially common among Bahima chiefs and gave rise to a class
of half-castes known as Abambari. From the point of view of
legal status, the Abambari were classed as Bairu, but personal

THE KINGDOM OF ANKOLE IN UGANDA 13 x

consideration often modified the strict rigour of the rule. A
Muhima chief or cattle-owner without sons by a Muhima marriage
would make an Illegitimate son his heir. In time entire lineages
were formed from such unions. These half-caste sub-clans the
Bahima distinguished by calling them the people of a certain man
instead of the children of a certain man, which is the name given
to a sub-clan of pure Bahima stock. The effect of concubinage is
quite noticeable when one compares the physical types of the
chiefly class with those of the ordinary herdsman of districts more
remote from the agricultural sections. One finds a markedly
larger percentage of dark Bantu types among the chiefly class.

The military organization, we have seen, was In the hands of
the Mugabe, who instructed certain chiefs to form warrior bands
for the protection of the borders. No band could be formed with¬
out. the express wishes of the king. While every Muhima was
liable for military service, the Bairn were, on the contrary, barred
from serving in these bands. The Bairn thus lacked the military
training and discipline necessary for effecting any change in their
status.

High official positions were likewise barred to the Baku. No
Mwiru, for Instance, could become an enganzt or an omugar&gwe.
The abaktmgu, however, appointed Baku assistants who aided
them in the collection of tribute in the various districts. These
assistants were also called abakungu and were considered by the
Baku as district chiefs. The Bahima, however, claim that these
individuals never had chiefly status.

Perhaps the most outstanding characteristic of Balm serfdom
was the rule that under no circumstance could a Mwiru kill a
Muhima. The right of blood revenge which was exercised by the
extended families of the Baku among themselves could not be
extended to the Bahima. If a Muhima killed a Mwiru, the
extended family of the murdered man could not claim blood re¬
venge, although it sometimes was able to exact compensation
through the agency of the Mugabe. The Bahima, on the other
hand, could avenge the death of a kinsman if he were murdered
by a Mwiru without consulting the Mugabe.

The Baku had no political status. They had no recognized
means by which they could alter the inferior legal rank imposed
upon them. The exploitation of the Baku by the Bahima took
the form of tribute payment in food and labour, and for this

x 3 2 AFRICAN POLITICAL SYSTEMS

purpose the Bahima endeavoured to keep them in subjection.
But the Baira were not outside of the law, inferior as their status
was to that of the Bahima. Within its own sphere, this status had
its positive aspects. Tribute gathering was so organized that only
chiefs could exercise it at will. Small cattle-owners had no legal
right to exact tribute from the Bairu.

This does not mean, however, that the chiefs alone benefited
from the tribute collected. It was a common practice for Bahima
herdsmen to visit their chiefs, sometimes remaining for several
days at the chiefs 1 kraals. It was a chief’s duty to provide these
men with beer and millet porridge. Thus food that came to the
chiefs kraal in the form of tribute Was later distributed among
the Bahima as a whole. Every chief had a number of Baira crafts¬
men who made spears, milkpots, and watering pails. These
articles, too, were obtained by the ordinary Bahima from the
chiefs. Herdsmen, of course, could obtain these articles direct
from the,Baira through barter and they did so to a limited extent.
But organized tribute and its distribution checked exchanges
which would otherwise have been quite extensive.

Unauthorized tribute collection was considered robbery and
was punished by the Mugabe. Any Mwira could go before the
Mugabe or one of his chiefs and complain of ill treatment and
could claim compensation for damages. To make his claim more
effective, a Mwira would take special gifts to the Mugabe and
thus claim protection. In other words, although the Bairu
system of rights was narrower and more restricted than the
fuller status of the Bahima, this system was still protected by
the Bahima State.

Another class which formed a part of the Banyankole kingdom
consisted of conquered Bahima Abatoro who had formerly con¬
stituted chieftainships or parts of other kingdoms. Over these
people the Mugabe would appoint an overlord who forced them
to pay tribute in cattle and who put down any attempts at rebellion.
These people being of the same race and economic status would,
in time, amalgamate with the Bahima of Ankole. An Omutoro
could become the Mugabe’s client by paying obutoizha, after
which he enjoyed the full rights of a Muhima.

The Abatoro , although not having equal status with the
Abatmzha or clients, fared better than the Baira serfs. There was
no bar to intermarriage and blood revenge could be exacted, this

THE KINGDOM OF ANKOLE IN UGANDA

133

right being derived from the underlying racial similarity and clan
rights. The breaking up of the Ahatoro as a class was gradual
and went through a process of individual shift of allegiance to
the Mugabe. On the other hand, the Mugabe's representative in
these districts sometimes endeavoured to establish himself as an
independent chief by rebelling against the Mugabe. Repeated
rebellion often meant the complete confiscation of the cattle of
the Ahatoro and the killing off of the cattle-owners, the women
and children of these cattle-owners being taken and distributed
by the Mugabe among his followers.

The Ahahuku , or slaves, were another class of subject people
in the Banyankole kingdom. Very little could be learned about
slavery in Ankole. Slaves were owned by the Mugabe, the leading
chiefs, and the wealthier cattle-owners, and they consisted entirely
of Bairn captured in raids made upon neighbouring kingdoms.
Slaves had their ears cut off so that if they ran away they could
be recognized and recaptured. Slaves were used as hewers of
wood, drawers of water, and as butchers. There is nothing to
indicate that slaves were sold or exchanged, although chiefs gave
each other slaves as presents. While the slaves performed menial
tasks, it cannot be said that their lot was any harder than that of
the Bairn craftsmen who formed a part of every chiefs household.
Being a prisoner of war, the slave had no legal status in the
community and was the private property of the person who owned
him and who had the right to do as he pleased with him.

From all accounts, slavery was restricted to the very wealthy
and slaves were restricted in numbers. Only those individuals
who had sufficient surplus wealth could afford to keep slaves.
When the Bahima are asked why they did not keep slaves as
herdsmen, they answer that they could not trust them and that
they would have had to accompany them while herding. Slaves,
they say, were used only to clean the kraals and to bring wood
and water. In agriculture, with Bairu tools and techniques,
slavery would not pay. Neither agricultural technique nor craft
specialization had developed far enough to make slavery on
a large-scale economically profitable.

Although supreme political and judicial authority was invested
in the Mugabe as the representative of the politically organized
Bahima, a certain amount of judicial and political power was left
to both the Bahima and Baira extended families. The function

*34

AFRICAN POLITICAL SYSTEMS

of the Mugabe and Ms chiefs lay more in giving judgements than
in meting out punishments. Moreover, there was no police
organization to guard life and property. How then did members
of the community guard themselves against criminal actions and
aid in the carrying out of sentences passed by the Mugabe ? It is
just here that the extended family fulfilled its important role. As
a political unit, it discouraged attacks upon its members by indi¬
vidual malefactors. Once a crime had been committed, the head
of the extended family took the matter before the Mugabe. In
cases of murder, the Mugabe would grant the right of blood
revenge, which, however, had to be carried out by the members
of the injured extended family. In lesser offences the judgement
of the Mugabe was generally sufficient to settle a dispute. The
extended family, therefore, guaranteed the rights of its members
in the community against the attacks of individual offenders of
customary law and practice. In matters concerning an extended
family alone, judicial authority was left almost entirely in the
hands of the head of this group. Murder within the extended
family was not a matter for the Mugabe to decide, but was settled
by the nyinyeka , or head of the extended family.

In summary, we might say that from the standpoint of political
and legal status the members of the Banyankole kingdom did not
form a homogeneous mass, but were distinguished by a wide
range of rights and prohibitions, resulting in a stratification of
society into classes. At the top was the Bahima State with its
governing nucleus centring around the Mugabe. Below were the
subject classes of the Baira, the Abatoro and the Abahuku. The
caste nature of this stratification was pronounced, resting ulti¬
mately on racial and economic differences.

The complex working of this political society becomes intel¬
ligible, not only by determining the roles played by the various
parts, but by observing the genetic relationship of these parts.
The status of the Baira, for instance, as a subject class, is not fully
explained by stating that they paid tribute and were prohibited
from possessing cattle, but by showing that this status was imposed
and maintained by the Bahima as a militarily organized group.
The Bahima-Baira relationship was a Bahima invention. If we
contrast this class difference with the political relationship existing
among the Bahima, the distinction becomes clear. The politically
organized Bahima State was an association of free men expressing

THE KINGDOM OF ANKOLE IN UGANDA 135

their unity in terms of clientship, this unity arising as a spontaneous
response to well-defined external conditions. Clientship, as has
been shown, can be described- by its functions. Clientship as
a system of co-operation carried out such collective enterprises
as raiding, conquest, and domination, and served as a system of
mutual insurance against the risks inherent in a raiding community.
Exchanges of cattle among the Bahima were free exchanges
depending upon the mutual interests of the parties concerned.
The Bairu-Bahima relationship, or Bairn serfdom, on the con¬
trary, was not a system of co-operation of this kind. The Bahima
and the Bairn did not co-operate in collective activities, economic
or political, nor can tribute payment be termed Tree exchange'.
We might contest that the Bairn received protection for the
services which they rendered to their masters, the Bahima. Yet
if we carefully analyse this protection, it appears to be no different
from that which the Bahima provided for their cattle, land,
chattels, and slaves. And, moreover, the Bairu had to be protected
from the Bahima of neighbouring kingdoms and not from other
Bairu.

On the other hand, the distinction should not be pressed too
hard, for serfdom is not slavery. The Bairu had well-defined rights
which the slaves did not possess. Furthermore, Banyankole
society was not static. The sharp differences between the Bahima
and the Bairu which have been stressed in the preceding analysis
were subjected to a steady pressure of social forces making for
their obliteration. In spite of the prohibition of intermarriage,
miscegenation took place. A class of half-castes arose known as
Ahambari, whose status, although not clearly defined, was not
always that of the Bairu. An omwambari whose father was a chief
often came into the possession of cattle and was recognized as
a man of importance, if of uncertain status. In our description
of the kinship organization, we had occasion to refer to a number
of Bahima sub-clans of pure descent. It is also said that the
present Mugabe's father established a Bairu band of warriors in
order to counteract the determined effort of the Banyanraanda
to conquer Ankole. From reports given to me by the natives of
Toro and from Roscoe's account of the Bakitara, it appears that
the Bairu-Bahima amalgamation had proceeded much farther m
these kingdoms than in Ankole. In spite of these forces making
for uniformity,.the traditional political structure of the Banyankole

AFRICAN POLITICAL SYSTEMS

136

kingdom was essentially stratified, depending upon the Bahima
as a dominating power.

III. The King and the Royal Kraal

So far we have stressed the forces which brought the ethnically
different Bahima and Bairn together and the resulting social
stratification with its formalization into strictly defined political
relationships. The king, or Mugabe, we observed, formed the
centre of this system of relationships. The exercise of power
demanded still further developments. A system of government
grew up round the king’s person, consisting of office holders, the
military bands, and the host of servants and specialists to uphold
the king’s dignity and authority and to carry out his orders as the
leader of the politically organized Bahima ruling caste.

The position of the Mugabe was exalted, his authority supreme,
his leadership all-embracing. ’ As high status was sanctioned, in the
first place, by his descent from Ruhinda, the originator of ■ the
Abaninda dynasty, and, in the second place, by his possession of
the symbols of kingship—the royal drum, Bagyendanwa y and the
beaded veil, Rutare. Both descent and the symbols of kingship are
said to date from the times of the semi-mythical Abachwezi kings.
The word "Mugabe’ is derived from the verb okugaha , to give, and
seems to imply that the Mugabe was a giver, although many
Banyankole describe the Mugabe as one to whom the Mugabeship
was given by the Abachwezi. The power of the Mugabe extended
over the free, cattle-owning herdsmen of Ankole who were bound
to him by mutual ties of defence and aggression, over conquered
herdsmen who paid him tribute, and over any Bairu peasants who
lived upon the tribal territory. Even to-day, when kingship in
Ankole has lost its essential purpose and much of its colour, its
original form is revealed to us by countless songs and stories which
are sung and told around firesides in Bahima kraals.

# Physical, magical, and religious powers were invested in the
king’s person. In song and in address he was called the "lion’, the
fiercest and most courageous of animal cattle-raiders. He was
called the leading bull’, for cattle increased through him by raid
and gift. He was called the "territory of Ankole’ for he had
"eaten’ the pastoral lands at his accession and defended them
against agression. He was called the "drum’, for like the drum
he maintained the unity of the men under his power. He was

THE KINGDOM OF ANKOLE IN UGANDA r 37

called the ‘moon’, for through the moon he had power to drive
away evil and bring fortune to the tribe. Power, then, both physical
and spiritual, was the inherent quality of kingship. And when the
physical powers of the king waned, through approaching age, these
kingly powers were believed to wane with them. No king, there¬
fore, was permitted to age or weaken. When sickness ’or age
brought on debility, the Mugabe took poison, which was prepared
for him by his magicians, and died, making way for a new, virile
king who could maintain the unity of the kingdom and wage
successful wars against external enemies.

The legal status of the Mugabe gave him the highest political
authority. Appointments to office were ultimately in his hands,
as was the decision for war or peace. From among his rela¬
tives, the Mugabe appointed the leaders of his military bands
and his favourite chief, or enganzi. Even those functionaries
which custom decreed should be selected from certain clans, as his
drum-keepers and personal servants, the Mugabe could refuse to
recognize. In other words, while the clan held the office, the
Mugabe selected the individual who was to fill that office. More¬
over, the Mugabe could demand the services of any individual in
his kingdom as he could demand any woman for his wife or could
claim any cattle he wished. As one would suspect, the Mugabe
could dismiss, office-holders for incompetency, personal incom¬
patibility, or because they brought him bad luck.

The Mugabe’s legal status gave him also the position of
supreme judicial authority. He had the right to p unish individuals
by death, exile, beating, torture, and cursing. He could con¬
fiscate the cattle of any of his subjects. He could prevent the
execution of his people by his chiefs for criminal offences and could
override the judicial decisions of the kinship-groups. In disputes
involving two lineages, the Mugabe alone could grant the right of
blood revenge. Excepting among rebellious subjects, the Mugabe
did not initiate legal action. All other cases had to be brought
before him.

Although the political and judicial powers of the Mugabe were
great, they were in the last analysis circumscribed powers. The
Mugabe, like all individuals in his kingdom, with the exception,
perhaps, of slaves, was bound by custom. It was his duty to
defend the cattle and lives of his subjects, to perform certain
magical and religious rites, to offer economic help to people in

AFRICAN POLITICAL SYSTEMS

138

distress and to pay, like any one else, a bride-price to the father of
any woman he wished to marry. He acknowledged the rights of

clans to certain offices and took the advice of his supporters in
political affairs. In judicial matters, his mother and sister could
veto his decisions. No man, it is said, could be executed by the
Mugabe until the consent of these two women had first been
obtained.

We must be careful to note, on the other hand, that while the
Mugabe was bound by custom, he was, strictly speaking, above the
law?. No man could take legal action against him, for there was no
authority higher than the Mugabe before which he could be
accused, tried, and sentenced. Political action could be and was
taken against him. The legal relationship which bound subjects
to the Mugabe could be broken by the subjects moving to another
kingdom or by refusing to pay homage until the king fulfilled his
obligations.

While the Mugabe was the unquestioned head of the State, he

did not stand alone. He was supported in his kingly duties by a
large number of individuals, who, together-with the king, formed
what might be called an effective government. Among these indi¬
viduals, the king’s mother and sister were the most important.
They lived in separate kraals and maintained establishments
almost as elaborate as that of the Mugabe. Next in rank came the
Enganzi, or favourite chief, who lived with the Mugabe and acted
as his adviser. Then there was a large group of individuals known
as the abagaragwa , or king’s relatives, who had a variety of duties
to perform in the king’s kraal Finally, there were the executive
chiefs, or abakungu , comprising war leaders and tribute collectors.

We shall for the moment postpone the discussion of the mother
and sister of the Mugabe and deal with the dignitaries who derive
their positions through royal selection. The Enganzi has been
variously called the ‘prime minister’, the ‘head chief’, the ‘beloved
one’ and the ‘favoured one’, but we shall here call him the ‘favourite
chief. When during the new moon the Bahima see the new moon
and the evening star together in the western sky they say that the
Mugabe and the Enganzi are in conference, the moon representing
the Mugabe and the evening star the Enganzi. When relations
between the Mugabe and the Enganzi are strained, the people are
afraid, for they say ‘power and wisdom’ are quarrelling. The
Enganzi is selected by the Mugabe with the advice and consent of

THE KINGDOM OF ANKOLE IN UGANDA 139

his mother and sister. The first act of the Enganzi after the acces¬
sion war is the establishment of the new Mugabe. In this sense he
is a king-maker. Although the Enganzi was a rich and powerful
man, he was always selected from a clan other than the Abahinda
and, therefore, could not lay claim to the Mugabeship himself.
The Enganzi was the chief military adviser and with his advice
every new Mugabe selected the leader of his military bands.
During war the strategical movements of these bands were
decided by the Enganzi. After a successful cattle-raid, the
Enganzi was responsible for the first distribution of cattle. After
the death of the Mugabe, the Enganzi would support the favourite
son of the Mugabe in the struggle for the Mugabeship. In this
struggle, his power would often turn the scales against the other
sons. The Enganzi then often formed a link between the two reigns
and was thus instrumental in checking the worst excesses of the
accession war.

In the Mugabe’s kraal there was a group of young men collec¬
tively known as the abagaragwa, or king’s relatives. These men
were selected from among the sons of the prominent men in the
kingdom and followed the Mugabe in all his movements from one
part of the country to another. It was from among these young
men that the future Enganzi * and the future abakunga were
selected. The younger men were known as abashongore, or singers.
They sang praise songs to the Mugabe, amused him by wrestling,
and accompanied him when he went hunting. Men older than
these youths were known as the abakazhwarangwe , or warriors, who
accompanied the Mugabe on cattle-raids, acting as his body-guard
and as messengers. Older men who had not received official posi¬
tions from the king were known as the emtkyeka, or councillors.
They attended the meetings of the Mugabe and the Enganzi. They
were at once respected and feared by the executive chiefs, respected
because they had great influence with the Mugabe and feared
because any failure was at once reported to the Mugabe by them.
The carrying out of the Mugabe’s orders was in the hands of a
number of chiefs known as the abakungu, or prominent men. The
majority of these abakungu were abatware y . leaders of military
bands. It was their duty to guard the borders of Ankole against
raiders; they were almost constantly away from the Mugabe’s
kraal. Each omutware gathered a band of warriors around him
who lived with their cattle near his quarters. The abatware were

14© AFRICAN POLITICAL SYSTEMS

men of power and importance and would sometimes rebel against
the king. Another class of abakungu consisted of entuma , or tribute
collectors. Whenever the Mugabe needed extra cattle, it was the
duty of the entuma to go to every cattle-owner, select a number of
cattle and take them to the Mugabe’s kraal. The Bairn do not make
distinctions between the abatware and the entuma , but call them
all abakungu , the reason being that all the Mugabe’s officers were
just tax collectors to the peasants.

Another important class of individuals involved in the manage¬
ment of the State was the king’s relatives known as the abanyi-
gyinye. As will become clear later, these relatives were almost
always relatives of his mother’s side. The king’s mother’s brothers,
who helped the king to secure the Mugabeship, were later given
positions as military leaders, and the sons of these men became
members of the abagaragwa. The king’s barimi (mother’s brothers)
had the same status in the State as the king’s mother and sister.
They were permitted to collect tribute from the Baira and to
demand cattle from the Bahima without the king’s permission.
The members of the king’s ekyika (sub-clan) were given special
status, if they had supported the Mugabe in the accession war;
otherwise they were treated as ordinary Bahima. One of our
nearest neighbours was the son of the present Mugabe’s brother,
who claims that he was too young to be involved in the accession
war. To-day he is just an ordinary herdsman in possession' of a
small herd and in no way distinguishable from the average
Muhima kraalsman. He stated that he had no right to demand
chieftainship or other offices and privileges, as his relatives had
not supported the Mugabe at the time of his accession. On the
other hand, the Mugabe supported the wives of his father’s
brothers and those of his own brothers after these men had been
either killed or driven into exile.

Besides these individuals who were directly concerned with the
management of the State, the Mugabe had a large following of
wives, guards, magicians, and servants, who formed the permanent
membership of his kraal. This kraal or residence was known as
the omrembo and was made up of a number of enclosures. Like
other Bahima, the Mugabe moved about the country. His move¬
ments were partly determined by the needs of his herd and partly
by magical considerations. If he were in poor health, the diviners
might decide that he must go to one or other of the sacred places in

THE KINGDOM OF ANKOLE IN UGANDA

141

Ankole to offer to Ms emandwa spirits, in which case the whole
orurembo would move. Moreover, as the Banyankole were almost
constantly on a war footing, the orurembo had to be ever ready to
move to places of shelter.

The orurembo consisted of the ekyikari, Mugabe’s private
enclosure, and the amachumbi kraals, in which lived the abagaragwa
or retainers, his privateunilitary band, and the abahuku, his Bairn
servants. Of these kraals, the ekyikari was the largest and formed
the centre of the orurembo. It was made of the same materials and
in the same way as the ordinary BaMma kraals, the only difference
being that the walls of the enclosure were Mgher and the huts were
larger. The main entrance to this enclosure was called the mugaba
and was guarded night and day by the abarizi, gate-keepers. In¬
side the kraal and to left of the gateway there was the ekyikomi,
main fireplace. All visitors, messengers, and litigants had to
remain here until their wishes were heard by the Enganzi. It was
at the ekyikomi that the Mugabe received Ms men, tried cases, and
held meetings of lesser importance. The ekyikomi was, therefore,
the public part of the ekyikari and took up about one-fourth of the
kraal space. The rest of the ekyikari was separated into five dis¬
tinct enclosures. The most important of these enclosures was the
nyarubuga, wMch housed the Mugabe’s women. The nyarubuga,
in turn, was divided into five lesser enclosures. Within the
ekyiniga were the huts of the Mugabe’s favourite wives. These
women were known as the enkundwakazi and were waited upon by
immature girls and guarded by the ebishaku, castrated Bairn
servants. In another of these lesser enclosures within the
nyaruhuga was the rwemhunda , in which the Mugabe kept the
immature girls, enshorekye, who were later to become Ms concu¬
bines or wives. The Mugabe had the right to take any girl in his
kingdom if he wished. It was one of the duties of Ms retainers to
inform the Mugabe of pretty girls in Ms kingdom and to bring
them before Mm. If the Mugabe was pleased with the appearance
of a girl and was assured that she was a virgin, he would include
her with the enshorekye. Parents whose daughter was taken by the
Mugabe in this manner did not deem it an outrage; on the con¬
trary, they looked upon it as an honour. Many Bahima and even
Bairu would offer their daughters to the Mugabe. Girls accepted
or taken by the king were not always an economic loss to their
parents, for if the Mugabe decided to make a girl Ms wife he would

142 AFRICAN POLITICAL SYSTEMS

pay the girl’s father the enzhugano , or bride-price, and, of course,
there was always the. chance that she would become the future
Mugabe’s mother. These girls were taken care of by the Mugabe’s
brothers 3 widows, who taught them to dance, sing, and play the
harp. It was the custom of the Mugabe to spend his evenings in
the rwemhunda % where the enshorekye entertained him.

When a girl became mature and pleased him, the Mugabe
would select her as his next concubine. She was first, as was
customary among the Bahama, put into a separate enclosure to be
fattened. This enclosure, which formed part of the nyaruhuga>
was known as the wayetwoka. Here one of the older women forced
the girl to drink large quantities of milk. When she was so fat that
she walked with difficulty she was considered fit to sleep with the
Mugabe. She then became an ekyinyashunzhu and was quartered
with the rest of the ekyinyashunzhu in the kagyerekamwe , the enclo¬
sure for the king’s concubines. The ekyinyashunzhu were most
closely guarded by Bairn eunuchs and were waited upon by the
enshorekye girls. Any man caught in the quarters of the king’s
concubines was put to death instantly. From among the ekyinya¬
shunzhu concubines, the Mugabe selected his wives. Any of the
girls whom he did not wish to marry he gave as gifts to his friends
and retainers. Older wives who were bringing up children lived
in another enclosure which had no special name and which was not
very closely guarded. The Mugabe did not neglect these women,
however, for they had already produced children and one of them
was destined to become the nyamasore, mother of the future
Mugabe; they were, therefore, already respected by the people.
The Mugabe was anxious that his sons should grow up to be
strong and capable men and took an active part in their training.
No matter how intimate the Mugabe might have become with his
wives and children, he never ate with them. His cooked food was
prepared for him by a Mwiru and served to him by one of the
enshorekye girls.

The next place of importance in the ekyikari was the large
meeting hut, nyarunzhu rweterekyero. It was in this hut that all
important meetings took place and before which the Mugabe
entertained his special guests. Near the meeting hut there was a
large beer-store and a number of smaller huts for visitors. When
a large cattle-raid had been planned, the men who were going to
take part in it gathered before the nyarunzhu rweterekyero and

143

THE KINGDOM OF ANKOLE IN UGANDA

swore before the Mugabe to come back with cattle or to die in the
attempt. It was before this hut that every new Mugabe was
invested with office, before which cases of murder and treason
were tried and punished, and where the peace ceremony was
performed. All important meetings were accompanied by beer¬
drinking, the beer being served in individual calabashes by Bairn
servants. The ekyikomi (great fireplace), we saw, was the common
meeting place where minor cases were tried, where entertainment
took place, and where every commoner had the opportunity to do
homage to the Mugabe. The nyarunzhu rweterekyero , on the other
hand, was the official centre of the Banyankole State, where only
the leading men met to discuss and transact State business.

The Mugabe w T as never completely free from danger. Not only
foreign enemies, but rebellious subjects threatened his position.
Chiefs who had fared badly in a distribution of captured cattle or
who had had their possessions and positions taken from them were
ever ready to revenge themselves upon the king. In the accession
war, it sometimes happened that one of the Mugabe's brothers
would flee to another kingdom and later endeavour to return and
slay the king. Against these external and internal enemies, the
Mugabe maintained a strong guard, permanently quartered in the
orwekubwo ;~this enclosure w T as built next to the women’s quarters,
the nyarubugq, and was the enclosure into which the Mugabe
retreated when the alarm was sounded by the gate-keepers. In
the orwekubwo there was a special hut for the spears which were
made by the Mugabe’s blacksmiths. As a rule, the command of
the king’s private guard was in the hands of the king’s mother’s
brother, who owed his high rank to the king and was, therefore,
believed to be loyal and trustworthy.

As we shall see later, religion played an important part in the
Mugabe’s life. .Offerings had to be made to his ancestors and to his
emandwa , not only for his bodily welfare, but also for the success
3f his enterprises and for the health of his cattle. A special enclo¬
sure, the kagondoy was set aside for this purpose. In this enclosure
there were the endaro , spirit huts for the ancestral and emandwa
spirits. These endaro were so large that the spirit wives of the
Mugabe were able to live permanently in them. The emandwa
huts were the same in form as those used by the commoners, i.e.
they consisted of a sheaf of grass tied near one end and set up to
form a conical hut into which a pot of milk or beer could be

144 AFRICAN POLITICAL SYSTEMS

deposited, in the kagondo the ohibandwa ritual for the emandwa
spirits was performed and here cattle were killed that were to be

used as offerings and to act as the leader in the okubandwa ritual.
Also that part of the moon ritual which entailed okubandwa was
performed in the kagondo.

The milk and the meat which the Mugabe personally consumed

was obtained from his own herd. For this purpose, a considerable
number of cattle were kept in the ekyikari in an enclosure called the
eka y'enkorogyi. The word enkorogyi means the herd which
remains with the owner and differentiates it from the enshubt, or
herds which are dispersed throughout the land. These terms are
used generally by all Bahima when speaking of their herds. The
Mugabe’s herd, like the herd of every Muhima, was made up of
cattle, some of which were set aside for the ancestral and emandwa
spirits and others which served purely economic purposes. The
Mugabe’s herds were noticeable for the fact that they contained
many black and white cattle. As cattle of these colours were used
for special ritual purposes, any Muhima bringing the Mugabe a
black or white cow would be well received and rewarded with
cattle of other colours. This special herd was kept in the ekyikari,
the Mugabe’s private enclosure, and was clearly separated from the
large herd belonging to the Mugabe, which was kept in one of the
numerous enclosures surrounding the royal kraal. The cattle of
this large herd were used to support the Mugabe’s retainers and
were given away as gifts to visiting Bahima. It was constantly
being replenished by cattle which were confiscated from rebellious
subjects, came in as fines, payment for trying legal cases, or in the
form of okutoizha (homage payments).

So far we have concerned ourselves with the internal form of the
ekyikari (royal enclosure). Within it we found the nyarubuga,
with all the various enclosures for the Mugabe’s women, the
nyanmzhu rwetirekyere (meeting hut, the guard quarters), orwekubo
(the ritual enclosure), kagondo, and the eka y y enkorogyi (cattle
enclosure). Just inside of the gateway, mugaba, there was the
ekyikmd, or great fireplace, where the Mugabe’s subjects gathered
to ask favours and to pay him homage. The ekyikari, then, was the
centre of the orurembo or royal place; around it were scattered the
subsidiary kraals called the amachumhi.

In one of these amachumhi kraals lived the king’s private warrior
band consisting of several hundred men, commanded by a

THE KINGDOM OF ANKOLE IN UGANDA 145

favourite retainer. These men had sworn to defend the Mugabe
until death and were picked from among other warrior bands for
their courage. This band was constantly near the ehyikari and its
members had their wives and such cattle as they needed with
them. They remained until age made them unfit for military duty.
The Mugabe’s private band was used only as a last line of defence
when an enemy force invaded Ankole. When danger threatened
they would move the Mugafre’s cattle and people to a safe part of
the country, scattering his cattle in small herds and taking special
care to hide the royal drums and beaded veil, Rutare.

The king’s magicians had a kraal to themselves. Most of these
magicians (< abafumu ), were Bairn and were forced to serve the
Mugabe during his lifetime. Any omufumu who had won fame
might be called upon by the Mugabe to serve him. Not only were
all departments represented, as divining, sorcery, white magic,
and the smelling out of bad medicine, but each department had
its own specialists. There were diviners who foretold the future,
using the entondo, a small insect, others who read the signs in the
entrails of a white cow or sheep, others, again, who divined with
cowrie shells. There were sorcerers who practised with their horns
filled with secret medicines; others who used the bow. There were
practitioners in white magic who were experts in purifying, in
casting spells against evil influences, or in ma k ing charms for use
against disease and bad luck. Of particular importance was the
omutsiriM (cattle magician). The Mugabe himself did not possess
magical paraphernalia. Each magician procured his own medi¬
cines. In divining, however, it was sometimes necessary for the
Mugabe to be present. While some of the Mugabe’s magicians
were busy from morning till night protecting the king’s person
from harm, it was during war-time that the majority were most
busy.

IV- Tribute

The labour required for the upkeep of the royal establishment
was considerable. Menial tasks such as wood-cutting, water¬
carrying, and butchering were performed by the abakuku. These
men were slaves and had their ears cut off to prevent them from
permanently escaping. They were said to be peasants who were
taken for this purpose from the neighbouring kingdoms. They
lived near the royal enclosure and worked under the supervision

i 4 6 AFRICAN POLITICAL SYSTEMS

of a Mwiru headman. Besides these menials, the Mugabe, nad

his expert craftsmen. The foremost of these were the king s
blacksmiths, abahesi^ who made spears, knives, axes, and ankle
and arm bands out of iron. Important also were the carvers, who
made milk-pots, drums, wooden spoons, and carved decorations
out of wood, ivory, and bone. Then there were the skin-dressers,
bark cloth-makers, sandal-makers, beer-brewers, and sanitary
attendants. Some of these crafts were the special prerogative of
certain clans. The king’s sandals were made from the skin of
a duiker by a man selected from the abaigara clan, who also grew
and prepared the king’s tobacco. A man of the ahasingo clan
had the duty of washing the Mugabe every morning, after which
a man from the abararira clan gave the Mugabe a magical potion
to drink. The Mugabe’s musicians were men from the Koki
district who had learned to play the Baganda flute. His hunters
came from Buwhezhu and Bunyarugura with their nets and dogs.
The labour required by the Mugabe thus fell into two classes:
slave labour and oruharo> or forced labour. The slaves were the
property of the Mugabe and no payment was- made to them. The
craftsmen, magicians, and servants whom the Mugabe called to
Ms service were rewarded by a form of payment known as the
engabirano. This payment, however, was not made until a servant
was given permission to leave. This permission was given on
account of old age or if a servant left a son or some other trained
person in Ms place. The engabirano consisted , of barren cows,
bullocks, sheep, and goats when the servants were Bairn and cows
when the servants were Bahima. Oruharo was also used by the
BaMma chiefs and wealthy cattle owners, but only with the
sanction of the Mugabe.

Besides labour, the Mugabe required large quantities of food
and beer, not only for the upkeep of the royal kraal, but also for
feasting Ms chiefs and visitors and to help such followers and
subjects who were in need. The essential foods, such as milk,
meat, and blood, came from the private herds of the Mugabe and
his principal chiefs. But other foods like millet and beer came
from the Baira peasants in the form of tribute. The duty of tribute
collection for the royal kraal was placed upon the Enganzi (favourite
cMef), who appointed Bairn collectors, who, in turn, were respon¬
sible for the actual collection. These subsidiary collectors were
called abakungu . Each omukungu appointed local collectors, who

THE KINGDOM OF A.NKOLE IN UGANDA 147

brought the necessary beer and millet to certain local centres
ready for transportation to the King’s kraal.

As tribute collection was exercised by the chiefs, there was, of
necessity, a division of the country into areas. Every Muhima
chief had, while in a given locality, the right to collect tribute,
but part of his collection must be sent to the Mugabe. Besides
the tribute sent in by the chiefs, the Mugabe levied tribute directly
from the peasants in the Shema district. In this locality the
Mugabe had two Bairn tribute collectors who collected by the
moon. When the moon was on the increase, one man collected
the tribute which was called orubabo . When the moon was waning
another man collected it and this was called ekyirabamu . The
quantity collected depended upon the needs of the Mugabe’s
establishment, the collectors being informed of the amounts
necessary. Before and immediately after a cattle-raid, when
feasting took place at the king’s kraal, more tribute was necessary
than during normal times. Failure on the part of the peasants to
provide the necessary amount was followed by destruction of
property and by beating. Persistent neglect of the tribute obliga¬
tion often resulted in the execution of the rebellious peasant.

There was considerable variation in the quality of millet in
Ankole and in the knowledge of beer-making. Whenever the
Mugabe found a brew that was to his liking, he selected its makers
as his private brewers. Such peasants had to take special pains
over the Mugabe’s beer and were forced to take it in person to the
king. These private brewers often became favourites and were
eventually rewarded with an engabirano payment.

It is difficult to-day to assess the amount of tribute gathered,
the hardships which it brought to the Bairn, and the reaction of
the Bairu to the tribute burden. The peasants are unanimous in
stating that this burden was heavy, but ‘it was better to pay the
tribute than die’. The old men complain most about the collectors,
who, they claim, exacted more than the Mugabe demanded,
keeping the surplus for themselves. When the collectors became
too bold, the peasants would complain to the Mugabe, who would
then appoint new collectors. It is said that both the peasants and
collectors practised sorcery upon one another and that a particu¬
larly evil collector would be speared to death. The Bairu, then,
were more concerned with the abuses of tribute collection than
with the existence of tribute itself. The payment of tribute, like

i 4 8 AFRICAN POLITICAL SYSTEMS

the payment of poll-tax to-day, they accepted as part of the order
of things. It was an admitted burden, but it had to be made in
order that life could be carried on.

Okutoizha , or the payment of homage by the client to the
Mugabe, was a source of considerable income. Here, again, the
exact amount is beyond investigation. Every Muhima, upon
becoming an omutoizha , or client, presented the Mugabe with
from one to three head of stock, depending upon the size of his
herd. Poor herdsmen who could not afford to give away cattle
brought milk, butter, or calf-skins. The payment of okutoizha
was made periodically and as long as a Muhima wished to be the.
Mugabe’s client. Okutoizha differed from tribute in that it was
freely given by the client, who believed that the protection received
warranted the payment.

Although okutoizha was essentially a political instrument, a
means for setting up the Mugabe-client relationship, we are here
concerned with it as an economic measure, a specific institution
for the maintenance of the State structure. As the cattle came to
the Mugabe’s kraal and were presented to him, they became his
personal property; he knew the names and appearance of these
cattle and knew also the increase which they constituted to his
herd. The Mugabe, himself, however, did not use these cattle
for his own food, but sent them to swell his herds distributed
throughout the country of Ankole. For the purpose of keeping
a tally upon his cattle, the Mugabe had special men called entuma,
who knew exactly where every cow was stationed and from whom
it had been received.

From the purely economic standpoint, cattle received through
okutoizha formed a savings fund, a surplus upon which herdsmen
in distress could draw. Any of the Mugabe’s clients, when in
need of cattle, could come to the Mugabe and explain his plight.
After carefully hearing the matter, the Mugabe would present the
man with a number of cattle in order that he could establish a new
herd. The number of the cattle which the Mugabe would give
to a client depended upon the man’s former wealth and his rela¬
tionship to the Mugabe. If the man had performed many services
for the king, he would be given more help than if he were unknown.
This differential treatment among the Mugabe’s favourites was
a source of ill will among the Bahima and often led to open rebel¬
lion on the part of dissatisfied herdsmen. It was the particular

THE KINGDOM OF ANKOLE IN UGANDA 149

duty of the Enganzi, or favourite chief, to see that equal treatment
was extended to all followers of his master.

The surplus fund of okutoizha cattle was also used by the
Mugabe for making engoMrano payments to servants, magicians,
and other followers. Important chiefs like the war leaders
(abatware), were given extensive herds by the Mugabe on their
retirement. Exceptionally successful cattle-raiders were given
great numbers of cattle as a reward for increasing the king’s herds.
The numerous marriages of the Mugabe demanded many cattle
for the marriage-prices. Large feasts, before and after cattle raids,
were supplied with meat from the Mugabe’s herds. Finally, as
hostile as the Mugabe’s relations were with the neighbouring
kings, there were times of peace in which the kings exchanged
gifts of cattle, during which time they aided one another against
other kings or rebellious subjects. The Bahama have a saying,
‘Darkness makes the mountains touch’, meaning that, unknown
to the commoners, the kings have dealings with one another in
which cattle pass from one monarch to another. Okutoizha cattle
were not used for ritual purposes by the Mugabe. All cattle which,
he set aside for the spirits of his ancestors or to those of his emandwa
or which he permitted to be used in divination came from his
private herd, the enkorogyi.

Another form of economic income to the Mugabe, which was
not, however, very extensive, was the payment of okutoizha by
the Bairn. With the political aspects of this form of gift we shall
deal later. Whenever a Mwiru visited the king’s kraal, he would
bring with him a goat or a sheep, millet, beer, maize, beans, &c.,
as presents. These articles the Mugabe used for making payments,
especially to his Bairn diviners and sorcerers, and for feeding his
large following of Bairu workmen and slaves. Any Mwiru, more¬
over, who had consistently visited the royal kraal and made pay¬
ments of this kind to the Mugabe could claim his assistance if
he found himself in economic distress.

We come finally to a form of income known as ehyitoro, As the
name indicates, ehyitoro cattle were derived from the Abatoro,
conquered herdsmen. The king’s entwna , cattle collectors, went
periodically among the herds of the Abatoro taking as many cattle
as the king required. Very little attention was given to the needs of
conquered herdsmen and very often a man’s entire herd would be
taken from him. The Bahama look upon the payment of ehyitoro

AFRICAN POLITICAL SYSTEMS

150

as a terrible event and say that 'the entuma are like lions which
attack at night when the men are drunk’. Ekyitoro was open to the
worst phases of abuse in tribute collection. The entuma took what
they wanted, keeping many cattle for themselves and if a cattle-
owner threatened the collectors with exposure to the Mugabe he
w r as simply speared to death. The Bahima also apply the word
ekyitoro to a form of compulsory tax levied upon their cattle by the
Mugabe. If through disease or raid the Mugabe had lost many of
his cattle, he claimed the right, as supreme protector of all the
cattle herds of Ankole, to send out his entuma to bring in as many
cattle as were needed in the royal kraal. I have never heard the
Bahima object to this levy. They claim that this right was seldom
exercised by the Mugabe and was always practised with due con¬
sideration to the needs of the herdsmen. Ekyitoro was a royal
privilege and was extended to the Mugabe’s mother and sister and
the mother’s brothers. The greatest honour which the Mugabe
could confer upon a chief was the right of ekyitoro. Very few men
received this privilege for life but many able warriors were given
the right temporarily. While a man had the right of ekyitoro he
could take what cattle he wished within the kingdom, excepting
only those of the king. Along with this privilege went the right to
kill any one who resisted the confiscation of his cattle. The
Bahima claim that any man who had been given this right used it
to damage his enemies by taking their cattle and by killing any
people who had formerly harmed him.

V. The Cult of Bagyendanwa

A visitor to the royal enclosure on Kamukuzi Hill, near Mbarara,
to-day would be shown an old ramshackle, mud-walled, grass-
roofed hut, the shrine of Bagyendanwa. If he were to enter into
the dim, smoke-grimed interior of this shrine, he would see on a
raised platform or altar a number of drums surrounded by milk-
pots and partly covered with bark cloth robes. Before the drums
he would see a number of bleary-eyed natives squatting beside
a fire which, he would be told, is never permitted to go out except
upon the death of a Mugabe. A European acquainted with the
Banyankole would tell him that these drums are the royal drums of
Ankole and would add that no white man has been able to solve
their mystery. He would gain little, if any, insight into the true
meaning of the drums to the Banyankole, the tremendous magical

THE KINGDOM OF ANKOLE IN UGANDA 151

power which the natives attribute to Bagyendanwa and the part
which it, along with other objects, plays in the life of the people.

Bagyendanwa is the tribal charm or fetish of the Banyankole.
In the past, it is said, that at the accession ceremonies human
sacrifices were made to it. 4 So long as Bagyendanwa remains in
Airhole, the people say, 'so long will the country and the people
prosper. The Banyankole do not think of Bagyendanwa as a
symbol of abstract unity, but as a concrete power capable of
helping men in need. £ Bagyendanwa is like the Mugabe, only
greater, Ankole is the land of Bagyendanwa and we are the people
of Bagyendanwa. The Mugabe is his servant’, is the way in which
a Munyankole describes the power of the drum over the king and
the people.

It is difficult to understand the beliefs which the Banyankole
nold about Bagyendanwa. They will deny that the dram has a soul
like human beings, but will say that it can see and hear and that it
knows what is going on in Ankole. The notions held about
Bagyendanwa are akin to the beliefs which they hold about the
magic horns of the magicians. Like these medicine-filled horns,
Bagyendanwa has the power to perform acts, but, unlike these
horns, the power in the dram is inherent and not due to the appli¬
cation of medicines. The Banyankole have no special word for this
power, but describe it as a capacity to perform certain acts. This
power or capacity, although inherent, can be reduced by the evil
influences of men, things, and events, and the dram has, therefore,
to be periodically purified and protected. Furthermore, the dram
requires cattle, milk, meat, millet, and beer for its welfare.
Although these offerings are given to the drum as offerings by Indi¬
viduals who require its help, the Banyankole believe that the dram
must have food to remain strong. Bagyendanwa must be kept
warm, so it is usually covered with a bark cloth and the fire is said
to add to its comfort. Bagyendanwa is considered a male, and a
female drum has been selected for him which is always kept by
his side. Attendants must not speak loudly in the presence of the
drum, as he is believed to punish such levity.

The Mugabe is a Muhi'ma and has the interests of the Bahima
at heart; the Bairn are his serfs. Bagyendanwa is impartial. He
is as much interested in the Baira as in the Bahima. The con¬
quered herdsmen, Abator 0, also had the right to offer to Bagyen¬
danwa and used this practice as a way of getting into the good

152

AFRICAN POLITICAL SYSTEMS

graces of the Mugabe as a preliminary step towards clientship.
Slaves who had no legal status were barred from worshipping the
royal dram. While Bagyendanwa showered his blessings upon the
Bahima and the Bairn alike, he still distinguished between them.
The Banyankole say that, 4 the Bahima are the cattle of Bagyen¬
danwa and the Bairu are his goats'. ‘The Bahima must offer cattle
and the produce of their cattle and the Bairu must offer the produce
of their gardens.' Thus, while Bagyendanwa was the tribal charm,
it would be untrue to say that he considered his ‘children 5 of equal
status.

The power of the dram is apparent in the activities which it
performed. When a chief decided to ask a favour, or to ask for
advancement from the Mugabe, he would first go to the shrine of
Bagyendanwa and offer a cow. He would take the beast in person
before the dram and say, ‘I have brought a .cow; one of the
Abachwezi, they who have gone before, may you take this cow, this
red one of mine, one that I have herded, a' clean one in the orurembo
[kraal], so that the king will not refuse me, so that the king will not
walk towards his nyarubuga [private quarters] 5 .

Once an offering had been made, a man felt encouraged to make
his request. This does not mean, of course, that no other magic
was. resorted to, but that the offering to Bagyendanwa was an
essential element in uncertain enterprises. If the request was
granted, the chief would take another cow to Bagyendanwa as a
thank-offering. ‘I have brought you this one, my king, for you •
have heard me. The great ones have heard me; they shall have
what I have . 5

Similarly, any man undertaking, a*cattle! raid, in the past, would
always offer to Bagyendanwa, asking the dram to protect him from
the spears of Ms enemies. ‘We are making a raid for you. We are
going to increase your herds. We are going to make your land
strong 5 , they would say. Not only in cattle raids would the
Bahima ask for the help of the drum, but also if they were moving
into another part of the country, digging a new water hole, or
launching any enterprise in wMch there was great danger. The
Baku would also ask the dram for success when they moved to new
parts, when going on a hunting trip, or beg for help when thek
crops failed or thek children died. In the case of the Bairu, beer
and millet would be offered, and if they were successful a second
offering would be made to thank the drum for its solicitude.

THE KINGDOM OF ANKOLE IN UGANDA 153

Not only did Bagyendanwa help people in their endeavours, but
he was also believed to punish evil-doers and to avenge wrongs.
If a man felt that he had been wronged by some one, but could not
prove his case before the Mugabe, he would go to the dram and
beseech it to punish his enemy. The common occasions for thus
appealing to the dram were theft, adultery, sorcery, and slander.
The Baira, it is claimed, sought justice more often from the drum
than did the Bahima, for the Mugabe was 4 often deaf to the
complaints of his serfs’. Bagyendanwa punished people by making
them ill, letting their cattle die and by causing wild animals to
destroy their cattle and crops. If through divination a man found
that the drum was punishing him because he had wronged some
one, he would go to the person whom he had wronged and com¬
pensate him for the loss or damage he had incurred. Sometimes the
two men who had come to terms thus would go to the shrine of
Bagyendanwa and offer to him and swear by the dram not to harm
each other again. Such men would continue to offer to the drum
for some time afterwards, for, they said, ‘he had brought peace
where there had been hate’. For all requests and answers offerings
had to be made.

Even though nothing had gone wrong, the people would some¬
times take offerings to the dram in order to solicit protection
against the evil devices of men and spirits and the malignant forces
which every Munyankole believes to reside in the world at large
and which are revealed to him through omens and signs. The
wealthier a man is, the greater is the danger around him and the
greater and more frequent must his offerings be to the drum.
Wealthy chiefs who were envied by rivals were particularly careful
to make large offerings of cattle in order that evil would not be
spoken about them to the Mugabe.

Bagyendanwa is also said to induce fertility in barren women.
In the past, women who had no children would take an offering to
the drum and ask it to make them fertile. The Abaruru, clansmen
who were the dram-keepers, also had the power to induce such
fertility, and upon request supplied charms made from plant
medicines which had been prepared in the shrine and which con¬
tained powers associated with the dram. Besides having the power
to induce fertility, Bagyendanwa looked with favour upon mar¬
riages and showered gifts upon important people after their
marriage feast. When the son of a chief married, he went with his

*54

AFRICAN POLITICAL SYSTEMS

bride to Bagyendanwa , where an omururu would show him the
drum-stick, omurisyo , as a symbol of fertility and would give the
groom bark cloth, milk pots, and millet. When a princess was
married, the sacred spear, nyamaringa, was shown to the newly
married couple and the groom would be given a cow called ‘cow
of the sacred spear’. Any couple who had a long and successful
marriage with many children would go to the drum and thank it
for its help and make an offering of beer and millet or a cow.
Children who had been born to a couple through the goodwill of
Bagyendanwa were called bene Bagyendanwa, or children of the
drum. They were considered more fortunate than other children
and certain to accumulate large herds and to be successful
raiders.

Bagyendanwa, like the Mugabe, provided a certain amount of
economic help to people in dire distress. Offerings of cattle and
food accumulated at the shrine of the drum. Some of the food
was consumed by the Abaruru drum-keepers and the slaves who
fetched wood and water, but much of it found its way back to the
people of Ankole. Cows were milked, bull calves were slaughtered,
and the beer and millet accumulated in greater quantities than
were needed to supply these attendants. At marriages food and
cattle were given away, as we have seen. But more important
than these gifts of the drum were the cattle which were given to
Bahima who had lost their herds through raids or disease and the
food which was given to Bairu who had suffered from crop-failure.
The case of a person in distress was heard by the head drum-
keeper, who decided whether the person had a just cause or not.
It was said that no person was helped if he had rich relatives who
could help or if he were a favourite of the Mugabe. Here, again,
we see the power and importance of the Abaruru drum-keepers.
They were believed to have, not only the magical power of
Bagyendanwa, but also the capacity for justice and the discern¬
ment of human wrong-and weakness. The shrine of Bagyendanwa
provided a centre for the saving of surplus wealth and for the
redistribution of it in times of economic stress.

It has been mentioned that the cult of Bagyendanwa acted as
a unifying agent in the political organization of Ankole. How,
specifically, did the drum cult perform this function ? The
particular teleological purposes carried out by the drum do not,
in themselves, explain this integrative action. The drum, through'

155

THE KINGDOM OF ANKOLE IN UGANDA

its magical power, was believed to contribute to the welfare of
the people as a whole, to enable individuals to rise in social posi¬
tion, to add to their strength in war and to the acquisition of
material goods, to right wrongs and to punish evil-doers, to
increase the fertility of women and cattle, and to protect men
from evil powers resident in the world. But other spiritual and
magical agents were also instrumental in furthering the interests
and endeavours of men, such as the emandwa spirits, ghosts, sacred
places, charms, and magical practices of various kinds. The power
of Bagyetidanwa , then, lay not so much in what the dram did,
but rather in the fact that the dram did these things alone and
for the entire tribe.

First of all, there was only one Bagyetidanwa , while the spirit

cults, the ancestor cult, magical charms, and shrines were very
numerous and therefore differentiating influences. The beliefs
and practices associated with these agents formed associations, it
is true, but there was nothing about these groups which empha¬
sized and supported the unity which the political structure repre¬
sented. But Bagyetidanwa was common to all men in Ankole—
as common as the land of Ankole and the king of Ankole. Its
shrine was the tribal centre, where individual and tribal interests
were furthered through ritual performances, and Bagyendanwa
was the focus of all those beliefs which made for the well-being
of men. 'Bagyendanwa is ours. We are the children of Bagyen¬
danwa', the Banyankole say in expressing their common aspira¬
tions and allegiance to a unifying agent that is at once concrete
and a source of power. In the second place, Bagyendanwa belongs
to Ankole and to the Banyankole. It differentiates the kingdom
of Ankole from all other kingdoms. 'Bunyoro 5 , the people say,
‘has its Ruhuga ; Karagwe has its Nyabatama ; Ruanda has its
Karinga ; but Ankole has Bagyendanwa.' Here, again, other cults
are of little value as buttresses for political unity, for they extend
beyond the borders of politically differentiated territories. The
people of all these kingdoms had the ancestor cult, and the
emandwa cult was common to Bunyoro, Toro, Karagwe, and
Ruanda. Thus while, on the one hand, the cult of Bagyendanwa
formed a common centre for belief and practice in Ankole,
overriding sectional beliefs and rituals, it differentiated, on the
other hand, the people of Ankole from the Inhabitants of
neighbouring kingdoms.

i 5 6 AFRICAN POLITICAL SYSTEMS

To say that the cult of Bagymdanwa supported the tribal unity
is not enough in itself to show that this cult contributed to the
maintenance of a particular form of political organization. Tribal
cults emphasizing the unity of a group, by relating that group to
its mythical origins, are common enough in Africa and elsewhere.
The unity represented by Bagyendanwa was not of this general
character. The dram cult was specifically a king cult. It sanc¬
tioned the particular complexion of political relationships which
existed in Ankole by relating these relationships to their legendary
origins, namely, to the Abachwezi. Bagyendanwa was the dram
of the Abachwezi, and as such is concrete evidence that they once
lived and founded the kingdom of Ankole. It does not matter
whether the particular beliefs held about the Abachwezi are fact
or fancy. The belief that the Abachwezi established the kingdom
of Ankole a recognized number of generations back is to the
Banyankole a fact and the belief upon which their political
structure rests.

To the Banyankole, Bagyendanwa represents the Abachwezi;
the Abachwezi, in turn, sum up the beliefs and values inherent
in Ankole kingship. From what has been said about the functions
of the drum, it has become clear that the dram performed the
actions of an ideal king. Besides fulfilling the duties of leadership,
the Mugabe has magical power which protects the people from
evil. The drum has this same power to an even greater degree.
Both king and drum derive this power from the same source, the
king by being a member of the Abahinda dynasty which links
kingship by descent to the Abachwezi, the dram by being a relic
of those ancient times which represent the values embodied in
Banyankole kingship.

To the Banyankole, Bagyendanwa is greater than the person of
the king. "The Mugabe dies, but Bagyendanwa is always with
us*, they say, stressing the permanence of the drum as compared
with the temporary nature of the individual ruler. The Mugabe
is also the ‘servant of Bagyendanwa! in that he guards it and
watches over it. In the succession rights, as we shall see, it is the
drum which makes the successor a Mugabe, which gives the final
stamp and seal. The accession war is for the possession of the
royal drum, and many Banyankole claim that if a foreign king
were able to capture the royal drum he would automatically
become King of Ankole. In their tales of former wars, the

J57

THE KINGDOM OF ANKOLE IN UGANDA

Banyankole constantly stressed the importance of hiding Bagyen-
dmwa 9 so that it would not be captured. Perhaps the most con¬
clusive evidence to the statement that Bagyendanwa is greater
than the Mugabe is the power of the drum to provide sanctuary.

If, after being condemned to death by the Mugabe, a Munyaakole

were able to dash to the shrine of Bagyendanwa and to touch the
dram he would not be killed. The Mugabe would forgive him;
he would be freed and given his former rights. This sanctuary
was effective only in protecting a man from the death penalty
and only when this sentence was passed by the Mugabe. When
a father or head of a family passed such a sentence upon one of
his subordinates, the drum provided no sanctuary.

VI. Succession

The emphasis which the Bahima placed upon the health,
strength, and courage of the Mugabe was so extreme that it
affected his tenure of office and the selection of his successor.
This excessive concern about the physical virtues of the king’s
person is explained partly by his position as a permanent war
leader, and partly by the magical powers attributed to him in his
capacity as a protector of the tribe from evil influences. As has
already been mentioned, no Mugabe was permitted to die of illness
or of old a|je. As soon as his wives and followers observed signs
of weakness, the Mugabe was given a poison which brought about
his death. The Bahima compare the Mugabe to the leading bull
in the herd. They say, ‘The Mugabe is like the leading bull.
When the engundu [leading bull] is beaten by a younger bull, we
kill the engundu and let the strongest of the younger ones take
his place’.

After the king’s death a successor must be chosen. Two rules
governed this choice. First, the new Mugabe must be in the royal
line; second, he must be the strongest of the last king’s sons.
Patrilineal descent fulfilled the first requirement. The second
depended upon some method by which the strength and courage
of the Mugabe’s sons could be tested. Primogeniture and
favouritism, both important factors in the selection of a successor
in the extended family of the commoners, also played their part
in the royal family, but were overbalanced by the political and
ritual demands of kingship. The Bahima demanded that the
strongest of the king’s sons should be their leader and that the

I 5 S AFRICAN POLITICAL SYSTEMS

test should be one of war. The brothers must fight among them¬
selves until one of them alone remained alive in Ankole to claim
the drum and the Mugabeship.

In the king’s kraal the sons prepared to fight for the Mugabeship
and to find Bagyendanwa . In the meantime Ankole could not be
left without a king. After the mourning ceremonies, a mock battle
took place in the royal kraal between common herdsmen, and the
winner was chosen as mock king. He maintained a semblance of
order in the royal kraal until the accession war ended. This mock
king was called ekyibumbe. The word ekyibumbe has a variety of
meanings. In common usage, it describes a stupid, foolish person.
Sometimes it is used to indicate a person who is the butt of jokes
and tricks. A small, toothless baby, who must be taken away from
the kraal upon the death of its father, is also called an ekyibumbe.
The royal brothers watched this mock battle, but after the person
had been chosen they chose their own followers and went out to
look for Bagyendanwa . If they met on the way they fought and
each tried to kill the other. If one brother had fewer followers than
the other, he generally got killed or fled to another country. On the
other hand, strategy often made up for lack of followers. The
brothers spied upon one another in order to creep up during the
night and get the other unawares. They put poison in each other’s
food or stabbed one in his sleep. Magic and the help of foreign
allies were both resorted to. Each son was aided by his mother
and sister, who practised magic against his enemies and protected
him from the spirits of his slain enemies.

During the accession war which might last for several months,
the country was in a state of chaos. Every man resorted to his
kinsmen for protection. It is said that there was much cattle
stealing and people who had a grievance took advantage of the
chaotic condition of the country to take revenge upon their
enemies. But the great chiefs who guarded the borders of Ankole
did not take part in the accession war. They endeavoured to keep
as much internal order as possible and to guard the country from
foreign invaders.

One by one, the princes were either killed or driven into exile
until only one remained. The hidden son then came out of his
hiding place and fought with the one remaining son for the
possession of Bagyendanwa. The late Mugabe’s favourite son did
not always win, but he usually had the most powerful magicians

159

THE KINGDOM OF ANKOLE IN UGANDA

and a large following. When the accession war was finally over, the
new king went back to the royal kraa! with Bagyendanwa , his

mother and sister, and the Enganzi, killed the ekyibumbe, and was
finally proclaimed and accepted as the new Mugabe.

Several days later there was an accession ceremony, after which
the king went on a long purification journey through the land.
With him went a number of special magicians, a small herd of
cattle, and a group of expert hunters. Upon his return to the
royal enclosure, the most thoroughgoing changes took place among
the office-holders immediately surrounding the king’s person. A
new group of retainers would be selected from among the king’s
friends. These were usually men who had fought for him in the
accession war. In the selection of the most important function¬
aries, the old Enganzi acted as adviser along with the king’s
mother and sister. After the principal appointments had been
made, the old Enganzi retired and was rewarded for his long
service with many cattle. The king then selected a new Enganzi
from among his followers and the governmental machinery was
again complete. In the change from the old to the new reign, the
strongest link was the old Enganzi. In a sense, he was the king¬
maker. During the accession rites, he was the one who announced
the new king to the Bahima chiefs and who aided in the selection
of the next governmental personnel. His retirement w r as due to a
stipulation which said that ‘the Enganzi [evening star] must set
with the Mugabe [moon]’.

Succession in the Banyankole kingdom was regulated by a
particular body of beliefs and practices, the general function of
which was to maintain the continuity of kingship as an essential
part of political co-operation and to eliminate, as far as possible,
competition and discord as permanent elements of political
leadership. The dynastic principle, by restricting kingship to the
Abahinda clan, at once ruled out general competition. The
dynasty found its source in the legendary past, in the person of
Ruhinda, the descendant of the Abachwezi. Patrilineal descent
further restricted the range of candidates. The accession war,
which at first appears as chaos and anarchy, in the long run serves
the purpose of eradicating likely rivals. After the accession war,
the Mugabe stands alone in the kingly line. The accession war,
therefore, is a way of defining the succession, similar in general
function to the rule of primogeniture or the rule of the favourite son.

AFRICAN POLITICAL SYSTEMS

160

The particular features of the succession rites, like the accession
war, elaborate purification, and the,, importance of the king's
mother and sister, are understandable only in terms of the particu¬
lar nature of Bahima political structure and Bahima magic. Once
we grasp the importance, to the Bahima, of the king as a military
leader, as a symbol of unity and magical power which is amply
illustrated by the ritual surrounding his daily life, practised in
order to enhance this magical power, and the belief that a physi¬
cally weak or ailing king makes the people of Ankole weak, we can
readily understand the special stress laid upon getting the strongest
and ablest scion of the dynasty as king. Elimination through a
trial of strength certainly provided a more practical method of
choosing the best son than any specific rule of succession could
have done. As far as I was able to discover, there is no myth
sanctioning the accession war. The Abachwezi did not practise
it, nor did Ruhinda, the only survivor of the Abachwezi in Ankole,
need to establish a precedent, for he had no brothers and no
rivals to the Mugabeship. In this case, we can scarcely say that
the accession war, although formally and traditionally sanctioned,
was a periodic re-enactment of any myth. But once we recognize
the importance of the physical strength and the magical power of
the king to the Banyankole, we can comprehend the purpose of the
accession war as a means of obtaining the desired end.

As we might suspect, the accession war had far-reaching conse¬
quences on the family connexions of the Mugabe. In theory, if
not always in practice, the king had no living brothers or father's
brothers. The intimate religious, magical, and judicial duties
generally performed by the father or the eldest brother in Banyan¬
kole society were performed for the king by his mother and sister.
The king s mother and sister, in the past, had no special titles, but
were called simply nyvnya omugabe, king's mother, and omunyana
omugabe , king's sister. Their status was practically equal to that
of the Mugabe himself. They both had their private kraals with
cattle, herdsmen and warriors, and they both had the right of
levying ekyitoro on Bahima cattle. They also received a share of
all cattle taken in raids. The principal duty of the king's mother
was the making of offerings to the king's emandwa spirits and the
practising of magic against the ghosts of men whom the king had
killed. Although the king himself made offerings to his ancestors,
his mother was said to have occasionally sent a white cow to Ishanzi

THE KINGDOM OF ANKOLE IN UGANDA 161

Forest as an offering to the dead Mugabe. In the kraal of the king’s
mother there was a shrine for the four emwidwo of the king—
namely, Wamara, Mugasha, Kagoro, and Nyakiriro—where
during every new moon she made offerings of cattle and of meat.
If the diviners said that it was necessary for the king to go through
an emandwa ritual ( 'okuhandwa ), he was said to have gone to his
mother’s kraal for the rite. Besides these ritual duties, the king’s
mother had judicial and administrative functions. No man could
be executed without her consent. She sat beside the Mugabe at
all important judical cases and helped in deciding questions of
war and peace. If messengers came from foreign kings, they had
first to go to the king s mother, her consent being necessary
for an audience with her son. The function of the mother as a
protector is in these cases a better indication of her status than
any hypothetical assumption of a former matriarchate. But the
fact that the mother assumed these duties seems to be correlated
with the fact that the king had no living brothers nor father’s
brothers.

VII . Conclusion

In the brief analysis of the political organization of the Banyan-
kole given above, I have tried, not only to describe the form of the
kingdom of Ankole in its political aspect, but also to point out the
underlying forces which contributed to its formation and main¬
tenance. We can readily see that this kingdom falls into that
larger class of political structures known as conquest states,
wherein ethnically different groups come into contact, resulting in
a stratified society and a mechanism for maintenance.

The political relationships of clientship, serfdom, and slavery
may be classified on the basis of their origin, as contractual and
compulsory, differing in this from the relationships based on
kinship which formerly were predominant and still play a funda¬
mental role in Banyankole society. As to their nature or constitu¬
tion, we might say that clientship was a well-balanced relationship
arising from the need for political co-operation. Serfdom and
slavery, on the other hand, were unbalanced relationships and
exploitational in nature.

In my treatment of the Abachwezi myths, the drum cult, and
the succession rites as forms of political ideology and practice, I
have stressed the fact that even their particular form is explicable

AFRICAN POLITICAL SYSTEMS

162

in terms of the political situation and the age-old magical concep¬
tions present in the culture. In other words, the political struc¬
ture of the Banyankole is understandable only when we know the
objective situation out of which it grew and the cultural material
out of which it was created.

The imposition of British rale, of course, has brought about
great changes. Clientship, serfdom, and slavery as political rela¬
tionships have disappeared. The Mugabe, although still part of
the picture, is no longer a political leader and magical power as
of old. The effects of British rule have altered, not only the
political relationships, but also the fundamental nature of the
kinship relationships, besides introducing new relationships of a
legal nature between the Native and the white man, on the one
hand, and between the Native and the Indian, on the other. A
significant discussion of these new bonds as they touch personal
relationships, land, economic activities, and governmental
machinery requires more space than this paper will allow.

CEG0*H
LEA8A Jl

Oonko *\\

THE COUNTRY OF THE KEDE

THE KEDE: A RIVERAIN STATE IN

NORTHERN NIGERIA

By S. F. Nadel

L Introduction

T HE Kyedye or Kede, 1 with whom this paper is concerned,
are a section of the large Nupe tribe of Northern Nigeria,
whose general culture and social organization I have described
elsewhere. 2 The Kede have many cultural traits in common with
their mother tribe; their kinship system is the same; they speak
the dialect which is spoken to-day by the majority of the Nupe
sub-tribes and has become the acknowledged language of the
Emirate, Nupe "proper"; they have also adopted the religion of
Nupe kingdom, Islam. Yet combined with this cultural affinity
we find certain marked divergencies. The Kede are a riverain
group—the only purely riverain group among the various Nupe
sub-tribes. Their economic pursuits and general social life centre
round the river on which they live and from which they derive
their livelihood. This means already that their social and cultural
life must present certain features which are absent in Greater
Nupe. Their political organization, moreover, contains certain
distinctive and unusual traits—unusual even for Africa at large.
It is for this unusual nature rather than for its relation to practical
problems of African administration that I have chosen the political
organization of this small Nigerian sub-tribe for the subject of this
contribution.

II. Demography

The main body of the Kede lives to-day on the Rivers Niger and
Kaduna between 8° 30' and 9 0 40' North Latitude, inhabiting a
narrow strip of land on both banks. The Kede share their terri¬
tory with a number of other tribal sections of the Nupe, which
lead a semi-riverain life, pursuing—unlike the Kede—agriculture

1 The proper Nupe name is Kyedye. But the Hausa, Yoruba, and other
neighbouring groups (as well as, to-day, Government officials) prefer the more
easily pronounced Kede. We shall adopt, for the sake of simplicity, this latter
name.

2 See Africa, viii, 1935, and also my forthcoming book, A Black Byzantium.

166 AFRICAN POLITICAL SYSTEMS

beside fishing and canoeing. Their villages and hamlets are
scattered between the Kede settlements all along the river banks , 1
The sharp division, cultural and tribal, between the Kede and the
other semi-riverain groups is illustrated in the linguistic usage:
for Kede the word eyapaciyi (canoemen), is used almost synony¬
mously, while the other groups are referred to, collectively, as
laticiji (farmers). Tradition, too, has its contribution to make:
it represents the Kede as alien immigrants who have come from
outside into their present habitat and settled there among the
‘aboriginal’ population. This tradition (to which we shall return
later) is again reflected in linguistic usage, the different semi-
riverain groups which to-day are the neighbours of the Kede in
the river valley being spoken of, collectively, as kintsoji (owners of
the land—that is, original inhabitants).

We possess detailed population figures only for one part of
Kede country, for what is to-day the Kede District of Bida
Emirate. But we may take these figures as representative of the
whole area inhabited by the Kede. 2 In a total population of
12,066, the Kede number 2,225, an <i the kmtsoji (comprising
various sub-tribes) 9,742, the small rest (99) being made up by
non-Nupe strangers who live in Kede District. The Kede thus
form a minority in their own country—the country which bears
their name. But it rightly bears their name and rightly is called
‘their’ country, for the Kede minority represents the ruling group,
and their chief the ruler of this whole territory and the different
groups which inhabit it, Kede as well as non-Kede.

But Kede country is itself part of a larger political system, the
Nupe Emirate. In pre-British times, the country of the Kede lay
almost entirely on Nupe territory or, more correctly, on the terri-
tory ruled by the Etsu (king) of Nupe, under whom it enjoyed the
status of a semi-autonomous, vassal State. Under British adminis¬
tration, Kede country, greatly affected by the re-alignment of the

1 They comprise sections of the following Nupe sub-tribes: the Gbedegi on
the upper stretch of the Niger; the Bataci, or Marsh Dwellers, on the lower
reaches; a few groups of Beni near the confluence of Niger and Kaduna; the
Kupa round Eggan in the south; Dibo or 3 itako near Katcha and Baro; and,
finally, a group of Nupe from Gbara, the ancient capital of Nupe kingdom, on
the Kaduna and on the Niger near Patigi.

2 These figures are taken from an official, unpublished, provincial census,
for the use of which I am greatly indebted to the Administration of Niger
Province. While the figures are perhaps not correct in every detail, they are
reliable enough for the purpose of this argument.

- the KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 167

political boundaries, came to lie in three different provinces and
six different (modern) Emirates or Divisions , 1 However, we
shall see later that this distribution of a comparatively small group
over so many political divisions is not due entirely to the re-
alignment of political boundaries. It is also a result of movements
of Kede groups in recent times, after the political boundaries had
been fixed by the present Government.

By far the largest section of the Kede lives on the left bank of the
Niger and Kaduna, in Bida Emirate, In this area the Kede have
also maintained their political status of a separate political unit,
with their own chief as the administrative head. In all other areas
the Kede communities are absorbed politically in the districts on
whose territory they lie, and live under the local, non-Kede chiefs
and district heads . 2 The modem political situation has not, how¬
ever, obliterated the other features of their social life; the charac¬
teristics of their political organization, more specially, live on,
though on a smaller scale, in the one area where it has been given
official recognition. Although many of our political data will of
necessity be derived from this one area, we may again take them to
be representative of Kede country at large, and when speaking in the
following of Kede, Kede culture and social system, we shall mean
the group as a wdiole, disregarding the modem political subdivisions.

There exists, however, one subdivision of a different nature,
deep-seated and of old standing, which we may not ignore. I have
spoken so far simply of the Kede. But there exist in reality two
Kede groups: the Kede Tifin , or upper-stream Kede, and the
Kede Tako, or down-stream Kede, the boundary between the two
groups lying roughly at Jebba Island (the two groups overlap for
a short stretch north and south of Jebba). 3 Now, what I have said
about the specific features of the Kede social and political system
applies only to the down-stream group. The upper-stream Kede
show none of the traits which give to the culture of the sister group

1 The Kede on the right bank of the Niger belong now to Harm and Kabba
Provinces, and to the Emirates, or political divisions, of Ilorin, Lafiagi, Patigi,
and (in the south-west comer) Koton-Karifi. The Kede on the left bank of the
Niger and on the River Kaduna belong to Niger Province (formerly Nupe
Province) and to the Emirates of Bida, Agaie-Lapai, and Kontagora.

2 In one place (Ogudu) a certain compromise has been effected, the head of
the fairly numerous Kede community acting as a titled ‘second-in-command*
to the village chief.

3 The upper-stream Kede are also called Kede Gbede , after the Nupe sub¬
tribe (Gbede or Gbedegi) with which they share their territory.

168 AFRICAN POLITICAL SYSTEMS

its distinctive character; nor do they share the position of political
supremacy which the down-stream Kede have assumed. They are,
culturally, a semi-riverain group, like all the other Nupe sections
whose villages lie interspersed between Kede settlements, and
politically, again like these sections, subjects of the (down-stream)
Kede ruling group. Thus when we shall speak in the following of
the Kede political organization, we shall refer to this latter group
only. The upper-stream Kede will be classified under one heading
with the other semi-riverain Nupe sections, as. Indeed, in native
eyes, they are closer to the kintsoji than to the ‘alien 5 immigrant
group on the lower river. But the comparison of the two Kede
groups will become of special significance at a later stage of this
analysis. This strikingly unequal development in what appear to
be two sections of the same tribal group should help us to isolate
the factors, social or otherwise, that have moulded the political
structure which we are studying.

Ill . Economic System

A demographic and political constellation such as we have dis¬
covered in Kede country is clearly the result of considerable group
movements, possibly covering a long period. To be fully under¬
stood, such situation demands, first of all, an analysis—historical
analysis, if possible—of tribal settlement. Before discussing Kede
settlement, however, it is necessary to give a short description of
the economic situation in the country. For, as I propose to show,
the nature of Kede economics has decisively influenced the
planning of Kede settlement and, indirectly, the whole political
development of the tribe.

We can be very short as regards the economic system of the
semi-riverain groups. They are in the main farmers, who cultivate

the fertile marshland areas in the river valley. They are, Besides,
fishermen on a small scale, fishing from their small dug-out canoes
In the backwaters and creeks of Niger and Kaduna—never in the

main river, where the Kede alone are entitled to fish.

The Kede, on the other hand, are fishermen and canoemen of
renown. Their name Is known all down the River Niger, in

districts far outside Nupe Country ; 1 and in their own part of the

• r Their familiarit y wit!l the river has led to a considerable number of Kede
being employed as sailors, skippers, and pilots by the Royal Niger Company
and the Nigerian Marine Department.

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 169

world they have almost monopolized river traffic , 1 Let me say a
few words about the river trade which Kede canoes have carried
for centuries up and down the River Niger: comparatively few of

the trade goods are destined for inland trade; the majority come
from and go to places outside Nupe country. The Kede ship
south: gowns (of Hausa and Nupe make), horses from Hausa,
potash from Lake Chad, Nupe-made mats and straw 7 hats, fish and

rice from the Niger; and north: kola nuts from the markets in
Southern Nigeria, European salt, and palm oil.

In this river traffic the Kede canoeman assumes two different
roles: he is either a contractor who hires his canoe and crew
(consisting of himself and his family members) out to a trader
for a specific journey or he is both trader and contractor in one,
carrying his own goods on his own craft. A variety of the first
kind of occupation is the extensive ferry service which the Kede
have established in certain places; here the Kede canoemen carry
people, goods, and animals across the river for a small payment. 2

The work of the Kede canoemen, though lucrative and admit¬
ting of big profits, is strenuous, exacting, and not infrequently
dangerous. They must be prepared to go on long expeditions,
which still to-day may be expeditions into the unknown. 3 * * * These
river voyages mean long absence from home, often of many
months, not only because of the distance of their destination, but
also because of the many stops and long waits which they involve:
thus the canoemen may have to wait in a certain place till the river
becomes navigable again; or till they have filled their canoes with

1 The division between fishermen and canoemen among the Kede is rarely
rigid—members of a fisherman family may take to canoe-work or canoemen
spend their leisure time fishing; we will ignore this division in the following
and concern ourselves only with the canoemen, who are of most interest to us.
Originally the Kede also used to build their own canoes. The deforestation of
the river banks forced these craftsmen to move south, where to-day they form
a small colony of Kede canoe-builders near Onitsha.

2 The Very lucrative 1 ferry service at certain riverports is mentioned in Laird
and Oldfield, Narrative of an Expedition into the Interior of Africa (1837), ii,
p. The most important ferry service of this kind to-day is at Jebba Island,
where the Kede compete with the railway bridge, offering a cheaper service to
the people who want to save the bridge toll.

3 The journey from Kede country to Onitsha and back—one of their regular

tours—takes two to three months. In 1936 I saw ten Kede canoes being loaded

in Jebba with petrol for the French Air Service in Fort Nyameh. The river

journey up the Niger was new to the canoemen, and was expected to take three

months.

170

AFRICAN POLITICAL SYSTEMS

trading goods, or till they have found another 'passenger 5 (in native
phraseology) to engage them for a profitable return journey.

Here we grasp the significance of this productive system for
the organization of settlement in Kede country. The long canoe
expeditions necessitate fixed stopping and resting places, where

the canoemen must be sure to find shelter and food and oppor¬
tunity to refit their craft. The 'termini 5 of their river routes more
specially will tend to become at least seasonal 'colonies 5 of Kede.
The stopping and waiting places will naturally be chosen in
accordance with commercial considerations. And, finally, there
must be some system of political protection, which could ensure
the safety (speaking of pre-British times) of these far-flung routes,
the stopping places with their valuable stores, and the seasonal
or permanent trading posts: 1 trading posts which, of necessity,
tend to become political colonies—this gives us the formula of
Kede settlement.

IV. Settlement

We are fortunate in possessing data which allow us to trace in
detail the history of settlement and population movement in Kede
country. Our data are derived only partly from oral tradition.
The history of Kede settlement reaches into the well-documented
era of Nigerian exploration and British occupation; its last phases
are happening under our very eyes. In addition to these historical
records, we possess evidence of a different kind, which, indirectly,
contributes considerably to our understanding of population
movement in Kede country: it lies, as we shall see, in the lay-out
and organization of the present-day Kede settlements themselves.

According to Kede tradition, their tribal home was near the
confluence of Niger and Kaduna, near Muregi, which, some time
later, became their political capital. From there they are said
to have extended their settlements, and rule at the same time,
gradually over the river banks towards the north and the south

1 A case in point is the uninhabited right bank of the Niger above Jebba,
where constant raids by inland tribes (from Borgu) made settlement impossible.
The left bank, on the other hand, could be adequately protected, and here the
Kede established a settlement at Buka (it was later moved to Jebba). In the
south, Kede canoes did not travel beyond Eggan in pre-British days. The river
south of Eggan was the domain of the Kakanda, a warlike riverain tribe of non-
Nupe stock, which refused Kede canoes admittance into their area. Trade
goods for the south had to be trans-shipped at Eggan to Kakanda canoes.

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 171

till their villages covered the Niger Valley between Eggan and
Jebba. At the time when the first European travellers appeared
in Mupe country, we find the Kede already firmly entrenched in
this part of the Niger Valley. Now, of the ten main villages of

the Kede which exist to-day only one represents an independent,
purely Kede settlement—the Kede capital, Muregi. It is a well-
built village, with solid houses of sun-dried mud bricks, each
compound walled in in Nupe fashion, with a big mosque and an
imposing chief s house. It is, as I have said, a pure Kede town,
a town inhabited entirely by the ‘ruling race—^ de ialakaji a\
say the Kede (it contains no poor*—meaning persons belonging
to the subject groups). All other Kede settlements, without
exception, are built on or near the site of a village of ‘original
inhabitants . In most cases the Kede settlement occupies the
river bank itself, and the ‘native’ village the stretch of country
immediately behind; in a few cases we find the Kede settlements
on an island off the bank occupied by the ‘native’ village or on
the opposite side of the river. The result is something like a twin-
village, half ‘native’ and half Kede. The scene of Kede tradition,
a tribal home and emigrant settlements, seems indeed visible in
the present-day organization of Kede settlement, with its one
all-Kede town and its many ‘twin-villages’ along the river valley.

The more recent history of Kede settlement remains true to
this picture of a gradual territorial expansion. We know that
towards the end of last century the Kede settled for the first time
on the Kaduna River. Later, under the Royal Niger Company,
the Kede were encouraged to extend their settlements still farther
on Niger and Kaduna. It is easy to trace these new settlements,
which were founded near the European trading posts and other
places which had similarly gained commercial importance. Only
some thirty years ago the Kede founded their latest ‘colony—
on Jebba Island

The villages differ greatly in appearance: some villages boast
solidly built permanent houses, while others consist largely
of more flimsy buildings, grass-walled huts, suggestive of
temporary occupation rather than permanent settlements. The
habitations of the Kede reflect the flexible, mobile nature of their
system of settlement. The degree of permanence attempted in
the buildings at the same time betrays the age of the settlement
as well as its (past or present) importance as an economic or

i7* AFRICAN POLITICAL SYSTEMS

political centre. The Kede settlement at Raba is a typical instance
of the more solidly built settlement. It was founded about 1840,
when Raba was the capital of Nupe Emirate. The first settlers

were five men with their families from Muregi, who belonged to
one ‘house’; to-day the Kede colony at Raba (which now also
possesses a Niger Company storehouse) numbers seven ‘houses’.

The Kede settlement at Katcha is an example of the other,'
poorly built, type, which gives the impression as if the settlers
had not yet found time, or had not yet quite decided, to build
themselves permanent quarters. The present Kede settlement
dates back to 1905-10, at which time a group of Kede who had
settled previously at Eggan, on the main river, abandoned this
colony for Katcha, attracted by the opportunities of the place.
Katcha, on a tributary of the Niger, owed its rise to importance
to the introduction of steamer traffic on the Niger and the building
of the first Nigerian railway from Baro, through Katcha, to Minna.
Its five or six ‘houses’ of original Kede settlers have now increased
to seventeen.

Our list of Kede settlements would not be complete without
the mention of the purely temporary riverside camps, meant
to last for one season or a few weeks only, which the Kede
put up in the larger villages where they are wont to stop on their
journeys up and down the river. During the main trading season,
Jebba, Patigi, Wuya, or Katcha are crowded with these lightly
built shelters, grass-huts, tent-like structures of grass-matting or
—crudest of all—the wattle-and-matting awnings of the canoes
simply pulled ashore. Kede settlement has not yet come to a
standstill. Places which are gaining in importance still attract
new^groups of immigrants; and the seasonal shelters in a busy
trading centre may at any time be turned into permanent quarters,
as, indeed, it has happened repeatedly in the course of time.

Reviewing the history of Kede settlement, we find its depen¬
dence on economic factors fully confirmed. We may conceive of
it as of a progressive realization of the dictates of the productive
system of the country. The Kede, as we have seen, did not occupy
new, uninhabited country, but settled in places where an existing
population^ had already established a certain level of social and
economic life. In the choice of places for settlement and in their
subsequent development, the Kede were invariably guided by
commercial considerations. They were, moreover, not satisfied

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 173

with the position of immigrants who are dependent on the good¬
will of their hosts, but claimed with the territory in which to settle
also the political rule over it. If they are thus not ‘colonists’ in
the strict sense of the word, not caring for settlement in new,
virgin country, they are ‘colonizers’ in a more special, political
sense, being settlers, immigrants, and representatives of a ruling
race in one.

Here a final point remains to be cleared up: the exact relation
between the elastic territorial expansion of the tribe and the
necessarily more inert and rigid expansion of political domina¬
tion. From the history of modern Kede settlement we learn that
there is a certain time lag between the first occupation of a new
place by Kede immigrants (the nucleus stage of the seasonal camp)
and its eventual rise to the status of a political outpost. The Kede,
as a rule, attribute the founding of their various colonies to
particular chiefs. The initiative taken by the Kede chief in the
colonial enterprise refers both to the early growth and the final
political incorporation of the new colony. He might himself send
out settlers from Muregi to a new place which looked a likely
centre for Kede activities (Raba is an instance of this), or he might
direct settlers from other places to a promising new settlement
(as was the case in Katcha). But not until a settlement was firmly
established and numbered several families would he delegate an
official representative of his to take charge of it, thus proclaiming
the political incorporation of the new Kede dominion.

We learn, further, that Kede territorial expansion did not
proceed step by step, in continuous stages, but rather in a series
of leaps, which may leave gaps between outpost and outpost, or
mother-country and new settlement. Thus there are ‘uncolonized*
spaces on the Kaduna River between Gbara and Wuya or Gbara
and Muregi—that is, stretches of country with native villages
which (unlike the interspersed kintsoji villages in the ‘old* Kede
country) have not been absorbed politically by the Kede. We may
assume that the early growth of the Kede community followed the
same line of development. If this was so, the compact political
unit in the ‘old* Kede country proves that political rule was later
brought up to the new outpost, and the territorial gaps eventually
absorbed in the extending political unit. To-day the firmly fixed
boundaries of provinces and districts forbid, of course, a similar
sequel to the founding of new settlements. The new economic

i?4

AFRICAN POLITICAL SYSTEMS

development of the country, moreover, has attracted Kede
emigration to places far outside the orbit of their political organiza¬
tion: thus at the busy trading-place on the confluence of Niger
and Benue, Lokoja, a large and prosperous Kede colony has
grown, which combines all the 'stages' of Kede settlement, per¬
manent and semi-permanent houses in which Kede families have
made their home as well as a large encampment on the river bank
for the Kede canoemen from the north. But it seems that even
in pre-British times the territorial expansion of the Kede had to
leave certain 'gaps’ in its network of settlements and political out¬
posts. Thus the large Nupe town of Eggan on the right bank of
the Niger, flanked to the north and south by (presumably old)
Kede settlements, remained a powerful, independent political
unit, placed directly under the King of Nupe. It is this charac¬
teristic scheme of Kede expansion which allows for territorial gaps
and the founding of (at least temporarily) isolated outposts that
justifies our speaking of Kede 'colonies’ and 'colonization’.

V. Political Organization

The political system of the Kede corresponds in all important
points to the concept of the State. On its small scale, it fulfils the
essential conditions of State organization: its dominion is terri¬
torial and non-tribal (or inter-tribal); its administration is cen¬
tralized ; its machinery of government is monopolized by a special
ad hoc appointed or selected body, which is separated from the
rest of the population by certain social and economic privileges. 1
The first of these three features we have discussed already; as
regards the second, we have learned that Muregi, the traditional
home of the tribe, is at the same time the political centre of the
country; and as regards the last point, we have seen that in a
broad sense the Kede themselves represent, corporatively, the
ruling group of the country. But among the Kede we find another,
more precisely defined ‘privileged group’, in whose hands the
government of the country is concentrated. This ruling group, in
a narrow sense, consists of the Kede chief and his titled councillors
and emissaries.

1 With regard to this definition of the State, see Africa, vol. cit., and my
forthcoming Nupe book. R. Lowie, The Origin of the State (1927) (chaps, iii
and iv), recognizes territorial sovereignty and centralized authority as essential
to the structure of the State; the factor of the special ‘ruling group’ has been
elaborated by F. Oppenheimer, The State (1926).

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 175

The Position of the Chief. The Kede chief, or Kuta, resides in

Muregi. In his hands used to lie the ultimate decision on all
matters concerning country and tribe as a whole—above all, war
and the founding of new colonies. The Kuta was also the judicial
head of his country. The larger part of taxes, duties, and other
revenues used to flow into his private treasury. He was (and still
is), finally, the official representative of his country vis-a-vis the
overlords of Kede, the Emirs of Nupe. Impressive paraphernalia
and forms of ceremonial serve to display the exalted position of the
Kede chief, the most imposing perhaps being the enormous State
canoe, propelled by twelve paddlers (two to three is the normal
crew of the Kede canoes), in which the Kuta travels. The
authority of the Kede chief rests in the main on three facts:
first, a moral sanction of Kede chieftainship lies in its hereditary
nature and the fact of its being derived, in a straight line, from a
mythical first Kuta, who had been invested with the rale over the
Kede by Tsoede himself, the ancestor-king and culture-hero of
the Nupe. (We shall hear more of him later.) Another support
of Kede chieftainship, of more practical order, lies in the
overwhelmingly strong economic position of the Kede chief. His
resources allow him to acquire a large fleet of canoes—not only
the chief means of livelihood, but also the mainstay of all military
action in this riverain country 3 —and to attach to his household a
host of followers and henchmen. The position of the Kede chief
is made finally secure by the fact that the most important political
offices in the Kede State are allotted to his blood relations.

The "Offices of State'. Political offices among the Kede fall into
two categories. One comprises a small group of rank-holders,
ticiji ("titled ones’), who reside in the capital and represent the
councillors of the Kede chief. A second categoiy comprises titled
official emissaries of the Kuta, egbagi ("delegates’), who are in
charge of the various Kede settlements and colonies. To these two
groups of ‘real’ office-holders we must add a third group of what
the Nupe call "private’ or ‘household’ ranks, which the Kede chief
bestows on faithful and able followers. The majority of them live
with the Kuta in Muregi, acting as his messengers, councillors of
a lesser order, and such like; a few are entrusted with emissary-
ships.

1 Laird and Oldfield (op. dt., ii, p. 279) mention that the Kede chief ‘had
twenty canoes of retinue*.

AFRICAN POLITICAL SYSTEMS

176

The councillors of the Kuta number five ranks, three of which
correspond to more or less specialized offices: the administration
of Muregi town, the guardianship of the sacred relics of Tsoede,
and—in pre-British times—the leadership of military expeditions.
In modem Kede, the District Alkali, the Mohammedan judge
appointed by the Bida Native Administration for the District, is
virtually a member of the chiefs council in the same capacity of a
‘departmental’ official. With the exception of this last office, the
ranks of the chiefs councillors are hereditary and ‘belong’ to the
various families which have held the ranks from times immemorial.
The succession to a vacant office is not, however, automatic, but
admits of a certain latitude, as for every vacancy there are bound to
be several candidates of approximately equal seniority and equal
claim. In the appointment of a new rank-holder by the chief
and his councillors, due weight is given to the reputation of the
candidate, his experience, intelligence, and economic success (as
canoeman or river-trader). The position of a councillor appears
to have carried no regular emoluments with it, except in the case
of the official in charge of the town .administration and thus the
collection of taxes in Muregi. Occasional gifts from the chief and
a share in the booty made on raids and warlike expeditions consti¬
tuted their official income.

The rank list of the ‘delegates’ is both larger and more flexible
than that of the councillors. It is frequently altered, increased, or
decreased, according to the demands of administration. The

members of this order of ranks are all recruited from the family
of the Kuta. Their ranks are graded, and follow a strict system of

precedence and promotion. Every promotion means a larger
measure of power and influence, for it goes hand in hand with
transfer to another, more important, and also more lucrative post.
A new as yet untitled member of the chief’s kin will as a rule be
appointed to one of the lower ranks; the higher ranks and more
responsible offices can only be reached by gradual promotion.
Promotion and first appointment are, again, decided by the chief,
in consultation with the other tribal notables; here, however, the
personal preferences of the electors count more than rigorous
qualifications: in this system of ranks based on promotion, long
experience is not regarded as a condition for a lower appointment,
nor already achieved economic success for a rise in rank which
carries with it increased economic benefits.

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 177

Succession to Chieftainship . Here we must turn back to the
position of the Kede chief and the question of succession to Kede
chieftainship. For the highest in the series of promotion which
we have just been discussing, the ultimate promotion open to
members of the chief's family, Is promotion to chieftainship. The
rank next to the chiefs, Egba (here meaning 'deputy'), is in fact
regarded as the rank of an 'heir apparent', and is as a rale held by
the most senior among the titled relatives of the chief, his younger
brother, or elder brother's son. Succession to chieftainship, more
rigid than succession to the other political offices, thus allows the
chief-to-be to consolidate his position in advance of his actual
appointment. The tribal notables exercise a certain indirect
influence: for in every one of the repeated routine decisions on
the promotion of a 'delegate' they already decide to some degree
his future chances as a candidate for chieftainship. But then, the
ruling chief is himself one of the 'electors’ and can easily turn the
decision in favour of the candidate whom he supports . 1 Here
becomes clear what I have said above about the kinship relation
between the Kede chief and his ‘delegates' tending to strengthen
the position of the chief. The mere tie of kinship between them
may conceivably prove a weak and unreliable support of his
authority; but the fact that the delegates remain dependent on
the favour of the chief for their promotion and political career in
general turns it into a bulwark of chiefly power.

With ‘councillors' and 'delegates' both dependent upon his
goodwill, the Kede chief exercised an almost absolute authority—
more absolute, I may add, than any other chief of Nupe. Repeated
promotion and transfer, all decided in the capital, tied the delegates
closely to Muregi and prevented them from making for them¬
selves too independent a position in their temporary dominions.
The 'absolute' power of the Kede chief thus appears as a necessary
element in the control of this mobile political system which, with
its scattered outposts and colonies, yet depended so much on
smooth co-operation and concerted action. The weakness of the
system lay in the fact that it allowed no legitimate check on the
power of the chief. A more equitable balance of power could
only be achieved by illegitimate means—that is, by feuds and

1 It is significant, in this connexion, that the present Kuta introduced a new
rank for his son when the list of traditional ranks was exhausted (see the chart
of Kede ranks).

278 AFRICAN POLITICAL SYSTEMS

factional splits within the hereditary ruling class. The recent
history of the Kede contains indeed evidence of such rivalry
between the Kede chief and his officers of State, or between
different claimants to Kede rulership, and the ensuing division
of the country into rival factions . 1 Attachment to one of the rival
factions is also the only means by which the subject classes, the
commoners with no rank and office, could exercise an indirect
influence upon the political management of their country.

Administration of ike *Colonies '. The following chart illustrates
the different grades in the rank list of Kede delegates. The ranks
are given in order of precedence, each rank together with the
district to which its holder is posted. The last five ranks on the
list do not belong to members of the chief’s family, but to that
group of 'household’ ranks which are also occasionally vested
with emissaryships.

Rank

Relation
to Kuta

Posted to
at present

Posted to
formerly

Egba

Elder brother’s

son (classific.)

Gbara

Kpatagban
(right bank)

Sonfara

Son

Lives in
Muregi, no
emissaryship

(A newly

introduced

rank)

Kofie

Younger brother

(classific.)

Raba

Raba

Ekpd

Younger brother
(classific.)

Kpacefu

—

Tswadiya

Younger brother

Ketsogi

Ketsogi

Lefiti

Younger brother
of Egba

Muregi, no
emissaryship

Kpacefu

Liman Gyedwa

Distant relation

>>

Kpasha
(right bank)

Tswadyagi

»

>>

Kpasha

™ * ff ct i. onal s P llts of this kind were occasionally utilized and fostered by

thev fn„nT erS ', T uf R ° yal Niger supported an Egba in whom

chilfLiTw * . valuable aU y agamst the ruling Kuta, promising the former the
r m ? turn for the support of his faction. And once or twice the
‘offi r baCked the rival cIaimant of Kede chieftainship against the

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 179

Rank

Relation
to Kuta

Posted to
at present

Posted to

formerly

Sodi

Younger brother

Katcha

(unofficially)

Muregi, no
emissaryship

Sheshi Kuta

Household-rank

Katcha

Eggan

(right bank)

Tsowa Kuta

Muregi, no
emissaryship

Egbagi
(right bank)

Capa Kuta

»>

s>

Wunangi

Mijindadi Kuta

Jebba

Buka

The list shows that the assignment of posts to political ranks
has undergone certain changes. They were due partly to the
re-alignment of modern administrative divisions, which placed
some of the former Kede areas outside of present-day Kede
country (e.g. the Kede settlements on the right bank of the Niger).
But partly also to changes in the economic and political importance
of certain places and the corresponding change in their official
‘appreciation’. Jebba and Katcha are illuminating examples. In
Jebba we find a man of comparatively low rank in charge of the
Kede community. He was posted there when Jebba was only just
beginning to become the important place it is to-day. The
Mijindadi is to-day a very old man, almost blind, and only
nominally in charge of the ‘colony’; he is generally expected to
be succeeded by a high rank-holder whose rank would do justice
to the importance of present-day Jebba. Katcha is, officially, in
charge of the Sheshi Kuta, another ‘household’ rank; he is,
however, unofficially assisted in his work by the Sodi, a relation
of the Kuta, who also lives in Katcha and, in fact, only waits till
this important Kede community will be handed to him as to the
more suitable representative of the Kuta.

The dominion of the delegate varies in extent and composition.
His district (especially if he resides in one of the twin-villages)
may comprise both tribal groups, Kede and non-Kede; or the
boundary of his dominion may be drawn round the Kede settle¬
ment, while the ‘native’ village (which would be some distance
from the Kede settlement or on the bank if the latter is on an
island) belongs to the country and political district inland. In

i8o AFRICAN POLITICAL SYSTEMS

either case, the dominion of the Kede delegate would stretch some
way up and down the river, comprising hamlets and villages on
(formerly) both banks. The ‘native’ villagers live under their own
chief and elders, and are on the whole left to themselves, except
for the political obligations towards the Kede rulers. In the
carrying out of these obligations (to be examined presently),
village chief and elders become mere subordinates of the Kede
government. The Kede families, on the other hand, which live
in the district of the delegate are his subjects in a different sense.
They too have their obligations towards the government which
he represents. But they share to some extent his privileged official
position vis-a-vis the ‘natives’ of the country. The family heads
of these Kede families assume certain titles in Nupe fashion, which
are to mark them as ‘elders’ of their community. In this case,
however, these are not the usual village ranks, nor are they tici
n ya Kuta (‘ranks of the Kata’), but are of the order of personal
or ‘household’ ranks, which the Kede delegate may confer upon
the family heads in ‘his’ town.

The official duties of the delegate relate to the three main
concerns of Kede administration. He is charged with the collec¬
tion of taxes on behalf of the Kede chief, the maintenance of law
and order in the districts, and he finally acts as the agent of the
chief in all matters requiring concerted action of the tribe as a
whole. The first two duties have undergone but comparatively
minor changes under modern administration. The last duty, the
most important aspect of which used to be contribution to' the
fighting expeditions of the tribe, is reduced to-day to such relatively
irrelevant activities as the arrangement of the periodical tours of

inspection through Kede country of the Kede chief or the District
Officer.

Taxation. The present system of taxation is based on an income
tax on sliding scale, assessed among the Kede on the basis of the
number of canoes owned. The tax is collected locally by the official
Village Head, and is then transmitted by the District Head to the
Native Administration Treasury in Bida. For Kede District read
delegate for Village Head (a certain number of Kede ‘delegates’
having been made Village Heads under the Native Administra-
ti°n)» and Kuta for District Head—in all other respects taxation
mnong the Kede is the same as in the inland districts of the
Emirate. Thiswasnotsoinpre-Britishtimes. The kintsoji villages,

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 181

like the other inland villages of Nupe, paid a certain annual
money tax, assessed p^r village, which was collected by the Kuta
and his delegates on behalf of the King of Nupe. The tax which
the Kede themselves paid, on the other hand, w r as of two kinds:
first, there was the tax proper, paid locally, at the village to which
one belongs. It was an income tax in the modem sense, consisting
in a percentage of the money Income of every canoe-owner (Le.
profits from trade and transport). Second, there was the albarka
(lit. ‘blessing’), a tribute voluntary but in name, which canoe-
owners were expected to pay to the delegates at whose place
they were stopping and doing business. The tribute varied in
amount: in Jebba it amounted to a 10 per cent., in Muregi to a
20 per cent, duty on all goods bought and sold. Failure to pay
meant forfeiting the permission to call and trade in the district.
The delegate returned to the Kuta half of his revenue from the
tax proper, and one-fifth of the revenue from the albarka.
The Kuta, in turn, handed about one-fifth of his total tax
revenue (including tax on his private trade and canoe profits) to
the Etsu in Bida.

Jurisdiction . The modem system provides for a professional

Mohammedan judge (Alkali), who holds court in Muregi, and to
whom all legal cases from the district have to be submitted.
Native Administration police assist him on the executive side. The
courts in the capital, Bida, are higher courts and courts of appeal
for Kede as for all other districts of the Emirate. Under the new
system, chief and delegate are allowed no judicial and only a limited
executive authority.

In pre-British Nupe the maintenance of law and order devolved

in different degree upon all existing political authorities, the local
delegate, the Kede chief, and the Emir of Nupe, in accordance
with the nature of the offence. The local delegate could only deal
with very minor offences of the kind which involves no restitution
and only domestic punishment, if any. All other aspects of public
security were regarded as being of direct concern to the State—
Kede or Nupe. Even the smallest theft came before the Kuta in
Muregi; adultery, litigation about bride-price or ^ inheritance
similarly fell under his jurisdiction. Certain major crimes, on the
other hand, were under the jurisdiction of the King of Nupe him¬
self. The list of these ‘crimes of the king’ (as they are still called)
comprises: highway robbery (including robbery on the nver

iSa AFRICAN POLITICAL SYSTEMS

highways of the country), homicide, seduction, and the crime Use
majeste (‘abuse of the king’, in Nupe phraseology). In securing
a criminal, the Kuta’s own men would act as a police force; if
the crime came under the heading of a ‘king’s crime’ the same
police force would take the prisoner to the capital, where
judgement would then be passed upon him by the king and his
councillors.

We can say that with the exception of offences against kinsliip
rules (e.g. the incest taboo) and offences against religious rules—
the various religious rules of the Mntsoji which were of no concern
to the Kede rulers—no serious breach of established rules of con¬
duct was left to private settlement or to the informal sanctions
of ‘public opinion’. The comprehensiveness of the claims of the
Kede State to enforce conduct within its boundaries must yet
admit of one significant limitation: the Kede must surrender
jurisdiction over certain major crimes committed in their territory
to their overlords; they must, in other words, agree to a limitation
of their autonomy in the interest of the Emirate at large. We
shall presently encounter another aspect of the maintenance of
‘law and order’ in which the prerogative of the Kede chief must
give way to that of the central government of the country, yet in
which the relation between the two is, in the very nature of things,
much less clearly defined: the territorial rights of the various
subject groups which are united under Nupe rule.

Territorial Rights. The Kede, as the overlords of the kintsoji,
guarantee to their own subject groups certain corporate territorial
rights. The semi-riverain kintsoji, as we have seen, derive their
livelihood to a large extent from fishing in the backwaters and
creeks of the River Niger. The fishing rights of the different kint-
soji villages were upheld by the Kede rulers and enforced if need
arose by the full military power at their disposal. Nupe kingdom
equally guaranteed corporate territorial rights to its various
subject groups. The Kede, as one of these subject groups, enjoyed
these territorial rights with respect to their undisturbed possession
of the whole riverain area.

But the presence of a strong and expanding group like the Kede
in the political framework of the Nupe State must invite conflicts
over the territorial rights of other subject groups of Nupe which
are neighbours of the Kede in the river area. As the more recent
history of Nupe shows, conflicts of this kind have frequently

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 183

arisen. Their occurrence or avoidance is clearly bound up with the
interpretation put on Kede political autonomy—a point which we
will discuss under a separate heading*

VL The Claim to Autonomy

Let me say, first, that the administration of Kede country
presents an exact analogy to the administration of Nupe kingdom.
Emissaries from the capital, delegates recruited from the ruling
house, are in charge of the administrative districts of the Emirate
as they are of the district of Kede country. This analogy reflects
the analogous political evolution in both countries: the rise to
power of a small group over a large country with heterogeneous
population. The only exception from this rule of administering the
districts of the Emirate through royal delegates is Kede country
itself, which remained under its own chief, who acted as the repre¬
sentative of the Eisu Nupe. Even for this exception there is a
parallel to be found in Kede organization: the emissary system did
not apply to the area of the upper-stream Kede. Their whole area
was regarded as one sub-district of Kede country and was placed
in charge, not of an emissary from Muregi, but of one of their own
chiefs, the village chief of Bele, their southernmost settlement. The
explanation for this privileged position of the two groups, the
Kede Tifin under Kede and the Kede under Nupe, is most
probably the same—namely, that their rulers would have found
it difficult to control themselves effectively the territory of
these subject communities: the Kede the country of the upper-
stream group, which was not easy to reach with their big canoes
(note that they made the chief in the place farthest down-stream
the ‘deputy’ for the whole group), and the Nupe kings the wiiole
river area.

To the down-stream Kede their autonomous position in Nupe

Emirate is a sacred trust, dating back to their first chief who
received, with the chieftainship of Kede, the ‘Rule over the
Water’ from the mythical Tsoede. The Kede chiefs still style
themselves Etsu nya nuwS (‘King of the Water), and, as their
history shows, have always taken this title very literally. It meant
to them more than merely the formal concession of assigning to a
Kede chief duties normally discharged by a royal delegate, and
they have in the past frequently attempted to acquire a larger

184 AFRICAN POLITICAL SYSTEMS

measure of independence in taxation, jurisdiction, and political
management in general . 1

In this connexion, we must point out an important development
in the history of Nupe kingdom, which in turn greatly affected
Kede history. The royal dynasty of Nupe, which traced its
origin back to the mythical Tsoede, was in power till the earlier
half of last century, when Nupe kingdom, like most of the Native
states in Northern Nigeria, fell under the rule of Emirs of Fulani
stock who had conquered the country and deposed the indigenous
kings. We have no data on the relation between Kede autonomy
and the Nupe State under the old dynasty. But it is certain that
under Fulani rule conflicts constantly arose, partly perhaps be¬
cause the Kede (like many other Nupe sections) resented the alien
rule, but to a large extent certainly because the Fulani kings, much
more than the Nupe rulers before them, had to curb in their own
interest the autonomous leanings of their Kede vassals. The
Fulani, whose wars at that time were directed chiefly against tribes
in the south, could not afford to let an all too independent section
control the river—the southern boundary and at the same timp the
vital artery of trade and traffic of the country. Slave trade, the
economic mainstay of the pre-British rulers of central Nigeria,
traffic in arms and powder, and troop transports for their military
expeditions, all had to cross the River Niger on Kede territory.*
The Fulani overlords enforced their sovereignty in a number of
wars and punitive expeditions, in the course of which (if the
reports are true) hundreds were killed, thousands of Kede sold as
slaves by the victorious Fulani, Kede notables executed in Bida,
and whole districts of Kede country devastated. A typical instance
is the Katcha War—the answer of the Fulani to the first attempt of

1 There exists certain evidence to show that the Kede chiefs succeeded in
enlarging their judicial power at the cost of their overlords and usurped a certain
judicial machinery of the Nupe;State which had 'evolved in the river area.
To-day, at any rate, the Kede claim that this judicial machinery was under the
authority of the Kuta, while other, non-Kede, informants state that it repre¬
sented entirely a prerogative of the kings of Nupe. I am referring to the Ledu
(ht. prison) villages on the banks of the Niger, which were so called because
they^ served as prisons and places of execution for criminals convicted, by the
^gs a crime of the king*. I have described this system and the

att 2 n? 1 * -° f the Kede t0 claim ** as their own in Man (1935), 143.

The importance, for example, of Raba as a river port for slave traffic to the
south is pouted out by Lander (Journal of an Expedition (1832), ii, p. 298).

for Fuwl? f Spea u 0 lr°° Kede Canoes - <aI1 of may be employed

for Ful troops to cross the Niger on war expeditions* (op. cit., ii, p. 315).

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 185

the Kede to force the tributary of the Niger on which Katcha is
situated and to occupy the town. The people of Katcha, under
allegiance to Bida, claimed protection for their territorial rights
from the central government. The Kede were routed by a Fulani
army, and the political rights of the Katcha people restored. 1

The feud between the Kede and the Emirs of Bida, and, above
all, the strategic position in the defence of the kingdom which the
river tribe occupied, was utilized in the Royal Niger Company
campaign against Bida in 1897. With promises of political
autonomy to the tribe and the chieftainship to the ambitious
nephew of the ruling Kuta, Sir William Wallace, the commanding
officer of the Niger Company troops, secured the support of the
Kede. Their canoe fleet was placed under the command of a
gunboat and massed on the river south of Bida; with their help, the
relieving force of the Nupe, which at the time was engaged in a
war in the south, was cut off and Bida captured. The Kede
received the promised reward and were granted semi-autonomy,
being made responsible directly to the British Administration
instead of to their former overlords, the Emirs of Bida.

The political status of the Kede was later changed again. The
Kuta lost a considerable portion of his territory when the new
provinces and divisions were mapped out (1900-1905). At the
same time, his autonomous position was curtailed, and Kede
country placed again under Bida. A last element of autonomy is
preserved in the regulation that the District Headship of Kede
District should remain vested in the Kede chief.

The adoption of this curtailed autonomy for the purposes of
modem government has not been an unqualified success. Admin¬
istrative officers have had cause to complain of the declining
authority of the Kede chief. The more evident supports of Kede
chieftainship—the dynastic title, the privileged economic position
of the chief (a comparatively high salary having taken the place of
the former share in taxes and tributes), and his control over the
officials of his State—have not been affected so fundamentally

1 This first attempt was carried out in the early days of the Royal Niger
Company. The second, successful, attempt of the Kede to gain a foothold in
the Katcha River was carried out under an administration favourably disposed
towards Kede expansion, and led to the peaceful occupation of the riverside at
Katcha mentioned previously. This ‘peaceful penetration’, however, is hardly
less resented by the Katcha people, who have lost through it their formerly
undisturbed fishing grounds.

z 86 AFRICAN POLITICAL SYSTEMS

by the political changes as to account for the decline of chief!;
authority. What has changed is rather the spiritual background
the conception, of Kede autonomous chieftainship. It has los
its place in the scheme of Kede political life, which is sti
growth and expansion. It has been shorn of the qualities of self
responsible and exclusive leadership, which, in this dynami
society, constituted the raison d'etre of its once absolute powei
What remains of the leadership vested in the Kede chief is
leadership moving in the narrow circle of modern administrate
boundaries and under the patronage of another, higher, authority
It had to surrender its prerogative of guiding and protecting th
movements of the group as they followed, farther and farthe
afield, the lead of the general economic and cultural developmen
of the country.

This change in the whole political existence of the Kede is mos

conspicuous in the Kede groups which now lie outside Bid
Emirate and the area ruled over by the Kuta. Here the liquidate
of Kede autonomy could not have been effected without disturb
ing consequences had Kede chieftainship still held its old meaning
Sentimental bonds between the split-off groups and the mother
country are still conspicuous; but on the whole I have gained th
impression that these groups do not (or no longer) seriously mis
or resent the separation, or contemplate the possibility of som
future reunion. The weakened appeal of political autonom;
reflects the general development, political and cultural, that ha
overtaken the country and separated autonomy from its vita
counterparts in social life: economic and cultural unity, and th
solidarity of community life. To put it differently: culture
self-realization has become possible outside the narrow politics
boundaries. I shall have to return to this point; but let me men
tion here a certain proof of the statement just made—namely, th
continued, and even increasing, emigration to those distant Ked
settlements which could under no circumstances be anything els
but ‘minorities" under alien rule.

VII. Social Stratification

The division of political rights and duties in Kede country anc
implied in it, the unequal economic advantage enjoyed by th
different ^sections of the population constitute the basis of a rigi
social stratification. Ethnic and cultural differences add thei

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 187

weight to it in that they coincide with, and are interpreted in the

sense of, distinctions implying social gradation.

On the top of the social structure, we have the hereditary ruling
class formed by the chiefs family group, and below it the second
stratum, which comprises the Kede of common status. Though
the latter exercise no direct influence upon the policy of the State,,
they can yet in certain respects range themselves beside the ruling
class: if the members of the ruling class decide the fate of the
country and derive the main benefits from all political exploits,
war or peaceful acquisition of territory—the Kede commoners take
an active part in all activities on a tribal scale, even bearing the
brunt in the most significant activity, colonization, and benefit to
a considerable extent from the political successes of their rulers.
Indirectly, by attaching themselves in subservient position to the
households and factions of the ruling class, they attain a larger
measure of political influence and certain economic advantages not
warranted by their hereditary status. Moreover, their racial and
cultural affinity with the class above, and the proportionate remote¬
ness from the lowest class beneath, place the Kede commoners on
one plane with the ruling group of the country. The lowest
stratum comprises the ‘original inhabitants’, whom tradition
paints as having been the subjects of the Kede from mythical
times and having had no culture worth mentioning before the
advent of the Kede. In fact, they are still, seen with Kede eyes,
a ‘primitive’ and slightly contemptible group; they fail by the
cultural standards which the Kede recognize, being pagans,
inferior canoemen, and relatively poor. They were excluded from
the tribal activities of the Kede and all benefits accruing from
them; yet they suffered more than the other groups from the
failures of their rulers’ policy, unsuccessful wars or revolts, for
their country might become the battle-field and their farms and
villages be destroyed, while, unlike the ruling classes, they had no
reward to expect in the event of success.

The question arises by what machinery the Kede maintained
this rigid division into hereditary classes, kept intact the solidarity
of their small minority, yet were able at the same time to integrate
and bind to loyalty the heterogeneous sections of their State. 1 Our

1 The strong bond of loyalty that existed between the Kede and their subject
groups is illustrated in the historical fact that in the Katcha War certain kintsoji
groups fought on the side of the Kede against the Emir of Bida.

x88

AFRICAN POLITICAL SYSTEMS

examination of the political organization has revealed one such

machinery: the machinery of coercion. We must ask now how
far other, non-political, forms of integration lend their support,
or possibly are made to lend it, to political unification. We can
distinguish three types of such supporting integrative agencies: (i)
integration through actual co-operation between the sections of
the population; (2) integration in the spiritual sphere—in other
words, through ideologies teaching or preaching unity; (3) in¬
tegration based on both. Instances of the first type are economic
co-operation and community life, of the second, tradition and
mythology, and of the third, religious practice.

VIII. Integrative Mechanisms

Economic Co-operation and Community Life. The propinquity of
the Kede and kintsoji settlements, combined with the difference
in their productive system, invites a certain measure of co-opera¬
tion in the economic field. The Kede buy farm-produce from
their peasant neighbours, who, in turn, use to some extent Kede
transport to dispose of their surplus on the large river markets.
This co-operation is by no means exclusive and does not lead to
complete dependence upon each other. The peasants also sell
some of their farm-produce inland or take their fish directly, on
their own canoes, to riverside markets; similarly the Kede buy a
certain amount of their food on the various outside markets which
they visit on their river voyages.

In their community life, the two groups hardly achieve a more
intensive, or less casual, co-operation. The difference in occupa¬
tion and in the main interests of their lives is not balanced by any
other strong ties. The young folk of the Kede and kintsoji
frequently join in each other’s dances; here and there friendships
are struck between individuals from the two groups. But apart
from these contacts the two sections keep to themselves. The age-
grade associations do not stretch across the tribal boundary and,
above all, there is almost no inter-marriage: the kintsoji arrange
their marriages with their tribal relations inland, and the Kede
many among themselves. 1 In these marriages between Kede
villages distance plays no part, which is rather significant,-for the
(inland) Nupe generally dislike marriages between distant places.

1 is true of all the older Kede settlements; in the more recent settle¬
ments (e.g. Katcha) I found a few cases of inter-marriage with the kintsoji.

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 189

The various Kede groups are thus all related, not only, rather
vaguely, by common descent, being all emigrants of common
stock, but also very concretely, by repeated inter-marriage. Per¬
sonal contact between fellow tribesmen, friends, and relations
among the Kede, whatever the distance of their habitations, is
maintained by occasional visits, and revived regularly every trad¬
ing season, when the Kede canoes travel from place to place.
Finally, the solidarity of the Kede community is strengthened
periodically, when new groups of immigrants come out from ‘home’
to join fellow tribesmen who have emigrated earlier in their
‘colony’.

Tradition and Mythology. The Kede share their traditions with
their mother tribe, and, like the Nupe, trace their origin back to
the mythical Tsoede or Edegi, of whom we have heard already
(see p. 175). This Tsoede is said to have been a Nupe man who
lived about 1400, and who was sent as slave to Idah, to which
country Nupe was tributary at that time. He won the favour of
the King of Idah, so much so that he evoked the jealousy of the
sons of the king, and eventually had to flee the country. He set
out in a bronze canoe, loaded with gifts from the king and manned
by other Nupe slaves, to return to his country and to make himself
king there,. the rule of Nupe having been the King of Idah’s
parting gift to his favourite. On his flight to Nupe, Tsoede was
helped by two men whom he met on the river—one sitting on
a stone in midstream (kuta ), 1 and one fishing with a fish trap
(ekpa). When Tsoede established himself as King of Nupe he
rewarded these two men by making them chief over the whole
river and its tribes (Kuta), and a high rank-holder in the new
river State (Ekpa), respectively. 2 These men were the ancestors
of the present Kede, and the first to exercise the ‘Rule over the
Water’, which has remained in the possession of the Kede since.

It Clearly does not come within the compass of the present study
to examine how far these legendary data might contain a kernel
of historical truth. Their importance to us lies rather in the
sociological significance of the ‘truth’ which they announce—that

1 Kuta, stone, seems to be an old form of Nupe; it occurs in two obsolete
Nupe dialects, and also in Gbari, a related language. The modem Nupe word
for stone is taku (the syllables of kuta reversed).

2 The Nupe are very fond of such puns. There exists another version of
this legend which derives the title Kuta from the fact that the man on the stone
was wearing two gowns ‘on top of each other’—in Nupe, ku ta do$i.

i9o AFRICAN POLITICAL SYSTEMS

is, in the influence upon actual social life which the belief in this
‘truth’ entails. The Kede tradition of origin, as we have seen,
concerns itself above all with the beginnings of the political history
of the tribe. In the fashion of such myths of origin, it anchors
the existing system in a dim past, which, by its very remoteness
and its supernatural and sacred associations, endows the present
with an immensely convincing validity. The Kede tradition of
origin ranges its strongest spiritual support behind what we now
recognize to be the two fundamental features of their political
system: the fact that the extent of Kede rule is defined on the
basis of locality, and not tribe, and the fact that, within the larger
political unit of Nupe kingdom, the Kede claim a semi-autonomous
position. Let me emphasize that the knowledge of this tradition
is not limited to the Kede, but is common to all Nupe sections,
and among them the various kintsoji groups. The common
possession of the myth represents a spiritual link of utmost
importance: for with the myth the subjects of the Kede also
accept the system which it is meant to guarantee— the overlordship
of the Kede.

Religion. As I have mentioned, the Kede are to-day Moham¬
medans— very eager Mohammedans, in fact. In many settlements
you will find mallams teaching the Koran to boys and adults, often
pupils who are able to attend the classes only when their travels
happen to take them to this place and allow them a short sojourn.
The kintsogiy on the other hand, are still largely pagans. More¬
over, against the solid religious unity of the Kede, the kintsoji
show a certain diversity of religious rites and beliefs, which reflects
the composite nature of this group. 1 However, there exist two
rituals, both essentially river rituals, which are common to all the
semi-riverain groups, and a third ritual, linked with the memory
of Tsoede, which exists in all the larger riverain (and partly also
inland) villages of Nupe. The first of these rituals is the Nddduma
(the Nupe name for the River Niger), an annual sacrifice to the
spirits of the river, performed in all the different river villages

1 We cannot say what the religious situation in the river valley has been in
the earlier periods of Nupe history. The probability is that the Kede, as most
Nupe sub-tribes, possessed certain special rites and beliefs, which were not
shared by the other tribal groups. The characteristic difference between Kede
and kintsoji to-day, between a solid religious unit on one side and a hetero¬
geneous religious group on the other, would thus have been equally marked
m the pre-Islamic era.

THE KBDE: A RIVERAIN STATE IN NORTHERN NIGERIA 191

and believed to procure, and at the same time to keep within limits,
the annual flooding of the Niger. The second ritual is the Ketsd,
a sacrifice to the spirit of the huge rock of the same name which
rises abruptly from midstream near Jebba (known to Europeans
as Juju-rock); this sacrifice is believed to cure illness and barren¬
ness and to secure luck in fishing or trade. The third ritual is the
sacrifice to the Chain of Tsoede , a sacred relic which secures fertility,
prevents illness, and is also used as an instrument of ordeal.

Now, these three rituals also occur among the Kede; the first
two have in fact been, as it were, usurped by them, and made to
some extent their own concern. The Kede chief holds himself
responsible for certain special, most conspicuous, performances of
these river ceremonies. Once a year a special Nddduma is per¬
formed in a place called Bazumagi, north of Jebba, the Kuta
himself providing the sacrificial food, a white bull and honey.
Unlike the local Nddduma rituals, the Nddduma of Bazumagi is
enacted on behalf,, not of a particular village, but of the whole
population of the river valley. The chief’s Nddduma also shows
another feature which is absent in the local rite: the priest climbs
a rock in midstream and throws a stone towards the bank; the
spot where the stone falls is believed to mark the line to which
the river will rise that year. The Ketsd becomes a chief’s ritual
at the apppintment of a new Kuta. About one month after his
accession the n ( ew chief sends a bull (preferably white) to Jebba
Island to be sacrificed there, by the local priest, again on behalf of the
whole riverain community, to inaugurate and secure a prosperous
reign. A special Chain of Tsoede finally is kept in Muregi, in the
house of one of the chief’s councillors (see p. 176). He performs
the annual sacrifice of beer and the blood of a sheep, beer and
animal being again provided by the Kuta himself.

It is interesting to note that in none of these rituals may the
Kuta or any member of his family be present. Thus the double
nature of these rites, which, though stamped chief’s rituals, are
essentially rituals of the kintsoji , is conspicuously symbolized.
The interests which they voice—the securing of safety and liveli¬
hood to the people on the river- —are adopted and made their own
by the Kede rulers. A single belief and a common cult comprise
both rulers and ruled, notwithstanding the religious barriers which
otherwise separate the two, and add to the political dependence
of the kintsoji another, spiritual, dependence.

I9 2 AFRICAN POLITICAL SYSTEMS

Conclusions. We may say then, in conclusion, that economic
co-operation and community life only underline the political and
tribal divisions which cut across Kede society. As integrative
forces which could foster a solidarity pertaining to the political
system at large, they fail. Only the ideological influences, myth
and religion, succeed in this. They anchor the external political
unit in more deeply rooted interests and sentiments. They add
to political coercion the more subtle persuasion of supernatural
arguments, of beliefs in the necessity and fore-ordination of the
existing system.

I have been using the present tense with some liberty. It was
applied correctly if we are thinking of the last ten or fifteen years,
but inaccurately if we consider the immediate present and, above
all, the future. The rite of the Chain of Tsoede is still performed
annually, in completely Mohammedanized Muregi. The bull
sacrifice at the Ketsd was performed when the present Kuta
succeeded to the chieftainship; whether it will be repeated for
his successor is open to doubt. The Nddduma y at any rate, although
still existing as a local rite, is no longer carried out by the Kede
chief. We have spoken before of the decline of authority of the
Kede chief. It may seem surprising that a chieftainship which
had been forced to give up most of the qualities from which it
formerly drew its strength should so easily discard these ‘binding
forces* of religion. The explanation lies, again, in the changed
conception of Kede chieftainship. Kede rule has exchanged its
dynamic and expanding nature for the secure, aquiescent authority
under the Pax Britannica. It can dispense with the binding forces
of religion, which used to uphold the autocratic rulership of a
small minority; it can, above all, afford- to discontinue a practice
which, to the Mohammedan chiefs of Kede country, appears as
a concession to their less enlightened subjects. Kede rule has thus
paradoxically weakened itself in its new-found security. And in
this the chiefs of Kede do not stand alone; this paradox is, I believe,
a not uncommon feature of modern, static, Government-backed
chieftainship in Africa.

The incipient dissolution of the ‘binding force 5 of religion in the
Kede State is only following in the wake of the general dissolution
of the solidarity which it was meant to uphold. The economic

development of the country led, as we have seen, to an extensive
co-operation with outside groups and to the founding of colonies

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 193

ill areas to which political rule could never be expected to follow.
Common interests and culture, community life, intermarriage
stretch as far as these outposts. Moreover, cultural assimilation
and intensified contact have removed some of the barriers which
formerly separated the Kede from neighbouring groups. The facts
of common culture, economic co-operation, and community life
thus no longer converge to cement the solidarity of a ruling group,
but merely outline the much vaguer unit of a scattered ethnic
group which has abandoned all claims to political self-realization.

IX. The Evolution of the Kede State

Our examination of the Kede State and its history does neither
support nor refute the theory of the origin of the State which,
accepted to-day by many students of society, derives every State
organization from an original invasion and eventual conquest of
one ethnic group by another. 1 Our data have shown political
domination of one ethnic group by another to be a factor of para¬
mount importance; they also revealed the occurrence of clash and
conquest—though not on the comprehensive scale implied in this
theory of the State; but they do not prove an original group
invasion, beyond that which we must relegate to the era of
mythical and thus unverifiable events. Our data can, in fact, also
be taken as evidence for an ‘internal diversification’ 8 and a gradual
emergence to political supremacy of one out of a number of ethnic
groups.

Our data relating to Kede expansion in recent times, on the

other hand, tend to confirm another sociological theory which
concerns itself with social origins—namely, the theory which holds
that migration and colonization are never a result of over-popula¬
tion, but rather an expression of that ‘spirit of hope’ and ‘enter¬
prise* which is absent in countries with over-population. 8 Can we
accept this description of the growing Kede society as a final state¬
ment ? Is the emergence of a tribal section to a dominant political
role fully explained by this reference to psychological charac¬
teristics ?

Our analysis of Kede political organization has emphasized a
somewhat different aspect; it outlined the political system in its

1 See F. Oppenheimer, op. cit.

2 R. Lowie, op. cit., p. 40.

* A. M. Carr-Sauraders, The Population Problem (1922), pp. 299

I94 AFRICAN POLITICAL SYSTEMS

dependence on economic factors and, tied up with them, environ¬
mental conditions. The impetus to colonization and expansion, as
we saw it, was already implied in the way in which the people gain
their livelihood. We can render this argument even more conclu¬
sive; the cultural situation in the river valley itself offers us the
comparative material from which to draw our deductions. We
remember the upper-stream Kede, who, neighbours, fellow
tribesmen and in many details of culture close relations of the
down-stream Kede, yet do not share their political achievements—
and also do not share their productive system and economic life in
general. We can even go further and point to the environmental
factors on which the economic system of the Kede in turn depends.
The environment in which the upper-stream Kede live indeed
forbids a development of river trade and traffic similar to that
evolved by the sister group. The river in their area is narrow,
barred in several places by rapids and rocky, frequently dan¬
gerous, passages, and generally impassable during half the year.
On its banks there are few settlements, and the hinterland is thinly
populated, inhabited largely by comparatively poor and backward
groups. Hold against this the down-stream country: a broad
river, navigable all the year round, the banks covered with
numerous villages and the hinterland a rich, populous country
with highly developed trades and crafts.

It will have been seen that I am speaking here of environment in
a rather wide sense, including, beside the physical constellation of
the country, also such factors as distribution of population and the
existence of a certain type of civilization. Methodologically, these
facts stand in the same category as physical environment proper.
True, they are essentially historical facts, representing the results
of various historical developments. These developments them¬
selves, however, are beyond our line of vision; their results—the
fact that the Kede area happens to lie in the centre of a rich and
powerful kingdom—are to us ‘unique events’ in the phraseology
of the historian, extraneous factors of chance, which, like environ¬
mental facts, we have to accept as ultimate data.

Let me admit that this causal chain, environmental conditions—
economic enterprise—political system, is not fully conclusive. It
is weakened by another aspect of our comparative evidence—
namely, the fact that the kintsoji, who are also close neighbours and
tribal relations of the down-stream Kede, and who share with

THE KEDE: A RIVERAIN STATE IN NORTHERN NIGERIA 195

them the same 'stimulating 5 environment, appear to have remained
unresponsive to it. 1 have no explanation to give. Does this mean,
then, that we have to fall back on the psychological interpretation

of the kind quoted above ?

But the environmental and economic ‘determinism 5 which these
conclusions have put forward was not meant to exclude com¬
pletely the contribution of psychological factors—that is, the social
motive power that may lie in the temperamental and general
psychological dispositions typical of a group. Nor was it intended
to minimize the decisive part that enterprising and far-seeing indi¬
viduals must have played in the creation of the Kede State. The
river colonization was undoubtedly the work of a people fully
deserving the attributes ‘adventurous 5 , ‘courageous 5 , or ‘possessed
of a spirit of enterprise’; it must also have been closely bound up
with the leadership of certain outstanding individuals: think of the
man who was responsible for the Kede throwing in their lot with
the British, or the Kede chiefs who so successfully utilized the
encouragement of the Niger Company for the expansion of their
country. But two facts must be borne in mind when defending this
psychological and ‘individualistic 5 theory of social origins. First,
the psychological characteristics which might be made responsible
for the achievements of the group are not racially determined (i.e.
by heredity)—the dissimilar social system of the upper-stream
Kede proves this to the full; they remain an expression of, and a
perfect adjustment to, environmental conditions. And, second,
these psychological characteristics do not reflect the effectiveness
of some spontaneous, as it were, self-contained, psychological force,
but are fostered and formed planfully by the existing social system
and its cultural demands, to which the individuals, generation after
generation, learn to adjust themselves. How much in the gradual
social development of the Kede State was due to the selective
effects of environment, and how much to the spontaneous actions
of exceptionally gifted individuals who, at one point of Kede his¬
tory, may have shown their people a new way of life, is one of
these questions of social origins to which, again, we have no
answer.

SKETCH MAP OF KAVIRONDO

THE POLITICAL ORGANIZATION OF THE BANTU OF

KAVIRONDO

By Gunter Wagner
I. Introductory

T HIS essay proposes to deal with the political organization

of two tribes living in the westernmost part of Kenya, the
Logoli and Vugusu, which, together with a number of others, are
usually referred to as the ‘Bantu Kavirondo’. The tribes grouped
under this name are politically quite independent, though cul¬
turally and linguistically closely related with one another. They
do nbt call themselves by a common name. The name Kavirondo,
although its origin and etymology are not entirely clear, was ap¬
parently given to them by Arab and Swahili traders. 1 It refers to
the rolling, open plains that extend, broken by the Nyando Valley,
from Mt. Elgon in the north to the Kenya-Tanganyika boundary
in the south, as well as to the various Bantu and Nilotic tribes
inhabiting these plains.

Migratory accounts and minor differences in language and
custom make it possible to distinguish several larger divisions
among the ‘Bantu Kavirondo’, each comprising a number of tribal
groups. 2 Such a larger division has no common name either, but
the tribal groups of which it is composed are conscious of their
similarities and explain them either by a vague relationship in the
distant past or by a long period of neighbourliness and intermar¬
riage. Between tribes belonging to different divisions of this kind
a more or less permanent state of warfare, broken by seasonal
periods of truce, was the rule. There is no record, however, of
combined warfare of one group of tribes against another. War

1 cf. Johnston, Sir Harry, The Uganda Protectorate (London, 1902), vol. ii,
p. 722 f.

2 Two such larger groupings are (1) the Kitosh group, comprising the Vugusu
of South Kitosh and Kimilili, the Tadjoni and some small groups of Milo-
Hamitic origin, such as the Ngoma (Ngomanek) and Lago (El Bawgek), and
(2) the Wanga group, comprising the Wanga proper, the Marama, and the
Tsotso.

:oS AFRICAN POLITICAL SYSTEMS

expeditions or raids were limited to the immediate neighbours, and
tribes living more than twenty or thirty miles away were regarded

as too remote to be either friends or foes.

The tribes immediately surrounding the ‘Bantu Kavirondo’, who

are all of non-Bantu stock, lived in a constant state of war or ten¬
sion with the Bantu. The Nilo-hamitic tribes to the east of Kavir-
ondo, 1 who are predominantly pastoral, attacked the Kavirondo
mainly for the purpose of raiding their cattle. The Teso and Luo
in the west were bent on the conquest of territory and, as far as
place-names and tribal traditions reveal, have gradually pushed the
Bantu tribes eastwards. Their successive fronts of retreat run,
generally speaking, parallel to the present Kenya-Uganda
boundary. The pressure exercised by surrounding tribes upon the
‘Bantu Kavirondo’ does not, however, seem to have been excessive,
for, in spite of their great similarity in language and culture, it
did not w T eld them into a political and military unit. It may even
be that their eastward retreat was at times and in some areas
voluntary, as their present territory is at least as fertile and more
healthy than the country from which they have moved.

In none of the approximately twenty tribal groups which make
up the 300,000 Bantu of Kavirondo has political integration
reached a very high degree, but it differs sufficiently in the various
tribes to make generalizations from conditions in one area impos¬
sible. The following analysis, therefore, claims to apply to the
two sub-tribes only of which a detailed study has been made, viz.
the Logoli in the south and the Vugusu in the north. Both tribes
have neighbours of non-Bantu stock along part of their boundary:
the Logoli the Nilotic Jaluo and the Nyangori, a Nandi-speaking
group, and the Vugusu the Teso, the El Kony, an offshoot of the
Nandi living on the lower slopes of Mt. Elgon, and the IJasin-
Gishu Masai, who frequently raided their country. In defence
against these raids, the Vugusu lived in walled villages, the con¬
struction and maintenance of which demanded the co-operation
of a large number of people, while the Logoli, like most of the
other tribes, lived in isolated homesteads that were scattered over
the whole countryside. The Logoli, who at present number
approximately 45,000, inhabit an exceedingly fertile and well-
watered stretch of country which permits of a very dense popula¬
tion. The Vugusu number about 40,000, but are dispersed over
1 The Nandi, Uasin-Gishu Masai, and El Kony.

THE BANTU OF KAVIRONDO

199

a less fertile, grassy plain, about seven times the size of the area
occupied by the Logoi !. 1 Like all Bantu Kavirondo, both tribes are
mixed pastoral-agricultural. They practise hoe-culture to a fairly
even extent, but while the Logoli own two or three head of cattle
only per family, the average among the Yugusu is nine head per
family, and individual herds of sixty to eighty head are not
exceptional among them.

IL Definition of ike Politico! Unit

The logical starting-point of any study of political organization is
the demarcation of the political unit as the group and area of refer¬
ence. In so far as the concept of the political unit involves the notions
of power and authority, it would have to be defined as constituting
that group of people which submits persistently and in an organ¬
ized manner to leadership for the purpose of maintaining itself as
a unit. It is thereby distinguished from other groups over which
it exercises no authority and in contrast to which it recognizes
and promotes its own unity. It may or may not maintain relation¬
ships with such other groups, and these relationships may either
be friendly or hostile, depending upon the preponderence of
common or mutual or of conflicting interests between them. The
political structure of the unit thus defined would consist of the
system of political institutions which maintain the unit as an
entity, protecting it against disintegration from within as well as

Maragoli) and 73 among the Vugusu (of North Kitosh or Kimiiiii district).
Census of 193a.

200

AFRICAN POLITICAL SYSTEMS

unit as long as these smaller groups are subordinated to it, i.e.
derive their authority from the central government. In the case
of the Kavirondo tribes, however, no such definition of the
political unit on the basis of internal and external sovereignty can
be given. As regards submission to political leadership, the largest
groups, both among the Logoli and the Vugusu, are the
exogamous, patrilineal clans 1 or clan groupings, consisting of one
larger and several smaller clans, but not the whole tribal society.
The tribal unit is marked by the belief in the common descent of
all clans from one remote tribal ancestor, Murogoli and Muvugusu
respectively, and by. the occupation of a continuous stretch of
territory. In addition, there are numerous institutionalized forms
of co-operation and interdependencies between the different clans
of the tribal group which distinguish inter-clan relations from
inter-tribal relations, but there is no tribal authority which over¬
rules clan authority, either in its dealings with foreign tribes or in
the management of its internal affairs. In terms of the definition
given above, the clan would thus have to be regarded as the only
political unit.

A number of considerations, however, make it seem more

adequate to widen the definition of the political unit so that
it comprises the tribal unit rather than the clan. In the first
place, the fact that clans are exogamous and that marriage is
regulated in such a way that all clans of the tribal group inter¬
marry, establishes a close connexion between the clans. The
kinship bonds, maintained between every member of the clan and
his maternal kin as well as his affinal relations, are so numerous
and so strong that they establish bonds between the clans which,
as we shall see, are in many ways as binding as if there were a
central authority overruling that of the clans. Furthermore, the
cult of a common tribal ancestor, to whom the Logoli sacrifice
on a tribal scale at regular intervals, and the performance of the
circumcision rites on a tribal scale create a feeling of unity which
serves as a sanction for close co-operation in all matters affecting
the tribal group.

The concept of political structure*, likewise, requires a wider
definition than is customary to become applicable to Kavirondo

1 The clmjoluhia, luyia) is named after its real or supposed founder and its
members tend to form a territorial unit. The various characteristics of the clan
are discussed later on in this chapter in their respective contexts.

201

THE BANTU OF KAVIRONDO

society. There is no political structure as distinct from the kinship
and social structure; that is, there exists no system of institu¬
tions that serve explicitly and exclusively the purpose of maintain¬
ing the tribal unit as a whole. To enable one to understand the
organization of the tribal unit, the emphasis must, therefore, be
shifted from the concept of the political institution to that of the
political function. The assumption that each function in a culture
must have its corresponding institution—religious, economic,
political, &c.—would cut short an understanding of the way in
which cultures are integrated into a body politic, the institutions
of which are not yet clearly differentiated according to different
aspects, but which serve many functions at the same time. Even
in advanced communities where institutions are highly differen¬
tiated, they present only a visible superstructure, while their bases,
the forces that sustain them, extend throughout the whole
structure of society.

The political unit must thus be defined in terms of a conscious¬
ness of unity and interdependence rather than in terms of sub¬
mission to a central authority. The tribe, as a political unit, is a

group of internally and externally ‘sovereign* clans, which are
conscious of having sprung originally from a common ancestor
and which are interconnected by the bonds of intermarriage as well
as by common practices and beliefs in such a way that they
consider themselves a unit in contrast to surrounding groups
with whom they do not maintain such bonds. This tribal political
unit does not necessarily act as a body in all its foreign relations,
but it is merely the largest unit of people which feels as a unit- and
which—on certain occasions—acts as one.

The term ‘political* will, accordingly, be used with reference to
any form of socially sanctioned behaviour which, directly or
indirectly, strengthens the unity of the tribal group, whether that
be its primary purpose or not. An institution thus has political
significance if it fulfils a political, function, regardless of what
other functions it may perform besides. The political structure,
in this sense, is the sum total of all forms of sanctioned behaviour
which serve, directly or indirectly, intendedly or not, to integrate
the political unit.

We shall.examine now where in the cultural life of the ‘Bantu
Kavirondo* this political structure resides.

202

AFRICAN POLITICAL SYSTEMS

III. The Internal Political Structure

From the point of view of political integration a clear distinction
between the external and internal aspects of the political structure
is hardly admissible, as both aspects are closely interdependent.
External pressure or other stimuli affecting the political unit from
without form the strongest incentive for internal unity and, vice
versa, the external action of the group depends very largely upon
the nature and degree of its internal cohesion. For the purposes
of an analysis of political functions, however, a distinction between
the external and internal governmental functions of the tribal unit
must be made, as both have different situations to meet. The
internal maintenance of the tribal society involves three major
political or governmental functions: (i) the enactment of laws,
(2) the maintenance of law and custom, involving both their
perpetuation in periods where they are inoperative and their trans¬
mission to succeeding generations; and (3) the restoration of
breaches of the law.

(1) Enactment of Laws. The general body of tribal norms is,
in native opinion, as old as the tribe itself. There exists no
historical or legendary account of a law-giver. Law and custom
are believed to have been handed down from unknown times
from ancestor to ancestor, and it appears to be the cumulative
weight of ancestral authority which serves as the most general
sanction for the observance of traditional norms . 1 Normally, the
suggestion of questioning the validity of tribal norms is rejected
both by motives of fear and of suspicion. Fear is felt that
deviation from established norms will evoke punishment by the
ancestral spirits. Such punishment is not thought to be limited
to the action of the immediate ancestors from whom one
normally has to fear unfriendly acts, but to consist in a
general, although vaguely defined, displeasure of the spirit-
world which might have disastrous consequences of any kind.
The suspicion of practising witchcraft is felt towards any person
who deliberately and persistently defies established norms.
If such a person cannot be brought back to reason by the

1 Remarks such as: ‘This is the rule since long ago wnich all our grandfathers
followed, It has been ordered (okulaga) by our grandfathers ( avadaddf , or
‘Our forefathers never did like that’ are frequently made by the elders when
discussing a case in the present-day tribal courts.

203

THE BANTU OF KAVIROXDO

performance of sacrifices, he is socially ostracized and, in an
extreme case, killed . 1

There is, accordingly, no recognized authority which wields
legislative powers; law is, in theory at least, unchangeable, and
the degree to which an action, a claim, or an obligation are in line
with the age-honoured tradition is the only criterion of their merits.

Nevertheless, there are a number of legendary traditions and
even historical data which indicate that the system of laws and
customs was not as rigidly closed against changes and innovations
as one is at first led to believe. Apart from norms observed
throughout the tribal group, there are numerous clan rules which
differ from clan to clan, but which, within each clan, are made
valid by the same type of sanctions as tribal law. Most important
among such clan rules are certain food taboos or rules of avoidance
concerning certain forms of behaviour. While in many cases the
origin of these rules is unknown, in others the time is still remem¬
bered when the rule was not yet in force, and the account given
for its origin often bears the stamp of a true historical tradition.
Thus among the Vugusu a certain clan refrains from wearing
finger-rings of coiled iron wire. In explanation of this rule, it is
said that some generations ago a member of that clan suffered
from a sore and swollen finger caused by his ring, which gradually
became worse until he died without having been able to remove
the ring from his finger. Before his death, he is supposed to have
said that ‘it was a bad thing for the people of his clan to wear iron
finger-rings, and that all who would do so in future should die
as he was now going to die’.

Numerous ceremonial rules, especially details of ritual in
connexion with sacrifices, purification rites, Sec., which are
observed by some clans but not by others have probably been
enacted in the same way. Although neighbouring clans are aware
of the existence of such sacrificial rules, it would not ordinarily
occur to them to copy them ‘as they were not theirs’. If, however,
an exceptional situation arose, e.g. the repeated and conspicuous
failure of the traditional ritual procedure to achieve its desired
ends, a man would either tentatively vary the procedure or copy
a ritual detail from a neighbouring clan or tribe. More probably

1 This attitude differs from that adopted towards a person who breaks a
particular norm for obvious purposes; cf. J. H. Driberg, ‘Primitive Law in
Eastern Africa*, Africa t vol. i (192.8), p. 66.

204 AFRICAN POLITICAL SYSTEMS '

he would go to a diviner (< omufumu ), who, after consultation of
his various oracles, might suggest an entirely .unconventional
procedure which, if successful, would set a precedent and thus
be gradually adopted by the whole clan. The fact that some types
of sorcery which were formerly restricted to certain tribes are now
recognized and ‘detected’ by the diviners of other areas and
counteracted by new devices, shows that there are some loop¬
holes in the generally rigid convention.

A wider power to enact laws and customs and to induce people
to take an unprecedented course of action seems to have been
wielded by the dream prophet and by certain clan elders who had
gained a reputation as warriors and successful arbiters in disputes
and thus stood out as leaders among their age-mates. As far as
I could discover, no dream prophet (omung'oU) of any importance
is living in Kavirondo at present, but stories are told of men
who for years ahead predicted the coming of droughts, disastrous
raids, epidemics, and even the advent of the white man and of
the railway and motor car., When one of their predictions came
true, they gained, of course, in prestige, and their advice to
engage in unusual or refrain from customary actions was followed
by the whole tribe. A former prophet among the Vugusu is said
to have persuaded the whole tribe to migrate from its former
home in the Bugishu country to its present domicile. The Nyole,
a tribe that lives next to the Logoli, refrained from circumcision
for a few generations, paying heed to a curse uttered by a man
called Masava. His prohibition was observed for six age-classes
until, in the year 1917, the old men of several clans decided to
call the circumcision operator again. They performed a ceremony
at Masava’s grave, neutralizing the curse (xukavusia emirnoa ), and
at first circumcised a few people only to see if they died. They
survived and nowadays circumcision is again performed by the
majority of Nyole clans.

Thus although there is a strong resistance to breaking away
from traditional norms and although there is no governmental
organ which possesses recognized legislative authority, law
and custom are by no means entirely rigid. Changes were
brought about either by the initiative of a strong personality
whose word carried much influence, as he had gradually gained
confidence by the display of courage and wisdom, or by the
example set by any ordinary tribesman if particular circumstances

THE BANTU OF KAVIRONDO

205

or successive events justified his action. It appears that diviners

and dream prophets whose advice was supposed to be the outcome
of supernatural inspiration had more influence over their tribesmen
than ordinary people.

(2) The Continuity of Law and Custom . Law and custom which
in their totality make up tribal culture are not merely an inventory

of rules of conduct, but a coherent system of relationships between
individuals and groups. These relationships do not merely entail
the observance of certain actions and the avoidance of others, but
ideologies and values, mental and emotional attitudes as well.

Thus 'family law’, in the fuller sense of the word, comprises
the totality of relationships, as expressed in actions and atti¬
tudes, that knit the members of the family together into a social
unit, while the formulated 'laws’—such as regulate paternal
authority, the rights and duties of husband and wife, inheritance
and succession, &c.—demarcate the main lines and limits only
along and within which these relationships work. The mainten¬
ance of law and custom is thus equivalent to the maintenance of
effective relationships. 1

The continuity in time or perpetuation of these relationships
tends to be disrupted by two factors inherent in the conditions of
social life. One of these is that most relationships and the institu¬
tions of which they form part operate, not continuously, but at
certain occasions only. Between these occasions there may be long
intervals during which the relationship remains latent. This is the
more so the wider the group between the members of which a
particular relationship exists. Clan solidarity, for instance, comes
into operation only when challenged by the murder of a clan
member or some similar occasion, but the specific type of rela¬
tionship between the members of the clan on which this solidarity
is based has to be permanently maintained, so that the law of
solidarity may come into action whenever the need for its realiza¬
tion arises.

The other potentially disruptive factor is the coming and going
of the generations. Matrimony, parenthood, kindred, clanship,
&c., are permanent relationship patterns or institutions, but they

1 cf. B. Malinowski, Introduction to H. I. Hogbin’s Law and Order in
Polynesia, pp. irec-xxxv. The theoretical approach in this present study of the
political organization of the "Bantu Kavirondo 5 has been greatly stimulated by
this and other writings of Malinowski (e.g. Crime and Custom in Savage Society ),
even where this is not particularly acknowledged.

2 o6 AFRICAN POLITICAL SYSTEMS

derive their reality only from the fact that they are actually lived
by human beings. The fact that individuals grow old, die, and are
replaced by others involves the need for the transmission of law
and custom, for the constant re-knitting of institutionalized rela¬
tionships in view of the changing personnel.

In advanced and especially in literate communities the con¬
tinuity of law and custom in the face of these two disruptive factors
is attained by a complex system of legal, scientific, and educational
institutions. In a primitive and illiterate community there are
no such distinct institutions. There is no codification of law
and very little education in the sense of an organized imparting
of knowledge and moral values. We have to examine, therefore,
by what other means the need for a continuity of law and custom
is satisfied and how these means are embodied in tribal life.

(a) The Perpetuation of Relationships over Periods during which
They are Inoperative. An analysis of the various cultural institu¬
tions and the behaviour of the individuals who partake in them
reveals a number of devices which serve to maintain the effective¬
ness of relationships, rights, and obligations over periods during
which they do not come into play. These devices take, as we shall
see, mutatis mutandis, the place of codification of the law in more
differentiated cultures.

The most general way of keeping a relationship alive consists in
the exchange of gifts and visits between the persons concerned and
in the participation in common feasts by all persons who form a
social group for some purpose or other. The same motives, if
might be said, underlie hospitality and gift-making in any society.
This is true, but the much greater formality and regularity of such
observances in a society like that of Kavirondo than, e.g. in a
modem European community, indicates that they serve this
purpose of keeping relationships effective far more definitely and
exclusively. Formal visits are clearly distinguished from mere
sociability. The visitor announces his intended visit beforehand,
and the host instructs his wife to prepare a proper meal while he
himself looks for an appropriate gift which his visitor can take
home. A person exchanges such formal visits at more or less
regular intervals with members of his maternal bn (especially his
maternal uncle), with his wife’s brothers and sisters’ husbands,
and with those members, of his paternal kin which belong to the
same ‘gate’ or lineage. These are precisely those persons to whom

THE BANTU OF KAVIRGNDO

207

he can turn for economic support, for help in a quarrel or dispute,

for a share in garden land, or on whose goodwill he depends to
conduct his marriage successfully. The frequency and formality of
the visits are graded according to the type of the relationship and
depend, to some extent, on the personal factor. But even now¬
adays, where new conditions have loosened traditional behaviour
considerably, the formal exchange of visits is still observed with a
regularity that allows only little room for personal likes and dislikes.
The refraining from mutual visits, on the other hand, is equivalent
to the absence of an effective social relationship. When the laws of
exogamy were stated to me, the persons that may intermarry were
frequently defined as ‘the people who do not visit one another 1 .
This absence or cessation of mutual visits is taken as a clear indica¬
tion of the absence of any social bond which might come into
conflict with the establishment of the marriage bond. People who
have had a serious quarrel break off their mutual relations by
strictly avoiding common participation in a dance, meat-feast or
beer-feast, even if they meet accidentally at a third man’s place. If
their quarrel has been settled, the relationship is resumed again
by a ceremonial exchange of visits, accompanied by certain ritual
observances. The same attitude prevails between a newly married
man or woman and their respective parents-in-law. The initial
avoidance between them is not personal, but extends to their
respective houses. After the birth of the first or second child, they
terminate the avoidance ceremonially by paying formal visits to
each others’ houses.

The exchange of gifts fulfils'the same purpose and is usually
linked with the exchange of visits, although it is here more diffi¬
cult to distinguish between the exchange of gifts as a means of
maintaining a relationship and as the fulfilment of that relationship.
Smaller gifts, such as accompany the ordinary exchange of visits,
clearly belong into the first category. They are ‘real’ gifts in the
sense that they are given voluntarily to a measure and that reci¬
procity is not strictly observed and checked up.

The larger gifts—of stock or grain—which are exchanged at
definite occasions between definite categories of persons are rather
mutual obligations than gifts, as they are not voluntary but strictly
reciprocal. In case of refusal, the gift is either fetched by force or
the relationship ceases to exist, as the reciprocal gift will, of course,
be likewise refused. As, however, years may legitimately pass

208 AFRICAN -POLITICAL SYSTEMS

between gift and countergift, and as the occasions at which the gifts
become due are usually such that the recipient can make good use
of them, they are not merely means of strengthening a relationship,
but they are also ends in themselves. The persons between whom
such mutual gift obligations exist are primarily the closer relatives
within the paternal kin-group, maternal uncle and nephew, wife's
brother and sister's husband, and circumcision friends.

Common feasts, finally, serve to maintain the feeling of unity
within the clan and age-group and the bonds that exist between two
clans. The obligation to give such feasts rests primarily with the
old men of the clan, who are expected to kill an ox for the benefit
of their clansmen and for chosen representatives of other clans
whenever they are in a position to do so. If they neglect this duty
persistently, they lose standing among their clansmen and, in
extreme cases, are publicly ridiculed by young men, who, on certain
occasions, climb the roofs of the huts and shout remarks of abuse
or who sing songs of mockery and derision at beer-feasts and
dances which quickly spread through the country. The animal
slaughtered for a common feast is known as the ‘ox of splitting’,
and the purpose of the feast is explicitly stated to be a demonstra¬
tion of the clan’s strength and unity. The killing of the ox and the
distribution of the meat do not take place at the homestead of the
person who has supplied the animal, but on the public place, the
oluhia . Each clansman may attend and, although strangers and
especially children who happen to pass by are all gjiven a share, the
bulk of the meat is divided among the clansmen and those persons
from' other clans who have been told to come or to whom the meat
is sent in recognition of previous hospitality enjoyed at their place.
While the maternal kin and the in-law relations of the owner of the
ex receive the largest share among the non-clansmen, it is signi¬
ficant that the distribution of meat is not restricted to relatives, but
extended to the old and influential men of neighbouring .clans.
They represent their respective groups and in turn distribute the
meat received among their own clansmen, who jealously watch that
they receive their share in due course. On some occasions the
animal slaughtered for a clan feast is not taken from the herd of an
individual clansman, but from the spoils of war or from the com¬
pensation received by a group of clansmen on behalf of the whole
clan for the death or injury of a clan member. In the case of the
circumcision feast, the ox that i* killed and distributed among all

THE BANTU OF KAVIRONDO

209

the initiates on the "day of coming out of the hut of seclusion* is
taken from an old man’s herd, but it is secretly selected and then
abducted by the initiates and thus becomes a "public animal*, over
the distribution of which the former owner has no say.

Beer-feasts on a large scale, which, as a rule, take place separately
from beef-feasts, likewise do not serve the purposes of individual
aggrandizement and sociability alone, but of maintaining relation¬
ships between the clans and the sub-groups of each clan, and of
promoting unity within these groups. After each harvesting
season, the elders of one olukia decide to hold a large beer-feast
and choose from among them a man at whose place the beer
will be brewed and drank. All people of the neighbourhood
contribute basketfuls of grain, each according to his means, till
the necessary amount has been collected together. When the
beer is ready for consumption, all who have contributed grain
assemble at the chosen place to join in the common drink. The
beer is shared with others who have not contributed any grain,
but the different people who take part in the feast are kept in
clearly distinct categories: The "owners* occupy the favourite
seats and get the best quality and the largest quantity of the beer
brewed. The "helpers’, who have given no grain, but have lent
a hand in the preparations, are assigned a place for themselves;
the "beggars 1 , poor old men too weak to help, are also given their
own place, and the "servants* of the "owners* may drink the re¬
maining beer ne^t day and sing songs of praise to their masters.
Besides, each ‘owner* may bring his friend or relative along with
him; he will then in turn be invited to a similar beer-feast in his
friend’s olukia.

The particular way in which a feast is organized serves as an
incentive for the display of a social attitude and a co-operative
spirit, as the individual members of the group participate in the
feast exactly to the extent to which they have contributed their
share of grain or labour. The feeling of unity in the group is thus
strengthened, and at the same time a safeguard is provided against
exploitation by parasitic elements.

The social groups which maintain and strengthen their unity
by such common feasts are the sub-clans (< dzimbia ), the main
clans, the age-grades and, in connexion with circumcision, also
the tribal society. Whereas beer- and beef-feasts by the clan or
sub-dan are given whenever an ox or sufficient grain are available.

310

AFRICAN POLITICAL SYSTEMS

distributions of meat between the members of one age-class
take place on the occasion of their sons 5 circumcision only.
The beef is not eaten jointly, but the boy’s .father kills an ox
or two and sends meat to all those of his age-mates (vagogi)
whom he knows personally, while others may come or send their
wives to ask for a share of meat on the strength of the fact that they
were circumcised in the same year as he. As "circumcision takes'
place every few years and as all the members of one age-class have
to slaughter an animal after each son’s circumcision, the occasion
of stressing the age-grade relationship by the distribution of,meat
recurs frequently and for each age-class extends over a large
number of years. Circumcision also provides the occasion for
feasts on a tribal scale. Among the majority of the Bantu Kavir-
ondo tribes, both the performance of the operation and the various
feasts that terminate the convalescence and seclusion of the
initiates are the occasion for the gathering of thousands of people
comprising many different clans. Although the entire tribe does
not actually assemble at one place, the various sub-groups cele¬
brate the different phases of the feast on the same days and in the
same manner .

Other feasts on a tribal or at least on an inter-clan scale were
wrestling matches (now almost entirely replaced by football games)
in which the best wrestlers of one clan fought against those of
another, the procedure being hedged round with rules very much
like those connected with European sporting events.

Religious observances that involve the assembly of people on
a tribal scale I have been able to record from the Logoli only. In
a cave on the slope of a wooded lull they perform a semi-annual
sacrifice to an ancestral spirit (Mung’oma) and to an apparently
vaguely conceived tribal deity (Asai). The purpose of the sacrifice
is to evoke an ancestral and divine blessing for the quick ripening
of the crops or, on other occasions, to pray for help when a calamity,
such as a drought, an epidemic, or a series of ill-fated raids,
endanger the whole tribal society. The ceremonial rules on these
occasions, such as refraining from any kind of garden work,
apply to all tribesmen, and members of all clans join in the com¬
mon singing of ceremonial songs feukelemana) at the bottom of
the hill, while the sacrificers, together with their ritual assistants,
perform the sacrificial ceremonies in the cave, to which only they
have access.

THE BANTU OF KAVIRONDO

211

Tributes to the ram-maker are rendered on a tribal scale by
;he Logoli, who do not have a rain-maker in their own tribe, but
lepend upon the goodwill of a powerful rain-maker living among
i neighbouring and hostile people, the Nyole. The decision to
collect offerings of gram and animals from the people is made by
:he elders of the different clans, who set a day (either in a general
meeting or by communication through messengers), when a
ielegation of the principal dan elders will go on an expedition
;o the rain-maker to implore Mm to send rain and to negotiate the
imount of the tribute which they will have to pay.

The exchange of visits and gifts and the participation in common
feasts and ceremonial observances serve to maintain relationships
and thus law and custom in a general way. But the absence of a
codification of the law and especially of distinct legal authorities
cr bodies which could safeguard the validity of an agreement or
claim, creates the necessity for some corresponding arrangements
by which rules and agreements, concluded between particular
groups or individuals, can be maintained valid. This need is met
in a variety of ways. Whenever a dispute arises which is not
settled between the immediate partners concerned, the case is
discussed by the elders of the sub-clan at great length. Previous
cases which have a bearing on the case under dispute are recalled
and the settlement which was then reached is restated. It is
significant that the restatement of previous cases is not limited to
those which directly bear on the present one, but that it usually
embraces a much wider range of cases. Each legal dispute—pro¬
vided that it is complex enough to offer scope for a difference of
opinions—thus furnishes an occasion for recalling the juridical
traditions of the tribe. Thus what to the casual observer appears
as a straying away from the point and an indulgence in telling
‘irrelevant’ stories, actually serves the very important purpose of
keeping known the body of traditional law.

Similarly, agreements between particular individuals are kept
valid by their public restatement on occasions where people are
assembled who may later be called upon as witnesses. Every
economic transaction, such as the sale of a cow or the division of a
garden, to begin with, takes place in the presence of witnesses,
who are invited to a beer-feast which marks the transaction.
Later on, when the non-fulfilment of some other obligation is
discussed before the elders or when property is re-distributed

212

AFRICAN POLITICAL SYSTEMS

after somebody's death* various people make use of the occasion
to restate their own claims* not because they want to realize them,
but because they want to keep them generally known. The elders
then nod in agreement and the validity of the claim is thereby ■
publicly recognized. The surest way in which a liar is detected is
by his inability to provide witnesses who can vouch for his claim.

(b) The Transmission of Law and Custom to Succeeding Generations .
The second factor that disrupts the continuity of relationships
—the coming and going of the generations—raises the problem
of transmission of law and custom to succeeding generations. This
problem has two aspects: It involves, on the one hand, the impart¬
ing of practical and theoretical knowledge, of ethical and moral
standards of behaviour, and of general rules of etiquette which
are common to the whole tribe or even to wider groups. On the
other hand, It involves the initiation into the successive phases of
life, i.e. the acceptance of the individual into different and ever-
widening social groups or types of relationship, each of which is
governed by its own set of customs, rules, and values.

The first, aspect of the task of transmission, the imparting of
general knowledge, values, and manners, chiefly’takes the form
of a general education through example and precept which Is
accomplished without organized effort by the up-bringing of the
child in the family and its adjustment to its everyday surroundings.
It does not concern us in this analysis of the political organization.

The second aspect, however, requires a closer examination as
it demonstrates the process by which the individual gradually
gains his place in the tribal structure. As every individual enters a
new phase of life and thereby attains a new status, the rights and
duties and the new types of relationships, implied in his new
status, are marked by a ceremonial initiation. If we follow up the
life-cycle of the individual among the Bantu Kavirondo, we can
distinguish six major phases, the entrance Into all of which is
marked by a very similar ceremonial procedure: (a) Earliest
infancy during which the child is not yet socially acclaimed, (b)
Later infancy and early childhood, the entrance into which is
marked by the ‘feast of washing the child' and the name-giving
ceremonies by which the child is acclaimed as a member of the
individual family and the father's clan, (c) Boy- and girlhood,
marked by the teeth-knocking rite and the child's formal accept¬
ance into its maternal kin-group as well as its admission to the

the principle of seniority operates in all relationships*

initiated into membership of an existing group, a common feast

by a close paternal or maternal relative, in the course of which
he is instructed in the standards of conduct which henceforth are
expected from him. While these commandments are given, the
initiate observes definite rules of ritual, and the person who speaks
the commandments drives home each sentence by spitting beer

214

AFRICAN POLITICAL SYSTEMS

on face and body of the initiate. In the case of infants and small
children, e.g. at the naming ceremony, the place of the command¬
ments is taken by the uttering of words of well-wishing over the

child (okugasidzd). (e) In connexion with the exchange of gifts
or the common feast the goodwill of the ancestors is invoked by
the offering of meat and blood to the spirits and the utterance of
a prayer in which they are asked to come and partake of the food
that has been offered to them.

We see from this survey of initiation rites that the task of trans¬
mitting tribal law and custom from generation to generation is
not performed by distinct institutions, controlled by a central
authority, but that the different social groups and relationship
patterns continue themselves by handing down their own systems
of values and standards of conduct through the formal initiation
of new individuals into them. The different groups and relation¬
ships, however, do not exist side by side, detached from one
another, but they overlap in various respects. In the first place,
as they are based partly on kinship, partly on marriage, partly on
age, and partly on common family status, every individual belongs
to several groupings. A person is not only a member of his lineage
and clan, but at the same time of his age-grade; he entertains well-
defined relationships with his maternal kin as well as with his
wife’s kin and that of his brothers’ wives and his sisters’ husbands,
and he shares the common status of a married man or a father
with many other members of the tribe, irrespective of clan or
marriage bonds. This overlapping of the different groups and
relationships, as regards their personnel, clearly acts as a force
that maintains and promotes the feeling of tribal unity and the
homogeneity of law and custom and that counteracts the tendency
towards rivalry and competition between clans.

In the second place, effective kinship bonds and membership
of various groups increase in number and importance as the
individual gets older. Each successive phase in life means, there¬
fore, a rise in status, i.e. an increase in rights and privileges, but
also, of course, in duties and obligations. This fact, along with
the principle of seniority as observed in the family relationships
with regard to the holding and transferring of property, likewise
tends to integrate the relations between the groups, as it places
the greatest authority into the hands of the old men who, by virtue
of the wide net of kinship bonds and group affiliations in which

THE BANTU OF KAVIRONDO 215

they are enmeshed, are best fitted to overcome group interests
and jealousies and thus to plead for the unity of the tribe.

(3) The Restoration of Breaches of the Law . Before discussing

the seat or seats of judicial authority in tribal life and the manner
in which it works, we have to examine the different types of

breaches or nonconformity that are distinguished and the nature
of the restoration that is aimed at. From the linguistic evidence, it
appears that offences are and were classed into four groups:
(i) amagovi, (2) amagoso (or amahiolo ), (3) amatava, and (4) endgilu .
A person commits an eligom 1 when he refuses to pay a debt or to
fulfil a customary obligation, such as may be demanded by the rules
of kinship or that may result from partnership in cattle or some
other possession. The term eligoso denotes a range of offences
that in Europe would be classed as both civil and criminal,
such as adultery (by or with a married woman), theft, assault,
arson, &e. The word elihioh is often used with the same meaning,
but seems to imply that the offence is of a more serious nature,
such as rape, murder, or witchcraft. It nowadays has the connota¬
tion of a sinful and morally strongly condemnable act, but it is
uncertain to what extent this connotation is due to mission or
other recent influences. Amatava are offences against property
or life which have been committed accidentally or at least without
the full intention and responsibility of the offender, such as
physical injuries inflicted by carelessness in handling weapons,
or the accidental destruction of a neighbour’s house by fire or of
his crops by cattle. Endgilu y finally, are violations of important
taboos or rules of ceremonial conduct, pre-eminently of such rules
as the prohibition of incest and the avoidance of one’s mother-in-
law or the desecration of objects used in the ancestor cult.

The distinction between these different types of offences and
recognition of a particular form of conduct as constituting an
offence is common to the whole tribal society, with the exception
of a few rales that are observed by some clans but not by others.
The fact that the body of rules and customs that demand con¬
formity is, generally speaking, valid in the whole tribe does not,
of course, entail the existence of a tribal judicial authority.

What does the distinction of these four types of ‘breaches’" of
the law mean from the point of view of the restoration of the
breach? In the case of amagovi , no actual wrong has been done

1 The class-prefix eli- denotes the singular, ama- the plural.

AFRICAN POLITICAL SYSTEMS

which needs to be undone* The fact that a claim has becoi
overdue does not involve the notion that the creditor is entitl
to compensation for the delay suffered in materializing his claii
The dispute merely has to be, decided in favour of one of t
disputing parties. When amagoso or amatava have been cor
nutted, the offence has to be undone by inducing or forcing !
accused person to compensate the damage he has caused to ti
accuser. If the offence was unintentional (elitava), he merely h
to restore the damage or pay cattle equivalent in value, while
the case of an intentional offence (eligoso) the double amount h
to be replaced or a fine in cattle is levied'that is considered amp
compensation. This double compensation was interpreted by n
informants as aiming to restore the goodwill of the wronged perse
towards the offender and not as a fine in the sense of a pen
sanction. This interpretation is corroborated by the fact that aft
the settlement of an eligoso or elitava the plaintiff, if he has wc
the case, is expected to make a counter-payment, or rather gif
to the defendant as a sign of his satisfaction and reconciliatic
with him. Breaches of taboo (emigilu), finally, are restored l
performing the appropriate sacrifice or purification ceremon
after which social relations are resumed with the offender by h
fellow beings, who have avoided him in the meantime. If tl
broken taboo involves another person—as in the case of a violatic
of the mother-in-law taboo—the purification rite is usual]
followed by the participation of both persons in a common mea
It appears, from this brief survey of the types of breach*
recognized and the methods employed in dealing with them thi
the restoration of law and order aims at the settlement of clain
and the reparation of damages rather than at the punishment c
the offender. This fact has two important consequences. Th
one ^ is that jurisdiction takes place only when solicited by th
victim or victims of the offence, as all offences are conceived a
being injurious to the interests of a particular person or group c
persons,, but not to the tribal society as a whole. It is a logic?
implication of this conception that, not only every material objec
but also every human being has its ‘owner’ or ‘owners’. 1 Atypica

1 The LogoH term ovwene refers both to ‘ownership’ of objects and <
persons. It differs, of course, from our concept of ownership in several respect
Ihus the ormvene (owner) of a person is the one who has not only the foremo*
rights over him but also the foremost obligations towards him.

the individual immediately affected by a breach of the law,

2x8 AFRICAN POLITICAL SYSTEMS

clan or any of its sub-groups of which the individual is a member
and the specific interests of which are affected, in both cases that
group acts, in a sense, as the accuser and as the judicial authority
at the same time.

A distinction, however, must be made between law administered
within a given group and between different.groups. When a legal
dispute or an offence involves two clans, the clan of the wronged
person tries to obtain justice by negotiation with the clan of the
offender, which, in turn, stands behind the offender and either
rejects the claim or assumes responsibility for his action. When,
however, a breach of law occurs within a given group and the
nature of the offence is such that it affects those interests which are
specific to the group and not further divisible, a restoration of the
breach by compensation is obviously impossible, as it would neces¬
sitate the splitting up of the group into two units, the one which
gives and the other which receives the compensation, an action
which would destroy the solidarity of the group. Thus, if a person
commits adultery with one of his father’s or brother’s wives, the
usual compensation of a heifer is not paid, as a father and his sons
form a property-holding unit in which the payment of compensa¬
tion by one member to another member would be pointless. If a
person kills a member of his sub-clan, no compensation would be
paid either (among the Vugusu), as they say that the loss of life
affects the whole sub-clan and not merely the immediate kinsmen.
The action taken in such and similar cases furnishes a clear cri¬
terion of the nature of the solidarity that prevails within the
group in question. Where it is deemed that no legal action (i.e. the
imposition of a compensation) can be taken owing to the indi¬
visibility of the common interests of the group, merely a sacrifice
is performed to propitiate the spirits and a purification ceremony
which frees the offender from his ritual impurity and renders it
safe for his relatives and neighbours to resume social relations
with him.

In the case of repeated offences, the only possible procedure is
to expel the offender from the group and to withdraw from him the
right for protection by the clan as well as the clan’s responsibility
for his deeds. The attitude towards an habitual offender thus
differs fundamentally from that towards an occasional offender.
Whereas the latter—no matter how serious the offence committed
by him—is considered to have acted within a set of particular

THE BANTU OF KAVIRONDO

219

conditions and circumstances and against the interests only of the
people directly or indirectly affected by the offence, the habitual
offender becomes a source of danger to everybody in the tribe. He
is, therefore, placed outside legal protection of the clan and may be
killed by anybody when he is caught in the act of committing his
next offence. Thus persons who have come to be regarded as
dangerous witches or incorrigible thieves are first driven away from
their clan and then, at the next provocation, put to death by the
method of lynching carried out by as large and mixed a group of
persons as possible to avert the possibility of a blood feud from
arising. Such group action in the face of threatening danger, taken,
spontaneously, i.e. without a hearing of the case and often on the
spur of the moment, is clearly not the same as institutionalized
jurisdiction by the tribal society through recognized judicial
authorities. It is rather that in such cases the person of the accuser
becomes multiplied and that the tribal group by being accuser and
public opinion at the same time cuts short the usual judicial pro¬
cedure. It will be seen, therefore, that the occurrence of indis¬
criminate group-action in the face of dangerous witchcraft and
habitual crime does not invalidate the basic principle of jurisdic¬
tion, viz. that it comes into force only when solicited by the victim
or victims of the offence and takes place only within and between
those groups of persons whose common interests are affected by
the offence. 1

A brief discussion of judicial procedure will show how this
principle of group solidarity along clan lines works in practice,
comprising ever-widening groups as the significance of the
offence widens. In the initial stages of each dispute over a material

1 A similar view is taken by J. H. Driberg in his paper cited on p. 203
supra. Driberg, however, constructs a difference between two distinct
categories of offences, those against individuals (or individuals represent¬
ing a family, community, or association), and those directed against or affecting
the whole ‘body politic*. With the first category he classes such offences as
homicide, theft, adultery, injuries, slander, &c., and with the second one—
which he calls anti-social—witchcraft, incest, and sexual perversion. With
reference to Kavirondo law this classification would not apply, as (1) from the
point of view of their social repercussion, offences show a wide range of shades,
affecting ever-widening groups as they increase in seriousness, and not the two
distinct types which he labels ‘anti-individual* and ‘anti-social*, and (2) the
amount of social disruption caused by an offence does not depend on the kind of
offence (whether murder or witchcraft), but upon the particular conditions
under which it has been committed, one of them being the previous record of
the offender, another one the motives underlying his action.

Z 20

AFRICAN* POLITICAL SYSTEMS

object or the non-fulfilment of an obligation an attempt is made to
reach an agreement between the two immediate parties concerned.
If a debt is not returned or the obligation not fulfilled, the claimant
at first takes resort to self-help, fetching the object in question—
usually an animal— by night, assisted perhaps by one or two
brothers or friends. If a neighbour encroaches upon his garden,
lie goes there at night or while the neighbour is away and re¬
marks the boundary-line by digging new ditches or by pulling out
Ms neighbour's seedlings wrongfully planted on Ms land. TMs
form of self-help is still frequently resorted to, especially among
the illiterate section of the community, and gives rise to much
disciplinary action by the elders of the new tribal courts, who are
anxious to settle every dispute in court for fear of losing their
court dues. In other cases, where self-help of tMs nature is im¬
practicable or impossible, the aggrieved person takes recourse to
the spelling of a curse or to engaging the services of a rain-maker to
obtain justice. Curses are mainly employed by older people, as
their efficacy is supposed to increase with the age of the person who
utters- them, or in cases where the offender is unknown or where
the evidence against Mm is not conclusive. I have myself witnessed
a few cases where stolen property was secretly returned within a
few weeks after a curse had been uttered. Threats to engage the
services of the rain-maker to divert the rain from the offender's
garden or to devastate it by directing a hailstorm towards it are
said among the Vugusu to have been the most common means of
pressing the payment of a debt.

When disputes or quarrels could not be settled by self-help, the
person who believed himself wronged appealed to the old men of
Ms sub-clan, and the accused person, if’he belonged to the same

sub-clan, was called by them or he came on Ms own account to
defend Ms case. The old men then listened to the case as pre¬
sented by the two disputants and any witness. The decision could
be announced by any of the elders present as, with the facts ascer-
tained, there was only one possible judgement wMch was common
knowledge to all. Nor was there any organized judicial assembly.
The elders of the sub-clan met every morning on a pasture,
where, sitting round the fire (oluhia), they discussed the news and
the gossip of the previous day. These informal gatherings provided
the main occasion for dispensing justice within the sub-clan. If
the evidence could not be established by hearing the two parties

221

THE BANTU OF KAVIRONDO

and the witnesses, an ordeal was administered in the presence of
the elders of the olvhia. No further action was then required, .as
the ordeal was supposed to administer justice automatically.

An alternative to the appeal to the elders for arbitration was
the continuation of mutual provocations between the disputing
parties until a fight ensued. It was then the duty of the
elders to intervene by separating the fighters and by per¬
suading them to return to reason and settle their dispute peace¬
fully. The ability of certain men in the sub-clan to carry through
such intervention successfully was one of the main requirements
for political leadership. ‘The head of the clan [omugasa] is the
man who talks gently and who can make the people listen to him
when they quarrel or fight’ is the usual definition given of a tradi¬
tional ‘chief.

If the dispute or offence was of a more serious nature, if it in¬
volved serious injury or even the death of a person, the news of it
would quickly spread to all clansmen, and the elders of the dif¬
ferent sub-clans would rush to the scene of the offence and hold a
meeting ad hoc (eMnmzo) to prevent fighting and discuss the
situation and finally give judgement. In such assemblies of the
clan council (i.e. of the elders of all sub-clans) there was again no
hereditary or formally appointed ‘chief judge’' or leading sub-clan,
but strong personalities who Had gained prestige as warriors and
givers of feasts were recognized as authorities whose opinion
carried more weight than that of the ordinary elders.

The only sanction which supported the legal decisions given
by the elders of the sub-clan or by the larger ad hoc assembly of
all clan elders was the solidarity of the members of the respective
groups in backing these decisions. Native statements assert
that whenever the verdict was supported by all or the great
majority of the clan elders, the defendant would not have tried
to oppose. Were he to have done so without finding any support
among a section of his clansmen, the verdict would have been
enforced by the elders by appointing a number of men who
would take the cattle, or whatever the compensation would
be, by force. Besides, the fact that a person who evaded justice
in his own clan could only with difficulty migrate and settle in
another clan, because he would not have been hospitably received
if the reason of his secession became known, would force him to
submit to the decision of the elders.

222

AFRICAN POLITICAL SYSTEMS

A different situation arose when a dispute between two
members of the same clan led to a dissension among the clan
elders, each backed by a section of the clan, and neither section
was willing to yield to the other. In such a case, the weaker
section w r ould secede from the clan land and join another clan or
migrate elsewhere, and a new clan would thus come into being.
The origin of many present clans is accounted for by such quarrels.
The main clan and the section that has split off would at first avoid
all social relations with one another and also continue to observe
the rule of exogamy . 1 After a few generations, when the quarrel
has been forgotten, a member of the seceded sub-clan would
tentatively marry an omwiwana (a niece) of the main clan, i.e.
a girl whose mother comes from that clan. If this marriage pro¬
duces offspring and the children survive, the two clans begin to
intermarry directly. Thus the independence of the seceded clan
becomes fully established and the original clan name, which at
first is maintained along with the name of the man under whose
leadership the clan group seceded, is dropped.

But the judicial functions of the clan were not limited to the
restoration of law and order within its own ranks. They also
comprised the settling of disputes that arose between members
of different clans. In such a case, the wronged person and his
immediate kinsmen would go to the elders of the defendant’s
clan. If the case was serious enough to affect the common interests
of the whole clan, the plaintiff was supported by all the elders of
his clan and a number of warriors, who would accompany him
in his search for legal satisfaction to demonstrate to the defendant
and his clan that they backed his claim. This was deemed to be
the case when the life of a clan-member had been taken or seriously
threatened by assault or attempted sorcery, or when a quarrel over
property had reached the point where peaceful negotiations
between the immediate parties concerned had come to an impasse
and the dispute threatened to lead or actually had led to fighting.

If no agreement could be reached between the two clans, which
would happen when one side demanded exorbitant compensation
or when the inter-clan relations had been strained by a series of

1 In this case the observance of exogamy is based partly on the consciousness
of still being related and partly on the existence of a feud between the two clans.
Thus, close co-operation and its opposite, a state of hostility, both act as a bar
to intermarriage.

THE BANTU OF KAVIRONDO 223

previous disputes, the two dissentient elans would break off their
relations with one another and enter into a state of vendetta. This
continued until the warriors of the aggrieved clan had taken a life
equivalent in status or until several lives had been taken on both
sides and the equilibrium had thus been restored. It w*as then
the task of the elders in each clan to work for reconciliation by
lamenting the deplorable consequences of the feud and by appeal¬
ing to the former neighbourliness and the common ancestry of
the two hostile clans in their talks with the young men of the tribe.
If both clans were willing to terminate the avoidance a feast of
reconciliation was arranged (okuhololizana) which entailed a
common meal and sacrifice, and the former relations between the
two clans were resumed again.

Thus the restoration of law and order, when infringed by a
member of another clan, was ultimately achieved by a showdown of
force between the two clans concerned. To render this effective,
the clan had to have a high degree of solidarity within and at the
same time had to be of sufficiently large size. As, however, owing
to the nature of the bonds that make up clan solidarity, an increase
in the number of clansmen beyond a certain point renders the
occasions for co-operation between them too rare and too vague,
there must have been an optimum size for a clan. This was reached
when a fair balance existed between its external power, as expressed
in the number of warriors, and its internal strength, as expressed
by the degree and frequency of co-operation between its members.
The need for this balance explains why young clans which were
still small in number sought affiliation with larger clans in the
form of ceremonially confirmed clan friendships and alliances and
why, on the other hand, large clans tended to split up into sub¬
clans which gradually become independent of one another.

As regards the restoration of breaches of the law the clans were
thus sovereign groups, as there was no tribal judicial authority
which could be appealed to in the case of inter-clan conflicts. The
fact that numerous bonds of kinship and marriage existed between.
the members of all clans in the tribe and that strife between clans
weakened tribal co-operation in warfare, served as an inducement
for the elders of clans, not directly involved in a clan-feud, to
intervene as arbiters, but there was no legally binding force behind
such arbitration.

224

AFRICAN POLITICAL SYSTEMS

IV. External Political Structure

Before discussing warfare as the predominant type of political
relation between the tribes, I shall briefly review other types of
mter-tnbal relations, as they lead up to the formulation of the
problem of warfare as a political problem.

Trade relations between the tribes were only weakly developed
as natural resources were fairly evenly distributed over the whole
of Kavirondo and technical skill and knowledge were of such
similarity m all the Bantu tribes that there was no incentive for
a regular and organized exchange of goods. The occasional
^enng of crops for live stock and of the products of certain
crafts (pottery, iron-work, and ornaments) was too erratic to lead
to the establishment of permanent political relations between the
tnoes on economic grounds.

Individuals who had gained repute as rain-makers, circumcision
operators, diviners, or herbalists were consulted by clients from
neighbouring and even hostile tribes and also called to other tribes
tioSf ^ ^ Servkes - Thus the Logoli sent frequent expedi-

TJS Ny C f m ' m f ker whose P° wers were acknowledged
by half a dozen other tribes. Such relations, however, are said
to have been maintained during periods of truce only and, besides
r ° f S , UCh a nature -tkat was in the interest of both tribes
oncemed not to interfere with them by committing hostile ac 2

V !f It0rS neighbouring tribes. Specialists were
r„nhl 7 m I s 0f , a hoStile tribe iDecause their services were

visitor n rr T u thC tribC 2nd they were therefore welcome
to fs. Clients who went to consult experts in another tribe
likewise, were immune from attack, as they brought gifts to pay
for the information that they received. I could obtain^o records
of any o<xasions at which such relations were abused nor how
ever, is there any evidence to show that they ever led to a’closer
political alliance between the tribes concerned.

Intermarriage between generally hostile tribes was limited to
rfsi^mf ° f j C who were taken aT^age

bad captared y them n m 0Pted ^ ** of ^ warrior who

^ grown up, they were married off in

*as ^or dZ ^ /^Shters. Capture of adult women

,, accordm g to information given by the old men. not at all
>nly exceptionally practised, as they would have arted as spies or

THE BANTU OF KAVIRONDO 22$

attempted to ran back to their own people at the first opportunity
for escape that offered itself. Regular intermarriage, based on
a mutual agreement between the parties concerned, took place
only between tribal groups who maintained friendly relations
based upon a recognition of a remote relationship, such as existed
between the Vugusu and Gishu, the Wanga and Marama, and
the Tadjoni and Nyala (Kabras). A one-sided incentive for
marrying women of a neighbouring tribe existed if it was poor
in cattle—especially after an epidemic had ravaged—and therefore
willing to marry off its ‘daughters’ for a low bride-wealth. How¬
ever, as far as genealogical records show, intertribal marriages
appear to have been exceptional between tribes that had marked
differences in custom, as these caused difficulties in the marital
relationship itself, as well as in the social relations between the
husband’s and the wife’s group. Even nowadays intertribal
marriages are disliked on these grounds, especially by the girl’s
kin and clan, and most of the cases noted down by me concern
teachers, clerks, or others whose occupation takes them away from
their own tribe. It seems, therefore, that intertribal marriages did
not constitute a strong factor in establishing social relations
between the tribes and in breaking down cultural differences
between them.

The attitude towards any neighbouring tribe as a whole was
chiefly characterized by a feeling of suspicion, to which was

added either fear or contempt. Fear of other tribes is never
openly admitted, either in personal talks, or in migratory accounts,
or in texts on warfare. It is, however, implied in the absence of
contempt and ridicule. The Uasin-Gishu Masai, who frequently
raided the country of the Vugusu without ever being raided in
return by them, are always spoken of in a solemn manner, and
I could not discover that any terms or songs of derision were
current with regard to them. On the contrary, many Masai terms
relating to the conduct of warfare, as well as some of their weapons -
and the attire of their warriors, have been copied by the Vugusu.
The Logoli refer to their chief enemies, the Nandi and Tiriki, as
avafumbwa (‘enemies who come by night’), a term which is
obviously intended to convey a feeling of horror.

Contempt and derision are far more common attitudes towards
other tribes and find expression in numerous sayings and pro¬
verbs. Wi thin the group of Bantu tribes, expressions of contempt

22 g AFRICAN POLITICAL SYSTEMS

centre chiefly round minor differences in custom: Foods eaten
or refused, peculiarities of dress or ornament, and mannerisms
such as particular ways of walking or talking. To tribes of
non-Bantu stock, despicable qualities and sinister Intentions
are usually attributed. To the Logoli,'the term avavo refers
to their westerly neighbours, the Luo, and has the connotation of
something evil and inferior. The Luo are ridiculed for differences
in their tribal customs. They are derided because they do not
practise circumcision, and cowardly boys, in a circumcision song,
are told to run off to the Luo country and marry there, as‘ the Luo
women are satisfied to marry uncircumcised men. Among the
Vugusu, an analogous attitude towards their non-Bantu neighbours
prevails. The El Kony are merely ridiculed for their weakness
in warfare, while numerous derogatoiy comments are passed on
the Teso who, next to the Masai, were the chief enemies of the
Vugusu.

With such relations and attitudes. prevailing between the
different tribal groups, what were the motives for warfare ? From
accounts of war expeditions and an analysis of all the accom¬
panying circumstances, it appears that the two immediate aims
in attacking other tribal groups were to capture cattle and to
inflict a loss of life upon the group attacked. While the first
of these two motives is clearly economic and requires no further
interpretation once the significance of cattle in the tribal economic
organization is known, the second motive cannot be considered an
end in itself, as it leads to the further question: ‘Why do people
want to inflict a loss of life upon a neighbouring tribal group ? >

To a certain extent, the taking of life might be considered a
necessary consequence of the raiding of cattle, as the owners will
try to defend their cattle, so that violence naturally results. The
conduct of warfare by the ‘Bantu Kavirondo" shows, however, that
open encounters with the enemy were not avoided whenever
possible, but actively sought after. Apart from single-handed
cattle-stealing, they employed very little cunning in their raids,
but openly challenged the enemy once they had approached him.

In war and victory songs, the killing as such is praised as the
main purpose and achievement in a fight, and undoubtedly the
thrill or ‘sport" which warfare furnishes and the fame and prestige
which the display of courage and daring bestows upon a warrior was
a powerful motive and perhaps the strongest immediate incentive

THE BANTU OF KAVIRONDO 227

for waging war. But beyond this immediate motive the taking
of life served other purposes, which become apparent when the
whole organization of warfare and the groups that waged war with
one another are subjected to a closer analysis. Although between
most tribal groups in Kavirondo—especially those who lived in a
permanent state of hostilities—an uninhabited zone of a few miles'
width was maintained, the tribal territory did not remain static,
but either expanded or contracted as the result of extended periods
of warfare. Among the Logoli the conquest of new land for culti¬
vation is expressly stated to have been one of the chief motives for
warfare. It does not, however, become apparent as an immediate
motive, as a war expedition was never terminated by the annexa¬
tion of a given area by the victorious side and a readjustment of the
boundary line confirmed by the vanquished or any similar pro¬
cedure. This would have required a much firmer military organi¬
zation than existed and an organized protection of the borders, for
which the political structure of the tribal groups was much too
loose. The immediate result of a raid was rather to weaken and
intimidate the neighbouring tribe and to induce its members
gradually to retreat, so that the uninhabited zone would widen and
the grazing of stock and the cultivation of gardens could safely
be carried on in what was formerly no-man’s-land. The territory
thus gained by a very gradual process came under the control of
the clan whose warriors had driven the enemy tribe into retreat
and was shared out among them.

Whether such, a conquest of territory involved a real expansion
of the tribal area or whether the gaining of land on one side was
always accompanied by a loss on the other side is very difficult to
decide. As far as traditions go, it appears that changes in the
territory held by the various tribes were due partly to a general
tendency of an eastward migration caused by a pressure of the
Nilotic and Teso-speaking groups, the ultimate reason of which
would have to be traced back to the upper Nile Valley, and partly
to a real need for expansion. This need, again, arose from a variety
of causes. Of these the most important seem to have been (a) a
natural increase in population which, in view of the fecundity of
the Bantu people, must at times have been considerable even in
pre-European days; (b) an increase in the wealth of cattle, either
by natural increase or by conquest, requiring larger grazing areas;
and (c) the deterioration of the soil, owing to various forms of

228

AFRICAN POLITICAL SYSTEMS

erosion caused by shifting cultivation and excessive burning of

bush for grazing purposes. Expansion, based on these causes,
cannot, however, have heen continuous. The first two of these
causes were checked by epidemics and by adverse luck in war.
The third was limited by the checks on the increase in population
and live stock and also by the fact that, under traditional conditions
at least, the deterioration of the soil was not permanent but
temporary, so that periods of expansion were followed by periods
of retreat to lands formerly cultivated. The system of land tenure
indicates that neither the tribal group nor the clans attached value
to the possession of land apart from those stretches actually used
or reserved for cultivation and grazing.

The fact that the two ultimate motives in warfare were the raid¬
ing of cattle and the conquest of territory has a definite bearing
upon the conduct of warfare, as it involves conflicting aims. While
it lies in the interest of expansion to carry on aggression in a
ruthless manner which drives the enemy away as far as possible,
the aim of raiding cattle clearly requires the presence of enemy
groups in the neighbourhood. Owing to the necessity of balancing
these two aims, warfare tended to be conducted with certain
restrictions, above all with provisions for te rminating a period of
hostilities and with generally observed rules regarding the treat¬
ment of slain warriors and of women and children. Such ‘rales of
warfare’ were more pronounced in the conduct of hostilities
between the various Bantu groups than between Bantu and non-
Bantu. In the latter case, the mutual destruction of the hostile
groups was the prominent aim, while in the encounters between
groups of Bantu stock the hostile groups conceded their mutual
rights of existence and maintained a type of relationship with one
another in which warfare functioned chiefly as a regulating and
balancing force, making for an approximately even distribution of
power and wealth between the tribes.

Secondary motives of warfare, the relative importance of which
differed in the different areas, were the taking of captives and the
raiding of crops. The first was limited mainly to the taking of
small boys and girls between the ages of six and ten years, who
were adopted and brought up in the family of the warrior who had
captured them. As, under traditional conditions, children were of
economic value to the family, the adoption of war captives meant
a welcome addition to the family and the clan.

THE BANTU OF KAVIROXDO 229

The raiding of crops was customary only in the densely
populated areas and among predominantly agricultural tribes,
such as the Logoli and Nyole, while the more pastoral Yugusu

ridicule it as below the dignity of warriors.

It becomes apparent, therefore, that there was no need for a
tribal military organization, but that each dan or a group of
neighbouring clans would conduct their war expeditions, as well
as their defence against attacks, on their own account. Actually,
both for the Logoli and the Vugusu, there are no records of any
‘wars’ on a tribal scale, but only of fighting and raiding expeditions
undertaken by the different dans. Since, of course, only the
larger clans could venture to undertake raids and provoke an
open fight, as only they possessed a sufficient number of warriors,
the smaller clans either had to stay behind or to associate them¬
selves permanently or temporarily with a larger clan for co¬
operation in raiding. Warriors of other small clans accompanied
on their own initiative groups of warriors of larger clans, par¬
ticularly those with whom they were related in the maternal
line or by marriage or whose circumcision age-mates they were.
Where hostile tribes lived on all sides of the tribal territory,
as was the case both with the Logoli and the Vugusu, the clans
living nearest those sections of the border which were most
suitable for raids and attacks acquired leadership in warfare. Thus
among the Logoli the clan of the Mavi which fought against their
western neighbours, the Nyole and Luo, and the clans of the
Yonga and Tembuli, which fought against the Nandi and Tiriki in
the east, appear to have been the clans which were foremost in
taking the initiative in fighting. Till to-day they boast of more
famous warriors than the smaller clans.

Such leadership of certain clans in warfare did not, however,
necessarily entail any political domination over the smaller clans.
There was no ‘calling up’ of warriors, but participation in a raid
was voluntary and the spoils of war were divided among all warriors
according to the degree of their participation in the raid. The
initiative to embark upon a raid came either from the young men
or from the elders who, sitting on the oluhia, incited the young
men to go out and capture cattle, ‘as they had not seen meat for a
long time’; or it arose from an incident, such as the murder of a
tribesman by members of another tribe with whom peace had been
concluded. If the raid was likely to prove difficult and to require

230

AFRICAN POLITICAL SYSTEMS

a large number of warriors, messengers were sent round to the
villages or homesteads of other clans to invite them to come to a
given place where the further procedure was discussed and a leader
(omwemiliU) was chosen, whose duty it was to direct the movements
of the warriors and to co-ordinate the action of the different sec-
■ tions. All raids were one-day affairs or, if the attack was under¬
taken at night, as in the case of the Vugusu raids upon the Teso, of
a day and a night.

Apart from seasonal periods of truce, which, by tacit agreement
between all parties concerned, were observed during the times of
hoeing and clearing the fields, a more lasting peace was concluded
by the performance of complex peace ceremonies. The main rite
had the significance of an oath, and is supposed to cause death and
other misfortune to the party which first breaks the peace. The
conclusion of peace never involved ‘peace terms’ in the sense that
one party would pledge to pay tribute or surrender cattle pre¬
viously raided, or even territory. Peace, agreements are said to
have been made by the Logoli with the Nyole, Tiriki, and Luo,
and by the Vugusu with the El Kony, Nyala (Kabras), and Kaka-
lelwa, but never with their chief enemies, the Masai and Teso.

V. The Nature of Political Authority

Although, as has been stated at the beginning of this chapter,
there were no individuals or bodies which wielded clearly defined
political authority entailing explicit rights and duties, the preced¬
ing discussion of the various aspects of political organization has
indicated a number of ways in which individuals could gain promi¬
nence over their tribesmen or clansmen and find recognition as
leaders by certain groups within the tribal unit and with regard to
certain activities. We shall now review these different ways of
acquiring prominence and then try to define the nature of political
leadership as it existed among the Logoli and Vugusu.

(a) The Privileges of Primogeniture . As primogeniture carries
with it a number of privileges, there is a tendency in every family
for the oldest son to be recognized as the person next in importance
to the father. His authority is based on three factors mainly: the
first is that he is in a privileged position to acquire wealth in
cattle. Although ultimately all sons are entitled to an even share in
the father’s legacy, the oldest son has a preferential claim to make
use of family property, a fact which gives him a much quicker

THE BANTU OF KAVIRONDO 231

"start" In life than is offered to the other sons. He Is the first to be
given cattle to marry and, If the father is poor In cattle, the second
son may have to wait for many years until the father's herd has been
replenished before he can take a wife. The oldest son also has the
first claim to the father’s land, to the inheritance of his junior
wives, and to the management of the family cattle* The younger
sons have to wait till the cattle taken over by the oldest son have
Increased sufficiently to permit of an equal division, and it is
usually only in long-drawn-out Instalments that they can obtain
their share In the father’s legacy. It will be seen that the privileged
economic position of the oldest son tends to have a cumulative
effect—up to a point—as one generation succeeds the other. As
a consequence of this tendency, the line of first-born sons often
becomes the wealthiest in the lineage.

The second factor is that, through his management of the
father’s legacy in land and cattle, the oldest son exercises authority
over his younger brothers, who depend upon his friendship and
goodwill for the realization of their share in the father’s property
that is ultimately due to them.

A third factor is that family tradition and the knowledge of law
and custom and, in particular, of .outstanding claims to property
are always passed on from the father to the oldest son, so that
in the lineage group the senior line becomes the chief guardian
of tradition and its members the performers of rites and sacrifices
for the while lineage or even the sub-clan.

The stress on primogeniture, however, is not so marked that in
each clan there is necessarily one leading family, viz. the descen¬
dants in senior line ‘ of the founder of the clan, although such
families are found in a number of clans. In most clans there are
several elders who trace their descent in senior line back for eight
to ten generations, but they are not able to link up their genealogy
with the name of the founder of the clan. The privileges of primo¬
geniture thus constitute one factor that makes for economic and
ritual differentiation within the clan and thus for leadership.

(b) Wealth. The wealthy person, whether he has accumulated
his possessions through inheritance or through personal effort, has
means of gaining prestige and influence both within and outside
his clan. In the first place, by his ability to offer everyday hospi¬
tality in the form of beer, his homestead becomes the gathering-
place of the elders of the neighbourhood. In addition, he gains a

232 AFRICAN POLITICAL SYSTEMS

more definite influence over particular individuals in his clan by
lending them a goat or sheep for a sacrifice, a heifer for their mar¬
riage cattle, or basketfuls of grain if they ran short of food. The
person who often receives such support thereby assumes the obli¬
gation to praise his creditor, to oblige him by small services, or, if
he cannot return the loan, to render more substantial help by
herding his benefactor’s cattle, clearing his gardens, and keeping
his huts in repair. The traditional type of ‘retainer' or servant who
is found in some wealthy homesteads usually has come into that
position as a war captive, as a widower without children, or as a
debtor who could not return his debt.

Moreover, by giving feasts on a clan scale, especially by killing
the ‘ox of splitting’, a wealthy person has a means of gaining popu¬
larity among all his clansmen. Through his right of directing the
distribution of meat, he can favour those who respect and honour
him and who, at the discussion of clan matters, submit to his
views. ■ Finally, as elders of other clans are invited to these feasts,
the wealthy persons also become, in a sense, the representatives of
their clan. When elders of other clans kill the ‘ox of splitting’ they
are invited in turn, or gifts of meat or beer are sent to them, which
they apportion to their own clansmen. They have thus an oppor¬
tunity of gaining influence among their clansmen even when they
are the recipients and not the givers of feasts.

(c) The Quality of being an Qmugasa. Among the Vugusu, the
leading elders of a clan are called avagasa, i.e. men who talk gently
and wisely and who can make the people listen and return to reason
when ^they want to quarrel or fight. The possession of these
qualities is usually quoted, as the most important condition of
leadership. A son who as a herdboy begins to show reason and the
capacity of making his age-mates follow him in the various
activities in which herdboys indulge is pointed out by the elders as
a future omugasa , and they welcome his presence when he sits
near them and listens to their stories of long ago. When' he has
become an old man he acts as an omusmi , i.e. he is called to the
people to speak to them and comfort them when they assemble
after a funeral to distribute the property of the deceased, to
decide who should inherit the widows and to settle outstanding
claims and debts. The death of each clansman is a critical moment
for the preservation of peace within and between the clans, as it
invariably leads, to accusations of witchcraft or sorcery as being the

THE BANTU OF KAVTRONDO a 33

cause of the death that has occurred. It is then the duty of the
omusetd to curtail all such accusations by pointing out "that all
people are bom into this world to die and that people should not
harbour grievances and accuse one another of sorcery, as such an
attitude would merely increase the sorrow that had befallen them.
The omuseni usually winds up his speech with a review of the great
deeds of the clan and with exhortations to live up to that tradition
and to forget petty quarrels for the sake of peace.

Similarly, the omugasa is expected to talk for unity when legal
disputes are discussed before the elders of the oluhia. When homi¬
cide or murder has occurred and the kinsmen of both parties insult
one another and show impatience to fight, he persuades them to
give and accept compensation. The degree to which he succeeds
in such efforts determines his recognition as a leader.

(d) Reputation as a Warrior . A further quality that in the past
made for leadership was the reputation gained as a warrior.
Success in warfare served as a means of gaining wealth, but it also
brought prestige in itself. Both among the Vugusu and the
Logoli the names of clan heads of the past that are remembered
are associated with accounts of their deeds as warriors, their
success being measured in terms of the number of enemies they
have killed and the head of cattle raided by them or under their
leadership. Whether the choice of a successful warrior as a leader
in raiding expeditions was linked with a belief in his possession of
superior magical powers is to-day difficult to determine; he is said
by the old men to have been chosen for his courage and his ability
of inducing the others to follow him in an attack. Since, as has
been said above, war expeditions were frequently undertaken
jointly by several clans, leadership in fighting, more than that
acquired in other ways, tended to be recognized by several clans
and thus to establish a superiority of one clan over others.

(e) The Possession of Magico-Religious Virtues, Although the
persons most commonly called to offer private sacrifices to the
ancestors are the members of the senior line of a lineage, this duty
can also be performed by any classificatory father or elder brother,
if his qualities of character are such that he is considered a suit¬
able person. He must be known for his kindness and honesty;
he must be past the age of sexual desire; and he must be some one
‘who can feed the people’; in short he must be a person without
etnbala, i.e. without any failures and blemishes in the record of

234

AFRICAN POLITICAL SYSTEMS

his past and present life, if his sacrifice is to be favourably accepted
by the spirits. That elder in the clan who possesses these qualities
to the highest degree is recognized as the omusalisi munene (‘the
p-eat sacrificer’) who is called to perform private sacrifices of great
importance and, among the Logoli, also the omoaU, the public
sacrifice to the tribal ancestor and deity. His office is not
hereditary, but dependent upon personal qualities.

Among the various ‘experts’, the dream-prophet and the rain¬
maker appear to have wielded political power of a kind, as, through
the practising of their arts, they could influence the activities, not
only of single individuals, but of larger groups of people. The
dream-prophet, as has been stated, was consulted on the probable
outcome of war expeditions, on the advisability of migrations, the
probability of epidemics, and on similar matters of wider concern
while the rain-maker, through his alleged ability of withholding
ram and of directing the rainfall, not only in general, but in respect
of particular gardens, had the power of an executive organ in the
administration of justice. These two experts, however, form
categories of their own. Their knowledge is by virtue of inherited
secret medicines and spells, and they wield it independently of
their clan—and, in the case of the rain-maker, even of their tribal
affiliations. Their special knowledge, therefore, does not appear

to have lent them authority beyond that implied in their specific

practices. r

[0. A f e ' P ld , a S e ’ finely, was the most general condition of
political leadership and was socially marked through the institu¬
tion of circumcision age-grades. The recognition of primogeniture
for the regulation of inheritance and succession lends seniority a
superior status in all kinship relations. Generally speaking it is
always the oldest member of a group of kinsmen whose opinion
cames the greatest weight on matters concerning that group.
Adult sons show more obedience and respect to their father’s
oldest brother than to the father himself, and after their father’s
death his authority is not immediately transmitted to the oldest
son, but first to the next oldest brother who is still alive.

The authority implied in old age is further strengthened by
notions connected with the ancestor cult. One of these is that old
age is regarded as a necessary condition of performing sacrifices
as it requires a mind that is free from sexual desire and that
possesses other qualities characteristic of old age, such as wisdom,

THE BANTU OF KAVIRONDO 235

gentleness, and freedom from greed and jealousy. The other
notion is that spirits remember the treatment received while they

were still living persons and that they spare or trouble their living
relatives according to the treatment received. Old men, therefore,
are more than others feared as potentially troublesome spirits, a
fact which considerably adds to their authority. Their power of
uttering a curse, and especially a dying curse, is an all-powerful
sanction at their disposal.

This review of the different ways of gaining prominence in the
clan and tribe shows them to be of such a nature that they are
not mutually exclusive. The more qualities of leadership came
together in one person, the higher was his authority and the
wider the group that recognized it. While primarily based on the
organization of the patrilineal kin-group, leadership could, as we
have seen, extend to embrace the clan and even a number of clans
through the channels of wealth, warfare, and sacrifice. If there
were several people in the clan who possessed the different
qualifications of leadership, it was divided between them, but
such a division does not appear to have led to an institutionalized
distinction between different types of leaders, such as war-leaders,
judges, and priests. Provided that he possessed the other necessary
qualities, the war-leader, as he became old, was recognized as an
arbiter in legal disputes and called as a performer of sacrifices, as
he had increased the power of the clan and pleased the ancestors.
There was a division of authority only in the sense that the leader¬
ship of the old men in matters of jurisdiction and sacrifice was
paralleled by the leadership of the active warriors in the conduct
of fighting.

Political authority thus remained inarticulate. It was not linked
up with clearly defined rights and privileges, such as are usually
associated with institutionalized chieftainship. The leading elders
of a clan or sub-clan were merely those persons whose opinion
carried most weight when public matters were discussed on the
oluhia and who were called to perform sacrifices. They had no
rights that were inherent in their office, such as to collect tribute,
to enact laws, to call up warriors for a raid, or to grant or refuse
residence of strangers on clan lands. There is no generally
accepted term for a clan or tribal head, but a leading elder is
referred to by a variety of terms which can also be used with

*36 AFRICAN POLITICAL SYSTEMS

regard to any respected and honoured person. Finally, there was
no formal appointment and installation of the head of a clan or
sub-clan. Only when a leader in warfare was chosen, he is said
among the Logoli, to have been shaved and anointed with ghee in
the presence of the elders of the clan and to have been presented
by an old warrior with a head-dress of cowrie shells, a ribbon of
colobus skin, and a cloak sewn up of pieces of the skin of various
animals, a ceremony which, aside from lending distinction to the
war-leader, had a magic significance. Finger-rings, rare feathers
wristlets, ivory armlets, and spears are similar ‘insignia’ of this
kind which were ceremoniously given to a man recognized as a
war-leader. They were kept by the person upon whom they had
been bestowed, and when he had reached old age were passed on by
him to his oldest son or to another worthy successor within the
clan. Such insignia of leadership seem, however, rather to have
been charms than proper regalia implying a clearly defined status
as they were not outwardly distinguished from similar ornaments
worn by ordinary elders. They were neither limited in number
nor clearly graded in importance.

SKETCH MAP OF TALELAND

(Principal Namoo or part-Namoo settlements underlined thus : Tongo)

THE POLITICAL SYSTEM OF THE TALLENSI OF THE
NORTHERN TERRITORIES OF THE GOLD
COAST

By M. Fortes

I. The Country and the People

T HE Northern Territories of the Gold Coast are inhabited by
nearly three-quarters of a million people of negroid stock.
They are part of a great congeries of peoples spreading far into
French West Africa which speak related languages and share the
same basic culture. To this congeries belong the Tallensi, who
speak a dialect of Mole-Dagbane, a language prevailing in the
eastern half, roughly, of the culture area under consideration.
South of them, across the White Volta River, dwell the Mamprusi,
speaking a dialect hardly distinguishable from theirs, but exhibit¬
ing a somewhat different variant of the culture. Economically and
demographically, the Mamprusi show many contrasts to the
Tallensi. 1 The other tribes adjacent to the Tallensi—the Gorisi
(or Nankansi), Namnam, and Kusaasi, as they are commonly
named—differ so little from them that they might all be regarded
as subdivisions of a single cultural unit. Together they number
some 170,000 people in British territory.

The Tallensi total about 35,000. To describe them as a tribe
suggests a cohesive or at least well-defined political or cultural
entity differentiated from like units. Actually, no ‘tribe’ of this
region can be circumscribed by a precise boundary—territorial,
linguistic, cultural or political. Each merges with its neighbours

1 The Mamprusi have a population density of twenty-three to the square
mile, whereas the administrative district which includes the Tallensi has a

density of 171 to the square mile. The Mamprusi live in villages often widely
scattered and varying in size from tiny hamlets to places with several thousand
inhabitants. Their country, relatively low-lying by contrast with the high,
well-drained plateau north of the White Volta, is reduced to swamp over con¬
siderable areas in the rainy season. Their economic system is much more
complex than that of the Tallensi, and their religion has been influenced by
Mohammedan communities settled in their midst. All population data are
cited from the 1931 Census.

240

AFRICAN POLITICAL SYSTEMS

In all these respects. In the transition zones between two ‘tribes’
dwell communities equally linked by residential contiguity and
by structural ties to both. Indeterminate frontiers roughly demar¬
cate the Tallensi as an aggregate of communities speaking one
dialect and having more cultural nuances In common and more
social bonds with one another than any of them have with neigh¬
bouring ‘tribes’.

Any inhabitant of Taleland calls himself Tabq by contrast with
Mamprusi, Gorisi, &c. Among themselves, however, they dis¬
tinguish the clans dwelling on and around the Tong Hills as the
‘real Tails’. 1

These and other Tale clans had to be subjugated by military
operations which engendered a lasting respect for the power of
the white man. Since then (c. 1911) a system of administration
has been evolved under British control intended primarily to
maintain peace and to provide the labour, and materials necessary
for opening up the country. Based in a rough and ready way on
native institutions, It nevertheless endowed native functionaries
with powers, backed by the sanction of force, both different and
superior to any permitted them by custom. 2 Alongside of the
system thus Imposed by the Administration, though partially
submerged by it, the native political institutions still flourish,
however, because the ends they subserve vary considerably from
those of the modem innovations. Apart from the Administration,
no other contact agents are active within Taleland. The founda¬
tions of the native social system remain intact. 3

II. Character of the Political System

Twenty-five years ago there was no one who had authority over
all the Tallensi; no one who could exact tax, tribute or service from

1 } Tallensi , the form current in the dialect (Gomi) of the Gorisi for all
the inhabitants of Taleland. In their own dialect (Talni), they speak of them-
selves as TaJis (sing. Tahp), a form which, for the sake of clarity, I shall keep
for the ‘real Talis* only.

2 1x1 x 933 this system of administration began to be replaced by a new
machinery of government based on the principles of Indirect Rule. This political
expnment has already produced marked and valuable changes, but a con¬
sideration of them must be left for a later publication. I confine myself here to
the period preceding 1934, the date of my first visit to the Tallensi.

on ‘Culture Contact as a Dynamic Process*, Africa, ix, 1,1936.
Reprinted in Methods of Study of Culture Contact in Africa, Mem. XV, Int.
Inst, of African Languages and Cultures.

241

THE TALLENSI

all. They never united for war or self-protection against a common

enemy. They had, in short, no ‘tribal 5 government or ‘tribal 9

citizenship, no centralized State exercising legislative, adminis¬
trative, juridical and military functions in the interests of the
whole society. Until British rule made them the subjects of a
foreign State, obliged to render certain services and to obey certain
laws and entitled in return to protection and freedom of move¬
ment, it was dangerous for anybody to travel outside his own com¬
munity, except under the safe-conduct of kinsmen in other clans.

The indigenous political system of the Tallensi has a different
character. It is based on a social structure which determines
the status, rights and obligations of individuals and defines units
—both territorial and associational—that transcend the domestic
group and outlast changes in membership due to birth and
death. A differentiated constitution provides for formal leadership
and authority within each unit, and there are institutions binding
them together in mutual dependence, compelling their co¬
operation for the common good and regulating their inter¬
relationships, hostile or friendly. Finally, there are explicit sanc¬
tions maintaining the system.

IIL Warfare

Formerly, war used to break out from time to time in Taleland.
During the last three generations three large-scale ware occurred,

involving almost all the Tale settlements, but, significantly, no
neighbouring ‘tribes’. Small fights were more frequent, both
between Tale clans and between Tale and neighbouring non-Tale
clans.

In general wars the alignment of forces always followed the
major cleavage which runs through the whole society. The same
communities always fought on the same side, assistance being
rendered to one another on the grounds of clanship, local, com¬
munity, or ideological ties. Armed conflict between units which
supported one another in war sometimes occurred; but this was
not regarded as war.

A general war consisted of a series of local skirmishes without
organized methods of collective attack or defence or any military
leadership, and lasted only two or three days. It ceased as abruptly
as it usually began—as soon as the clans which initiated hostilities
made peace.

242 AFRICAN POLITICAL SYSTEMS

Whatever the reason, it was a sin to instigate war. War occurred
when members of one clan committed a grave injury (e.g. murder)
against members of another from which theirs was divided by
social barriers more powerful than any ties uniting them. It was
not an instrument of policy, but an act of reprisal. Punishment,
not conquest, was its purpose. Territorial annexation was in¬
compatible with the social structure, nor could captives or booty
be taken. It was a stem taboo to retain any of the food-stuffs or
livestock pillaged in w r ar. All had to be destroyed or immediately
consumed. Yet war was clearly distinguished from armed self-
help. To kill a man in w r ar, though mystically dangerous, was not
homicide, as it would have been if he were slain in a private
quarrel.

War could occur only between contiguous communities or if
the settlements intervening between the opponents were the
attackers 1 allies. It implied the absence of clanship ties between
the opponent clans, which could therefore intermarry. The
highest frequency of marriage is with neighbouring communities;
hence a man’s kindred and affines would be amongst the enemy.
Great vigilance was necessary, for it is both a sin and a serious
breach of kinship ties to kill or injure such relatives. To take
captives w r as impossible since they would generally be kinsfolk of
the captor clan—people for whtose welfare and on whose behalf
elders of the captor clan must sacrifice to their ancestor spirits.

IF. The Network of Clanship and the Fundamental Cleavage of

Tale Society

Tale warfare illustrates the basic principles of the native political
organization. Any analysis of it must begin with a more precise
definition of the units I have called clans. They vary significantly
in their actual constitution, but the ground-plan of all is the
same.

A settlement is referred to as a Zep, a word which means pri¬
marily the Earth in its material aspect. It denotes also the Earth in
its mystical aspect (see below, p. 254). Ti ttrjo may mean the
whole country of which Taleland forms a part, or Taleland only,
or the settlement, according to the context of discussion. Tzy in
the secular sense is not a territorial concept, but indicates always
a localized social unit, a community, or part of a community.

The skeleton of every residential aggregate is a clan, a part of a

THE TALLENSi

243

clan, or a group of clans. Such a local clan is conceived by the
natives as an expanded agnatic lineage, all the members of which
are ‘kinsmen by consanguinity 5 (doyam) to one another. Actually,
the composite clan comprising two or more maximal lineages, 1
each occupying roughly its own section of the settlement, is the
commonest type. Co-members of a maximal lineage are bound
by the rule of exogamy. The correlative of this rule is the right of
male members to inherit one another’s widows, if kinship rules are
not transgressed thereby. 2 These, the basic norms of clanship, are
extended to other maximal lineages with which ties of clanship
exist, whether of the same clan or not. The constituent maximal
lineages of a composite clan, though relatively autonomous in rela¬
tion to one another, are bound by clearly defined reciprocal duties
and privileges, obligations and rights which emerge in ceremonial
situations, economic and legal affairs and in the religious insti¬
tutions.

A clan is referred to by outsiders as ‘the people of such and
such a place 5 , e.g. Toryism. From the graves of their ancestors it
can be inferred that the older settlements have been inhabited by
the present local clans for at least three centuries. According to
native theory, bonds of consanguinity, and therefore lineage
membership, can never lapse. New maximal lineages cannot arise
through fission of those in existence.

Like the constituent maximal lineages of a composite clan,
though less so, the major segments of a maximal lineage are

1 By a maximal lineage, I mean the most extensive group of individuals
tracing agnatic descent from a single common ancestor. It comprises, there¬
fore, all the agnatic descendants, male and female, of the remotest ancestor
(eight to eleven generations back) known to members of the group. A maximal
lineage has an hierarchical structure. It consists of two or more major segments ,
each a lineage of lesser span than the (inclusive) maximal lineage, whose
members have a common ancestor one degree less remote than the founder of
the maximal lineage. Each major segment comprises lesser segments con¬
stituted on the same principle; and so on down to the minimal lineage consisting
of the children of one man. A composite clan therefore has no single common
ancestor. The natives speak of a lineage as the ‘house’ ( yir ) or the ‘children’

biis) or the ‘room’ (dug) of So-and-So. There is no term for what I call a
‘clan’. The lineage system operates completely independently of numbers.
A maximal lineage of two members has the same status as one of 2,000 members
in the same clan. I use lineage as the general term for a lineage of any order of
segmentation and of any span.

2 A father (son) may not marry the widow of a son (father). The classificatory
extension of this rule differs slightly from one maximal lineage to another,
according to its structure.

AFRICAN POLITICAL SYSTEMS

244

relatively autonomous genealogically, ritually and jurally, ye
closely united by bonds of the same kind. The expansion of

maximal lineage through the course of generations, involvin

often territorial dispersion, enhances the centrifugal force
that promote the relative autonomy of its segments; but th
centripetal forces of clanship, common religious cult, and ofte:
political interdependence continue to hold it together. In it
temporal extension, every lineage represents a balance of thes
contrary forces. At a given time, it is a system of mutual!
balancing segments in which are vested the rights and dutie
through which the structural equilibrium is sustained. Thi
tendency towards an equilibrium is characteristic of every phas
of the social structure.

Descent being patrilineal and marriage patrilocal, the cor
tinuity of the lineage depends on its male members. Only the
inherit property like land or cattle, succeed to office, and transm
the ritual and moral observances (totemic avoidances, &c.) dis
tinctive of that unit.

Clanship has a further extension of political importance
Maximal lineages belonging to different, usually adjacent clans ai
asymmetrically linked by ties of clanship identical with those ths
unite constituent maximal lineages of the same clan, and ci
across the latter ties. 1 Between such units, as within the clai
war was impossible.

This ramification of clanship ties corresponds closely to tl
local distribution of clans. Its greatest elaboration occurs amor
the ‘real Tails’. Numbering about 10,000, they have some twent]
five territorially adjacent, composite clans interlinked by a netwoi
of clanship ties that embraces some clans of Goris as well. Thi

1 Thus, for instance, three adjacent clans, A, B, and C, are interlinked
follows: A has three maximal lineages, Ai, A2, A3; B has four, Bi, B2, B
B4; C has two, Ci and C2. Lineage Ai has ties of clanship with lineages I
and Ci, but not with the other B or C lineages, nor have Bi and Ci ties
clanship with A2 and A3. Members of Ai, Bi, and Ci may not intermarr
they may inherit one another’s widows and have the reciprocal ceremoni
obligations of clansmen. Members of Ai marry into the other B or C lineage
and Bi and Ci intermarry with A2 and A3. Similarly, A2 has ties of clansh
with B4, but not with the other B or C lineages, and A3 has ties of clanship wi
B3. Maximal lineages linked in this way have the same relationship towar
one another as the constituent units of a single clan, but the rights and duti
pertaining to it are less rigorously effective than within the clan. Clan C h
similar criss-crossing linkages with clan D, D with E, and so on.

THE TALLENSI

24s

every maximal lineage has its specific field of clanship,, overlapping
but not identical with that of any other maximal lineage of the same

clan. It is both a constituent unit of a local clan and an intercalary
unit linking that clan to another; and no clan is a closed unit. Thus
any given sector of this network manifests an equilibrium of clan¬
ship and local ties balanced against genealogical and local cleav¬
ages. Loyalty to the local clan is balanced by the contrary loyalty
to a component unit of a neighbouring clan. The resulting articu¬
lation of clan with clan gives the Talis a loose cohesion. They
often speak- of themselves as a unit differentiated from non-Talis
by distinctive ritual and ceremonial observances. In fact, not all
Talis have all these usages, and some non-Talis share them;
characteristically, the Talis overlap with neighbouring aggregates
of clans.

Interlocking with this nexus of clanship ties is an equally ela¬
borate network of ties of ritual collaboration in the Great Festivals,
the cult of the Earth and of the external boyar (see below, p. 262).
Ritual collaboration implies joint mystical benefits and responsi¬
bility and therefore amity and solidarity analogous to that of clans-
folk. The two sets of ties, though not congruent, reinforce one
another.

Fights between Talis clans were never, in consequence, regarded
as war. Mediators linked to the combatants by ties of clanship,
contiguity, or ritual collaboration immediately intervened. War
to the Talis meant fighting their traditional enemies, the people
of Tongo and their allies. Yet two Talis clans usually supported
Tongo (see below, p. 257) and the Talis were and are bound to
Tongo by stringent ritual ties, as will appear later.

By contrast with the Talis, their neighbours, the people
of Tongo, are called Namoos. This nomenclature reflects the
fundamental cleavage in Tale society. It is universally accepted
that the founder of Tongo, Mosuor, was a fugitive from Mam-
purugu, where he had been forcibly ousted from the paramount
chiefship. Mosuor found the primordial ancestors of four of the
Talis clans occupying the country. Chief among them was the
primordial Gbizug tmdaana, who, the myth relates, terrified of
the red turban, the flowing gown, the horses and the guns of
Mosuor-^-these are the insignia of chiefs—fled to the Tong Hills.
By a ruse, Mosuor caught him and declared that he had come
to settle peaceably and to bring benefits to the community.

246 AFRICAN POLITICAL SYSTEMS

Therefore the tmdaana allotted land to Mosuor and swore a
covenant of eternal friendship with him and his descendants.

According to their lineal successors of to-day, the primordial
imdaanas ‘emerged from the Earth 5 or ‘descended from Heaven 5 .
Namoos scoff at these myths as physiologically absurd, but recog¬
nize that they express a claim to priority of occupation. The
myth of Mosuor and the myth of the primordial imdaanas are
complementary and are typical of the culture. Such myths concep¬
tualize and postulate a beginning for the political and. ceremonial
relationships of chiefs and imdaanas , which they invest with the
sanctity of unchallengeable antiquity.

Mosuor 5 s agnatic progeny spread toYamobg andSie inTaleland
and to Biuk on the frontier between the Tallensi and the Gorisi.
Formally, the clan of Mosuorbiis (the children of Mosuor) con¬
stitutes a single maximal lineage distributed in these four terri¬
torially and politically independent units. The Tongo branch is
senior to the others, as Mosuor’s grave and the shrine (boyar)
dedicated to him remain in its custody. Each of these branches
includes several accessory lineages united to it by some genealogical
fiction and linked by ties of clanship with other clans. Tongo is
linked thus to the Talis nexus, the other three branches to clans
in the vicinity of each.

All those clans which claim descent from immigrant Mamprusi
are designated Namoos. They have the same distinctive ritual
observances. Living in close juxtaposition with the Talis are
several other genealogically independent Namoo clans, each a local
unit. Some fall within the political orbit of Tongo; others lie
outside it. Around Yamolog and Sie dwell heterogeneous clusters
of clans, Namoos and non-Namoos in close juxtaposition, some
interlinked by clanship ties, others completely independent
genealogically. Namoos and non-Namoos frequently form con¬
stituent lineages of the same clan, holding complementary ritual
offices divided by the same structural cleavages as separate the
Talis and the Tongo Namoos, but inseparably joined by equally
strong structural ties and common interests. Most of these
groups of non-Namoos claim affinity with the Talis as the autoch¬
thonous inhabitants of the country, though they are genealogic¬
ally distinct from them, on the ground of similar ritual observ-
ances and prerogatives connected with the Earth cult. They have
a role in the political system analogous to that of the Talis, but

THE TALLENSI

247

are distinguished from them by differences in the date and form
of their respective Harvest Festivals.

The Talis and their congeners, claiming to be the autochthon¬
ous ‘owners of the land’, and the Namoos, presumed to be of
varied immigrant origin* are territorially mingled, genealogically
intertwined, and bound together by ineluctable ritual ties. But
they are also separated by profound cleavages of equal import
for their respective functions in the political system.

V. Limiting Factors: Kinship , Local Contiguity and the Economic

System

Clanship, the most significant tie determining mutual assistance
in war, did not operate automatically. Even segments of a single
clan sometimes refused to help another segment if it was thought
to have incurred just reprisals. Clanship also interposes genealo¬
gical barriers between units. For the individual, cognatic and affinal
kinship ties breach the barriers. Great importance is attached
to cognatic relationship, particularly to uterine (soog) kinship.
But political relationships, like war, cut across these ties. Kinship,
though it limits the insulation of lineage and clan and restricts the
extent to which conflict can develop between such units, is mar¬
ginal to the political system. This is obvious nowadays in the
political intrigues which rend the country. Kinship ties between
adherents of rival factions do not mitigate their political hostility.
Conversely, however, the political rivalry of their clans does not
deter individuals from the intercourse and reciprocities that
kinship entails. 1

Local contiguity also establishes ties and cleavages. The
economic system, the lineage structure, and ritual ideology all put
a premium on local cohesion as a factor of community solidarity.
Where adjacent clans are genealogically distinct, they usually have
ceremonial or community ties. The balance of ties and cleavages
produces a state of tension liable to explode into conflict if one
group infringes the rights of the other. Peace and non-provocation
are stressed as the ideal relationships between neighbours. In
this respect, contiguity imposes, constraints similar to those of

1 The web of kinship spreads so widely, both spatially and genealogically,
that a native can travel as much as twenty miles, across ‘tribal* frontiers, working
his way from settlement to settlement through the hospitality and good offices
of kinsfolk in each.

248 AFRICAN POLITICAL SYSTEMS

clanship, 1 but in consequence of the operation of different
sanctions. The threat of war or, nowadays, of suspending friendly
relationships is especially effective, since it endangers also
community or ritual ties of fundamental import for the general
good.

When ties of contiguity are added to those of clanship or of
ritual interdependence, communities wider than, the local clan
emerge in certain situations. It is a matter of degree, of balance
and contrast. The constituent units of such a community are

more closely interconnected inter se than with other, similar units.
Co-ordinated action for one end may be succeeded by inde¬
pendent, even conflicting action, following the lines of local and
genealogical cleavage, for other ends.

Finally, the economic system is a limiting factor in the political
organization. The Tallensi are peasants farming mainly cereal
crops. The essential feature of their agriculture is fixed cultiva¬
tion. They till the land surrounding their homesteads {soman)
continuously, supplementing this with the less intensive cultiva¬
tion of bush farms (poog) some distance from the settlement. The
livestock they keep, though indispensable, is of subsidiary signifi¬
cance in their economy. Very few are wealthy enough to possess
livestock equivalent in value to more than one or two head
of cattle.

Peace and the introduction of British currency have brought
about a tremendous expansion of local trade, but commerce is
still mainly a casual occupation ancillary to agriculture. This is
the case also with the few domestic crafts of the Tallensi. The
only division of labour is that according to sex. Agriculture and
animal husbandry are predominantly men’s work; women attend
to the domestic duties and engage considerably in petty trade.

Hunting and fishing, though pursued with zeal, contribute
little to the subsistence level. The sylvan products of the untilled
bush, however, supply commodities indispensable for their

1 Thus it is an extremely heinous and therefore almost unknown offence to
abduct a clansman’s wife. It jeopardizes the very foundations of clanship as
a factor both of political solidarity and of cult unity. The whole clan would be
outraged and the elders would resort to the most drastic measures to set the
matter right. The reaction is similar in the more frequent case of a man’s
abducting a woman married to a neighbouring clan; but this is due to fear of
violent retaliation. In both instances, ritual reconciliation is necessary (cf. below,
p. 270, where this is referred to again in another context).

THE TALLENSI 249

domestic economy—firewood, shea (Butyrospermum parka) fruit,
&c. The products of the locust-bean tree (Parkia filicoidea) are
exceptionally valuable, but not always freely obtainable (cf.
below, pp. 258, 259).

The hazards of agriculture are enormous. The rainfall is pre¬
carious. An inopportune dry spell during the rainy season 1 may
ruin the crops and create widespread privation. A generation ago,
a prolonged drought spelt famine, when men in desperation seized
their own or their neighbours’ children to pawn or sell t he m into
slavery among the Mamprusi for food. Nowadays such catas¬
trophes can be averted by purchasing grain from more fortunate
areas. Locusts are another unpredictable menace. Food is
chronically insufficient; for even in an excellent season few people
have the surplus to lay up supplies against a setback.

Fixed cultivation entails permanent and stable settlements and
thus profoundly influences the political system. In the older
settlements, the core of the society, an economically independent
man forms land transmitted to him from his forebears, whose
graves are beside his homestead. Security of tenure approximating
to full proprietorial rights is the rule. In some settlements,
farm-land (kuo) —i.e. the rights of tillage—can be aliened
subject to the consent of potential heirs. 2 Elsewhere the sale of
land is a sin against the Earth. In any case only extreme necessity
will force a man to sell a farm. The home farms {soman) are a
precious patrimony sanctified by the labour of former generations
and held in trust for future generations. To sell this land is little
short of sacrilege. To a lesser degree, this applies also to bush
farms.

An essential element in the ecological adjustment of the natives
has been a steady expansion into the uncultivated tracts bordering
the older settlements. Pressure of population and low technical
efficiency appear to have been the main causes of this process of

1 In common with other parts of the Sudanese climatic zone, the Northern
Territories of the Gold Coast experience two well-defined seasons, a rainy
season yielding, in the region of Taleland,amean annual rainfall of about 43 in.,
which lasts from April to mid-November, and an entirely rainless dry season
lasting from mid-November to the end of March.

2 Land is only alienated to clansfolk, kinsfolk, or co-members of the same
local community, never to complete strangers. This is a consequence of the high
degree of congruence between local grouping and genealogical grouping. Tale
agricultural economy is more fully dealt with in M. and S. L. Fortes, ‘Food in
the Domestic Economy of the Tallensi’, Africa, ix, 2, 1936.

250 AFRICAN POLITICAL SYSTEMS

local colonization, which has been greatly accelerated and ampli¬
fied in range by the establishment of peace. It is a cycle in the
history of a lineage. Young men shift to the periphery for a period;
then, as the older members die, some return to take over the
patrimony now left to them. The lineage system and the ancestor
cult are the centripetal forces. The original home {daboog) of one’s
father is sacred; to abandon it is to incur the wrath of the ancestor
spirits. New colonists, often younger members of the same
lineage, replace those who return home. Gradually a permanent
nucleus may be formed of descendants of men who did not return
to their natal homes, and a new settlement arises. Such settlements
are genealogically more heterogeneous than the older settlements.
The dispersal of Mosuorbiis must have occurred in this way.

A maximal lineage, however widely it may be dispersed, never
ceases to regard the original home {daboog) of its founding ancestor
as its true home, very particularly associated with the spirits of
its ancestors. Though dispersed, it remains anchored to a definite
locality.

No one has an over-right to the farm-land a man holds by right
of inheritance or purchase. No one can dispossess him of it,
prevent him using it as and when he wills, 1 or resume any that
he leaves untilled. Land can be borrowed; it cannot be rented.
Chiefs and Undaanas (see below, p. 255) have no over-riding rights
of ownership entitling them to rent, tax, or tribute for land. They
have, indeed, no more land than they have acquired in the same
way as any other elder.

Economically, therefore, the Tallensi are a homogeneous,
sedentary, equalitarian peasantry. Every settlement has a few men
of more than average wealth, due usually to the fact that they
have many sons to farm for them. But no social privileges attach
to wealth, though it is admired and envied. Wealth cannot be
accumulated. It is partly utilized to add to the number of wives
in the joint family, thus progressively increasing the drain on its
resources, and is eventually distributed by inheritance. Thus it
has only a temporary advantage. There are no economic classes
cutting across and detracting from the solidarity of lineage, clan,

1 This was remarkably demonstrated when the Tong Kills were re-settled
in 1935-6. After twenty-five years, the people returned to take possession of
their ancestral lands without a single boundary dispute or a single disagreement
as to the ownership of plots.

THE TALLENSI

251

and local community , 1 a fact of utmost importance for the political
organization.

VI. Authority and Responsibility in the Lineage System

The principles of Tale social structure appear most conspicu¬
ously in large-scale activities like funeral ceremonies, the Great
Festivals, hunting expeditions, &c. Rights and duties, privileges
and obligations are vested in corporate units; and any authorized
member can act on behalf of the unit. The principle of repre¬
sentation, rooted in the identification of lineage members with
one another, is inherent in Tale social structure.

The range of participation determines what units emerge in a
particular situation—the maximal lineages in clan activities, the
constituent segments in lineage affairs, the clans in activies involv¬
ing many communities. Concerted action is achieved by a balanced
and symmetrical distribution of functions among the units
involved. The solidarity of a unit varies accordingly. Segments
bitterly opposed over divergent interests unite vigorously on
matters of common interest. Co-members of any unit have
a common interest in one another’s welfare and in safeguarding
one another’s rights. Any of them will take reprisals for a wrong
done against another.

The corporate identity and solidarity of the units thus delimited
by agnatic descent and locality are functions of a differentiated
constitution sustained by definite sanctions. Every lineage is
subject to the authority of its senior male member (kpeem ) 2 In
a lineage of narrow span, i.e. with common ancestry placed four
or less generations back, he is the most senior by generation; in
lineages of wider span, age is the criterion, since generation
seniority is no longer determinable. Throughout the social
structure seniority confers authority.

The authority wielded by a lineage kpeem varies with its span.
In the lineage round which a joint family is built up, the head

1 Chiefs and headmen have become exceptionally wealthy through the
exactions now within their power. They remain the individual beneficiaries
of the new dispensation. No social cleavages based on differences of wealth
have as yet grown out of this, though conflicts due to pecuniary competition are
assuming a political complexion in some parts of the country.

2 Wives are never assimilated into their husbands* lineages, though they
gradually come to share the loyalties and interests of the latter. They are under
the authority of their husbands, and a fortiori under that of any one exercising
authority over their husbands.

35* AFRICAN POLITICAL SYSTEMS

has complete moral and ritual authority; he has the right to dispose
of his dependants’ labour, property, and persons; and he can use
force or ritual measures to assert his authority. In a lineage of
greater span, the head has only moral and ritual authority over
his co-members other than his own dependants. The greater the
lineage span, the older is the kpeem, the more prestige, respect,
and honour attaches to his status, though it confers purely moral
and ritual authority. Thus if the head of a major segment invites
his co-members to hoe for him, it would be disrespectful but not
a breach of sanctioned obligation to refuse, and they must be
generously recompensed. The hierarchy culminates in the kpeem
of the maximal lineage.

Every important transaction, secular or ceremonial, between
constituted units brings into operation every grade of authority
in the maximal lineage. The legal and economic rights of the
family head cannot be exercised without the consent of the lineage
heads senior to him. Each grade of lineage head has corresponding
rights—e.g. in the gifts that legalize marriage. A specific right,
though it is rigorously defined, is an element in the total con¬
figuration of rights.

Rights imply responsibilities. Every grade of right and authority
is matched by an equivalent grade of responsibility. Those who
can exact economic services from their dependants are economic¬
ally and ritually responsible for their welfare and publicly liable
for their actions. The head of a major segment has ritual and
moral responsibilities towards and for the unit. He must provide
the ceremonial contributions due from the segment on important
public occasions, e.g. funeral ceremonies. He has no economic
or jural liabilities for his co-members severally or jointly. The
head of the maximal lineage has still more general moral and
ritual responsibilities. This hierarchy of rights balanced against
a hierarchy of obligations is the foundation of Tale jural
relations.

The kpeem is the principal representative of the lineage, the
focus of the forces maintaining its corporate unity and identity.
All inter-lineage transactions are conducted formally through
lineage heads; but whatever the issue, the whole unit must be
consulted. Every member may express his opinion, greatest
weight being attached to that of any one directly implicated,
economically or jurally.

THE TALLENSI * S3

Ultimately, a kpeem's authority rests on a moral basis—the
bonds of mutual dependence and common interest which unite
co-members of a lineage, accepted, as axiomatic in virtue of
community of descent and most explicitly conceptualized in the

ancestor cult. Every lineage, whatever its span, worships the

shrine (boyar) 1 , of its ancestors separately. This is the primary
index of its differentiation from other like units and the arch-
symbol of its corporate identity and relative autonomy. To the
hierarchy of segments constituting a maximal lineage corresponds
a hierarchy of ancestor shrines. At sacrifices to a particular
ancestor, every segment of the lineage sprung from him must be
represented. Thus segments of a lineage sacrifice separately to
their respective founding ancestors, jointly to their common
lineage ancestor.

VII . Tale Religion

A man becomes head of his lineage by succeeding to the custody
of the lineage ancestors’ shrine (boyar). He sacrifices to it on
behalf of the lineage or any of its members and in his own name,
especially at sowing, harvest and festival times. The Tallensi both
fear and venerate their ancestors, seeking to placate and coerce
them with sacrifices, so that health, fruitfulness, and prosperity
may prevail.

This is native belief. Objectively, Tale religion is a potent
instrument of social control. People who sacrifice together,
whether as kinsfolk or through ties of ritual collaboration, must
be at amity with one another, else they offend the ancestors.
Because of this, death and the extinction of his issue is the mystical
retribution falling on a man who murders a kins man or clansman.
For this reason, too, dissension amongst people thus united must
eventually give place to reconciliation.

The custodian of any shrine must be treated with respect by
those dependent on his priestly offices, else he may reject their
sacrifices. This is the most effective sanction of a lineage head’s
moral authority. The ancestor cult, the supreme sanction of
kinship ties, is a great stabilizing force counteracting the
centrifugal tendencies inherent in the lineage system. However

1 A boyar is a particular catagory of bay or. Any object or animal which has
ritual significance may be called a bay or. A boyar is the bayor which is the seat
of all the ancestors of a lineage as far back as the one who founded it.

254

AFRICAN POLITICAL SYSTEMS

widely a lineage may be dispersed, its members can never
escape the mystical jurisdiction of their founding ancestor.

Cognatic descendants, too, fall under this jurisdiction as

individuals.

The ideological framework of the lineage system is the ancestor
cult; that of locality is the cult of the Earth (tsy). It is not easy

to formulate briefly the connotation of the word tsy, in its mystical
aspect. The natives distinguish between kuo, the arable surface,
tarn, the soil, and tsy, the community, the locality, the land, or the
Earth in its mystical aspect, according to context. Unlike ancestors,
who differ from one genealogical group to another, and whose
influence is confined to their own descendants, the Earth is single
and universal; in theory, all peoples are subject to the mystical
power of one and the same tsy. Yet the Earth is manifold, too.
Sacrifices must be offered to it at particular sacred spots (toy-
gban), and the word tsy (pi. tss) is applied to such places, meaning
the sacred spot and its precincts. Toygbana, and hence tss in this
restricted sense, have names and, like genealogical groups,
differentiating taboos.

Between tsy the universal and tsy the particular sacred spot
lies tsy the locality embracing a number of sacred spots, but
conceived as the widest precincts of one, the principal (kpeem)
among them, and under the priestly jurisdiction of a single
tsndaana. This principal toygban is the ritual hub of the locality,
the shrine of tsy , the universal Earth, where important sacrifices
are performed, especially at the times of the Great Festivals, and
ritual atonement is made for sacrilege. As neighbouring tsndaanas
are usually connected by ties of clanship or of ritual collaboration,
so neighbouring toygbana and tss are regarded in ritual and
doctrine as ‘kinsfolk’ (mabiis). In conformity with the social
structure, all the tss of Taleland and of some neighbouring non-
Tale communities are regarded as ‘kinsfolk’, a metaphor which
reconciles the diversity of sacred spots with the singleness and
universality of the Earth.

The greatest sacrilege against the Earth is to shed human blood
in strife. Atoning sacrifices must be made by both parties or they
and their issue will perish. It is only less sacriligeous to keep
anything found on the Earth (tsyonpiima), especially stray animals,
objects of metal, or vagrant humans. Cloth may not be worn
when sacrifice is offered to the Earth.

THE TALLENSI 2-55

The Earth is impersonal, but ‘alive’ (honvor )—that is, a control¬
ling agency in the lives of men. Incalculable, like all mystical
agencies, the source of prosperity, fertility, and health as well as
of drastic retribution for sin or sacrilege, witting or unwitting, it
is regarded with great awe. As lineage and locality are inextricably
intertwined in the social structure, so tsp and ancestors {yaanani)
are indissolubly associated in Tale religion. The concepts mark
two poles of the system, the ancestors being concerned primarily
with the good of their descendants, the Earth with the general
good. Every lineage worships its ancestors, but the priestly offices
connected with the Earth are confined to particular maximal
lineages.

VIII. Chiefship and Tmdaanaship

The head of any lineage greater than the minimal is at the apex
of a hierarchy of lesser heads. The head of a maximal lineage alone
is not subordinate to one of higher degree nor balanced by the head
of another segment of equal order. His status in the unit is unique,
epitomizing its genealogical and corporate exclusiveness in com¬
parison with all other similar units. He is the fulcrum of its
relations with other units. Most heads of maximal lineages hold
special offices sanctioned by the religious system and defined by
myths of origin or descent. Through them a range of political
relations transcending the limits of agnatic grouping is achieved.

In native thought, these offices comprise two major institutions:
na'am , chiefship, and tmdaan, the office of Custodian of the
Earth. Na'am is the prerogative of one set of clans and lineages,
predominantly the Namoos; tmdaan that of the Talis and their
congeners, in accordance with the major cleavage of the social
structure. Actually, these two categories overlap. Several clans have
both types of offices, held by different maximal lineages, and some
offices (e.g. those held by the Hill Talis) have attributes of both.
Homologous though they are in many respects, na!am and tmdaan
are polar functions indissolubly coupled together though opposed.
This is the central factor in Tale political organization. The same
configuration is found, with many local variations, throughout
Taleland and the neighbouring areas. Its most precise elaboration
occurs in the Tongo district. The Master of Tongo ( Toyraana ) is
the most eminent chief (na’ab) in Taleland. He claims nowadays
to be the ruler of all the Tallensi; but this is a distortion in

AFRICAN POLITICAL SYSTEMS

256

terms of the modem privileges of chiefship, of a status which
is merely that of primus inter pares in the native system. His
pre-eminence of rank is apparent from the deference shown to him
by all other lineage heads in Taleland, as well as from the special
taboos observed by him alone. Yet he has no administrative
authority over any other community than Tongo, not even over
the junior—but locally and politically autonomous—branches of
Mosuorbiis. The Tongo na'am may be taken as a paradigm, for
it differs from lesser chiefships only in degree.

‘The chiefship belongs to all of us’ (‘Na’am la a ti waabi bon")
is a maxim often cited. The office belongs to the maximal lineage,
the clan, the community; a particular chief is only its temporary
incumbent. This conception, expressing the identification of the
whole group with the na'am, their loyalty to and pride in it, is the
basis of a chief’s moral and political authority. A chief cannot be
deposed, nor will dissident segments secede, however objectionable
a particular chief may be. They know that their turn to hold the
chiefship will come round; for every member of the maximal lineage
which has the prerogative of a particular chiefship is eligible for it.
Rank is temporary in a given segment. Conversely, only agnatic
descendants of the founder of a na"am may hold it, as only they can
directly invoke the beneficence of the chiefly ancestor spirits . 1

The Tallensi say that na'am is purchased {da). Theoretically,
any eligible man, young or old, may compete for it. Actually men
of junior status are considered to be unsuitable. Indeed, only
elders commanding the services of many dependants could,
formerly, raise the ‘price ’ 2 enabling them to compete, partly from
their own resources, but largely by borrowing from clansmen and
kindred. Competition for a chiefship was a contest of segments,
not of individuals. The higher a man’s prestige and standing, the
wider would be the span of the segment supporting him with loans
and by their presence on the election day. The general level of
economic equality made the purchase of na y am an indirect ballot.

1 Thus theoretically every agnatic descendant of Mosuor is eligible for any
of the chiefships held by Mosourbtis , and members of accessory lineages are
ineligible. Actually, the political independence of each branch is asserted by
the restriction of competition for its chiefship to its own members. Some minor
chiefships may be held by members of two or more clans on the same principle.

Minor chiefships ‘cost’ eight or nine cattle, important chiefships up to
seventy head of cattle, as well as large sums of cowries—not to speak of the
many presents that must be given to the elector’s elders to obtain their good
offices.

THE TALLENSI 3S7

Na'am brings Taleland into the political orbit of the Mamprusi.
The prototype and fountain-head of all na'am is the Chief of Mam-
purugu. To be valid, the mystical attributes which constitute its
essence must be ritually vested in the holder by him or by someone
endowed with na'am by him. Na'am is also an ancestral heritage
and therefore most appropriately held by those who belong to the
same stock as the Chief of Mampurugu, i.e. the Namoos. Accord¬
ingly, the elector of most Tale and many Gore chiefships is one of
his sub-chiefs, the Kuna’aba, The Torjraana alone among Tale
chiefs elects subordinate chiefs, on the same princple. A hierarchy
of chiefships results, all miniature replicas of, the fountain-head
na'am in structure and participating in its mystical virtue.

But the analogy of a feudal system 1 would be mislea din g The
chiefs appointed by a single elector—and by extension their clans¬
men—speak of themselves as ‘kinsmen’ (malms). In Taleland they
assisted one another in war, sometimes to die detriment of ties of
real clanship or of ritual collaboration . 2 They would also protect
one another’s clansmen from illegitimate molestation by their own
people. Similarly, Tale chiefs refer to Kuna’aba as their ‘father’
(ba), implying that they owe him loyalty, respect, and ceremonial
deference. They would not make war on his settlement nor he on
any of theirs. They would tiy to protect clansmen of his travelling
in Taleland from molestation, as he would their clansmen trav elling
in Mampurugu. But Kuna’aba has no economic, juridical, admin¬
istrative, or military rights sanctioned by the native political system
over any Tale chief. His ceremonial investiture by Kuna’aba is the
crucial act conferringchiefship ona man (even if he isactuallyselected
by anadministrative officer). Nevertheless, Kuna’aba’s modem judi¬
cial and administrative authority rests solely on the sanction of force
represented by the Administration. It is significant that he was

1 The Administration has always regarded Taleland as part of the ‘Mamprusi
State’, unde* the ultimate rule of the Chief of Mampurugu, through his sub¬
chief and deputy, the Kuna’aba, who was considered to have full jurisdiction
over ‘Kurugu Division’. Kuna’aba and his councillors were created a Native
Authority and Court in 193a as the only official court in the Division vested
with judicial and administrative authority. In the native political system,
Kuna’aba’s sphere of electoral authority does not correspond to a political or
‘tribal’ unit.

2 e.g. the people of Sii, Talis by clanship ties and ritual observance but hold¬
ing a chiefship from Kuna’aba, together with their clansmen the people of
Yindu’uri, usually supported Tongo in war against their fellow Talis for this
reason.

AFRICAN POLITICAL SYSTEMS

258

formerly treated with derision and his settlement often plundered
by Tale clans which have no ties through chiefship with him.

An elector has, correlatively, no direct ritual, political, or mili¬
tary responsibilities for a clan whose chief he appoints. He is
morally and ritually bound to select, in consultation with his
elders, the best candidate for a chiefship, recking less of his
pecuniary offer than of his reputation and pedigree. He must see
that the office circulates, so that it is not monopolized by any one
segment, thus stressing the common interest of the whole unit in
the no? am . 1 An immoral choice would evoke the wrath of the
ancestors. To this extent only has an elector a moral responsibility
for the well-being of the community whose chief he appoints.
He is the repository of na’am, so when a chief he appoints dies
the insignia with which he was invested must be ceremonially
restored to the elector pending the appointment of a successor.
Thus the cycle is completed.

IX. The Complementary Functions of Chiefs and Tmdaanas

Every Tale chief says, ‘ Man so nteg; ndame 1 (T own my land.
I bought it’). In certain respects the rights and authority implied
in this are precise. He owns the products of all locust-bean trees
(Parkia filicoidea) growing within the precincts of his clan settle¬
ment, as well as certain stretches of river and of hunting bush, the
exploitation of which is prohibited until they have been com¬
munally fished or hunted at the chief’s instance. Big fish and
special portions of any animal slain or found dead in his bush must
be delivered to the chief. Stringent ritual sanctions uphold these
rights. Vagrant humans (da'abr), stray dogs or cattle, and brass or
copper ware found lying about must be delivered to a chief. These
prerogatives have some, though limited economic value even to-day . 2

1 Owing to the advanced age at which chiefs were often appointed, the average
duration of a chieftaincy was only about ten years. This conduced to a fairly
rapid circulation of a chiefship amongst the segments of a clan.

2 The right to vagrant humans has, of course, been abolished. Some minor
privileges of chiefs corresponding to it in political significance have not been
mentioned, as they are also falling into abeyance. In several cases, chiefship
reduced entirely to these rights until the pacification of the country led to the
establishment of permanent settlements on the fringes of what used to be merely
hunting bush. Thus the chiefs of Biuq and Gbiog used to live amongst their
clansfolk as members of communities dwelling within the zone of authority of
other clan heads, though vested with rights over their respective tracts of river
and bush, &c.

THE TALLENSI

*59

The products of locust-bean trees, river, and bush are luxuries not
accessible to most commoners. Vagrant humans were sold; dogs
and cattle sacrificed to ancestor spirits. The modem privileges of
chiefship are sometimes described as substitutes for these tradi¬
tional rights.

But to the natives their crucial significance lies in the correlative
duties and responsibilities they involve. It was a grave moral
responsibility, subject to mystical penalties, for any one but a
chief to sell a wandering stranger into slavery. Fishing and
hunting expeditions are dangerous. Only a chief can fire the bush.
The fault for a serious accident falls on him. He must perform
precautionary magic before an expedition, and offer placatory
sacrifices to render river or bush safe again after an accident . 1

These rights and responsibilities are indices of the complex
configuration of rights and responsibilities through which chief-
ship accomplishes what the natives regard as its supreme end—‘to
prosper the community' (maal %). Na'am is a medium through
which the mystical forces conceptualized in Tale religion are
mobilized to ensure the welfare and fertility of humans, animals,
and crops—the common good, in so far as it is determined by
natural forces beyond pragmatic control. A chiefs death brings
famine upon the community. His blessing is as potent for good as
his curse is dangerous. His office is sacred, imposing on him
observances and taboos—very rigorous in the case of the Torjraana
—symbolizing his mystical powers and responsibilities. He is the
guardian of the community, responsible for the organization of
and major contributions towards sacrifices made by it to preserve
the beneficence of the ancestors and for the conduct of the annual
ceremonies of the Great Festivals. He is the custodian of the
sacred objects that symbolize the continuity and perpetuity of the
na'am. When a natural calamity threatens, the elders appeal to
him for intercession with the ancestors. Most important is his
power to regulate the rainfall.

These capacities, derived from the chiefly ancestors, are vested
in a chief as the highest representative of his maximal lineage. He
cannot exercise them arbitrarily, for his own ends, but only in con¬
clave with representative elders of the clan or community for the

1 Other aspects of communal fishing expeditions are discussed in my paper
on ‘Communal Fishing and Fishing Magic in the Northern Territories of the
Gold Coast*, J.jR. A 67 , 1937.

*6©

AFRICAN POLITICAL SYSTEMS

common good. But he is also bound to exercise them; they are obli¬
gatory because, as trustee of the ancestors, he benefits by the rights
of na'am bequeathed by them, which belong really to the mavim-ii
lineage. Hence a proportion of the economic goods he obtains
through them must be distributed amongst the segment heads.

This configuration of rights, responsibilities and mystical powers
binds a chief and his community in reciprocal dependence. Poli¬
tical boundaries are an innovation frequently causing acrimonious
disputes between chiefs and headmen. A chief, in the native
system, is the pivot of a community consisting of a series of zones
of increasing amplitude and diminishing integration. At the
centre is his own maximal lineage and clan. One or two contiguous
clans may be closely associated with it, forming part of this central
community in all but the genealogical sense. Its area of residence is
approximately that within which the chief owns all the locust-
bean trees. Beyond this stretches a zone of unrelated H a n ?
acknowledging the chiefs mystical value for the common good and
his correlative right to vagrant humans, but otherwise inde¬
pendent of, sometimes even hostile in war to, his clan. Divided by
local, genealogical and ideological cleavages which may precipitate
open conflict over divergent interests, such a cluster of Hang
emerges as a community in ritual collaboration for the common
good, especially during the Great Festivals or if a natural calamity
threatens. It represents a balance, usually, between Namoo and
non-Namoo units, the pivot of which is the bond between
chief and tsndaana.

Without the blessing of the Earth, a chief’s mystical powers are
void. Thus the final phase of his investiture is his ceremonial
reception by the tmdaanas of the community in turn, who present
hun to their Earth shrines (toygbana) with pleas for blessings on his
chieftaincy. Frequently thereafter he must send animals to thpr n to
be sacrificed to the Earth. He is powerless to ensure the welfare of
the community without their ritual collaboration. He cannot hunt
or fish his bush and riverwithout a tsndaana’s blessing. Finally, on
his death a chief is buried secretly by a tsndaana. A community,
whether it is a single clan or a group of clans, is politically defined
by the complementary functions of chief and tsndaana.

The relationship of chief and tsndaana is one of polar opposition
and mutual constraint limited by and maintaining their joint
responsibility for the common good, validated by myths like that

THE TALLENSI a6l

of Mosuor and symbolized in the taboos and prerogatives of each,
those of tzndaanas being mainly the exact contrary of chiefs’. As
throughout Tale society, the structural relationships are concep¬
tualized in and sanctioned by the ritual ideology.

A tmdaana— Custodian of the Earth—is primarily a religious
functionary. His office is homologous with chiefship, but oriented
towards the Earth. He ‘prospers the community’ by ensuring the
beneficence of the Earth for it. His ritual relationship with the
Earth imposes certain taboos (e.g. he may never wear cloth, but
only skins) on him and enables him to accept the responsibility of
dealing directly with it. Hence all lost property not the preroga¬
tive of chiefs must be delivered to a tmdaana. Lest the Earth be
offended, a tzndaana must pierce the soil for a hew grave and turn
the first sod for making a farm or building a homestead on vi rgin
land. Portions of the animals sacrificed on such occasions belong
to him. Tmdaanas may not sell men; but if a chief sold a vagrant
person he gave a cow to the tmdaana of the area where the man
was found as a piacular offering to the Earth. Because the Earth
abhors bloodshed, tmdaanas have ritual power to stop fighting and
to mediate in disputes. They perform the sacrifices offered to the
Earth to expiate murder.. Their curse or blessing is more potent
than a chief s, since the Earth is universal and can punish or bless
a man anywhere.

Tmdaanas, therefore, have great moral and ritual authority.
But they cannot ‘prosper the community’ without the collabora¬
tion of chiefs for they have no mystical power over rain. Thus in
the Tongo area, if flood or drought threatens, the representative
tmdaanas of the Talis call on the Topraana and exhort him to
avert it. Though ancient animosities and structural cleavages
divide their clans from his they are bound to collaborate ritually
for the common good.

X. Tmdaanas and the Wider Community

The office of tmdaana is vested in a maximal lineage. Any male
member of the lineage may at times deputize for the tmdaana, and
all its members must observe certain of his taboos. Tzndaanas
succeed by right of seniority or are chosen by divination from
amongst the segment heads . 1 They are ritually install^ by

1 The Hill Talis have a variant procedure which is a compromise between
this method of selection and the way chiefs are elected.

z6z

AFRICAN POLITICAL SYSTEMS

fellow tmdaanas of lineages linked to theirs by ties of clanship or
of ritual collaboration. This is one of many ritual ties that unite
neighbouring tmdaanas.

Composite clans have several tmdaanas , one of whom is senior
in status to the rest, but in ritual matters which concern the whole
clan all participate, the key roles being distributed equally amongst
them. Tmdaanas of neighbouring clans, whether interlinked by
clanship ties or not, usually have ritual ties, and any one of them
can represent all in ritual relationships with a chief or another such
group of clans. In the Great Festivals, the key functions are dis¬
tributed amongst tmdaanas representing such groups of clans and
an equilibrium of mutual dependence is attained which is an
extremely powerful sanction of solidarity, counteracting the
conflicts due to divergent loyalties.

Every tmdaana has his own tzy, the area within which he
sacrifices to the Earth shrines (torjgband) and exercises his other
ritual prerogatives. He allots any unowned land in this area for
farming or building, in return for gifts which have a ritual and not
economic significance. People of other clans than his may dwell
there, acknowledging his ritual rights, but not paying him rent
or tribute.

A tmdaanas's tmj is roughly demarcated by certain natural land¬
marks, but it has no precise boundaries, since it is only a sub¬
division of the single, unitary Earth. Since neighbouring
tmdaanas usually have clanship or ritual ties, they regard their
respective rights and responsibilities as specific devolutions of
what are really common rights and responsibilities shared, in the
last resort, by all tmdaanas. Frequently neighbouring tmdaanas
have one or more Earth shrines in common.

The Earth cult, therefore, is at the same time a factor in the
differentiation of structural units, accentuating their relative
autonomy in relation to one another and their divergent interests,
and a factor in the integration of the community. In the wider
community, it balances chiefship; but in the narrower unit of a
local clan or interconnected group of clans like the Talis, which
has no chiefship, it is balanced by the cult of the ‘external (yeyha)
boyar . Among the Hill Talis, this is an esoteric cult into which
their youths are initiated, as well as a fertility cult which attracts
pilgrims from places far beyond the borders of Taleland. Its
devotees could visit these places safely, and this was a channel

THE TALLENSI a 6 3

for both economic and cultural exchange. Such e boyar is the
core of the Harvest Festival of its congregation, which consists of a
group of maximal lineages generally of different clans, each having
the prerogative of one ritual office connected with the cult. This
grouping cuts across the grouping in terms of ritual collaboration
in the Earth cult. A boy at is the seat of its congregation J s ancestors,
the opposite pole to the Earth in the religious scheme. Thus ritual
sanctions and interlaced loyalties are counterbalanced to maintain
the social equilibrium. Among the Hill Talis, the boyar has the
same mystical value and function as na'am in the wider community,
and its principal officers are referred to as ‘chiefs’ among them¬
selves, though they are tmdaanas in relation to the Chief of
Tongo.

The most conspicuous mechanism through which the ritual
interdependence and joint responsibility for the common good of
chiefs and tmdaanas is maintained is the cycle of the Great
Festivals . 1 Its centre is the Tongo area, but it embraces all the
Tale settlements as well as several neighbouring non-Tale settle¬
ments each entering the cycle in its proper sequence. These
festivals are periods of ritually sanctioned truce, when all con¬
flicts and disputes must be abandoned for the sake of ceremonial
co-operation. In each phase of a festival, every corporate unit
involved has its specific ceremonial role, vested in its head and
indispensable for the propitious outcome of the whole set of cere¬
monies ; and in each the crucial act is the meeting of chief and
tmdaanas, or their deputies, jointly to perform ritual for the bless-
* n f> Qf the community. The chief on whom the most important cere¬
monial duties devolve is the Toyraana ; but the rites and celebra¬
tions show that he represents all the chiefs whose common
heritage is no!am derived from the Chief of Mampurugu and
whose rights and responsibilities are interlinked through this
fountain-head. Similarly, the principal tmdaanas concerned repre¬
sent all tmdaanas.

In this festival cycle, therefore, the widest Tale community
emerges; but it is so loosely articulated that for the members of any
particular clan it forms merely a remote frame of social reference.

It is not a fixed political entity but a functional synthesis. It brings
out the common allegiance and ideological fraternity of all chiefs,

1 See my paper on ‘Ritual Festivals and Social Cohesion in the Hinterland
of the Gold Coast’, Amer. Anthropologist , 38 , 4, 1936.

264 AFRICAN POLITICAL SYSTEMS

the kinship of ail tmdaanas evinced in their common cult of the
Earth, and the complementary functions of these offices. It means
the dominance, for a period, of the forces of integration ever¬
present in the social structure—in kinship, clanship, neighbour¬
hood ties, chiefship and tmdaanaship —but generally submerged
by the sectional interests, springing from these same institutions,
that divide Tale society into a multitude of independent corporate
units. The festivals are annual events reputed to be of immemorial
antiquity. This is proof of the relative stability of Tale society over
a long period of time and of the well-adjusted balance maintain
in the long run, between the forces of integration and those of
differentiation.

The mainspring of this synthesis is Tale ritual ideology. Prin¬
cipally, it is the notion of the common good as referring to human
welfare and prosperity in their most vital and universal aspects,
superseding all sectional interests rooted in the social structure!
The mystical determinism postulated for it raises the common
good above all mundane issues and subjects the obligation to
collaborate for it to unchallengeable and eternal sanctions which
it would be inconceivable to flout. It stands for the widest body of
established custom, the ‘rule of law’ as the Tallensi understand it,
which regulates their social life.

XI. The Secular Authority of Chiefs and Tendaanas

The secular powers and authority of chiefs and tmdaanas have
been radically altered by the advent of British administration.
Chiefs are now the agents of the Administration, exercising
judicial and executive authority in its name and with its
backing. Tmdaanas have no political status under this dispensa¬
tion. Clans which have no chiefs are governed by headmen
calling themselves ‘chiefs’ and exercising the modem powers of
chiefs. They form part of the administrative machinery which has
grown up to meet modem requirements—the provision of labour
and materials for public works, such as roadmaking, formerly non¬
existent, and especially the maintenance of peace and the enforce¬
ment of law.

The significant characteristic of the new order lies in the
Administration s monopoly of the sanction of overwhelming force.
Chiefs and headmen nowadays exact taxes, tribute, and labour
from their people which have made them fabulously wealthy

THE TALLENSI i6$

compared with their predecessors. They stress jealously their
territorial and political independence, instead of, as in the native
system, the common basis of their rights and responsibilities As
agents of the Administration, they place their private interests
first. To perform the duties and exact the rewards of their present
administrative status, they rely on the assistance of their close
agnates and of a new class of subordinate officers appointed by
themselves; for the new sanctions deriving from the backing of
Government could not operate through the hierarchy of lineage
elders, who have no power to coerce fellow members.

The people accept the new powers and exactions of chiefs and
headmen with a mixture of resentment and resignation. This does
not affect their valuation of their traditional political institutions
For the new system and the native political organization are still
largely discrete, though focussed partly in the same personages
The sanctions inherent in the native social structure are not
effective in the framework of the new administrative organization.
Chiefs and headmen are not restrained from what would be illicit
extortions according to native values by the sanctions to which
they submit unconditionally as members of the native social
structure. Friction is inevitable when chiefs or headmen attempt
to assert their administrative rights in situations defined according
to the native political ideas by others; and factions coveting the
wealth and power of office under the Administration are arising
with claims to recognition based on their status in the native
system.

At the same time, tendencies conducing to the fusion of the
two systems are operating. Chiefs’ and headmen’s tribunals are
especially important in this connexion. Though not recognized
as part of the official judicial machinery , 1 they were encouraged
by Administrative officers. They are rapidly becoming an integral
part of the native political structure, though their authority is
derived from the power of the Administration to enforce the peace

1 The only officially recognized Native Tribunal in Taleland was, in 1034
tlm of the Chief of Kurugu (Kuna’aba). Its jurisdiction was limited to civil
wrongs, with the exception of actions relating to land, inheritance, or damages
or debts of over £5. Crimes fall under the jurisdiction of British courts, the
District Commissioner sitting as magistrate. Actually, most cases dealt with
by the Chief of Kurugu’s court came on appeal from the unofficial chiefs’ and
headmen s tribunals, and appeal was allowed from this court to that of the Chief
n Mamprusi, to which all land cases also went in the first instance.

266

AFRICAN POLITICAL SYSTEMS

they have been established to maintain. Their vigour is due to
the abolition of traditional methods of obtaining redress for
wrongs. Chiefs and headmen are the judicial officers, assisted
sometimes informally and for reasons of etiquette and of ancient
habit by a few elders. Their jurisdiction is confined to civil
wrongs, and though they could until recently inflict fines in
special cases, they lacked penal sanctions for the enforcement of
their verdicts. The best of them tried, therefore, to arbitrate
justly, so as to gain the acquiescence of both litigants. Generally
a chief deals with cases in which members of his community are
defendants. The hearing fees paid were a lucrative source of
income until recently.

These judicial powers have enhanced enormously the prestige
and authority of chiefs, especially within their own clans. Their
judgements are influencing the development of Tale law and
custom. Yet their administrative authority is still bounded by the
cleavages of the native social structure. The Totjraana,ior instance,
though recognized by the Administration as chief of the T ab's ’
has no effective administrative control over them. The l imi t of
his effective authority is the close-knit community consisting of
his own clan and two adjacent clans, which have always been
intimately united to Tongo by local, kin, and ritual ties.

In the native system, the secular authority of a chief or a
tmdaana is derived, on the one hand, from his ritual status, and
on the other, from his supremacy in the hierarchy of lineage elders.
Chiefs and tendaanas, especially those who are considered to be
of senior rank, are always treated deferentially. Their ritual
prestige and status in the lineage hierarchy has always enabled
them to command individual or communal assistance from the
whole clan in return for the customary recompense. They had no
right to tax, tribute, or service. They were and are morally obliged
to be hospitable and generous, especially to their clansfolk, but
they have never had economic obligations towards them severally
or collectively.

. ^ he ? d of the maximal lineage, a chief or tmdaana must be
informed of all important affairs that concern it. His assent is
necessary m the conduct of many, especially if they involve rela¬
tionships—jural, ceremonial, or economic, whether pacific or
hostile—with other clans. A chief cannot, for example, allocate any
land except his own to a new settler, but his consent and blessing

the tallensi 3 6 7

are essential to permit the man to join the community and prosper . 1
If the common interests of the clan or of a close community like
Tongo and its neighbours are infringed, e.g. if a member is
murdered or a member’s wife abducted, the action to be taken is
decided and often carried out by a conclave of the elders presided
over by the chief or tmdaana, or by a committee of all the tmdaanas
and elders, in a composite unit. It was a grave sin for a chief or
tmdaana to instigate war, but if an individual or a segment went
to war, help would only be given by the rest of the clan if the chief
or tmdaana consented, since his blessing and intercession with
the ancestors and the Earth were indispensable for victory.

In these ways, chiefs and tmdaanas have always exercised
considerable authority in the affairs of the clan; formerly, they
had no judicial or administrative powers comparable to those of
contemporary chiefs and headmen.

Associated with every chiefship—integrally part of it, according
to native ideas is a number of titled elders (kprni) appointed by
the chief in the same way as he is himself elected. Appointment
to one of these titles is a signal distinction, though their value is
mainly honorific. A conscientious chief distributes them fairly
amongst all the segments of his clan, as well as amongst neigh¬
bouring clans closely bound to his. These elders never formed
a regular council. In the affairs of the unit, the lineage elders
played as great or greater part than they; but they, and through
them their respective segments, have direct bonds of loyalty to
the chief, independent of the lineage structure and counteracting
the centrifugal forces of divergent segment loyalties. Na’am is
thus, as it were, distributed amongst all the segments of the
maximal lineage.. In keeping with this, some of the titled elders
have special duties and compensatory privileges connected with
the chief’s rights over locust-bean trees, bush, and river. The
Ytdaana, the most important titled elder, acted as the chief’s
deputy and spokesman in public matters. In the interval between
the death of a chief and the appointment of his successor, when
the na'amvfas carried on by his brothers and sons, a Yidaana could
formerly exercise great influence on the conduct of affairs. These
offices are not found in clans which have only tmdaanas and

1 A Undaana can allocate unowned land within the locality over which he has
ritual jurisdiction as fanning plots or house sites. The ‘tenant’, however, owes
nim only a ritual tithe in return, but no political allegiance.

268 AFRICAN POLITICAL SYSTEMS

analogous ritual lineage heads, who are more closely identified
with the lineage structure than chiefs.

How, then, were the rights and interests of individuals or groups
protected against injury thirty years ago? The general principles
of what we should call legal procedure are the same now as they
were then; for a case reaches a court only when other methods
of effecting a settlement fail—when, formerly, the injured party
would have resorted to armed self-help or drastic ritual sanctions.

The action taken to redress a wrong or assert a right depended
on the structural relationship in which the parties stood to each
other. The distribution of rights and responsibilities in accord-
ance with the lineage structure makes every dispute an issue
between groups—segments, lineages, clans, communities. Settle¬
ment by negotiation between the lineage heads, conducted through
a privileged intermediary like a tmdaana, or a kinsman of both
units, or the head of a lineage connected with both, would be
attempted, to begin with, and often succeeded. A dispute between
clansmen creates intestine hostilities, disrupts co-operation, and
undermines the clan’s corporate unity. A misfortune attributed
to the displeasure of the ancestors may intervene and compel
a reconciliation to be effected. But if it is acute or involves two
major segments, now as formerly it may be brought before the
chief or Undacma, who with the elders threshes it out. Chiefs
nowadays treat these like ordinary cases, but are swayed in their
judgements as much by concern for the solidarity of the group
as by considerations of strict justice. In any case, the weight of

^" e f SSUre k^S^t to bear on the disputants is usually
sifificient to settle the matter; but instances are known where
clansmen fought one another as a result of a dispute.

b 4 tWeen members of clans linked by ties of clanship,
eighbourhood, or ritual collaboration were dealt with similarly.
The mjmed group might force a rapid settlement by threatening
to cut off reciprocal good relationships and to take to arms.

injured n a ^ mg 5 ** d “?“ ,oe betwfien ^eir settlements, the
Zff IS 7 T n0t direct ^Presentations or threaten
then opponents, recourse might be had to arbitration Chiefs
assi^ed by their dders, were the usual arbitrators. The injured
party would appeal to the head of their opponents’ maximal
meage, who would send the disputants to a neighbouring chief

d t0 enSUre 311 ^Partial hearing. This chiefs recompense

THE TALLENSI z6g

was a gift from the party for whom he found. He had no means
of enforcing his verdict. The heads and lineage elders of the group
adjudged wrong might attempt to do so in the interest of future
good relationships. Sometimes, as still to-day, the issue would be
left to mystical arbitrament. The disputants swear to the justice
of their respective claims by the chief’s skins or the Earth and the
prevaricator will, it is held, perish in due course. A chief could not
impose fines even on members of Ms own clan or expel any one
from the community. Like any head of a maximal lineage, he
might, if he were gravely affronted, or if some one were a source
of continual discord, curse the offender, who might migrate for
fear of the ancestors’ wrath. Public indignation might ha Ve the
same effect; for such people endanger the co mmuni ty’s welfare
Compensation plays no part in Tale methods of adjusting
wrongs. Homicide was and is regarded as equally a grave sin
against the Earth and the ancestors, and an injury against the
coiporate unity of the victim’s lineage and clan. If a man killed
a clansman, whether accidentally or deliberately, the elders of the
murderer’s segment sent to beg the forgiveness of the chief or
tzmaana for this act which threatened to ‘destroy the community’
{rjma tty). The chief or t&idaana and the clan elders would then
determine the number of cattle and sheep which must be offered
by the culprit’s family as expiatory sacrifices to the ancestor spirits
and the Earth. The victim’s family, too, must contribute animal.
to these sacrifices; for they serve not only to expiate the blood-
shed but to reconcile the two hostile segments. Vengeance is
forbidden and, if necessary, forcible restraint or a ritual inter¬
diction by the chief or tt.ndaa.na would be used to quell hot
tempers. The procedure was the same if the victim belonged to
a different clan from the murderer. But if the two clans were
traditional enemies in war, vengeance would be taken if oppor¬
tunity offered by any clansman of the victim against any rlancm^
of the culprit. Expiatory sacrifices would be made again, but no
further reprisals ensue. Such murders, however, might formerly
have led to war.

A thief caught in flagrante was severely beaten and publicly
disgraced if he were a clansman of the sufferer. If not, his eyes
were put out or he was otherwise mutilated. The disgrace was
considered to be so great that no reprisals would be attempted.
Matrimonial rights are far more jealously guarded than property

270

AFRICAN POLITICAL SYSTEMS

rights, for they are the concern of the whole clan, since all have
leviritic rights to one another’s widows and all children replenish
the clan. This is a consequence of the elaborate differentiation of
Tale society according to agnatic descent, and of the strength of
exogamy as a factor of social cleavage. The abduction of wives—
inconceivable, as we have seen, if ties of clanship exist between the
abductor’s group and the husband’s—was and is regarded as a
serious violation of a clan’s rights. The injured clan would threaten
to suspend ritual co-operation, or to retaliate in kind, or to go to
war, and the lineage elders of the abductor’s clan would imme¬
diately take steps to return the woman. This was indeed the most
frequent cause of both small and large wars in the old days, as it is
of much litigation to-day. Disputes over bride-price debts or over
the possession of children form the largest proportion of cases
brought to chiefs’ courts. Formerly, they were a prolific source of
armed conflicts and of cattle raids. Adultery provokes similar
reactions, though it did not formerly precipitate war, since it does
not usually break up marriage. If the adulterer belongs to the
same clan as the wronged husband, a neighbouring clan, or one
which has any ties with it, a ritual reconciliation is necessary The
lineage heads, sometimes with the aid of chief or tmdaana nego¬
tiate and arrange this. No compensation is exacted.

In all such cases, territorial remoteness from one another or
wide social cleavages between the two groups concerned made it
almost impossible to obtain redress for wrongs. The injured
group had to await an opportunity to retaliate in kind.

In the background there lurked always the ultimate sanction—
the right to resort to self-help, nominally permissible only if there
were no ties between the two groups concerned, but sometimes
employed even against clansmen. The commonest method was by
raiding (pA), especially if claims to goods or livestock were at
issue The creditor alone or aided by members of his lineage,
would, at Ae risk of being shot, seize livestock belonging to any
clansman of the debtor in payment of the debt. The latter would
have to retrieve his loss from the actual debtor by putting pressure

chief or tmdaana for support. He was entitled to receive only the
mX r °, fl K VeSt0Ck ,° riginalIy owed ^ !<« in excess he right

thfcreTt 3 ^ raid; ° r hC might 3 PP eal the head of

the creditor s maximal lineage, through an intermediary, to order

271

THE TALLENSI

the return of the excess livestock raided in the interest of peace
between neighbours—an appeal as likely to fail as to succeed.

Tale jural notions and procedures are in conformity with the
elaborately segmented character of the social structure. As there
was formerly no completely dominant social unit or association,
there could be no constituted legal machinery backed by irre¬
sistible force. Every region of Tale society, from the joint family
to the whole vaguely delimited aggregate known as the Tallensi
exhibits a dynamic equilibrium—of like units balanced against one
another, of counterpoised ties and cleavages, of complementary
institutions and ideological notions. At every level of Tale
social organization—kinship, clanship, economic relations, local
relations, and the nexus of ritual interdependencies—the ten¬
dency towards an equilibrium is apparent. Overlapping and
interlocking, these different orders of social relations reinforce
one another. The principal mechanism by means of which this
equilibrium is maintained is the balanced distribution of authority
and prerogative, on the one hand, and of obligations and responsi¬
bilities—economic, jural, moral and ritual—on the other. Through
this mechanism the component elements of any segment of the
society control one another.

This does not mean that Tale society was ever stagnant. Tension
is implicit in the equilibrium. It might explode violently when the
specific interests of a unit were violated. But conflict could never
develop to the point of bringing about complete disintegration.
The homogeneity of Tale culture, the undifferentiated economic
system, the territorial stability of the population, the network of
kinship ties, the ramifications of clanship, and especially the
mystical doctrines and ritual practices determining the native
conception of the common good—all these are factors restricting
conflict and promoting the restoration of equilibrium. War was
the ultimate sanction against the violation or submergence of the
specific rights of the corporate units constituting Tale social
structure, and the ties of ritual collaboration the sanction prevent-
ing the complete disintegration of this structure into anarchically
independent fractions. Social relationships in Taleland fluctuate
between amity and discord, co-operation and conflict, for forces
engendering both are always active; but in the long run an equi¬
librium is maintained. The political system of the Tallensi hinges
on this principle.

THE NUER OF THE SOUTHERN SUDAN
By E. E. Evans-Pritchard

T WRITE shortly of the Nuer because I have already recorded a
considerable part of my observations on their political constitu¬
tion and the whole is about to be published as a book . 1 They have
nevertheless, been included in this volume for the reasons that
their constitution is representative of East Africa and that it pro¬
vides us with an extreme political type.

i. Lhstrumtton

To discover the principles of their anarchic state we must first
review bnefly the oecology of the people: their means of livelihood
Aeir distribution, and the relation of these to their surroundings.’
ihe Nuer practise cattle-husbandry and agriculture. They also
fish, hunt and collect wild fruits and roots. But, unlike the other
sources of their food supply, cattle have more than nutritive in-
terest being indeed of greater value in their eyes than anything
else. So, although they have a mixed economy, Nuer are pre-
dommantly pastoral in sentiment.

Nuerland is more suited for stock-breeding than for agricul-
re: itis flat, clayey, savannah country, parched and bare during the
fought and flooded and covered with high grasses during 5 the
ms. Heavy rain falls and the nvers overflow their banks from
June to December. There is little rain and the rivers are low from
Ju “:. The thus comprises two seasons of about

T jT S f sonaI dichotom y> combined with pastoral
interests, profoundly affects political relations.

kuST^V^ N T Uv f in vilk ^ P erched « backs of
“ an A ndgCS or do 1 tted over stretches of slightly elevated
ground, and engage m the cultivation of millet and maize. The
country which intervenes between village and village, being more

1933 to 1938 Papers in Sudan Notes and Records from

inately by the <£££££ 0 72 J^. ?*&**» and was financed
a Leverhulme Fellowship Rather , 8 °‘ Egyptla ® Sudan and partly through
already SZfe ^ than describe again what I have

form than would be Dermiwihl V * Pre *f nte ^ my materla l in a more abstract
man would be pernusstble were a descriptive account not accessible.

THE NUER OF THE SOUTHERN SUDAN 2 y 3

or less flooded for six months, is then unsuitable for habitation
agriculture, or grazing. Anything from five to twenty miles may
separate neighbouring villages, while greater distances may divide
sections of a tribe and tribe from tribe.

At the end of the rams, the people bum the grasses to provide
new pasture and leave their villages to reside in small camps.
When the drought becomes severe, the inmates of these inter¬
mediate camps concentrate on permanent water supplies. Although
these moves are made primarily for the sake of the cattle, they also
enable the Nuer to fish, which is generally impossible from village
sites, and, to a lesser degree, to hunt and collect wild fruits and
roots. When the rains set in again, they return to their villages,
where the cattle have protection and the higher ground permits

The distribution of the Nuer is determined by the physical
conditions and mode of life we have outlined. During the rains,
villages are separated, though by no means isolated, from their
neighbours by flooded stretches of grassland, and local com¬
munities are therefore very distinct units. During the drought,
people of different villages of the same district eventually concen-
trate on permanent water-supplies and share common camps. On
the other hand, some families of a village may go to one camp and
some to another, though the majority form a local community
throughout the year.

Nuer seldom have a surplus of food and at the beginning of the
rains it is often insufficient for their needs. Indeed, it may be said
that they are generally on the verge of want and that every few
years they face more or less severe famine. In these conditions, it
is understandable that there is much sharing of food in the same
village, especially among members of adjacent homesteads and
hamlets. Though at any time some members may have more
cattle and grain than others, and these are their private posses¬
sions, people eat in one another’s homesteads at feasts and at daily
meals, and food is in other ways shared, to such an extent that one
may speak of a common stock. Food is most abundant from the
end of September to the middle of December in a normal year,
and it is during these months that most ceremonies, dances, &c ’
take place.

The Nuer have a very simple technology. Their country lacks
iron and stone and the number and variety of trees are small, and

*74 AFRICAN POLITICAL SYSTEMS

they are generally unsuited for constructive purposes other than
building. This paucity of raw materials, together with a meagre
food supply, contracts social ties, drawing the people of village or
camp closer, in a moral sense, for they are in consequence highly
interdependent and their pastoral, hunting, fishing, and, to a lesser
degree, their agricultural activities are of necessity joint under¬
takings. This is especially evident in the dry season, when the
cattle of many families are tethered in a common kraal and driven
as a single herd to the grazing grounds.

Thus, while in a narrow sense the economic unit is the house¬
hold, the larger local communities are, directly or indirectly, co¬
operative groups combining to maintain existence, and corpora¬
tions owning natural resources and sharing in their exploitation.
In the smaller local groups the co-operative functions are more
direct and evident than in the larger ones, but the collective func¬
tion of obtaining for themselves the necessities of life from the
same resources is in some degree common to all local communities
from the household to the tribe.

These local communities are the monogamous family attached
to a single hut, the household occupying a single homestead, the
hamlet, the village, the camp, the district, tribal sections of varying
size, the tribe, the people, and the international community the
limits of which are a Nuer’s social horizon. We regard the family,
the household, and the hamlet as domestic, rather than political'
groups, and do not discuss them further in detail.

The distribution of these local communities is very largely deter¬
mined by physical conditions, especially by the presence of ground
which remains above flood-level in the rains, and of permanent
water which survives the drought. In any village, the size of
population and the arrangement of homesteads is determined by
the nature of the site. When perched on an isolated knoll, home¬
steads are crowded together; when strung out along a ridge, they
a ^ e , m ° re Wldel y separated; and when spread over a broad stretch
of higher ground, several hundred yards may intervene between
one hamlet and the next. In any large village, the homesteads
are grouped m clusters, or hamlets, the inmates of which are
generally close kinsmen and their spouses. It is not possible to
give more than a rough indication of the size of a village popula¬
tion, but it may be said to vary from 50 to several hundred souls.
As explained, villages are separated by several miles of savannah.

375

THE NUER OF THE SOUTHERN SUDAN

An aggregate of villages lying within a radius which allows easy
inter-communication we call a ‘district’. This is not a political
group, for it can only be defined in relation to each village since
the same villages may be included in more than one district; and
we do not regard a local community as a political group unless
the people who comprise it speak of themselves as a community
by contrast with other communities of the same kind and are so
regarded by outsiders. Nevertheless, a district tends to coincide
with a tertiary tribal section and its network of social ties are what
gives the section much of its cohesion. People of the same district
often share common camps in the drought and they attend one
another s weddings and other ceremonies. They intermarry and
hence establish between themselves many affinal and cognatic
relationships which, as will be seen later, crystallize round an
agnatic nucleus.

Villages, the political units of Nuerland, are grouped into tribal
sections. There are some very small tribes to the west of the Nile
which comprise only a few adjacent villages. In the larger tribes
to the west of the Nile and in all the tribes to the east of it, we
find that the tribal area is divided into a number of territorial
sections separated by stretches of unoccupied couiitry, which

intervene also between the nearest habitations of contiguous
tribes. * &

As all Nuer leave their villages to camp near water, they have
a second distribution in the dry season. When they camp along
a river, these camps sometimes succeed one another every few
miles, but when they camp around inland pools, twenty to thirty
miles often separate one camp from the next. The territorial
principle of Nuer political structure is deeply modified by seasonal
migration. People who form separate village communities in the
rains may unite in a common camp in the drought. Likewise,
people of the same village may join different camps. Also, it is
often necessary, in the larger tribes, for members of a village to
traverse wide tracts of country, occupied by other village com¬
munities, to reach water, and their camp may lie close to yet other
villages. To avoid the complete loss of their herds by rinderpest
or some other misfortune, Nuer often distribute the beasts in
several camps.

In western Nuerland, where the tribes are generally smaller
than to the east of the Nile, there is usually plenty of water and ‘

376 AFRICAN POLITICAL SYSTEMS

pasturage, and it is possible, therefore, for village communities
of the rains to maintain a relative isolation in the drought. But
where, as in the Lou tribe, for example, scarcity of water and
pasturage compels more extensive movement and greater con¬
centration, people who are very widely distributed may have more
social contact with one another than is the case in western Nuer-
land. The isolation and autonomy of local communities are broken
up by economic necessity and the size of the political group is
thereby enlarged. This fact has to be considered in relation to
the further fact that to the east of the Nile wider stretches of
elevated ground allow larger local concentrations in the rains than
is usual to the west of that river. Moreover, seasonal concentration
offers an explanation, though by no means a full one, of the loca¬
tion of tribal boundaries, since they are determined not only by
the distribution of villages, but also by the direction in which the
people turn in their move to dry season pastures. Thus the tribes
of the Zeraf Valley fall back on the Zeraf River and therefore do
not share camps with the Lou tribe, and that part of the Lou tribe
which moves east and north-east make their camps on the Nyand-
ing River and on the upper reaches of the Pibor and do not share
their waters and pasture with the Jikany tribes, who move to the
upper reaches of the Sobat and the lower reaches of the Pibor.
Furthermore, that some of the larger Nuer tribes are able to
preserve a degree of tribal unity without governmental organs
may in part be attributed to seasonal migration, since, as explained
above, the different local sections are forced by the severity of
the latitude into mutual contact and develop some measure of
forbearance and recognition of common interests.

Likewise, a tribal section is a distinct segment, not omy because
its villages occupy a well-demarcated portion of its territory, but
also in that it has its unique dry-season pastures. The people of
one section move off in one direction and the people from an
adjoining section move off in a different direction. Dry-season
concentrations are never tribal, but always sectional, and at no
time and in no area is the population dense.

The total Nuer population is round about 300,000. I do not
know the total square mileage of the country, but to the east of
the Nile, where there are, on a rough estimate, some 180,000 Nuer,
they are said to occupy 26,000 square miles, with the low density
of about seven to the square mile. The density is probably no

THE NUER OF THE SOUTHERN. SUDAN 277

higher to the west of the Nile. Nowhere is there a high degree of
local concentration.

Although dry-season movement produces more social inter¬
relations between members of different tribal sections than the
rainy season distribution might lead us to expect, these contacts
are mainly individual or, when they concern groups, only smaller

local communities, and not the larger tribal sections, are brought
into association. This is probably one of the reasons for the lack
of structural complexity and of great variation of types of social
relations among the Nuer. Outside small kinship groups and
village and camp communities, there are no co-operative economic
combinations and there are no organized ritual associations.
Except for occasional military ventures, active corporate life is
restricted to small tribal segments.

278

AFRICAN POLITICAL SYSTEMS

II. Tribal System.

What is a Nuer tribe? The most obvious characteristic is its
territorial unity and exclusiveness, and this was even more marked
before European conquest than to-day. The population of a tribe
v ^ n 5 s r. 0 7 m a . few hundreds among some small tribes to the west
ot the Nile—if these are rightly regarded as tribes, for very little
research was conducted in that area—to many thousands. Most
tabes have a population of over 5,000 and the largest number
betaeen 30,000 and 45,000 souls. Each tribe is economically
self-sufficient, having its own pastures, water-supplies, and fishing
reservations, which its members alone have the right to exploit
ft has a name which is the symbol of its distinction. The tribes-
men have a sense of patriotism: they are proud to be members
of their tribe and they consider it superior to other tribes. Each
tribe has within it a dominant clan which furnishes a kinship

r ^, e J° rk ° n Wh ' ch t)* political aggregate is built up. Each also
egulates independently its age-set organization.

None of the above-mentioned attributes clearly make a formal
distinction between a tribe and its divisions. The simplest defini-

that? “r 3 tnbC 18 1116 largest commun ity which considers
Jrr hi™ ltS members should be settled by arbitration
nd that it ought to combine against other communities of the
same kind and against foreigners. In these two respects there is

no larger political group than the tribe and all smaller political
groups are sections of it. pouucai

Within a tribe there is law: there is machinery for settling

I a mltus “n t0 C ° ndude ™ner or S

/ ?u a fe l0W tnbesman > it is possible to prevent, or
curtail, a feud by payment of cattle. Between tribe and tribe there
is no means of bringing together the parties to a dispute and
compensation is neither offered nor demanded. Thus if a man

* n of “ oth “ ^ d,,;r

ftuiTthk'aT? 1 ™ rfare - 11 be su PP°sed that

contror o verr,t l p- are t0 COnclude ‘ There is considerable
control over retaliation within a village, but the larger the local

ommunity the more difficult settlement becomes § When two

toge divisions of a .rib. are concerned in a feud. Ae iTces™

ir“vS«thTd T d Settlement ■“ The force of

vanes wth the distance ,n tribal structure that separates the

THE NUER OF THE SOUTHERN SUDAN 279

persons concerned. Nevertheless, so long as a sense of community
endures and the legal norm is formally acknowledged within a
tribe, whatever may be the inconsistencies and contradictions that
appear m the actual relations between tribesmen, they still consider
themselves to be a united group. Then either the contradiction
of feuds is felt and they are settled, the unity of the tribe being
maintained thereby, or they remain so long unsettled that people
give up all hope and intention of ever concluding them and finally
cease to feel that they ought to be concluded, so that the tribe
tends to split and two new tribes come into being.

Nor must it be supposed that the political limits of the tribe are
the limits of social intercourse. People move freely all over Nuer-
land and are unmolested if they have not incurred blood-guilt.
They marry and, to a small extent, trade across tribal boundaries,
and pay visits to kinsmen living outside their own tribe Many
socialrelations, which are not specifically political, link members
of different tribes. One has only to mention that the samp plana
are found in different tribes and that everywhere the age-sets are
co-ordinated. Any Nuer may leave his tribe and settle in a new
tribe, of which he thereby becomes a member. In time of peace,
even Dinka foreigners may visit Nuerland unharmed. Moreover,’
we must recognize that the whole Nuer people form a single
community, territorially Unbroken, with common culture and
feeling of exclusiveness. Their common language and values
permit ready inter-communication. Indeed, we might speak of
the Nuer as a nation, though only in a cultural sense, for there is
no common political organization or central administration.

Besides being the largest group in which legal obligation is
acknowledged, a tribe is also the largest group which habitually
combines for offence and defence. The younger men of the tribe
went, till recently, on joint raiding expeditions against the Dinka
and waged war against other Nuer tribes. Raids on the Dinka
were very frequent; war between tribes less so. In theory, if
two sections of different tribes were engaged in hostilities, each
could rely on the support of the other sections of the same tribe,
but in practice they did not always join in. Contiguous tribes
sometimes combined against foreigners, especially against the
Dinka, though there was no moral obligation to do so, the alliance
was of short duration, and the allies conducted their operations
independently, even when in collaboration.

2§0

AFRICAN POLITICAL SYSTEMS

At the present time, Nuer are to the west and south bordered by
Dinka, who appear to have very much the same kind of political
system as their own, i.e. they comprise a congeries of tribes with¬
out centralized government. From the earliest times the Nuer
have been fighting the Dinka and have been generally on the

DINKA NUER SHILLUK ANUAK

offensive. We know that during the first half of the last century
waves of Nuer broke from their homeland to the west of the Nile
on to the Dinka lands to the east of that river and that they con¬
quered and absorbed the inhabitants in most of what is now eastern
Nuerland (the Nuer distinguish between Nath cieng , the home¬
land , or western Nuer, and Nath doar> the ‘migrated’, or eastern
Nuer). Fighting between the two peoples has continued till the
present time but there does not appear, if maps made by early

THE NUER OF THE SOUTHERN SUDAN a8l

travellers are to be trusted, to have been much change of territory
during the last fifty years. This eastwards migration is a fact that
has to be taken into account, with those related earlier, if we wish
to know why the eastern tribes are larger, territorially and numeri¬
cally, than the western tribes, for it may be assumed that the
struggle of conquest and settlement, and absorption of Dinka on
an unprecedented scale, had some effect on the migrating hordes.

To the north the Nuer are in varying degrees of contact with
Arabs, the peoples of the Nuba Hills, the powerful Shilluk king¬
dom, and certain small communities in Darfung (Burun and
Koma); while to the east and south-east they are bordered by the
Galla of Ethiopia, the Anuak, and the Beir. Wherever the Nuer
have direct relations with these peoples, they are hostile in character.

Arab slave-raiders from the Northern Sudan intruded here and
there into the more accessible portions of Nuerland in the second
half pf the nineteenth century, but nowhere did they gain the
upper and or, indeed, make a marked impression on the Nuer
who opposed them as strongly as they resisted later the Egyptian
overnment, which undertook no serious operations against
them. The Nuer likewise treated British rule with open dis¬
respect till, as a result of lengthy military operations between 1028
and 1930, their opposition was broken and they were brought
under effective administration. With the exception of this last
episode m their history, the Nuer may be said to have reached in
their foreign relations a state of equilibrium and of mutual
hostility which was expressed from time to time in fighting.

A tribe is divided into territorial segments which regard them¬
selves as separate communities. We refer to the divisions of a
tribe as primary, secondary, and tertiary tribal sections. Primary
sections are segments of a tribe, secondary sections are segments
0 a primary section, and tertiary sections are segments of a
secondary section. A tertiary section is divided into villages and
villages mto domestic groups. A member of Z 2 tertiary division
0 ® sees himself as a member of Z 2 community in relation

to Z 1 , but he regards himself as a member of Y 2 and not of Z*
in relation to Y 1 . Likewise, he regards himself as a member of Y,
and not of Y 2 , in relation to X. He regards himself as a member
of tabe B, and not of its primary section Y, in relation to tribe A.
r®’ , on a structural plane, there is always contradiction in the
definition of a political group, for a man is a member of it in virtue

28 a

AFRICAN POLITICAL SYSTEMS

of his non-membership of other groups of the same type which he
stands outside of, and he is likewise not a member of the same
community in virtue of his membership of a segment of it which
stands in opposition to its other segments. Hence a man counts as
a member of a political group in one situation and not as a member
of it in a different situation, e.g. he is a member of a tribe in
relation to other tribes and he is not a member of it in so far as his
segment of the tribe is opposed to other segments. In studying
the Nuer political constitution, it is therefore essential that we

A B

X Y

DIAGRAM No. II

view it together with those of their enemies as a single political
system, for the outstanding structural characteristic of Nuer
political groups is their relativity. A tribal segment is a political
group m relation to other segments of the same kind, and they
jointly form a tribe only in relation to other Nuer tribes and to
adjacent foreign tribes which form part of their political system,
and without these relations very little meaning can be attached to
the concepts of ‘tribe’ and ‘tribal segment’. That the distinction
and individuality of a political group is in relation to groups of the
same kind is a generalization that embraces all Nuer local com¬
munities, from the largest to the smallest.

The relation between tribes and between segments of a tribe

THE NUER OF THE SOUTHERN SUDAN 283

•which gives them political unity and distinction is one of opposi¬
tion. Between tribes, or federations of tribes, and foreign peoples
this opposition is expressed, on the Nuer side at any rate, by con¬
tempt and persistent raiding, often carried out in a reckless and
brutal manner. Between Nuer tribes, opposition is expressed by
actual warfare or by acceptance that a dispute cannot, and ought
not, to be settled m any other way. In intertribal warfare, however,
women and children are neither speared nor enslaved. Between
segments of the same tribe, opposition is expressed by the institu-
tion of the feud. A fight between persons of the same village or
camp is as far as possible restricted to duelling with clubs. The
hostility and mode of expression in these different relations varies
m degree and in the form it takes.

Feuds frequently break out between sections of the same tribe
and they are often of long duration. They are more difficult to
settle the larger the sections involved. Within a village feuds are
easily settled and within a tertiary tribal section they are concluded
sooner or later, but when still larger groups are involved they may
never be settled, especially if many persons on either side have
been killed in a big fight.

A tribal section has most of the attributes of a tribe: name, sense
of patriotism, a dominant lineage, territorial distinction, economic
resources, and so forth. Each is a tribe in miniature, and they differ
from tribes only in size, in degree of integration, and in that they
unite for war and acknowledge a common principle of justice.

The strength of the sentiment associated with local groups is
roughly relative to their size. Feeling of unity in a tribe is weaker
than feeling of unity within its sections. The smaller the local
group, the more the contacts its members have with one another
and the more these contacts are co-operative and necessary for the
maintenance of the life of the group. In a big group, like the tribe,
contacts are infrequent, short, and of limited type. Also the
smaller the group the closer and the more varied the relationships
between its members, residential relations being only one strand
in a network of agnatic, cognatic, and affinal relationships.
Relationships by blood and marriage become fewer and more
distant the wider the group.

It is evident that when we speak of a Nuer tribe we are using a
relative term, for it is not always easy to say, on the criteria we
have used, whether we are dealing with a tribe with two primary

284 AFRICAN POLITICAL SYSTEMS

segments or with two tribes. The tribal system as defined by
sociological analysis can, therefore, only be said to approximate
o any simple diagrammatic presentation. A tribe is an exempli
ncation of a segmentary tendency which is characteristic of the
political structure as a whole. The reason why we speak of
Nuer political groups, and of the tribe in particular, as relative
groups and state that they are not easily described in terms of
potacai morphology, is that political relations are dy^if
ey are always changing In one direction or another. The most
evident movement is towards fission. The tendency of tribes and
triba! sections towards fission and internal opposition between
eir parts is balanced by a tendency in the direction of fusion
the combination or amalgamation of groups. This tendency
towards fusion is inherent in the segmentary character of Nuer
political structure, for, although any group tends to split into
opposed parts, these parts tend to fuse in relation to othe^groups
Hence fission and fusion are two aspects of the same segmentary

stood P C an< ! Ae ^ Uer tribe 311(1 its divisions are to be under^
stood as a relation between these two contradictory, yet comple¬
mentary tendencies. Physical environment, way NeHh 00 d
mode of distribution, poor communications, staple'^economy
&c to some extent explain the incidence of politicafcleavage but

the tendency towards segmentation seems to be inherent in
political structure itself. mnerent in

III. Lineage System

Tribal unity cannot be accounted for by any of the facts we

rrr d ' taten alone ° r in i;

reterence to the lineage system. The Nuer clan is not an un

ffirsS.ES;1 p t r

luiibmp, as are many African clans, but is hkhlv *e*m^t*A mu

s~4^t‘rr s ' md - ” d Jt

the NUER of THE SOUTHERN SUDAN 3 g 5

into major lineages D, E, F and O Tn

S'dTM'N’ J ’ d"^ K

and L M, N, and O are mimmal lineages which are segments of
mmor lineages H and J. The whole dan is a genealogicau“re
i.e. the letters represent persons to whom the clan and its segments
trace their descent and from whom they often take theLSTes
There must be at least twenty such clans in Nuerland wiAotit
taking into account many small lineages of Dinka origin

e Nuer lineage is a group of agnates, and comprises all living

1

D

B A r

E 1

F Q

1

fj

1 J

1 K |

0

i n

DIAGRAM No. Ill
persons descended, through males only, from the founder of that

from7h ar f ne '/T aIIy ’ k ak0 indudes dead Persons descended
from the founder, but these dead persons are only significant in

the lidS g Tr ° gl ? P ° Slti0n f Xplains . the relationship between
, , , Wlder agnatic kinship is recognized the further

back descent has to be braced, so that the depth of a lineage is always

m proportion to its width. s. always.

chJrart NU ? ^ ^8% segmented, has many of the

are dS-Ino, T ^ f ° Und intribal stru cture. Its lineages

are distinct groups only in relation to each other. Thus, in the

diagram, M is a group only by opposition to L, H is a group only

TW? 81 T n t0 /’ DlS a / r ° Up onl y b y position to E, and so on.
There is always fusion of collateral lineages of the same branch in

Wr?el E f f? Ch ’ e ‘ g - in thC dia § ram ’ LandM are no

onger separate minimal lmeages, but are a single minor lineage,
H, m opposition to I, and D and E are no longer separate major
meages, but are a single maximal lineage, B, in opposition to C.

rektionT ? ea f S W ? Ch arC e( 3 ual and OPPOSE are composite in
elation to a third, so that a man is a member of a lineage in relation

386

AFRICAN POLITICAL SYSTEMS

to a certain group and not a member of it in relation to a different
group. Lineages are thus essentially relative groups, like tribal
sections, and, like them, also are dynamic groups. Therefore they
can only be described satisfactorily in terms of values and situations.

Nuer lineages are not corporate localized communities though
their members often have an association with a locality and speak
of the locality as though it were an exclusive agnatic group. Every
Nuer village is associated with a lineage, and though the members
of it often comprise a small proportion of the community, it is iden¬
tified with them in such a way that we may speak of it as an aggre¬
gate of persons clustered around an agnatic nucleus. The aggre¬
gate is linguistically identified with the nucleus by the designation
of the village community by the name of the lineage. It is only in
reference to rules of exogamy and certain ritual activities that one
needs to regard lineages as completely autonomous groups. In
social life generally, they function within local communities, of all
sizes from the village to the tribe, and as part of them. We cannot
here discuss the ways by which residential groups become a net¬
work of kinship ties—marriage, adoption, and various fictions—
but the result tends to be that a local group is a cognatic cluster
round an agnatic core, the rules of exogamy being the operating
principle m this tendency. 6

Nuer clans are everywhere much dispersed, so that in any
village or camp one finds representatives of diverse clans. S mall
lmeages have moved freely over Nuerland and have settled here
and there and have aggregated themselves to agnatically unre-
Iated elements in local communities. Migration and the absorption
° Dmka have been circumstances favouring the dispersal and
mixture of clans. Being a conquering, pastoral people and not
havmg an ancestral cult, the Nuer have never been bound to any
particular spot by necessity or sentiment.

Nevertheless there is a straight relation between political
structure and the clan system, for a clan, or a maximal lineage, is
associated with each tribe, in which it occupies a dominant position
among other agnatic groups. Moreover, each of its segments tends
to be associated with a segment of the tribe in such a way that there
is a correspondence, and often a linguistic identification, between
the parts of a clan and the parts of a tribe. Thus if we compare

tribf B nS tE Iand 111 an , d , SUpP ° Se cIanAt0 be the dominant clan in
tribe B, then maximal lineages B and C correspond to primary

THE NUER OF THE SOUTHERN SUDAN 287

sections X and Y; major lineages D and E correspond to secondary
section X^and X 2 ; major lineages F and G correspond to secondary
sections Y 1 and Y 8 ; and minor lineages J and K correspond to
tertiary sections Z 1 and Z 2 . r

We speak of a clan which is dominant in a tribe as the aristocratic
clan although except on the periphery of Nuer expansion east¬
wards, its predominance gives prestige rather than privilege. Its
members are in a minority-often a very small minority-in the
tribe. Not all members of the aristocratic clan live in the tribe
where it is dominant, but many are also found in other tribes Not
all clans are associated with a tribe in this manner. A man is only
an aristocrat m the one tribe in which his clan is dominant. If he
lives in another tribe, he is not an aristocrat there.

There is consequently in every tribe some social differentiation.
There are aristocrats, Nuer of other clans, and Dinka, but these
strata are not classes and the second and third are properly to be
regarded as categories rather than as groups. The Dinka who have
been absorbed into Nuer society have been for the most part incor¬
porated into their kinship system by adoption and marriage, and
conquest has not led to the development of classes or castes. This
is, perhaps, to be attributed, in part at any rate, to the fact that the
Dinka, like the Nuer, are chiefly pastoralists and that in other
respects their ways of life are very similar.

Without presenting all the evidence and without making every
qualification, we attempt an explanation of why Nuer clans, espe¬
cially the dominant clans, are segmented into lineages to a far
greater degree than is usual among African peoples. In our view,
they are segmented because the political structure to which they
correspond is segmented in the way we have described. Social
obligations among the Nuer are expressed chiefly in a kinship
idiom and the interrelations of local communities within a tribe
are defined in terms of agnatic relationship. Therefore, as the tribe
segments the clan segments with it and the point of separation
between the tribal sections becomes the point of divergence in the
clan structure of the lineages associated with each section. For, as
we have seen, clans and their lineages are not distinct corporate
groups, but are embodied in local communities, through which
they function structurally. Such being the case, it is not surprising
that they take the form of the State which gives them corporate
substance.

288 AFRICAN POLITICAL SYSTEMS

Those clans which are associated with tribes have generally
greater lineage extension and depth than those which are not so
associated, and the larger the tribe the more significance this
association has for the Nuer. It is in the largest tribes, territorially
and numerically, and those which have expanded most and
assimilated most foreigners, like the Lou and Eastern Gaajak and
Gaajok tribes, that we find the greatest attention paid to the dis¬
tinct and dominant position of the aristocratic clans. Indeed, not
only do political relations affect the clan structural form, split¬
ting it into segments along the lines of political fission, but also the
clan system may be said to have a corresponding action on the
political structure. In a confusion of lineages of different clan
origin and in an amorphous network of cognatic links, the political
structure is given consistent form, in the language of kinship, by
one clan—a single system of lineages—being made to correspond
to the tribe and to its structure of opposed segments. Just as a mari
is a member of a tribal segment opposed to other segments of the
same order and yet also a member of the tribe which embraces all
these segments, so also he is a member of a lineage opposed to
other lineages of the same order and yet also a member of the clan
which embraces all these lineages, and there is a strict correspon¬
dence between these two sets of affiliations, since the lineage is
embodied in the segment and the clan in the tribe. Moreover, the
distance in clan structure between two lineages of a dominant
clan tends to correspond to the distance in tribal structure between
the two sections with which they are associated. Thus the system
of lineages of the dominant clan enables the Nuer to think of their
tribe in the highly consistent form of clan structure. In each
segment the network of kinship ties are given unity and coherence
by their common relationship to the lineage of the dominant clan
that resides there, and as these separate lineages are composite in
relation to other clans so the whole tribe is built around an exclu¬
sive agnatic framework. Though the sections may tend to draw
apart and to split, a common agnatic value, shared by the dominant
lineages contained in them, endures.

IV. Age-set System

Another tribal institution is the age-set system, which is socially
more significant among the Nuer than among other Nilotic peoples
of the Sudan. Nuer boys pass into the grade of manhood through

THE MUER OF THE SOUTHERN SUDAN 289

a severe ordeal and a series of rites connected with it. These initia¬
tions take place whenever there are a . sufficient number of boys of
from about fourteen to sixteen years of age in a village or district
All the youths who have been initiated in a successive number of
years belong to one age-set, and there is a four-year interval be¬
tween the last batch of initiates of one set and the first batch of the
next set, and during this interval no boys may be initiated. The
initiation period is open for about six years, so that, with the four
years of the closed period, there are about ten years between the
commencement of any age-set and the commencement of the set
that precedes or succeeds it. The age-sets are not organized in
a cycle.

■ Nuer age-sets are a tribal. institution in the sense that, in the
larger tribes at any rate, all the sections of a tribe have the same
open and closed periods and call the sets by the same names. They
are also the most characteristic of all Nuer national institutions,
for initiation scars are the sign of their communion and the badge
of their supremacy. Moreover, though each big tribe has its own
age-set organization, adjacent tribes co-ordinate their sets in
periods and nomenclature, so that the Western Nuer, the Eastern
Nuer, and the Central Nuer tend to fall into three divisions in
this respect. But even when a man travels from one end of Nuer-
land to the other, he can always, and easily, perceive the set which
is equivalent to his own in each area. The age-set system, there¬
fore, like the clan system, whilst having a tribal connotation, is
not bound by lines of political cleavage.

There is usually in each tribe a man whose privilege it is to
open and close initiation periods and to give each set its name.
This man belongs to one of those lineages which have a special
ritual relationship to cattle and are known as ‘Men of the Cattle’.
He opens and closes initiation periods in his own district, and
other districts of his tribe follow suit. Once a period has been
opened, each village and district initiates its boys when it pleases.
The age-sets have no corporate activities and cannot be said to
have specific political functions. There are no grades of ‘warriors’
and ‘elders’ concerned with, the administration of the country, and
the sets are not regiments, for a man fights with the members of
his local community, irrespective of age. In the rites of initiation
there is no educative or moral training. There is no leadership
in the sets.

290

AFRICAN POLITICAL SYSTEMS

There are probably never more than six sets in existence at
any time, since six sets cover about seventy-five years. As each
set dies its name is remembered only for a generation or two.
Each set becomes more senior as the years go on, so that a man
rises from a junior to a middle, and from a middle to a senior
position in his community as a member of a group. The stratifica¬
tion of the age-set system is thus a further exemplification of the
principle of segmentation which we have seen to be a characteristic
of the political and kinship systems. There is further stratification
within each set, but this is not of great importance, for the set
sees itself, and is seen by others, as an undivided group in relation
to other sets, and its divisions become merged as the set becomes
more senior. A set once complete does not change its membership,
but the sets are constantly changing their positions in relation to
the whole system. There is also a certain relativity about these
stratified sets similar to that we noted about tribal sections and
clans, for, while they keep their distinction, there is often a
situational fusion of two sets in relation to a third. This is
especially apparent at feasts. Whether a set is regarded as
junior or equal depends not only on its position in the age-set
structure, but also on the status of a third set concerned in any
situation, a tendency due to the connexion between age-sets and
generations.

The most evident action of age-sets in determining behaviour
is the way duties and privileges are effected by a transition from
boyhood to manhood. Also, in virtue of the position of his set in
the structure, every male Nuer is in a status of seniority, equality,
or juniority towards every other Nuer man. Some men are his
‘sons’, some his ‘brothers’, and some his ‘fathers’. Without
entering here into further detail, we may say that the attitude of
a man towards other men of his community is largely determined
by their respective positions in the age-set system. Hence age
relations, like kinship relations, are structural determinants of
behaviour. The age-set system may, moreover, be regarded as
a political institution, since it is, to a large extent, segmented
tribally and since it divides a tribe—as far as its male members are
concerned—into groups, based on age, which stand in a definite
relation to each other. We do not consider, however, that it has
any direct accord with the tribal structure, based on territorial
segmentation, which we have recorded. The politico-territorial

THE NUER OF THE SOUTHERN SUDAN 291

system and the age-set system are both consistent in themselves
and to some extent overlap, but they are not interdependent.

V. Feuds and other Disputes

The political system operates largely, we think, through the
institution of the feud which is regulated by a mechanism known
as the ‘leopard-skin chief’, a title we retain, although the appella¬
tion of chief is misleading. This person is one of those specialists
who are concerned, in a ritual capacity, with various departments
of Nuer social life and of nature. Leopard-skin chiefs belong to
certain lineages only, though not all members of these lineages
utilize their hereditary ritual powers. In most of Nuerland, the
lineages are not branches of dominant clans.

When a man has killed another, he must at once go to a chief,
who cuts his arm so that the blood may flow. Until this mark of
Cain has been made, the slayer may neither eat nor drink. If he
fears vengeance, as is normally the case, he remains at the chief’s
home, for it is sanctuary. Within the next few months the chief
elicits from the slayer s kin that they are prepared to pay com¬
pensation to avoid a feud and he persuades the dead man’s kin
that they ought to accept compensation. During this period
neither party may eat or drink from the same vessels as the other
and they may not, therefore, eat in the home of the same third person.
The chief then collects the cattle—till recently some forty to fifty-
beasts—and takes them to the dead man’s home, where he per¬
forms various sacrifices of cleansing and atonement. Such is the
procedure of settling a feud, and before the present administration
it had often to be used, for the Nuer are a turbulent people who
esteem courage the highest virtue and skill in fighting the most
necessary accomplishment.

In so brief a description, one may give the impression that the
chief judges the case and compels acceptance of his decision.
Nothing could be further from the facts. The chief is not asked
to deliver a judgement: it would not occur to Nuer that one was
required. He appears to force the kin of the dead man to accept
compensation by his insistence, even to the point of threatening
to curse them, but it is an established convention that he gball
do so, in order that the bereaved relatives may retain their prestige.
What seems really to have counted were the acknowledgement of
community ties between the parties concerned, and hence of the

AFRICAN POLITICAL SYSTEMS

29a

moral obligation to settle the affair by the acceptance of a tradi¬
tional payment, and the wish, on both sides, to avoid, for the
time being at any rate, further hostilities.

A feud directly affected only close agnatic kinsmen on both
sides. One did not avenge oneself on cognates or on distant
agnates.. Nevertheless, we believe that the feud had a wider
social connotation and that therein lies its political significance.
We must first recognize that feuds are more easily settled the
smaller the group involved. When a man kills a near kinsman or a
close neighbour, the matter is quickly closed by compensation,
often on a reduced scale, being soon offered and accepted, for
when a homicide occurs within a village general opinion demands
an early settlement, since it is obvious to every one that were
vengeance allowed corporate life would be impossible. At the
other end of the scale, when a homicide occurs between primary
or secondary sections of a tribe, there is little chance of an early
settlement and, owing to distance, vengeance is not easily achieved,
so that unsettled feuds accumulate. Such homicides are generally
the result of intertribal fights in which several persons are killed.
This not only increases the difficulty of settlement, but continues
between the sections the mutual hostility that occasioned the fight,
for, not only the close agnatic kinsmen of the dead, but entire
local communities are involved. Feud, as a -choice between direct
vengeance and acceptance of compensation, without the necessity
of immediate settlement, but requiring eventual conclusion, is
especially a condition that flourishes between villages of the same
district. The kinsmen of the dead man are near enough to strike
at the kinsmen of the slayer and far enough from them to permit
a temporary state of hostility between the local communities to
which the parties belong. For whole communities are of necessity
involved, though they are not subject to the rigid taboos that a
homicide imposes on close agnatic kinsmen of slayer and slain,
nor are they threatened with vengeance. Nevertheless, their members
are, as a rule, closely related by cognatic or affinal ties to the
principals and must assist them if there is an open fight. At the
same time, these communities have frequent social contacts, iso
that eventually the mechanism of the leopard-skin chief has to be
employed to prevent their complete dislocation. The feud thus
takes on a political complexion and expresses the hostility between
political segments.

THE NUER OF THE SOUTHERN SUDAN 293

The balanced opposition of political segments is, we believe,
largely maintained by the institution of the feud which permits a
state of latent hostility between local communities, but allows also
their fusion in a larger group. We say that the hostility is latent
because even when a feud is being prosecuted there is no unin¬
terrupted endeavour to exact vengeance, but the kinsmen of the
dead may take any opportunity that presents itself to accomplish
their purpose; and, also, because even when compensation has
been accepted the sore rankles and the feud may, in spite of settle-
ment, break out again, for Nuer recognize that in sentiment a feud
goes on for ever. The leopard-skin chief does not rale and judge,
but acts as mediator through whom communities desirous of end¬
ing open hostility can conclude an active state of feud. The feud,
including the role played in it by the chief, is thus a mechanism by
which the political structure maintains itself in the form known to us.

The leopard-skin chief may also act as mediator in disputes
concerning ownership of cattle, and he and the elders on both
sides may express their opinion on the merits of a case. But the
chief does not summon the defendants, for he has neither court
nor jurisdiction and, moreover, has no means of compelling com¬
pliance. All he can do is to go with the plaintiff and some elders of
his community to the home of the defendant and to ask him and
his kinsmen to discuss the matter. Only if both sides are willing
to submit to arbitration can it be settled. Also, although the chief,
after consultation with the elders, can give a verdict, this verdict
is reached by general agreement and in a large measure, therefore,
arises from an acknowledgement by the defendant’s or plaintiff’s
party that the other party has justice on its side. It is, however,
very seldom that a chief is asked to act as mediator, and there is
no one else who has authority to intervene in disputes, which are
settled by other than legal methods.

In the strict sense of the word, the Nuer have no law. There is
no one with legislative or juridical functions. There are conven¬
tional payments considered due to a man who has suffered certain
injuries—adultery with his wife, fornication with his daughter,
theft, broken limbs, &c.—but these do not make a legal system, for
there is no constituted and impartial authority who decides on the
rights and wrongs of a dispute and there is no external power to
enforce such a decision were it given. If a man has right on his
side, and, in virtue of that, obtains the support of his kinsmen and

294

AFRICAN POLITICAL SYSTEMS

they are prepared to use force, he has a good chance of obtaining
what is due to him, if the parties live near to one another. The
usual way of obtaining one’s due is to go to the debtor’s kraal and
take his cattle. To resist is to run the risk of homicide and feud. It
seems that whether, and how, a dispute is settled depends very
largely on the relative positions of the persons concerned in the
kinship and age-set systems and the distance between their com¬
munities in tribal structure. In theory, one can obtain redress from
any member of one’s tribe, but, in fact, there is little chance of doing
so unless he is a member of one’s local community and a kinsman.
The force of ‘law’ varies with the position of the parties in political
structure, and thus Nuer ‘law’ is essentially relative, like the
structure itself.

During the year I spent with the Nuer, I never heard a case
being conducted, either before an individual or before a council of
elders, and I received the impression that it is very rare for a man
to obtain redress except by force or threats of force. And if the
Nuer has no law, likewise he lacks government. The leopard-
skin chief is not a political authority and the ‘Man of the Cattle’
and other ritual agents (totemic specialists, rain-makers, fetich-
owners, magicians, diviners, &c.) have no political status or func¬
tions, though they may become prominent and feared in their
locality. The most influential men in a village are generally
the heads of joint families, especially when they are rich in
cattle, of strong character, and members of the aristocratic
clan. But they have no clearly defined status or function. Every
Nuer, the product of a hard and equalitarian upbringing, deeply
democratic, and easily roused to violence, considers himself as
good as his neighbour; and families and joint families, whilst
co-ordinating their activities with those of their fellow villagers,
regulate their affairs as they please. Even in raids, there is very
little organization, and leadership is restricted to the sphere of
fighting and is neither institutionalized nor permanent. It is
politically significant only when raids are controlled and organized
by prophets. No Nuer specialists can be said to be political agents
and to represent, or symbolize, the unity and exclusiveness of local
groups, and, apart from the prophets, none can be said to have
more than local prominence. All leaders, in this vague sense of
influential persons in a locality, are adults and, except for an
occasional prophetess, all are men.

THE NUER OF THE SOUTHERN SUDAN 29S

Owing to the fact that Nuer prophets had been the foci of
opposition to the Government, they were in disgrace, and the
rcmre influential of them under restraint or in hiding, during my
visit to Nuerland, so that I was not able to make detailed observa¬
tions on their behaviour. Nuer are unanimous in stating that they
did not arise much before the end of the last century and there is
some evidence to suggest that their emergence was connected
with the spread of Mahdism. However that may be, there can be
no doubt that powerful prophets arose about the time of Arab
intrusion into Nuerland and that at the time of British conquest
they were more respected and had wider influence than any other
persons. No extensive raids were undertaken without their sanc¬
tion and often they led them, received part of the spoil, and to
some extent supervised the division of the rest of it. Though
there seems to be good evidence that the earlier prophets were no
more .than ritual agents, some of the later ones appear to have
begun to settle disputes, at any rate in their own districts. How¬
ever, their chief political importance rather lay elsewhere. For
the first time a single person symbolized, even if in a mainly
spiritual form, the unity of a tribe, for the prophets were essentially
tribal figures, though—and this fact is also of great political signi¬
ficance their influence often extended over tribal boundaries and
brought about a larger degree of unity among adjacent tribes than
there appears to have been hitherto. When we add that there was
a tendency for the spirits which possessed prophets to pass, at
their deaths, into their sons, we are justified in concluding that
development was taking place towards a higher degree of federa¬
tion between tribes and towards the emergence of political leader¬
ship, and in explaining these changes by reference to Arab and
European intrusion. Opposition between Nuer and their neigh¬
bours had always been sectional. They were now confronted by
a more formidable and a common enemy. When the Government
overthrew the prophets, this development was checked.

VI. Summary

We have briefly described and analysed what we regard as Nuer
political structure: the relations between territorial segments
within a territorial system and the relations between that and
other social systems within an entire social structure. We have
examined intertribal relations, and the relations between tribal

2 9 6 AFRICAN POLITICAL SYSTEMS

segments. It is these relations, together with the tribal and inter¬
tribal contacts with foreign peoples, that we define as the Nuer
political system. In social life the political is combined with other
systems, particularly the clan system and the age-set system, and
we have considered what relation they bear to the political struc¬
ture. We have also mentioned those ritual specialisms which have
political significance. The political system has been related to
environmental conditions and modes of livelihood.

The Nuer constitution is highly individualistic and libertarian.
It is an acephalous state, lacking legislative, judicial, and execu¬
tive organs. Nevertheless, it is far from chaotic. It has a persistent
and coherent form which might be called ‘ordered anarchy’. The
absence of centralized government and of bureaucracy in the
nation, in the tribe, and in tribal segments—for even in the village
authority is not vested in any one—is less remarkable than the
absence of any persons who represent the unity and exclusiveness
of these groups.

It is not possible from a study of Nuer society alone, if it be
possible at all, to explain the presence and absence of political
institutions in terms of their functional relationship to other
institutions. At best we can say that certain social characteristics
seem to be consistent. Environmental conditions, mode of
livelihood, territorial distribution, and form of political seg¬
mentation appear to be consistent. So do the presence of clans
with genealogical structure and a developed age-set system seem
to gc together with absence of political authority and of class-
stratification. Comparative studies alone will show whether
generalizations of such a kind are true and, moreover, whether
they are useful. We cannot here discuss these questions and will
only say, in conclusion, that the consistency we perceive in Nuer
political structure is one of process rather than of morphology.
The process consists of complementary tendencies towards fission
and fusion which, operating alike in all political groups by a series
of inclusions and exclusions that are controlled by the changing
social situation, enable us to speak of a system and to say that this
system is characteristically defined by the relativity and opposition
of its segments.

SUPPLEMENTARY STUDIES BY CONTRIBUTORS
TO THIS BOOK

I. Schapera: Tribal Legislation Among the Tstcana of the Bechuanaland
Protectorate. L.S.E. Monographs on Social Anthropology, No. 9,

1943.

Land Tenure in the Bechuanaland Protectorate, Lovedale Press,

1945.

S. F. Nadel: A Black Byzantium, Oxford University Press, 1942.

M. Fortes: The Dynamics of Clanship Among the Tallensi, Oxford
University Press, 1945.

The Web of Kinship Among the Tallensi, Oxford University Press,

* 949 *

E. E. Evans-Pritchard: The Nuer. Clarendon Press, 1941.

G. Wagner: The Bantu of North Kavirondo, Volume I, Oxford Uni¬
versity Press, 1949.

INDEX

Abachwezi, 123 #., 136, 156, 160.
Abahinda dynasty, 124-5, I 59 *
abakungu, 139-40, 146.
abatoro, 131-2, 149-50, 151-2.
abatzvare, 139-40.

Accession war, 140, 143, 156, 158#.
Administration, Bemba, 107-10.
British, and Banyankole, 162.
and Ngwato, 56, 78.
and Nuer, 281.
and Tallensi, 264-6.
and Zulu, 26-7, 36, 46#.
European, 15-16, 65 if.

Ngwato, 78, 80-2.

Tallensi, 240-1.

Age-classes (or age-grades), 188, 209,
210, 213, 214, 234.

Age-sets, Nuer, 288 if.
Agriculturalists and Agriculturalism,
8, 94-5, 121-62, 199, 248, 249.
Ancestor spirits. See Spirits, ancestral.
Ankole, 121-62.

Army, Ngwato, 70, 73.

Zulu, 26.

See also Regiments.

Bagyendamva , 150#., 158, 159.

Bahima and Bairu (or Muhima and
Mwiru), 121 #.

bakabilo (hereditary state councillors),
99, 100, 108-9, no, in, 112,
119-20.

Bakiga, 130.

Bantu, 52, 83-5, 94, 95, 101, 108, no,
121.

Kavirondo, 197-236.

See also Bemba.

Banyankole, 5, 8, 9, 121-62.

Barotse, 116, 120.

Basuto, 26.

Basutoland, 65.

Bechuanaland, 56-82.

Beer, 104, 106, 143, 146, 147, 154,
191, 213, 231.

feasts, 209, 211.

Bemba, 5, 7, 8, 9, 14, 22, 83-120.

Birwa, 57, 58.

Bisa, 86, 94, in.

Blood revenge, 129, 131, 132, 134,
137.

Bride-price (or bride-wealth), 36, 49,
50, 69 w.i, 77, 130, 138, 142,
149, 225.

British South Africa Companv, 113.

Cattle, 33, 38, 45 , 48, 49 , 59 , 76, 77 ~S,
95, 121, 124, 129, 130, 133,
133 , 135-6, 137, 140, 144 , 145 ,
148-50, 152, 153 , 154 , 159 ,
160, 199, 216, 231, 248, 272,
278, 289, 291, 294.
raids, 25, 78, 127, 129, 135, 139,
142-3, 147, 149, 152, 198, 226,
228, 229-30.

Ceremonies, 19, 22, 30, 31, 59, 70, 99,
105, 109, no, in, 212-13,
214, 218, 230, 236, 259.
Accession, 151, 159.

See also Feasts ; Ritual.

Cetshwayo, King of Zululand, 26, 30,
32, 35-6, 38.

Cewa, 86.

Chiefs and Chieftainship, 10, n-12,
13, 25, 30, 56, 13°, 132, 167 et
passim.

and their subjects, 43-4, 49, 62 ff.,
68#., 81-2.

Banyankole, 130, 132, 133, 134,
138,139-40,143,150, 152,158.

abakungUy 139-40, 146.

Bemba, 83, 91, 95, 96, 97, 98, 99#.,

duties of, 104-6.

paramount, 87, 91, 92, 98, 99,
101, 108, 109, no, 112.
sub-chiefs, 91, 101, 103, 108
territorial, 91, 92, 99, 105, 108,
109, 112, 120

Kede, 173, 175, 177#*, 185-6, 191,
192.

‘leopard-skin* chief, 291, 292, 293.
Ngwato, 59, 60, 62, 63, 64-5, 66#.

duties of, 69-70.
paramount, 116, 116 n. 1.
relatives of, 79-80, 92-4, 97, 99-
100, 175, 176, 177 ,

salaries of, 78, 116/ 117, 185.
Tallensi, 255#.
tribute and, 146#.

INDEX

299

Chiefs and Chieftainship— contd.
Zulu, 30#., 4°#-, 54 - 5 *

duties of, 38-9.

See also enganzi; Tribute.

Children, 93, 96, 154, 228.
Christianity, 51, 59, 70, 76, 115-16.
Circumcision, 204, 209, 210, 213, 226.
Clans and Clanship, 122, 137, 138,
146, 241 et passim .

Bantu Kavirondo, 200, 201, 203,

205 #-

sub-clans, 211, 221.

Bemba, 84, 87, 91#.

Nuer, 284j(f.
patrilineal, 25.

sub-clans, 131, 209 et passim.
Tallensi, 242#.

Zulu, 28, 28 n. 1, 29, 30, 37.
Clientship {okutoizha), iz&ff.y 135,
148, 149.

Concubines, 130-1, 141-2.

Councils and Councillors, 12, in.
Bantu Kavirondo, 221.

Banyankole, 139.

Bemba, 108, 109, 119-20.

Kede, 174, 175, 176.

Ngwato, 65, 68, 71.

Zulu, 33, 39-

See also bakabilo; kgotla.

Courts, Native, 32-3, 63-4, 67, 69,
72, 105, 108, 116-17, nS, 181,
181, 220. See also Law.
Crimes, 64, 69, 134, 181-2. See also
Offences.

Cults, 155-6.

ancestor, 200, 234, 253, 254.

Bagyendamva,

Earth, 246, 254-5, 260, 261, 262,
264, 269.

emandwa , 125, 143, 144, 155, 161.
See also Spirits.

Culture, 3, 9-10.

Curses, 204, 220, 235, 259, 261, 269.
Custom, Bantu Kavirondo and, zozff.y
212#.

Descent, dogma of, 96^.

Dingane, King of Zululand, 26, 34,42.
Dinka, 10, 279, 280, 286, 287
Dinuzulu, 26-7, 36, 46.

District Governors, Ngwato, 61, 63,
73 -

Divination and Diviners, 145,149,153,
161, 204, 205, 261.
Dream-prophets, 204, 205, 234*
Drums, royal, 136, 150jff.

Earth, cult of the, 246, 254-5, 260,
261, 262, 264, 269.

Economics (or economy), 8-9.

Kede, 168-70, 188.

See also Wealth.

Education among Ngwato, 56.
ekyibumbe, 158, 159.
ekyitoro, 149-50.

El Kony, 198, 226, 230.

Elders, 180.

Bantu Kavirondo, 204, 209, 211,
212, 220, 221, 222, 229, 231,
232-3, 234, 235-6.

Ngwato, 59, 63.

Tallensi, 256, 266, 267, 269.
emandwa, 123, 143, 144, 155, 161.
engabirano, 146, 147, 149.
enganzi (favourite chief), 137, 138-9,
146, 159 -

Exogamy, 200,207, 222, 243,270,286.
Family:

extended families, 128, 131, 133,
134 -

groups, 58, 59, 63.

See also Kinship; Lineage.

Feasts (or festivals), 208-10, 213, 214,
223, 232, 245, 247, 254, 263.
Great Festivals, the, 245, 254, 259,
262, 263-4.

See also Ceremonies; Ritual.
Fertility in women, 153-4.

Feuds, 41, 76, 223, 278-9, 283, 291-3.
Fines, 29, 34, 38, 59, 74 , 76, 116, 144,
216, 266, 269.

Food, 146,151, 154, 249, 273.

Force, organized, 14-15.

Fulani, 184-5.

Gaajok, 288.

Gifts (or presents), 34, 38, 39, 132,
149, 206, 207-8, 213, 214, 262.
See also Offerings; Tribute.
Gishu, 225.

Gold Coast, the, 239-71.

Gorisi (Nankansi), 239, 244, 246.

Headmen, 58, 59-60, 62, 71, 73, 80,
89, 98, 264.
bafilolo, 107, 108.
bailhanka , 77-8.
common, 77-8.
duties of, 59, 103-4.

Herero, 57, 58.

icalo (district), 91, 92, 105, no, in.
Indirect Rule, 1, 15, 85, 113, 117,
240 n. 2.

indunas, 30, 33, 34, 37 , 38, 39 , 4 °, 4 1 ,
44 *

3 oo

INDEX

Inheritance, 69 n. 1, 231, 243.

chieftainship and, 74, 88, 99-100,

. x 75 *

official offices and, 50, 60, 74-5, 90,
96#., 176.
o<?e o/jo Succession.

Initiation and the ceremonies, 73, 94,
289.

Jaluo, 198.

Jikany, 276

K**, 57 , 58, 76.

Kakalelwa, 230
Kakanda, 170 n. 1.

Kalaka, 57, 58.

Kede (or Kyedye), 5, 7, 8, 9, 22,

165-95-

two groups of, 167-8.

Kenya, 197-236.

Kgalagadi, 57, 58,77.

Kgama III, 56, 61, 69 n. 1, 70, 76, 77,
79, 81

*§? tla (council-place), 58, 69, 72.
Khurutshe, 57, 58, 76.

Kings and Kingship, 11, 12, 13 21,

_ I2 5 *

Banyankole, 128-9, 131#.; 156#.
mother of, 1 56 jf., 1 60-1.
relations of, 138-9, 140, 160.
wives of, 141-2.

women relatives and, 138-9. 140,
141-2, 150, 160-1.

Zulu,/ 2 6, 29, 30 jf., 52.

See,also Chiefs; Mugabe.

Kinship and Kinship Systems, 67,
200, 206-7, 214, 234, 247.
among the Bemba, 83, 87-90
92#., no. ’

among the Kede, 165.
among the Zulu, 28, 29, 35, 36, 45.
groups, 28,36,39, 208.
terminology, 29, 88, 98.

See also Lineage.

kpeem (senior male of a lineage)
251-3. *

Kraals, royal Banyankole, 138 Jf.
Kuba, 57, 58. J

Kusaasi, 239.

Kwena, 57.

Labour, 53.
division of, 248.

service (tribute labour), 34, 49, 68,

73 ~ 4 > 9 L 104, 106, 118-10!
145-6.

See also Work.

Lala, 86.

Land, 227-8, 249, 250, 262.

Bemba and, 103.
political rule and, 29-30
See also Earth.

Law (or jurisdiction), 38.

Bantu Kavirondo and, 212 If.

Kede and, 181-2.

Nuer and, 293-4.

See also Courts; Crimes; Offences.
Leadership, 42, 44, 58-9,' 9S , 186,
232#.

Lineage and the Lineage System, 6 7
n. 243#-, 267#., 284#.
groups, 83,88,94,231.
maximal, 243#, 251#., 261,266.
See also Family; Kinship.

Logoli, s, 6, 7, 8, 9, I4 , i S , 16, 2I>
197 - 236 .

Lou, 276, 288.

Luba, 85.

Lund, 86.

Lungu, 86.

Luo, 198, 226, 229, 230.

Magic and Magicians, 30, 31, 70, 96,
98, 106, 145, 151, 158, 159,
160, 259.

Magistrates, Zululand,47-8,49,50,52.
Mamprusi, 239, 246, 249, 257.
Marama, 225.

Marriage, 95, 149, 153-4, 200 , 222j
242.

intermarriage, 130-1, 135, 188-9,

t r ,- 00 * ’

Kede, 188—9.

Muhima and Mwiru and, 130-1,

_ T

Ngwato, 76.

Zulu, 26, 28, 35, 44, 46.

Masai, 126, 127, 198, 225, 226, 230.
Matrilinearity, 87, 89, 96-7.
Matrilocality, 87, 80.

Mavi, 229.

Medicines, 43.

Messengers, royal, 38, 72, 73, 105,
107-8, hi.

Military bands and service, 129, 139,
. 1 44-5 • ^ e e also Regiments.
Missions and Missionaries, -70, 72,
114-16.

Mosuor, 245-6.

Mosuorbiis, 246, 250, 256.

Mpande, King of Zululand, 25, 26,
, . , 30, 33, 34, 35, 38, 42*

Mugabe, 129, 131-2, 133, 134, 136#.
duties of, 137.

See also Kings.

Muhima and Mwiru. See Bahima

and Bairu.

302

INDEX

Settlement, Kede, 170 ff., 188.
Tallensi, 242-3, 249-50.

See also Villages.

Sex intercourse, 98, 104, 109.

Shaka, 13, 26, 31, 34, 36, 37, 43.
Shona, 57.

Shrines, 99, 109, 112, 253, 260, 262.
Bagyendanwa , 150 ff.
village, 92, 97, 104.

See also Relics.

Slaves and Slavery, 04, 106, 137

c .145-6,152,184,249,259.

Sociology: social structure, 17, 161-2.

186-8.

of Bemba, 87^.
of Kede, i86jf.
of Tallensi, 245, 251-3, 271.
Sorcery and Sorcerers, 31, 39, 43, 45,
66, 145, 147, 204.

Sotho, 107, no. |

Northern, 57. j

Western, 56.

South Africa, 25-55.

Spirits, ancestral, 30, 75, 88, 92, 97-8,
104, 125, 141, 143-4, 149, 155,
160, 161, 202, 210, 214, 235,

. 242, 250, 253, 255, 259, 269.
Subia, 57, 58.

Succession, 95 ff., 101-3.
among Banyankole, 156, 157 ff.
Kede, 177-8.

Ngwato, 74-5.

Zulu, 35-6.

'See also Inheritance.

Sudan, Southern, 272-96.

Swazi, 26, 52, 84, 92, no, 120.
Swaziland, 65.

Symbols, political structure and,

i7-i 8.

Taboos, 105, 203, 215, 216, 242, 254,
256, 259, 261.

Sex, 98, 108, 261.

Tadjoni, 225.

Talaote, 57, 58, 76.

Taleland, 240 ff.

Tallensi, 5, 6, 7, 8, 9, 10, 14, 15, 16,
18, 22, 239-71.

‘real Tabs’, 240 rt. 1, 244 ff.

Taxation, 66, 78, 113, 150, 180-1.
Tembe (Thonga), 26.

Tembuli, 229.

tmdaanas , 246, 255, 260 ff.

Teso, 198, 226, 230.

Tiriki, 225, 229, 230.

Tongo, 245, 246, 255, 206, 267.

Trade, 224.

of the Kede, 169-70.

Tribes, 56, 83, 84, 86, 92, 94, 100,

197, 198 et passim.

Nuer, 27% ff.

Nupe, 165, 166 n. 1.
territorial segments, 281 ff.

Zulu, 25, 30, 31-2, 36-7, 40-1, 43,
47, 51-2.

See also tribes by name.

Tribute, 59, 61, 68, 76, 78, 86, 95,
104, 106, 127, 128-9, 131-2,
_ , . yS, 3<40, i44> 145#., 181, 211.
1 shekedi, Regent of the Ngwato. 66,
67-8, 71, 73, 79-80, 82.
Tsoede, 175, 176, 184, 189.

Chain of, 191, 192.

Tswana, 56, 57, 58, 84.

Tswapong, 57,^8.

Uganda, 121-62.
ulupioa, 88, 89, 90, 93.

Venda, 84.

Villages, Bemba, 89 ff., 99, 104, 107.
Kede, 171-2, 179-80.

Ngwato, 57-8, 60.

Nuer, 272-3, 275, 281.

Visiting, 206-7.

Vugusu, 197-236.

Wamara, 123-4.

Wanga, 225.

Wards, Ngwato, 58 ff.

Zulu, 30, 41.

Warfare, 226 ff., 233.

Tallensi and, 241 ff.

Warriors, 135, 139, 144-5, 233.

Zulu, 26.

See also Army; Military Bands;
Regiments.

Wealth, 8, 9, 44-5, 49, S4 , 76, 77, 9S)
231-2, 250.

Witchcraft, 66, 105, 202.

Wives, 74, 141-3, 270.

‘great wife’, the, 74, 99.

See also Concubines; Marriage.
Women, 153-4.

Ngwato age-regiments, 73-4.
of royal rank, 93, 97, 99-100, 138-9,
I 4°> 141-2, 150, 160-1.

Work of age-regiments, 26, 74.
of Kede, 168—70.
of slaves in Ankole, 133.
of Zulu, 28, 46, 47.

See also Agriculturalists; Labour;
Pastoralists.

Yonga, 229.

Zulu, 5, 7, 8, 9, n,.25-55, 84, 92, no.
Christian and pagan, 51.