2015.262278.Caste-In India J H Hutton - Text

Download as pdf or txt
Download as pdf or txt
You are on page 1of 333
At a glance
Powered by AI
The book discusses the nature, function and origins of caste in India. It provides historical and anthropological context around different castes and tribes found across different regions of India.

The book is about the social anthropology of caste in India. It discusses the different racial elements, tribes and castes found across India and analyzes the structure, strictures, sanctions and functions of the caste system.

Chapter I provides an introductory overview. Chapters II-IV discuss different regions of India. Chapters V-VIII analyze the structure, rules, control and role of caste. Chapters IX-XI discuss analogous systems elsewhere and various theories around the origins of caste.

CASTE IN INDIA

Its Nature, Function, and Origins

J. H. HJJ T T O N C.IE.. M.A.. D.Sc.


Professor Enseritus of SocM Anthropology
in the University of Cambridge

FOURTH EDITION

BOMBAY
OXFORD UNIVERSITY PRESS
DELHI CALCUTTA MADRAS
Oxford University Press
OXFORD LONDON GLASGOW
NEW YORK TORONTO MELBOURNE WELLINGTON
NAIROBI DAR ES SALAAM CAPE TOWN
KUALA LUMPUR SINGAPORE HONG KONG TOKYO
DELHI BOMBAY CALCUTTA MADRAS KARACHI

y
I"

First published (by the Cambridge University Press) 1946


Second edition 1951
Third edition 1961
Fourth edition 1963
Fifth impression 1980

*
^khiktkuS'^

Printed in India by N. K. Dhote at Uniserve, 11 D, Thacker Industrial


Estate,N. M. Joshi Marg, Bombay 400 011 and published by R. Dayal,
Oxford University Press, Oxford House* Apollo Bunder, Bombay 400 039
CONTENTS
Preface vii
Preface to the Third Edition ix
Preface to the Second Edition xiii
Foreword to the First Edition xv

Part L The Backgrount


Chapter L Introductory 1
Racial elements in India; the diversity of India's population.

Chapter IL Southern India 8


The forest tribes; Telugu and Tamil; Malabar; Coorg; the Nilgiris; the
Deccan.

Chapter III. Western, Central, and Eastern India 18


The west coast; Maharashtra; the tribes of central India and Chota Nagpur;
Orissa; Assam; Bengal.

Chapter IV. Northern India S8


Upper India; Rajputana; the Punjab? Sind; the mountaineers of the
north-west. The significance of diversity.

Part II. Caste


Chapter V. Its structure 46
Caste as a constructive system of endogamous units of society; subcaste,
hypergamy, exogamous division; mma; castes of the Right and Left Hand.

Chapter VI. Its strictures 71


The avoidance of pollution through water, food or contact; use of temples;
birth and death taboos; marriage rules; clothing and ornaments; language;
houses; inheritance; travel.

Chapter VII. Its sanctions 92



Caste a social unit ^respective provinces of social and religious authority;
control by secular rulers; by religious authorities; by the caste council.
The territorial limits, jurisdiction, procedure, and punishments of caste
councils.

Chapter VIII. Its functions 111


The functions of their social, economic, political, and religious
caste in
aspects from the point of view of the individual member; from that of the
caste as a communal unit; from that of the state and society as a whole,
Caste as a stabilizer; as an' organism; its religious sanction; its drawbacks,
political, economic, and social, The genetic function of caste.
vi CONTENTS
Part III, Origins
Chapter IX, Analogous institutions elsewhere 133
Plural societies; compulsory functions; social hypergamy; ritual occupations;
hereditary occupations; stratified society in ancient Egypt; African analogues;
analogy in Burma; in Japan.

Chapter X, The traditional origin and its implications 149


The^ lotir voTna; castes of mixed
origin, anuloma and pratiloma', matrilineal
survivals; Mediterranean cults; devadasi, busaxn, murli; castes
of the Riirht ^
and Left Hand.

Chapter XI, Other theories i factors in the emergence of


*
caste 1.70
Caste as an creation; as originating in occupation, in class dis-
artificial
tinctions, in colour; caste as originating in family worship,
in race and
hypergamy, in guild and in tribe, in ritual; caste as originating in primitive
beliefs in magic, mana, totemism, and in the fear of pollution.
Effect of
geographical and political divisions on caste, of the doctrine
of karma-
marginal survival of primitive conditions; life-matter; rmna; taboo

Chapter XII, Conclusions 183

Part IV, Appendices


A. The position of the exterior castes
192
B. Hinduism in its relation to primitive religions in India
223

Part V
Bibliography
263
Glossary
276
Index
299
Maps folded at end
A. Areas referred to in the text
B. Places and Physical Features referred to in the text
j

PREFACE
In this edition of Caste inIndia little new matter will be found.
One or two minor mistakes have been corrected, an omission or
two rectified. A few slight alterations have been made in order
to conform to some of the internal changes of administrative
boundaries in India, e.g. of the partition of Bombay into Gujarat
and Maharashtra; but N.E.F.A. and Nagaland will be found
alluded to as if they were still part of Assam. They never had
very much in common
with the Assam Valley proper, but the
same applies to other hill tribes like the Khasi and the
Lushei who are still included administratively in the state
of Assam. Otherwise there has been no change. That is not
intended to imply that the caste system itself has suffered no
change, but it is likely that such change as has taken place since
the third, or perhaps even since the first edition was published,
is hardly yet perceptible in day-to-day social life, at any rate in

the villages, in which the great majority of Indians live. It is


another matter when it comes to election time, whether for
general or local elections, for Professor Srinivas has pointed out
that some weakening of the ideas regarding pollution which have
permeated the caste system in the past has been accompanied
by a greater activity of caste in politics.^ This need not be
regarded as an unfavourable prognostic for India's nationhood;
patriotism, like charity, begins at home and the lesser loyalties
lead to the greater. Consciousness of their legal rights as citizens
has increased among members of low or exterior castes, though
this does not imply a consequent increase in status; such an
increase has however undoubtedly taken place in some, perhaps
inmany cases as a result of improvement in economic condition.
The present position of caste in India has been admirably
reviewed by Professor Srinivas in the volume of essays already
referred to, and there are other recent works on caste to which
attention ought to be drawn. In the last edition of this book
Dr W. H. Newell's article on The Brahman and Caste Isogamy
'

in North India in the Journal of the Royal Anthropological


Institute vol. Ixxxv, pp. 101-10, was overlooked, and there have

1 Caste in Modern India and Other Essays, pp 5 sqq. and Essay I passim.
viii PREFACE
appeared since Aspects of Caste in South India, Ceylon
and
North-West Pakistan, edited by Dr E. R. Leach, Dr McKim
Marriotts Caste Ranking and Community Structure
in Five
Regions of India and Pakistan, Professor Irawati Karv^s
Changing India: Aspects of Caste Society, her Hindu Society:
an
Interpretation (see Man, 1962, 254), and Mrs Taya Zinkins Caste
Today. The two last items taken together will give an admirable
idea of the practical working of caste, in the past
and in the
j^esent, to anyone unfamiliar with it at first hand;
they make
an excellent supplement to Professor Srinivas account of the
current political aspects of caste. All trf these are of
course more
up-to-date than I can be. and Professors Karve
and Srinivas at
any rate have the advantage of seeing thie caste system from
inside. ,

My acknowledgements are due again to a number of


kindly
critics, and to the indefatigable co-operation of the
Bombay
branch of the Oxford University Press.

MEW MmoR H. H.
J.
luly xs63
uv uT
published. The 7 of*ndia
aijiyal
since Caste in India was first
independence and a purely Indian
^eminent has i^ble legislation which wo^ild live been
imptt^per ewn a i e^
possible to an alien administratitm,
and the leaders of thou^t and policy in India seem
generally
to hare o the decuion that the caste system must
go. If
that be indeed the caM it 1$ perhaps
jMotly on the ground that
die caste syrtem is a hindrance to industrial
development and
eccHUwic advance, but more particularly under the urge
of an
altruistic and humanitarian desire for
an improvement in the
condition of the depressed classes, and a feeling that the
caste
system is bound up with the untouchability of the exterior
castes and is discreditable therefore to a
modern society. Legis-
lation then has been undertaken in an attempt to deal with this
situation.
In 1931 it^ was redtoned that the exterior castes numbered
over fifty million; in 1951 nearly fifty-two million 'untouch-
ables were counted; in i960 the scheduled castes are spoken
ofi as nuinbering fifty-five million. The Indian Constitution
makes special provision for the scheduled castes, and in 1955 the
Indian parliament enacted the Untouchability (Offences) Act,
which laid down severe penalties for anyone acting on ancient
custom to prohibit to excluded castes the use of temples, wells,
schools, shops, eating-houses, or theatres, or to treat them as in
any way separate or inferior. In the agencies of central and of
state governments one post in eight of those filled by compe-
titive examination is reserved for them, and standards and age
limits are adjust^ to their poorer education; government
schedarships are likewise reserved for students from their com-
munities. Caste^ Hindus are encouraged by the offer of free
board and lodgii^f to enter hostels intended to accommodate
exterior castes, ^d
aU hostels accepting government aid are
^

required to take in a utinimum complement of sudi castes, and


seats are reserved for them in the central and state legislatures.
There are at the time of writing 76 exterior caste members in
of Tims, a* May i^ea.
X PREFACE TO THE THIRD EDITION
the Lok Sabha; the Cabinet Minister for Railways is of their
community, and the Chief Minister of the state of Andhra
Pradesh also. It is not easy to see what more can be done
towards abolishing untouchability by legislative action, parti-
cularly as it is still practised by many of the exterior castes as
between themselves. The experience of Japan, where some
similar discriminations against certain castes were abolished in
law eighty years ago but still subsist in practice, suggests that
such an end can only be attained by very slow degrees and the
continued pressure of advancing education and rising standards
of living.
Most writers on
do seem to regard untouchability as
caste
inseparably bound up with the
caste system, and some as a
necessary condition of the survival of Hinduism, but I cannot
see it in either of these aspects myself. One writer says: It
seems that members of a multi-stratified society must feel supe-
rior to one group, and assert that superiority if their subser-
vience to another is to remain tolerable. And this apparently
human need finds full and inhuman expression in untouch-
ability.* ^ It is possible that this is so, and that the bottom
group of all has to take it out in superiority over dumb animals,
but the u ntouchability of certain groups as groups does not
apply to the majority of graded castes and cannot therefore be
essential to the system.It is a principle which can be observed
also in non-Hindu societies which are not graded to anything
like the same extent, while societies stratified in some degree
at any rate have existed throughout history without it.
However that may be the intelligentsia of India today seem
to have made up their minds that the caste system should be
got
rid of, though seems clear that in general they fail to compre-
it

hend the and perhaps the dangers, of such an under-


difficulties,

taking one indeed which would, if carried out at a stroke,


wreck the edifice of Hindu society and destroy a growth of some
^hree thousand years or more. In the autumn of 1955 an
Indian
Conference on Social Work organized a Seminar on Casteism
and Removal of Untouchability * in order to suggest concrete
measures for the removal of untouchability and to combat the
ubiquitous menace of casteism It met under the able chair-
manship of Professor M. N. Srinivas, and many and various were
2 The Delhi correspon^pit of The Times on 2 May 1955.
PREFACE TO THE THIRD EDITION XI

the views expressed, but it may be doubted whether any but


the trained anthropologists and a psychologist or two appre-
ciated the formidable nature of an attempt to jettison the whole
structure of their society. The anthropologists included, besides
the chairman of course. Dr Smt. Irawati "Karv^ and of the
psychologists it may be noted in passing that Dr Pandharinath
Prabhu agrees with me in regarding the problem of untouch-
ability as quite separable from that of caste. Several speakers
pointed out the dangers of too much protection for the un-
touchable classes, and indeed there had long before been
rumours of caste Hindus declaring themselves untouchable to
enjoy the untouchables special facilities. But it appeared that
on the whole there had been little efEective change of attitude,
and it was left to Professor Srinivas to point out, when summing
up, that although everyone says that he wishes caste to go
'
actions are more or less at complete variance with professed
beliefs.
There have been a number of individual contributions to the
this book was prepared,
study of caste since the second edition of
and some of them of the first importance. Bruce Ryans Caste
in Modern Ceylon (1953) is perhaps only relevant for com-
parative purposes, but no student of caste in India can afford
to neglect Dr Karv^s Kinship Organization in India (1953);
Kapadias Hindu Kinship was published in 1947, Stevensons
Status Evaluation in the Hindu Caste System in 1954, and
Marriotts Village India in 1955. Dr Ghaus Ansari has written
in Vienna of caste as it affects Muslims (1954) and Professor

Mario Cappieri in Italian of the untouchables (1947). Then


there have been valuable studies of specific areas ^Professor
Srinivass Religion and Society among the Coorgs of South
India, 3 model of its kind, appeared in 1952, Professor L.
,

Dumonts Hierarchy and Marriage Alliance in South Indian


Kinship in 1957, F. G. Baileys Caste and the Economic Frontier
in the same year, and Mayers admirable Ca^te and Kinship in
Central India in 1960. There have likewise been many mono-
graphs on individual castes such as Furer-Haimendorfs The Raj
Gonds of Adilabad (1948), Dubes T/ie Kamar (1951), and Fr
Fuchs The Gond and Bhumia of Eastern Mandla in 1960, and
that is to mention merely a small part of all that has been
xii PREFACE TO THE THIRD EDITION
published since I wrote, which is r^vant to the study of the
caste system.
If I were to take all these into account I should have to go
back to India and start again, but, as before, I have done my
best to rectify in this edition any mistakes of fact that have been
pointed out and come to my notice, and to remedy the fault of
one or two omissions. One of my critics chid me gently for
saying nothing of the future of caste, but I have no claim to be
a seer, and prophecy has no part in my purpose in writing. All
that I have attempted here is to achieve a general but factual
view of the caste system as 'a working whole and the principal
cohesive factor in the society of Hindustan as I knew it, and
to otamine briefly the various views put forward as to how it
came into being. I write as an ethnographer and perhaps, in
a very modest way, as an historian. Not for me the misty visions
of Durkheim nor even the more objecdve horizons of Radcliffe
Brown. I am neither a psychologist nor a sociologist, and
general thecffies of human society and religion are little relevant
to my In so far as such theories are valid their validity
theme.
miut be of an universd natwe; but the caste system as observed
in India is an unique pheamnentm, and can only have arisen
thereto: J as the result of local historical
and particular impacts.
Moreover I am far from feeling confident of the possibility of
mans ascertaining from inside human society any permanendy
valid rules of human behaviour tlierein. In moments of pes-
simism I see mankind devolving into monstrous, communes of
fixed pattern comparable to societies of ants, wasps, termites, or
dung-beedes. In such an event it would no doubt be possible
(for apes or angels) to codify its rules of
behaviour and predict
its reactions to stimtdi. And perhapsonly a Victorian and
it is
non-evidentiary belief that in the very ltmg run the better pre-
vails over the worse, which unseats the nightmare. Meanwhile
the chief characteristics of human society, as I see it,, are its
incredible adaptability to environment, its extreme unpredict-
ability in behaviour, its infinite variety, and its innate folly.
Instances a plenty will be found in the pa^ that follow.

Midsummer, 1^0
PREFACE TO THE SECOND EDITION
In this second edition of Caste in India the author has to
acknowledge with thanks his indebtedness to the many kind
reviewers of the first edition for improvements in this one. He
has done his best to take advantage of all criticisms except one
or two which he found it impossible to accept, and one which
required something beyond his powers: having no Sanskrit he
was compelled to continue to rely on translations of the vedas
and puranas instead of consulting the original texts. He has
moreover made little attempt to recast his material in the terms
of India and Pakistan as they appear after their achievement of
independent status, but has thought it wiser to retain the termi-
nology of the India with which he was personally familiar. In
the few passages cont2dning words, references, or quotations from
Greek literatmre Roman script has been substituted for the Greek
lettering used in the original edition. Otherwise he has added
a little, he has altered a litde, and he has to thank Mr Martin
Rowlands of the School of Geography in Cambridge for two
admirable maps, one showing areas, the other places and geo-
graphical features. The general aim in these maps has been to
insert all places mentioned in the text, and, with the object
of keeping the maps as dear as possible in a small space, such
places only.
It should perhaps be added that it was largely the appreciative
reception of the first edition by Indians that encouraged the
author to publish a second, and to get it done in India itself,
a country to which his lasting gratitude is due. And in this
connexion the author has to acknowledge n<^ only the generosity
of the Cambri<%e University Press in allowing him to transfer the
work to another press when the first edition was sold out, but
also the courtesy and the efficiency of the printers and puHisbers
in3ombay.
J. H. H.
NEW EADNOn
November
FOREWORD TO THE FIRST EDITION
It isprobable that some sort of an apology is needed from
anyone who is bold enough to add to the great mass of literature

that already exists upon the subject of Indian caste, A recent


Indologist in America claims to have compiled a list of over
five thousand published works on this subject; so obviously some
justification is needed for adding to their number, Nothing like '

that quantity, however, will be found in the list of works cited


in this volume, which does not claim in any sense to deal
exhaustively with the subject. Only an encyclopaedia could do
that. It does attempt, however, to oflEer a brief conspectus of
the various aspects of caste, since, when trying to give to my
classes a general idea of the nature of the problems involved,
I was unable to find any single book of moderate size to which I
could send students for what I regarded as a satisfactory outline
of the subject. Two compact books in English there certainly
are, both fairly recent and each admirable after its kind; they
have been freely cited in this volume, particularly in chapter vii.

But the one phenomena omitting,


deals merely with existing
generally speaking, quesdons of origin and of wider significance;
the other has its treatment of origin based on theories of race
which are no longer valid in the light of present knowledge,
while its treatment of existing facts is limited to northern India,

though the caste system is probably found at its strongest in the


south. In this outline some attempt has been made in brief not
only to consider origins, but also the place of caste in the social
and economic order of Hindustan.
Besides the two authors referred to above there are a number
of others whose works have been quoted freely. All will be
found in the bibliography at the end of the volume. Even where
facts referred to are within my own observation and experience,
I have, wherever I could, given a reference to some printed
authority in which the matter can be followed up.
An apology is also perhaps required for certain inconsistencies
in the pages that follow. . .In the text I have [not often used
.

diacritical marks] but in the glossary I have endeavoured to ijaark


f GREW ORB
all long where a knowledge of the quantity is needed
syllables

to afford some approach to a correct pronunciation. Again, the


termination -an in South Indian caste names is sometimes re^-

placed by the honorific -ar, or the final n may be omitted; and


there are almost inevitably inconsistencies in transliteration from
Indian alphabets into Roman. But the inconsistency which has
proved the most disconcerting to the proof-readers in the Univer*
sity Press has been a tendency to fluctuate between singular and
plural forms for the names of castes, and above all my unpr;r
dictable and unrepentant practice of using singular and plural
verbs indiscriminately where the subject is the name of a caste.

As however

caste *
is a collective noun I hold that I am
entitled to use either number in the verb that follows, and I have
deliberately declined to be unnecessarily (as it seems to me) con-
sistent in this matter. If thereader is offended thereby, I offer
him my apologies forma.
I have to acknowledge the kindness of the Government of
India in giving me leave to reproduce as appendices to this
volume an appendix (ii) from my report on
and a chapter (xi)

the Indian census of 1931; form part iv of


the two together
this volume. I have to thank Mr J. Brough, of St Johns

College, Cambridge, now of the British Museum,^ for his impor-

tant help in the matter of pr&uir&, and Rajkumar Sahib Prafulla


Chandra Bhanj Deo also for some assistance in the same field.
Dr Alma Witdin has helped me in the matter of references, but
I have to thank in particular Miss Maureen OReilly for reading
my proof sheets and stiU more for the laborious and exacting
work of verifying almost all the references and checking my
bibliography. must thank Professor Raymond Firth also for
I

very kindly reading throt^h my proofs, and not the least of my


acimowledgements are due to my friend Dr Mary Edwards, but
for whose kindly and persistent instigation I should probably
uever have set my hand to plough this furrow at all.

CAMBRIDGE
p October

1 Since 1948 Professor of Sanskrit in -the University of London.


PART ONE
THE BACKGROUND

CHAPTER I

INTRODUCTORY
The subcontinent of India has been likened to a deep net into
which various races and peoples of Asia have drifted and been
caught. It is of course divided from the rest of Asia by the high
barrier of the Himalayas and from the other parts of the outside
world by sea. This isolation has not prevented numbers of races
from migrating into India, but it has led to the development of
society on peculiar lines of its Together with Pakistan
India ^ contains nearly a fifth of the world's population, some 530
millions of people, and being a natural geographical unit there
is inevitably much
that this great population has in common.
At the same time it is composed of all sorts of different elements
of great diversity, of different creeds, different customs, and even
different colours. All these varied peoples have been enabled
to live together, in conditions of comparative stability and
forming what may be described as a. multiple society, by the
caste system, which must probably be regarded as having deve-
loped as a sort of organic response to the requirements of the
particular case. Geographical circumstances have imposed a
certain unity on the inhabitants of the peninsula, whereas the
diverse origins of the people have dictated variety. The view
put forward in this volume is that it is caste which has made
it possible %r both requirements to be satisfied within a single

social system, a system, moreover, which has proved historically


to be very stable. It has proved capable of absorbing any intru-
sive society, and no intruders have yet succeeded in revolution-
izing it, though it is not so rigid that a caste cannot rise in the
social scale. Some castes, it is true, are so low in the scale that
1 This work having been originally written before their separation, the
term India is hereafter used generally in its geographical and historical
*

values to refer to the whole peninsula in which India in its more recent
sense and Pakistan are included.
1
3 THE BACKGROUND
they are regarded as outside the ordinary pale of society and
suffer disabilities accordingly,
but that is perhaps an incidental
rather than an essential feature of the caste system.
Each The customs by which it
caste is a social unit in itself.
lives are generally different in
some respects from those of other
castes, and are sometimes in marked contrast to those of
any
other caste at all. Persons of one caste do not marry those of
another. The extent to which persons of one caste will eat or
drink with those of another is strictly limited by unwritten laws
and everybody knows who is affected by them. Even a change
of religion does not destroy the caste system, for Muslims,
who
do not recognize it as valid, are often found to observe it in
practice,and there are many Muslim castes as well as Hindu;
and when some reforming body breaks away from Hinduism
and repudiates caste, it becomes something very like a
new caste
of its own.
Jews and Christians also in India often form castes
or bodies analogous to castes.
There are thought to be some 3,000 castes in India,
and it
would need an encyclopaedia to deal with them all.
Some are
derived from tribal or racial elements, some are
occupational,
being of the nature originally perhaps of guilds
of artificers or
craftsmen, some are territorial, some religious, and so fortl^.
The
original bond which united the members of a caste ^as
often
been different in different cases, but it is
suggested that the
society of the country as a whole has
been buiU up by the inte-
grauon of these different units into an orgafiic
community, and
one into which a new unit is fitted without
difficulty, while any
^oup that breaks ^way from other upTts can as easily
form itself
into a new cell within the structure,
and will indeed find it
almost impossible to do anything else.
To get some idea of the vaii-ious peoples which
have contri-
buted to this society one must
consider what is known of
^pauons into India. To judge by other areas of
south-east
Asia It seems likely that the
earliest human inhabitants of the
Indian peninsula of which any trace
survives were Negritos--
vtxy dark pygmies with peppercorn
hair, marked steatopygia,
and generally between 4 and 5 feet high.
There are several sLh
tribes still surviving in Malaya
and the Philippine Islands, and
there are somewhat kindred tribes in
New Guinea. They have
survived in an almost unmixed form in
the Andaman Islands in
INTRODUCTORT 3
the Bay of Bengal, though in rapidly dwindling numbers, and
traces of their blood are possibly to be seen in some of the
forest tribes of southern India, like the Kadars of the Cochin
forests, in occasional individuals in remote and hiUy areas such

as the Rajmahal Hills just south of the Ganges river in Bihar,


where such a case has been reported,^ and in occasional remote
Naga villages on the north-east frontier between Assam and
Burma. But of these Negritos nothing more than the merest
trace remains on the mainland, and one can only conjecture that
at some remote date they may have inhabited southern and
eastern India. In the extreme north-east of India, in the Naga
Hills of Assam again, there are traces of other negroid tribes
akin to Papuans and Melanesians, traces which can be seen
readily in culture and in art, and which occasionally appear in
physical traits.

Whether or not we are to count the Negritos among the races


which have contributed to India's population, there is no doubt
about the Australoid race which is often alluded to as pre-
Dravidian or proto-Australoid. This is a racial element which
may be ultimately allied to the so-called Brown Race of south-
'
'

east Europe, or which may have developed from a subhuman


ancestry in south-east Asia itself. The type is allied to the
aborigines of Australia and to primitive ^races spread over the
Indian archipelago and is very widely spread all over India, par-
ticularly among the lower castes and humbler classes of society.
It is found in the least adulterated form in some of the more
primitive of the southern Indian hill and forest tribes, but
generally appears mixed with other elements in all grades of
society, though it is much less evident among the higher classes
than the lower. It is characterized by very wavy, even curly,
hair,a broad nose and rather coarse features generally, and by a
darkish skin. There can be little doubt but that the ancestors
of people of this type formed the bulk of the population of the
Indian peninsula at some very early date, and probably before
anyone else, except perhaps Negritos, had entered it How and
2 Guha, Racial Ethnology of India,
p. 130 and PI. VIII, fig. 4. It may be
pointed out here that on the previous page Dr Guha mentions frizzly liair
as occurring not infrequently among the Angami Nagas. This statement
needs some correction. Such hair is very rare among Angamis, less rare
among Kachha Nagas, and not infrequent among Konyak Nagas, where
occasionally it seems to be characteristic of certain dans or villages.
4 THE BACKGROUND
when the subsequent types came to arrive in India is a matter

o conjecture almost until we reach historic times. What light

we have on the matter is supplied by inferences drawn from


languages, culture and physique.
There are some thirty different groups of languages spoken in
I ndia each of
,
which consists of from one to five or more different

vernaculars. Half a dozen of these belong to the family known


which since it the most widely dispersed is probably
as Austric, is

one of the oldest families of human speech. The peoples who


stilluse branches of this family of speech in India, or branches
of other families which show traces of afi&nity with Austric
tongues, are to be found in hilly and afforested areas where they
have been comparativdy free from external contacts. Several
tribes whose language is now classed with other families seem
to have formerly used Austric languages. In physique such
tribes have a good deal that suggests that they contain strong
pre'-Dravidian or proto-Australoid elements, but with consider-
able admixture of later comers of a more refined type, and an
appreciable dash of something that is common to Mongoloid
types, for without showing pronounced Mongolian features they

suggest that sub-Mongoloid type which is sometimes met with


on the fringes of purely Mongolian areas and may represent
an ancient marginal and less highly specialized Mongoloid stock.
As there are two distinct families of Austric languages in use in
India it is possible that there have been two routes of migration
from centrd Asia, one round the west end of the Himalayas, the
other somewhere to the eastward of them. The languages used
may be spoken of as Kolarian in the former case, Mon-Khmer in
the latter, which is confined in India to one area in Assam, while
the Nicobar Islanders in the Bay of Bengal also speak a language
intermediate between the two branches, but probably derived
from the east.
The speakers of Kolarian languages in India are few in
number and scattered in distribution, but languages of the
Dravidian family are widely spoken over large and contiguous
areas. Not only does the whole of India south of the Godavari
river ure these languages, but there are more or less isolated
survivals like Brahui in the north-west and Malto in the
Rajmahal Hills, as well as certain words and sounds used gene-
rally in northern India which indicate that Dravidian languages
INTRODUCTORY
5
were once spoken over the whole peninsula. No certain affinities
have been established outside India, but it seems likely that
these languages entered India &om the north-west coming
perhaps from Asia Minor, or at any rate from that direction,
via Mesopotamia. The physical type particularly associated with
these languages is probably closely related to that which is pre-
dominant in the Mediterranean, though it is generally found
now in a state of mixture with other types, particularly with a
round-headed type suggestive of Syria, Armenia and the Iranian
plateau. There are many cultural parallels between south
India in particular and the older civilizations of Mesopotamia,
Syria, and Crete, and it is probable that many waves of immigra-
tion have taken place, all more or less of the same type of
people. Probably the earliest of these were of a more or less
unm^ed long-headed type, the later comers being increasingly
mixed with rounder-headed peoples from the higUands of Asia
Minor and perhaps Iran. .

At some later date, probably in the third millennium b.c.,


when there seems to have been much disturbance of races and
peoples in the southern Russian steppe and in Iran, a wave of
round-headed peoples seems to have come in from the north-
west and passed down the west coast as far as Coorg. From
there some no doubt trickled through what is now the Mysore
state into the Tamil-speaking area of southern India, missing
the Malabar coast, which is cut off by the Western Ghats from
the rest south India, and reaching the Telugu-speaking
of
Andhra country just south of the Godavari river in even less force
than the Tamilnad. A section of these round-headed immigrants
seems to have crossed India to what is now Bengal, probably
moving south of the Jumna and Ganges rivers for most of the
way. It is possible that these immigrants introduced what is now
known as the outer band of Indo-Aryan languages, and it is

possible that they were also the first to introduce a patrilineal


form of succession; but these are very much matters of conjecture.

The first invasion of India of which we have any sort of record

was that of the Aryans' of the Rigveda somewhere about


*

,1500 B.c. These people definitely used an Indo-European


3*
Aryan is really a linguistic term and ought not to be used as if it
*

were a racial one; it is only used here for want of a better term, and
because it is familiar.
6 THE BACKGROUND
language, and it is not unlikely that they had, at any rate to
some extent, fair hair and blue eyes. Their religion was a form
of nature worship, expressed in the hymns to their nature deities
which we know as the Rigveda. These people occupied and
settled in the Punjab in the north-west of India, later sending
colonies farther east and filtering through central India into the
south. Over the whole of northern India Aryan languages have
generally superseded all others, and though they have never
succeeded in doing this in southern India, they have naturally
succeeded in modifying some of the languages of that part of the
peninsula.
These Indo-Europeans were followed in subsequent centuries
by many invasions on a smaller scale. Between 600 b.c. and the
beginning of the Christian era northern India was invaded by
Persians and Greeks, and after them by the Sakas or Scythians
and by the Kushans, who were a nomad tribe from Central Asia.
Again in the fifth century a.d. another horde of nomads from
central Asia, the Huns, invaded northern India. In the eighth
century the Muslim invasions began with an Arab invasion of
Sind, and Muslim invasions of one sort or another continued'
until the establishment of the Mughal Empire in the sixteenth
century, by which time the Portuguese had already founded
their settlement at Goa. From the other direction the
Mongoloid Shans had invaded Assam, and Mongolians no
doubt already occupied the higher mountains all along the
northern fringe of India.
It is from such diverse elements that the peoples of India are
derived, ahd although the development of the caste system and
its stabilization have crystallized the peoples of India into a
large number of fairly 'watertight* communities, the barriers
between them were less rigid while the system was in its early
development than they are now, and even now they are not
all immutable. Consequently the differences between caste and
caste, in so far as they are racial, are rather differences of degree
of mixture than absolute differences of race and type. In some
cases it is easier to differentiate between Indian racial types
hy
area than by caste,^ and of course it is possible that in the course
of time different geographical environments may have played some
part in establishing physical differences between the inhabitants

^ Vide Guba, op. cit., particularly with reference to his map on p. lag.
INTRODUCTORY 7

of different areas, since environmental differences, e.g. between the


Thar Desert and the humid coasts of southern India, are extreme.
It in the south probably that the most ancient types survive,
is

for, while the climate has favoured the growth of heavy forest
most in the south, the shape of India has tended to drive early
inhabitants southwards under pressure, no doubt, of migrations
and invasions from the north. In this introductory glance at
the background of caste, then, it will be convenient to start with
the south of India, and the most primitive of the people to be
found there, working northwards and then across India from
west to east, and thence north-westwards again into the Hima-
layas which form the northern boundary of the Indian penin-
sula, It will be understood that in a survey so restricted of an
area so vast and a population so varied it is only possible to
generalize on broad lines.

Note, In line 2 of page 3 above I have substituted possibly*


for the ' probably of the first two editions in deference to the
*

criticisms of several Indian reviewers whose knowledge of


physical anthropology is likely to be less superficial than mine.
At the same time it seems to me on purely general grounds that
the Negrito is unlikely to have been absent from the mainland
of the Indian peninsula; traces of him have been reported from
the Persian Gulf (Giuffrida-Ruggeri, First Outlines of a Systematic
Anthropology of Asia) and he is visibly present in the Bay of
Bengal, the Golden Chersonese, and Indonesia. If he was pre-
sent in India he is likely to have left some traces in the existing
population, for man is an incorrigible miscegenator. Of some
such physical strain in the Naga Hills, with its many cultural
links with Melanesia, I feel confident from personal observations.
Since writing the above a team of five physical anthropologists
of the University of Calcuttas Anthropological Department have
made a detailed physical survey of the Kadar, and conclude that
the features which some anthropologists (including myself) had
taken to indicate a Negrito strain are really due to a negro
admixture. See A Physical Survey of the Kadar of Kerala, by
S. S. Gautamsankar Ray, M. R. Chakravartti, A. R.
Sarkar,
Banerjee, and Papia Bhattacharjee, Memoir No. 6 (1959) of the
Anthropological Survey of the Government of India.
CHAPTER II

SOUTHERN INDIA
Perhaps the most primitive of the south Indian forest tribes is
that of the Kadar of Cochin State, a tribe in which the proto-
Australoid element is definitely predominant. For an un-
questionable Negrito we have to go outside the mainland of
India to the Andaman Islands. The Eadar have only
got some suggestions of similar physical characteristics, but,
like the Andamanese, they live by hunting and the collection
of forest produce. They bury their dead, observe no period
of taboo after a death has occurred in the family, and erect no
memorial. They file or chip the incisor teeth to a point. Like
the Andamanese the Eadar are very rapidly dying out, largely,
apparently, as the result of their having taken to opium and
distilled liquor, which has been given them by alien contractors
employing their services.^
Another tribe of southern India living in an extremely primi-
tive way by hunting and digging for roots is the Mala-Pantaram
tribe of Travancore. Their principal or only weapon and
implement is with the point hardened in the fire, and
a stick
they have no domestic animal except the dog. Being entirely
dependent on hunting and food collecting they need about two
square miles of forest per head to support themselves. Their
form of burial is very simple; the dead are just buried where
they die and the group leaves the place and moves elsewhere.
They arq nomads within a limited area, and their dwellings
either rock shelters or flimsy huts of leaves. Unlike most of
their neighbours they are not matrilineal, but reckon descent
through the father. A very similar tribe is that of the Paliyan
of Madura district. They, too, do not cultivate, and live
entirely by collecting forest produce and snaring and trapping
wild animals and birds; they own no personal property except
the clothes they wear, a few of the simplest utensils, a digging
stick like that of the Pantaram, and a bill-hook. They depend
largely on wild honey and on the sale of the wax from its
combs. To get the honey a man is commonly let down over
^ Report on the Census of Indta^ 1951, vol. 1 (India), pt. iii-B, pp. aij sqq.
SOUTHERN INDIA %

the cliflE by a rape and dislodges the bees*


of twisted creepers
nest with his digging stick. Most Paliyar cannot count beyond
the numbers of their fingers and toes.^
A little less primitive in their mod^ of life than the Pantaram
and Paliyan are Malavetan, Thantapulayan, Urali and other
similar tribes, mostly of hybrid origin though predominantly
proto-Australoid, who practise a shifting or migratory cultivation
in addition to hunting and food collecting. Many retain indi-
vidual primitive traits of social behaviour of one sort or another.
Some of the Malavetan, for. instance, chip their incisor teeth
to a sharp point like the Kadar;^ the Thantapulayan women
wear (or did till recently) a skirt of leaves or grass; the Urali
are particularly given to making tree houses 60 feet or so above
the ground, and it is in such houses that their women seclude
themselves at adolescence, menstruation and even at childbirth,
which takes place in isolation, advice and instructions being
shouted from a distance to the woman in travail. The hills
and forests of Travancore, Cochin and the other parts of
southern India contain a number of tribes of this kind, many
of them the Mala-Arayan for instance ^ma trilineal, others
patrilineal, and some, like the Ulladan, in an intermediate stage
in which the property of a man passes half to his sons and half
to his sisters sons. All of them, probably, practise cross-cousin
*
marriage, in the form of a union between a man and his mothers
brother's daughter. The weapons and implements of the most
primitive ones consist of a digging stick and a wooden spear,
and nowadays often a mild steel bill-hook purchased in the
towns. The less primitive use also the bow and the pellet-bow,
and in a few cases, that of the Muthuvan of the Palni Hills
for instance, the blow-gun. Some, like the Kannikan, have the
institution of the bachelors' hall, a hut where all the unmarried
men live, or at any rate sleep, in common, and most of them
have a headman of some sort who settles disputes, and performs
minor administrative functions, with the help of a council of
elders. All are regarded as more or less outside the pale of
respectable Hinduism,

2 Ibid.,
pp. 195 sq.
3 L. A. K. Iyer, Travancore Tribes, i, p. 159, and illustrations at pp. 145,
149 -
10 THE BACKGROUND
The peoples of southern India, other than the primitive jungle
tribes of pre-Dravidian race, are roughly divisible into four
groupsthose of Telingana or the Andhra country, of the
Tamilnad, of the Carnatia and of Kerala. The first of these
corresponds roughly with the distribution of the Telugu lan-
guage, which is spoken along the east coast of India from the
Pulicat lake, a little north of Madras itself, in the south to
Kalingapatam at the mouth of the Vamsadhara and the Orissa
boundary in the north. Inland it stretches back into
Hyderabad (formerly the Nizam's Dominions), to Sandur
and to the eastern districts of Mysore south of that. The
second occupies the whole of the eastern side of India
south of Pulicat lake, from the Coromandel coast, that is, to
the Western Ghats from Cape Comorin to the Nilgiri Hills.
This is the home of the Tamil language. The term Carnatic *

is liable to some misinterpretation. It has become obsolete


possibly on account in part of the confusion to which its use
may give rise. Historicdly it has been used to signify the low
country between the Eastern Ghats and the Coromandel coast,
while its one-time synonym * Canara probably of the same
derivation etymologically, has been similarly limited to the low-
lands betw'^en the Western Ghats and the sea.^ Here the term
Carnatic is taken to cover the Tulu- and Kanarese-speaking
countries of Kanara, Mysore, and the adjacent parts of Bellary
and Dharwar, together with the little plateau of Coorg where

Kodagu is the vernacular that tract of Dravidian-speaking
country, in fact, which lies south of Maharashtra and north of
Kerala, and extends eastwards towards the meeting-place of
Telingana, or the Andhra country, and the Tamilnad on the
east coast. Kerala may be said to extend from Kanara down
the Western Ghats and the Malabar coast to Cape Comorin.
On this coast, including the states of. Cochin and Travancore,
the language is Malayalam. To the north of Kerala is the real
Karnataka, where the language is Kanarese, and the great culti
vating castes of OkkalTga and Lingayat, a caste of sectarian origin,
co-exist in the same sort of rivalry as the Kapu and Kamma
castes of Telingana.
All languages mentioned here belong to the Dravidian family.
Each linguistic area contains people and castes of all grades of

^ Vide Yule and Bumell, Hobson-Jobson, s.w. '


Canara *
and
Camadc
SOUTHERN INDIA 11

society, but while very many of them correspond to some more


or less equivalent body in another area there are differences of
character and custom as well as of language which distinguish
them. The Telugu language has often been compared in
mellifluence to Italian, and the upper-class Telugu approxi-
mates to the south-eastEuropean in physical type. This is parti-
cularly marked in the case of the Telugu Brahman. The Telugu
caste of Kapu, which uses the familiar title Reddi ", is probably
*

the most numerous single caste in the south of India. It num-


bers two or three millions, mosdy cultivators, farmers and
landowners. Socially they take a place next to Brahmans in
many districts, and claim an origin from northern India and
kinship with the Rathor dan of Rajputs. The caste is split
up into a large number of sections, some of which show traces
in their customs of the former existence of a matrilineal system,
while one section, the Morasu, is noticeable for a custom by
which at the marriage of the first daughter, or alternatively at
the birth of the first grandchild into a family, the bride's mother
or the wife of the grandfather's eldest son must amputate at
least one joint of the third and fourth fingers of her right hand
at a temple of Bhairava. The custom was practised at any
rate up to the end of the last century. Some of the Panta
Rcddis of Conjeeveram have totemistic septs which taboo the
totems with which they are associated. Generally speaking; the
inhabitants of the Andhra country appear to be more volatile
and * temperamental than their neighbours, and in particular
'

are noted for the outbreaks of rioting between the castes of the
Right Hand and of the Left, two ancient factions in which the
low castes of Mala and Madiga are respectively prominent as
protagonists. The Chindu dance performed by Madigas with
bells on their legs at marriages, and festivals generally, has had
to be prohibited in several districts on account of its infuriating
effect on the Malas, and of the resulting riots. The Tamil
seems generally to be of a harder-headed, more practical turn
of mind than the Telugu, similar musical gifts being combined
with a peculiar aptitude for mathematics and physics. The

Tamil community Brahmans in particular have produced
^ Vide Thurston, Castes and Tribes^ v, pp.
75 sqq. The same practice
used to be followed by the Okkaliga of Mysore and Kanara. Vide Iyer and
Nanjundayya, The Mysore Tribes and Castes, rv, pp, sts^ sqq.
Vide infra, p, 70.
U THE BACKGROUND
whose names are well known in Europe. Among the
physicists
lower Tamilnad are the criminal tribes known as
classes of the
Kallar and Maravar, more or less professional cattle thieves
who have a notable bull cult very suggestive of the Mediterra-
nean, whether of ancient Mycenae or of modern Portugal.
Bulls with sharpened horns are driven to fury, and the young
man who wishes to distinguish himself and prove himself a
desirable match for the lady of his choice must jump on the
bulFs back and recover a doth from its horns.^ Another caste
of the Tamil country with rather unusual customs is that of the
who live in a walled fort of some 20 acres extent,
Kottai Vellala
insidewhich no male of any other caste is allowed to enter
and outside which no woman of the Kottai Vellala caste is
allowed to go. The family dwelling-house passes to the
youngest daughter, and u daughter who marries continues to
live in it until her younger sister marries. The caste is a very
small one, a contrast in this respect to the Kallar. Tamils of
the poorer classes have emigrated in large numbers as labourers
to Ceylon and also to Malaya and elsewhere.
The caste known as Kammalan or Panchala should be men-
tioned, as it consists of the artisans of five occupational sections--
goldsmiths, braziers, carpenters, stonemasons, and blacksmiths.
They daim descent from Viswakarma, the architect of the
gods, and equality with Brahmans. They intermarry (though
the goldsmiths will not intermarry with the blacksmiths) and
have most elaborate guild organizations. When, as they some-
times do, they bury their dead instead of cremating them, they
use a stone-lined grave, in which apparently they place the
deceased in a sitting position, in contradistinction to the culti-
who do not line the grave with stone and generally
vating castes
bury the body extended. The Chetti, a caste of bankers,
brokers, shopkeepers, moneylenders and traders, have a large
number and of much
of subdivisions; being in general wealthy
businessacumen their social importance is greater than mere
numbers would warrant. One other Tamil caste must be
mentioned here and that is the more humble one of the Kuruba,
whose traditional occupation is the keeping of sheep and the
weaving of woollen doth and blankets. They seem strictly to

^ Thurston, Cast^ and Tribes, ni, p. 75.


Thurston, CasUs and Tribes, iv, pp. 33 sqq.
SOUTHERN INDIA i3

belong to the Kanarese country, but are found in both Andhra


and the TanGiilnad and are remarkable for the fact that they
still erect cromlech-like rough stone dolmens over the graves

of their dead. In choosing a bride they are very particular that


the whorls or natural partings in her hair shall be such as are
accounted fortunate, a prejudice common enough in India and
Arabia in estimating the value of catde or horses, but perhaps
unusual outside south India as applied to human beings,
While the society of Kanara is generally speaking not dis-
similar to that of the Andhra and Tamil countries, the physical
type of at any rate the higher classes varies from it and shows
some affinity to that of Bengal, possibly because both areas
constitute the limitsmore or less of the main infusion of round-
headed stock before the invasion of the Rigvedic 'Aryans'.
Kerala, on the other hand, has been the stronghold of an ancient
type of society which predates any 'Aryan' invasion and was
probably at one time virtually universal in India. The Malabar
coast is a long strip of very fertile tropical country cut off on
the east by the high mountains and thick forests of the Western
Ghats from easy communication with the Telugu- and Tamil-
speaking areas, a strip narrowing in the north in the neighbour-
hood of Goa and in the south at Cape Comorin, so that it is
virtually shut off by mountains from land communications.
Such external contacts as it has experienced until comparatively
recent times have been mostly by sea, and the influjcnces of
Arab traders and of immigrants from Malaysia via Ceylonthe
route by which came also the coconut, a staple producf of

Malabar are readily traceable in certain sections of the popu-
lation. It is no doubt the landward isolation of the Malabar
coast which has been responsible for the long survival there of
the matrilineal system. Until the passage of special legislation
within the last fifty years, the great majority of the castes of
Malabar lived under a system of society in which all real pro-
perty and the family name were inherited through the women
of the family instead of through the men. Any property acquir-
ed by a man went not to his own children but to his sister's
children, for his own children inherited the family name and
property of their mother and hmr brothers, not of their father.

Thurston, Castes and Tribes, m, pp. 133 sqq.; Ethnographic Notes,


pp. 146 sq.
14 TUB BACKGROUND
This system is not without advantages of its own, but succumbed
in other parts of India to a patrilineal one which probably
came in with Indo-European tongues or earlier. In Malabar
itwas retained by nearly all castes, the most important being
that of the Nayars, a
more or less military society in which the
men tended to go off on fighting or marauding expeditions
leaving their wives to live in their own maternal homes, a state
of society familiar also in Malaysia. The Nayars formed
the
and the lower castes were and are mostly
aristocracy of Malabar,
organized on a similar social scheme, but the priestly caste
of
Brahmans, called Nambudri, hold a peculiar position in the
country. They seem to have come originally from northern
India, and to have entered Malabar at a much later date
than
the Nayars, and they observe the rule of primogeniture
and a
patrilineal system of inheritance. The eldest son marries a
Nambudri Brahman wife and their children inherit the paternal
estate. The younger sons form matrilineal alliances with Nayar
women, and their
children inherit from their mothers and
mothers brothers on the Nayar pattern. At least that
was the
rule until recent legislation made it possible for the
paternal
property to be divided and inherited by younger sons
who can
now hand on property to their own children. The Nambudri
Brahman women marry Nambudri Brahman men only, so that,
as all the younger sons of
Nambudri families married Nayars,
they used to find it difficult to get husbands
at all. The family
system of the Nayars naturally tended to
preserve a much
greater social freedom for women than is
common elsewhere in
India, and female education has gone
much further in Malabar
than in most other parts. This is not only the
case in Nayar
society but in many other Malabar
Christians, for instance, for Malabar
communities among Syrian
contains a large number of
Christians, the earliest converts having
been converted by
Nestorian missionaries of the Syrian church,
possibly in the first
century a.d. and traditionally by St Thomas
the Aposde, but
probably not later than the fourth in any
The Cochin case.
State on the Malabar coast also contains some
very ancient
Jewish communities which daim to have been
in south India
at least since the Roman destruction of Jerusalem.^ They are
^ Hebrew Insciijption from Chennamangalam , in I.A., July
SOUTHERN INDIA 15

divided into the White Jews of pure Hebraic descent and the
Black Jews who .nclude local converts and persons of mixed
descent. Linschoten remarks that the White Jews speak good
Spanish and says that they came direct from Palestine, and
Crooke classifies them with Spanish Jews as Sephardim.^ The
Moplahs (Mappila) of Malabar are generally regarded as partly
of Arab extraction, much reinforced by local converts. The
more southerly have the reputation of being exceptionally
fanatical Muslims, and in frequent rebellions have earned a
reputation for complete disregard of personal safety in their
which has also shown itself in attempts at whole-
religious zeal,
sale conversion ofHindus to Islam by force.
Among the non-Nayar Hindu castes of Malabar mention must
be made of the Tiyar, a caste corresponding to the Shanan of
the Tamil country, whose traditional occupation is the extrac-
tion of toddy from palm trees, and the Iruvan, who seem to
have reached Malabar by sea from the Indian archipelago by
way of Ceylon; the Panan caste also is of interest for its
picturesque practice of the curing of sickness by dancing in the
guise of demons about the body of the sick, a practice used also
by the Malayar, as well as by the Parava caste.
The landward isolation of the Malabar coast from the rest
of the Indian peninsula has led to the survival there of ancient
forms of society. It has also led to the survival of ancient
physical types, for the immigration of broad-headed elements,
probably coming down the west coast of India from the Iranian
plateau somewhere about the third millennium b.c., has hardly
penetrated this area at and the Nambudri Brahmans from
all,

north India were also a long-headed breed, so that the Malabar


head is almost universally narrow. The litde mountain plateau
of Coorg, however, presents a great contrast in this respect. The
Kodaga (orKoraga ) population here, from which the plateau
'

takes its name, is tall, stalwart, broad-headed and comparatively


fair-complexioned, and though Hindu, like that of Malabar,
has preserved and developed a distinct culture of its own with
greater freedom from the restrictions of caste and a long tradi-
tion of martial independence and self-reliance which may be
compared in some respects with that of Switzerland among the
nations of Europe. On the birth of a Kodaga boy a miniature

Things Indian,
i6 THE BACKGROUND
bow and arrow put into his hands as a symbol, and athletic
arc
contests of one kind or another form a
common feature of
ceremonial or festive gatherings/^ The uniform brachycephaly
of these mountaineers, in marked contrast to the dolichocephdy
of most of Madras, indicates Coorg as the southern extremity

of a prehistoric penetration down the west coast of India of


an intrusive broad-headed element which has retained on this

plateau its original character with comparatively little admixture


of other racial stocks. It is in marked contrast to the next
most numerous people now living in Coorg, the Yeruvas, who
are dolichocephalic and of markedly Australoid affinities. The
Kodaga population is marked by a straight and narrow nose,
whereas the Yeruva, who is incidentally of much lower stature,
has a broad and flat one.^^
Associated with no other community in southern India in their
physique or in their society is the rapidly vanishing Toda tribe
of the Nilgiri Hills. This remarkable people subsists by the
keeping of buffalo herds and dairies in extremely primitive
conditions. Though inheritance is through the male, the social

system involves joint families in which one woman has several


husbands. The cow buffalo is never eaten, but a bull buffalo
calf is killed and .eaten sacramentally annually and the office of
priest is indistinguishable from that of dairyman. In fact, the
dairy practice involves a ritual which makes it amount to a sort
of religion. In physical type the Toda is a contrast to the
neighbouring tribes of the Nilgiris by his hairiness, and is other-
wise more suggestive physically of tlie Nambudri Brahman
perhaps than of any other type of southern Indian, though poles
apart from him culturally and socially. In general, of course,
the Toda seems to approximate more to northern than to
southern Indian physical type, and it is conceivable that he
represents a very ancient migration of cattle-keepers from Iran
hybridized with some local type, and that his culture embodies
traditions going back to the prehistoric time when dairying was
a new invention. In contrast to the pastoral economy of the
Todas are the intensive agrictflture* and gardening of the Badaga
tribe, a very much more numerous people, some 50,000 in
12 Richter, Manual of Coorg, p. 141.
18 Holland, Coorgs and Yeruvas, p. 59.
i^Ibid., p. 81.
SOUTHERN INDIA 17

number. Though primarily agricultural, they follow all kinds


of pursuits JErom that of lawyer to that of scavenger. By origin
they seem to be immigrants from Kanara who have migrated
south-eastwards, but they show litde or no trace of the broad-
headedness characteristic of the Konkan, being, on the contrary,
extremely long-headed. Among their ceremonials is that of
fire-walking, in which numbers of persons walk across a pit
some 15 feet long filled with glowing embers to the
sound of
music played by Kotas, a caste who supply the musical as also
the artisan element in the Nilgiri Hills. The fire-walk is under-
taken as an act of worship of the god Jadeswami and with some
reference apparently to the fertility of nature.
In the centre of the northern half of southern India is the
plateau known as the Deccan (literally the South'), most of
which, with Berar to the north of it (Berar belongs rather to
central or to western than to southern India), constituted the
dominions of the Nizam of Hyderabad. The population is
mainly Hindu and differs little from that in British India on
its borders, the cultivators in the east, for example, being
indistinguishable from those of the Andhra country, and those
in north being largely Gonds as in the adjoining districts of
t|ie

the Gentral Provinces. In the forests of the south we find the


primitive tribe of the Chenchus, living by hunting, and, though
nowadays helping out their existence by a litde agriculture of
the most \ primitive kind, really dependent otherwise on the
collection bf forest produce. They occur also in the reserved
forests of thV neighbouring districts of what was British India, a
very primitive tribe of proto-Australoid type like those already
alluded to farmer south.
Although the population of the Nizam's dominions was Hindu,
the Nizam was a Muslim ruler and the nobility of the state were
of Arab extraction, many of them with pedigrees that went back
to the ancient aristocracy of Arabia in the time of the Prophet.

2
CHAPTER III

WESTERN. CENTRAL, AND EASTERN


INDIA
To the west of the Nizams dominions lay the province of
Bombay occupying the west coast from south of the Portuguese
colony of Goa to the gulf of Surat in the north. All down
this west coast the physical type is more markedly broad-headed
than elsewhere in the plains of India, though brachycephaly
is also strong in Bengal. Trading and business castes have
brought manufacturing towns like Bombay and Ahmedabad to
a high level of prosperity, particularly by cotton spinning, and
trade connexions overseas go back to a very early date. Among
the trading castes, particularly in Gujarat, the northern part of
the former Bombay province, Jains are well represented, as well as
Hindus, the Agarwalas, originating from the United Provinces,
largely Jainby religion, Oswals and Jaiswals from Mewar in
Rajputana, Khatris from the Punjab, and Bhatias, and the
Muslim Khojas. Bohras, a caste from Baroda State, do, or used
to do, a good deal of business with Spain, and in 1931 there
were some two hundred of them in business in Bilboa, Malaga
and other towns there. Among the seafaring castes the Kharvas
of Kachh, Kathiawar, Cambay, etc., seem to be of Rajput
extraction. They hardy sailors, and besides manning the
are
country craft that ply up and down the west coast of India, to
Aden and Zanzibar, and as far east as Singaporfe, they are familiar
as lascars in the liners running between Bombay and
Europe.
Prominent in big business
' in Bombay is the Parsi commu-
nity. The Parsis, who
are Zoroastrians by religion, emigrated
from Persia about the eighth century a.d. to avoid compulsory
conversion to Islam. They have been foremost among Indians
to adopt European manners and customs, and as a community
are one of much wealth, culture and refinement. They have
an unusual way of disposing of their dead, who are not buried
as by Muslims, nor burnt as by most Hindus, but exposed on
the tops of Towers of Silence where the bodies are eaten by
vultures and the fleshless bones disposed of in a well below.

^ aofphysical
lity
type the are more
Parsis
Indians of Bombay than another
from the genera-
distinct
community of foreign
WESTERN, CENTRAL, AND EASTERN INDIA 19

extraction, the Ben-i-Israel, who seem have arrived a couple


to
of centuries earlier from Arabia. These Jews are claimed to be
representative of the lost ten tribes, as they are said to have
no names which can be shown to be later than the Captivity.^
Gujarat and Maharashtra are coastal states, and while overseas
trade is a natural activity of some of their population other sea-
faring activities occupy many more. The numerous and humble
caste of Koli, dark in complexion and coarser in feature than the
trading classes, often cultivate the soil, but are also much occupied
with fishing and with manual labour in the mills. Physically they
approximate in type to the jungle tribes and Bhils referred to
below. Often employed as labourers, their tribal name has
been taken over in English as a synonym for the unskilled
labourer of India and elsewhere in the tropics ^the *
coolie *.

From a tribe the Koli has become a Hindu caste, but retains in
his appearance the physique of the earlier inhabitants of India.
Above him in the social scale is the widespread cultivating caste
of the west coast, the Kunbi, the typical peasant of western India
who probably contributed largely to the rank and file of that
fighting race the Marathas. The heard of in
Marathas, first

the thirteenth or fourteenth century, came into prominence as


a fighting race in the seventeenth century and nearly became
masters of India in the eighteenth. Their leaders rank as
Kshatriyas, though racially probably of the same extraction as
the shepherds and cultivators who supplied the rank and file

of their armies. In fact the word



Maratha in its narrower'

use is applied to a society which Rajputs or quasi-Rajputs,


iri

at the top, with Kunbis, Dhangars (shepherds), and Goalas


(cowherds) practise hypergamy, each group taking wives from
the one below, causing a superfluity of women at the top and a
scarcity at the bottom of the social scale. One sect of Kunbi,
known as Kadva Kunbi, claim to have been created by the
goddess Parbati, wife of Shiva, from the perspiration at her waist
(ked); they worship the goddess Umia Mata at Unja in the
Kadi district, and marriages take place only once in nine to
twelve years when the goddess speaks an occasion determined
*

by divination. Even unborn children are married on such an


occasion, the mothers performing a marriage ceremony on an

1 Auseful account of this caste will be found in Enthovens Tribes and


Castes of Bombay, i, p. 67.
THE BACKGROUND
understanding that it shall be valid if the two infants prove to
be of different sex. Girls for whom no husband can be found
when the goddess speaks, are married to a bunch of flowers
which is thrown into a well or a river, after which the girl can
remarry as a widow when occasion offers.^* Several of the
Maratha dans seem to be totemic, and the Marathas in general
worship, among others, a god spoken of as Khandoba (khanda =
a sword), a warrior incarnation of the Supreme Deity who is
depicted on horseback with two women, a wife, a Baniya (mer-
chant caste), in front of him and a mistress, a Dhangarin
(shepherd caste), behind him.* The spring festival was celebrat-
ed by a foray. The Marathas are noted for the part ,played
by women in public life, for their simplicity of living and the
affection of their leaders for their original village. The Mara-
than rise against the Mughals affords perhaps the sole instance
in earlier Indian history of a movement which was
primarily
rather national than religious, in spite of the appeal it
made to
anti-Islamic feeling. The term Maratha in its wider aspect
*

is used to indicate a number of castes


from Brahmans down
wards. Of the west of India Brahman castes <Mie may
mention
the Chitpavans with their fair complexions and
occasionally
even grey eyes, while of equal intellectual attainment
probably
is the important caste of Prabhu,
occupationally writers, who
have often held important administrative and official
posts.
Maharashtra, the name of the coimtry of the Marathas,
where
the Marathi language is spoken, is generally
associated with the
word Maratha, but it has been suggested that it is feally
derived
from the word Mahar which is the name for a widespread
exterior
caste of that country. Its social status is degraded on
account of the occupations it follows, but it is
commonly asso-
ciated with certain vUlage services, and the
fact that the Mahars
are commonly regarded as authorities on
village boundaries has
been cited as evidence that they are the original
inhabitants of
Maharashtra. This is not impossible, as they
represent an older
stratum of population than the broad-headed
Kunbi or the
fair-skinned Brahman and are numerous and ubiquitous.
Though outcaste the Mahars are Hindus, but this is more
Castes etcpp. 43 sq.,- Enthoven, Tribes and Castes,

Enthoven, Tribes and Castes, iii, do in sno
Russell, Tribesand Castes of the Central Provinces, iv,
s For the expression * exterior caste see p 504.^
^

Appendix A.
WESTERN, CENTRAL, AND EASTERN INDIA si

doubtful in the case of some of the jungle tribes that still survive
in small numbers in the west of India, tribes such as the Chodhra
or the Katkari, who up a meagre and nomadic living by
pick
down as labourers
collecting forest produce, or sometimes settle
on the land of their employers. One famous tribe of western
India, however, is still numerous and vigorous, and that is the
Bhn tribe, an important section of which is to be found in the
former Bombay province, though it extends widely into Rajputa-
na to the north and into central India eastwards. Bhili, the tribal
language, is an Indo-Aryan tongue, but it probably replaces an
older Dravidian tongue, and perhaps a Kolarian language before
that. The tribesmen, partially Hinduized now in most districts,
but in some professing a rather unorthodox form of Muslim
and tribal conscious-
faith, still retain their tribal organization
ness. Most them are now peaceful cultivators, but in the
of
past they have been renowned for brigandage and lawlessness.
They are a smallish, swarthy race, very expert in the use of the
bow. There are a number of different sections of the Bhil tribe
each divided into many exogamous and totemistic dans. They
have been said to venerate the horse above other animals, but
theiir chief festivals are those in honour of the dead. They are
inveterate believers in witchcraft, and their women, who have
much freedom and influence, used in former times to accompany
their menfolk into battle, using slings with great effect.

Associated like the Bhils with western India, with Rajputana


and with central India, as well as with the Deccan, is the caste
known as Banjara or Lambadi, a nomad caste of traders and
cattle herdsmen
with reputed criminal propensities. Their
former occupation consisted largely in supplying grain for
campaigning armies. They moved about, as many of them still

do, using their oxen as pack animals for transporting grain and
living in temporary camps. They used to do not a little business
in the disposal of loot and in the purchase and sale of children.
Their women are noted for good looks, domestic freedom, and
a remarkable headdress in which a shawl is- raised above the
head by a wooden horn resting on the back hair.

Turning eastwards from Bombay to central India there is little


marked change in the higher classes of society until the centre
of the peninsula is passed. Indeed, the influence of the Marathas
^2 THE BACKGROUND
carried their language at one time right over to the confines

of Orissa^ where many of the common people use a dialect called


Halbi which is based on Marathi. The Bhils of Bombay also
reappear in the Central Provinces, where they occupy the
Vindhya and Satpura ranges, but soon give way to other tribes
more typical of the central forests and highlands. Before the
Marathas, Rajputs from northern India had pushed into the
centreand south through the forest passes of the Vindhya Hills
and set up local dynasties in small states, often intermarrying,
itwould seem, with aboriginal rulers, and central India held till
1948 numbers of Rajput or quasi-Rajput states. Of the abori-
ginal tribes the Baiga are possibly the most ancient survival
that still retains a tribal consciousness. They practised until
recently, and perhaps still, in places, practise a shifting cultiva-
tion of a primitive type, using only the hoe and abjuring the
plough; and their marriage system has features that suggest a
survival from some long-abandoned system of marriage classes
analogous to those of the aboriginal Australians. It is still not
excessively uncommon ^ for a Baiga grandparent and grandchild
to be married
the difference in age of course may be even
lessthan 30 years in a people where unions are normal at the
age of 15 and not unknown at that of 12. Possibly there is a
belief in the reincarnation of a dead man in his graridson or
of a dead woman which makes it only
in her granddaughter
proper for. a couple to be reunited when the one who has died
young is reborn while the widowed partner still lives. At any
rate, a belief in reincarriaiion of this kind has been widely held
in many parts of the world.
The Baiga now speak a dialect of Hindi, but the Gond tribe,
who constitute the most numerous element among the forest
I

tribes of central India, have a language of their own that belongs


to the Dravidian family. Over large areas they have settled
down as ordinary villagers cultivating the soil with the plough,
though in forest areas they retain their primitive life and cus-
toms. Their national weapon is the tangi, a type of light axe
with a narrow, curved vertical blade, which every Gond man
Though the States Reorganization Act was not passed till
1956, most of
the smaller States were merged into larger units soon after Independence.
^ Vide Elwin, The Baiga, 180 Fuchs an
pp. sq. reports analogous custom
of the Gond and Bhumia vide his The Gond and Bhumia of Eastern
Mandla, pp. 168-9. f Vide infra, pp. ^52-3.
[
WESTERN, CENTRAL, AND EASTERN INDIA
carries in the forestand with which they have been kndwn to
kill even the tiger in single combat* They are oiganized in
totemistic clans with a ruling clan of Raj-Gonds as socially the
most important. This clan has given dynasties to many Gond
states and claims Rajput status. The Gonds had at one time
a dynasty of considerable importance which had suzerainty over
all the Gond country of central India, and the mins of its
palaces are still to be seen. Like the Baiga they are great
believers in and practisers of witchcraft, and they have family
and tribal gods which vary in number. The legs of their women
are elaborately tattooed, but where Gond and Baiga live as
neighbours the tattoo is put on (with soot as dye, and thorn

prickers) by women of the Baiga tribe, from whom no doubt


the Gonds adopted the practice. Human sacrifice used at one
time to be practised, but this was far from being limited to any
one people, or even any one group of peoples in India.
Although they differ in language the Baiga and Gond are not
dissimilar in physical type, and the same applies to Korku,
Korwa and other jungle tribes of this area, but the Chamar of
the Chattisgarh plateau in the east of it is of a distinctly finer
physicalmake-up than the generality of Gonds, though this
might not apply to the Maria Gonds of the Bastar State who
hunt the great gaur or bison of their forests to wear their

horns as headgear. The Chamar of Chattisgarh, although


belonging to that exterior caste of leather-workers whose touch
is polluting to caste Hindus, are here cultivators tilling the land

whose women have a great reputation locally for their handsome


features. There are other people too on the Chattisgarh plateau
of more primitive type similar to those in the adjoining plateau
of Chota Nagpur. Ghota Nagpur, while not actually part of
the Central Provinces, for it fdls in the province of Bihar, is

racially closely bound to central India, and geographically part


of it, and the tribes which people it, having remained in their
comparative isolation for longer than their neighbours in the
plains, still retain their tribal society. Some of them, the
Oraons, for instance, like the Gonds, speak a Dravidian lan-
guage, but more of them use one of the Kolarian languages
which have elsewhere been superseded either by a Dravidian or
by an Indo-Aryan tongue, or first by the one and then by the
other. Several of them retain the institution of the bachelors'
THE BACKGROUND
dormitory, where the boys of the village spend three
years in
educative discipline, the younger boys working under the
direc-
tion of the elder. Except in language and in some of
minor institutions, they do not differ greatly from their Kolarian-
s^aking neighbours, who inhabited Chota Nagpur before
they
did and who retain their distinctive tribal organization,
even
though their villages are now territorially mixed up
with those
of the Mundas. The Mundas are the most
important of these
other tribes, perhaps, but really form one
people with the
IQimas, Hos, Kols, Bhuiyas and others, all of whom
have a
simQar territorial organization, which is shared by
the Oraon,
and make a tribal feature of the Spring Hunt for
which a
nuinber of villages assemble annually. The Munda
tribe in
particular puts up rude stone monuments for its
dead somewhat
resembling the so-called cromlechs of
Britain. Another
tribe very closely akin to this group is
the Santal tribe, weU
known as hunters and labourers, who rebelled- in 1855 largely
M the result of the exactions of moneylenders
and dispossession
frorn land under dvil court decrees.' This
vigorous tribe,
ongmally located in the Santal Parganas of Chota
Nagpur, has
colonized the empty areas of the Rajmahal
Hills and is now
widespread in the Central Provinces, Bihar,
Bengal and Assam.
The pre-Santal oaupants of the Rajmahal Hills are
the Malers,
a tribe using, like the Oraons, a
Dravidian language, and
apparently like the Brahui of Baluchistan,
an oudying remnant
of the former Dravidian-speaking
population of northern India
generally.
On the southern side of the Chota Nagpur plateau
and in
toe east of the Central Provinces the
forests and broken hills
muse other remnants of Kolarian-speaking tribes
such as toe
Juangs whose women, at any rate till recently,
used petticoats
of leaves, which in hot dry weather
shrivel so fast thatfrequent
departures to toe forest for a change
of garment are said to be
nectary Another nomadic jungle tribe of this area
and with
sm^ar affinities is the Birhor, who subsist
without agriculture
Md are reputed to represent toe
Padaeans of Herodotus and
to have had the practice, shared
with toe Massagetae and appar-
Hwin, The Muria and their Ghotul, Pt. 11
; Roy. The Oraons,
pp.

pp.
to' vn'^w ^' r
fIndus of an Indian
and the West (el OMalley).
WESTERN, CENTRAL, AND EASTERN INDIA 25
ently some of the ancient Irish, of displaying their affection
for their deceased relatives by eating their bodies in piety.
If the Birhors ever did have this custom, they have abandoned
itfor at least a couple of generations, but similar
customs are
not unknown in other parts of the world, and there
is nothing
inherently improbable in the tradition of its
having existed
among the Birhors.

From the Chota Nagpur and Central Provinces border


the hilly
country exten^ towards the east coast and continues
to be the
home of primitive tribesmen until the coastal plain is reached.
Some of them, like the Gadaba and the Bondo, retain their
Kolarian language, others like the Bhatra now use either
Halbi, a
dialect much influenced by Marathi, or Oriya,
the language of
the Orissa plrin, or both. As in the case of the
Maria Gonds,
there a noticeable suggestion of the Mongolian about many
is
of
these tribesmen. Perhaps, as it shows little trace of the epicanthic
fold in the eye, it should be called palaeo- or
proto-Mongoloid,
and is more typical of the fringes of the Mongolian area than
the centre and may be due to an early distribution
of the
Mongolian type before the more pronounced features were
developed. There it is, however, for several trained
observers
have noticed it, particularly in the tribe known as Sawara which
has a distinctive language and culture in the Orissa hills.
The
language is Kolarian, and the culture has much in common
with other Kolarian-speaking tribes, particularly in the setting
up of stone monoliths, but differs from that (ff its neighbours
in the use of irrigated terraces for the growing of rice, a
feature
more suggestive of Assam than of the Orissa or central Indian
hill tracts.The funerary cult of the Sawara involves cremation
of the body, accompanied by the construction of a temporary
house for the soul and the erection of a menhir, in a special
grove reserved for that purpose, when the ghosf goes to join
his ancestors. Periodically a secondary funeral feast is
held to
commemorate all those who have died since its last celebration.
A people more numerous than the Sawara and cme which has
attracted much more attention in the past is the Kond or Kondh
(commonly written Khond) tribe, nearly allied to the Gonds of
central India both in speech and cultureindeed, it is probably
Dalton, Ethnology of Bengal, p. isto.
?6 THE BACKGROUND
to be regarded as a branch of that people, and the names are
perhaps really identical. The Konds are notorious for the
trouble they gave by opposing the suppression of infanticide and
still more by their obstinate addiction to the use of human
out often in a very cruel manner. Here the
sacrifices carried
lifeof the dead victim was associated with the harvest, and the
harvest consequently liable to fail if no life had been provided
to quicken and vivify the seed. The victims were sometimes
kidnapped from the adjoining areas, sometimes provided by
families in whom the duty was hereditary from among their
members. The Kond women tattoo themselves like Gond and
Oraon women, and the warriors used to dress like the Maria
Gond with a flamboyant headdress of bison horns and feathers.
The inhabitants of the Orissa plains no doubt share, in the
cases of many castes, the same racial origins as the tribesmen of
the hills. One caste now scattered throughout India, that of
the migratory earth-workers known as Odh, has a traditional
origin from Orissa, and, it has been suggested, has affinities with
Kolarian tribes. Thequasi-military Khandait caste of Orissa
now claims the rank of Kshatriya, but probably has closer local
ties than such a claim would suggest. The same may apply to
the nume? ms cultivating caste of Kalta. a famous Orissa is
land of shrines and. pilgrimages, the most famous being the
temple at Puri where Vishnu is worshipped in his incarnation
of Krishna, theLord of the World (Jagan-Nath). This is the
temple from which the car of Jagannath is dragged by pilgrims,
some of whom have in the past immolated themselves under the
wheels. The festival is probably a Hinduized survival of an
ancient fertility ceremony, but it is associated with a temporary

relaxation of the more rigid restrictions of caste, and it is


probably the catholicity of the worship of Krishna under the
form of Jagannath which has made it such a popular place of
pilgrimage for Hindus. Jagannath is a god of the people, or
t any rate has been, and although outcastes were excluded
from
the temple, even washermen could go into the temple court where
they could catch a glimpse of the idol, and rice once offered to
the god can no longer be a vehicle of any contamination, so
that outside the gate a hundred thousand pilgrims of all castes

12 Eiithoven, Tribes and Castes, ra, pp. 138 sqq.; Thurston, Castes and
Tribes, v, p. 424. ^
WESTERI^ , CENTRAL, AND EASTERN INDIA
can eat together and no one be defiled. Sir William Hunter,
indeed, recorded that he had seen a Brahman priest accept
this food from the hand of a Christian. Pilgrimage to Jagan-
nath secures all sorts of earthly blessings and gives liberation
from sin. Indeed, on the entry of the pilgrims by the Lion
Gate of Puri a man of the scavenger caste used to strike them
with his broom to sweep away all sin and compelled them to
promise not to disclose the secrets of the shrine.^*

Geographically Assam is separated from Orissa by the great pro-


vince of Bengal, much larger than either, but culturally these
two small provinces have had much in common and probably
retain remnants of a common culture which once spread over
both sides of the Ganges delta, but which has been greatly
modified at the centre by later intrusive elements both of race
and culture. While Orissa consists of a coastal plain with
jungle-clad hills behind it, Assam proper is a riverine plain
with hills almost all round it. Like Orissa it contains Hindu
shrines of great antiquity and survivals of ancient and probably
pre-Hindu fertility cults. But it differs from Orissa in having
a much stronger Mongolian element in its population. Traces
of this are to be seen not only in the ordinary cultivating castes
like Kochh and Kalita but occasionally even in the higher
castes whose forebears migrated from the United Provinces,
while the Ahoms of the Assam valley represent the Shans from
Burma who conquered the country in the thirteenth century
ftL.D. Many of the Brahmans and Kayasthas of the province have
names which embody Ahom titles, and the Assamese language,
though belonging to the Indo-Aryan family, contains many Shan
words in its vocabulary. The hills round the valley contain a
great variety of tribes with probably more diversity of custom
and language than any comparable area in the world. The
valley is drained by the Brahmaputra river, which after debouch-
ing southwards from Tibet turns westwards and runs roughly
from north-east to south-west. On the north bank the very
Mongoloid tribes are linguistically and physically allied to the
Tibetans and Bhutanese. The Daflas, Abors and Mishmis all
inhabit inaccessible forest-dad mountains subject to an extremely
Orissa, I, p. 86. ^ >
Ibid., p. 142. It is probable that by now (1946) outcastes are again
admitted to the temple of Jagannath.
the background
high rainfall. Accustomed for centuries to descending on
the
plains for loot and blackmail, it is
only in comparatively recent
years that they have been reduced to
behaviour which is more or
less peaceable, and no very
detaUed accounts of them have
been published. The Apa Tani tribe
in this area has an
daborate system of terraced irrigation, but
the agriculture of
the other tribes is much more primitive.
On the south bank of
the Brahmaputra and occupying the
hills running thence south-
wards to the Bay of Bengal are the two
groups of tribes known
to us as the Nap
tribes and the Kuki-Chin tribes,
together with
pme interme^ate ones which have cultures with something
m common with both these groups. Between them
they use
some forty diflEerent
languages or more. In some Naga villages
two quite different languages are used on
opposite sides of
Ae village street, and though one of the two will be understood
by everyone the other wiU remain a private
language unknown
to half the village. Customs and cultures
differ almost as much
as languages, although all Nagas are
head-hunters by predilec-
aon at any rate. Some use a primitive method of shifting
culti-
vation, clearing the forest, sowing crops
for a couple of years
and then letting the land go back to jungle till the growth is
ready to cut and burn again; others, on the
other harid^ have
a careful and elaborate system of irrigation,
leading water round
the contours of their steep hills for miles
to flood their rice
terraces. While some have sacrosanct chiefs so sacred that they
must not touch the common sofl, others are governed
by councils
of elders, and others again are extreme
democracies. All are
more or less ancestor worshippers. Like the hill tribes
of Orissa
they erect monoliths as memorials for their
dead and vehides
for the operation of the life-essence of
the dead in the fertiliza-
tion of the crops. Indeed, the head-hunting
cult itsdf is based
on a bdief in a regular cyde of life which is a material
essence
residing particularly in the head and which can thus be
abstracted from one owner and carried off to
benefit the village
and family of the head-taker. The Kuki-Chin groups
of tribes,
on the other hand, are almost all organized under
the rule of
hereditary chieftains and used to be slave-raiders rather than
Since Aeabove was written Sachin Roy has made a
(rood start on the
iWam-Afinyong Culture and Furer-Haimendorf has
nnhfiX!?
published The Apt Tents and their Neighbours.
W EST ERN CENTRAL, AND EASTERN INDIA
, 39

head-hunters. They are very dannish indeed, with great loyalty


to fellow-dansmen and to their chiefs. The Mongolian element
is them than in the Nagas, who, in spite of their
stronger in
Tibeto-Burmese languages, have racial and culturd links with
the peoples of the Indian Archipelago and the Pacific Ocean.
The Nagas and Kukis are patrilineal, but that is not the case
with the Khasis and Garos who inhabit the hills to the south
of the Brahmaputra running westwards towards Bengal. Both
these tribes reckon descent through the mother rather than
through the father. Among the Khasis property in land or
homestead passes from a mother to her youngest daughter, who
also inherits the office of family priestess.^ Where the head of
the state there were twenty-two small Khasi states ^is hereditary,
the chieftainship passes to the chief's youngest sister's eldest
son, but several of the Khasi states have elected rulqrs.^
Although the women own the property, only the men have votes
at election time, that is, when a local ruler is being elected, for
if it is member of the Assam Legislative
a matter of electing a
Council, women, under a recent electoral law, have a vote as
well. The Garo tribe, like the Khasis, inherit name and pro-
perty through the mother, but a woman's daughter is expected
to marry her father's sister's son as a matter of course, so that
while the ownership of land passes from mother to daughter,
the management of the estate passes regularly from, unde to
nephew in a different family, and it is a rule that when his
uncle dies the nephew, his sister's son, that is, must marry his
mother-in-law as well as her daughter whom he has already
married.^
Half-way between the Himalayas and the Bay of Bengal is
the state of Manipur. The Maharaja and the prindpal
inhabitants of this upland valley, a place of great natural
beauty, are Hindus, but before the state was Hinduized it had a

culture of own, probably Shan or Chinese in origin, and


its

survivals of this are still to be seen in the distinctive costumes


and customs of the inhabitants. Among other distinctive
Gurdon, The Khasis, p.83, Ibid., sect, in (pp. 62 sqq.).
Playfair, Gatos, p. 68. The Caribs make marrying the wife's
The
mother (as well as her daughter) an alternative to complete avoidance
(Briffault, The Mothers, i, p. 264, quoting La Borde, and Quand), and
Spencer cites an instance of a Tierra del Fuego native marrying his wife's
mother during her daughters life (Spencer's Last Journey, p. 72).
30 THE BACKGROUND
features are its national sports of polo, played between whole

villages on small ponies on their grazing grounds with unlimited


numbers a side (it was from here that Great Britain adopted
the game in a modified form), and boat racing, an annual
event in which elaborate dragon-boats are rowed, as in Ghina,
by crews in special gala dress while the spectators on the banks
also take part by helping to push off the boat they support when
it runs aground on the muddy bank.^*

There are many immigrants in the plains of Assam. Some of


them are labourers in tea gardens, the bulk of whom are Hindus,
as are most of the Assamese proper, while some have been
converted to Christianity. They come from all parts of India,
but mostly from Chota Nagpur and the neighbouring parts of
central India. The majority of the more recent immigrants into
Assam, however, were Muslim cultivators from Bengal, sturdy,
industrious, but rather turbulent people who tended to oust the
Assamese from the south and west of the valley. The Sylhet
and Kachar districts of southern Assam, which belong geogra-
phically to Bengal, had, long before the emergence of Pakistan,
already become in effect Bengali colonies, the Kachari, who once
ruled there, having become quite unimportant numerically.
They are allied in origin to the Garo, Kochh, and the people of
the north bank of the Brahmaputra river, and in the hills still
speak their own distinctive language.

Between Orissa to the south-west and Assam to the north-east the


population of Bengal seems to contain a brachycephalic element
of population comparatively lacking in the other two, which
has perhaps come down the Ganges valley, but which is much
more prevalent in Bengal than it is in Upper India generally.
There is also in Bengal a brachycephalic Mongolian element
to be seen in the people of the Chittagong Hills on the
Burma
border and at the foot of the Himalayas, but the broad-
headedness which is typical of the Bengali Brahman and found
also in other castes of high Standing like Kayastha and Bajdya
isof a different character and to be associated with that found
in western India. It seems to be strongest in the highest
caste
and to decrease with desicent in the social scale. Thus the
Chandals or Namasudras who are low in social position are
Hodson, The Meitheis, ^, "
^ ^st. '
f.
WESTERN, CENTRAL, AND EASTERN INDIA 31

longer headed than the Brahman, or Kayastha of Bengal and


the Rajbansi and Bauri still more so.^ On the other hand,
the Namasudra is often fair-skinned and apt to be leptorrhine
insome contrast to the bulk of the Muslim population of Bengal
which is dark-skinned and rather flat-nosed, suggesting a consi-
derable strain of the Australoid type, a type also apparent in
many of the lower-class fishing and cultivating castes such as
Bagdi, Bauri, or Kaibartta.
It is the cultivated and educated Bengali who has been
probably more than any other Indian the interpreter of the
Englishman to the Indian. Although no more intellectually
acute probably than the educated Madrasi, Parsi, Maratha, or
north Indian, the receptivity of the Bengali has made him very
quick to appreciate and to reproduce the British point of view,
and often to carry it with characteristic subtlety to a logical
conclusion to which the English themselves would be unlikely
to take it, particularly in the interpretation of legal principles.
Bengali literature, too, in an English form is probably more
familiar to Englishmen than that of other parts of India, largely
as a result of the activities of Tagore, a Pir-Ali Brahman of
Bengal, and of Bengali novelists such as Bankim Chandra
Chatterji and Romesh Chandra Dutt. The area known as the
Bikrampur pargana in the Dacca district is famous for the
extremely high proportion of lawyers, clerks, and literati
generally whom it supplies or used to supply widely to northern
and eastern India, but particularly, of course, to Calcutta, which,
taken with the contiguous municipality of Howrah, is one of
the most populous cities in the British Commonwealth. Calcutta,
although naturally predominantly a Bengali city, contains
people from all over India and Pakistan to say nothing of for-

eigners, and had a population of one and a half million as long


ag6 as 1931. The commercial abilities of the Bengali are dis-
played by such castes as those of Shaha and Teli, but it would
be a mistake to think of Bengal as a purely Hindu province. In
point of fact the Muslims actually outnumber the Hindus, their
stronghold being eastern Bengal,*^ where a very high proportion
of the cultivators are Muslims. The bulk of the Muslim popu-
lation are no doubt of more or less aboriginal extraction
20 Risky, Tribes and Castes of Bengal, Anthropometric Data, i, vii.
21 Now, of course. East Pakistan.

M ./
3* THE BACKGROUND
converted to Islam, but the aristocracy indudes a number ojE an-
cient families of Arabic or Mughal lineage. It is the Bengali
Muslim of eastern Bengal and of Sylhet who supplies a very high
proportion of the Indian seamen and lascars serving in British
merchantmen. In the extreme east of Bengal the Chittagong
Hill Tracts contain more or less Mongolian
tribes, some of
them Buddhists like the Maghs and Mru, others like
>

the
Chakma more or less Hinduized, but all really having more
in common with the lull tribes to their north
in Assam than
with the inhabitants of the plains of Bengal. In the
Himalayan
and sub-Himalayan areas again many Mongoloid tribes,
like
the Lapcha, are found, and in the delta areas of
eastern Bengjal
a number of Feringhees, now mostly cultivators,
who
Hair^i
descent from the Portuguese pirates who once
occupied Chitta-
gong and thence in the seventeenth century raided
Bengal as
far west as Dacca and Calcutta.
CHAPTER IV

NORTHERN INDIA
Proceeding up the Ganges one passes from Bengal into Bihar,
and thence into- the province of Awadh (Oudh). These areas
form a meeting place of the outer and inner bands of Indo-
Aryan languages and are likewise, at any rate in the case of
Bihar, the area of change from the broad-headed Bengali type
to the long-headed type characteristic of Upper India. Here-
abouts was the stronghold of Buddhism before it was overthrown
and extinguished by the Brahmanistic reaction, while farther to
the west, between the Ganges and the Jumna, was the centre
of Hinduism par excellence, the sacred land from which
Brahmans were distributed to many other parts of the peninsula.
There are of course many castes and classes of Brahmans, and
they are not all so intellectual and refined as the high-class
pandits of Benares. Moreover, it is only a minority of the
Brahman castes which exercises priestly functions. Brahmans
practise in all the learned professions as lawyers, land agents,
clerks, etc., and act not infrequently as cooks, since most Hindus
will eat food cooked by a Brahman of good standing. It is
probable that more than one racial element has gone into the
composition of most Brahman castes, and perhaps they represent
the priesthoods of what were once different, even antagonistic,
religions, but there is no gainsaying the fact that the Brahmans
of Upper India in general have preserved their purity of blood
and their intellectual ascendancy for a very long period.
Brahmans like the Gayawals of Gaya, who maintain themselves
in idleness on the offerings of pilgrims to their holy city, or the
Chaubes of Mathura, who are more celebrated for their skill in
wresding than in learning, are generally held in comparative
disrepute;^ Apart jErom Rajputs, the castes which probably
stand next to Brahmans in social position and are in some ways
allied to them are the Babhan, and the Bhat. The Babhan
are land-holders particularly associated with Bihar and daim
to be Brahmans who have given up the priesdy function for a
life of agriculture. Their social divisions, however, suggest
aflBinities with Rajputs rather than Brahmans. The Taga caste
1 Crooke, Northern India, pp. 99 sq.

3
34 THE BA CKGROU ND
of the Upper Ganges valley and the neighbourhood of Agra
have similar traditions and a similar status. The Bhats, who
daim a Brahman extraction, as the similar caste of Charan does
a Rajput one, are hereditary genealogists and heralds who used
in particular to act as escorts to protect persons and property
passing through areas of freebooters in northern India where
brigandage was almost a profession. The Bhat secured a safe
passage for his convoy by committing suicide if it were plun-
dered, when the firm belief that the plunderers would be for
ever haunted by his outraged ghost was enough to deter -them
from becoming the cause of his death.
The ^eat cultivating caste of Bihar and Awadh is that of the
Kuirmi, a caste equivalent in status to the Kunbi of the Maratha
country; while the Ahirs, graziers, occupy much the same sort
of position. Very numerous also in
this part of India, though
absent elsewhere, the caste or tribe known as Bhar who have
is

sunk to the class of labourers and small cultivators, but who


seem at one time to have been the ruling people of Bihar.
Many old forts and reservoirs ^e attributed to them. The
Kalwar who distils spirit from molasses or from the flowers of
the mahwa tree has a status as low as the Bhar, not because he
represents a conquered and dispossessed tribe, but on account
of his occupation; and occupation again
is also in part at any
rate the reason for the low standing of the Pasi who taps the
toddy palm for liquor, the Dhobi, who is a washerman, the
Khatik who is a butcher, or the Chamlr who tans hides and
works in leather. Below all these are the scavenging castes, the
various grades of Bhangi or sweeper and the numerous caste
Dorn. The Dorns, though serfs and outcaste for centuries
if not millennia, possibly represent a once
respectable tribe.
Przyluski would identify them with the Odumbara, a more
or
less legendary tribe of shepherds in the Punjab. They are here-
ditary executioners, and their assistance is necessary at funerals
to hand the wisp of lighted straw with which the chief mourner
ignites the pyre, but theyperform any menial work, are often
nomadic and are most unorthodox in their willingness to take
food from anyone except a Dhobi, and food, at that, of any
kind at all, on which account the Dorn is very offensive to
the
*Ibid., pp. 107 sqq.; Malcolm, Centrallndia,
pp. isi-q.
^ Les Udumbara,

N ORTHERN INDIA $5

orthodox Hindu, though he acts on occasions as a musician. The


Dorns in the Himalayan foothills, in Kumaon and Garhwal, have
retained a considerably higher status and something of their tribal
organization, and are subdivided into various artisan groups
practising different crafts. These hill Dorns now repudiate any
connexion with the scavengers of the plains.^ Going westwards
from Upper India into Rajputana one enters the most romantic
area in the peninsula, a land of ruined forts and ancient strong-
holds, famous for deeds of valour and self-sacriffce, crowning
rugged heights above old and picturesque towns. Many are
the monuments to long-dead warriors, stones carved with a
horseman and his weapons and the heavenly bodies, and still
frequent are the stones carved with a single upraised arm indi-
cating the place where a widow burned herself upon her
husbands funeral pyre. There are many inhabitants besides

Rajputs the famous Bhil tribe has been already mentioned;
the Bhilala are hybrids of Bhil and Rajput ancestry; the Mina
are a tribe of marauders, for that is what they used to be,
closely akin to the Bhil in character. There are other tribal
communities too, Meo, Mer, and Merat for instance, and many
castes of cultivators and others often claiming quasi-Rajput
Lodha, and the great Baniya or trading castes
status, like the
Agarwala, Oswal, Parwar and so forth, who likewise claim to
be of Rajput stock like the Charan or bards; but the Rajput
clans are the people who matter here.
Both by and in physical characteristics the Rajputs
tradition
appear to be of mixed origin. Popularly they are regarded as
the lineal descendants of the Rajanya warriors who invaded
India in the Rigvedic age and who at that time took precedence
even over Brahmans. A recent writer has associated some
features of their culture with a still earlier race of invaders with
a matrilineal culture,^ while from many tradi-
it is pretty clear
tions and inscriptions that the Rajput dans were augmented
by Saka, Kushan, and Hun invaders at a much later date. It
is possible that the traditional division of the Rajputs into

^ Turner, Caste in the Kumaon division and Tehri-Garhwal Slate


*

in the Report on the Censm of India, 1931, vol, i, pt, iii a, pp. 17 sqq.;
Crooke, and Castes, ii, s.v. Bom.
5
Ehrenfels, Mother-tight in India, pp. 1:58 sq.
Crooke, Tod's Annals and Antiquities
of Rajasthan, pp. xxxi sqq.;
Baines, Ethnography, p. 31.
36 THE BACKGROUND
descendants of the Sun, of the Moon, and of Fire, reflects such
a threefold origin and that, while the descendants of the Sun
and of the Moon go back to the Rigvedic age and earlier, the
Fire-descended Rajputs were of later origin and became Rajputs
in virtue of a ritual of purification by and rebirth from fire
after the Hindu reaction against Buddhism had devastated its
Rajput supporters. There are generally said to be thirty-six
royal clans of Rajputs, but the lists vary and some include the
Huns, and also the Jats, which latter are more often regarded
as a lower caste and who are perhaps derived from
comparatively
late immigrants. Indeed, the distinction between the Rajput
and the Jat is in some degree more social than racial, and the
same applies again to the distinction lower down the scale
between the Jat and the Gujar. The principal clans of the
Solar line are those of Sesodiya, Kachwaha and
Rathor; of the
Lunar line those of Yadava and Tonwar; and of the Fire-
descended lines those of Ponwar, Chauhan, Parihar and
Solanki.
Some inscriptions record the marriage of Rajput princes, among
the Sesodiya, for instance, with Hun wives. Many of the
dynasties ruling in Rajputana states in 1946 can be
still
shown to
have been established between the seventh and eleventh
centuries
A.D. The earliest Sesodiya inscription in Rajputana is dated
A.D. 646, when that dan migrated eastwards from Gujarat to rule
in Udaipur. The Bhatti, a sept of the
Yadava or Jadu dan,
came in the eighth century to Jaisalmer and still
rules there,
and about the same time the Chauhan settled in
Sambhar, but
later migrated southwards to Sirohi,
Bundi and Kotah states.
The Kachwaha came from Gwalior to Jaipur early in the
twdfth century, and the Rathor settled in
Jodhpur in the
thirteenth. The Princes of Bikaner and of Kishengarh also
belonged to this illustrious dan. It has
been said of this clan,
which IS freer than most from taboos
on food and
tonk, that they are soldiers first and
Hindus after that.
The Rajputs, with their quasi-feudal tribal system, their
romantic character and chivalrous valour,
and their punctilious
regard for personal honour, have
been well described by Tod
in his Annals and Antiqvities
of Rajasthan, which, although his
speculations as to origins and affinities
are out of date has
rightly been described as the
most comprehensive monograph
ever compiled by a British officer describing one of the leading
NORTHERN INDIA 37

peoples of India The


adventurous disposition of the Rajputs
was not confined to the men, for their women often shared the
dangers of war and sport and often exercised a healthy influence
on public as well as on domestic affairs. It was usual for them
to burn themselves on their husband's pyre, and there are
many authentic stories of the final sacrifice of their lives in the
rite known as johar in order to save themselves from dishonour
in the hour of defeat. The Bhats, already mentioned, acted as
genealogists, bards and preservers of tradition for the Rajputs,
while the Baniya classes largely provided the revenue and minis-
terial administration required by Rajput chiefs.^

The population
of the Punjab the land of Five Rivers ^which
has often proved the best of all recruiting grounds for the Indian
army, falls generally into three groups Hindu, Sikh and
Muslimbut the differences between them are rather religious
than racial, at any rate if the Pathans and Baluchis of the
western Punjab be excluded. For the Punjabi Muhammadan
who provided such an important element in the Indian army is
commonly of Rajput extraction, as are many of the Sikh rulers,
while the Jats who form perhaps the most important element
in the population of the Punjab may be either Hindu or Sikh
or Muslim, though the last are in the minority, and the typical
Punjab Jat is probably a Sikh. The Jat is a typical yeoman,
devoted to agriculture* and not particularly concerned in satisfy-
ing the requirements of orthodoxy, at any rate as a Hindu, so
that he takes a lower social level than the Rajput. His industry
is unceasing and every member of his family shares the work
in the fields.His reputation for stolidity and reticence has
caused him to be described as having grown grave and impassive
*

like the great white oxen which he prizes so highly. The


'

Sikhs of the Punjab started as a sect within the Hindu pale


about the end of the fifteenth century. It was the great Guru,
Govind Singh, about the end of the seventeenth, who organized
them on a military and political basis and who emphasized the
social aspect of the first Guru's reforms. Taught by him the
Akali Sikhs at any rate discard the sacred thread of the 'twice

Cole, Rajputam Classes, ch. iv ; Rmsell, Tribes and Castes, jv, pp.
414 sqq.; Enthoven, Tribes and Castes, ni, pp. i?9q sqq.; Rose, Tribes and
Castes, III, p. 300 ; Ibbetson, Panjab Custes, pt. iii Crooke, Tribes and
;

Castes, III, p. Crooke, Northern India, p. 93.


TBE BACKGROUND
barn * (i.e. initiated) Hindu, repudiate caste, eschew the use
of alcohol and tobacco, the wearing of hats or caps, and the
eating of all meat not killed by decapitation; the Akali Sikhs
pay no make no pilgrimage to the great Hindu
bride-price,
shrines, forbid witchcraft and the taking of omens, the observance
of social impurity at births and deaths, and Brahmanical usages
in worship. On tjie other hand, they accept the Hindu ways
of attaining communion with the deity. Strict Sikhs not only
always wear
the five K's *, but will not eat food not cooked by
a Sikh who is himself wearing them at the time of cooking.
These five K's are kesh (uncut hair), kirpan (a dagger worn in
the hair), kanga (wooden comb), kachh (shorts) and kata (iron
bracelet). Sikhs are fond of athletics and games; they make
very good mechanics, and their fighting qualities are too well
known need comment.
to
Two important castes in the Punjab who are much more
often Hindu than Sikh or Muslim are the Khatri and Arora
castes, which stand in a reciprocal relationship very similar to
that of Rajputs and Jats. The Khatri at any rate claim to be
of Kshatriya origin, but the principal pursuit of both these
important and numerous castes is trade.^ Much of the com-
merce of Afghanistan and central Asia is in Khatri hands, and
they have at all times shown ^eat administrative ability. Some
of them are Muslims, when they are known as Khoja. These
Punjab Khojas are generally Sunni, but the Khojas of Bombay,
derived largely from the Hindu Lohana caste of Sind, are
Shiah and followers of the Aga Khan.
As the Rajput is to the Jat so is the Jat to the Gujar,^^ but
the Gujar, though physically and in his social system not so
very different from the Jat, is a poorer cultivator and by pre-
ference a herdsman, The Gujars seem to have become estab-
lished in the Gujarat district of the Punjab before the seventh
century a.d. and to have come in with one of the Hun invasions.
They are often associated with the Ahir caste, which is also a
caste of graziers rather than cultivators, but is quite distinct
from the Gujars, Very low down, almost at the bottom of the
social scale inHindu society, comes the numerous caste (or
tribe) of Chuhra, which does the scavenging of the Punjab and
* Rose, Tribes and Castes^ i, pp. 676 sqq.
Rose, Tribes and Castes^ ii, pp. 501 sqq., 507.
Ibid., pp. 306 sqq.; Ibbetson, op, cit,,
pp, i8f sqq,
NORTHERN INDIA g9

which is possibly here the parent caste of the Chamar m


tanner. The Chuhra has been held to be the remnant of an
aboriginal tribe, but in point of fact his physical type diflEers
but little jErom that of other inhabitants. His outcaste position
is due largely to his occupation, and when he turns Sikh he
becomes a Mazhbi, and the taint of his origin prevents his ever
being admitted to full social equality with the Sikh of Rajput
or Jat origin, despite the Sikh repudiation of caste. When he
turns Muslim he becomes a Musalli and again retains some
taint of his despised origin.^^
It has already been pointed -out that the Muslims of the
Punjab, except for the Pathans and Baluchis of the west, are
generally of similar race to the Sikhs and Hindus and are
commonly Jat or Rajput in origin. There are, however, others,
the Awans for instance who claim to be Arabs by extraction and
claim descent from Ali, the Prophets son-in-law. They are
exclusively Muslim and probably the descendants of some of
the earlier Muslim invaders of the tenth century or earlier.^*

The earliest Muslim invaders appeared in the eighth century in


Sind, and this small province has since been predominantly
Muslim in religion, but the blood of the inhabitants is mixed
indeed. Sind seems at one time to have been occupied by
Rajput clans, and it is possible that in sonae cases clans were
still Buddhist when forcibly converted to Islam by conquering

Arabs from the west. There are, however, clear traces of the
Sakas and Huns in Sind, and the clan known as MahSr probably
preserves the name of the Epthalite Mihirs. The Rajputs of
the tenth century succeeded in expelling the Arabs, but were
converted to Islam by the later Muslim invaders from the north.
The Rajput title Jam is still used by Muslim descendants of
the old Rajput houses. Of the non-Muslim castes the Bhatias
and Lohanas are among the most important. They are very
similar in custom, both are trading castes, and the Bhatias at
any rate probably of Rajput extraction. The Bhatia are mostly
Hindu, but the Lohanas arc many of them Sikhs.
Of the Muslims of Sind, many are of Baloch origin, parti-
cularly in the west, and they are often organized as the followers

Rose, Tribes and Castes, n, pp. i8s sq.


rs sq.; Ibbctson, op. cit., pp. 169 sq.
pp.
40 THE BACK G.R O U N D
of some particular pir or saint holding office as
a religious
and hereditary leader. These pirs exercise great influence on
their adherents and obtain very considerable
revenues from
the offerings of their clientele over whom they used to enjoy
quasi-judicial and executive powers. In some cases, in spite of
their saintly
title, they have been little better
than leaders' of
banditti. A
notorious case ,in point is that of the Pir
Pagaro
or Pir of the Turban. This pir is a
hereditary religious
leader descended from a family which entered
Sind with the
Arabs in a.d. 711., whose followers, known as
Hurs, form a sect
of quasi-religious criminals. Aposition not uiilike that of a
pir is also held by the head of the Muslim
conununity known
M Meman or Momin who in Sind consist of Lohanas converted
from Hinduism to Islam by the descendants
of a Muslim saint
who died at Baghdad in the twelfth century a.d. They
still
retain certain non-Islamic customs such
as that of not allowing
a daughter or widow to inherit
property and of consulting
astrologers, but are very punctilious
about performing the pil-
grimage to Mecca.

West of Sind lies the mountainous country of Baluchistan.


The dominant race is the Baloch, which is probably
derived
from Iranian stock that migrated into its
present country
between the mid-thirteenth and mid-fift^nth
centuries a.d.,
having been originally located somewhere
about the shores of
the Caspian Sea. The Baloch tribes
may be divided roughly
mto two groups, the northern or Sulaimani
Baloch and the
southern or Makrani Baloch, separated
from each other by the
Brahui tribes. Baluchistan is at present
an arid and inhospit-
able country of rock and sand
sprinkled with oases and the
remnants of bygone earthworks suggesting
that it was once a
country of woo^, springs, and plenty,
a land of fountains and
of depths, drawing water from the
rain of heaven , though now
steady turning into desert under the encroaching
sand and
needing to be irrigated by means
of underground waterways,
mts of It contain as few as one
inhabitant to 50 square miles,
he Rind tribe is said to be of Arab
stock, but differs little
physically from the other Baloch
tribes, all of which have
been
0/ MacGregor, Report on the Census

NORTHERN INDIA 41

modified by admixture with the Brahui. The Baloch are orga-


nized in tribes, each with a chief whose post is hereditary in a
particular family; they are nomads by instinct and where it is
possible still wandering pastoral life, though circum-
prefer the
stances nowadays often compel them to setde permanently in
one place as cultivators. Where they are still nomadic, their
villages consist of little enclosures of rough stone masonry S or
4 feet high over which a temporary roof of matting is spread
while the hut is occupied. When the time comes the occupants
move off to another grazing ground, taking the roof with them.
Their wealth consists in camels, catde, sheep and goats, and
making of carpets
their life is extremely primitive, although the
is an important industry. By inclination they are bandits and
catde-lifters, and the tide Rahzan (=' Highwayman') is a tide
of honour.
The true Baloch type is brachycephalic and the Baloch
dans are patrilineal, but the Brahui, the other important indi-
genous people of Baluchistan, were probably long-headed origi-
nally and traced their descent and kinship through the mother.
They speak, too, a Dravidian language completely different froro.
Balochi and allied to those in southern India. It is no doubt
an isolated remnant of what was once the general linguistic stock
of northern India as it is |iow of southern. The Brahui have
largely absorbed Baloch blood and so much admire the short
Iranian head that they mould their children's skulls to make
them shorter headed; they bind their foreheads with a smooth
tight bandage, give them soft pillows of millet, and in the case
of a girl, the shape of whose head is more important than that
of a boy's, they bore three or four holes in the ears so that if by
chance the infant turns on the side of her head the pain makes
her turn on to her back again, till the back of her head is per-
manently flattened. Balochis and Pathans, however, also follow
similar practices, and the Brahui even goes so far as to attempt to
mould the head-forms of his domestic animals ^lambs and kids
while horses are bandaged to give them a slender foreleg above
the knee. But the Brahui does not seem to practise female
(as well as male) circumcision as many of the Baloch tribes do.
Of the other tribes in Baluchistan the humble nomad Lori
travelling tinkers, and workers in gold and silver, minstrels,
15 c. Dames, The Baloch Race, p. 11.
4, THE BACKGROUND

musicians, midwives and menials call for mention. They seem
to correspond in some respects to the Dom of northern India,
but claim descent from the youngest son of the Prophets unde
and an origin from Aleppo. They are gipsies and most at home
on the Makran coast, but nomadism is typical of all Baluchistan.
A third of the population live in tents or temporary huts, and
a very great proportion of the population migrate with the
seasons either from one part of Baluchistan to another or from
Baluchistan to the plains of Sind and back again.^*

On the North-West Frontier proper, as in Baluchistan, we are

almost outside the range of caste. Many of the nomads who


are to be met with in Baluchistan are Pathan pawindas or
caravanners from Afghanistan. No satisfactory distinction can

be made between the terms Pathan and Afghan, but the

term Pathan is commonly applied to several millions of


mountaineers occupying the mountains in the north-wtest of


India on both sides of the Afghanistan frontier and speaking
the Pashto language. Their emigration towards India seems to
have begim about the fifteenth cehtury and probably started as
an annual migration leading permanent settlement. Their
to
migration into the Kurram valley is recent and two tribes of
Pathans are still nomads. They a^ now bounded roughly by
Persiaon the west, Dardistan on the north, Baluchistan on the
south and the Punjab to the east. The Pathans are not orga-
nized politically but tribally in anumber of well-known tribal
groupsGhilzai, Mohmand, Afridi, Wazir, Orakzai, Yusufzai,
Mahsud, and so on. The tribes are subdivided into septs and
families, each group having its leading manmalikwhile the
malik of one group, generally the eldest member of the tribe,
acts as chief of the whole tribe. He possesses, however, influence
rather than power, and the real authority, in so far as there
is any, rests with the jirgah or council of elders, that is, of all

the maliks of the tribe. Each tribe, generally speaking, has a


more or less corporate existence as a geographical, as well as

genealogical, unit. The blood feud is a well-established custom,


and the internal jealousies and Pathan groups
dissensions of
are well illustrated by the fact that the same word in Pashto
serves both for *
cousiii
and for
enemy and a derivative both

Bray, Report on the Census of Baluchistan, igii, pp. 4s sqq.


N ORTBERN INDIA 43
for *
cousinhood ' and for *
enmity The Pathan has been
described as treacherous, bloodthirsty, cruel, and vindictive in
the highest degree, but he has a code of honour of his own to
which he adheres strictly enough and which imposes on him
three sacred obligations : hospitality to all vi^ho may ask for it;

protection, even for an enemy who comes as a suppliant; and


the duty of revenge by retaliation. On the frontier they live in
fortified villages with stone houses in commanding positions
which serve as watch-towers or as places of refuge. The Pathan
tribes formed a very valuable source of recruits for the Indian
army of high-class fighting quality.^
Towards Dardistan on the edge of the Pathan country are
found what is left, after constant harrying by the Afghans, of
the Red and the Black Kafirs. These tribes were, till recently
at any rate, head-hunters, speaking a very early form of the
Sanskrit language, using carved wooden vessels more suggestive
of Scandinavia than of India, and tripods described as Grecian
in type. One account describes them as burying their dead
erect in the snow, as extracting the heart and liver, burning
them on an altar and sometimes eating the ashes out of piety.
They erect carved wooden memorial figures of their dead and
used to be given much to dancing with great activity, though
now most of them have been more or less compulsorily converted
to Islam. In view of their language and their physical type with
lightish hair and eyes, tall stature and long heads, it seems not
unlikely that the Red Kafirs at any rate may represent some-
thing like the original stock of the Indo-European invaders of
India in the second millennium b.c.^
Going eastwards among the Khos of Gilgit, the Dards, Baltis,
Brokpas, and Burusho, the physical type is broader headed and

darker, and among the Kanets of the Kulu valley, and the
Garhwalis, who again form an important element in the Indian
army, a shorter dark long-headed type is found akin to that of
the Indian plains. In Ladakh and in Lahaul and Spiti the
inhabitants become definitely more Mongoloid, and this latter
type predominates among the Gurkhas of the independent state
of Nepal as it does likewise in Sikkim and Bhutan to the east

Rose, Tribes and Castes, ni, p. siQ n.


Barton, Indians North-West FronUer, Rose, op. cit., iii, s.v. Pathan.
3L8

Robertson, Kafirs of the Hindu-Kush Biddulph, Tribes of the Hindoo


X9

Koosh, pp. i8 sqq.; Guha, Racial Ethnology of India, p. 136.


PART TWO
CASTE

CHAPTER V

ITS STRUCTURE
Let be understood at the outset that it is not intended to give
it

here an exhaustive account of individual castes, their cere-


monies, and their machinery for regulating their relations with
other castes, nor of their own internal conduct, nor of the
historical developments of caste as reflected in Brahmanical
scriptures, but rather to examine the nature and origins of the
caste system, its function in Indian society, and the broader
relations of caste to race and religion in India.
A great deal has been written about caste and more or less
from this aspect, but most of those who have so written have
approached the question from a single viewpoint, theoretical
or social, and have tended to regard caste as an institution
which, whatever has been contributed from other sources, has
in the main a single and a simple origin. And the origin ascribed
to the institution has varied with the writer. To the present
writer, however, caste appears to be an institution of highly
complex origin, an origin so complex indeed that in its very
nature it must be limited to a single area; and that, no doubt,
is why it is only found in India. For although social institutions
that resemble caste in one respect or another are not difiicult
to find elsewhere, and some of them undoubtedly have some
association with caste in their ultimate origin, yet caste in its

fullest sense, caste, that we know it in India, is an exclu-


is, as
sively Indian phenomenon. No comparable institution to be
seen elsewhere has anything like the complexity, elaboration and
rigidity of caste in India. Indeed, on reflection, it is apparent
that caste as known in India might be expected to prove to be
unique. If it were a simple institution, it could hardly fail
to be more widely distributed; and accordingly we do find a
wider distribution for certain important or striking features of
ITS STRUCTURE 47

caste. But highly complex as it is, caste could only arise within a
limited area in which all the elements contributing to it were
associated over a long period of time. It is virtually incon-
ceivable that the association of circumstances necessary to pro-
duce so complex an institution as caste is in India could ever
be found in more than one area of the earth*s surface; and it is
probably significant that the geographical limits within which
the institution is manifest are such as have offered in the past
very considerable obstacles to perennial communications or easy
contacts of any kind.
The word caste comes from the Portuguese word casta,
* *

signifying breed, race or kind; homem de boa casta is a man *

of good family. The first use of this word in the restricted


sense of what we now understand by caste seems to date from
1563 when Garcia de Orta wrote that *no one changes from
his father's trade and all those of the same caste (casta) of shoe-
makers are the same Yule and Burnell, who quote this
passage,^ follow it by another from a decree of the sacred Council
of Goa in 1567 describing the Gentoos (Hindus) as dividing
themselves 'into distinct races or castes (castas) of greater or
holding the Christians as of lower degree, and keep
less dignity,
them so superstitiously that no one of a higher caste can eat
or drink with those of a lower
To define a caste is harder than to give the derivation pf the
term, Risley defines itas ' a collection of families or groups
of families bearing a common name; claiming a common descent
from a mythical ancestor, human or divine; professing to
follow the same hereditary calling; and regarded by those who
are competent to give an opinion as forming a single homo-
geneous community'. He goes on to add that the caste name
is generally associated with a specific occupation and that a

caste is almost invariably endogamous, but is further divided,


as a rule, into a number of smaller circles each of which is
endogamous, so that a Brahman is not only restricted to marry-
ing another Brahman, but to marrying a woman of the same
subdivision of Brahmans. This definition is open to question
on the ground that descent from a mythical ancestor is claimed
rather by the gbtra, the internal exc^amous division of the
endogamous caste, than by the caste as a whole; further,
1 Yule aiad Burnell, Hobson-Jobson (1903), s.v. Caste.
48 CASTE
although the endogamous subdivision of a caste is a genuine
and common feature of the institution, it is doubtful if the
unqualified term Brahman can rightly be used as the name of
a caste at all. It is probably too vague to be of value, for
though the Brahman belongs to one of the four original varna
or " colours into which Rigvedic society was divided, a varna is
'

very far from being the same thing as a caste, the Hindi word
for which' is jati or jat. The relationship between varna and
jat is dealt with later on.
KetKar defines caste as a social group having two characteris-

tics: (1) membership is confined to those who are born of


members, and includes all persons so born; (2) the members
are forbidden by an inexorable social law to marry outside the
group. Each one of these groups has a special name by which
it is called, several of such small aggregates are grouped together

under a common name, while these large groups are but sub-
divisions of groups still larger which have independent names.
Thus we see that there are several stages of groups and that
the word caste is applied to groups at any stage. The words
caste and subcaste are not absolute but comparative in
signification. The larger group will be called a caste while the
smaller group will be called a subcaste. A group is a caste or a
subcaste in comparison with smaller or larger. When we talk of
a Maratha Brahmin and Konkan Brahmin, the first one would
be called a caste while the latter would be called a subcaste;
but in a general way both of them might be called castes. . . .

These divisions and subdivisions are introduced on different


principles. In this way two hundred million Hindus are so
much divided and subdivided that there are castes who cannot
marry outside fifteen families. ^ Here again the definition is
not entirely satisfactory as there 'are a number of castes for
instance, particularly in southern India, which have been
recruited, and continue to be recruited, from the mixed offspring
of other castes which are not members of the caste itself, as well
as JErom those who, having a similar origin, are already members
of the caste; such castes are the Ambalavasi caste of Malabar,
the Shagirdpesha of Orissa, the Chasa caste of Orissa, and the
Karan caste of Orissa. It is perhaps noticeable that the first of
the above-mentioned four castes is one of temple servants, while

^ Bistory of Caste in Indidj p. 1^,


ITS STRUCTURE 49
the latter three are all occupational, the Shagirdpesha being
domestic servants, the Chasa cultivators and the Karan writers,
corresponding to the Kayastha farther north. Generally speak-
ing, however, the definition is valid, though there are still parts
of India where caste is fluid enough to make it possible for
persons to acquire a caste into which they were not born. Again,
in eastern Bengal there is a subcaste of the Baidya caste which
intermarries on equal terms with the Kayastha and it is said with
Sunris also, the children always taking the caste of the father,
Further, while all castes and many subcastes are strictly en-
dogamous, many subcastes are not, but intermarry with other
subcastes within the same caste. N. K. Dutt refrains^ from
defining caste, but describes its features: members of a caste
cannot marry outside it; there are similar but less rigid restric-
tions on eating and drinking with a member of another caste;
there are fixed occupations for many castes; there is some
hierarchical gradation of castes, the best recognized position
being that of the Brahmans at the top; birth determines a man's
caste for life unless he be expelled for violation of its rules;
otherwise transition from one caste to another is not possible;
the whole system turns on the prestige of the Brahman.
This description of Butt's is a normally accurate description
applicable to India as a whole. If it needs qualification, it is
in the matter of the possibility of transition from one caste to
another, for besides the case in Orissa of Chasas who become
Karans, there are places on the fringes of northern India where
the ofepring of a high-caste father by a low-caste mother can
in the course of a few generations be accepted as belonging to
the higher caste, as provided for by the laws of Manu, though
the provision has been a dead letter in orthodox Hinduism for
hundreds of years. There were also Indian states in which the
secular ruler still had power to bestow caste as well as to take it
away, for a Loi of the Manipur State could be made into a Mani-
puri with the status of a Kshatriya and could wear the sacred
thread by the authority of the Maharaja. Senart's description
is more comprehensive: *Figurons-nous un groupe corporatif,

and Castes, s.v. Baidya (i, p. 47). Cf. also Bhattacharya,


3 Risley, Tribes
Hindu Castes and Sects, p. 161, and Gait, Report on the Census of Bengal,
1901, 611 (p. 379). ^ Origin and Growth of Caste in India, p. 3.
Rose, Tribes and Castes, i, pp, 41 sqq.; and cf. Forsyth, Highlands of
Central India, p. 416.

4
50 CASTE
fenni, et, en thferie du moins^, rigoureusement h6r<5ditaire,

muni d'une certaine organisation traditionnelle et ind^pendante,


dun chef, d'un conseil, se r^unissant k Toccasion en assemblies
plus ou moins pliniires; uni souvent par la cilibration de cer-
taines fites; relii par une profession commune, pratiquant des
usages communs qui portent plus spidalement sur le manage,
sur la nourriture, sur des cas divers d'impureti; armi enfin,
pour en assurer Tempire, d'une juridiction de compitence plus
ou moins itendue, mais capable, sous la sanction de certaines
pinalitis, surtout de Texclusion soit difinitive soit rivocable,
de faire sentir efficacement Tautpriti de la communauti: telle
en raccourd nous apparait la castev The truth is that while a
caste is a social unit in a quasi-organic system of society and
throughout India is consistent enough to be immediately identi-
fiable, the nature of the unit is variable enough to make a
concise definition difficult. If it be enough to define the
system, the folloi^ing formula is suggested a caste system is
one whereby a society is divided up into a number of self-

contained and completely segregated units (castes), the mutual


relations between which are ritually determined in a graded
scale would be hard to daim that to define a caste
But it

as one of a number of such units was completely satisfactory.


One of the dilEculties in defining caste is caused by a certain
fluidity which shows itself most often perhaps in fissiparous
tendendes, but sometimes of recent years in a tendency towards
the amalgamation of analogous castes with a view to the exer-
cise of social and political influence. Thus in 1931. there was
a widespread movement on the part of various cattle-keeping
castes ^Ahirs, Ahars, Goalas, Gollas, Gopas, Idaiyans, etc. to
combine under a new caste name as Yadava, though how far,

Les Castes dans VJnde (1927), p. 35. This passage may be


Senart,
somewhat freely translated as follows : We
must conceive of a group,
united, closed, and at least in theory hereditary, provided with a measure
of organization which is traditional and independent, with a headman,
and with a council; a group that meets when need be in more or less full
assembly ; that is often combined in the keeping of certain festivals that ;

is bound together by a common occupation, and shares common customs


in regard particularly to marriage, to the consumption of food and drink,
and to various cases of pollution; finally, a group which has the power to
maintain its authority oy means of a jurisdiction which, though rather *

attenuated, is capable of making the authority of the community effectively


felt by the imposition of various penalties, the most important being
permanent or temporary expulsion from die group such in epitome, as it ;

seems to us, is a caste.


ITS STRUCTURE
if at all, such a movement would involve reciprocity in the

matter of marriage is uncertain. On the other hand, the ten-


dency of many castes to split up and throw off jEresh endogamous
units has persisted from very early times. Risley draws

attention to this and ascribes to the fissiparous nature of caste


decadence and sterility in Indian art. The medieval guild, he
says, was capable of expansion and development and could give
free play to artistic inspiration, whereas a caste is an organism
of a lower type; it grows by fission; and each step in its growth
detracts from its power to advance or to preserve the art which
it professes to practise Attention has also been drawn to the
process of the splittingup of castes into various endogamous
sectionsby W. Kirkpatrick,* who regards the various gipsy castes
of the south-east Punjab and the western part of the United
Provinces as endogamous sections of a single original family.
Change of occupation is also a common cause of the splitting
up of castes into subcastes. Thus the Khatiks, a caste of
butchers, has subcastes of Bekanwala, pork butcher ( <
bacon *);
'

Rajgar, mason; Sombatta, ropemaker; Mewafarosh, fruiterer,


etc. Formerly fissions of this kind were often the result of
migration or of political or social environmental factors, but
latterly they have more often perhaps been the result of attempts
by the well-to-do elements in a despised caste to cut adrift from
their humbler caste brethren and raise themselves in the social
scale by finding a new name and a dubious origin associating
them with some higher caste.*
This process of separation is generally the sequel to the
segregation of part of the caste into a subcaste which for a time
has accepted wives from other subcastes while refusing to give
daughters to such subcastes, thus establishing first of all the posi-
tion of a superior subcaste, the claim to superiority generally
being based on a change of occupation. The final step is to
refuse to take wives from other subcastes and then to adopt a
new name and deny all connexion with the caste of origin. Con-
versely the Kaibarttas of Bengal, perhaps a tribe originally, were
long regarded, and regarded themselves, as a single caste ^vided
occupationally into Jaliya Kaibarttas who practised, the calling

People of India, p. :*7o. And see Wise, Notes on the Races, Castes
and Trades of Eastern Bengal, pp. 194 sq.
Primitive Exogamy and the Caste System*, Proc. A,S.B., viii, no. 3,
iguj. *Vide infra, pp. us sq.
58 CASTE
of fishermen and Haliya (or Chasi) Kaibarttas who lived by
agriculture. But as the latter is a respectably regarded mode
of life while fishing is a despised calling, in the course of time
the practice arose on the part of the Haliya Kaibarttas of exact-
ing high prices for their daughters when married by Jaliya
Kaibarttas while themselves refusing to marry Jaliya women.^
Eventually the Haliya Kaibarttas broke away entirely, banning
all intermarriage with the Jaliya Kaibarttas, and succeeded in

getting recognition as a separate caste under the name of


Mahishya.^^ The same caste probably affords another instance
of the tendency of castes to split up, as the Kewat caste appear
to have belonged originally to the same caste as the Kaibarttas,
and the names were synonymous in Orissa after the two were
quite distinct in Bihar and Bengal.^^
With reference to the respective status of subcastes, it should
be made clear that that status no less variable than that of
is

castes in different localities. Thus in the district of Cawnpore


in the United Provinces the caste of Dhanuk has five subcastes:
Laungbarsa, Badhik, Kathariya, Hazari and Taihal. In the
eastern part of the district each of these subcastes has become
an endogamous unit of its own. In the southern part of the
district the three latter are endogamous, but Laungbarsa and
Badhik intermarry to the extent that Badhik men take Laung-
barsa wives, but Laungbarsa men do not take Badhik wives.
In north-west Cawnpore only Hazari and Taihal are closed
endogamous units; the other three subcastes practise some form
of intermarriage with one another. In north-east Cawnpore
only the Taihal subcaste is strictly endogamous, the other four
practising again some form of hypergamous or reciprocal inter-
marriage.^ This case is typical of the extreme variability of
status found as between subcastes, and of the very variable
pattern of the caste system in general. In one instance hyper-
gamy is practised in alternate generations: a Brahman of the
Asht-bans subcaste of the Bunjahi section of Saraswat Brahmans
in the Punjab may marry a wife from the Bans-puj subcaste of
that group; but if the son by such a marriage does the same,
he loses status for he must marry an Asht-bans. Hypergamous
10 Risley, Tribes and Castes^ i, s.v. Kaibartta; Wise, op. cit,, p. sgg.
11 Gait, op. cit., Risley, People of India, pp. 117 sq.
612, p. 380 ;
12 Risley, Tribes and Castes, loe. cit.
13 Blunt, Caste System
of Northern India, pp. 49, 128.
ITS STRUCTURE 53

marriages between the Bans-puj and Asht-bans subcastes are


thus limited to alternate generations.^^
The marry wives
practice of the higher subcaste in refusing to
of the lower while allowing and even encouraging the man of
the lower to marry wives from the upper, as mentioned above
in referring to the Kaibartta caste, is contrary to what might
be expected in a system in which the principle of hypergamy
is so deeply rooted, but it is not unique. Nevertheless, the
much more usual pattern shown by the Sadgops for
is that
instance, in which the superior subcaste takes wives from the
inferior but refuses to give daughters to men of the inferior
in marriage.^ In Assam a Rabha man can marry a Kachari
girl without any obstacle, but a Kachari man can only marry
a Rabha girl after undergoing a special ceremony of purification.
The Rabhas generally speaking are probably stricter Hindus
than the Kacharis. The practices of hypergamy, that is, of seek-
ing a bridegroom in a higher caste or subcaste than that of the
bride, of isbgamy, or marrying a spouse from the same caste,
and of hypogamy, or seeking a bridegroom in a caste or sub-
caste lower than that of the bride, will have to be referred to
later when dealing with the origins of caste, but it will be
convenient here to make the workings of hypergamy clear as it
is a widespread feature of the caste system and extends even to

matrilineal castes like the Nayars, whose women may never


marry husbands of inferior status to themselves. A good illus-
tration of the working of hypergamy may be taken from the
practice of the Rarhi Brahmans of Bengal. This caste is divided
int6 subcastes known as Kulin and Srotriya, the latter being
again subdivided into Siddha, Sadhya and Kashta Srotriyas. A
man of the Kidin subcaste can take a wife from the Kulin,
Siddha Srotriya or Sadhya Srotriya subcastes; a man of the
Siddha Srotriya from his own or from the. Sadhya Srotriya sub-
caste: a Sadhya Srotriya man or a Kashta Srotriya man can
only take a wife from his own subcaste. Conversely, while a
Kashta woman can marry only another Kashta Srotriya, a Sadhya
can marry Sadhya, Siddha Srotriya, or Kulin, and a Siddha
Srotriya woman can marry a man of her own subcaste or a
Kulin, while a Kulin woman can marry a Kulin only.^

Rose, Tribes and Castes y 11 , p. i6.


Tribes and Castes^ ii, s.v. Sadgop.
15 Risley, i Ibid., i, p. 146.
54 CASTE
The Rarhi Brahman pattern of hypergamy is clearly liable to

lead to an esccess of unmarried girls in the Kulin subcaste, since


their choice of bridegrooms in that subcaste is much more limit-

ed than is the choice of brides, who may be taken from three


subcastes. This superfluity perhaps, together with the cluty
incumbent on respectable Brahmans of getting their daughters
married before puberty, has led to the practice known as
'
Kulinism by which a man of a Kulin subcaste
would often
marry a large number of brides, whom he never intended to
support, in order to remove from their parents the risk of failure
to get their daughters married. Bhattacharya states that the
practice might be considered brought to an end by a ruling of
the Calcutta High Court that Kulins were bound to maintain
their wives, but this opinion was probably unduly optimistic.
Many Brahmans would no doubt forgo any daim to have
their daughters maintained in order to get them respectably
married by the proper time, and OMalley mentions the
case of the Kulin Brahman who as recently as 1911 had married
sixty wives, but no doubt the abuses to which Ktdinism led at
one time are now very much reduced.
Hypergamy among the Rajputs is of a different pattern to
that of the Rarhi Brahman. The Rajputs as a whole are divided
into a large number of exogamous clans of which some are of
higher rank than others, while generally speaking the Rajputs
of the west are regarded as of purer blood and better standing
than those of eastern India. Thus while theoretically any
Rajput can marry any other Rajput who belongs to a dan
other than the dans of his parents (and sometimes grandparents),
generally speaking a bridegroom of superior status is sought
for a girl, while a boy can marry into a dan of lower standing
than his own; and while a boy looks eastwards for a bride of
lower rank the girl goes westwards for a husband of higher
rank, than her own. One result of hypergamy is to put a price
on bridegrooms instead of on brides, and it is significant that
while a high price for bridegrooms is paid where hypergamy
IT Kulinism is not confined to Brahmans. There are Kulin Kayasthas,
for instance, and Kulin Sadgops. The origin of hypergamy is dealt with
by Rivers in a paper with that title in The Journal of the Bihar and Orissa
Research Society fat Mxrch igai. Hindu Castes and Sects, p. 41.
Indian Caste Customs, pp. 9 sq. Cf. Wilson, Tndtan Caste, n, pp. *06
s^.; Risley, Tribes and Castes, i, pp. 440 sq.; Risley, People of India (1915),
pi 180, cites an interesting parallel to Indian hypergamy from Madagascar.
ITS STRUCTURE 55
prevails, hypogamy, on the other hand, is associated with a
bride-price. Hypergamous marriages, in which the status of the
bridegroom is higher socially than that of the bride, are known
as anuldmd, i.e. with the grain (lit. with the hair'); hypo-
gamous marriages, on the other hand, are pr&tildmS, against
the grain, that is, against what is natural or proper, since the
status of the bride is in this case higher than that of the bride-
groom.
According to Risley, among the Rajputs of Bihar the gdtra
is tending to supplant the cl^n as an exogamous unit. Like the
clan, the gdtra or gdt, by derivation a
is an cxo-
cowshed ',^

gamous unit from a single


of individuals theoretically descended
ancestor. A gotra is in fact a dan, and, indeed, a Rajput clan
is spoken of as a got, but in their existing state dan and gotra

are not by any means conterminous in Rajput society. In castes


other than Rajput the gotra is the normal exogamous subdivision
of the endogamous whole, but, as in the case of the composition
of a caste by subcastes, the composition of a caste or subcaste
by gotra is extremely variable and often anomalous. Blunt, for
instance, speaks of castes or endogamous subcastes in which
there is only one gotrti, the Kashyapa, mentioning specifically the
Ojha Lohar and the Kalwar. Whatever the case be in the
Kalwar caste, the Ojha Lohar are stated by Crooke to have
what seems to be a minimum complement of eight Brahmanical
gotras. For it will often be found that the gotras of the
Brahman castes appear under the same names in other castes
also, however humble, though many castes, of course, have many
gotras the names of which are not common to the Brahmans
and are often peculiar to themselves. Clearly the nomendature
of the gotra has often been borrowed from the Brahmanical
system and has not by any means always been applied to a
similar exogamous imit by the borrower. In the case of a caste
with a single gotra for instance, marriage must be regulated not
by a true gotra system but by degrees of prohibited kinship which
are observed in the caste marriage system. What might be called
the type-form of the gotra system is the Brahman one in which
20 Tribes and Castes, s.v. Rajput.
21 At one time, no doubt, the family cattle shared the family dwelling,
as until comparatively recently in parts of Great Britain (Feate, The Welsh
House, ch. iv).
22 Caste System, p. 44 ; Majumdar, Pseudo-Rajputs, p. 169.
23 Tribes and Castes, ni, p. 376.
56 CASTE

the seven patriarchal sages or rishis, traditionally the seven sons


of Brahma, founded seven or rather eight gotras. The list of
rishis varies in diflPerent sacred books, but their gotras axe as fol-

lows: Kashyapa, Vasishtha, Agastya, Bhrigu, Gautama and


Bharadvaja (the two sometimes bracketed as Angirasa), Atri, and
Viswamitra.^^ In addition to these there are ten gotras founded
by Kshatriyas who became Brahmans and innumerable other
gotras founded by descendants of the original seven (or eight)
gotrakara rishis. The Brahmanical gotras are to be found in
many castes other than Brahman. The usual explanation given '

-i

is that whereas among Brahmans the gotra represents actual

descent from the original rishi, in other castes the descent is I

from an individual for whom such a rishi acted as family priest.


In many cases it may be taken as certain that the name of a
Brahmanical gotra has merely been adopted by some pre-existing
exogamous unit for the sake of appearance; it is conceivable
that in a few cases the existence of identical gotras in diflPerent
castes may go back to a time when the constitution of society
'

.was less rigid, and individuals of different varnas (mde infra,


pp. 64 sqq.) were enabled sometimes to change their varna and
status in society. Generally speaking the gotra is a strictly
exogamous group among Brahmans, though even here castes are
found in which the gotra, though existing, is not the exogamous
unit. For instance, the Sakaldwipi Brahmans of Bihar are
divided into purs or sections which do not coincide with their
gotras; here the pur is the exogamous unit, and provided
marriage takes place outside the pur it may be within the gotra,
and a similar system is reported of the Saraswat Brahmans of
the Punjab.^ Among many of the lower castes and more recent
adherents to the Brahmanical religion, the gotra may actually
represent a pre-Hindu exogamous unit, or it may be a purely
arbitrary division which has little practical bearing on exogamy.
Sometimes it represents a territorial rather than a kinship unit,
and it is to be noted that in
many parts of tribal India also
village or locality tends to supersede- the clan or sept as the
exogamous unit. Thus many of the gotras of the Golapurab
caste of the Agra district are named after villages and some
after occupations,^^ while the Bhuinhar or Babhan caste of
Wilson, Indian Caste, n, pp. 14 sqq. 25 Blunt, Caste System,
26 Bhattacharya, p. 43.
Hindu Castes and Sects, pp. 48 and 56.
27 Crooke, Tribes
and Castes, u, p.
ITS STRUCTURE 57

Bihar are divided into territorial or functional units as well as


into Brahmanical gotra, and where exogamy based on the terri-
torial unit (mul) conflicts with that based on gotra, must
as it
frequently do, since everyone belongs to both sets, the mul pre-
vails over the gotraJ^^ The Khangars of the Central Provinces
have four subdivisions, Rai, Mirdha, Karbal, and Dahat; the
firstthree intermarry hypergamously, Rai taking wives from
Mirdha and Karbal but not giving wives to them, while the
fourth is endogamous, being looked down upon for keeping
pigs. But, apparently of these hypergamous
independently
divisions, the tribe as a whole is divided into some thirty or
more gotras which are really exogamous totemistic clans.^
The Dumal, an agricultural caste of Orissa, has three sections
(harga) based apparendy on occupation, three sections (mitti)
based on territorial origin, and three sections {got) based on
totemic dan. The rule of exogamy takes all three into account,
and it is necessary for the bride and bridegroom to have at least
one of them different but not necessarily more than one. The
Gonds, or some branches of that race, and some allied tribes
and castes base their exogamy on clans worshipping different
numbers of household gods, a system possibly analogous to the
pravara^^ system in origin, and the Nattukottai Chetti trading
caste of Madura names its nine apparendy exogamous divisions
after the nine temples in which they respectively worship. Three
of these kdvih have exogamous subdivisions, and all advocate
marriage with the fathers sisters daughter.^ The Satmulia
Goalas of Bihar are organized in territorial mul, and a man is
debarred from marrying not only into his owii mul but into
that of his mother and from five to seven of his grandmothers
and great-grandmothers for four generations back, in addition
to which some take account of the muls of the female ancestors
of the bride as well as of the bridegroom.^ Quite dearly the
internal exogamy of Hindu castes is infinitely variable and every
sort of system seems to be represented. Indeed, strictly speaking

gotra itself is not inherited but is acquired by the twice-bom

28 Blunt, Caste System,


p. 45 ; Risley, Ethnographic Appendices, p. 176.
28 Russell, Tribes and Castes of the Central Provinces, in, pp. 44^ sq*;
Luard, Ethnographical Survey of the Central India Agency, m, pp. 9 sqq.
30 Russell, op. cit., n, Vide infra, p. 58.
p. 530.
32 Thurston, Castes and Tribes of Southern India, v, pp. 261, ^65.
33 Risley, Tribes and tastes, i, p. 385.

58 CASTM
castes only at the initiation ceremony {up&n&y&nS) at which the
sacred thread is assumed. In the case of an adopted son whose
gotra is not that of his adopting father he will be regarded as
belonging to two gotras for marriagfe purposes, unless, that is,
he is adopted in infancy before he has undergone the upanayana
ceremony. In the latter case his gotra should be that of his
adoptive parent.^
The gdtra then, though normally the exogamous group within
the endogamous caste or subcaste, is very far from being a stable
institution of consistent pattern, even among Brahman castes
themselves, and the only safe generalization that can be made is

that the eridogamous caste is subdivided into exogamous groups,


usually spoken of as gdtra or got, and theoretically derived
either from the gotrakara mAw of early vedic times or from
the gotra of some Brahman priest who has ministered to a non-
Brahman caste; these gotra, however, are often in reality exo-
gamous units of various kinds, territorial, occupational, totemistic
and so forth, masquerading as or actually overriding Brahmanic
gotra, and very often spoken of by quite a different term illam,
kul, mul, phaid, pal, pangat, bani, that, nukh, hurt, khel, and
what not. In Gujarat a kula is a subdivision of the gotra denot-
ing a kindred or group worshipping the same family god; inter-
mediate to the kula, and the gotra is a larger subdivision of the
latter known as nukha, or sept, which is the de facto exogzmous
unit.* One fact remains more or less constant: the endoga-
mous unit contains within it a number of exogamous units.
Even that generalization does not hold good for all of them,
as within the endogamous units marriage may be regulated
merely by a formal schedule of prohibited degrees of kinship.
Before coming to prohibited degrees of kinship, however, there
is a restriction on marriage within the caste which is closely

associated with the Brahmanical gotra which needs some further


explanation, and that is restriction on marriage according to
pr&v&rd. A pravara is a passage, or rather perhaps the names
contained in such a passage, in the vedic formula used in the
worship of his ancestors by a follower of any particular vedic
school; and the use of 'fyravara involves the recital, at specific

Karandikar, Hindu Exogamy, p. 76.


Thoothi, Vaisnavas of Gujarat, pp. 140 sqq.
ITS STRUCTURE 59
points in the sacrificial ritual^ of the names of certain quasi-
deified ancestors who followed that particular school of vedic
worship also followed by the descendant reciting the pravara,
and who are commemorated in the act of worship. These
ancestors, known as rishi or sages, are believed to have been
the original founders of the family in the case of the Brahmans
who recite their names. The gotra again is subdivided into a
number of gana (kindreds), each having its own distinctive
pravara, though all such gana will normally have at least one
pravara name, the chief name of those used, in common. In
a case where two Brahman gotra have two chief pravara names
in common, this is explained by a tradition of adoption, but it

is possible that the pravara names were not always those of


actual but sometimes of
spiritual
fathers, since a distinction
is made between arsa-gotra and is, between
laukika-gotra, that

and profane dans, for arsa-gotra means gotra as


spiritual '

determined by the names of the rishis (rsi) used in the pravara.**


If this be the case, there might well be found identical rishi
names in the pravara of different gotra. It has been suggested,
however, that pravara has been used to determine the exogamous
marriage unit by Brahmans because the term gotra has been
loosdy used with inexact meaning.*^ Be that as it may, any
identity of names in the respective pravaras of two families
operates as an effective bar, to intermarriage, whether or no the
gotta constitutes such a bar by itself. Castes with no pravara
ancestors of their own use the pravara of their hereditary purohit
or family priest who, being a Brahman, must have pravara, but
it is the Brahman castes only which attach primary importance

to pravara. It should further be pointed out here that certain


gotras are in any case forbidden to intermarry, presumably on
account of some relationship of their respective ancestors. Thus
intermarriage is barred, for instance, between the Kashyapa and

Vasishtha gotras. The ban on intermarriage varies .in some


gotras, and In the Bhrigu and
the regulations are meticulous.
Agastya gotras, for instance, the colt of more than one comm<Mi
rishi out of three in the pravara, or two out of five, is enough
to make such pravara a valid bar to intermarriage. In the other

** Brough, Early Brahtnanical System of Gotra and Pravara.


lbid.
CASTE
i

i gotras apparently a single rishi ancestor common to two pravara


will act as a bar to marriage between them.
While gotra 2ind pravara regulate exogamous marriage as
: between agnates whether the kinship be real or hypothetical,
there are also rules of consanguinity which govern marriage be-
tween afl&nes as well as agnates, and on rules of this kind,
it is

of course, that castes which have only one gotra depend, though
castes with multiple gotra observe them no less. The most fami-
liar of these rules is the one known as that of sapindH.^^ Under
the sapinda rule, which is generally observed in castes of good
status, marriage is prohibited between any two persons who
possess a common ancestor within a certain number of degrees
on the fathers side and a smaller number of degrees on the
mothers. The actualnumber of prohibited degrees varies
in different castes
and in different parts of India seven degrees
on the. fathers side or five on the mothers is sometimes given
as the rule; six on the fathers side and four on the mothers
is a frequent standard, but the degrees on the mothers side

prohibited under the sapinda lule are sometimes as low as


three. A simplified form of this rule which is commonly
observed is that no marriage shall take place between two
families while the memory of a previous intermarriage lasts.

Sometimes this is less exactingly stated in the form that a man


not take a bride from any family to which his family is
shall
known to have given one, while conversely again a daughter
may not marry into a family from which a son has taken a
bride. Various other rules are in use in different castes. The
Ahir are reported to prohibit a man from taking a bride from
any of the gotras of his four grandparents. Some castes forbid
marriage with any descendant of a fathers or mother's brother
or sister, thus barring all marriage with a first cousin of any

kind or his or her child, marriage with a brothers or sisters


child being likewise forbidden. Other castes again forbid mar-

Brough (op. cit.) says that to bar marriage a majority of rishi names
must be identical, but it is possible that he is thinking in terms of pravara
containing the names of three rishts only.
39 The original meaning of s&pindd seems to be one who is under an
obligation, shared with the person using the term, to perform funeral cere-
monies for a father, grandfather, great-grandfather, and their wives, in
common. The term is derived from the Sanskrit word pinda, meaning the
cake of food offered to the departed at the funeral ceremonies.
tTS STRUCTURE 6i

riage with all patrilineal kindred and with the kindred of the
mother's brother, though permitting marriage with cousins
derived from a fathers or mothers sister, while others again
allow marriage with all but patrilinear cousins, i.e. have no
real sapinda rule at all. Some castes forbid a daughters being
married into her mothers family, a rule which Blunt explains
on the ground of h^jpergamy, but which may be a survival of
a matrilineal tradition.^ The generally accepted simple, rule
is that no marriage shall take place between a person and a

child or descendant of his or her fathers or mothers brothers


or sisters. Many castes, however, have far more elaborate
rules, a good instance of which is to be found among the
Goalas of Bengal. Here the subcaste known as Satmulia for-
bids marriage within seven groups of relatives as shown in the
following table, while the Naomulia subcaste forbids marriage
within nine groups:

(i) Fathers fathers Fathers fathers Fathers fathers


fathers father = fathers mother (7) mothers mother (9)

(1) Fathersfathersfather=Fathersfathersmother(6) Fathers mothers


modier (8)
I

(1) Fathers fatller= Fathers mother (5)

I
(5*) Mother*sfathersfather=Mothersathersmother(4)

(11) Mothers father = Mothers mother (3)

(1) Fathers brother (1) Father = Mother (2) () Mothers brother

(1) Fathers brothers child ego (55) Mothers brothers child

Here ego may not marry into any of the he be lines (1) (7) if
a Satmulia, nor any descendant of he be a (8) or (9) either if

Naomulia.^^ The opposite extreme was to be found at any rate


until recendy in one district in northern Bengal where the
Rajbansi caste permitted marriage with a uterine half-sister.^^
While all Hindu castes forbid marriage between ortho-cousins
-^0
Brough, op. cit., mentions a verse attributed to Satatapa which forbids
any marriage into the mothers gotray apparently Jor either sex.
-*1
Vide Risley, Tribes and Castes, i, p. 587. 42 ibid., p. 494.
6s CASTE
on the male side, between the children of two brothers, and
i.e.

while the great majority at any rate forbid it between ortho-


cousins on the female side also, between the children, that is,

of two sisters, cross-cousin marriage ^marriage,


between that is,

a man and daughter


his mother's brother's or father's sister's
is widespread and common, particularly in south India, marriage

with the mother's brother's daughter beings especially favoured.


In some cases this type of marriage is virtually compulsory.
Among people following the matrilineal system -of the Malabar
coast a man often has by custom a prescriptive right to marry
his mother's brother's he wishes. This custom has
daughter if

often been interpreted as a conventional compromise to retain


matrilineal inheritance under patrilineal forms, but since it is
followed in the matrilineal castes as well as in many patrilineal
ones it may equally well, if not more likely, be in the nature
of a concession on the part of matriliny to patrilineal influences.
On the other hand, the Baiga, Agariya,Gond, Pardhan, and even
Ahir, Dhulia, and Dhimar castes of the Central Provinces, who
practise this form of marriage, speak of the inarriage of a
daughter to her mother's brother's son as giving back the milk "

(dudh lautdna), the idea being that the taking of a woman from
her family is repaid by returning another in the next genera-
tion, which may imply a quasi-economic origin for the custom
its economic advantages in the way of keeping property in the


family are obvious or possibly the results of contact between
patrilocal tribes who wished to remove their wives from .matri-
local families, who insisted on a return in kind to the matrilocal
fold. However this may be, this very un-Brahmanical custom
is followed in many respectable castes of southern India; Iyer
remarks that the practice is creeping in to some sections of
*
'

Brahmans; Thurston mentions some Deshasth Brahmans and


the Ayyar Brahmans specifically in this connexion; Steele

records that in the south Maratha country a brother's child


may marry a sister's child provided the' gotra is different;
inquiries made in the Bombay census of 1911 showed that thirty-
one castes in this area allowed a man to marry his cross-cousin,
paternal or maternal, three to marry his mother's sister's
daughter, and fifteen to marry his mother's brothers daughter,
^3 Lectureson Ethnography, p. 130.
Ethnographic Notes in Southern India, p. 54.
^5 Law and Custom of Hindoo Castes within the Dekhun Provinces,
p. 163.
ITS STRUCTURE 6$

but not his


father's sister's. Some castes go still further, for the
Kammavan, and probably some others, add a sister's daughter
to the category of suitable wives.^* Marriages of these kinds are
so strongly regarded as proper sorts of match that on account
of the relationship a boy will often be married to a woman
much older than he In some of the central Indian tribes
is.

and castes the relationship system seems to suggest that at some


time in the remote past a marriage class system analogous to
that of Australia may have prevailed, for grandparents and
grandchildren are found to have a joking relationship
*

impossible between parents and children, as if, as in the four-


class system of several Australian tribes, for instance, grand-
parents and grandchildren fell into respectively marriageable
classes. Indeed, in the Baiga tribe marriages actually occur
between grandparent and grandchild, specific instances being
given by Elwin,^* who quotes one case of a woman who had
two children by her own grandson, so perhaps the list of pro-
hibited degrees in the Book of Common Prayer does not really
start with such comic irrelevance as one is inclined to think. In
the case of grandparent-grandchild marriage it is possible that
the basis of the practice is a belief in reincarnation. If the
grandchild merely the grandparent reborn, what could be more
is

natural than for him or her to be reunited to his or her former


spouse? It is possibly significant that these grandparent-
grandchild marriages among the Baiga seem to be normally per-
formed with the informal rites used for remarriage generally.
The same tribes and castes that have a joking relationship be-
tween grandparent and grandchild generally have one between
a man arid his wife's younger sisters and between a woman and
her husband's younger brothers, which is surely a relic of the
levirate and sororate, and, indeed, in the Rengma tribe in Assam
the relationship between a woman and her husband's younger
brother is often (or was till recently) very much more intimate

than mere jocularity or even romping. This joking relationship


between brothers and sisters-in-law obtains very widely in India,
and in Bengal at any rate among castes high up in social status.
Restrictions on marriage then vary very considerably according
to caste and and many of the rules observed can hardly
locality,

46 OMalley, Indian Caste Customs,


p. 7.
47 Russell, op. cit., Ill, p. 7ij. 4 Baiga, p. 180.
6^
CASTE
be said be exactly relevant to the question of caste. Yet to
to
this extent caste does govern
marriage: the caste is almost
invariably endogamous and the subcaste
very frequently so in-

deed; the caste either contains within itself


exogamous sub-
restricting the choice of
divisions, or else observes certain rules
mates; and when those rules, whatever they may be, are broken,
the culprit is dealt with by his caste fellows and punished
according to the code of his caste. This control is an aspea of
caste to be treated later, but it may be stated here that generally

speaking, and apart from recognized hypergamous and hypo-


gamous subcaste systems, a person marrying outside the caste is
excommunicated. Very often a man or woman marrying into a
caste which is definitely recognized as unquestionably inferior
in status will be received into the caste of his or her spouse.*
Very rarely in the case of a man marrying a bride from a lower
caste, the original status may be recovered in the course of
several generations.* in accordance with an ancient
This is

practice, which had the sanction of the lawgiver Manu, and


also of the Mitakshara, but which has generally ceased to be
observed.'*

Something must here be said about varna, colour , a term
which is often confused with caste (jati, jat) though it is far
from having the same meaning. The varna seem to have been
originally the four classes into which Rigvedic society was
divided; that is, the three categories of twice-bom. Brahman,
Kshatriya and Vaishya, and fourthly the Sudra, below whom
were the outcastes. The twice-born (dvija) classes are so-called
on account of the ceremony (upanayana) at which
initiation
they are ceremonially reborn and assume the sacred thread, a
ceremony not permitted to Sudras. It is probable that in vedic
times the lines between these classes were not impassable.
According to tradition Kshatriyas could and did become
OMalley (Caste Customs, 33) cites the practice of the
Bauri caste
as a common instance of this, but I have known it happen in other castes.
so Rose, Tribes and Castes, i, p. 4a ; Laws of Manu (tr. Buhler), x, 64.
51 The Code of Manudated by Macdonell between aoo b.c. and A.p. soo,
is
but is founded on material of an earlier date. The Mitakshara is a juristic
work of the eleventh century a.d. based on two Yajhavalkya-smriti of
c. 300 A.D.; its authority is accepted in most of India, but is not necessarily
followed in practice, and in this instance may be regarded as generally
speaking a dead letter. Vide Macdonell, Indias Past, pp. 58, 163, 165,
and Dowson, Classical Dictionary of Hindu Mythology, s.w. Manu-Sanhita
and Mitakshara.

ITS STRUCTURE 65

Brahmans, Viswamitra, who from a Kshatriya became a


Brahman and a pravara rishi by his austerity, affording an
example. According to the purana * Harivansha two sons
of Nabhagarishta who were Vaishyas became Brahmans. Even
Sudras seem sometimes to have become Brahmans, for the
ancestor of the Pokhar Sevaka Brahmans of Rajputana was a
Mer who was Yajurveda by a muni,^^ and the
taught the
Vyasokta Brahmans of Bengal are reputed to be the descendants
of a Sudra who was made a Brahman by the rishi Vyasa, him-
self, according to the Buddhist writer of the Vajrashuchi, the

son of a fishermans daughter.^ The barrier between these


classes is still perhaps not entirely impassable, for the important
caste of Kayastha, which ranks next to Brahman in Bengal, is
now commonly regarded as twice-born and itself claims to

be Kshatriya, though it was perhaps more often regarded as


clean Sudra a hundred years ago and its status as twice-born is
still disputed.^ They are probably really an occupational caste
in origin, and as ^ writers may have been recruited from more

than one varna.


At the time of the Rigvedic invasion the four varna must be
held to represent a fourfold division of society into classes
Brahmans who acted as priests, Kshatriyas or Rajanyas who
were rulers, nobles and fighters, Vaishyas ( vish) the people <
generally,, ordinary householders so to speak, and Sudras, the
servile classes drawn from the people of the country. The last
term does not even occur in the early parts of the collection of

the Vedas while the Brahman was the utterer of prayer and
conductor of ceremonial, but apparently had no privilege on
that account and attained his greatest dignity as the purohita or
family priest of kings and princes, of the Rajanya that is, who
were the holders of the raj or government and possessors of
kshatra, authority. The highest religious authorities seem to
have been the sages or rishis, the poetical authors of the vedic
hymns, and these rishis came from any varna and belonged in
some cases to indigenous families; both Rajanya and Vaishya
52 Wilson, Indian Caste, i, p. 454.
5^ Ibid., p. 307.
^
I,
55 Vide Gait, Report on the Census of Bengal, 1901, pp. 379 and 381, and
cf. Colebrooke,
Enumeration of Indian Classes in Asiatic k Researches
*

(1807), V, pp. 58, 66.


56 Vish is translated *
commonalty '
by Keith, vide Rapson, Cambridge
History of India, pp. 91-4. 5t Wilson, Indian Caste, 1, p. 204.
5

66 CASTE
had the privilege of conducting certain sacrifices at any rate as
well as Brahmans, whose predominance does not appear until
the later or derivative vedas. Certain colours are associated

with the four varna ^white with the Brahman, red with the
Kshatriya, yellow with the Vaishya and black with the Sudra;
varna, as has been stated,means colour. It is possible that this
colour distinction some way associated with race, as one is
is in
reminded of the ancient Egyptian convention which showed
Egyptians red, Asiatics yellow. Northerners white and negroes
black; Hocart has strenuously attacked this hypothesis and
maintains that the traditional association of the colours of the

four varna with points of the compass ^white with the north,
red with the east, yellow with the south and black with the
west ^has a ritual, not a racial significance, and refers to the
four quarters of an enclosed town allotted respectively as dwell-
ing places to the different varna, outcastes having to live, as
they Still do, outside the village fence.
However that may varna are often claimed not only
be, the
to be of the nature of castes but up to a point to be castes
Brahman and Kshatriya at any rate are terms which seem at
first sight to be still in use as labels associated with particular

castes, while the term Vaishya has come back into use of recent

years, generally with some qualifying adjective and associated


with some particular caste group. As a matter of fact none of
the four terms for varna now represents anything but groups of
castes. All Brahmans do not intermarry, but there are many
endogamous Brahman castes. The Rajput does seem to answer
more nearly to the Rajanya or Kshatriya varna, but here again
there are very many castes who claim Kshatriya origin and wear
the sacred thread of the twice-bom, but do not intermarry; thus
the Khatri, a caste of the Punjab, of the United Provinces
and of Bihar claiming Kshatriya origin, do not ordinarily inter-
marry with Rajputs and indeed contain a number of endoga-
mous subcastes. The probability seems to be that in vedic times
the varna were classes rather than castes, and that post-vedic
scholars, looking for authority for the caste system in the earliest
vedas, have interpreted the nature of the varna in terms of the

Les Castes, pp. 46 sqq.


For a very full account of the Khatri caste see Crooke, Tribes and
Castes, mand Rose, op. dtj n.
ITS STRUCTURE 67

caste system as they knew it. At any rate the varna of the
present day is not a though it may be regarded as a group
caste,
of castes, and there a tendency among social reformers to
is

advocate the substitution of four varna for three thousand or


more endogamous groups of the caste system, a movement which,
if successful, would tend to fix a hard and fast line of social

cleavage between classes, and which probably underestimates the


power and misunderstands the significance of the caste system,
its origin, and the important function it still performs in the

integration of Indian society.


Like caste varna is to some extent fluid. In its earlier history
there seem to have been Brahmans by works as well as Brahmans
by birth; the whole of the Kshatriya varna is claimed by
Brahmans to have been extirpated by Parasurama,* but if so
it has been replaced by manufactured ELshatriyas, and in any

case Kshatriya rank is claimed by many whose title is one of


function or of creation rather than of inheritance. Numbers of
Sudra castes have taken to wearing the insignia of the twice-
born, and some of them gain acceptance, after a time, as doing so
legitimately.*^ In short, varna, like caste, is a somewhat indeter-
minate expression, and persons belonging to one varna in one
locality may be otherwise classified in another. It is wider and

vaguer than caste, like which it is a variable social unit.


One other aspect of caste must be briefly referred to here.
That is, the curious tradition only found in southern India by
which certain castes are regarded as castes of the Right Hand
and others as castes of the Left Hand. Between these two
divisions there is a very strong sentiment of factious rivalry,
leading to frequent clashes, often with riot and violence, gene-
rally occasioned by some real or supposed encroachment by
castes of the Left Hand on privileges claimed as belonging
exclusively to the Right. All castes are not concerned in this
dual division and its rivalries and animosities. Brahmans, for
instance, are spoken of as Mahajanam ( of great birth ') and
are often regarded as belonging to neither division.^ There
are generally said to be eighteen castes of the Right Hand and
nine of the Left, but when it comes to stating the names of these
See Dowson, op. cit., pp. 30 sq.
For example see Majumdar, Pseudo-Rajputs, pp. 156 sqq.
Schoebel, VHistoire des Origines et du Divelhppement des Castes de
ITnde, p. 94.
68 CASTE
nine or eighteen much difi&culty is experienced, and accounts
are found to vary. Buchanan gives a list, but the terms

are Kanarese, do not always tally with the Tamil, and are in
some cases apparently purely occupational. A list is given by
Aiyangar,^ but that of the castes of the Right Hand is clearly
incomplete. Oppert gives a list which shows at least fifty-eight
castes of the Right and at least five of the Left, but it is dfficult
to identify all the castes in his lists, which contain other obscur-
ities also. The principal castes of the Left Hand are the fol-
lowing: Beri Chetti, Vaniyan (who yoke two bullocks to their
oil-press), Devanga (weavers), Golla (cowherds), Panchala (five

artisan castes carpenters, masons, goldsmiths, coppersmiths, and


blacksmiths), Chakkiliyan (leather-workers), Bedar, Palli, and
Pallan (cultivator and soldier castes). In the Chakkiliyan caste
the men belong to the Left Hand, but their women-folk to the

Right Hand. One authority ^Dr Macleane, quoted by Oppert^
as stating it in the Administration Manual, vol. I, p. 69' ^has
stated that the male Palli belonged to the Right and the female
to the Left Hand. The same statement is recorded by Thurston
as having been made by the Census Superintendent of 1871,
but he adds that this has not been confirmed by recent investi-
gations. Indeed both earlier and later authorities return the
Palli as belonging unequivocally to the Left Hand. Buchanan
so mentions them in 1800, and the famous Chingleput judge-
ment of George Coleman in 1809 does the same. Fully Right-
Hand castes include at any rate the following: Balija, Banajiga
and Komati (trading castes), Vaniyan (who yoke one bullock
only), Chaliyan and Seniyan (weaving castes), Janappan (hemp
dressers), Kusavan (potters), Melakkaran (musicians), Shanan
(toddy drawers), Ambattan (barbers), Vannan (washermen),
Idaiyan (shepherds) Vellala, Paraiyan, Mala, Holeya (cultivating
castes), Kurava, Odde, Irula, Vedan and Vettuvan (mostly

Francis Buchanan, Journey through Mysore, Canara and Malabar (1870),


PP* 53 sq. [ch. II]. M. Srinivasa Aiyangar, Tamil Studies, p. 95.
Gustav Oppert, On the Original Inhabitants of Bharatavarsa or India,
5

pp. 51 sqq*
Other references to this division into Right and Left will be found in
Thurston, Castes and Tribes, e.g. i, p. 35 ; ii, pp. 4, 11, ii ; m, pp. 40, 117
sq.; IV, pp. 295, 330, 335} ; v, pp, 474 sq.; vi, pp. 15, 91 ; in Sonnerat,
Voyages aux Indes, i, pp. 54 sqq.; in Hayavadana Rao, Indian Caste System,
pp. 73 sqq.; and in Love, Vestiges of Old Madras, i, pp. 118 sqq,
67 Op. cit., p. 65 n. 68 Castes and Tribes, vi, p. 15.
6 Loc. cit. 70 See Oppert, op. cit., p. 6s, n.
59.
ITS STRUCTURE 69

labourers). In addition to the above there seem to be, as


already stated, the women of the Chakkiliyan or Madiga caste,
but the Right- or Left-handedness of castes differs apparently to
some extent with locality. Thus the Kaikolans,^^ a weaver caste
associated with Devanga, belong to the Left Hand, but the
devadasis (dancing girls) and nattumns (teachers of dancing
girls), drawn from the Kaikolan caste, belong to the Right
Hand. and
Conversely, there are sections both of Komatis
Chaliyans which belong same applies
to the Left, while the
to some of even the Mala in at any rate the Kurnool and
Guddapah districts. There the normal Right-Hand Mala are
reported to be spoken of as Pekinati-vandhlu or as RedduBhumi-
vandhlu; the Left-Hand Mala call themselves Murikinati-
vandhlu, a name said to be applied also to a class of Brahmans.
These divisions of the Mala caste are endogamous; the women
of the Right Hand do not wear glass bangles nor pierce the
left nostril; they remove the bodice at marriage, and they
forbid the remarriage of widows. The women of the Left Hand
wear the paita over the left shoulder except at marriage.^^
It may be mentioned in passing that one or two authorities
seem to have suffered from a tendency to transpose Right and
Left. Thus Thurston quotes the Madura District Manual as
stating that the Chakkiliyan men belong to the Right Hand and
their women to the Left, while Oppert likewise quotes the
Madura Manual but in the reverse sense.^ The latter seems
to be the correct version, but Hocart has followed Thurston's,
and sees in this division between the sexes an indication of a
former dual system of exogamous moieties. Oppert refers to
an account which associates the Right and Left Hands with
the Saivite and Vaishnavite sects respectively,^ and himself sees
in it a reflection of objections to Brahman authority,'*
Later immigrants into south such as Muslims,
India,
Marwaris, and Gujaratis, are, like Brahmans, generally regarded
71 Thurston, Castes and Tribes, in, p. 40. 7a ibid,, i, p. 35.
73 Ibid., II, p. 11.
74 Personal information from the Venerable P. B. Emmet, formerly
Archdeacon of Nandyal in the Kurnool District of Madras. Murikinati
prohMy ssMurakanddu, a sect of Brahmans following the White Yajurveda
(Wilson, indfan Caste, 11, p. 52). Pekinati possihly^Veginadu ; see Wilson,
loc. cit., p. 54, and Thurston, Castes and Tribes, iv, po. 345 sq.
7fi Castes and Tribes, ii, 76 op. cit., p. 60.
p. 4,
77 Xe3
Castes, p, is 3. 76 Op. cit., p. 58.
79 Ibid., p. 61.
70 CASTE
as outside this dual system entirely. The subjects of contention
are generally purely social or ceremonial matters, such as the
right to use twelve pillars in a marriage booth, to wear certain

gold ornaments on both arms, to carry a flag with the emblem


of a monkey on it, to ride on a horse or to use a palanquin in
wedding processions, or to indulge in certain specific sorts of
music or dance. These may seem trivial matters, but the
disturbances resulting from such disputes have often been
widespread and serious. Certain parallels, in some degree at
any rate, are to be found elsewhere in India, though without the
dual division into Right and Left. Thus iii the Laccadive
Islands, where the inhabitarits are Muslims, certain dans are
not allowed by tradition the use of certain ornaments and
symbols, the use of shoes, or the ownership of sailing vessels;^
again, in central India the use of certain gold ornaments by
Chamars has led to riots between them and Rajput neighbours.
Srinavasa Aiyangar says^ that 'similar distinctions may still
be found among the Sakti worshippers of Bengal", though he
does not regard them as having any connexion with the Madras
divisioii. The origin of this peculiar division into Right-Hand
and Left-Hand be discussed in another chapter, but
castes will
it may be noted here that while trading and weaving castes are

found in both divisions, the artisan castes belong almost exclus-


ively to the Left and the cultivating castes mostly to the Right
Further, although this division of castes is familiar in the
Telugu-, Tamil- and Kanarese-speaking districts of south India,
it is not found in the Malayalam-speaking districts, where the

marumakkathayam system, by which descent is reckoned in


the female line, prevails against the makkathayam (male line)
system of the rest of the peninsula. This fact is possibly not
without significance in regard to the origin of the system, though
the Abb^ Carr^ in the seventeenth century ascribed it ^ to the
use, or disuse, of the left hand in working.

Vide, for instance, Molohy, Census of Madras, 1911, i, p. 16a.


Per contra the Malacheri class which suffers these disabilities has the
exclusive right to perform the functions of barber, goldsmith, etc. and
of cooking and baling aboard sailing vessels. Vide The Statesman (Calcutta)
of 6 October 1954 and R. H. Ellis, Laccadive Islands, ch. iv.
!! P* 96- See chapter x (infra), pp. 149 sq.
Travels in India and the Near East, ii, p. 595.
'

CHAPTER VI

ITS STRICTURES
The marital restrictions of caste have already been treated, as
they are vitally involved in the relationship of a caste to society
as a whole, bnt they are very far from being the only restrictions
which isolate or, rather, insulate one caste from another.
Indeed, seems possible that caste endogamy is more or less inci-
it

dental to the taboo on taking food cooked by a person of at any


rate a lower, if not of any other, caste, and in the view of the
writer this taboo is probably the keystone of the whole system.
It is not uncommon in some parts of India for a man of one
caste tokeep a concubine of a lower caste, or even a non-Hindu,
and he is not ou teas ted by his caste fellows on that ground,

though he may be, and often is, on the ground that he has eaten
food cooked or served by her or taken water from her hands.^
This suggests that the taboo on marriage is the necessary and
inevitable outcome of the taboo on food and drink, rather than
the cause of it. Roscoe, indeed, has pointed out the efiEect of
food taboos in preventing intermarriage between pastoral and
agricultural peoples in central Africa.^
Now the taboo on food and water as between caste and caste
is subject to many gradations and variations It is often stated

that the test of a *


clean caste that is to say, a caste of respect-
able and non-polluting status, lies in whether or not a Brahman
can accept drinking water at its hands. Here, of course, as in
the case of marital restrictions, there is room for much variation
between one locality and another, and any generalization that
can be made must be made subject to local variations of custom
which may now and then be very striking. Thus in northern
India there are a number of Sudra castes from which men of
higher caste can take water, whereas in southern and western
India the higher caste at any rate will as a rule only take water
from men of their own caste or a caste higher than their own.
1 Russell, op. cit., i, p. 179 O'Malley, Caste Customs, p, 64 ; and cf.
;

Blunt, Caste System, pp. 153 and i4 n.


2 Immigrants and their Influence in the Lake Region of Central Africa

in The Frazer Lectures, p. 33.


3 In Bengal a Brahman, or any man of good caste, was forbidden to drink

water from the hands of any woman who had no tattoo spot, though this
taboo was breaking down in 1883. Vide Wise,<op. cit., p. 1*3.
7
CASTE
Ganges water, however, can be taken apparently even from
untouchables, on account of its sacred character which is beyond
pollution. There is a saying also that water is purified by air,
alluding to the common practice of pouring out water from
one main vessel into the drinking vessel of another. No one in
practice drinks out of a vessel belonging to another caste, though
theoretically a man
can drink from one that has been used only
by a higher than his own. This would, of course, make it
caste
useless to the higher caste owner. In northern India a Brahman
will take water poured into his lota (drinking vessel) by men of
several Sudra castes regarded as clean, e.g. Barhai (a carpenter
caste, claiming an origin from the god Viswakarman, the Archi-
tect of the Universe), Nai (the barber caste, the services of which
are important in much Hindu ritual), Bharbunja (grain-
parchers), Halwai (confectioners), Kahar (fishermen, well-sinkers,
and growers of water-nuts). The southern Brahman is more
particular, and in any case the water distributors at railway
stations are always Brahmans, so that anyone can accept water
poured out by them. As an instance of the variation shown
from place to place the Goala (cowherd) caste will serve. In
Bihar Brahmans can take water from them, but not in Bengal,
or at any rate in parts of Bengal. In the case of castes lower
than Brahman the restrictions on taking water may be somewhat
less rigid, at any rate in northern and north-eastern India, but
are approximately the same, a caste recognized as dean by
Brahmans being similarly recognized as such by other castes.
Restrictions in regard to eating are generally speaking more
severe than those which govern drinking, but do not depend, as
in that case, on who supplies the food but rather on who cooks
it. The cooking is very important, and a strangers shadow, or
even the glance of a man of low caste, ^ falling on the cooking pot
may necessitate throwing away the contents. Members of the
same exogamous unit can, of course, share each others food.
So, too, as a rule can members of different exogamous groups
who can intermarry, for a man must be able to eat food cooked
by his own household. Blunt maintains that this commensality
is a result of intermarriage, and that until such intermarriage

had taken place the two groups could not eat each others food.
^Thurston, Omens and Superstitions, p. 109.
0 Caste System, p. 89.
ITS STRICTURES 75

The Other way round, however, is perhaps the more likely, if one
comes before the other at all, and intermarriage takes place
because there is no taboo on interdining, A Kahar employed by
a superior casteBrahman, Rajput, Kayastha, etc. ^may eat
their leavings so long as he himself is r')t married; after
marriage he may not do so. Some castes v;ill not take food from
their own daughters once these daughters are married, even to
men own caste. ^ The ordinary cooked meal has to be
of their
prepared with much ceremony and care, rice boiled in water, or
chSpatij that is, bannocks, cakes of unleavened flour or meal
mixed with water and baked upon a griddle, form the staple
food of most castes, and these must be cooked, as Blunt says,
'
with the precautions of a magic ceremony. If away from the
regular household cooking-place, each man marks off his own
cooking-place, makes his own mud oven and cooks apart from
his fellows. He may cook for others of his own caste or subcaste
also,but so particular are most castes that a number of sarcastic

proverbs attach to their scruples Thirteen Rajputs, thirteen
cooking places Three Tirhut Brahmans and thirteen cooking
'

places ', are quoted by O'Malley. A. K. Forbes mentions eastern


Brahmans who forbid brothers to use a common cooking-place
or even take fire from one another to cook with. He gives details
of the elaborate ceremonial of a Brahman meal, and he describes
the precautions required of a Nagar Brahman, who must first
bathe, dress in clean clothes, which, if cotton, must first have
been washed and dried where nothing contaminating could
touch them. He must not touch an earthen vessel that has con-
tained water, nor a piece of cotton touched by anyone in a state
of ritual impurity unless it has been decontaminated by dipping
in ghi (liquid butter), nor leather, nor bone, nor paper unless
it has Hindu characters on it. He must not touch or be touched
by a donkey, enough to eat solid food.
pig, dog, or a child old
(Such a child being uninitiated would not be bound by
caste rules as to food.) Before he begins to eat he must not
touch a Brahman who is eating or has eaten. He must not
read a printed book while eating, nor a manuscript book
unless bound with silk and pasted with pounded tamarind

Risley, Tribes and Castes, 1, p. 374. ^ Russell, op. cit., i, p. 179.


8 He must himself use metal utensils only, of course. Cf. Stevenson,
Rites of the Twice-Born, p. 540.
76 CASTE
some cultivators, such as Kisan and Koiri, fall into the first

category on all counts. The Kalwar caste, distillers and liquor-


sellers, on the other hand, fall into number (i) for kachcha food
and (iii) for pakka; the Halwai themselves fall likewise into (i)
for kachcha, but into (iv)for pakka food; the Kayastha (writer)
caste fall into (ii) for both purposes; the Kahar fall into (iv) for
kachcha and into pakka food; and the Agariya, more
(iii) for
or less tribal iron-smelters, fall into (iii) for kachcha food, but
.
into group (i) for pakka food. These instances show clearly
enough that each caste almost has a law to itself. Further, as in
the case of intermarriage, the taboos of any given caste or sub-
caste are liable to vary from one locality to another, and no
general rule can be laid down, nor any scale of correspondence
between social position and strictness of observance. A Kurmi,
for instance, in west Bengal, will not take food cooked by any
Brahman except his own guru, and his wife will not take the
food her husbands guru cooks. A Kurmi again will take water
from a Santal and smoke from the same huqqa, but cannot take
the food he cooks, though the Santal can take the Kurmis.^^ As
regards southern India someone has pointed out that while a
woman can eat food cooked by her daughter-in-law, whose
mother can also eat food cooked by her own daughter, the mother
and the mocher-in-law cannot eat of each others cooking.^
Besides the restrictions on drinking and eating, similar restric-
tions are observed on smoking.
Where a common pipe is used,
it is passed from hand to hand
Earthenware bow!
(and thus from mouth to
mouth) in turn, a practice at
least as intimate as drinking
or from the same
eating
H Bamboo tube
dish. Smoking, in fact, is nor- ^16 for-
mally spoken of as drinking ^hahng^ Coconut
water
tobacco *.
Sometimes, it is true, container
a differenceis made between

smoking in which the mouth-


piece is put to the mouth
and smoking through the
hand or hands folded so to

md Castes, 1, 56.
Molony, A Book of South India (i96), p. 109.
ITS STRICTURES 11

make a funnel, the thumb end of which is put to the sucker's


mouth while the hand only comes into contact with the
pipe ^^
or cigarette. Generally speaking, however, smoking
comes into the same category as taking water or kachcha food,
and the usual expression for suspending a man's caste privilege
is huqqa pani hand karna, to deny tobacco-pipe and water',

which prevents a man from associating with his caste fellows.


Before leaving the subject of food it should be men-
tioned that there are restrictions on the material of which

eating and drinking vessels afe made earthenware, for instance,
is tabooed by all higher castes, the reason usually given being


that it cannot be made really dean as well as on the use of
certain animals for food. The only castes that will eat beef
are untouchables like the Chamar (leather-worker) or some of
the scavenging castes like the Dom. The Dom is reputed to
eat tiger's flesh, while the monkey is probably eaten only by
primitive tribes, most of whom would taboo tiger flesh. Some
castes are at any rate in theory purely vegetarian, but this seems
to depend largely upon sect, Vaishnavas being vegetarians,
whereas Shaivas and Saktas eat meat. Even where meat is eaten,
most respectable castes eschew the domestic fowl and even more
the domestic pig. Mutton, goat and game, whether ground or
winged, is generally eaten freely, the superior Rajput eating
the wild pig as many other castes do. As regards fish, custom
varies greatly, more perhaps by locality than by social position.
Thus most respectable castes eat fish in Bengal, whereas in the
dry and sandy deserts of Raj pu tana the idea of eating fish causes
disgust, and traders coming from there to live in Assam refuse
to allow their lorries carrying goods to transport the disgusting
creatures. A
Marwari baniya (trader) has been heard to remark
that to carry fish for food was as bad as carrying snakes. Some
castes distinguish between fish with scales and those without,
and some, the Kewat, for instance, who will not eat fowl or pork,
will eat crocodile and tortoise. Certain vegetables are also
tabooed in some cases: Agarwalas will not eat turnips or
carrots; they and some others bar the onion, and according to
Blunt some subcastes bar turmeric and are called Haldiya
The pipe * is often just a coconut water-containex having an earthen-
*

ware tobacco bowl, connected with it by a vertical bamboo tube, and a hole
in the side of the upper part through which the smoke is inhaled.
Caste System, p. 96.
78 CASTE
(= Turmeric ones) accordingly, a fact which is rather suggestive
of totemism.
Since pollution may be incurred by contact through food or
drink it is not surprising that it should be carried by mere bodily
contact. Thus contact with a woman during her monthly
period, a woman within the tabooed period after childbirth, a
man who had lit a funeral pyre and is therefore tainted by death-
pollution till purified, or persons in a similar state of ceremonial
impurity or taboo, cause pollution and make it incumbent on
a Hindu of caste to bathe and wash his clothes before eating or
before undertaking any act requiring ceremonial purity. Similar
purification is strictly speaking necessary as a result of contact
with certain low castes whose traditional occupation, whether
actually followed or not, or whose mode of life places them
outside the pale of Hindu society. Such castes are those com-
monly spoken of as outcastes or untouchables. Thus Chamars
(they work in cowhide). Dhobis (they wash dirty, particularly
menstruously defiled, clothes), Dorns (they remove corpses),
sweeper castes, and many others who are impure because they
eat beef or the flesh of the domestic pig, all pollute a Brahman
by contact. Castes lower than a Brahman are generally speaking
less easily defiled, but the principle is the same, and contact with
castes or outcastes of this category used to entail early steps to
remove the pollution. Thus if a Cheruman, or Pulayan, be
touched by a Paraiyan, he is defiled and must wash his head
*

and pray.^ According to Barbosa a Nayar woman touched by


a Pulayan is outcaste for life and thinks only of leaving her home
for fear of polluting her family.
As a result df the increase of travelling in public vehicles on
the railway or bus routes, pollution of this kind has become so
common and its frequent removal so inconvenient that it is no
longer treated very seriously by the majority of high<aste
Hindus. Indeed, some of the lowest castes in Hindu society are
more particular than many high ones; a Kuricchan of Malabar
plasters his house with cowdung if it is polluted by the
entry of
a Brahman. This is an extreme case, but Blunt has worked
out " for the United Provinces a table of which castes regard
which as untouchable, which is analogous to the lists already
Buchanan, op. cit. (1870), n, p. 151 [ch. xiiil.
20 Aiyangar, op. cit., 21 Caste System,
p. gi. p. 102.
ITS STRICTURES 79

referred to of the varying incidence of taboo on kachcha and


on pakka food. Li southern India this principle of untoucha-
bility has been carried a good deal further in the observance
of what is known as
distance pollution The following of a
degrading occupation, or membership of a caste traditionally
associated with a degrading occupation, is a cause of pollution
by contact, so that Mahabrahmans (i.e. Great Brahmans '),

though Brahmans, are not accorded the respect due to Brahmans,


but are regarded as polluting, because they ofl&ciate at the cre-
mation of corpses, no less than the Dom who removes the dead
bodies. But while in the north of India prejudice or pride may
prevent a Rajput who does not lay hand to a plough from
dining with a Rajput who does, mere proximity would not
pollute him. In southern India, on the other hand, many castes
are regarded as polluting by proximity, and are not allowed to
approach within certain distances of Hindu temples; so much
is this the case that common expressions of spatial measurement
are, or were, Tiyapad, Cherumapad, etc., indicating a distance
equivalent to that within which a Tiyan or a Cheruman, as the
case may be of course, may not approach a man of high caste.
At the same time this spatial measurement must be anything
but exact, as not only is the polluting distance less for a Nayar,
for instance, than for a Brahman, but difEererit standards are
mentioned by different authorities. Jonathan Duncan, in his
Remarks on the Coast of Malabar says that a Nayar may
approach a Nambudri Brahman, but must not touch him; a
Tiyan (toddy-drawer) must remain 36 paces off; a Malayan (i.e.
Panan, exorcist basket-maker) must remain three or four paces *

farther; a Pulayan (cultivator and untouchable) must keep 96


paces from a Brahman. A Tiyan must not come within 12 paces
of a Nayar; a Malayan (Panan) must keep 3 or 4 paces farther
oflE, and a Pulayan must still keep his 96 from a Nayar as well

as a Brahman. A Panan may approach but not touch a Tiyan,


but a Pulayan must not even approach a Panan. If he wishes
to address a man of higher caste, he must stand afar oE and cry '

aloud \ If a Pulayan touch a Brahman, the Brahman must at


once bathe, read much of the divine books', and change his
Brahmanieal thread. A Nayar, on the other hand, or any other
caste polluted by a Pulayan's touch, need only bathe to purify

22 Asiatick Researches, v (1807), p. 5.


8o CASTE
himself. In any case, of course, a Nayar, being a Sudra, has no
sacred thread. Francis Day,^ writing in 1863, says that an
Ilavan must keep 36 paces from a Brahman and 12 from a Nayar,
while a Kaniyan (astrologer caste) pollutes a Nambudri Brahman
at 24 and a Nayar by touch. Mateer,^^ in 1861, gives 36
ft.

paces as the distance within which a Shanan must not approach


a Brahman, and 96 as the distance for a Pulayan; from a Nayar
a Shanan must keep a distance of 12 paces and a Pulayan 66
paces. Wilson,^^ writing some 70 years after Duncan, says that
a Nayar must not come within 3 ft. of a Nambudri Brahman,
an Ilavan or Shanan (equivalent to, or identical with, Tiyan)
within 24 paces, and a Pulayan or other untouchable within 36
paces, while C. A. Innes, in the Malabar Gazetteer ot 1908, says
that artisans must keep about 24 ft. from a Brahman, while a
Nayadi (a member of a more or less nomad, outcaste and quasi-
aboriginal tribe) must keep 74 ft. away. There may naturally,
of course, have been some changes during the hundred years
covered by these reports. Aiyappan,^ in 1937, gives a scale of
distance pollution for several castes: a Nayar must keep 7 ft.
from a Nambudri Brahman, an Iravan (Ilavan, Izhuvan, Tiyan)
must keep 32, a Cheruman 64 and a Nayadi from 74 to 124.
The respective distances between these lower castes are calculated
by a simple process of subtraction the Iravan must keep 25 ft,
:

from the Nayar and the Cheruman 32 ft. from the Iravan. The
Nayadi when travelling has to avoid not only people of other
castes, but dwellings, tanks, temples, and even certain streams
when people are bathing in them. If a Nayadi touches the
water in which men of higher castes are bathing, the water loses
its purificatory qualities so long as the Nayadi is in contact with
it within the sight of the bathers. At Vilayur there is a tintal
para or pollution rock, which marks the limit within which
Nayadis may not approach the village. It is three furlongs from
there to the nearest Hindu house. An Ernadan must not come
within 400 yards of a village or 100 yards of a man of high
caste.-"^ If a man is polluted by a Nayadi, he must bathe in
23 Land of the Permauls, pp. 322, 323.
C/iarity, pp. s, 46. 25 Op. cit., 11,
::
2 p. 74,
Anthropology of the Nayadis, pp. i8 sqq.
27
The Emadan are a small jungle tribe of Malabar which is reported t6
have a r^arfable custom by which a man marries his
eldest daughter as
his second wife (Thurstop, Castes and Tribes, n, p. giy)
ITS STRICTURES 8i

seven streams and seven tanks and let blood from his little finger.
Ulladans and Paraiyans are mutually polluted by each other's
approach. Innes^ records that Nayadis had to avoid walking
over the long bridge over the Ponnani river and go miles round,
because if they walked over it they would pollute it, or any at
least who might make contact with their footprints, while the
Ande Koragas of Mangalore District had to carry round their
necks a small spittoon since they must not expectorate on the
public road for fear of polluting a passer-by who might all
unknowing tread where they had spat. Mateer quotes a say-
ing that a Vedan pollutes the road while he is upon it, but a
Pulayan pollutes the road by which he has gone. Indeed, in
some cases, as already indicated, mere sight might be enough to
cause pollution, for apart from the case he mentions of bathers,
Aiyappan says: 'Some believe that low caste people should
not be seen by them on days when they have to be specially
pure.* A correspondent of The Hindu reported (24 December
1932) that in the district of Tinnevelly there were a class of

unseeables a caste of washermen known as Purada Vannan,
who washed the clothes of untouchable castes and were therefore
doubly polluting. They had to work between midnight and
daybreak and were not allowed to come out during the daytime
because the very sight of them was polluting.^ ^ Such distance

pollution *
as that described is still observed in ritual situations,
though no longer enforced in ordinary day-to-day secular life.
Since indirect contact and even sight can be polluting, it is
clear that the use of the same wells by caste and outcaste would
give rise to trouble, though A. K. Forbes mentions wells from
which outcastes drew water on one side, and Brahmans 'when
they are gone from the other. A well defiled by the corpse
*

Gazetteer of the Malabar and Anjengo Districts^ p. los.


Native Life in Travancore, p. 337. 3o Lqc. cit.
Hutton, Report on the Census of India, 1931, vol. i, pt. i, p. 483.
OMalley (Caste Customs, p. 147) quotes James Forbes as saying that a
Maratha proclamation issued at Baroch in 1783 forbade the untouchable
castes of Halalkhor, Dhed and Chandal to come out of their houses after
9 a.m., and Ho bson-Jobson mentions the passage under the heading
Halalcore referring to Forbes, Or. Mem, iv. 53*. This reference is to the
first edition, and the passage has been omitted in the second. It prohibits
the castes referred to from coming out of their houses after 9 a.m.
upon
any consideration . lest they should taint the air or touch the superior
. .

Hindus in the streets *. It should be added that Forbes anticipated that


Baskar Raos durbar would cancel the edict on the receipt of satisfactory
bribes to do so. 32 Mdld, ii, p, ^40.
8 CASTE
of a dog, etc., had to have water drawn from it five times and

Ganges water or cow's urine poured into it. Where there is


plenty of water, as in Assam and Bengal, it is easy enough to
have different wells, and occasional difficulties can be sur-
mounted by water being drawn from the well by some caste
Hindu and poured out for the untouchable. In the drier parts
of India the difficulty is more serious, and the outcastes have
often had to put up with the most indifferent water supply.
When they have succeeded in asserting their right to use public
wells, the higher castes have usually given up these wells.
Much the same has happened in the case of schools, cremation
grounds, and inevitably restaurants, hairdressers and so
etc.,

forth catering for caste Hindus have found it necessary to debar


outcaste Hindus from using their services.
In temples there are (or have been) regular scales of distance
beyond which certain castes must remain, and just as caste-
polluting distances vary, so the distance at which temple
pollution is involved varies much in different places. No Iravan
or Tiyan must come within 325 ft. of the curtain wall of the
temple of Guruvayur in Malabar. The sides of the square
enclosed by this wall are each 350 ft. long and the temple is in
the centre of that enclosure. The public road which skirted
the temple at Vaikam was forbidden to be used by untouchables,
and the same applied to that which passed the Sachindram
temple, likewise in Travancore. The former road had to be
realigned to prevent the pollution of the temple. Similarly,
court houses were tabooed to low castes who had
keep from to
40 to 100 paces off, and at one court house near a temple
Pulayans used to have to keep 200 paces off and so cannot give *

their evidence with convenience as Mateer puts it. Mateer


also records^* how the temple of Kottayam had to be recon-
secrated because a European passed along the pathway between
the temple, and the tank attached to it, and also how in 1873
the Nagerkoil temple had to be purified because it had been
entered by the children of aBrahman who had allowed his virgin
daughter to he remarried after the death of her husband married
in -infancy. Generally speaking, Muslims and Christians are

Native Life, pp. 343 sqq. 3^ Ibid., p. aaa.


Ibid., p. 331.
ITS STRICTURES B3

regarded as inferior to Brahmans and Nayars in Malabar, but


as less polluting than the lower castes, but Thurston records
how he touched the ladle in a pot in which an Odh woman
was cooking her meal and later found that she had been out-
casted for subsequently touching* the cooking pot.' On the
other hand Syrian Christians in Cochin and Travancore seem
to have ranked as equal if not superior to Nayars. Thurston
also records that in Travancore the breath of a courtier may
pollute the king, while a low-caste man at a temple must wear
a bandage over mouth and nose lest his breath pollute the idol,
and a Kudumi woman in her menstrual period must keep 7 ft.
away from anyone, cover her mouth and nostrils with her hand,
and take care that her shadow falls on no one. So too a potter
making a household deity for the Kurubas must cover his mouth
with a bandage; and a Brahman may be polluted by a whifE
of smoke from a funeral pyre.^
This question of indirect pollution has been treated at some
length, since it shows so very clearly how intimate is the associa-
tion between caste and taboo. There are of course a number
of other taboos, similar in nature, which are more or less inti-
mately associated with caste. Menstruation of course renders
a woman polluting, but the degree of pollution varies with the
caste. The case of the Kudumi woman in Travancore has been
mentioned, and she must take care to prevent contamination
being carried by her breath, or even by her shadow. Childbirth
is no less polluting, but whereas the period of pollution for a
Brahman in southern India is ten days, it is eleven or twelve
for a Kshatriya or an Ambalavasi, fifteen for a Nayar, twenty-
eight for a Kuricchan, twenty-eight for a Cheruman in the south
but forty-two in the north of Malabar, and four months for a
Kadar. On the other hand, the social position of a caste
is not necessarily an indication of the length of its birth-

taboo, as Nayadi and Paniyan observe ten days only. In some


of the more primitive castes the labour has to take place in a
separate hut, or in a segregated part of the dwelling-house, as in

36 Aiyappan, op. cit, p, rS u.


S' Omens and Superstitions of Southern India (i9i), p. no.
38 L. K, A. Iyer, Syrian Christians,
53 sqq.
39 Omens and Superstitions, p. 56.
^0 Dubois, People of India (1879), p. 110 (pt. 11, ch. iv).
Innes, op. cit., p. 170.
S4 CASTE
the case of the Vettuvan who dig a hole in one corner and segre-
gate the expectant mother there alone with some water until the
cry of the child is heard. In the case of birth pollution all the
members of the household are involved for periods which,
generally speaking, increase as the social status of the caste
descends. an extension of this pollution to
In Malabar there is

the tarwad, that is branches of a joint family


to say, to all the
who recognize their descent in the female line from a common
ancestor, and the extension applies to death pollution as well as
to birth pollution.^^ The periods of death pollution are gene-
rally spoken of as being varied according to varna. The
Brahman must observe mourning for ten days, the Kshatriya
for twelve, the Vaishya for fifteen, and the Sudra for a full
month. This general rule, however, cannot be regarded as
covering all castes, and in southern India at any rate there is
a good deal of variation from it except perhaps for Brahmans.
Thus the pollution period is said to vary in dijferent sections of
the Nayar, while Tiyan, Mukkuvan and Cheruman observe only
fourteen or fifteen days, though some of the latter have a method
of postponing pollution to a convenient season by sealing a
ball of cowdung in a pot and suspending the taboo till the pot
isopened again, when the period to be observed after post-
ponement becomes forty days instead of fourteen.^ The per-
formance of certain funeral ceremonies also varies according to
caste, Brahmans performing them after certain intelligence of

. .death has been received but some other castes, e.g. Prabhu
.

(a western India writer caste of high standing corresponding to


Kayastha), wait for twenty-five years, and others, e.g. Sonar (gold-
smiths), for twenty years, others again for twelve or fifteen
years.^^ During periods of ceremonial pollution, in Malabar at
any rate, a characteristic service known as mdttu (=* change) is
performed for each caste by whatever caste performs for it the
office of washerman. Thus when a Nayar woman is delivered
of a child a woman of the Mannan caste, which washes otherwise
for exterior castes but not for Nayars, brings a change,
mdttu,
of dean clothing (belonging to the washerwoman) and
puts it in
the courtyard where she finds her perquisites of grain
and a
lamp. A woman of the caste serving the Nayars as barbers and
mid p. 103. 43 Ibid., p. 189.
Steele, op. cit., p. *8.
^ ^
ITS STRICTURES %
midwives (in north Malabar the Marayan caste, which also per-
forms certain functions at Nayar funerals, in the south the
Velakkattalavan) takes the dean dothes in and the Mannan
woman removes those previously worn by the woman in child-
bed. On the seventh and fifteenth days the washerwoman is
accompanied by a man of her caste who goes through a cere-
mony before the barber-woman takes the dean clothes indoors.
It is essential that this change should be furnished by the
appropriate caste in each case.^* The mattu is worn for the
purificatory immersion which marks the end of the period of
pollution, whether by menstruation or by childbirth, or as a
result of death, which causes the pollution of the entire house-
hold in which it takes place.
Attention has already been drawn in the first chapter to the
variation of marriage rules according to caste; and although the
practice of castes in the matter of the age of marriage, and the
remarriage of widows, and in particular the taking to wife of
the widow of an elder brother, or the practice of cross-cousin
marriage, may have varied considerably in the course of years,
it is, generally speaking, the caste rather than the individual
which determines what custom shall be observed, or at any rate,
approved; the behaviour of individual caste men is determined
by the decision of the caste as a whole, though the practice
approved may vary in different localities.
Another question to which much importance is attached in
some parts of India is the right of a given caste to clothe them-
selves with certain garments, wear certain ornaments, and use
certain articles of show or luxury in public. Different materials
are prescribed for the sacred thread for different t/arna: that of
the Brahman is nowadays usually of cotton, of the Kshatriya
hemp, the material used for bowstrings, and of the Vaishya
wooL^ In southern India the wearing of clothes above the waist
was formerly a privilege of the twice-born castes, while the Sudra
castes themselves until quite recently insisted that it was for-
bidden to the exterior or untouchable castes.^ The same prohi-
bition extended to the use of gold or even silver ornaments, of
umbrellas, and even of shoes. Farther north cases have occurred
of Chamars, for instance, being beaten up for dressing like

Innes, op. cit., p. 170 ; Thurston, Castes and Tribes, vn, p. 518.
For a detailed account vide Dubois, op. cit., pt. n, ch. i and Crooke,
Things Indian, s.v. Thread, sacred. ^^Matecr, Land of Charity, p. 45.
S6 CASTE
Rajputs or for wearing gold ornaments in a similar way.*^ This
on the use of ornaments seems to be of very ancient
restriction
and widespread observance, non-Hindu
as certain clans in the
Ao Naga country in Assam, for instance, are not allowed to wear
the heavy ivory armlets in use there on both arms, and similar
restrictions are in force in the Islamized Laccadive Islands on
the extreme opposite fringe of the Indian area.^ In both
cases again failure to observe the prohibition has frequently
resulted in violence. Similar prohibitions are found against
the use by low castes of horses as mounts
bridegrooms in for
marriage processions, a common cause of violence or boycott, or
of palanquins, the use of which at marriages has often led to
disturbances in Madras when used by low castes there and has
had the same result in Bengal when used by Namasudras
(Chandals). The right to use ornaments, etc., of certain kinds
may vary not only as between caste and caste, but also as be-
tween castes of the Right Hand and castes of the Left, and
Dubois mentions great commotion caused at a festival by a
v"^hakkiliyan*s wearing red flowers in his pagri which the
Paraiyans denied his right to wear.
In keeping with these restrictions as to dress and ornament are
the restrictions of language still perhaps obtaining on the Mala-
bar coast, and at one time probably a good deal more widely
observed, which compelled members of what may be conveniently
called the
exterior *
castes, asbeing outside the pale of respect-
able Hindu society, to use special language when referring to
themselves or their possessions; "when speaking of their bodily
members, such as an eye, or an ear, to a superior, they (as must
also the Chogans [=:Tiyan, Izhavan or Iluvan], and
those in-
ferior to them) prefix it by the epithet old, such as ""
old eye ,
old ear They are obliged to call their children ""
calves ,
their silver "copper, and their paddy chaff. They com-
mence speaking by saying, ""your slave has received permission
to observe . Nairs they must call Kings , and Brahmans
they
must not approach etc.^ The Pulayan *
dare not say ""
I ,
but your slave ; he dare not call his rice ""c/zorw
but karikadi dirty gruel. He asks leave not to take food,
Hutton, Census of India,
pp. 485, 486.
and R. H. Ellis, loc. cit.
"S,
Day, Lana of the Permauls, p,
ITS STRICTURES 87

but '*
His house is called madam
to drink water a hut,
and he speaks of as monkeys , or calves
his children and
when speaking he must place the hand over the mouth, lest the
breath should go forth and pollute the person whom he is
addressing/ Conversely, a Brahman may use different expres-
sions when returning the greetings of persons of different varna.
Mateer says that if a Pulayan tried to build himself a superior
type of house, infuriated Sudras' would soon pull it down.
From Day we learn that the higher castes also are not without
restrictions, obviously as a result of the possibility of breach of
taboo, in the matter of building materials: In building for
those of the higher castes, great care is necessary, as a piece of
wood, clothing or drapery will convey pollution, from the lower
to the higher castes: as will also coir matting should it contain
even one thread of cotton upon it although it is of itself un-
pollutable, a piece of new doth may be thrown to a high caste
person, who can look at it and toss it back without having been
defiled, but should it be old it causes pollution. Floors must
be made of chunam [lime plaster], stone, or earth, which are
non-conductors, and not of planks: and for the same reason
no carpets or mats can be spread. The chequered black and
white chunam floors, are therefore usually seen, in the houses
of the higher castes,^
It is hardly necessary to labour the differences between caste
and caste in matters of custom.Specific taboos are assodated
with particular occupations, like the taboo on the pan culti-
vations of the Barui (;^5n-growing) caste, which the owner may
not enter without bathing and purification and which the
Brahman may not set foot in at all; or like the impurity which
attaches to the potters oven: or the taboo which the Let sub-
caste of Bagdi place on bamboo fishtraps, which they will not
use although they fish with nets; or the Sarak caste place on
the use of the word meaning to cut .^ Such differences extend
to every kind of ceremonial activity, and many instances have
already been given of the sort of variation found. To say
52 Or, according to Logsm, Malabar, i, p. 85, ' dung heaps
53 Mateer, Land of Charity, p. 45. 5 ^ Day, op. cit.,
p. 405 (sic).
Kisley, Tribes and Castes, !, pp. 72, 73.
56 Gait, Report on the Census
of Bengal, etc., p. 421.
57 Ibid.,
pp. 428, 430. The Saraks of Orissa are Buddhist, and those of
Bengal still extremely sensitive about the taking of animal life, whence,
no doubt, the taboo on a word assodated with slaughtering.
88 CASTE
nothing more of difiEerences in etiquette and ritual, customs such
as those of inheritance, for instance, vary very greatly by caste.
In Malabar, at any rate, the matrilineal system of inheritance
still survives, according to which a mans property is inherited
by and there are traces of it elsewhere in
his sisters children,
India. Among
pinely patrilineal castes observing primogeniture
the rules of precedence between brothers vary. Some castes
hold the senior son to be the first son of the first wife married;
others the first son bom, irrespective of the wife; others again
regard as senior the son first seen by his father. Similarly in
the case of twins, while some castes count the elder of the two
to be the one first seen by his father, many more count the first
to be born the elder son, but many more still, apparently, count
the later bom to be the elder son.
One restriction on Hindus in general may perhaps be men-
tioned here, as it applies, or used to apply, with much greater
force to higher than to lower castes, and that is the
prohibition
against going overseas. The causes of the prohibition
can only
be guessed at, but it has possibly arisen from the feeling that
the act of aossing the sea and living in a strange land makes the
observance of caste rules so difSicult that they are certain to be
broken and therefore the mere act of such travel has itself
become
taboo. However that may be, it entails among strict
Hindus
the purification ceremony which involves drinking the
panchgavya\h3.t is, the five products of the cowmilk, clari-
fied butter, curds, urine, and dung all mixed together, than
which no remedy is more efficacious for purifying
the body from
defilement. Cows urine is likewise a potent
cleanser of external
defilement, and Dubois * noted having
seen it used for that
pujTJOse. Nowadays, of course, it is common enough for
many
strict Hindus to ctoss the sea,
and it is probable that the punch-
pvya is reduced to a purely
ceremonial minimum for those who
have to cTOsume it on return to India.
Restrictions on occupations have
already been mentioned in
connexion with taboo infections, and
are to be generally asso-

PP/ 37*

'* vide infra p. 108.


ITS STRICTURES 89

ciated rather with the Hindu creed in general than with the
caste system specifically. Since certain occupations are unclean,
e.g. scavenging or flaying cattle, the persons following these
occupations become untouchable, and anyone adopting them,
unless in company with his caste, must neceR^arily be ou toasted
to preserve the whole caste from pollution. Often there is a
distinction between the occupation of different subcastes of the
same caste. Thus there are two divisions of the Teli caste in

Bengal sometimes distinguished as Tili and Teli one of which
only deals in oil while the other presses it. There seems to be
no doubt that both spring from an original caste which pressed
oil seeds (til) and sold the oil. The pressing of oil seeds, how-
ever, is stigmatized as a degrading occupation in the Code of
Manu because it destroys life by crushing the seed.
This seems
to have 3 'd to the division of the caste into two, one of which
is treated as untouchable, the other not, and the Telis who

only sell oil will outcaste a member who should venture to press
it. Similarly, those Rajputs of the Kangra valley who refrain
from ploughing hold themselves distinct and superior to those
who plough,*^ while Blunt records that the bad reputation
for chicanery acquired by the patwaris (keepers of village land-
revenue records and maps) nearly led to the formation of a
separate and inferior subcaste of Kayastha by the refusal of
Kayasthas in general to have connubial and commensal relations
with patwari families of that caste. Clearly, therefore, caste may
lay occupational restrictions on its members, although mere
occupation other than that traditionally associated with caste
will not of itself be an offence. The traditional occupation of
Brahmans is the teaching and interpretation of the scriptures,
but very many Brahmans are employed as cooks, since anyone
can take food from them, but they are not therefore put out
of caste.
Sometimes occupational restrictionsimposed by caste may have
a purely economic purpose: O'Malley mentions a case of the
Kasera (brass-founder) caste expelling a man who tried to steal
a march on his fellow-castemen by working on a day which the
caste had decided to keep as a holiday; the Sonars (goldsmiths)
of a district in the Central Provinces have a feast at which the
castemen take oath that they will not reveal the amount of alloy

*Ibbetson, Panjab Castes, p. 156. Caste System, p, stsisi*


90 CASTE
decided to be mixed with gold by the Sonars on pain of being
outcasted.'^ Most professional castes also have regular
clienteles
persons for whom individuals perform regular services and
from
whom they receive fixed dues.
Poaching on the practice
of a fellow-casteman would be a proper subject for the
caste
panchayat to adjudicate upon. As regards actual change
of
occupation Blunt sums up the existing position by saying that
a change of function will no longer result in a change
of caste
unless it involves a change of status, and when such a
that
change of status does occur it will take one of three
forms
segregation into a new caste, or affiliation of the
new group to
another already existing caste, or the creation of a
new endo-
gamous subcaste within the original caste. (The last
is easily
the most frequent result.) He instances the
creation of the
Singhariya caste as a new caste formed by
Kahars (domestic
^rvants) who had taken to the growing
of water-nuts, of the
Cual Nats, wandering singers and dancers who
took to trade
and now call themselves Badi Banjaras (the Banjara
is a wander-
ing carrier and trade caste), and of the
occupational sT-.bcastes of
Kahar known as Dhimar (fishermen), Mahar
(womens servants),
and Kamkar (drawers of water). Progress,
according to Blunt,

has weakened and is slowly killing the


functional caste, at all
events on us purely occupational side.
The road lies open to
trade unionism. As far as the
purely occupational side of
caste concerned Blunts verdict may be accepted,
IS
but caste
has never been a purely occupational
institution, and in so far
as It has been occupational
has long performed many of the
it
functions of trade unionism, as Blunts
chapter on caste in rela-
tion to occupation shows clearly
enough.
Wilson sums up comprehensively the
extent to which caste
rules govern every member of any caste. Caste, he says,* gives

Its direcnons for recognition, acceptance,


consecration, and
sacramental dedication, and vice versa,
of a human being on
his appearance in the world.
It has for infancy,
pupilage,
and manhood, its ordained methods of
sucking, sipping, drink-
mg, eaung, and voiding; of washing,
rinsing, anointing, and
smearmg; of dothing, diessing, and
ornamenting; of sitting.
OMalley, Caste Customs, ap. i*k
Caste Systemj p. 336. Caste System,
Indian Caste, i, p. jg. p, 240.
JTS STRICTURES 91
rising, and reclining; of moving, visiting, and travelling; of
speaking, reading, listening, and reciting; and of meditating,
singing, working, playing, and fighting. It has its laws for social
and and occupations; for instructing,
religious rights, privileges,
training, and educating; for obligation, duty, and practice; for
divine recognition, duty, and ceremony; for errors, sins, and
transgressions; for intercommunion, avoidance, and excommu-
nication; for defilement, ablution, and purification; for fines,
chastisements, imprisonments, mutilations, banishments and
capital executions. It unfolds the ways of committing what it
calls sin, accumulating sin, and of putting away sin;
and of
acquiring merit, dispensing merit, and losing merit. It treats
of inheritance, conveyance, possession, and dispossession; and
of
bargains, gain, loss, and ruin. It deals with death, burial,
and
burning, and with commemoration, assistance, and injury after
death. It interferes, in short, vath all the relations and
events
of life, and with what precedes and follows life.
. . .
CHAPTER VII

ITS SANCTIONS
A CASTE has been described as a social unit, and it is in accord-
ance with its character as such that it is, generally speaking, the
guardian of its own rules, that it disciplines its members, expels

them from the community, or readmits them after penalties im-


posed and satisfaction exacted. It is true that religious autho-
rity, normally in the form of the Brahman, plays a part in the

proceedings, but that is only to declare if what pattern


necessary
of behaviour has scriptural approval, or what expiations may
be prescribed by religious authority or tradition. Ketkar points
out that an appeal against expulsion from caste to Brahmans is
useless. All that they can do is to specify and administer a
suitable penance; they cannot readmit the culprit to caste.^
The fact that their ministrations are necessary for all rites de
passage such as birth, marriage, etc., and that they can, if they
choose, refuse to perform the necessary ritual, naturally gives
them a position of very powerful influence, but it does not make
them the final authority in matters of caste.
Nor did the Brahman fail to make the most of his privileged
position as interpreter and arbiter of holy writ, at any rate
according to. the Code of Manu. A Brahman need observe
mourning for 10 days only, but a Kshatriya for 12, a Vaishya
for 15 and a Sudra for a month, A Brahman is initiated, and
in the process born again, in his eighth year, a Kshatriya in his
eleventh, a Vaishya in his twelfth, and a Sudra never, Sudra A
may use only the southern gate of a town for carrying forth his
dead, and his killing by a Brahman is equivalent merely to the
killing of a cat, a mongoose, a blue jay, a frog, a dog, a lizard,
an owl, or a crow. To serve a Brahman learned in the vedas
is the highest duty of a Sudra, and if he be pure and serve
humbly he may in another incarnation attain the highest class.*
The Brahman is by right the chief of this whole creation;
whatever exists in the universe is the wealth of the Brahman,
who is entitled to it by his primogeniture; in virtue
all

of which he is entitled to treasure trove and his property never

1 Ketkar, Essay on Hinduism,


3 JL^z(;5
0/ Mflnu (tr. Btlhler), IX, pp. 334, 335,
ITS SANCTIONS 93

escheats to the ki ig. He is the deity on earth by divine status


and the intelligent one by his innate comprehension. He may
without hesitation take the property of a Sudra for the purpose of
sacrifice, for a Sudra has no business with sacrifices. If a Sudra
mention the name and class of the twice-bom with contumely
an iron nail ten fingers long shall be thrust red hot into his
mouth.** he arrogantly teaches Brahmanas their duty, the
*
If
king shall cause hot oil to be poured into his mouth and into
his ears.* But no reciprocal punishments are prescribed for

cantankerous Brahmans though they follow mean occupations
Brahmans are to be honoured in every way for each of them *

is a very great deity *. Such are the injunctions of the Code of


Manu; the Padma Purana says that immoral Brahmans are to
be worshipped, but not Sudras, though subduing their passions,
for the cow that eats things not to be eaten is better than the
'

sow of good intent *. Vedic rites and prayers are required of


the twice-born castes, but they are prohibited to Sudras who
may learn only the Puranas and the Tantras. The Code of
Manu says that a king shall never execute a Brahman though

convicted of all possible crimes *, but may banishhim with all


*

his property secure and his body unhurt. No greater crime is

known on earth than slaying a Brahman; and the king, there-


must not even form in his mind an idea of killing a priest.*
fore,

A
Brahman, be he ignorant or learned, is a great divinity.*

With all this, secular power was hardly necessary.


It is probably significant of the true origins of the caste system
that the ultimate controlling authority is secular, and further
that the secular authority visualized as responsible in the ancient
Hindu authorities is the king. Vasishtha, quoted by A. M. T.
Jackson in his Note on the History of the Caste System *, says
*

(xix, 7-8): Let the King, paying attention to all the laws of
countries, castes and families, make the four varna fulfil their
particular duties. Let him punish those who stray.* Manu
prescribes the action of a conquering prince in regard to a
conquered realm as follows When he has gained the victory
:

let him duly worship the gods, and honour righteous Brahmanas,
let him grant exemptions, and let him cause promises of safety

3 Ibid., XI, p. 13. Wilson, Indian Caste, i, p. 22.


^ Ibid., VIII, p. 271. ^ Laws of Manu (tr. Biihler), ix, p. 317.
5 Ibid. ^J,A,S.B., vol. Ill, no. 7, July 1907, p. 510.
Ibid., IX, p. 319.
94 CASTE
to be proclaimed. But, having fully ascertained the wishes
of
all him place there a relation of the (van-
the (conquered), let
quished ruler on the throne) and let him impose his conditions.
Let him make authoritative the lawful (customs) of the
(inhabi-
tants), just as they are stated (to be), and let him honour the
(new king) and his chief servants with precious gifts.
Clearly
here no change in customs is envisaged as a result
of conquest,
and the new king is charged with governing according
to the
former customs.
If the secular ruler was the ultimate authority
during the early
period of Indian history, he certainly continued to
be so in the
middle ages. Ballal Sen, King of Bengal in the
twelfth century,
prescribed the order of precedence of
different castes of
Brahmans in his dominions, and raised the status of
some rqs fe s,
degradmg OMalley points out that the fourteenth-
others.
century prince Kara Singh Deva of North
Bihar settled the
respective ranks of three sections of the
Maithil subcaste of
Brahmans and made marriage rules for them; and
it should be
noted that he was not a Brahman but a
Kshattriya. Kara
Singh was later a ruler in Nepal, and there
the castes are still
under State control; law courts take cognizance
of cases involving
expulsion from caste, serious cases, that is, such as cattle-killing
or the breach of commensal rules, and
also determine the caste
of the offspring
of mixed unions. The hereditary
Prime
Minister, who was
de facto ruler, was the final court
of appeal in
such caste cases. Minor caste matters
are dealt with by caste
councils. At^the other end of the
peninsula the Maharaja of
Cochin, a Kshatriya, acted as the final
authority in caste matters
forNambudn Brahmans and had the power to raise
persons from
one caste to another, while final
expulsion from a caste required
his sanction. In Rajputana the position was
similar, for al-
cases the caste councils
disposed of caste matters,
cases that could not be settled
by them were referred to the
State
cour s or to the ruler himself, and
Ibbetson reports the creation
' Maharaja of
M^rwh^kM^
Mwar, who laid down at the same time rules
to govern hypeiwamv
be. new and old I Indore ufe ruler appototed
ITS SANCTIONS 95
a council of sastri:, that is, of persons learned in Hindu law, to
advise upon or decide on caste questions, subject to the powers
of the Maharaja to override their decisions. The Maratha rulers
exercised the same powers, both Gwalior and Baroda having
councils of sastris like Indore, though O'Malley says that the
Gaekwar no longer exercised his personal authority in castes
other than Maratha castes.^^ It is probable that the powers of
the secular ruler in caste questions have been allowed to lapse
in many states, and Ibbetson attributes this to the Mughal
conquest which he regards as having deprived the Hindus of
their natural leaders, the Rajputs, and so strengthened caste
rather than otherwise, by leaving matters in the hands of the
Brahmans and of caste councils acting under their influence. In
the marginal areas and in remoter fastnesses of hills and forests
the older order held good, and the Rajput princes in the Kangra
Hills classified Brahmans, promoted from one caste to another,
and readmitted expelled persons to caste partially at any rate
for money payments.^ The Rajas of the Simla Hill States,
of Bastar and Jashpur in the Central Provinces, of the Orissa
Feudatory States, all exercised the final power in caste matters,
including expulsion from and restoration to caste, and wielded
discipline over Brahman offenders as well as over those of lower
caste. The Maharaja of Manipur in Assam likewise had power
to admit to, to expel from and to restore to caste, and at some
time in the nineteen-thirties created a minor local sensation by
expelling from caste his own Maharani who had in some way
offended him. As in the Kangra Hills, so in Manipur, fines and
penalties arising from caste cases constituted a by no means neg-
ligible contribution to the privy purse. O'Malley describes the
wretched plight of some high-caste subjects of one of the Orissa
Feudatory States who refused to accept the decision of their ruler
in a caste case, and were themselves outcasted by him in conse-
quence. No priest,
barber or washerman could render them any
service, with the result that they had long beards matted with

dirt, their hair hung in long strands and was filthy in the extreme,

OMalley, Caste Customs, p. *70.


15
Op. cit., p. 15, and c, p. 101. Buchanan says of the Smartal Brahmans
of Kolar that they leave the punishment of transgression against caste
rules to the transgressors ' own hereditary chiefs ; at whose desire, however,
they reprimand and impose fines on obstinate offenders op. cit., i, p. 215
(ch. V, 8 July 1800). 16 Ibbetson, op. cit.,
pp. 16, 101.
96 CASTE
and their clothes were beyond description for uncleanliness
Somewhat similar cases are often to be seen in the completely
non-Hindu villages of the Naga Hills where some recalcitrant
individual has refused to conform to village opinion and has
been in consequence officially ostracized by his fellow-villagers.
Beards do not grow much in that region, but such a person's
hair, which can only be cut with someone else's help (for scissors
are not, and hair is cut by tapping it between a sharp dao blade
and a little wooden hammer), gets into an indescribable state.
Clothes are not much worn, but the effect of ostracism is seen
in his house, which decays and falls to pieces since no one will
help to rethatch it. unweeded and unharvested
His fields fall

for want of co-operative labour, and unless he be restored to


community life he soon relapses into dire poverty and ex-
cessive discomfort. As will be seen when discussing the origins
of caste, the parallel here is perhaps not entirely without
significance.
The principle that caste is ultimately a matter for the secular
or political authority is to be seen carried so far that landlords at
any rate in eastern India are apt to interfere in purely caste mat-
ters, either because their intervention is sought, or in other cases
because it increases their control over their tenants and because,
perhaps, the fines can thus be pocketed in whole or in part.^
So clearly has the principle that the secular power is the final
arbiter of caste been accepted in the past, that the Mughal rulers
of Bengal and their British successors have in turn found them-
selves in the position of judges of such matters. Nor have they
always adopted the attitude of the nonchalant Gallio. In Bengal
the Mughal governors retained the right to sanction readmission
to caste, and O'Malley quotes from S. C. Bose's The Hindus
as they Are (Calcutta, 1883, p. 167) a case in which a Nawab of
Bengal refused to restore a Brahman family that had been put
out of caste. He goes on to quote Verelst {View of the English
Government in Bengal [1772], p. 238) on the instructions drafted
in 1769 for British officers in charge of the revenue administra-
tion there: 'When any man has naturally forfeited his cast (sic),
you are to observe that he cannot be restored to it without
i'
Caste Customs, pp. 65, 66.
18 OMalley, Caste Customs, p. 7s, and ct. Blunt, Caste System, p. is?.
Caste Customs, p. 59 n.
ITS SANCTIONS 91
the sanction of Government; which was a political supremacy
reserved to themselves by the Mahomedans/ Although the
power was waived by Government, which probably misunder-
stood the historic reasons leading to its exercise by the Mughals,
a court known as the Caste Cutcherry was maintained in Calcutta
to settle caste questions, and the British Governor was nominally
head of it, though he deputed the chair to his Hindu Banyan or
commercial agent, who presided over it by virtue of the imme-

morial usage of the setdement fact of which Burke made


effective and, of course, misleading use in his indictment of
Warren Hastings.^o Since then the British Government has
declined to give decisions on caste questions except in so far as
it has been compelled to do so through judgements of the civil

courts when deciding cases of Hindu law. Nevertheless, the


Census Commissioner used to be grievously importuned at everj^
decade to decide on the claims of castes to be recognized for
something better than their neighbours would allow them to be.
So far from accepting the position of arbiter was the British
Government, that in 1921 the Chief Commissioner of Assam said
that every man's caste was to be returned as what he said it was.
In such circumstances accuracy cannot even be claimed for the
resultant figures, and the return of caste disappeared entirely
from the census schedules in 1941.
The fact that a secular authority is normally the final arbiter
of caste disputes and offences does not entirely rule out religious
jurisdiction. As might be expected in dealing with so varied a
phenomenon, exceptions occur to the general rule. The councils
of sastris already mentioned as functioning in Indore and
Gwalior, for instance, are advisory only, but the corresponding
body in Kashmir, known as the Dharma S&bha ' Religious

Convention perhaps appears to have delivered judgements and
to have deprived men of caste in that principality, except possibly
in the case of the Rajputs of whom the Maharaja himself was the
head.^^ In Mysore and in Kanara important religious leaders,
especially among Brahmans and Lingayats, seem to dispose of
questions of caste, the caste councils having little more than ' the
position of inquiring and reporting agencies under them So
too in the Assam valley, in the extreme north-east of India, the
20 OMalley, op. cit., pp. 6o sqq.. Modern India etc., p. 369.
21 Harikishan Kaul, Beport on the Census of the Punjab iqii, i, p. 4sr8.
22 OMalley, Caste Customs, p. 71.
7
98 CASTE
Gosains, who are the heads o satras or religious colleges or mon-
asteries and who are held in great respect, are referred to when
the caste councils find difficulty in dealing with a case. Some-
times the control of caste observance has fallen into the hands of
the gum who
is usually but not always a
or spiritual adviser,
Brahman. Malalodhi
Blunt reports this of the subcaste of Lodha,
and Buchanan in 1807 reported from Kerala that the gurus take

cognizance of all omissions of ceremonies and actions that are


contrary to the rules of cast (sic). Small delinquencies they
punish by pouring cowdung and water on the head of the guilty
person, by fine, and by whipping. For great offences they ex-
communicate the culprit; which is done by shaving his head. . . .

The excommunication may be removed by the guru; in which


case he purifies the repentant sinner by a copious draught of
cows urine. In some castes in northern India there is a
regular official, no doubt normally if not always a Brahman in
this case, who is dMrmadhikari ('religious preceptor),
whose duty it is to fix the punishment and who, in Garhwal at
any rate, purifies the outcaste for readmission to caste.^^
The vast majority of caste matters, however, are nowadays
disposed of by the caste councils which exercise the final powers
of expulsion and restoration to caste as well as imposing fines
and other penalties for less serious breaches of custom.
Now it is clear, of course, that with a widely spread caste the
ideal of a council for the whole caste is impossible of attainment.
The whole may, it is true, nowadays, have a sabha, Rn asso-
caste
ciation, that is, with branches all over India and even a central
headquarters. But such an organization, if it exist, must be a
recent development subsequent to the introduction of a cheap
postal system and rapid communications of various kinds, and
ithas not replaced the old system of caste control for the purpose
of the imposition of sanctions, though it may represent the caste
for purposes of social or political agitation. The caste council
can only act for a limited area, an area small enough for the
members of the council to assemble and for members of the caste
within the area to have some knowledge of each other as a gene-
ral rule. In practice the members of the caste in such an area
will usually form a nearly related group and are spoken of
23 Buchanan, op.
dt., i, pp. 101 sq. (ch. iii, 8 Tune 1800).
24 Blunt, Caste System,
p.
ITS SANCTIONS 99

collectively; in northern India, as a biradari or as bhaiband, that


is, as a brotherhood, an association of kinsmen. They may, in-
deed, actually constitute an exogamous unit within the embracing
endogamous caste, but none the less act for the caste as a whole
in enforcing sanctions on the caste members within their sphere
of action. In any case they will not consist of a group so large
as to embrace more than one endogamous unit, since an en-
dogamous subcaste will normally have rules varying from those
of other such subcastes and each must administer its own rules.
Local conditions, such as ease of communication, numbers of the
caste residing, and their concentration or dispersal, must deter-
mine the area within which the caste council functions. In con-
ditions of this kind there is obviously room for every kind of
variety in the nature, constitution and working of the caste
council, and since the variation will be no less than that already
encountered in the structure of caste and in the restrictions
enforced by it, it is both impossible to give here a complete
account of all of them, and unnecessary to do so for the purpose
of understanding caste. The general lines on which the system
works can be indicated by a few examples.
The caste council is commonly spoken of as a pdnchdydt, lite-
rally, that is, a body of five men, but in practice it is usually veiy
much larger. It may be a permanent institution with a conti-
nuous existence, or it may be an intermittent one only balled
into being when circumstances demand it and dissolved again
after its work is done. Similarly, the ofScials who perform its
executive functions, as well as sharingits judicial ones, may be

appointed ad hoc for the particular purpose in hand, or for a


much longer period, or they may be hereditary, or some may
be elected while others are hereditary. It has frequently been
observed that the lower the caste in the social scale, the stronger
its combination and the more efficient its organization. This
has been pointed out by Sir William Sleeman,^ OMalley and
Blunt,^ and probably by many others, since in northern India
at any rate the fact is patent. The panchayat oL the Baghel
Rajputs in Banda has not met for three generations.^ Indeed,
25 Ibid.
^
10^^ 1^8; OMalley, Caste Customs, p. 38.
pp. 10^
Rambles andRecollections of an Indian Official, i, p. 61 (ch. vin).
Cf, Molony, Report on the Census of Madras, 1911, p. 180.
27 Caste Customs, sq. 28 Caste System,
pp. pp. 104, 106, 1^5.
29 Ibid.,
pp. 155, 157.
100 CASTE
the high castes rarely have any organization strictly comparable to
that of the lower ones. They may well have a sahha, a loose asso-
ciationwhich may be India-wide, but a panchayat and oflScials,
permanent or impermanent, are rarely to be found. The rules
of the caste are maintained, in so far as they are maintained,
by public opinion and by the feeling of the caste members them-
selves. Informal ostracism is generally the sanction applied, and
it is some members of a caste will apply
clear that in such cases
the sanction while others wdl refrain from doing so.
Thus as
OMalley points out," Mr Gandhi after his first visit to England
was formally excommunicated by the leaders of his caste
in
Bombay and Porbandar, and was still excluded by them from
religious privileges in 1924, whereas the members
of his caste at
his home in Rajkot had duly received him back
into communion
on his return home.
Persons of influence can sometimes thus
afford to act in a manner contrary to the rules
and prevailing
sentiment of their castes, and can carry it off and often
persuade
many sympathizers to their point of view, and, indeed, behaviour
of this kind has often occasioned the formation
of subcastes in
the past, such as that of the Pir-Ali Brahmans
in Bengal to which
the Tagore family belongs.^ It is possible
that this generaliza-
tion does not apply with the same
force to Kerala as it does
to the rest of India, since the Nayar
caste organizations are very
stroifg, and more extended spatially, than those of
the lower
castes, which often refer their internal disputes to
Nayars.
On the other hand,
the caste councils of the lower
castes
in the south as well as in the north
are sometimes extremely
well organized. Among the Iluvans of
Tinnevelly, for instance in
south India, the caste in each village
has its own council and
elects two members to a council which controls
a union of vil-
ages. Each group of from five
to seven union councils elects
five
members of the divisional council which
regulates the affairs of
the caste throughout the nadu,
that is, the division or tract of
county. The Panikkan, likewise in
the Madras Presidency, have
a similar though not identical
system. A somewhat similar
system sometimes found in northern India,
is
where the Guiar of
the United Provmces, for
instance, have a small hereditary
caste
Caste Customs, p. ta,.quoting
uray and Parekhs Mahatma
Wilson, op. cit., Gandhi.
ii. p. *15- Ct Uolony, Census of Madras, iqii,
pp. 173 sqq- O Malley, Caste Customs,
p. 40.
ITS SANC TIONS 101

panchayat in each village, and a combined panchayat of several


villages under a headman, who is likewise hereditary, for the
trial of grave offences. Generally speaking, however, the caste
councils of northern India seem to operate for smaller areas
than in the south, and indirect representation as among the
Iluvan and Panikkan is not reported. Blunt gives details of
the panchayat system for twenty-nine castes, in all of which the
system differs, and even then the practices of individual castes
are apt to vary within themselves according to subcaste or loca-
lity. Each panchayat has a headman commonly known as
sarpanch
head of the five the members as a rule being known

severally as punch,though many other terms are in use for both


these The headman is sometimes hereditary, sometimes
oflSces.

elected, and the same applies to the members, who may also
include vice-chairmen and executive officers who may likewise
be hereditary or elected. Election may be for life or for each
meeting. Often two systems are combined, the headman being
hereditary and the members elected or vice versa. In the case
of the Lal-Begis, a scavenger caste in Benares, the organization
ismodelled on that of a military cantonment. The diversity of
constitution prevailing may be illustrated by the variations with-
in a single caste. Thus in the Khatik caste in Aligarh there
is an hereditary headman called chaudhuri with five punches
chosen for the occasion, though there is a tendency to choose
them on hereditary lines, and the Sonkhar subcaste in Gorakhpur
has aheadman and six or seven punches who are all hereditary.
The Poldar subcaste has a headman {chaudhuri) and a vice-
president (padhan) who are both hereditary, but the Saqba sub-
caste has a headman only who is elected for a single year at the
Dasehra festival. In Bulandshahr every village has an officer
known as muqaddam, who disposes of minor cases, and an
hereditary chaudhuri with two vice-presidents (diwan) to every
hundred or so villages.^ A
similar diversity holds good gene-
rally as between different and is common between different
castes,
subcastes of a single caste. Among the Koltas of Orissa, where
the headman is hereditary, women have been known to succeed
to the office.^ This is, of course, most unusual, but the Tawaif
of the United Provinces, who are dancing girls, have a head-
^3
Caste System, pp. 107 sqq, 34 Blunt, Caste System, pp. 109 sq.
OMalley, Caste Customs, p. 41.
102 CASTE
woman (chaudhuram) elected for life, and a panchayat of women
elected for the occasion. The chaudhurdin must have been born
within the jurisdiction of the panchayat. The headman of a
caste council often has privileges or insignia which possibly sug-
gest that the position is derived in some cases from that of the
chief of a tribe. In one section of the Vellalans of Tinnevelly
he has a seat on a dais, while everyone else stands, a.nd is
addressed by the title Irunkol meaning Please sit down

In
other castes the headman may on horseback, use an
ride
umbrella, wear a gold ring, etc., and the Bhoksa of the United
Provinces address the president of their council as takht,
* *
throne while badshah, * king and mehtar,
are prince
among the appellations used by other castes for their headmen.'
In some castes in south India *
the headman is like a chief with
practically absolute power', according to O'Malley, and he
mentions the Maravan caste as an instance.
The jurisdiction of caste panchayats is conceived as extending
to any matter in which the men of the caste consider that the
interests or reputation of the caste require action to be taken
against a member of the caste. Clearly there is vast room here
for the diversity of practice which has been seen to be charac-
teristic of the caste system as a whole. Here again, too, it is
obvious that the castes which have a permanent and continuous
system of controlling bodies are likely to exercise a much more
rigid control than those which have not, and control will
be
least among the higher and generally speaking better educated
classeswhose distribution is wide, who are less tied to a parti-
cular locality by individual family ties, and who
have learned
to appreciate the liberty which they enjoy by not
being in
tutelage to any particular caste authority. In fact a proposal
to revive caste panchayats for the socially
advanced Prabhus in
Bombay met with a critical opposition, which declared the caste
panchayats to be the greatest of all evils as being opposed

to '

individual growth and liberty. The


more territorially con-
centrated a caste is the greater will be
its consciousness of itself
as a coherent entity and the
more stringent is its control likely
to be over the individual
members of which it is composed.
^ Blunt, Caste System,
pp. 107 sq.
1). 43.
Mead and MacGregor, Cewuj
of Bombay, tgn, p. mi.
ITS SANCTIONS 103

Generally speaking, then, the offences of which a caste


panchayat will take cognizance are of the following kinds.
Offences against commensal taboos, which prevent members
tfie
of the caste from eating, drinking, or smoking with members of
another caste or at least of other castes regarded by the prohi-
biting caste as lower in social status than themselves, are
undoubtedly the most important, for the transgression of one
member of the caste if unknown and unpunished may affect the
whole caste with pollution through his commensality with the
rest. Offences involving sex or breaches of the marriage rules
of the caste really fall into two different categories: inter-
marriage contrary to the caste custom is obviously liable to affect
the commensal taboo no less than eating food cooked by a
$tranger, since for practical purposes that is what is involved;
on the other hand, marriage with a widow of the same caste
where the caste forbids the remarriage of widows is liable to
affect the good reputation of the caste or subcaste among its
Hindu neighbours, though no commensal taboo is broken.
clearly
Any offence in which sex is involved may come under one of

these two heads, and the nature and extent of such offences will
vary greatly with caste, and locality, and the conditions of life
in general. Some castes, for instance, forbid the remarriage of
any widow, while others allow remarriage of widows to widowers
only, and others again observe no restrictions. Often a subcaste
which wishes to raise itself in social status will place a ban on
the remarriage of widows, while other subcastes of the same caste
continue to allow it. Grave offences against widely accepted
precepts of Hinduism, such as insulting a Brahman or killing a
cow (i.e., generally, causing its death by neglect or ill-treatment),
of which the courts established under British rule will take no
cognizance^ or treat only as minor offences, are likewise subject
to caste discipline, though the nature of the reason is not always
easy to determine, as besides the feeling of injury to the repu-
tation of the caste the feeling of collective responsibility and
the contagious nature of taboo may also come into play. This
seems almost certainly the case where the caste panchayat takes
action, as it does in some castes, on account of the killing of a
dog or cat or ass. The same doubt attaches to the sins of being
But since the achievement of independence the government of what
were the Central Provinces has made the killing of a cow a cognizable
offence.
104 CASTE
deservedly or undeservedly beaten with a shoe or put in gaol/^
it is possible in their case that it is rather the fear of pollution
for
through contact with cowhide or through eating with someone
who has been the guest of the Inspector-General of Prisons that
is operative, than any sense of disrepute attaching to the stigma
of chastisement or of conviction in a criminal court. It may be
likewise such a fear of pollution by contagion that causes a man
to be put out of caste, as by the Sansiya, a criminal caste or tribe
of Rajputana, for happening to be touched by the petticoat of
his mother-in-law or daughter-in-law, or for being struck by
his wife's petticoat '
in the course of connubial strife Some-
thing of the same doubt attaches to offences against etiquette
or against caste custom in the, matter of feasts and entertain-
ments, while in the latter case again there may be an overlap
with offences against the economic interests of the caste such as
might be involved in a breach of caste trading conventions,
or customs respecting handicrafts exercised, or the collection of
wild produce. Where the caste panchayat takes the liberty of
retrying cases among its members already disposed by the
of
courts, no doubt it is regarded as in the interests of the caste
as a whole that a decision which it finds unsatisfactory
should
be modified in its effects by the action of the caste, even though
the actual decision of the courts of law must be endured.
Blunt
suggests that such retrials among criminal castes may
be con-
cerned really with the clumsiness which has led to the
offender's
detection.^* Caste panchayats will generally be very much more
likely to know the true facts of offences their castemen have
committed than the ordinary law courts are, and they un-
doubtedly resent matters closely touching their community being
taken to the wdinary courts, and when the
interests of a fellow-
casteman are in question the caste may combine to stifle evidence
which might lead to a conviction in a court of
law. O'Malley

Thugs and Dacoits, pp. 153 so. Similarly


in the Nasa HilU
of Assam, though no question of outcasting would arise, it
is a terrible mis
\
,

ITS SAJsfCTIOUS
gives two instances of evidence of murder being suppressed in
this way.Such instances could probably be multiplied indefi-
nitely by diving into the records of police stations throughout
India.
The procedure observed in cases tried by caste panchayats is

no doubt extremely simple, informal, and untrammelled by the



law of evidence^ ^more so probably than even a perusal of
O'Malley's account^ would suggest. At the same time Blunt
records a number of cases in which some special procedure is
laid down. The Lal-Begi panchayat already referred to have
a quite elaborate arrangement^^ for determining by indirect
consultation the general opinion of the members present. In
the Nai (barber) panchayat of Bulandshahr an inquiry is called
for in a specific formula, uttered when the offender is present,
in the words 'Look and flies (in your food) before
for hair
you eat '.* Among the Muslim Chamargaurs of Banda an out-
casted man desiring reinstatement has food prepared on a day
fixed by himself and summons his own caste brethren, the Hindu
Chamargaurs of the place, and some Brahmans; if they consider
that he can be reinstated they eat the food, but '
a single dissen*
tient amongMuhammadans, and
the three or four among the
Hindus, would be enough to make his restoration impossible
The Chandel and Kachwaha Rajputs of Cawnpore require the
person ostracized to approach a leading clansman, who calls
together the. brotherhood, appoints and holds an
assessors
inquiry; he will not act, however, on any person's motion except
that of the ostracized person, and the decision pronounced must
be either a confirmation of the ostracism suffered or a complete
acquittal.^ This latter procedure is probably typical on the
whole of cases where there is no permanent caste panchayat, and
it is necessary for some individual, who is anxious to have
one, to go to the expense of providing entertainment and
setting machinery in motion to get the members of the
caste together.
Evidence before a caste panchayat is frequendy given on oath,
and Hindu oaths by Ganges water, or the sacred tulsi (sweet
basil) plant, or by holding a cow's tail, etc., are commonly used.

See Blunt, Caste System, p. 113. *48


Ibid., loc. cit.
Caste Customs, p. 45. 49 ibid., p. 126.
47 Blunt, Caste System, Ibid., 10c. cit.
p. 114.
io6 CASTS
Blunt mentions a white stone tied round with black woollen
thread used by the Bhotiyas of Kumaon; no man, he says,
dare tell a lie with his right hand on this stone.^^ Ordeals of
a more exacting kind are also in use, and accused persons, or
the parties to a dispute, may be required to plunge an arm into
boiling cowdung and extract a coin or pebble, or to hold in the
hand a piece of red-hot iron. An
interesting account of the
administration of an kind with dramatic and satis-
ordeal of this

factory results the innocent mans arm came unscathed from
the boiling water, while the guilty mans was scalded at the

brink will be found recorded by an observer of complete reliabi-
lity in the Journal of the Royal Asiatic Society of Bengal, vol. n
(no. i), A
of 1936.^^ less drastic mentioned
form of ordeal is

by OMalley as used by a Brahman subcaste of Sambalpur in


which the accused takes oath that if he is guilty he shall become
blind or his children shall die within a specified time. The oath
is recorded on a palm leaf which is kept in the temple for the
period named and the accused is temporarily outcasted. If at
the end of the period he and his family have suffered no evil,
he appears before the panchayat with the leaf and applies for
readmission to the caste. Very close analogies with this form
of ordeal ae to be found in use among the Ao Nagas of Assam
who, from being Brahmans, are in no way Hindus, but as
far
they have no writing, their oaths are not written down and
deposited in a Hindu temple.^
The most familiar punishment imposed by caste panchayats
is that which deprives a casteman of the right to receive water,
or the tobacco pipe, jfrom the hands of his fellow-castemen and
forbids them likewise to receive it from him; huqqa pant band~
pipe water stoppedis the equivalent of the Roman inter-
dictio aquae et ignis. The prohibition on water atitomatically
and the culprit is
prohibits kachcha food, for practical purposes
excommunicated from communion with his fellow-castemen.
He cannot have the services of the Brahmans who conduct
ceremonies for his caste, nor of the barbers who shave for it
nor the washermen who wash for it, and if he die under the

Ibid., p. 114.
Norman Loftus Bor cC the Indian Forest Service in an article on
'
The Baflas and their Oaths Caste Customs, p. 46.
S'*
Mills, The do Nagas, pp. 195 sqq. Vide supra, pp. 74 sqq.
ITS SANCTIONS 107
ban his corpse must lack the funeral rites which alone ensure
a continued existence after death and subsequent reincarnation.
An excommunication of this sort may be a temporary penalty
for a stated period, for an indefinite time subject to the perfor-
mance of some required expiation for the fault punished, or
for life, in which case the culprit is virtually compelled to seek
acceptance in some other caste or to remain in communion
with other unfortunates in the same plight as his own. Other
forms of punishment made use of are the exaction of fines or
feasts to the caste or to Brahmans. The fines may be utilized
to buy sweetmeats for the assembled castemen or for putting
mto the fund which many castes maintain for communal
purposes. Other penalties inflicted may be corporal punishment,
or the performance of a pilgrimage, or the collection of a fine
by begging. Many penalties imposed are intended to humiliate
the culprit excessively
and often to fit tile crime as far as possible
somewhat in the manner advocated by the English philosopher
Jeremy Bentham at the end of the eighteenth century. Thus
corporal punishment may be given with a shoe, for the sake
of
humiliating the offender. The person responsible for the death
of a cow often has to make a pilgrimage with a
cows tail tied
to his staff or to beg for his living for a period in the same
manner, or accompanied by a cow the tail of which he holds;
or he may have to low like an ox at the same time or even
to
wear the dead animals hide, horns and hoofs, or have a rope
round his neck and straw in his mouth. Shaving the culprit
oil one side of his head and face and
leaving the other adorned
with hair is another method of humiliating, or the
victim is
paraded on a donkey with one side of his face blacked
and
the other whitened; or he may be compelled to
wear a string of
old shoes round his neck, while an unchaste woman
may have
to wdk round the village with a basket of
mud on her head
or with a grindstone hung round her neck. OMalley' men-
tions as a punishment making a man stand on one leg in the
sun with a pitcher of water balanced on his head;
also a
punishment used by the Nats of Bihar of making the culprit
sit in cold water for as long even as
twenty-four hours. This
penalty suggests an element of purification, though
the same
can hardly be said of one mentioned by Blunt,' who
speaks of
Caste Customs, r Caste System,
pp. 8o, 8i. p. 1*3,
io8 CASTE
a Thathera (brazier) of Azamgarh who was immersed for three
days up to the neck in a cesspit before going on a pilgrimage
to Puri and Benares and feasting the biradari. But bathing in
the Ganges or even swallowing some of its sand (or mud) very
often forms part of a penalty imposed, as does drinking water
into which a Brahman has dipped his toe. The element of
purification is clearly present, and
another punishment
in
described by OMaJley^ the proceedings probably involve a
more elaborate symbolism than mere washing, for a Ravulo
husband, who has ill-treated his wife, has to enter a fish-trap
shaped like a hencoop and his wife sits on it; both parties then

go to that part of the house in which corpses are washed, and


are there washed as if they had become corpses; the wife then

acts widowed, breaking a cooking-pot and her bangles,


as if

after which she returns to her fathers house and is free


to
marry again. This looks rather as though a permanent divorce
were effected by a symbolic birth of the husband from his wife
who thereby becomes his mother whom he cannot marry, a
severance which is illogically but emphatically reinforced by the
death of both parties and a widowhood as well. One may com-
pare with this symbolic birth, if so it be, the symbolism
by
which a Chamar who has abducted a married woman has to
suck her breasts as a sign that he henceforth regards her
as his
inother, and would therefore be guilty of incest if
he continued
his liaison with her, before he restores her
to her husband.*
Blunt records some two dozen varied punishments,
specific
f:ases, by the panchayats of various castes.
inflicted
Perhaps the most familiar form of expiation is
that of ingest-
ing what is known as panchgavya, the five
products of the cow, a
mixture of milk, curds, ghee, cowdung, and cows
urine. OMalley
calls It ' a sovereign purge for impurity
and mentions a modern
substitute spoken of panchamrata (five nectars) which sub-
stitutes honey and sugar for the
last two ingredients, a degene-
rate concession to civilized squeamishness.
Fees to administering
Brahmans are involved in either case.
Sometimes purification
IS very elaborate. Among the Urali of Trichinopoly, if a man
Caste Custams, p.

MM. fcmd to
ITS SANCTIONS 109

has seduced a girl o the caste they have to be married, and they
and by bathing in 108 different pools of
their relatives purified
water, by walking over the buried head of a sheep with the
blood of which they are smeared, by further bathing, by the
drinking of cow's urine, by bathing again, and finally by feast-
ing the panchayat^^ The sheep here is possibly associated with
the scapegoat idea, which, like the idea of sin-eating, appears in
some caste ceremonials for purification. Among the Uppiliyans
of Trichinopoly a man called
the man
two lights has to of '

eat a meal in the polluted house with his hands tied behind his
back, and in the Central Provinces several castes pay an agua

one who goes ahead to eat the first mouthful at a penitential
feast, by doing which he takes on the sins of others.^

Purification and expiation is followed by the readmission to


ordinary communion of the erring casteman, but life expulsion
may be the penalty for heinous offences, and it may likewise
follow errors or misfortunes for which the unhappy victim is
in no way to blame. If a Nambudri woman commit adultery she
is outcasted and a funeral ceremony is performed for her as if
she were dead,^ as, indeed, she is to her caste, but if a husband
take back an erring wife or a father receive home an erring
daughter, they, too, are liable to be outcasted. The Dombo of
Vizagapatam district outcaste a man who has been eaten by a
tiger, and although this may not affect him in his existence
just ended, presumably do so hereafter, and the practice
it will
throws some light on one of the points of view involved in the
observance of as does also the fact that in some castes
caste,
persons whounmarried are buried instead of burned.
die
Many Assamese girls forcibly abducted by Burmese soldiery in
the invasion of 1822 were turned away from their homes when
the Burmese withdrew. No less unfortunate are, or were, the
miserable wretches who were taken to the banks of the Ganges
to die in odour of sanctity and were ill-advised enough to
recover. To die half-immersed in the Ganges with Ganges mud
61
OMalley, Caste Customs, p. 77. 6^
63
Thurston, Castes and Tribes of Southern India, v, p. 523 performing ;

the funeral ceremony is the accepted form of excommunication in Kerala,


and is not restricted to Nambudris.
64 Molony, Census of Madras, 1911, i, p. 181.
65 Blunt, Census of the U.P.,
1911, p. istsi.
66 Benudhar Rajkhowa, Short Accounts of Assam,
p. 92.
no CASTE
in ones mouth Agni Purana, aeons of
secures, according to the
bliss hereafter, but if and mud be such
the virtue of the water
that the moribund escape death, then mother Ganges has
rejected her votary. He is dead and his property has passed
to his next of kin. He must live an outcaste, and his own
family may not consort with him and must know him no more.*^

According to OMalley Customs, p. 87) such persons are no


lo^er outcasted but are none the less regarded as disgraceful.
Greek practice mentioned by Plutarch (Roman
m fttons, V), but here the ostracism was remedied by a ceremony
In me Naga Hills of Assam it is, or was, an actionable
of rebirth.
offence to spread
^ an
untrue report of a persons death.
CHAPTER VIII

ITS FUNCTIONS
The previous three chapters have been concerned with the
manner in which caste operates, and the functions of caste in
regard both to the individual casteman and to the caste as a
whole will already be evident from what has been said. It will
therefore be necessary here to do litde more than recapitulate
and generalize on certain aspects of the foregoing. Something
more is perhaps necessary when it comes to considering the
society as a whole, and it will be convenient to treat of the
functions which the institution of caste performs first, very
briefly,from the individuals point of view, then from that of
the caste as a body, and finally, at somewhat greater length,
from the point of view of society and the State.
From the point of view of the individual member of a caste
the system provides him from birth with a fixed social milieu
from which neither wealth nor poverty, success nor disaster can
remove him, unless of course he so violate the standards of
behaviour laid down by his caste that it spews him forth
temporarily or permanently. He is provided in this way with
a permanent body of associations which controls almost all his
behaviour and contacts. His caste canalizes his choice in
marriage, acts as his trade union, his friendly or benefit sciciety,
his slate club and his orphanage; it takes the place for him of
health insurance, and if need be provides for his funeral. It
frequendy determines his occupation, often positively, for in
many castes the occupational tradition is very strong indeed,
commonly negatively, since there are many any rate
pursuits, at
in the case of all but the lowest castes, which he cannot follow,
or can follow only at the cost of excommunication from the
society to which he belongs. It must often happen that member-
ship of a caste will take the place of attachment to a political
party, since in such cases as disputes between castes of the Right
Hand and of the Left his views on the merits of a dispute and
the side he is to support are predetermined for him by his caste
membership.
Thus the practice of his caste dictates to each member customs
to be observed in the matter of diet, the observance of ceremonial
lU CASTE
uncleanness, and whether he may, or may not, marry or remarry
a widow. It prescribes to some extent (or at least limits his
choice of) ritual to be observed at birth, initiation, marriage
and death. It may state, for instance, whether or not his ears
shall be bored, and if so in how many places. With reprd to
the individual the function of caste is to predetermine his
pattern of behaviour in this world to a very considerable degree
of nicety, leaving much less to individual choice than is usual
in a casteless society.

From the aspect of the caste community itself caste determines


the membership of the community. It is true that in the case
of some castes, particularly those of sectarian origin, recruit-
ment from outside may be possible, but generally speaking a
caste cannot increase the membership of its body except by
means of an increase in the number of births within the caste.
Caste again determines to a very large extent the social status
of each caste in regard to other castes. That, of course, is not
to say that it is draw up a strict list of precedence
possible to
as between castes. This can be done perhaps for a small and
restricted area with some accuracy, but the variation between
different areas is too great for any list to be valid unless drawn
up in very general terms. Nevertheless, in any given area the
status of any given caste is roughly prescribed, with reference to
the others with which it is in contact, by the system as it obtains
locally.
The caste system enables the caste to act corporately and to
control the behaviour of its component members. It is able
in this way to raise its position in society. This may not be
easy to achieve and may perhaps be accomplished only in the
course of generations, but it been and can be
certainly has
done. By organization and propaganda a caste can change its
name and in the course of time get a new one accepted, and
by altering its canons of behaviour in the matter of diet and
marriage can increase the estimation in which it is held. Thus
the Chandals of Bengal, bearing a name despised if not abhorred
in Hindu tradition, have succeeded in getting themselves gene-
rally known by a name which at least is free from traditional
obloquy. Perhaps the most usual method of procedure is to
take the name of some much higher caste and qualify it by
an
ITS FUf/CTIONS 113

adjective which it is hoped no doubt will in the course of time


come to have less significance than the caste name chosen. Thus
the
Brittial of Assam are apparently Dom or Hari by
Baniyas

origin,though now sometimes spoken of as Baniyas simply,

though never confused with the high<aste trader of that name


from Rajputana. So, too, each census used to produce a decennial
crop of Viswakarma Brahmans (Panchala, artisans), Gahlot

Rajputs (Chamars), Nai Brahmans (barbers) and so forth,


'

recalling irresistibly W. S. Gilberts simian character who

christened himself Darwinian Man A number of such castes


who claimed to be some special sort of Kshatriya or Vaishya
at the 1921 census claimed to be some peculiar kind of Brahman
in 1931.^ The Haliya Kaibarttas, when they aspired to climb,
began first to segregate themselves from the Jaliya Kaibarttas,
then to get the newly discovered name of Mahishya accepted,
and by now may be regarded as a separate caste, the Kaibartta
origin of which may in time become completely forgotten. Les

sous-castes daujourdhui sont les castes du demain to quote ,

Bougld.** That important caste the Kayasthas, who stand next


to Brahmans in general estimation in Bengal, seem to have been
regarded in the eighteenth century simply as a clean Sudra caste,
a view of them which lasted in point of fact a good deal later;
but it is probable that nowadays they are more generally regarded
as a twice-born caste whose claim to Kshatriya status need not
be cavilled at. In a not dissimilar way the various cow-keeping
castes of northern India were combining in 1931 to use the
common term of Yadava for their various castes, Ahir, Goala,
Gopa, etc., and to claim a Rajput orig^in of extremely doubtful
authenticity. The status of a caste may also be raised by educa-
tion. Mr Cunningham, for many years Director of Public
Instruction in Assam, wrote, in a memorandum for the Calcutta
University Commission in 1917, as follows: It has been said
that in these parts the social order is a despotism of caste
tempered by raatriculation. by matriculating and
It is only
taking the part in the which has been reserved for
after-life
those who have matriculated that the lower castes can raise
themselves to consideration. Similarly, the Census Superinten-
dent of Assam in 1931 says:
The respectability of a community
1 Hutton, Census of India, 1951, pt. i, p. 431.
^ Le Rdgime des Castes, p. 117.

8
114 CASTE
in Assam can, in fact, be generally measured by the number of
persons belonging to that community who are in Government
service/ Caste then may in some cases serve communities (but
never individuals) as a ladder for rising in the social scale.
Itno doubt one of the functions of a social system to hand
is

on from generation .to generation that pattern of skill, knowledge


and behaviour which we speak of as a culture, the benefits and
advances achieved in man's struggle to control environment,
using that word in its widest sense, of course. The caste system

is peculiarly fitted to hand oii culture patterns and particular


items of culture, though from its isolating tendency it is pro-
bably ill-adapted to growth and change. Craft secrets are well
safeguarded and perpetuated by occupational castes, a|id the
caste generally functions as a guild or a trade union. The
caste acting as an occupational guild can effectively influence
the action of individuals or of corporate bodies outside its own
membership. In an eastern municipality dependent for its
sanitation on the manual labour of one of the scavenging castes
a strike of sweepers and dustmen may have the most devastating
effect on the comfort and health of an urban population. Com-
merce has been held up by strikes of buffalo-drivers in Calcutta
dissatisfied with regulations. There is nothing to prevent united
action by a big caste from having very considerable influence
on the election of political representatives, though interference
of this kind with elections would probably be anything but
beneficial to the political life of the community.

In the religious sphere the caste system so operates that the


caste can change or modify its social and religious observances
in accordance with changing ethical standards or the trend of
public opinion. It is familiar enough to find castes that are
wishful to rise in the estimation of the general public insisting
on their members giving up the keeping, and eating, of pigs
or chickens. The Kacharis of the remote and isolated Dhansiri
valley who in 1910 all kept and ate chickens had by 1920 almost
entirely eliminated the barndoor fowl from their economy as
the result of uplift propaganda received probably from
Kacharis
living in the Hinduized atmosphere of the plains of
Cachar,
Similar movements among the Gonds of the Central Provinces
are called purification , and the villages which have eliminated
fowls and pigs from their domestic economy are known as
ITS FUNCTIONS m
*shivraj\^ Another of the first steps usually taken by a low
caste wishing to rise in public esteem is to place a ban on the
remarriage of widows in order to bring their practice in that
respect into line with the orthodox Hindu teaching on the
subject and with the prevailing behaviour of the higher castes.
There seems, of course, to be no reason why caste should not be
activated in the opposite direction to liberalize the Hindu
tradition, but in point of fact it have
generally seems to
functioned hitherto in the direction of increasing restraint, and,
in view of the element of taboo in its origin, this is perhaps
only what could be expected,

The third of the functions of caste, and the most


aspect
important, is which has regard to the State, to the society
that
as a whole. The functions which caste performs for the indi-
vidual may be and are performed for individuals by other
institutions in other societies. The functions performed by caste
for the caste group as a body will be found performed in some
more or less analogous way in other systems by such social
groupings as exist in them. But caste is an unique institution
found only in India, and the functions which it has performed
and still performs for Indian society as a whole are not found
elsewhere. FurnivaU> has described British India as the out-
standing instance of a country where what he calls *a plural
society' has proved stable. Fumivall attributes this to castes
having given a religious basis to inequality, but it is undoubtedly
due largely also to the fact that the caste system has afforded a
place in society into which any community, be it racial, social,
occupational or religious, can be fitted as a co-operating part of
the social whole, while retaining its own distinctive character and
its separate individual life. Even Sherring, whose indictment
of the caste system is as savage as it well could be, admits that

caste promotes cleanliness and order, and is, in a certain sense,


a bond of union among all classes of the Hindu community
Many instances could be given throughout historic times of
the way in which this function has been performed by caste.
The Huns, it is true, are no longer recognizable under that

name, they seem to have been absorbed into Hindu society under
5 Information given personally by Dr Verrier Elwin.
^ Netherlands India, p, 464.
^ Hindu Tribes and Castes, in, p. 574.
116 CASTE
the guise of Rajput clans, but the Gujar caste has retained its
ancient tribal name, and that is the rule rather than the
exception. Many are the tribes who have now become castes.

The Meitheis of Manipur are still to some extent in the process,


or were until recently, for while part of the tribe have been
Hindus for some seven hundred years or so, perhaps, their ruler,
who is one of them, can take non-Hindu tribesmen into
still

caste as Kshatriyas, giving them the right to the sacred thread.


Indeed it is (in 1960) less than two hundred and fifty years since
the Maharaja abolished ancestor-worship and dismissed the indi-
genous priests from the temples of the local deities and super-
seded them by Brahmans. Many such instances can be given
of tribes or peoples which have apparendy within comparatively
recent times beconie castesKoli in western India, Chuhra in
the Punjab, Dom in the United Provinces and elsewhere, Turi,
Musahar, Agariya in Bihar, Rajbansi and probably Kaibartta in
Bengal, Kochh and Rabha in Assam, Badaga and Paliyan in
Madras, and so forth. Instances could also be given of many
tribes and peoples who, while not yet strictly speaking castes
within the pale of Hindu society, are, like the Gond and the
Bhil, rapidly assuming such a position. Both these tribes are
in a fair way to become integrated members of Hindu society,
even though they may retain part at any rate of their own
customary law. Customary law in India, as Bougie has rightly
pointed out/ attaches not to area as in the West, but rather to
groups, i.e. to caste.
Itnot so easy to give instances of occupational groups
is

taken into Hindu society from outside as a ready-made caste.


Instances could be more easily given of occupational groups
which have migrated from one part of India to another and
formed a new caste group in the new area which has become
completely separated from its original caste. This appears to
have happened more than once in the case of Brahmans as well
as of labouring castes like weavers or oil-pressers. New castes
have also been formed by the segregation of subcastes or of
occupational groups within castes, and these have been accepted
as independent castes just as tribes and peoples have been.
A recent case typical of the segregation of a new occupational
caste within the community is perhaps to be found in the

Hodson in Folk-Lore, xxm, p. 182. Op. cit., p, 184.


ITS FUNCTIONS 117

secretion of a caste of motor-car drivers. Enthoven writes * :

*
Modern India, having created a caste of chauffeurs from the
menials who tend motor-cars, is almost ripe for a Rolls Royce
caste rejecting food or marriage with the Fords. He should
rather have called it a Rolls-Royce subcaste, for at least it would

start in that way and, if true to pattern, reject first the giving
of daughters to Fords while not hesitating to take wives from
among them, secondly, the eating of food with them, and finally,
allconnexion of any kind, discovering a long-forgotten descent,
not shared by Fords, from some Brahman or Kshatriya ancestor
who drove the fiery chariot of Surya in the misty dawn of
mythology. Whether Enthoven be right or not in stating that
a caste of chauffeurs actually exists, he is most certainly right
in emphasizing ' the quite disproportionate importance attached
in India to trifling differences arising from one reason or
another in the conduct of small groups of individuals Work- *

men who wear pagris , says Bougie, will have nothing to do


with workmen in belts; the caste which mends shoes refrains


from making a pair of them and he goes on to quote Rhys
Davids to the effect that it takes three distinct craftsmen to
make a bow and arrows, though here we are not so far from our
own middle ages when the bowyers and the fletchers were sepa-
rate crafts. Such an attitude of behaviour is probably the in-
evitable consequence of a system in which each of many
component elements has been integrated on a principle of pre-
serving as essentials any features peculiar to itself and tending
to distinguish it from other groups. *For this is, in effect, what
the caste system has done and still does.
In a similar way religious groups, groups in many cases formed
in defiance of and in protest against the caste system, have
found themselves relegated, whether they wish it or no, precisely
into the position of castes in that system which they have aspired
to reform. Indeed, unless they were to succeed in destroying
the system, as the followers of the Buddha nearly did, it iS,diffi-

cult to see what other outconie such a movement could have


unless it broke away from Hinduism altogether. The Sikhs are a
case in point. The comparatively unimportant division of them
known as the Sahejdari Sikhs still claim to be within the Hindu
Reviewing Dutt*s Origin and Growth of Caste in India, vol. i, in the
7.R.v4.S. January 1932. cit p. 515.
10 October 1901, p. 863:
jjg CASTE
pale, but the more vigorous and independent Sikhs of the
Keshadhari persuasion have repudiated the claim to be called
Hindus, though still sometimes intermarrying with Hindu
families. A typical instance of the religious sect turned caste is
that of the Lingayats, a caste which started as the followers of
Basava, a teacher of the twelfth century who preached the
repudiation of caste. They wear as the symbol of their faith a
*

small silver box containing a stone phallus (Unga), the loss of


which is equivalent to spiritual death similar sectarian
caste is found in the Kabirpanthis, followers of
possibly to be
Kabir, who addressed himself to both Hindus and Muslims, and
was particularly popular among weavers; but the somewhat
similar sect of Satnami, largely recruited from Chamars, is care-
ful not to override the demands of caste.
While incorporations of occupational and sectarian groups
have been, as they are now, more or less automatic and in-
evitable, given the creation of the group in the first instance,
the incorporation of tribal, national, or political groups appears,
on the other hand, to have taken place as a result of deliberate
policy, and to have been used as an administrative device from
the earliest historical times. Jackson^ quotes Vasishtha to the
effect that the king should pay attention to all the laws of
countries, castes and families and make the four varna fulfil
their respective duties; he quotes Apastamba 'who says that if
those who have broken caste rules fail to perform the penance
prescribed by their spiritual guide, he shall take them before the
king* who shall 'reduce them by forcible means',
(to reason)
short of corporal punishment and servitude if Brahmans, but
with punishments which may amount to death in the case of
other castes. He also quotes Manu (vii, 201-3, Buhler's
translation) as follows: 'When he has gained victory let him
duly worship the gods, and honour righteous Brahmanas, let him
grant exemptions, and let him cause promises of safety to be
proclaimed. But, having fully ascertained the wishes of all the
(conquered), let him place there a relation of the (vanquished
ruler on the throne) and let him impose his conditions. Let him
make authoritative the lawful (customs) of the (inhabitants),
Enthoven, Tribes and Castes, n, p.
344.
Crooke, Tribes and Castes, rv,
pp. agg sqq.
'
Note on the History of the Caste System y.A.S.B.. \o\. in, no. 7.
IlllV
ITS FUNCTIONS 119

just as they are stated (to be). Jackson goes on to point out
that a large proportion of subcastes bear geographical names,
and while some take a name from the kingdom others, parti-
cularly the trading and artisan castes, take it from the capital
city, and that many traces are to be found of sets of subcastes

bearing specific geographical names, which is what we should


expect if each of the old tribal kingdoms had its own sections of
priests, traders, artisans, etc. He gives among other instances
those of Gujar subcastes among the castes of Darzi, Sonar, Sutar,
Charan, Dhed and its synonym Parajia)
Baniya; Kachela (or
among the Lohar, Sonar,Charan and Brahman castes (Kachela
means an inhabitant of Kachh); Modh among Brahmans,
Baniyas and Ghanchis (from the town of Modhera); Surati
among Ghanchis, Lohars, Mochis and Dheds (after the town
of Surat or its district), and a number of other instances all
from the former Bombay province. They could, of course, with
little difficulty be added to by parallel instances from all over

India. Clearly geographical, political and administrative con-


siderations have not been unimportant in the development and
operation of the caste system, though the essential concepts of
pollution, commensality, and endogamy are ritual rather than
economic in nature.
It will be understood then that one important function of
caste, perhaps the most important of all its functions, and the
one which above all others makes caste in India an oinique
institution, is, or has been, to integrate Indian society, to weld
into one community the various competing if not incompatible
groups composing it. Some of these groups have been occupa-
tional or religious. Others, and this is more important, have
been national, political, and tribal societies that must otherwise
have either been absorbed and transformed or remained as un-
adjusted and possibly subversive elements. Generally speaking
conquered peoples and their conquerors blend into one society in
which one or other element may in the long run predominate.
The Castilians who conquered Spain absorbed the Moors that
remained there; Anglo-Saxons, British, and Danes coalesced,
and absorbed their Norman coriquerori^. On the other hand the
negroes of the United States remain so far an unadjusted popu-
lation, while European and Bantu in South Africa remain no
less segregated in a different way; The caste system has effectively
ISO CASTE
dealt with problems such as these which other societies

have failed to solve* It has involved no doubt disadvantages


peculiar to itself which are indicated in the proper place, bu^
an integrator of peoples is perhaps
this aspect of caste as
without some bearing on problems facing the world today, where
the question arises of incorporating the various nations of the
world into some sort of integrated whole which is presumably
the only alternative to yet more internecine conflict.
India
writes Gilbert, * has developed a system of castes which, as a
scheme of social adjustment, compares rather favourably with
the European system of warring territorial nationalities.'

A second very important function of caste has been to act as a


political stabilizer. It was this function, apparently, of the caste
system which so moved the admiration of the Abbd Dubois,^
who considered the caste system of the Hindus
the happiest
effort of their legislation He regarded the caste system as
being responsible for the preservation of India from complete
barbarism, and as the sure basis of orderly government, as a
defence against despotism, and a means for preserving the arts,
and as a sure means of preserving the Hindu pattern of culture
under the regime of alien conquerors. At a more recent date
S. C. Hill ^ has taken a similar view. He points out that
whereas '
in Europe we are accustomed to think of the Political
and Social systems of a country as one and the same thing',
the Hindu regards them as distinct, and separable in fact. His

intimate life . . . the life which to the Hindu really matters, is


altogether independent of the political conditions which happen
to prevail.' He calls the caste system 'the only Social System
ever proposed upon a basis stronger
it puts into than Force';
successful practice the only instance of a
scheme of social life
which is 'entirely independent of any form of political govern-
ment, .The system is permanently stable because of the
. .

complete absence of any motive on the part of the ruled for


seeking any alteration.'
The admiration of Dubois was not without some justification,
and he has not been the only writer to draw attention to the
value of the caste system as a stabilizer of society. Indian society

P* 1"** .
^^opie of India, pt. i, ch. ii.
System in India
Oct x, xii, xiii, xiv (LA., March-
ITS FUNCTIONS isi

has survived a vast number of invasions, famines, revolutions


and social upheavals of all by invaders
kinds, including conquests
of alien religions essentially antagonistic to Hinduism, and there
can be no doubt but this is largely due to the caste system on
which that society has constructed itself, a system which has often
survived even conversion to Islam or Christianity. Muslim castes
are a familiar phenomenon; Momin or Jolaha, for instance, a
caste of Muslim weavers; Lalkhani, a caste of Muslim Rajputs;
Gaddi, a caste of Muslim graziers; Mirasi, a caste of Muslim
minstrels closely akin to the Hindu caste of Dharhi, and so
forth. Indeed, one Muslim caste, the Turkiya Banjara, have
definite gotras, three of which have Muslim namesBahlim,
Khilji and Shaikh.^ As for Christians, it is not very many years
since the Madras High Court had to adjudicate on the rights
of a Ctiristian congregation which sued its priest for breaking
down the wall which divided the aisle of the respectable from
the aisle in which sat Christians of the untouchable castes, while
Bougie mentions a church in the Tamilnad with separate
naves giving on to a common chancel to accommodate hostile
castes. Indeed, Iyer, writing of the Syrian Christians of the
Malabar Coast, says The average Indian Christian is a staunch
:

observer of castes. It is a moot point whether he is not stricter


in his observance of castes than the average Hindu. There are
a large number of Christians in the Southern Districts of the
Madras Presidency who even boast of their being firmer, and
truer adherents of the caste system than the Hindus. The
reason of the social stability conferred by the caste system is no
doubt its provision for the unlimited extension of the society by
the inclusion as integral parts of the structure of any number
of extraneous, particularizing, or segregative entities.
Again, the caste system does provide for the various functions
necessary to social life, functions ranging from education to

scavenging, from government to domestic service of the most


menial kind; and it makes this provision under the sanction
of a religious dogma, the belief in k&rm&, which renders the
superficially inequitable distribution of functions acceptable as
being part of the divine order of the universe and a transient
episode in the prolonged existence of the individual soul, which

Blunt, Caste System, p. 45. 1* Op. cit., pT. 26 n.


L. K. A. lytr. Anthropology of the Syrian Christians, p. 218.
by acquiring merit in one existence
may rise in the scale in the
from a degradation in caste
nkt, or which may be suffering
in a previous life. As an
merely by reason of its transgressions
stabilizing effect, caste, as Bougie
important element in this
has pointed a damper, so to speak, on the rise of the
out,** puts
in spite of the wealth
parvenu. The Shanan of southern India,
to build two-storied houses,
they have acquired, have no right
to wear gold ornaments, or to
sport an umbrella. ^ The very
and discourages attempts to
nature of the system discountenances
existing barriers of rank or occupation or to break
surmount
those barriers by intermarriage or by
freedom of social
down
caste has developed as
intercourse generally. The truth is that
caste stands to the society
a quasi-organic structure in which the
almost analogous to that of the indivi-
as a whole in a relation
dual cell which it forms part. The
to the greater organism of
many undoubted advantages which follow from a development
of this kind are more or less inevitably attended by certain
disadvantages to which be necessary to draw attention.
it will
-This view of the caste system asan organism as it were, a
composite unit of many individual cells each functioning inde-
pendently, must not be taken as precluding a certain fluidity in
the system, a power of mutability in caste itself within certain
definite limits. Hindu society may be classified very roughly

into four gradesthat of Brahmans, that of other twice-born

castes deriving in theory at any rate from the Kshatriyas and


Vaishyas of the vedic age, that of Sudras, and that of the exterior

castes. No one not born a Brahman can any longer hope to


become one, though that seems to have been possible in the
early stages of Hinduism. Castes in the third category can and
do succeed in establishing themselves in the second; castes in
the fourth can and do get accepted in the third. That is about
as far as the mutability of castes goes. Worldly prosperity and
a high level of education enable a caste to rise first in its own
estimation, later, and much more tardily, in the general estima-
tion of other castes, so that in the course of time it is able to
establish recognition, grudging at no doubt, but ultimately
first

more or less general perhaps, in a group higher than that to


which it belonged by origin. Indeed, Blunt insists that the

20 Op. dt, p. 212 :


*
Linde ne vent pas connattre la figure du
parvenu 21 Bhattacharya, op. dt.,
p. 259.
Caste System, rp. soB.
ITS FUNCTIONS ists

caste system is so mutable that no account of it remains accurate


for long: 'its evolution is still proceeding: new castes and
subcastes come into existence, old customs fall into abeyance/
This, no doubt, is true enough, but it is not the system that is
changing so much as caste details within it. Caste, no doubt,
keeps changing, and customs come and go; the pattern alters,
but the principles that govern it, the frames that hold the pat-

tern so to speak, are exceptionally constant for a human institu-


tion. It has been already suggested that there are certain
disadvantages probably inseparable from the advantages of the
caste system, and it is necessary to draw attention to these in
any attempt to describe the functions which caste performs.
Pramatha Nath Bose, in his History of Hindu Civilisation during
British Rule,^^ has stated that the caste system has acted essen-
tially to impose that attitude of mind needed to raise men
from
savagery but to stop them half-way on the road to progress. In
so far as this is true, and in a great measure
would seem to it
be true, it is probably the result of the passing on of crafts and
skills by means of an hereditary environment in
which the son
learns as it were unconsciously from the father in the process of
his upbringing. It is no doubt this method of transmitting a
craft which makes an innovation in method appear sometimes
as if it were a sin against the craftsman's ancestors. Sonnerat
found it useless to try to get Hindu carpenters to use improved
methods of sawing wood,^* and a similar difficulty is still expe-
rienced in the hills of Assam, where the local inhabitant prefers
the adze to the saw; and it is noticeable how prevalent in eastern
India is the employment of Sikhs as carpenters. The Sikhs, be
it noted, belong to an offshoot of Hinduism
which has repu-
diated caste. The fisherman in the play of Sakuntala who is
reproached with cruelty to fish replies that he cannot be blamed,
since it is wrong to give up the calling of one's ancestors however
degraded that calling may be.^ Caste makes no compromises,'
says Sherring. The most ignorant Hindu is able to compel
the obedience of the most intelligent.'
It is a possible further disadvantage of the caste system
.
eco-
nomically that it probably discourages organization from above
^3 Quoted by Bougie, op. 24
cit., p. 149. Voyage aux Indes, i, p. 104.
^=5
Quoted by Bougie, op. cit., p. si. 2 Op. cit., HI, p. 376.
W4 CASTE
by the entrepreneur, no less than it forestalls ambition on the
part of the workman. Durkheim has suggested that the function
of the division of labour is to give the individual more freedom
by substituting an organic for a rigidly mechanical economy,
but the organic structure created by the caste system would seem
to have provided for the division of labour on a plan ingeniously
calculated to avoid giving just that freedom; for occupation is
determined by status instead of contract, and transition from
status to contract, which Sir Henry Maine regarded as an essen-
tial feature of political progress,^ is nullified and frustrated. A
system of this kind must operate to stifle progress in economic
life, giving it an almost paralysing stability. The great strides
made in industrial progress in India in the twentieth century
have been made in spite of the caste system, not as a result of
it. It is likely also that a further political consequence of the
caste system has been to simplify the intrusions of foreign
invaders by opposing to them a society irreparably split up and
unable combine, a society that has for centuries lacked a
to
national life, so much so, perhaps, as to justify Bougies
apophthegm that for Hindus patriotism consists in attachment
to the caste system and they achieve the paradox of being unable
to unite except in the very culture pattern that divides them.*
This aspect of the caste system has a specific religious sanction.
It has been held to have been established by divine ordinance

or at least with divine approval, and the Bhagavad Gita, which


probably more nearly than any other sacred scripture fulfils for
Hindus the function performed for Islam by the Quran or for
Christians by the Bible, contains some well-known lines which
inculcate the supreme merit of performing ones caste duties.

Ones own duty [i.e. dharmS. or caste rules], though defective,


is better than anothers duty well performed Custom, says
the Code of Manu, is transcendent law. Perfection, we are told
in theMarkandeya Purana, can only be attained by the man who
does not deviate from the duties of caste. In the observance
of these duties therefore consists what is known as dharma, that
is, religious observance, moral obligation, righteousness; the
word is not easy to translate precisely, but God himself is spoken
Ancient Zazc (1880), p. 170. ^8 Op. cit., p. 194.
2* Quoted by Hill, loc. cit., March 1930.
so Wilson, Indian Caste, i, p, 437.
ITS FUNCTIONS 135
of as Dharmeswar, the Righteous One. At any rate it is dear
that the sodal habits of caste are inextricably tied up with
and their sanction is reinforced by the doctrine of
religion,
karma already alluded to. Under this doctrine a mans condi-
tion in this life is the result of his conduct in his last incarna-
tion; his high or low caste
is therefore the reward or punishment
of his past behaviour, and if unsatisfactory can be remedied by
acquiring merit in his present life which will raise his status in
the next. The Laws of Manu contain the following passage:
(A Sudra who pure , etc., and free from pride, and always
is)

seeks a refuge with Brahmanas, attains (in his next life) a hi her
g
caste. No doctrine is more likely to conduce to contentment
in this life. The ties of caste , writes Sherring,
are stronger
than those of religion. A man may
be a bad Hindu so far as
the practice of his religious duties goes; but caste rules must
be minutely observed, or he will have to reap the consequences
of neglect. With many Hindus the highest form of religious
observance is the complete fulfilment of the of caste;
and most of them conceive of sin as a breach of caste discipline
rather than of moral law.
The doctrine of karma must clearly help to palliate the appa-
rent inequity of the caste system, but it seems likely that it has
been grafted on to it in the course of its development, and is
not inherent in its origins. In any case it is quite dear that
the caste system has been effectively utilized by the Brahman
priesthood to maintain the existing form of Hindu society with
the Brahmans as a privileged class, and that caste has been taken
advantage of generally by the higher castes to keep the inferior
and exterior castes in their respective stations, if not to reduce
them to the lowest status possible. It is true that theoretically
greater responsibilities rest on the higher castes, but the Brahman
reception or rejection of water or food is the measure of the
status of any given caste in a given place. And in point of
fact a tribe which does not recognize Brahmans or Hinduism
at all generally in a better position socially with regard to
is

the Brahman than the exterior castes whose occupation is


degrading according to Hindu law and therefore polluting to
Brahmans. Dubois, an admirer of the caste system, has

The Laws of Manu (tr. Buhler), ix, p. 335.


Op. cit., HI, p. *76. as Webb, These Ten Tears, p. *oi.
is 6 CASTE
described their arrogance and finds it not unnatural after all

in a caste brought up with the idea that they have a right to


everything and owe nothing to anyone. It is not difficult to
give instances of the pollutability of Brahmans. In a story in
the Jatakas two Brahman girls playing at the gate of a city
happen to see two Chandals. They run away at once to wash
and decontaminate their eyes. The reason why it is the general
custom in India for Indian gentlemen to call on Europeans at
as early an hour in the day as possible is commonly believed to
be to enable them to decontaminate themselves so as to take
their morning meal in a state of ceremonial purity. It is not

surprising then that there were castes which had to carry warning
bells like a medieval leper, and that they were treated as one
might expect when they were completely without power them-
selves and inspired nothing but disgust in persons of higher caste,
a disgust so great that contact with them by sight or by the
passage of a breath of air necessitated ceremonial purification.
The theory that greater responsibilities and obligations rest on
the higher castes is no doubt put into practice in the form of

a more scrupulous cleanliness, a stricter observation of rules of


diet, a truer regard for ceremonial observances; but generally
speaking the responsibilities accepted are rather ceremonial than
moral, and although the Hindu sacred books lay down a rule
of strict asceticism for Brahmans, for instance, and of public
service in the form of be given by them
religious instruction to
to other castes, the rule is not taken too seriously by the gene-
rality of the Brahman castes. At the same time it should perhaps
be made clear that although there are many different Brahman
castes Brahmans do form a cohesive group with much solidarity,
and, to -some extent in spite of caste, constitute a sort of intellec-
tual aristocracy.
The caste system is also open, as a result of the particular form
which its religious sanction takes, to the criticism that it may
aCt as a saeen and justification for persistently anti-social con-
duct. The fisherman in the Jataka story justified his sin against

Op. dt., p. 36 (pt. I, ch. vi); and see Wilson, Indian Caste, i, pp. 390,
411, 426; Ketkar, History of Caste, pp. 138 sqq.; Dutt, op. dt., pp. 148
sqq., etc.
Dr J. Gonda, in a review of the first edition of this book in Erasmus,
refers to a text of Kautilya (Arthasastra, ch. xix) according to which kings
should receive intelligence, etc. before the daily bath and breakfast.
ITS FUNCTIONS igy

the fish and


against the teaching of the Buddha by an appeal
to the practice of his caste. In another case the identical argu-
ment is pleaded by a butcher for practising his hereditary
trade.*
Precisely the same justification has been used by many criminal
castes to justify their behaviour towards theL- fellow men. The
Thugs undoubtedly made a religion of Thuggee and strangled
their victims to the honour of their goddess Bhawani as well as
to the enrichment of their pockets. They worshipped also the
sacred pickaxe with which they buried the bodies of their
victims and from the position of which they read omens before
an enterprise; when hidden in a well this pickaxe rose in the
morning of its own accord and came of itself into its custodians
hand; its sound made in digging a grave could be heard by
none but a Thug; it was more sacred than Ganges water or the
Quran, and a Thug who took a false oath on it died within six
days. *' They considered that their victims were killed by
. . .

God, with them as his agents, their appointed job being to kill
travellers
to (juote the words of a Thug, just as a tiger feeds
upon deer . The caste had a life of at least more than five
hundred years, as the earliest historical mention of Thuggee
isin Zia-ud-Din Barnis history of Firoz Shah, written in 1356,
and refers to the capture, and release, of a thousand Thugs at
Delhi about a.d. 1290.^ Other criminal castes, like Korava,
regard their criminal practices as justified by if not enjoined on
them by their membership of the caste to which they belong
and by their common share in its ancestors and traditions.
Again, Dubois says of the Kalian caste that they exercise their
profession of robber,
without disguise, as their birthright . . .

and conceive their calling no way discreditable to themselves


or their tribe, as having legitimately descended to them by way
of inheritance. So far from shrinking at the appellation, if one
of them be asked who he is, he will coolly answer that he is a
robber.
The above considerations suggest that there may be good
reason for Fumivalls criticism of 'plural societies. In exa-
mining their economy he says that a plural society in its political
aspect
resembles a confederation of allied provinces, united by
38 Hopkins, Ruling Caste in Ancient India, p. ij?o.
3^
J. L.Sleeman, Thug or a Million Murders, pp. 8, ig, as, 57 and passim,
38 Hatch, Land Pirates of India, ch. vi.
39 Op. cit., ch. ^0 Op.
p. 3 (pt. I, I). cit., pp. 447 sqq.
laS
CASTE
constitution, merely for
treaty or within the limits of a formal
certain ends common to the constituent units and, in matters

outside the terms of union, each living its own life. But it
differs', he says, from a confederation
in that the constituent

elements are not segregated each within its own territorial


limits , so that The elements are so intermingled that secession

is identical with anarchy. This is not an unfair description


of the caste system, and Furnivall goes on to argue that such a
society lacks a common will in economic life, and there is in
consequence an absence of any common social demand; social

wants are sectional, and


there is no social demand common
to all the several elements. This, he maintains, raises the
economic criterion to a new place in the scale of social values ,
since the only place where the various sections of the community
meet on common ground is the market-place, where the highest
common factor of their wants is the economic one; the elements
of such a society come to regard the production of material goods
as the prime end of social life; the social needs of the country
as a wWe are not apprehended, social demand itself becomes
sectionalized, and within each section of the community the
social demand becomes disorganized and ineffective, so that in
each section the members are debarred from leading the full life
of a citizen in a homogeneous community; finally the reaction
against these abnormal conditions ... sets one community against
the other , enhancing the need for the society to be held
together by some force exerted from outside . Such a force is
clearly to be found in the law, but in a society which has no
common bond but law. Right is superseded by Legality, and
the only Duty is not to be found out . Furnivall clearly regards
the rule of law in a plural society as responsible for the litiga-

tion which is a curse of British India and the aristocracy


of native lawyers and money-lenders which is typical of British
India outside the Native States.^' One would say at first

thought that the caste system had the common bond of religion
as well as of law,but we have already seen that the first
injunction of religionis to obey caste rules, which suggests that

Hinduism fails to act as an overriding bond for the plural


societies represented by castes. At any rate those who have
known India will see much force iri what Furnivall has said.

Ibid., p. 466.
ITS FUNCTIONS 129

There are also drawbacks to the caste system in India


which arise not so much because of the nature of the system as
incidentally to its development. One of these is the hardship
which is entailed on generations of women in all those castes
that aim at raising their position in the social scale. Any caste
or subcaste that wishes to rise in the estimation of its fellows,
or to claim an origin from some twice-born varna to which it is
not generally conceded to belong, finds it essential to conform as
nearly as may be to the traditional standard of Brahman castes
in the matter of insisting on the marriage of girls' before they
reach the age of puberty, and in forbidding their remarriage even
if widowed in infancy. This is a very high price to pay for a
shadowy descent from some doubtfully historical rishi or for
the unsubstantial glories of Rajput blood and state. It is

perhaps, however, in its treatment of outcastes, that is, of castes


outside the pale of respectable society, the
exterior ' castes as
they have been called, that the working of the caste system is

most open to criticism. The position of these exterior castes


has already been indicated. It may be attributed to failure,
wilful or otherwise, on the part of the exterior caste to observe
the stricter prohibitions of Hindu society, the prohibition, for
instance, against the killing of cattle, which keeps outside the
pale tanners, cobblers, and all those who deal in the products
of their dead bodies; the prohibition, albeit less stringent,
against crushing the oil-seed, which makes the oil-presser an
exterior caste; or the ban on alcoholic liquor which makes the
toddy-drawer an outcaste. It may be attributed also to the
pursuit of some even necessary profession which involves contact
with some sort of pollution. One might have expected the
washermans to be a cleanly pursuit, but the fact that it brings
him into contact with menstrually polluted clothes is enough
to make him an outcaste no less than the scavenger who removes
night-soil or dead bodies. And in all cases the nature of taboo
is such, of course, that the contagion of the polluted occupation
contaminates all members of the caste whatever their individual
occupation may and to an infinite number of generations.
be,

xhis taboo is, of course, world-wide. There can be little doubt but
the reason why the adjectives bloody * and * bleeding are, or till recently
*

were, regarded as indecent in English is because they have been used as


swear-words with that particular association implied.
9
The number of these exterior castes in India was found at the

census of 1931 to be more than 50 millions, or 21% on a Hindu


population of 239 odd millions. Their position was gone into
in some detail in an appendix to the Report on the Census of
India, 1931. Extracts from that appendix have been included
in an appendix to this volume as being germane to the whole
question of caste as treated here. It is quite clear that the
unfortunate position of the exterior castes could be remedied
without destroying the caste system. Their position is the result
of prejudices and taboos which have been operative during
millennia, and cannot be got rid of in a night. That is a
process which can only be accomplished in the course of time,
and with the help of education in its broadest sense. To destroy
the caste system, even if it could be done, might be disastrous
to the society which is constructed on its framework. Religious
beliefs make it tolerable and even reasonable to the members
of that society. No doubt modifications in those beliefs, which
probably will take place in the course of time, which probably
are indeed already taking place, will bring about in their train
a modification of the caste system. At any rate they will make
modification easier, and will automatically improve the position
of the exterior castes. Meanwhile there seems to be no reason
why a great improvement in the position of those castes should
not be made v/ithin the system as it exists today, and without
any necessity for destroying a social order which performs a
number of useful functions both for the individual and for the
society. Caste is probably what Professor Bartlett would call
one of the hard points of Hindu culture,

and any attempt to
modify it by a direct attack on it is likely to provoke resistance
and discord, and reformers will need to aim at some soft point
some other feature of the culture, that is, through which the

hard point can ultimately be circumvented and isolated. The


*

mere fact that caste is one of the chief differentiating cultural


features between Western cultures and Hinduism, is, if Professor
*

Bartlett is to be followed, a reason for scrupulously avoiding any


direct attack on it.

^3 Anthropology in Reconstruction, p. 5 .
ITS FUNCTIONS 13,
There is yet another function which the caste system probably
performs, and that is a genetic one. India
is a country in which
the male sex generally outnumbers the female. The ratio in
1931 was 1,000 males to 941 females, or, if Hindus alone be
considered, to 953, and this inequality is likely to be felt most
acutely among Hindus on account of their ban against the
remarriage of widows. Now a good deal of fairly recent work
on sex ratios has pointed to the conclusion that an excess of
males is indicative of a declining population, but this is certainly
not the case in India. It is possible that the caste system itself
contributes to the preponderance of masculinity. Westermarck **
takes the view that a mixture of race leads to an increase
in the ratio of females to males. He cites a number of observa-
tions from various parts of the world to support this view:
he quotes Dr Nagels experiments in the self-fertilization of
plants, as producing an excess of male flowers, and several cases
of inbreeding herds of cattle in which bull calves greatly exceed
two independent experiments in horse-breeding
heifers, as well as
indicating that predominate among foals in proportion
fillies

as sire and dam differ in colour. George Pitt-Rivers quotes


Heape to the effect that experiments upon dogs indicate that
inbreeding produces a high percentage of males. Westermarck,
again, quotes the Talmud as stating that mixed marriages pro-
duce only girls,*" and mentions a number of statistics pointing
to an excess of female births as the result. of mixed marriages.
The causal connexion between inbreeding and an excess of male
births is not proved, but there is a good deal of evidence to
support the theory that pure-blooded societies produce an
of males, and Miss King's experiments with rats afforded evidence
that the normal sex ratio could be changed by breeding from
litters which contain an excess of males.*^ Sedgwick, in his
Report on the Census of Bombay, 1921,* pointed out that the
Indian caste system with its endogamous caste and its exogamous
gotras 'is a perfect method of preserving what is called in
genetics the pure line. The endogamy prevents external
hybridization, while the (internal) exogamy prevents the

History of Human Marriage (1901), pp. 476-8^.


-4S The Clash of Culture, p. 111, referring to Proc. Camh, Phil: Soc, xiv
pp. sx et seq. Op. cit., p. 470,
Pitt-Rivers, loc. cit. 48
p,
CASTE
pLbSUy o a teb P" 8 'whh
Witti
to the whole line.
isolation of any character not common
line the
the preservation of the pure
necessarily follows. Whether this
characters common to it

.Btement o th. position be
strictly

if once a caste,
whether as the, result ot
least be conceded that the
different factor, have acquired
inbreeding or of some quite condmon
of ha^ng an excess of
m^es, this
ntmal condition
maintained.
long as inbreeding is
is likely to be
perpetuated as
Caste, Lrefore' would
appear to be of definite ce ^o
There
for male offspring.
Hindu in his superlative anxiety
man.
is no heaven for the sonless
PART THREE
ORIGINS

CHAPTER IX

ANALOGOUS INSTITUTIONS
ELSEWHERE
It has already been said that caste is an exclusively Indian
phenomenon, but there are in various parts of the world ana-
logous institutions which resemble caste in one or other of its

aspects, and before proceeding to examine the question of how


caste originated it will be useful to glance at these institutions,
since they are likely to throw light on the origins of caste proper,
which are not one but many. It is no doubt the manifold
origin of caste in India that is the cause of its being unique,
since it Is unlikely that a complex combination of factors could
be exactly repeated in some other part of the world. It is not
to be understood, therefore, that there is any complete analogy
to caste in any non-Indian society, but there are a number of
social phenomena which offer a parallel to one or other element
in the caste system. How mixed in origin caste is is aptly
illustrated by Molony,* who remarks that the caste table for the
Madras census in 1911 suggested a division of the inhabitants
of England into families of Norman descent, clerks in Holy
Orders, noblemen, positivists, ironmongers, vegetarians, commu-
nists and Scotchmen. The difference is that whereas these
categories overlap in the English social system, corresponding
categories in the caste system would be mutually segregate.
Risley has elaborated the idea in his chapter on social types.*
An allusion has been made in the last chapter to FurnivaHs
treatment of plural societies.* There he treats caste as one
system by which the problems inherent in a plural society are
resolved, alternative resolutions being Law, Nationalism, and
Federalism, and he suggests that three colonizing peoples-the

1 A Book of South India, p. 106. * People of India, ch. 11.

* Netherlands India, ck, Jun, , v


134
ORIGINS
Portuguese, British, and Dutchhave adopted these
respectively

three solutions of what is primarily an economic problem.


For
a plural society is one in which different elements perform
different functions. Thus in Siam, he says, the Siamese, Chinese,
and Europeans have distinct economic functions, and live apart
'

i
as separate social orders \ Similarly, in South Africa and in the

United States there are white and coloured populations exhibit-


ing a not dissimilar phenomenon, and in the French provinces of
Canada, English and French Canadians who grow up separated

by race, language and religion in two distinct societies,


*
Even where there is no difference of creed or colour, a com-
munity may still have a plural character, as in western Canada,
where people of different racial origin tend to live in distinct
settlements and, for example, a northern European cannot find
work on the railway, because this is reserved for Dagoes or
''

M/ops. And in lands where a strong Jewish element is


regarded as alien, there is to that extent a plural society.
Clearly there is a social phenomenon in this which has much
in common with the caste system, particularly in its economic
aspect. At the same time it is clearly not caste in the Indian

sense, a truth well illustrated by FumivalFs citing Burma as


a country of plural society where different functions are dis-
charged by Europeans, Indians, Chinese and Burmese,^ but one
in which thereis no caste.^ Yet Burma from quite a different,
and a non-economic, aspect offers us yet another remarkable
be seen later.
parallel to caste, as will
Another institution in some ways analogous to caste in its
economic and also more particularly in its occupational aspect
is to be seen, as pointed out by Risley, in the ordinances of

the Theodosian Code of the Western Roman Empire in the


early part of the fifth century a.d. The intention of these
ordinances seems to have been partly an administra-
fiscal, partly
tiveattempt to ensure the performance of the various functions
needed for the social life of the community. The method by
which it was sought to do this was to make it compulsory for
a man to follow his fathers calling, and in certain cases at any
rate to compel a man to follow the calling of the family into
which he married. Thus owing to a scarcity of bakers not only
^ Op. dt., p. 467. Ibid., p. 465.
People of India (1915), pp. ^70 sq.
ANALOGOUS INSTITUTIONS ELSEWHERE 135
was a baker^s son compelled to be a baker, but if a man married
a bakers daughter he himself had to become a baker. All
public functions were made hereditary; even the curiales, the
quasi-priesdy persons responsible for the secular and religious
functions associated with what might be described as municipal
wardsaldermen we may perhaps call them, but they were
responsible for the collection of taxeswere ordered back to
their native cities and compelled to resume their original
functions,which they were forbidden to evade by taking govern-
ment They were eve.n denied the right of asylum in
service.
the Church. The wine merchant, oilman, butcher and pork-
butcher, the transport contractor and the corn merchant, the
waterman, the stoker in the public baths, even the senator were
all made hereditary and bound to their callings from one genera-

tion to another. The troops known as burgarii who garrisoned


the frontier posts were likewise bound to their military profession
by heredity. *
It was says Dill,^
the principle of rural serfdom
applied to social functions. ... A man
was bound to his calling
not only by his fathers but by his mothers condition. Men
were not permitted to marry out of their guild. Not even . . .

the power of the Church could avail to break the chain of


servitude. The latter point illustrates very clearly one aspect in
which the society sought to be created or perpetuated by the
Theodosian Code, and by similar legislation of the period, difEers
from caste; that is, in the matter of religious sanction. The
society analogous to caste of which the last century of the
Western Roman Empire affords an example lacked completely
the religious sanction which caste receives from Hinduism. It
lacked also the basis in taboo and magic, and other essentials
such as the gradual development over a period of centuries or
rather millennia. At the same time it affords a definite analogy
to the occupational aspect of caste.
A third analogy to caste, also pointed out by Risley, who lays
much more stress on
one than on the occupational one, is
this

the racial analogy. He argues that whenever the conquest of


one people by another has taken place it has been followed by
interbreeding, and by an initial stage of hypergamy; that where
the two peoples concerned are of the same race, or at any rate
*

of the same colour the initial stage of hypergamy passes away

f Roman Society, bk. in, ch ir. ^ People of India (1915), p.


136 ORIGINS
and a stage of complete amalgamation follows. On the other
hand, where marked distinctions of race and colour intervene

the tendency towards the formation of a class of half-breeds,


is

the result of irregular unions between men of the higher race


and women of the lower, who marry only among themselves and
are to all intents and purposes a caste As instance he men-
tions the southern States of the U.S.A., 'where negroes inter-
marry with negroes, and the various mixed races, mulattoes,
quadroons and octoroons, each have a sharply restricted jus
connubii of their own and are absolutely cut off from legal
unions with white races. He goes on to give as further
instances the half-breeds of Canada, Mexico, and South America,
and the Eurasians of India, and, a better instance, perhaps, the
Burghers of Ceylon who are of mixed Dutch and native extrac-
tion. By way of a precise Indian parallel Khas
he refers to the
of Nepal and the Dogras of the Kangra valley, similarly formed
by alliances between Rajput or Brahman immigrants and
women of the country. He fails to mention the cases of the
North American Indians and of the Maoris,, which perhaps tell
against his argument; but the analogy in any case fails in certain
important particulars.
It is true that the negro in the southern
States hasbeen in many respects kept segregated as a distinct
community, prohibited or at least prevented from using the
same public amenities as white men. It may be the case that
there are different grades of colour each with its own jus
connubii, and it is enough that there is a strong prejudice
certain
cm the part of the whites against mixed marriages, but the
question of taboo and pollution by touch hardly arises.* A
negro servant to a white man is no strange anomaly, but a
Brahman with a Chandal cook is unthinkable, and hardly less
so a Rajput with a Dhobi for valet. Conquest and the colour
bar may contribute to caste; they may be important factors
in the complex; and they help to account for its existence in
India; but they are not by themselves the cause of caste, and
conditions such as those under which negroes and whites live
together in the United States do not form a true parallel to
caste in India, and it is doubtful even if any system really
analogous to hypergamy has ever existed between whites, octo-
roons, quadroons, mulattoes and negroes.

Bot see below, p. 174, n, 4.


ANALOGOUS INSTITUTIONS ELSEWHERE 137

What Risky calls


a singularly complete parallel to the Indian
usage of hypergamy by him quoting van Genneps
is cited
Tabou et Totemisme d,The patrician caste, he
Madagascar.
says, is divided into six classes endogamous in theory, but in
practice allowing a man in' marry a woman of a lower class
while a woman is 'prohibited by strict taboo from marrying
a man of a lower class. If this is correct of any of the Mada-
gascar tribes, it does not apparently hold good of all of them,
as Linton, in his monograph on the Tanala, makes dear that
in that tribe, although there are distinct dasses of patridan
origin, of slaves descended from prisoners of war, and of an
intermediate commoner group, and although the patricians
disapprove of their women marrying commoners or slaves, there
can be no strict taboo on so doing, since such marriages did take
placeand cases of commoners marrying slaves were comparatively
common. In any case these classes seem to be social grades rather
than castes, and the prohibition on intermarriage was not one
of taboo so much as one of mere social prejudice, while there
was no commensal taboo as in India. It is probable that the
Madagascar dasses therefore can hardly be regarded as castes
in the Indian sense, even if the prejudice against social relations
between different classes was even greater than similar dass
prejudices have been and indeed are in most European countries.
**
This view is confirmed by the information given us by Ellis
and Further, it must be borne in mind that such
Piolet.
parallels asMadagascar may have to offer to Indian social struc-
ture may themselves be derived from India, since it is quite
clear that much of Malagasy culture has an Indonesian origin.
Indeed, the account given by the Abb6 Rochon of the origin
of the Malagasy castes is dosely analogous to the traditional
origin of the four varna in Hindu mythology.**
An Indian origin, of course, goes without saying in the case
of Ceylon, where caste appears in the ritual of the Buddhist
Temple of the Tooth. Hocart insists that royal courts cannot
be distinguished from temples, since the king represents the
gods, and therefore all the different participants in ritual who
are required to shave the king, wash his clothes, etc., are found

People of India (igis), f. j8o. The Tanula, pp. 137 sq., 143.
History of Madagascar, i, p. 164. Madagascar et les Horn, p. 97'
i"* Madagascar; or Robert Drurys- Journal etc,, and a Further description
. . by the Abb^ Alexis Rochon (ed. by Oliver, 1890), p. 368.
138 ORIGINS
at royal courts and at the courts of his vassals and subjects, who
imitate his behaviour, down to the mere cultivator of land.^*

This is a point to which allusion is made below in discussing


the origins of caste more directly. It is enough here to point
out the parallel between the ritual aspect of caste, which requires
certain purificatory services from certain sections of the people

in Ceylon, in the Buddhist temples, that is, and in India, where,


as we have seen, the services of barbers, washermen, etc., are
essential for certain ceremonial.
A somewhat similar parallel is to be found in Fiji, where there
are chieftains each associated withvhis own clan and each at the
same time an ofl&cial at the court of the High Chief. Each clan
has a specific function to perform in regard to the society of
the whole group. The clansmen of the Chief of the Land, for
instance, function as a sort of police; those of the Lord of the
Village Green preside over public feasts and allot the portions of

food; the herald clanmay also be the one which dresses the
High Chief's head in life, watches over his corpse and buries
him when dead; sometimes there is a crier clan which proclaims
the Chiefs wishes; sometimes even chieftains of the dogs ',
*

with clansmen who are the High 'Chiefs dogs, follow him
about, and bite the disrespectful. These castes have nothing
to do with specialization of crafts, but are graded according to
their function. There are what appear to be specialized manual
workers forming castes in some tribes, but they appear to be of

foreign extraction and are associated with the service of chiefs


only. They do not work for pay, are technically superfluous,
and are required only for the ritual or quasi-ritual service of
the High Chief or of the god of some temple. A great chief
may thus have a clan of King's Carpenters who make his official
canoes. These sacred canoes are built with a succession of
ceremonies unnecessary to the building of ordinary canoes, but
who know the rites. Similarly,
requiring the services of builders
the King's Fishermen undertake the ritual fishing for turtle
required for a special feast, and their chief is in charge of the
ritual which ensures that turtle and fish shall be plentiful.^
Ofl&ces and functions of this kind, hereditary in certain families

Hocart, Les Castes, p. 11). Since the publication of the second edition
ot Caste in India, a detailed account and penetrating analysis of caste
in Ceylon has been made by Bruce Ryan in his Caste and Modern Ceylon,
Hocart, Kings and Councillors, pp, 105 sqq.; Les Castes, pp. 140 sqq.
ANALOGOUS INSTITUTIONS ELSEWHERE 139

or dans, are or have been of course a fairly familiar feature in


many countries. Meek records functional groups of this kind

in Africa the Ba-Nando dan of the Jukun of Nigeria who
perform the burial rites of the kings of Wukari,^^ the Ba-t6vi,
who are the royal grooms.^* The hereditary ofiGices held about
the royal person in medieval courts in Europe were not very
dissimilar, for the office of Royal Butler, Constable of the Royal
Household, Master of the Buckhounds, King's Champion and
so forth became hereditary in certain families. Obviously this
hereditary functionalism does not constitute caste. But we do
find cases in Africa in which taboos associated with occupation
have operated to prevent not merely intermarriage but any
sexual relationship at all between pastoral and cultivating
groups.^
It is a not infrequent error, too, among Europeans, to suppose
that differences of caste are analogous to social differences in
Europe. Molony puts the matter in a nutshell when he points
out that in England the marriage of the daughter of the house
to the cook would be a mesalliance on her part, whereas it might
very well be in India that in the case of a ruling family of lower
caste employing a Brahman cook the discredit of such a union,
which could scarcely be styled a marriage , would fall on the
'

cook.^ In Iny case the essentials of the caste system are as


absent from the hereditary functions of the. Earl Marshal, or
from the families of Butler or Dymoke, as they are from the
functional village which affords a somewhat similar analogy to
certain aspects of caste. The functional village is not unknown
in India. References occur in the Jatakas to functional villages
of potters, smiths, wheelwrights, etc., and such communities may
be found even surviving today. Thus the Maharaja of Manipur
has a village which has the specific duty of providing him with
boats,^^ like the King's Carpenters in Fiji. But although the
Maharaja and his Meithei compatriots are Hindus, the boat-
making village is a Loi village, that is to say, it is a village of
one section of his people who have not been received into a
Hindu caste. An almost closer parallel to caste is to be found in
the medieval practice of segregating tanners on account of their

ASudanese Kingdom, pp. 50, 170. 1* Ibid., pp. 149, 338.


Roscoe, Immigrants and their Influence in the Lake Region of Central
Africa, p. 34. A Book of South India, p. 107.
21 Hodson, The Meitheis, p- 33.
malodorous occupation. Medieval guilds, too, with their craft
secrets and their hereditary nature
^for it was much harder for

a stranger to enter them than for a son of a member* offer


an obvious parallel to the functional caste, even though it is very
clearly incomplete. Another incomplete medieval parallel is to
be found in the case of some German executioners the son was :

compelled to succeed to his fathers ofi6ce, the holder of which


was so shunned that he had difficulty in getting priests to give
him the sacrament, or a fellow-citizen to let him marry his
daughter; the latter difficulty was met by granting him the right
to claim as a wife a female sent for execution.
The existence of a caste system in ancient Egypt has been
recorded by a number of writers, and
has been suggested that
it
here was to be found the nearest analogy to the caste system
of India. Herodotus tells us that there were seven classes
or
rather ' clans (genea) of the population in Egypt priests,

ing men, herdsmen, swineherds, tradesmen, interpreters,


fight-
navi-
gators.* The fighting men were divided into two categories
territorially, were not allowed to learn or practise
any craft or
trade, but must devote themselves entirely to military
pursuits
from generation to generation, the son succeeding to the father.
In regard the swineherds we are told that the pig was
an
impure (mioroj) animal, mere contact with which made
it
necessary to bathe; the swineherds, although native-born
could not enter any temple, and they married among
themselves, as no one else would be willing either to give a
daughter to one of them in marriage, or to receive a
swineherds
daughter as a wife.* The priests were also hereditary,**
and
we learn amongother things that they bathed themselves in
cold water four tiroes every twenty-four hours;
shaved themselves
completely except when in mourning; drank universally from
brass vessels
which they scoured daily; and regarded beans
as unclean (ou katharon) food.** When
one reads further of
the great reverence of the Egyptians for the
cow, one inevitably
recalls Brahmanism, in spite of the fact that male cattle could

M P Herodotus, n, 164.
* Cf. Stevenson, Rites
of the Twice-Born, p. *40.
brass, or at any rate metal, dishes. ^ Brahinans
.nranmans must use

edition of Kutorch's flomene


jSnsfpp.'tevrsq5"
ANALOGOUS INSTITUTIONS ELSEWHERE 141

be sacrificed (as indeed in vedic ritual). 1 he parallel with caste


is further emphasized by the statement that no Egyptian man
or woman Greek on the "'mouth, or use the knife,
will kiss a
spit, or cauldron of a Greek or taste the flesh of a clean (katharos)

ox if cut with a Greek knife.^ Strabo,^ it is true, mentions


only three but Diodorus records five priests,
classes of Egyptians,

soldiers, shepherds,husbandmen, and artificers and states that
these divisions are hereditary and compulsory.^ Plato (Timaeus,


24 A, b) mentions six classes ^priests, craftsmen (who are sub-

divided 'each keeps to its own craft without infringing on
another *), shepherds, hunters, farmers, and soldiers. It all
sounds as though a caste system not unlike that of India may
have existed in ancient Egypt. Revillout has gone into this
question carefully in his Cours de Droit dgyptien. He
instances the cases of a family of architects to the crown, so
to speak, who for several hundred years and under all the later
dynasties continued to exercise their hereditary calling, and of
a family of mummy-wardens in Thebes who followed the same
calling from 680 b.c. down to the Roman occupation. But he
comes to the definite finding that whatever the nature of these
so-called Egyptian castes there is nothing to show that there
'
'

was any caste system which really resembled that of India, nor
anything in the customary laws of Egypt which interfered with
social intercourse between these groups or prevented their inter-
marriage (no doubt with the possible exception of the case of
the pig-keepers), instances of which are known to have taken
place.^ Indeed, in so far as there seem to have been groups
analogous to caste in Egypt, they seem to have been rather of
the nature of administrative organizations like those of the
later Roman Empire, than of quasi-organic development like

Herodotus, n, 41. 29 xvn, 787.


Shepherds seem also to have been tabooed like the swineherds whom
so
Diodorus does not mention, for when Joseph's family came down into
Egypt he told them to tell Pharaoh that they were herdsmen of cattle for

every shepherd is an abomination unto the Egyptians (Genesis xlvi, 34).


'

Book I, ch. VI. 32 Cornford, Plato's Cosmology, p. 17.


33
p. 136.
34 ^ Sil existait des castes en Egypte, dvidemment elles n'dtaient pas imitdes
de celles de Flnde, et nous trouverions plutdt leurs analogues dans les
classes dtablies, daprds danciennes traditions locales par les lois [romains]
du Bas-Empire. Or, dans ces lois rien ninterdit les rapports sociaux et
mdme le manage avec des personnes appartenant k une autre caste.*
'

Revillout, Cours de Droit dgyptien, p. 135.


33 Revillout, op. dt., p. 141.

142 ORIGINS
castes in India, a view supported incidentally by Diodoruss state-
ment already quoted. At the same time there is a passage in
the book of Genesis which states that when Josephs brethren
came down into Egypt and were entertained by him incognito,
food was served separately for Joseph, for his brothers, and for

the Egyptians, which did eat with him, by themselves: because


the Egyptians might not eat bread with the Hebrews; for that
is an abomination unto the Egyptians.^ This is again dis-
tinctly suggestive of caste.
Some other analogies are to be found in modern Africa.^
Thus among the Somali of the East Horn there are certain out-
caste classes. One of these, as perhaps might be expected since
the craft is so often treated as taboo, is that of the blacksmiths
Tomal; another is the class known as Yebir, who* are credited
with supernatural powers, and feared accordingly, and are able
on the strength of that to levy tribute from Somali families
on occasions of birth and marriage. The third are tanners and
hunters known as Midgan, who are regarded as so unclean that
marriage with them is tabooed, and the Somali who breaks this
taboo is degraded to the class of Tomal. Similarly among the
Masai there is a despised tribe of hunters known as Wandorobo,
and a completely tabooed class of blacksmiths on which Wando-
robo as well asMasai look down. Although a Masai proper
may act as a blacksmith without suffering serious loss of status,
the hereditary blacksmiths live apart and a blacksmith may
be killed with impunity by a Masai (but not a Masai by a
blacksmith) and no Masai would stop at a blacksmiths encamp-
ment, nor marry a blacksmiths daughter; his products are
impure and must be purified with grease before use, and even
the very name of blacksmith must not be uttered at night lest

lions attack the camp.


In the Ruanda and Urundi regions adjoining Lake Kivu and
the eastern Congo a somewhat similar condition has been
observed even more suggestive of caste. Three racially and eco-
nomically distinct groups occupy co-extensive areas the pastoral
Tussi (or Tutsi), the Hutu, who are cultivators, and the Twa,
a hybrid pygmy tribe, who are hunters. With the latter neither
Hutu nor Tussi will intermarry or contract a blood brotherhood,
though a Tussi is not absolutely barred from such relations with
3 Book of Genesis, xliii, 37 See Lowie, Ongin of the State^ ch. ii.
A N A L 0 GOUS INSTITUTIONS ELSEWHERE 143

a Hutu. The Twa themselves, though despising the agricultural


Hutu (and likew^ise that section of their fellow tribesmen which
has given up^hunting for the setded life of professional potters),
acknowledge the superiority of certain Hutu chiefs and pay
tribute to the Tussi king through them. The Tussi look down
on the Hutu for eating mutton and goat-flesh, and both Tussi
and Hutu despise the Twa for eating poultry and eggs. The
Twa act as executioners and as the police force, carry the king
in his travels, and chant in royal processions, thus recalling the
Dorns of India who, are commonly minstrels as well as execu-
tioners, in spite of being general scavengers of the lowest social
scale, and who are also nearly allied to a number of hunting
or trapping castes. These instances, however, though analogous
to caste in some directions do not constitute a caste system, and
in Ruanda there is a secret religious society into which all three
classes mentioned are admitted. At the same time they may
be of much significance in regard to the origin of caste in India.
On the other side of Africa an analogy to caste is afforded
by the osu in Ibo society. The osu are the descendants of free-
born ancestors who were bought by a family or by an individual
at the instance of a diviner and offered as slaves to some deity
who could not be by any smaller offering such as that
satisfied
of a fowl or a goat. Osu became a class apart and sexual rela-
tions of any sort between them and the free Ibo are tabooed.
There are other restrictions on their intercourse with the free
Ibo; their houses are segregated; and to call anyone an osu
is a gross insult. The free Ibo regard the osu with repugnance,
and the explanation given is that they would in the old days have
been in fact sacrificed to the deity to whom they are now offered
as slaves, and are therefore ritually dead. One is reminded
not only of the pagoda slaves in Burma, a very close parallel
mentioned later in this chapter,^ but also of the outcastes
rejected by the Ganges.**^
To return to Asia, an analogy to caste, which is perhaps the
most significant of any of them, is to be seen in that
country which marches most closely with India, that is, in Burma.
And this analogy is important, for no one who knows both India
and Burma would for an instant go so far as to hold that the
38 Lowie, loc. cit., p. 35^. sa Green, Ibo Village Affairs, pp. 23, 24, 50.
^0 Fide in/ra, pp. 144 sqq. -*1 Fide swpm, pp. 109-10.
ORIGINS
there at all. Yet the Burmese
caste system, of India obtained
essence of caste as any
analogue is possibly as near to the
parallel elsewhere, and it does,
perhaps, throw an important

light on one of the origins,


perhaps the most important and
It may be noted in
primary origin, of that unique institution.
connexion that Burma was on the whole even
less affected
this
therefore to conserve
than India by external contacts and likely
form than India.
customs and beliefs in a more primitive
while Burma received Buddhism from India
when it
Further,
was there predominant, Buddhism remained as a permanent
this

part of its culture and there was no development of Brahmanism


as in India.
monarchy, seven
In Burma under the old rule of the Burmese
distinct classes of outcaste were recognized.
The rnost important
is regarded as
of these was probably that of Pagoda Slaves. It ,

very meritorious in Burma to build or repair a pagoda or even

to gild or regild a few square feet of its roof. But no such merit

attaches to the service performed by the slaves who keep the

pagoda in order. A
pagoda slave is such for and his
life, child-

ren and descendants are pagoda slaves in perpetuum: they

cannot be liberated even by a king. If a person who


is not a

pagoda slave marry or be married to a pagoda slave, even un-

wittingly, such a person and all her or his children, even by a

former marriage, also become automatically pagoda slaves in


perpetuity. Pagoda slaves cannot be employed in any
other

capacity than that of pagoda servant. It will be observed that

in the last two respects the disabilities suffered are even more
severe than those of outcastes in India, though the element of
untouchability is not stressed in at all the same degree. The
pagoda slaves were largely drawn from prisoners of war, and
from prisoners convicted of some offence, but also from quite
inoffensive Burmese villagers, who had been nominated for the
service by some royal official and were too poor to buy them-
selves off. The other six outcaste classes consisted of, secondly.
Professional Beggars, that is, of vagabonds compelled to live
such a life and prohibited from any regular occupation, drawn
in the main from the same sources as the first dass and to be
distinguished from free beggars, who beg by choice and not by
compision; thirdly. Executioners, Jailors and Police, three
functions normally combined, under the rule of the Burmese
ANALOGOUS INSTITUTIONS ELSEWHERE 145
kings, in the same persons, who though
feared, no doubt, when
alive, and treated accordingly with circumspection, were treated
as offal when dead; fourthly, Lepers, and others suffering
from incurable diseases, who were compelled to live with other
outcastes outside the ordinary village
community; fifthly.
Deformed and Mutilated Persons, the horror of
the maimed in
Burma being very strong, so that death is commonly preferred
to the amputation of a limb. The sixth
class of outcastes con-
sists of Coffin-makers and
other persons occupied with grave-
digging, graveyards and the disposal
of corpses; the Burmese
word for such people is sandala, which is clearly the same word
as the Hindi chandal. The seventh class consisted of Govern-
ment Slaves, who were cultivators of the royal lands and drawn
from the same sources as pagoda slaves.
Persons of these outcaste classes in Burma
cannot enter a
monastery nor become a hpungyi (Buddhist monk). Indeed,
they may not enter a monastery even to study,
though such an
entry forms a normal part of the upbringing of
a Burmese child.
The effect of marrying persons of these classes
has already been
stated as regards the pagoda slaves, and is the
same in the case
of the second class and of -the last two classes.
How far it would
affect the other three classes it is a little
difficult to say, as the
third class at any rate presumably disappeared
as such with the
British occupation. Up to that time at least
these seven classes
constituted an outcaste population distinct from the people with
whom social relations were possible.^2 constitute
castes, so that there were, so to speak, only incaste and outcaste
in Burma, and no '
caste system but the analogy to one aspect
of the caste system in any case is clear enough.
The untouch-
ability in Burma is obviously based on taboo.
This element
can be seen to be present in most if not all of the seven groups
above mentioned. In the case of the grave-diggers it is probably
the fear of death infection, a fear which is associated with the
disposal of the dead among so many of the neighbouring hill
tribes. A similar taboo probably operates in regard to lepers
and deformed persons, for among the Chin and Kuki tribes on
the western frontier of Burma, tribes nearly allied to the Burmese
in language and probably of the same racial stock, persons dying

>2 $uway Yoe [Sir J. G. Scott], The Burman ; His Life and Notions, ri,
ca; XVI.
10
,46
ORIGINS
of leprosy ox from a.ny similEX
deforming disease are buried witb
orientation from otber persons and regarded as
having
j,

death, which apparently makes it impossible, or


died a bad
at least undesirable, that they
should rejoin their families in the
future life." In the case of the Thado
Kuki, to have the lobe

of the ear split is enough to prevent a woman so mutilated from


joining her kindred in the after life. The
taboo on executioners
Buddhist prohibition on the
is no doubt associated with
the
takin life as well as on dislike and contempt for persons
g of
slaves might
following such a calling, while that on pagoda
that they are persons set aside
naturally follow from the fact
as belonging entirely to the deity. It is typical of taboo that

the emotions evokes are ambivalent, and there is no dear line


it

of distinction between the sacred and the accursed.


What is
holy may be either, but it is certain that it is dangerous. The
contagion by which their state infects any who might associate
with them with the same disability is also typical of primitive
doctrines of taboo. The untouchable condition of pagoda slaves
presumably springs from the fact that they are dedicated through-
out eternity to the pagoda or to the deity and have no separate
life of their own, being as it were part of the deity or of his

shrine. A similar untouchability might extend by analogy to


vagabonds and to the cultivators of royal lands, who were
drawn like the pagoda slaves mainly from prisoners of war
and from convicted criminals. In theory all the misfortunes
suffered by these dasses are the result of the working of
karma;** the unfortunates who composed them are (or
were) in a sense subhuman, they form a link between man
and the animals, housing souls suffering in the highest
grade of punishment for faults in former lives. Even such
taxes as were levied on them constituted vile and contaminated
funds, which were used to provide food and accommodation for
European foreigners as a subtle insult to the unconscious reci-
pients. This doctrine of taboo on which, it is urged, untouch-
ability in Burma is based, probably precedes in that country the
influence of Buddhist culture and! is rather a survival, like the
cults of the nats, from a pre-Buddhistic era, from a period of

more primitive bdiefs which India no doubt shared before it

*3 See Parry, The Lakhers, pp. 406-8 and 413, and Shz'w, Notes on the
^4 yide supra,
Thadou Kukis, p, 1, . isi sq.
ANALOGOUS INSTITUTIONS ELSEWHERE 147

became Hindu at all. There seems to be common ground here


for an origin of caste which, while developing in India into
an organic structure of society, has in Burma become stabilized
in an undeveloped form or even degenerated so as to afEect only
a limited part of society, and leaving the main body of the people
untouched. For the Burmese as a whole are as free from the
working of the caste system as the other peoples among whom
analogous institutions have been pointed out.
Analogous to the sandala of Burma are the eta of Japan.
Although no doubt primitive taboos have contributed to their
status of untouchability, it is perhaps in this case to be asso-
ciated primarily with the introduction of Buddhism, and its
prohibition on the taking of life, together with influences derived
directly from the caste system of India. The eta .form a com-
munity of outcastes, said to be about a million in number, who
live on the fringe of Japanese society. So strong is the prejudice
against them that the very word eta, if it must be uttered, is
only whispered. The name seems to have appeared first about
A.D. 700 and to be associated with the introduction of Buddhism.

Outcaste professions included executioners, butchers, leather-


workers, saddlers, cobblers, actors, jugglers, grave-diggers, mid-
wives, umbrella- and lantern-makers, because they handled

oiled
paper ', and, as in India in Buddhist times, when they were
despised occupations, basket-makers and potters. They were
considered subhuman; numbered with the termination -hiki
used for quadrupeds; lived in separate quarters in the village;
had to wear distinctive dress; could only marry among them-
selves; had no social intercourse with other classes, and could
only go abroad between sunset and sunrise. * In the small fish-
ing village of Mihoroseki ... nothing distinguished the eta
houses in appearance or marked the boundaiy of the outcast
community. Yet the children on either side never crossed an
imaginary line which marked the frontier of untouchability
halfway up the street.^ In 1871 the then Japanese Govern-
ment abolished all feudal discrimination, but the effect on the
eta was in some ways to enhance their depression, for they lost
their caste autonomy and such privileges as they had, and is
particular the monopoly of their special occupations. In cities

^sThe Tokyo correspondent of The Times writing in that paper on


9 September 1958.
148 ORIGINS
and towns it is no doubt possible for a stranger of eta origin

to conceal it, where it is known, he is discriminated against


but,
at school, in employment, in trade, and in marriage; eta tenant
farmers failed to obtain the benefit due to them under the post-

war land reforms, and the gulf between them and the ordinary
population has yet to be bridged. Their position is not without
analogies to that of the exterior castes of Hinduism, but
probably both here and in Burma what there is of caste is
closer to the Ceylon than to the Indian pattern. This is what
might naturally be expected, for Japan was once no less
Buddhist than Burma and Ceylon.
Indeed it is possible that light on the origin of the caste
system is really to be sought for in societies such as those of
Burma, Ceylon, and Japan, where traces of the Buddhistic
regime have survived unHinduized. Buddhism and Jainism
have much in common, and Professor Tucci has argued that
the Jain religion embodies a revival of very ancient rituals and
forms * probably even pre-Aryan The Buddhist movement of
the sixth century b c was a revolt against Brahmanism and
. .

probably in part at any rate an attempt to revert to an earlier


order. If that be so, it might well be that Buddhist societies
incorporated, and have retained, some form of caste earlier in
pattern than its systemization by Hindu priests and law-givers.
One recalls the fact that the first historical Tirth&nk&ra, Maha-
regarded by Jains as the twenty-third in succession, and
vira, is
that Parswa, who preceded him, is described as the son of a
King Aswasena, of Benares, though no Aswasena is known to
Brahmanical literature except a king of the Nagas We

should perhaps therefore be justified on these grounds alone in


seeking the ultimate origins of the caste system in pre-Rigvedic
India.
Manyother analogues no doubt exist in various parts of the
world. Those that have been mentioned in this chapter merely
afford a few instances of phenomena elsewhere that bear some
resemblance to caste. It is on the face of it unlikely that a
combination of many of them would occur more than once, and
it is not therefore surprising that we find a caste system in
India alone.
CHAPTER X
the traditional origin and its
IMPLICATIONS
According to Hindu tradition the caste system owes its origin
to the four varna,^ which are respectively derived from the
Brahman who sprang from the mouth of the deity, the Kshatriya
who was who was formed
created from his arms, the Vaishya
from and the Sudra who was born from his feet. To
his thighs,
the Brahmans were assigned divinity and the six duties of
studying, teaching, sacrificing, assisting others to sacrifice, giving
alms and receiving gifts to the end that the vedas might be pro-
tected; to the Kshatriya were assigned strength and the duties
of studying, sacrificing, giving alms, using weapons, protecting
treasure and life, to the end that good government should be
assured; to the Vaishya were allotted the power of work and
the duties of studying, sacrificing, giving alms, cultivating, trad-
ing and tending cattle, to the end that labour should be pro-
ductive; and Sudra was given the duty of serving the
to the
three higher varnaJ^ From mixed unions between the different
varna, or between the varna and the offspring of mixed unions
and so forth,^ come the various castes of which 2378 were actually
counted at the Indian census of 1901, some numbering millions,
others only a very few individuals. To say that there are some
3000 diflEerent castes in India is probably to run little risk of
exaggeration. The general nature of hypergamy and the mean-
ing of the terms anuloma and pratiloma have already been
explained. Marriages between members of different varna are
postulated in this traditional account, and these marriages were
either anuloma or pratiloma. Such unions, and the subsequent
ones between the mixed oflEspring of such unions, are held to
account for the various and the Code of Manu and the
castes,
treatises of Vasistha, Baudhyana, Gautama and others give us
lists of the castes so formed. Wilson has worked out a list
of 134 castes, including the four varna, in order of precedence
1 Vide supra, pp. 64 sqq.
^ Btihler, Laws of Manu (i, 31, 87), pp. 14, 34 ; Sacred Laws, pt. 11, p. 25
(Vasistha rv, 2), etc., and wd(? Wilson, Indian Caste, i, p. 17*
Biihkr, Sacred Laws, pt. ii, p. 199 ; Jolly, Institutes of Vishnu, p. 12.
^ Mann, x, 6 sqq. (Biihkr, Laws of Manu, pp. 403 sqq.).
Vide supra,
p. 55. Indian Caste, i, pp. 65 sqq.
J 50 ORIGINS
asmentioned in the ancient law books. In general principle the
pratiloma castes rank below the anuloma, and at the bottom
of the scale are the Chandala, begotten of a Sudra male on a
Brahman female, and other mixed castes of Chandal parentage.
The given by the lawgivers of the origins and functions of
lists

the various mixed castes do not always agree among themselves


and suggest very strongly a formalist interpretation of a state
of society the origins of which were already obscure when these
treatises were composed. Senart has made out a very strong
case for the view that the Code of Manu
an interpretation
is

in terms of varna, that is, in terms of an Indo-European social


system analogous to that of ancient Rome, of an existing social
order which the terms could not, at any rate at the time of
application, be made to fit satisfactorily. In any case one of the
questions raised by Manus scheme of precedence is the reason
why a caste of mixed origins should, if pratiloma, be so much less
respectable than an anuloma caste of inferior derivation. The
offspring of a Sudra man and a Sudra woman might, on the
face of it, be expected to rank lower than that of a Sudra man
and a Brahman woman. If we fall back on the principle of
hypergamy as an explanation, we are really no better off, as
some explanation is equally required for the excessive applica-
tion of such a principle, if not for its actual existence. It is
possible that the explanation is to be found in the results of the
impact of patrilineal invaders on an indigenous matrilineal popu-
lation. In such a case a condition of society would arise in
which the offspring of a woman of the patrilineal immigrants
by a man of the indigenous matrilineal peoples held an ano-
malous and ambiguous position. In the case of a patrilineal
father and a matrilineal mother the child would have status
and kinship with both his father's and his mother's families and
would inherit from both, or, whether the marriage were patri-
local or matrilocal, at least from one of them. But the issue
of an immigrant patrilineal woman by an indigenous matrilineal
male would have no place with either. He could claim no kin-
ship through his mother with her exogamous patrilineal dan,
nor through his father with his exogamous matrilineal family.
With no claims on kinship or inheritance his status would in-
evitably tend to be degraded, and the exact causes of this would

Le$ Castes dans VInde, di. lu


TRADITIONAL ORIGIN 151

very soon be forgotten when both adopted the patri-


societies
lineal system, but would formulae
call for expression in definite
at whatever date an attempt was made to record in the form of
a code the state of society as seen by the early lawgiver. There
are one or two passages in the Sutras which distinctly suggest
such an interpretation. Thus Gautama (iv, 25, 26) tells us
that those born in the inverse order (from fathers of a lower
*

and mothers of a higher caste) stand outside (the pale of) the
sacred law as well as those born in a regular order (anuloma)
from a Sudra woman, and the Institutes of Vishnu (xv, 37, 38)
tell us that children begotten by a husband of inferior caste on
a woman of a higher caste have no right to inherit, and *
their
sons do not even receive a share of the wealth of their paternal
grandfathers In passing, it is perhaps significant that the
Suta, who is the pratiloma offspring of a Brahman woman by a
Kshatriya man, takes a fairly high place in Manu's list, far
removed from the degraded position of a Chandal who is by a
Sudra out of a Brahmani. This would reflect, perhaps, the fact
that as between varna which were both patrilineal the question
of complete lack of status and inheritance would not arise, and
there seem to have been Brahman families both of invading and
of indigenous stock. It must also be pointed out here that
statements as to what constitutes a Chandal are not always
consistent, for there appear to have been Chandals by works
as well as by birth.
Besides the suggestion made above there are a number of other
survivals which suggest that a matrilineal system w,as once much
more widely distributed in India than it is now. Sir Denis
Brays account of the Brahui shows that the traces of a former
matrilineal system are clear enough; and the story recorded by
Lieut. Carloss of his visit to the cave dwellings of Gondrani
in 1838 March 1839), when he was shown the palace
of Badi-ul-Jamal,^ the princess married by the stranger, a son *

of the King of Egypt who became king of the matrilineal and


matrilocal kingdom, is exceedingly suggestive of Mediterranean
aflSnities, while the Shrine of Bibi Nani near Hinglaj, also in Las

Bela, is a place of pilgrimage * celebrated from the Euphrates


to the Ganges and the resort alike of Muslim and Hindu
pilgrims.^^ Sir Thomas Holdich says of it, the object of their

Biihlcr, Sacred Laws, pt. i, p. 197. Jolly, op. cit,, pp. 64, 65.
10 Minchin, Xfltf Gazetteer, p, 40, 11 Ibid., pp. 35 sq.
152 ORIGINS
venerationis probably the same goddess who was known to the

Chaldeans under the same old-world name (Nana) a thousand


years before the time of Abraham. Nothing testifies so strongly
unchangeable nature of the geographical link formed by
to the
Makrin between East and West than (sic) does this remarkable
ziarathidden away in the deep folds of the Maldn mountains/
Some of the features of south Indian culture suggest not merely
Asia Minor (e.g. the fire-walking ceremonies such as are common
in southern India and formed a feature of the worship of
Artemis in Cappadocia), but Crete. Here there was a cult of
snakes/3 which, indeed, were vehicles of the soul throughout
Greece, and worship of the mother goddess; here again the
vogue of slender waists for males is very suggestive of medieval
Indian sculpture. The popular sport among the Kalian and
Maravan of jumping on to enraged bulls with sharpened horns
to pluck off a cloth put there for the purpose and prove them-
selves men in the eyes of their womenfolk is also most remi-
niscent not only of the bull-baiting of Provence, where a rosette
must be snatched from the points of the horns of an infuriated
bull, but likewise of the bull-jumping scenes on Cretan vases,
though in southern India the practice does not extend to
somersaults nor to the fairer sex. Terracotta figurines of the
mother goddess have been excavated in Crete not dissimilar, it
seems, to those from Mohenjodafo, and to others excavated in
1926-7 at Buxar in Bihar from a site 52 feet below the present
surface and 13 feet below the Maurya stratum,^ which rather

12 Haldich, India, p. 45. Nana or Nanai was the mother of Attis and
identified with Ishtar, Astarte, Artemis, Anaitis, or Aphrodite. Alone of
her pilgrims devotees and unmarried girls pay no tax to the State* Naina
Devi of the Kulu valley, where her image is a black stone 3 feet high, and
of places in Sirmur and Bilaspur States, and probably elsewhere in the
lower Himalayas, is likely to be the same goddess.
13 The cult of snakes is also strong in the lower Himalayas and is found

there in the significant company erf Naina Devi, megalithic monuments, and
marriage customs not unsuggestive of Babylon, and of fertility rites which
are spoken of by Rose in Punjab Castes and Tribes as * Paphian * but
which are not described, Devi in the hills is often spoken of as Devi Mai

or Devi Mata the Mother Goddess.
14 In the case of the Kalian it is to be noticed that they include a hoorae-
fang among their wedding gifts, practise circumcision, and bury their dead
and perform their Karuppan worship with the face to the north, probably
indicating migration into India by land. The rite of circumcision is paid
for by the boy's fathers sister, mother of his potential wife.
13 Banerji-Sastri,Mother-Goddess cult

(anniversary number, 1939).


m Magadha '
in The Searchlight
traditional origin 153
extension of the Indus valley culture to the Ganges
valley. The cult of the bull is common to the early cultures
of Crete, of Egypt, of the Near East, of the Indus valley and of
Hindu and it may here be pointed out that Przyluski
India,
suggests anon-Indo-European origin for the name of the god
Vishnu and a Dravidian origin for* the god himself.
The culture of the early civilization of northern India may
perhaps be most conveniently described as pre-Rigvedic Hindu-
ism. Even if this culture disappeared entirely from the Indus
valley, it may well have survived across the Jumna with sufficient
vigour to react to the Rigvedic Aryans whose religious beliefs it
ultimately submerged in its own philosophy. Slater has aptly
pointed out that Krishna himself was of Naga descent, and the
traditional blue colour in which Hindu art depicts him possibly
represents the brunette colouring of the indigenes as distin-
guished on the one hand from the blond Aryans and on the
other from the dusky aboriginals. Slater again points out that
Sakra, the chief priest of the Asuras, is stated by the Maha-
bharata to have become the spiritual guide of both the Daityas
and the Devas thus recording the success of the pre-Rigvedic
priestly class inimposing their spiritual authority on the Aryans
also, and this same Sakra, or another one, according to the
Vishnu Purana, said mantras iox the success of the Asuras and
restored to life the Danavas slain by Indra. His father was the
rishi Bhrigti whose sons were Brahmans and priests of the
Daityas. Clearly there were Brahmans before the Rigvedic
Aryans, and we must look for the origin of that caste partly no
doubt in the priests of the presumably Dravidian-speaking
civilization who may well have shared the mathematical and
astronomical knowledge of contemporary Babylonia.
It seems generally to happen that people with a matrilineal
system substitute a patrilineal one where the two systems come
into contact, and it is likely that the change from a matrilineal
to a patrilineal system started to take place in upper India as a
result of immigration, while it is not unlikely that the same
process tended to substitute the worship of male for female
deities. The practice of Hinduizing the female village deities
of southern India by providing them with orthodox male
Archw Orientdlni, tv,St August 193s, vide Indian Antiquary,
1933. Dravidian Element in Indian Culture, 55
ORIGINS
husbands from the official Hindu pantheon is still perhaps going
on. In Madura the fish-eyed goddess Mihakshi is annually so
married with great pomp and iclat, but in the villages the
goddess is still the real deity, and protectress of the people, rather

than the recognized Hindu gods. So also in Bengal the Dharma-


puja-paddhati records that Adya the mother of the gods was
married to Shiva with traditional ceremonies not enjoined in
'

the Shastras but somehow or other accepted as inviolable by


them and known as stri-achara (lit. female custom) '
performed
by women before the Brahmans officiated,^ and the same author
maintains elsewhere that goddesses have been accorded (in folk-
custom presumably) a higher position than the gods. According
to the Mahabharata a matrilineal system survived in medieval
India in the kingdom of Mahishmati about the Narbada river,
where the women had liberty to choose a plurality of husbands,
and among the Arattas, somewhere apparently in the Punjab,

whose heirs are their sisters children, not their own With
the exception of the Nambudris, who follow the Rigveda,
Brahmans in southern India, many of whom at any rate follow
the Samaveda, Yajurveda or Atharvaveda, are accustomed to
marry the daughter of their mothers brother. This is opposed
to the letter and spirit of the Brahmanic code and is clearly
suggestive of a survival of a matrilineal system; orthodoxy would
appear to enjoin the patrilineal prohibition of such marriages.
The Pandyan dynasty seems to have been originally matrilineal,
as Tamil poems are said to allude to its founder as a woman,
and the tradition recorded by Megasthenes^ is that it was
founded by a daughter of Heracles, while Pliny describes the
people as * gens Pandae, sola Indorum regnata f oeminis The
worship of goddesses in whose honour annual fairs are held is
more important in the Himalayas of the Punjab and United
Provinces than that of the orthodox gods, and such goddesses,
though now regarded as incarnations of Devi, are frequently
associated with the worship of snakes, while
it seems likely that

the Nanagotri Brahmans of Tehri-Garhwal represent families


which originally traced their descent through the female line
(nana= mothers brother).

Sztkzt, Folk Element in Hindu


Culture, p. 5*31,
McCrindle, Ancient as described by Megasthenes and Arrian,
J'ldia
pp. 37, jo6, and Pintiani^s note on Pliny, vii, U, m
the edition of
30 Nat Hist,^ vt, xx.
1^,
vol. I, pp. 40, 403.
TRADITIONAL ORIGIN 155

This pre-Rigvedic. culture seems to have still been vigorous


enough east of the river Jumna to react to the stimulus of the
Rigvedic Aryans, when these invaders, near cousins probably of
the Kassites who about the same time overthrew the rulers
of Babylon and established themselves as lords of Mitanni,
occupied the Punjab about the middle of the second millen-
nium B.c. Their previous traditions as gleaned from the Avesta
are conveniently summarized by Peake and Fleure.^^ Driven
from their northern home on account of its becoming ice-bound
and uninhabitable they moved south to Sughda (Sogdiana) and
Mura (Merv, Bokhara). Locusts and hostile tribes drove them
to Bakhdi (Balkh) whence they moved to Nisaya. There they
divided, one part going to Haroyu (Herat) and the other pro-
bably to Kabul and thence to the Punjab at a date not later
than 1400 b.c. and probably a century or so earlier. Their
occupation of the country between the Indus and the Jumna,
where the Rigveda seems to have been composed, but not
written, must have taken the form of migratory movement and
was probably effected without much diflSculty. The description
of their enemies as
noseless ' suggests conflict with tribes of
proto-Australoid and there seems every probability that
affinities,

tribes such as the Bhils and the Chodhras would have continued
to occupy hill and forest areas in spite of previous migrations
into northern India from the west and north. If the pre-Rigvedic
civilization in the Indus valley had really declined, this may
have enabled such tribes to reoccupy parts of the open country.
The mention, however, e.g. in the 104th hymn of the first book
of the Rigveda, of the cities, casdes and great wealth of an
enemy whose womenfolk bathed themselves in milk suggests
that the ancient civilization was far from being extinct, and it
is not impossible but that many aborigines were employed as

servants and as auxiliary troops or as allies against the invader,


a practice which was common enough in medieval India.
Alternatively, the description of the Dasyus as
noseless ' may it-

self have been a tradition taken over from previous invaders and
*
telescoped in the matter of time and association. The deve-
'

lopment of a written literature dearly took place at the second


stage of their invasion, described by Rapson^ as a colonizing

Merchant Venturers in Bronze, p. 130.


Cambridge History of India, t, p. 47.
156 ORIGINS
Stage in contrast to that o mass migration. It is jnst at
such a stage that amalgamation with the pre-existing inhabitants
is most likely, and the influence o the latter can clearly be

traced in the change in vedic religion which appears even in the


tenth book of the Rigveda, and in the Yajurveda, as well as in
the Atharvaveda, for this last consists, principally at any rate,

of magic, while the pantheistic philosophy later developed in the


Upanishads is already apparent in the tenth book of the Rigveda.
Pargiter points out that though the name of the compiler of
the Rigveda was well known to the later epic and puranic tradi-
tion, the very mention of Vyasa is ignored or suppressed in vedic
literature. It is possible to infer from this that the immediate
post-Rigvedic Brahmans may have been the inheritors of the
pre-Rigvedic tradition which was adverse to the Rigveda and
supported the indigenous deities by preference. It is possible

that the contest between Vasisthaand Viswamitra, as a result


of which the Kshatriya became a Brahman, may symbolize the
amalgamation of the two cultures, and Viswamitras formal
renunciation of Kshatriya ways the final ascendancy of the pre-
Aryan religion. In any case it is highly probable that the post-
Rigvedic literature, in which the tenth book of the Rigveda
must b : included, would contain importations from pre-
Rigvedic indigenous tradition, while the cerebral dental con-
sonants apparently acquired by the Indo-Aryan from the
indigenous languages * play an increasingly important part in
the development of Indo-Aryan in its subsequent phases "
just as
the use of rice and the areca nut, equally of pre-Aryan origin,
has affected the subsequent developments on the cultural side.
The development of the art of writing may have gone on simul-
taneously and have been, as suggested by Macdonell,^^ no doubt
first used for purely secular purposes and not at first regarded

as proper for application to religious hymns and formulae.


This would fit in with a derivation from the signs on the
Mohenjodaro seals which seem to have been primarily used for
commerce, as cotton fabric bearing a seal impressed with an
Indus valley stamp has been recovered from a prehistoric site

Ancient Indian Historical Tradition, '^, 9.


Rapson, Cu..%hridge History of India, i, p. 49.
India's Past, pp. 50 sqq.
traditional origin 157

in Iraq.^ It seems anlikely in the


extreme that the language,
religion and
culture of the invaders could have been influenced
by the pre-existing civilization without an admixture of race
taking place, and the fact that later tradition and literature defi-
nitely describesome of the rishis and their descendants as non-
Aryan indicates that this admixture extended to the priesthood/^
as it probably did to
all other classes. Thus the legends of the
origin of the Baidyas of Bengal, a caste not only of acknow-
ledged respectability and the repository of traditional medical
knowledge, but also one which provided a ruling dynasty in
the eleventh century a.d., have been justly interpreted as indi-

cating a matrilineal origin,^ while another version definitely


describes their descent as pratiloma in spite of its being ascribed
to the TwinBrethren of vedic mythology. It is tempting to
see in the two strains the origin of the two great Rajput houses
of the Sun and of the Moon, the latter typical of Mesopotamian
cults while the former is more suggestive of the Rigveda.^ The
Agnikula branch was probably added at a later date to include
the conquering families of Hun or Saka origin. On the other
hand, the Brahman in the Rigveda seems to have been second
in social importance to the Rajanya, and it seems not impossible
that traditions of the conflict between the Brahmans and the
Kshatriyas and the extermination of the latter by Parasurama
represent a revolt led by a priestly class of mixed origin, which
would naturally have the support of the people in. general,
against the Rigvedic aristocracy.
A further reflection of this amalgamation of two cultures, one
patrilineal, the other matrilineal, is possibly to be seen in the
employment by Hindus of the daughter's or sister's son or
husband in certain rites as an alternative' to the employment of
Sayce, Antiquity, June 9^7, p. 505 n., and see also Marshall, Mohenjo-
31

daro mdthe Indus CiviUzation, voL ii, pp. 380, 4^55 sq. Both quote Schexl,
Revue d* dssyriologief xxii (igas). . ,

*7 Some of the rishis actually appear to have had a sort of totemistic


origin ascribed to them : * Achela Muni was born of an elephant, and
Kesha Pingala of an owl, and Agastya Muni from the agasti flower, and
Karnika Muni from the kusha grass, and Kapila from a
(quoted by Wilson, Indian Caste, x, p. 597, from the Vaj'ra^ Shucm^a.
Buddhist sutra). 28 See Risley, Ethnographic Appendices,
uas
3 (5), 7 (1), and 9 (^) SJ
29 Ehrenfels [Mother-Right in India, chh. B
ingeniously argued a continuity of culture between the moou-descenaeci
Rajputs, the Ivayars, and the naatrilineal peoples of the Indus civilization.
igS ORIGINS
a Brahman.^* It is hardly necessary to point out that such
circumstances, under which patrilineal invaders took wives from
matrilineal indigenes, would also operate very strongly towards
the erection of apurdah system. The woman under the matri-
lineal system has a freedom not dissimilar to that of the man
under the patrilineal. The woman taken from a matrilineal
society and having ties of language, kinship, acquaintance and
custom with that society, but expected to live according to
strange and probably repugnant domestic and marital rules,
could only be ejBEectively restrained to that end by cutting off
her freedom of movement in and association with the society to
which she belonged. It may appear at first sight that the case
of a Nambudri Brahman cohabiting with a Nayar wife is a con-
trary instance, sin,ce she does not observe purdah at all, but
the fact that in this case the children follow the matrilineal
system supports the argument that purdah was necessary to the
combination of a patrilineal system with the practice of taking
wives from a matrilineal society. In the case of the younger
son of a Nambudri Brahman his Nayar wife continues to live,
and brings up her children, in her own house, in which her
husband is only a visitor. She rarely if ever visits his home,
and even then is much restricted in the parts of the building
to which she can penetrate. Only sometimes in royal house-
holds has the Nambudri husband quarters of his own. That
purdah should exist so strongly in the case of the Nambudri
wives of the elder sons may be explained by the necessity for
maintaining a barrier against the encroachment of a matrilineal
environment and by the possibility if not probability that the
Nambudri already practised purdah when they first arrived in
Malabar. That purdah system was alien to the Rigvedic
the
Aryans when they invaded India the Rigveda itself is a witness,
while there is nothing whatever to associate it with the
Dravidian-speaking stock, which seems to have followed the
matrilineal system in which purdah has no natural place at all.
perhaps significant in this connexion that purdah is gene-
It is
rallyweaker in the south of India than in the north, and in
Madras at any rate gets weaker from north to south, where,
conversely, caste gets stronger. It is conceivable that the same

Rose, Punjab Tribes and Castes, i, p. 39 ; Turner, of the


Z7.F,, p. 561. Some Brahmans' also recognize the sisters son as the family
priest. Wise, op. cit, p.
TRADITIONAL ORIGIN 159

circumstances gave rise to the practice of the pre-puberty


marriage of girls as to that of purdah, and infant marriage again
is least prevalent in the extreme south.
Various views have been held as to the origin of the custoln of
child marriage. It seems never to have extended to the Malabar
coast and is not nearly so prevalent in the extreme south, along
the east coast, or in the north-west as in the west, in Bengal and
in the Deccan. Clearly it is very ancient, and the suggestion that
attributes it Muslim invasions is not even momentarily
to the
entertainable. Megasthenes presumably refers to the practice
when he records that the girls of the Pandaian kingdom bear
children at the age of seven, adding that they are old at 40.^
Risley regards the custom as due primarily to hypergamy which,
by limiting the choice of bridegrooms, impels parents to marry
off their girls at the earliest age possible. Whether hypergamy
has or has not contributed to the establishment of the custom,
it might equally well have been due, as it almost certainly is

in one tribe in Assam, the Lhota Nagas, to a paucity of women


and the necessity of making sure of a wife before someone else
snapped her up, and not to a paucity of bridegrooms in a country
where males exceed females. Gait regards infant marriage as
the result of the impact of Aryan and Dravidian and a device
* '

to guard against pre-marital communism, and this again has


possibly been a contributing factor. More weight, however, than
was given by him seems due to the late Sir J. Campbell's opinion
quoted on p. 271 of his (Gait's) 1911 Census Report, Campbell's
opinion was that early marriage was due to a belief in the danger
of dying with unfulfilled wishes combined with the great wish of
a Hindu to marry and have children. As stated this is probably
the wrong way round. The wish to marry and have children
is the cause of the fear of dying unmarried. The penalty for
failure to marry is extinction in a future existence. This belief

McCrindie, Ancient India as described by Megasthenes and Arrian,


Frag, u(p. 115) ; Fliny, Nat, Hist,^ vii, ii. What is presumably the earliest
recorded age of pregnancy anywhere is one at 6 years which was observed at
the Victoria Zenana Hospital in Delhi in June 1935, when a girl who
had only been born on 8 October, 1955, gave birth to a child. The case
was investigated and established by Dr Kean, the woman physician in
charge of the case. Since first printing this note, however, a case has been
reported from Lima, Peru, of a girls having given birth at the age of
five years and eight months. At the time of the reports appearing in an
English newspaper both mother, aged 16, and child aged 10, were said to
be attending the same school (Cambridge Daily A'etus, 16 May 1949).
160 ORIGINS
appears in the tradition of primitive tribes from Assam to Fiji,

of which it has been said that


no unwedded has ever yet
spirit

reached the Elysium of Fiji Duarte Barbosa records of the


Nayars that the woman who does not go to
dies a virgin
paradise and some parts and castes
this idea is still so lively in

of India that the corpse of a person dying unmarried is married


before cremation.^ In China this feeling is so strong that the
very souls of the unmarried dead are married and given in
marriage post mortem, and that the idea is not absent in Europe
may be inferred from the belief once current in England that
maids who died unwed lead apes in hell

It would be rash

to say that a superstition of this kind was the cause of early


marriage, but it seems more than probable that the underlying
idea which imputes blame to failure to marry and propagate is
the same as that which enjoins fertilization at the earliest
opportunity.
On the Malabar coast, owing to the prevalence of the
marumakkathayam system,^ the position of women generally
has been better than in any other part of civilized India. The
Nambudri Brahmans, however, follow the makkathayam system,
and the great majority of Nambudri women must have been
condemned by ancient custom to remain unmarried, as the eldest
son only in a Nambudri family married a wife of his own caste.
The others marry Nayar women with whom they live in the
form of union known as sambandham, easily dissoluble by either
party. Recent legislation has now enabled the younger brothers
of a Nambudri family to contract marriages within their own
caste, and has given legal recognition and binding effect to
sambandham marriages was done in Cochin State by the
(as
Nayar Regulation of 1920), whether hypergamous or endoga-
mous, while securing to the wife the right of divorce and, with
her children, the right of maintenance by the husband, together
with the right to inherit a moiety of undisposed self -acquired
property left by the husband. By the customary law all respon-
sibility for the maintenance of the children lay with the woman*s
family, and any property acquired by the husband reverted to
his family at his death. Obviously modern conditions called for
some relaxation of this customary law, for the marumakkathayam
32 vV'ilHams, Fiji and the Fijians,
i, p. S44.
^^ Thuxstoiiy Ethnographic Notes,
pp, 105 sqq.; Mvtrs, The Todas,
701, 34 Vide
PP- p. ^0, supra.
;!

TRADITIONAL ORIGIN l6i

system, perhaps the best in the world when taken by itself, has
possibly in the past been exploited by the Nambudris to their
own profit. But it would be as great an error to replace the
marurnakkathayam system by the ordinary makkathayam one as ;

to suppose that the former is either primitive, barbarous or in i

any way less respectable than the latter, although it is less widely
distributed. On the contrary, it seems likely that the marumak-
kathayam was the ancient and civilized system, which was >

replaced in most countries of southern Europe and southern ^

Asia by the makkathayam under the stress of conquest by a ruder


people from the steppes of southern Russia.
Another possible survival of matrilineal or at least of matri-
local customs is to be found in the practice, widely spread in
India^ under which a man, known as a ghar-jawai that is, as
a house-son-in-law
* *

or some such equivalent term, goes to live
in the house of his prospective father-in-law. It is commonest

perhaps when the girls father is well-to-do and has no sons, in


which case the son-in-law is expected to take up residence there
permanently, but it also occurs in cases where the girls family
is very poor and requires the help of an able-bodied man, and

likewise where the bridegroom is very poor and unable to pay


a marriage price in cash. In such cases the normal period for
which he must work to obtain his wife seems to be three years,
;
during which time connubial intercourse is theoretically at any
rate not allowed to him, and though the prospective son-in-law
is maintained by his father-in-law-to-be he has no claim to inherit

in case of his death. This practice is known in the United Pro-


Vinces as Una marriage and is particularly associated with castes
'

derived from tribes, castes such as Chero and Kharwar, which are
probably Kolarian in origin, but also Gonds. Blunt mentions
the following other castes as using hina marriage: Bind, Ghasiya,
Majhwar, and Parahiya. Turner adds the names of some of
the Kanjar and Nat castes and mentions one division of Nats
among whom a bridegroom married in this way must stay and
serve as long as his wifes parents are alive, failing which he
becomes liable for a money payment, which is fixed by the
tribal panchayat to meet the case. As Turner points out,^ the
practice, in the case of a poor son-in-law who goes to live

ss Ibid.,
Census of the U.F., 1911, p. tso. 1931, pp. |ii sqq.
37 Loc. cit.
11
ORIGINS
permanently in the house of a rich father-in-law and is virtually
paid to become a ghar-jawai, is widespread and found in all
castes from Brahman downwards; he names a score of such castes
in the United Provinces in which either this practice has been
reported, or the analogous one of ghar baithna, where a man
goes to live in the house of a widow with no male collaterals, or
both. Marriage by exchange, either direct or (the more popular
form) three-cornered, is practised in the same areas, and in one
or two castes of Nats and Kanjars the custom of adopting a
daughter survives, though under strict Hindu law only a son
can be adopted.*
A probable survival of the matrilineal culture of south-west
Europe, south-west Asia and the Indian peninsula is to be found
in the practice in India of dedicating girls to the service of
the god in Hindu temples. Girls so dedicated are known as
devadasis and commonly live as prostitutes, but it has also to
be borne in mind that a life of immorality is not necessarily the
consequence of such dedication. In several castes in Madras,
particularly in Bellary and the neighbourhood, it is the practice,
if a male heir be wanting, to dedicate a daughter in the temple.

Thenceforward she becomes by established custom the heir to


her parents' property and can perform their funeral rites as
if she were a son. She takes to herself a mate of her own
selection of any equal or higher caste, but continues to live in
her father's house, and her children take his name and belong
to his family and not to their father's. If she has a son, he
inherits the property and continues the family, while if she has
a daughter only the daughter will in her turn become a basavi
and renew the attempt. No social stigma attaches to her, but
perhaps rather the contrary, for her presence at weddings is
auspicious, probably because she cannot become a widow.

Though not apparently relevant as traceable to matrilineal institutions,


an unusual item of marriage ritual reported by Turner in the same context
may perhaps be mentioned here in passing. He writes of the curious *

custom observed by Goriyas of the Gorakhpur district (who appear to be


the outcome of fusion between a sub-caste of Mallahs and a sub-caste of
Kahars). Before the rite of sendhurdan (marking the parting of the bride's
hair with red lead) is performed the bridegroom leaves the wedding party
in assumed umbrage and goes and sits on a roof erected specially for the
purpose. The bride then goes to him and entreats him to marry her
saying My lord, come and marry me. You need do no work as I will
work and earn money for you.'* The bridegroom and bride then descend
from the roof and the marriage proceeds. The origin of this peculiar
rite is unknown. . .
TRADITIONAL ORIGIN 163

Clearlyno immorality is inherent in this custom, which is merely


a method of temporarily reverting to the marumakkathayam
system when the family lacks a male heir and the makkathayam
system proves irksome or inadequate. It amounts in fact to a
method of adoption, but, being quite other than the orthodox
Hindu form of adoption which prevails in northern India,
su^ests an origin in the very different form of society which
stillsurvives on the Malabar coast, and was no doubt at one
time more widely prevalent.
The western Asiatic affinities of the devadasi custom of the
Tamilnad, for it is perhaps only there that the custom is still
general, are clear enough. Besides the familiar account given
by Herodotus of the oiffering made by women of their chastity in
the temple of Mylitta at Babylon, a reference to which custom
is made also in the 43rd verse of the Epistle of Jeremy in the

Book of the apocryphal Prophet Baruch, Lucian mentions the


same custom at Byblos in Phoenicia; there the goddess of a
temple, connected at a very early date with Egypt, was a fish
from the waist downwards and had by her temple a pool of
sacred fish with a stont chattri in its midst, and a woman had
the alternative of shaving her head and offering her hair*
instead of her chastity. Lucian, as a matter of fact, puts it the
other way: the women who do not chuse to be shaved are
obliged in lieu of it to expose their persons, and submit to the
embraces of strangers in the public market-place for hire, during
the sp5:e of one whole day; the money arising from it is
consecrated to the service of the goddess and expended on a
sacrifice to her.*^ Hair was offered by Argive maidens to
Athene before marriage, and those of Megara offered clippings
of their hair at the tomb and
of Iphinoe; in Delos both youths
maidens offered their hair before marriage at a tomb of corn-
maidens in the sanctuary of Artemis; hair was polled and offered
to the goddess of health at Titane near Corinth whose image
was swathed in strips of Babylonish raiment and it is still

shorn and offered to the Virgin at Tenos as a thanksgiving for


recovery from sickness; at Troezen every maiden before marriage
dedicated a lock of her. hair, in ibis case to. a male deity, in
the temple oi tilppoiytus. In ail these cases, with the possible

Gf. pp, sqq., m/m.


On the Syrim,. Goddess, in The Works of Ludm, tr. Francklin,/ vol.
IV, p.. 354^
I (4 ORIG I jXS
exception oJF that of the goddess of health, the association seems
to be between hair and fertility of the soil and of the body,
and it is probably the same association in India that makes
the Abbe Dubois remark on the commonness of the offering
of hair in temples by men and women in fulfilment of a vow/^
This practice obtains now even at the darga of the Mirgan Sahib
at Nagore in Tanjore district, where there is a tank, such as
those commonly found at Hindu temples, to which Hindu
women come to bathe who have vowed their hair to the Muslim
Saint, The element of a vow seems also to have entered into
the practice of sacred prostitution at Comana in Pontus, where
Strabo records that people assembled on account of vows made
to sacrifice to the goddess, describing the place as
full of women
trafficking in their persons, the most of them sacred that is,

to the moon goddess. Similar customs were observed at the


Phoenician colony of Sicca on the north African coast; at Helio-
polis in Syria likewise women prostituted themselves from reli-

gious motives; and in Armenia, at the temple of Anaitis, Strabo


tells us they dedicate male and female slaves, a fact in
that *

no way remarkable, but further the most illustrious of the people


dedicate their virgin daughters, who according to custom prosti-
tute themselves for a long period in the service of the goddess
and are afterwards given in marriage Similarly, it appears that
in some Indian temples it is usual for a devadasi to serve for
a period in the temple and thereafter be kept as a concubine
outside. Consecrated hetaerae were associated with the worship
of Aphrodite at Corinth, and an inscription from Tralles in
Lydia is referred to by Frazer which shows that the custom
obtained in Asia Minor as late as the second century Am, It '

records he says, of a certain woman, Aurelia Aemilia by name,


'

not only that she herself served the god in the capacity of a
harlot at his express command, but
that her mother and female
ancestors had done
same before her/ Strabo records a prac-
the
tice dE dedicating girls in the temple of Ammon in Egypt not
dissimilar to that in Armenia, except that in the case of Ammon
the deity was male and the dedicated girls, when given to a
human husband after serving their term as the wives of the
People of India, m, iii.
Kai allot de kat* euchm aei tines epidemousi, thusias te epitelountes
te theo . plethos gunaikcn ergazomenon apo ton somatos, hon hat
. .

pleious eisin hierai (Strabo, xii, 559). ^3 xi^


Golden Bough, v, p, 38.
traditional origin 765
god, were mourned as dead.
This practice of marriage to the
god appears again, but at an earlier date, in Mesopotamia.
Marduk and the Sun-god Shamash both had female votaries who
were married to them and who had human children, and the
word used for these dedicated w^omen was the same as the
Hebrew word for a temple harlot. It is hardly necessary to
point out that the devadasi is likewise married to the
god,
but may have children by men. Dedications of virility were
also
made in the temple of the Syrian goddess whose priests
were
eunuchs who had dedicated themselves by castration, and it is
|

possible that an Indian survival of the same cult is to


be seen
in the dedication of natural eunuchs or otherwise
deformed
males to the goddess Huligamma, and in the cult of the goddess
Chatushringi, whose temple on a hill near Poona is served by
men (said to be natural eunuchs) who dress in women^s clothes
and spend their lives begging and worshipping the goddess,
at whose temple they collect in large numbers to celebrate the
Dasehra. In the Deccan besides the girls dedicated to the god
Khandoba as murli, boys (vdghyd) are also dedicated and are
;

brought up as temple servants and mendicants. This god,


Khandoba, is worshipped at the Dasehra in association with ,

Ekavira, who is definitely a fertility goddess


worshipped during
the Dasehra in Maratha households in little Gardens of
Adonis In any case the general parallel between the practice
in southern India and that in Syria and Asia Minor is too close
to be fortuitous, and offers another link between Dravidian
India and the eastern Mediterranean. It is true that the Asian
deity was more often a goddess' and that in India a god is served,
but there is much evidence to suggest that in India, as in Greece
and Italy and Asia Minor, the mother goddess and
as also in
a matrilineal system preceded a change to a patrilineal system
introduced by invaders. It is, of'course, impossible to dissociate
the custom by which all worshippers propitiate the deity once
iri their lives
by an offering of their chastity from that of
dedicating some individuals to do it for a period of their lives.
Indeed, the custom alluded to above of dedicating a daughter
as a hasavi for the sake of reviving the otherwise inoperative

Vide Fnzer, Golden Bough, n (The Megic Art and the Evolution of
ii), ch, xvm, The institution of kingship under a matrilineal system
seems to have accompanied the worship of 9ie mother goddess*
i66 origins
marumakkathayam inheritance rather suggests that the practice
of dedication in one form or another, real or symbolic, may
have at one time been the universal concomitant of the matri-
lineal system in India as in Asia Minor or in Cyprus. It is

possible that the talikettu ceremony on the Malabar coast points


in the same direction; the Brahman explanation of it as a
purificatory sacrament,^* read with Hamilton's account of the
Zamorin's nuptials, is not antagonistic. At any rate the
custom of dedication prevailed as far west as Cyprus, where there
was a custom (Herodotus mentions it) similar to that of Babylon,
and on the African coast to Sicca Veneris and probably to
Carthage herself. The Cyprian shrines were connected in form
in respect of their horn and pillar cult with Crete and Mycenae,
while the cones which were the emblem of the goddess at Byblos
and other places in Asia Minor have been found in the most
ancient sanctuary of Cyprus and as far west as Malta. There
can be little doubt but that the custom of consecrated prosti-
tution originates in a commerce regarded (on some principle
of sympathetic magic, perhaps) as essentially necessary to ensure
that life should be propagated and that the earth should fructify.
According to Bernier^ the virgin married to Vishnu at Puri
consulted the god as to the abundance of the coming harvest, an
association which is most significant. Marco Polo, in giving an
account of the custom of prostitution to strangers in the province
of Kamul, tells us how it was indignantly prohibited by the
Emperor of China. The order was obeyed for three years and
then rescinded, as it was found that the land became barren in
consequence and the earth no longer brought forth her fruit in
due season."*

The division of society into two groups, castes of the Right


Hand and castes of the Left Hand, has already been mentioned
in an earlier chapter. It is natural to suppose that it reflects
some acute division of society which, though it apparently con-
vulsed the other Dravidian-speaking provinces of southern India,
failed to affect the Malabar coast, since such a division is absent
in the Malayalam-speaking area, though lively enough in the
* Iyer, Cochin Tribes and Castes, ii, p. 39.
New Account of the Mast Indies, p. 510.
Voyages de Frangois Bernier, ii, p. 105.
Bk. I, ch. XU (Yule, 1, p. fio).
Vide supra, pp. 11, ^sqq.
TRADITIONAL ORIGIN 167

Telugu, Kanarese and Tamil countries. In the three latter the


makkathayam principle of inheritance through the male is the
prevailing order, in spite of occasional survivals such as the
basavi system referred to above. In the former, however, the
marumakkathayam system still prevailsand has prevailed almost
without exception until a recent date. The inference that the
factions of the Right and Left Hand arose as a result of the
introduction of the makkathayam principle, which some castes
were unwilling to accept, is inescapable. The fact that the
women of one or two castes belong to one faction while the
men belong to another does nothing to weaken this inference,
since there nothing more likely than that the women of a caste
is

might be opposed to a change of this kind while the men were


wishful to adopt it,^ and it is to be observed that in the only case
in which the women are certainly reported to differ from their
menfolk the women belong to the Right Hand and their men-
folk to the Left. Moreover, in the case of the Madiga caste the
women are often made basavi or are dedicated as devadasis, for
whom the matrilineal system is important, while in the Kaikolan
which the same practice obtains,
caste (a Left-Hand caste) also, in
the womensometimes belong to the Right Hand.^ Further,
it is to be noted that while the cultivating classes support the

Right Hand, which we may take perhaps to represent the form


of society which conformed to ancient practice, the artisan castes
all support the Left-Hand faction. These claim that their- privi-
leges were bestowed
on them by a goddess Kali.^ Brahmans
and other castes from northern India are generally regarded as
not concerned in the rivalry of the two factions. Further, there
is a strong tradition in southern India that where burial is

One exception in Malabar is the Chaliyan caste of cotton weavers, and


while some of them are of the Right Hand others are of the Left, the latter
being regarded as superior, though numerically much weaker. Some of
j
the caste follow the rmrumakkathayam rule, others (it may be the Left
Hand, but we are not told) the makkathayam. In any case the caste
appears to be an immigrant one, and this is confirmed by the fact that
1 some of them are patrilineal. See Thurston, Castes and Tribes, n, pp. 11 sq.
** Among the Jats men claim the right to marry a deceased brother's
widow, while women vehemently deny it. Wise, op. cit., p. las-
According to Rose (' Caste *) devadasis of the Right restrict meir favours
to the caste to which they belong, whereas those of the Left Hand are much
f more promiscuous, a distinction which seems to support the theory advanced
I here that the devadasis of the Right Hand are merely adhering to an original
:
matrilineal system.
Buchanan, through Mysore, Canara and Malabar, ch. n (1870,
vol. I, p. 54)-
168 t r
either on account of poverty
disposing of the aeaci.
resorted to for
other reasons, taming of course being the
or for climatic or
disposal for Hindus cultivators normally
preferred method of
the Panchala, that is the five artisan
Le a simple grave, whereas graves with stone slabs, a contrast
their
rustes line and cover
is suggestive
of two different cultures, a metal-
in practice which
perhaps been immigrant into an area of
usine culture having
still in the stone age.
primitive cultivators
^ Be that as it may, there is one obvious inference to be dmwn

of the origin of the Right- and Left-Hand


if the hypothesis
advanced be accepted; and it is an inference, more
factions here
other considerations;
over which not without support from
is

sort of a caste system, or something closely


that is that some
southern India along with a matrilineal
analoKOus existed in
to the infiltration of Rigvedic Hinduism from
culture antecedent
by Oldham
the north This would support the view advanced
we know it is the result of the imposition of the

that caste as
conditions found in India at the Rigveda
varna tradition upon

that 'the influence of a Brahmanic


period He suggests

hierarchy was exerted


with the object of embracing within
tribes and social groups ' already occu-
their fold the indigenous
establishing at the same time an
pying the country and of
broadly with the tradition of
organization that would conform
'

conclusion accords in some degree


the sacred hymns. This
above,'* but it would be rash to
with that of Senart referred to
developed, as a deliberate
regard the caste system, as it has since
construction so mudi as the result of the reactions of indigenous

institutions of taboo, pollution, purification and so forth to whaf


hierarchical policy adopted
was perhaps a conscious political or
successors towards the corn-
by the Rigvedic invaders and their
in' , Itics they found in the
land.

.^i|e complete and emphatic change over from the matrilineal


to have taken place
,
"
ae patrilineal system, which seems
whole of India except the Malabar coast,
iirghout almost the
'

.J^ps to be attributed to the sudden spread of some new


Such an
jl^Mnvincingly satisfactory, even if illusory, idea.
importance to be attached to
n idea of an exclusive
from the observation of
ty 'k breeding-may have arisen

Man, xxxn,
*In wadew rtt GhiByei 'Cute and Race in India' in
isonipw.
TRADITIONAL ORIGIN 169

some no'^.able or characteristic feature derived from the bull in


any herd in which a number of cows are served by a single bull.
Sir Arthur Olver, late Principal of the Royal (Dick) Veterinary
College in Edinburgh, and at one time a Colonel in the Indian
Army Veterinary Service, writes to me of *
the far greater
influence of a sire in a herd as a whole than of any particular
female. Siring as he may from fifty to a hundred or even more
calves in a season, the bull is bound to have a more profound
effect on the herd Among matrilineal peoples accustomed to
attach importance to paternity, an observation of that
little

kind might well spread with a conviction devastating to existing


theories of descent, and cause a swing over to an extreme
opposite view^ such as that now generally held in Brahmanic
doctrine which regards woman as merely the soil, a soil in which
the paternal seed germinates and grows, it is true, but which
contributes little or nothing to the new creation except
nourishment. It is possibly significant in this connexion that
a pastoral society has generally been found to be associated with
pronounced male domination.
CHAPTER XI

OTHER THEORIES; FACTORS IN THE


EMERGENCE OF CASTE
The European observers of the caste system were content
earlier
to regard as an artificial creation, as a device of a clever
it

priesAood for the permanent division and subjection of the


masses, or even as the creation of a single lawgiver. Thus the
Ahh6 Dubois, among others, speaks of it as the ingenious device
of Brahmans and clearly regards it as made by, and for, them.
But it is impossible to accept such a view. So deeply rooted and
pervasive a social institution as the caste system could hardly
have been imposed by an administrative measure. It is organic
rather than artificial, and the same criticism applies to the
occupational theory of caste advocated by Nesfield in his Brief
View of the Caste System. Nesfield regards occupation as the
exclusive basis of caste distinciion and emphasizes the fact that
artisans working in metals rank higher than basket-makers, for
instance, and other primitive callings which do not involve the
use of metals. Caste he regarded as a natural product of society
in the creation of which religion played no part at all. This
theory will hardly stand critical examination. It is true that
basket-making an occupation of one of the lowest castes; at
is

the same time the blacksmiths caste is very far from being more
highly esteemed, though it must presumably be later in point
of time, than that of the coppersmith or the goldsmith. More-
over, this scheme does not explain at all the varying positions
of agriculturists who arc of low castes in certain parts of southern
India but generally of respectable if not of high caste in northern
India. It should, however, be mentioned in this connexion that
it does appear to be a custom in southern India, or in parts

of southern India, to provide a craftsman, that is, a man of the


carpenters or blacksmiths or goldsmiths professicm, with a grave
made of stone slabs in a cist form in place of the simple unlined
grave which is given to a cultivator, and this does suggest either
a difference in status or in custom between the aborigines and
immigrants who brought in tools and aafts, or in the alternative
some difference in escLatology between the cultivator and the
craftsman. Moreover, if the cnigini of the. caste system were, as
FACTORS IN THE EMERGENCE OF CASTE 171

Nesfield maintains, totally independent of religion, they have


undoubtedly received a religious sanction since.
Nesfields view is, however, in some* measure supported by
Dahlmann in his Das Altindische Volkstum. Dahlmann regards
Indian society as organized originally into three natural groups
the priesthood, the nobility, and the bourgeoisie^ ^which are,
in his opinion, found in every civilized community, representing
the divisions of the people respectively concerned with the reli-

gious, the political, and the economic branches of the national


life,and which correspond to the three twice-born varna; these
groups he conceives as split up into a number of smaller groups
and communities, some based on relationship, others bn commu-
nity of occupation. Caste sprang not from the four varna but
from the infinite number of corporations and of groups of rela-
tives into which these four varna were divided. There was,
he considers, a steady progress of development from classes to
corporations and from corporations to castes. Agriculture was
originally the prime factor in the economic life of India; a
rival developed in the form of trade and industry, and the
principle of division of labour became so important that ii
became regarded as the duty of the ruler to base his economic
policy on the division of labour and distinction of occupations.
On the basis of the old division by classes, corporations gradually
arose and guilds of traders and hand-workers came into existence.
Community of interest among persons following the same craft
gave rise to a corporate organization, and technical skill was
passed on from father to son. Families of craftsmen thus
arose bound together by a community of interest which gave
rise to a corporate organization and formed a guild. It is this
guild which is really the basis of the caste system. Here again
this view fails to account for a great deal in the caste system.
The German guild is cited by Dahlmann as a dose parallel, but

it involved no ban on commensality between one profession and


another and no prohibition of intermarriage, nor is it conceiv-
able that a distinction based solely on function Could develop
into the vivid and lasting prejudices that accompany caste dis-
tinctions in India.
Blunt ^ follows Dahlmann and accepts also Risley's theory,
heading that the origin of caste mu^ be sought to in the peculiar

pf sqq.
'

1,7, ORIGINS
circumstances of a complex system of society
a society of classes

with a cross-division of guilds. The were largely a


classes

matter of colour, and the ultimate result was a society divided


shades colour, all degrees, that is, of mixed
into groups of all of
blood. Meanwhile guilds were growing up which contained
recruits of all classes and which developed strong self-governing
oiganizations; within each such guild would be found little
groups each with the same class prejudice against corrupting
their blood by intermarriage. Each guild would thus become
an endogamous caste with endogamous subcastes. In confirma-
tion of this theory Blunt finds first that the functional caste is

made up of subcasteswhich are themselves endogamous, whereas


the non-functional castes are endogamous as a whole, but
are not so generally made up of endogamous subcastes; secondly,
that the functional castes have much more powerful governing
bodies (panchayats) than the non-functional; and thirdly, that
while the non-functional castes claim descent ultimately from
a common ancestor, the functional ones do not. The instances
given by Blunt of functional castes include Brahman, Kayastha,
Sonar and Lohar, of non-functional Rajput, Khatri, Jat and
Bhar. These, he says, are made up of exogamous sections instead
of endogamous. The facts are, however, by no means so simple
as Blu:.t here suggests, for exogamous gotras are to be found
in functional as well as non-functional castes, while the latter,
though they may claim a common origin, do not as a rule claim
a common ancestor, whereas certain functional castes do claim
a common ancestor in Viswakarma.
Chanda also traces caste to race and function. Colour or race '

difference, real and fancied, together with hereditary function,


gave birth to the caste system. But as newer groups formed or
attached themselves to the Arya nations, the absurdity of regard-
ing them all as distinct colours or varnas was recognized, and
the theory of varnasankara or mixed caste was started to explain
*
their origins.

Senart, in his well-known work Les Castes dans VInde, seeks to
<j!r account for caste in an entirely different way. He does not main-
tain that the caste system springs only from primitive ' Aryan
'

elements but regards them as most important in the creation


of the caste system, as he stresses the importance of commensality

s Chanda, Indo-jSTyan JRocm, p. g6.


FACTORS IN' THE EMERGENCE OF CASTE 173

and derives the commensality of caste jErom the


exilusive
family worship and family meal of the primitive gens or clan,
although he compares it, not very satisfactorily, with the Roman
tribus and the Greek phyle. He regards the '
Aryan '
invasion
as having resulted in a mixed race with two orders of scruples
about purity, one order based on purity of descent and the other
on purity of occupation. These scruples led to the formation
of new groups among which the priestly class alone maintained
a solid feeling of espritde corps, using its moral power to
establish the caste system. This it proceeded to interpret in
terms of the traditional division into four classes, accounting
for the multiplicity of existing groups by a theory of mixture
and by ascribing the lower social level of certain castes to their
extraction from irregular unions. This theory also is open to
serious criticism. Dahlmann has attacked it as postulating a
completely non-existent simplicity ^of society at the time of the
Rigveda, when, he says, the majority of Indians had no longer
the rude customs of a pastoral people but a wide and, in some
respects, a highly developed civilization in which custom and
ritual were well developed. In any case Senart*s explanation
of the exclusive commensality of caste as derived from the family
meal and sacraments shared with the family ancestral spirits
cannot be regarded as satisfactory. A caste in India hardly ever
claims a common ancestor. It is the gotra which, like the gens,
does so, and it is the gotra which, like the gens, is an exogamous
unit. The caste is not a clan and is not analogous to a clan, and,
whatever the origins of this particular feature may be, identical
gotras are actually found in diiSerent castes, so that caste is not
analogous to a trihus or a phyle any more than it is to the gens.
Senart, however, has clearly seen that neither race nor occupation
or function is by itself enough to cause a caste system to come
into being, or to account for its restrictions on commensality
and marriage.
Sir Herbert Risley, in his book The People of India, on the
other hand, has relied mainly on theories of race and hypergamy
to explain the caste system, which he clearly regards as prinoarily
due to colour differences and to a system of hypergamy result-
ing therefrom. In order to base caste on hypergamy Risley finds
it necessary to presuppose a hypothetical point at which the result
of intermarriage between fair invaders and dark aborigines
174
ORIGINS
provides enough women for the society in question to close
its ranks and become a caste, although there still exist outside

it more women of the same community from which it has been

drawing its wives and with which it has been in more or less
intimate relation. The position of negroes in the southern states
of the U.S.A. has been cited as offering a parallel case, and the
view is supported by Westermarck;* it finds a certain measure
of confirmation perhaps in the laws passed in the Union of
South Africa against the intermarriage of white and coloured
races; but it fails to offer any satisfactory explanation of the
taboo on commensality. It is no doubt true that separate rail-
way carriages, separate restaurants, even separate townships, are
provided for negroes, but no pollution takes place ^ as a result
of employing negro servants, and there is no hard and fast line
which is really analogous to a caste division between, say, quad-
roons and octoroons, nor have the social factors which might
have tended to produce similar results in India ever succeeded
in. making Muslims, Anglo-Indians or Eurasians into a caste in
the Hindu and where Muslims do form a real caste it
sense,
is always one which has been converted to Islam from pre-Islamic
inhabitants while retaining its original caste organization. A
History of Human Marriage (1901), pp. $65-7.

At the same time it is necessary in this connexion to draw attention to
*
a passage quoted by Dr Little in * The Psychological Background of White-
Coloured Contacts in Britain in The Sociological Review, xxxv, 3, n. 4
'

from Deep South:


'
The belief in the organic inferiority of the negro reaches its strongest
expression in the common assertion that the negroes are unclean . .

There remains a strong feeling that the colour of the negroes is abhorrent
and that contact with them may be contaminating. There is generally a
strong feeling against eating or drinking from dishes used by n%roes, and
most of the whites provide separate dishes for the use of their servants.
The idea of uncleanlmess is also extended to any clothing worn by negroes,
as was dramatically shown when a negro customer returned a coat which
she had bought from a white clothing merchant. The clerk was unwilling
to accept the coat, and when the assistant manager accepted it, the clerk
said to another clerk: ** This is perfectly terrible; I think it is awful. We
can't put that coat back in stock." She hung it up very gingerly and did
not touch it any more than necessary.' {Deep South : A Social Anthropolo-
gical Study of Caste and Class, Allison Davis and B. B. and M. R. Gardner,
p. 16.)
The attitude to negroes indicated here is most suggestive of the caste
system.
But according to Professor Vesey-FitzGeraldexclusive Muslim commu-
nities having a caste organization exist outside India, the Tbadis of the
M'zab in Algeria, for instance, while Ismai'li sects in India tend to organize
themselves in secret societies with esoteric rituals and a dreadful penalty of
excommunication (review of Caste in India in the Bulletin of the School of
Oriental and African Studies, xii, p. atjfi).
FACTORS IN THE EMERGENCE OF CASTE 175

system of hypergamy is no doubt explicable as a result of the


impact of colour prejudice on a caste system, but it is difficult
to see how it can give rise to caste. In general its action pre-
supposes just such facilities for intermarriage and commensality
as the caste system tends to suppress.
Several other writers on caste explicitly follow Risky or are
content to take his theory for granted. Professor N. K. Butt,
while criticizing him, adopts his racial theory of origin in efEect,
though he attaches much more value than Risky to the account
of caste in theCode of Manu. Mr Hayavadana Rao ^ and Dr
Ghurye likewise regard caste as having arisen largely as a result
of racial differences.Ghurye emphasizes in particular the factor
of priesdy manipulation by Brahmans attempting to maintain
the purity of race of Aryan invaders. One cannot but believe,
however, that for priestly interference of this kind to be effective
in setting up so far-reaching and complicated a system as that
of caste, it is necessary to assume the pre-existence of certain of
the essential factors in that system which would predispose the
population generally to accept an extension of them. Colour
prejudice and racial exclusiveness have been common enough in
the history of the world, but they have nowhere else led to such
an institution as caste, and it would be rash to suppose that
they could have done so in India of themselves. On the other
hand, race and colour prejudice have no doubt made an impor-
tant contribution to the development of the caste system, and
it may be that they have played such a part in the crystallizing

of that institution as to justify the statement that it could not

have come down to posterity in its present form without having


been subjected to the reagent of racial prejudice and discrimi-
nation. Indeed, Risley's test of the nasal index as indicative of
the position of a caste in the social scale has been regarded by
both Ghurye * and S. C. Roy as holding good in a broad sense
for northern India. If indeed it does so, there are a number
of striking exceptions to the general rule.
Ibbetson has, like Ghurye, laid great emphasis on the exploi-
tation of their position by the Brahman caste which he supposes
Ori^n and Growth of Caste in India, ch, i.
r Indian Caste System, p. 66.
^ Caste and Race in India, ch, vil cit., pp. 108 sqq.
*
Caste, Race and Religion in Man in India, xiv, no, a, p. 79.
11 Fanfab Castes, pt, i.
17(5
ORIGINS
to have degraded all occupations
except their own and that

o their patrons of the ruling class He explains caste as arising


from a combination of tribal origins, functional guilds and a

Levitical religion and has laid the greater stress on the tribe.

It of course, clear enough that certain tribes are responsible


is,

for the formation of certain castes, and no less dear that certain
castes are, or have been in the past, restricted to certain occupa-
tions, but Ibbetson's explanation of the origin of caste is really
only a summary of certain of its obvious features. These features

and religious monopolies ^have no doubt con-
tribes, guilds,
tributed to the growth and extension of the caste system, and
done much to consolidate and perpetuate it, but they cannot be
regarded as causes. They are features which are not unique but
common to many sodeties in many countries, whereas caste is
found in India and nowhere else. Given caste, a tribe, a guild,
or a priestly order may very easily become a caste, but failing
the essentials of the caste system there is nothing in them, with
the possible exception of the priesdy order, which will produce
a caste.
Quite a different origin for caste is argued by Hocart, who
apparendy regards the whole system as originating in ritual. He
regards the four vama as a division of the people devised pri-
marily for ritualistic purposes; they represent the four points ot
the compass, as do the colours white, red, yellow and black. The
ritual appears to be a duplication of some ancient fertility rite
in which two divisions of the people were involved, representing
the sky and the earth, and to be focused on the king as repre-
senting the deity, or as being himself the deity incarnate, and
the functions of various castes are derived from the offices per-
formed by them in the daily ritual of the royal court. Hocarts
conclusions are, however, based on his observation of the existing
ritual of the Buddhist Temple of the Tooth at Kandy and not
on direct acquaintance with caste in India. The functions of
the washerman and barber necessary to purificatory ceremonies
have struck him, and he seems to regard the corresponding castes
as having been created by local imitation of royal courts on a
small scale by petty rulers and landed proprietors, each of whom
must have, like the divine or royal model, a barber and a washer-
Les Castes, passim. Bruce Ryan, in a note on p. 180 of his Caste in
Modem Ceylon, remarks: Hocarts attempt to interpret caste in purely
ritual terms must be utterly discarded.
FACTOKS IN THE EMERGENCE OF CASTE 177

man and other ritual functionaries, so that there grew up classes


of people each performing separate ritual functions, which classes
developed into closed castes. This view, however, offers no
explanation of the taboos on commensality and intermarriage
which are essential to the caste system, unless it is conceived
that any performance of ritual functions is vitiated if the per-
former is in any way contaminated by association with persons
who do not or cannot perform them. In such a case ideas of
taboo, mana, and magic must be postulated which might them-
selves be enough to account for the formation of caste without
bringing in at all the idea of ritual. In any case, it is perhaps
less unconvincing to suppose that the performance of the ritual
was first associated with particular groups because that was their

customary work than that association with the ritual led to the
formation of the caste. At the same time it must be admitted
that Hocart's theory does supply a unifying factor in the caste
system in that all over India castes are to be found with, up to a
point, similar ritual as well as social and economic functions
from Brahman down to washerman.
A theory of the origin of caste which combines both functional
and racial origins has been put forward by Slater in his
Dravidian Elements in Indian Culture. He emphasizes the
fact that caste is actually stronger in southern than in northern
India, and suggests that caste arose in India before the Aryan
invasion as a result of occupations becoming hereditary and
marriages being arranged by parents within the society of the
common craft because sexual maturity is early and trade secrets
were thus preserved. As a result of magic and religious cere-
monies also, exclusive occupational groups were built up,
marriage outside which became prejudicial and contrary to
practice. The Aryan invasion had the effect of strengthening a
tendency to associate difference of caste with a difference of
colour and of strengthening also a tendency for castes to be
placed in a scale of social precedence. Slater also maintains the
existence in the pre-Aryan society of India of an order of priest
magicians which he associated, rather unnecessarily perhaps, with
Egypt and a heliolithic cult. Dutt,^^ to whose views rrference
has been made above, takes a somewhat similar view of the
impact of an Aryan culture upon pre-Aryan occupational classes
IS op. cit., sqq., 106.
pp, i}8

12
,,8 ORIGINS
already endogamous on account of i
their own and ^that

the preservation of trade secrets. Ghui ^ caste as arising

lar line, and suggests that the southe 3r'*^^'^tional


guilds and a

their contact with Indo-Aryan culture,


of peoples about the- poWi^^rribes are responsible
primitive
qualities, while ideas of untouchability
clear that certain

monial purity first applied to aboriginal^^Wfcd to certain occupa-

ficial ritual and impurit'^^lgin of caste is really


the theoretical
Ghurye^s account of caste was publis M^^fires, Xhese features
that Bonnerjea, in an article in The ^Jl-have no doubt con-
explained caste as due to primitive beli^^j^c caste system, and
he credits both Aryan and pre- Aryan, ^
but they cannot be
view apparently of the operation of hypc#^^ are not unique but
Aryan who married into indigenous peCP^^iies, whereas caste is
enough women to close its ranks and iste, a tribe, a guild,

is the Aryan invader whom he regards caste, but failing

for the institution of caste. The cont*|thing in them, with


established, he regarded as resting on the which will produce
occupation being primarily that of being f
fore no doubt specially skilled in the craft **^^1 jd by Hocart,^* who
No doubt the common practice by craftsKrf|ating in ritual. He
the tools of their craft could be cited in s^p^people devised pri-
The tools, like the craftsmen, would the four points of
particular mana involved in the craft, ar Imf and black. The
or desirable to keep en rapport fmcitnt fertility rite

Rice in 1929 published an articj|;^^ I'olved, representing


he later elaborated in fi
the king as repre-
of caste to totemism a; leity incarnate, and
A
it. He emphasizes in fpm
.

the offices per-


and the principle thaf court. Hocart's
f',~

qualities in totemistic: / tion of the existing


Gilbert has pointed /at Kandy and not
to different peoples at li*
The functions of
divisions of the coast^;^the l^fc'/|catory ceremonies
mountain and desert areas, 'The Perava^^l^rresponding castes
by fishing, the Vellalar and Vdir cultiva| / royal courts on a

i-iOprAqit., p. 145. l^tors, each of whom


15 *
Possible Origin of the Caste -/rber and a washer-
i The Origin of Caste
%
Guha advanced a theory somewhat 180 of his Caste in

a thesis for Harvard University ^f/rpret caste in purely
consulted.
FACTORS IN THE EMERGENCE OF CASTE J7?

man and other ritual functionaries, so that there grew up classes


of people each performing separate ritual functions, which
< lasses

developed into closed castes. This view, however, olirrs si<

explanation of the taboos on coromensality and iiuerm.uriagc


which are essential to the caste system, unless it is coiueised
that any performance of ritual functions is vitiated if the per-
former is in any way contaminated by association with persons
who do not or cannot perform them. In such a case ideas of
taboo, mana, and magic must be postulated svhich might them-
selves be enough to account for the formation of caste svithinii
bringing in at ail the idea of ritual. In any case, it is perhaps
less unconvincing to suppose that the performance tif the ritual

was first associated with particular groups because that was their
customary work than that association with the ritual led to the
formation of the caste. At the same time it must be admitted
that Hocarts theory does supply a unifying factor in the t aste
system in that all over India castes arc to be fouiul with, up in a
point, similar ritual as well as social and ettmtunic fuiuiinri*
from Brahman down washerman.
to
A theory of the origin of caste which combines both fimt tintwl
and racial origins has been put forward }y ,Sla;cr in lii*
Dravidian Elements in Indian Culture, lie cniplusircs the
fact that caste is actually stronger in southern than iti muthein
India, and suggests that caste arose in India before the At van
invasion as a result of occupations becoming heirtiitaiv and
marriages being arranged by parents within the ioietv of the
common craft because sexual maturity is eatlv ami tiadr wuet*
were thus preserved. As a result of magic ;md religiom trie
monies also, exclusive occupatiotud groups were huils up,
marriage outside which became pieiiHiui.a and outuajs (
practice. The Aryan iiva.sin hati the effett of Jitfugihruiug
tendency to a-ssociate different e of caste with a thUnrmr ..i
colour and of .strengthening also a teiuirmy Un tasir. n. )w
placed in a scale of social* precedeme, .Slater also tuauuatut iltr
existence in the pre-Aryan society of India of an ortirr
of jtnrtj
magicians which he associated, rather tinnecessaiih
pnhap^ hidt
and a heliolithic cult. Uutt,*^ to whose cirws irfrim-r
has been made above, takes a soinewh.if
similar vjrw of j};-
impact of an Aryan culture upon pre,Aryan
ttctupaiimu! tUi-rs
Op. eit., pp. xH Mjij ,

at
i8o ORIGINS
jati applied to Brahmans, Kshatriyas, Chandalas, Nishadas,
etc.,
that is, to themrna and to tribal groups, whereas tradesmen and
artisans are termed kamma and sippa, and they appear to
be
organized in sreni or associations of persons following
the same
occupations which made their own laws or rules
and exercised
jurisdiction over their members. The occupations would tend
to become hereditary and their guilds, under
a system of restric-
tions due to fear of pollution, would become more and more
detached until they hardened into regular castes,
while racial or
tribal groups would follow a similar
course. Olden berg then
regards caste as derived from class, guild
and tribe segi^egated
into permanently separate groups by
heredity and by restrictions
on marriage and commensality and a fear of
pollution, which
were derived partly from the Aryan
invader and partly from
the aboriginal inhabitants.
This fear of pollution has also been
stressed by Ketkar, who
points out that the chief-principle

on which the entire system
depends is that of purity and pollution
a view which has
not escaped Ghurye and which
has been endorsed by Sarat
Chandra Roy.^^ Both these authors,
however, regard this
insistence on ceremonial purity as
inherently associated with the
Rigvedic Aryans, whereas it is not
improbable that it was taken
over m
part at any rate from, a more
ancient civilization, a view
which would probably receive the
support of Professor Murphy
w o suggests that the very ample provision
for washing and
civilization
itTdtTes that the ancient
indica inhabitants of India laid special
stress
oil ablutions and purification
long before the advent of the
emphasis on the

tiiem by the later Hinduism.^

" '>' tmory of the caste fysteni


lA u
which he T
refers to Colebrooles
in
wort and cinphasires the impor-

government of castes. He lays stress on the fact that under


" > < PP- nd 1.6 n,,.
"I- .

Uoly 1!5.
*' * ^ ' y-A S.B., vol. m. no. 7
FACTORS IJV THE EMERGENCE OF CASTE i8i

Hindu rule the authority of the king was frequently called in to


compel castes to keep to their proper functions. The effect of
the exercise of royal jurisdiction would be to establish gradually
a body of caste custom, and in the case of a caste spread widely
throughout several kingdoms several bodies of different customs
would be created, giving rise inevitably to subcastes or to
different castes. The political condition of ancient India was
particularly favourable to just this kind of fission; Megasthenes,
for instance, reports a hundred and eighteen different kingdoms.
As a matter of fact a very large number of subcastes (e.g. of
Brahmans) do bear geographical names, and the same geogra-
phical name name of subcastes of different
often appears as the
main number of instances of which are given by Jackson
castes, a
from Bombay. Thus both the Ghanchi and Muchi castes have
subcastes called Ahmadabadi; both Ghanchi and Kansaras have
a subcaste called Champaneri (Champaner was the seat of a
Hindu dynasty early in the fifteenth century, but has long been
uninhabited); Brahmans, Charans, Sonis and Lohars have a
subcaste called Kachela or its synonym Parajia, and the name '

Kachella occurs in inscriptions as early as the eighth century


for the people of Kachh. Jackson gives many other instances
of place-names for subcastes, and suggests that while the instances
he gives are all from Bombay province the number could be
multiplied indefinitely by research in the rest of India. It may
here be pointed out that recurrent famines are likely to have
had on castes not dissimilar to that exer-
a fissiparous influence
cised by political disarticulation.
Linton has drawn attention to the probability that India, or
at any rate part of India, formed at some time in its past part
of the Austronesian region in which there has been since
palaeolithic times a fundamental pattern of organization on
"

the basis of small endogamous groups. This pattern he says,

assumes various forms, but it is linked with a predominant


importance of kinship as the basis for organizing the reciprocal
behaviors of the groups members. In such a system, he
suggests, it would be an easy matter for caste to develop in
response to frequent invasions and to the emergence of an
25
Vide infra, 'p, 186.
In a review of Caste in India
26 (first edition) in J.A.O.S., vol. 68 , no. t,
April-June 1948.
^0xiG/~ys
provldh *a flexible mechanism for

TOrimMlatiHf fW^ *^^*"** ' *^ developing guild systems

m iheir
origin of caste so far referred to mostly
Thc theorie* of the
ome important contribution to the subject. At the
cawwln
nufoe tia* *t <*<**' phenomena
than the causes of the caste system, and in particular
radber
fcarc^y one of them lays enough emphasis on the importance
of the primitive conceptions of taboo, mana and soul-stuff in
contributing to its formation. This cannot justly perhaps be
said of Sarat Chandra Roy, who docs indeed accept the hypo-
thesis that a belief in mam
and soul-substance has contributed

'

to the development of caste, which he regards as the outcome of

the interaction between the Indo-Aryan varna system on the one


hand, and the tribal system of the pre-Dravidian and the occu-
pational class system of the Dravidian, on the other. These
he regards as having been cemented into the caste system by the
Indo-Aryan concept of karma and a certain taboo-holiness

that came to be attached Brahman for his accredited


to the
possession of a special spiritual energy (brahma-sakti)
born of
the predominance of the sattva guna [fundamental
quality of
revealing light] sustained and stabilized through
well-disciplined
continuance The emphasis
laid on the contribution of the
karma concept important, but generally speaking Roy seems
is

to attach too much weight to the


philosophic aspects of
mnduism and attempts to interpret the institution of caste
in terms of ethical principles and an
idealism which probably
had do with its primary origin. It is readily
little to
admitted
that nearly any unpromising material
even of the grossest kind
may be found sublimated to form the principle
of a noble
philosophy, but the institutions of human
society are generally
the result of conflicting principles
in which self-seeking and
economic considerations play almost as
great a part as super-
stition, while sublimation comes
in at a later stage when the
results are appreciated as being
far from ideal.

'
Caste, Race and Religion , in Mm in India, vol. xvn, no. 4, p. S54.
ofiAjpim XU
COMCLUSIOMS
Seeing that light is required on the origin of caste it would
seem not unreasonable, as in the ease of religion, to examine first
those cultures that survive in India least altered from antiquity
in case, they can illumine origins elsewhere obscured by changes
and developments due to growing civilization and to external
contacts. Caste, as it now is, is an institution which has grown
and developed through many centuries, but since it is so firmly
rooted in India, and since it is found nowhere else, it would
appear almost certain, on the face of it, that its first beginnings
are to be sought in India and not outside, and we have fortu-
nately in the more inaccessible corners of this vast country still
a few tribes whose primitive conditions of life have changed so
little in a thousand years as to be witnesses of value. Pliny the
Elder, writing in the first century a.d., mentions ^ a great valley
in the Himalayas called Abarimon inhabited by wild men; the
Assamese still speak of the hillmen who remain hostile and
have not come under the influence of the plainsmen in that pre-
cise term, abari manu, untamed folk; and Ptolemy,^ writing in
the second century, locates '
the Nanga-logae, that is the realm

of the naked precisely where the Naga log are found today,
some tribes of them still unclothed, still, in 1948, untouched by
contact with the people of the plains, tribes who had never seen a
white man nor a horse nor knew what
gunpowder, and whose is

language is still unspoken by anyone outside their own


community save some of their immediate neighbours. Hinduism,
Buddhism and Islam have never penetrated here, and caste as it
exists in the plains is unknown and undreamed of, but neverthe-
less institutions are found which seem to throw a definite light
on caste and religion as they have developed in another environ-
ment. Thus in the unadministered area to the east of the Naga
1Nat. Hist., vii, ii. It has been justly pointed out (Sachin Roy, Aspects
of Padam-Minyong Culture, p. i) that when Pliny wrote the Assamese
language, as such, had not yet been formulated. Whether such a derivation
could have been possible from the eastern Prakrit vernacular from which
Assamese evolved, I am not philologist enough to have the least idea.
2 Geography, vii, ii, 177.
s Cf, BiOdson, Naga Tribes of Manipur, pp. 82 sq., and
Gcnisa in
Assam ', 1906, p. 9.
184
ORIGINS
Hills, where each village is an independent political unit, there is

very often to be seen a distribution by villages of certain occupa-


tions. Thus some villages make pots but do not weave cloth;

others weave, and others again are occupied principally with


blacksmiths' work, the one village bartering its products with its

neighbours, when not prevented by mutual hostilities, in spite of


differences of language, customs and sometimes perhaps of race
between one village and another. Here we have clearly an
aspect of occupation distinctly suggestive of the caste system;
and, indeed, the remnants of such a condition seem to have
survived in northern India until the Buddhistic age, as the
Jatakas indicate that certain trades were localized in separate
villages, some containing potters, others smiths and so forth.
But this is not the only feature suggestive of caste. It frequently
happens that upheavals in village politics end in battle, vendetta
and sudden death, and that as a consequence part of a village
community, usually an exogamous clan or sept, is compelled
to migrate to some other village. It might be anticipated that
a group of weaving families would be welcomed in a pot-making
village which only obtained doth by barter, and vice versa, and
up to a point this is indeed the case; numbers are strength, and
such immigrants are generally welcomed and allowed to settle

and cultivate but not to ply their ancestral craft when that
differs from the occupation of their hosts. That is taboo, and
should the strangers insist on it they must again go elsewhere
to some village in which it is permitted. Instances of this have
occurred within the writers personal knowledge (compare also
Hodson, Naga Tribes of Manipur, p. 47), and the underlying
feeling seems to be that the practice of the tabooed craft will
affect the crops and the fruits of the earth generally, perhaps
because it is an offence to the ancestral spirits who are usually
regarded as the source of fructification, or it may be that the
particular form of mana or aren which enables the manufacture
of the artide made by the strangers is liable to neutralize the
corresponding magic on which the traditional village industry
depends. Here, however, there is,' generally speaking, no taboo
on commensality or on inter-marriage, and for the sources of
these aspects of caste we must look elsewhere.
For a possible source of the commensal taboo, however, we
need not look far from that of the occupational one. The same
CONCLUSIONS 185

Naga comraunities which we have been considering afford


abundant instances of taboos on certain foods, of a vivid belief
in mana or, as the Ao Nagas call it, aren, and of the magical
effects of food on the consumer. Certain foods are peculiar to
certain exogamous clans, or at any rate are in many cases
associated with clan ceremonial, and it may be offered as one
hypothesis that the presence of strange craftsmen practising their
craft is less dangerous by the prohi-
condoned or rather rendered
bition of intimate with them, reducing thus the
relations
inconvenient strictness of one- taboo by erecting another which
at the start may be less irksome.^ That ;this is in accordance
with the spirit of the primitive society under consideration is
demonstrated by the readiness with which in some tribes the
proximity of Christian converts is tolerated, even though they
cultivate on tabooed days, provided they live outside the village
fence and therefore form a more or less separate community,
though here again commensality is not barred, except indeed in
so far as the Christians refuse to eat such flesh as they are taught
to regard as meats offered to idols
'
Another hypothesis, and
there is no reason to suppose the suggested explanations to be
mutually exclusive, is the theory that the food of strangers is
itself dangerous. Senart's citation of the taboo on strangers at
the family meal would doubtless lead back to this, and Rice's
view of the taboo on commensality as derived from a belief in
totemism agrees in effect with the hypothesis here put forward,
since both depend for their force on the belief in mana and in
the resulting taboo on food or other contacts, which may be
infected with the dangerous soul-matter of strangers; this soul-
matter is particularly perilous if such strangers have new and,
what is the same thing, mysterious arts and therefore magical
powers. Thus when the writer was touring in previously un-
visited Naga territory in 1923-4 he found villages which not
only objected to accepting presents or purchase-money of any
kind from the strangers, or to parting with any possession to
them for fear of the influence to which they might thus become
subject by proxy as it were, but which actually destroyed mats
or other property lent to build shelters when the visitors who
had used them had gone, and threw away their tainted coins in
the jungle. The differentiation between cooked and uncooked
food as a vehicle of pollution so familiar to any observer of caste
m ORIGINS
in India is likely to be traceable to this view of the infection,
by the act of cooking, of the food cooked with the mana of
the cooker. Thus a Brahman can buy ghi, grain, or vegetables
in the market, even from Muslims, but once such food has been
'

cooked, becomes Brahma, and so must be treated with


it

sacramental care Similarly, among the Maori, to quote


Eldon Best, the most soul-destroying thing according to native

ideas is tamaoa, deprivation of tapu by means of cooked food.


In the case of the Maori, however, it is the low-caste man who


dare not eat of the chiefs dish, whereas in India it is the high-
caste Brahman who must not eat any food cooked by a man of
a lower caste. A taboo on intermarriage could easily be traced
to a similar source if not to the same one (among the Mafulu
of New Guinea no girl who is not a near relative of a bachelor
may even see him eat), since there is little use to a man in
a wife who cannot cook for him. In Orissa there is a whole
caste known as Chattarkhai which consists of descendants of
Hindus who were outcasted for having taken food at relief
kitchens in 1866. The caste consists of two endogamous sub-
castes, one of which is of Brahman derivation, the other being
derived from other castes. Famines must have been an impor-
tant factor in the splitting up of castes.
Ideas about soul-stuff or life-matter and of the necessity of
protecting it by the observance of taboos are common over the

whole area of south-east Asia and Australasia from India to


New Zealand, though how far they represent undeveloped primi-
tive ideas and how far they are the remnants of some more
developed philosophy it is difBcult to say. They have been
described by Mills for the Ao
Nagas, by Marshall for the Karen
of Burma, by Warneck for the Batak of Sumatra, by Kriiijt for
the Toradja of Celebes, by Keysser for the Kai of New Guinea.
Traces of these beliefs are very widely spread, and the present
writer has given a brief account of them elsewhere. Generally
speaking this life-matter
is finite and concrete, if invisible, but

liable to be attractedand drawn away by any more powerful


manifestation of the same principle or to be otherwise harmed
if not protected. Its transferability has made it the underlying
^ Stevenson, Rites of the Twice-Born,
p. 346.
OMalley, Caste Customs, p. 50.
In the Frazer Lecture lor 1938, at Oxford, on A Primitme Philosophy

CONCLUSIONS 187

motive in ceremonial and perhaps other cannibalism, in head-


hunting, and in human sacrifice, none of which is unknown
in India. Life-matter may be inherent in apparently inanimate
objects. Warneck says of the Batak: 'Soul is awarded to the
house, the hearth, the boat, the hatchet, the iron and many
other instruments not because they are fetiches, but because their
usefulness is proof of their soul-power.' ^ One is again reminded
of the Indian craftsman worshipping the tools of his craft.
These beliefs in life-matter, or soul-stuff, arc closely associated
with beliefs in mana and with the practice of magic. According
to the Kai of New
Guinea mere words have their own soul-
stuff, which we may perhaps here call mana, just like a person's

name, and a mere form of words may have its own efficacy
like a mantram. Wherever the belief in mana prevails a
corresponding belief in the value of taboo as a protective measure
is also to be found. It is possible that ideas of mana and taboo
have been distributed to Indonesia and the Pacific from the
Indian peninsula. Hocart has pointed out the analogy between
the barber, amhattan in Tamil, whose business it is to shave
heads, and the mbota, a clan in Fiji, who alone can touch the
chief's head. The Malayo-Polynesian word tabu itself seems
to contain the basic meanings of segregation and refuge,'* but
it is abundantly clear that the ideas of mana, magic, and taboo

have from a very ancient period been shared by India with


Indonesia and the Pacific. Food taboos have already been
mentioned, and they are particularly striking in connexion with
caste. Handy
According to the eating tapu in Polynesia '
'

requires men and women and persons of different degrees of


'

sacredness to eat separately, since a loss of mana would be


'
'

incurred by the more sacred person as a result of the shared


food's acting as a conductor This is precisely the outlook
'.

required for a taboo on commensality between caste and caste.


Roy^^ has criticized this argument on the ground that the
Brahman is at the head of all castes, yet his food can be taken
by anyone, whereas the Polynesian chief's food cannot be taken
by a man of lower position who would be destroyed by infection
Quoted by S. C. Roy, *
Caste, Race and Religion *, in Man in India,
XIV, no. 5 , p. 117. ,

s '
India and the Pacific % C.J.S,, vol. i, pt. iv, p. 176.
I. H. N. Evans,
Kempunan *, in Man, xx, 38 (May igiso).
Polynesian Religion, p. 49.
'
Ckte, Race and Reiigion *, in Man in India, xiv, no. s, p. 111.
188 ORIGINS
with the chief's more powerful mana. This criticism is perhaps
inoperative when one regards the two systems as extreme develop-
ments in different environments from a common and primitive
basis. not suggested that the caste system has developed
It is

from ideas of soul-stuff, mana, magic and taboo alone; only that
without these ideas it could not have developed. If these ideas
alone were enough, one might expect to see a caste system in
every island from the Nicobars to Easter, instead of only in
India. At the same time Roy himself supplies evidence of a
precisely similar outlook to that of theMaori within the system
of Indian caste. If a Brahman enter the quarters of the Holeyas,
'
they turn out in a body to slipper him, in former times it is

-!i
said to death while a party of Brahmans who passed through
a Paraiyan hamlet had to run to escape cowpats and broomsticks
because contact with them meant ruin to Paraiyans.^^ The
taboo on a Brahman's entering a Barui's panbari has been
I'
mentioned earlier.^^ Again, when a Brahman has been in a
Kuricchan's house, the moment he leaves it, the place where he
was seated is besmeared with cowdung to remove the
pollution Roy himself again tells us that the pre-Aryan
inhabitants of India had developed the art of magic to such an
extent that the Aryan immigrants into northern India called it

Asura Vidya, the science of the Asuras



Here, surely, is the
basis on which through a number of factors, too complex to have
occurred together elsewhere, the caste system of India has come
into being and developed into a unique social system found
nowhere else in the world. It is probably significant that one
observer reports that in Gujarat, at any rate, caste taboos are
stricter as one goes down in the social scale. High-caste Hindus
are less particular about their water than castes of lower social
position, and restrictions on intercourse increase as one goes
12 Handy
(Polynesian Religion, p. 318) sees a common element in mana
and the Vedic brahma and says: ' Similarly, 1 believe the tapu system as a
body of rules to prevent improper contact of the superior divine with the
inferior common or corrupt, to be an example of divergent evolution from
the same system out of which came also the Indian caste, with its iintouch-
ahleness, eating rules, purifications/
13 Caste, Race and Religion in Man in India, xiv, no. a, p. 136,
quoting Capt. J. S. F. Mackenzie in The Indian Antiquary, it, p.
65 (1873)*
1^ Ibid., p. 137.
13
p. 87 supra,
1 S. C. Roy, Caste, Race **nd Religion % in Man in Indian xvn, no. 4,
p. 24a n., quoting Thurston (Castes and Tribes, iv, p. 126).
1^ Ibid., p. 332.
CONCLUSIONS 189

from the top to the bottom of the caste system. The Brahman,
Vania, Rajput ana Kanbi castes must marry their girls within
their caste or subcaste, but they may eat together within
the large circle of their whole caste. . . . But the artisan and
depressed classes are both with regard to food and
more strict

marriage. Reasons have been given for supposing that castes


of some sort preceded the impact of the Rigvedic invaders, but
it is not suggested that these invaders did not play an important
part in developing the system as it exists today. On the con-
trary, it is urged emphatically that the Indian caste system is

the natural result of the interaction of a number of geographical,


social, political, religious, and economic factors not elsewhere
found in conjunction. No doubt ideas of magic, mana, taboo
and soul-substance were not wanting among the Indo-Europeans
themselves. Parsi priests have to undergo elaborate ceremonies
of purification, and while in a state of purity must eat no food
cooked and drink no water drawn by anyone but a man or
woman of the priestly class.^ With Parsis eating and drinking
*

are religious rites. The mana principle appears in the


Buddhist religion as iddhi and in Islamic beliefs as kudrat.^^
In Hinduism it is familiar as sakti. It seems likely, however,
that it was largely the social and political impact of the Rigvedic
invaders with their definitely graded social classes that was
responsible for introducing the principle of social precedence
into a society already divided into groups isolated by taboos.
The general Hindu feeling about the caste system is that
it has been
established by divine ordinance
or at least with
divine approval. Reference has already been made to the well-
known lines in the Bhagavad Gita which inculcate the supreme
merit of performing caste duties, which take precedence of all

other obligations, including even those of friendship and


kindred. Perfection, it is said, is only attained by the man who
18 Desai,
Hindu Families in Gujarat (Baroda, 193^), quoted by Roy in

Caste, Race and Religion (Man in India, xiv, no. si, p. 147). According

to Dalton (Ethnography of Bengal, p. 160 n.) some Kharias are excessively


particular about the rules of pollution in regard to cooking vessels: He

may not allow even his wife to cook for him, and if a stranger enters a
house in which he keeps his earthen drinking and cooking vessels, and
water pots, every vessel is polluted and the whole are destroyed or thrown
away.
18 Roy, Caste, Race and Religion, Man in India, xiv, no. s, p. 177.
20 Modi, Anthropological Papers, pt. u, pp. 63, 64, quoted by Roy, foe. cit.
21 Roy, loc. cit., 4). 155. 22 Abbott, Keys of Power,
p. 3.
IQO ORIGINS
does not deviate from the duties of caste. Observance of caste,
therefore, is equivalent to dharma, that is, to religious observance,
righteousness, moral obligation. Dharma is not easy to translate
accurately, but the term Dharmeswar the Righteous One? is

used for God. In any case it is clearly the result of this teaching
that social habits, caste customs, are inextricably tied up with
religion. The Brahmanic codes have insisted that every com-
munity should obey its own rules. That insistence took place
first a stage in social evolution at which law consisted, largely
at
at any rate, of a code of taboos. Hence the development erf
the caste system and the unusual success with which the growth
of a plural
society
' was attended in India. Inevitably
communities incorporated into such a society would overlap; a
person belonging to one group by tribal descent might belong
to another by occupation. Hence, perhaps, castes of mixed
origin like the Prabhu, Karan, or Kayastha, not to say Brahman.
Hence also the apparently irrelevant association of caste with
religion, coupled with a great variety of sanctions.
By way of conclusion an attempt may be made to recapitulate
a number of the more obvious factors which have been indicated
as probably contributing to the emergence and development of
the caste system.
The geographical isolation of the Indian peninsula as a whole
and of individual areas within it.

Primitive ideas about the power of food to transmit qualities.


Similar ideas of totemism, taboo, mana, and soul-stuff, or life-
*matter.
Ideas of pollution, ablution, purification, and ceremonial
purity with associations of ritual sacrifice.
Ideas of the exclusive family, ancestor worship, and the sacra-
mental meal.
Beliefs in reincarnation,and in the doctrine of karma.
Belief in magic associated with craftsand functions.
Hereditary occupations, and trade and craft secrets.
Guilds and associations of that character and various factors
in the development of economic life.
Clash of antagonistic cultures, particularly between cultures
with matrilineal and patrilineal modes of descent.
Clash of races, colour prejudice, and conquest.

Vide supra, pp. 115, sq.


CONCLUSIONS igi

The development of classes with exclusive religious and social


privileges.
Individual isolation of tribes and polities, and their inclusion
without absorption in larger administrative units.
Deliberate economic and administrative policies.
Exploitation by a highly intelligent but by no means entirely
altruistic hierarchy which had evolved a religious philosophy too
subtle for the mass of the people.
Even in this inclusive list, however, it is difficult to feel at
all sure that all more important factors contributing to the
the
emergence of that unique social phenomenon, the caste system,
have been included. Probably enough emphasis has not been
laid in previous theories of the origin of caste on the geographical
aspects of the situation. The fact is, many roads of migration

have led into India and have ended there. This has resulted
in the accumulation of a large number of societies of very
different levels of culture and very varying customs in an area
in which they have neither been mutually inaccessible nor with-
out some measure of individual isolation. The mere inescapable
necessity of finding a modus vivendi on the part of a number
of different cultures has probably played a not unimportant part
among the various factors that have combined to cause the caste
system to develop. At the same time it is pretty clear that in
previous attempts to account for caste insufficient importance
has been attached, not only to geographical and migrational
considerations and to the clash of matrilineal and patrilineal
societies, but more particularly to that complex of beliefs in
mana, taboo and magic which surrounds the primitive philo-
sophy of soul-stuff or life-matter which is so widely spread and
so deeply seated in the mainland of south-east Asia and in the
islands of Indonesia and the Pacific. To illustrate the strength
of those ideas in India up to the present time an appendix
on their relationship to Hinduism has been added to the account
of caste given in the foregoing chapters, though the higher
esoteric philosophy of Hinduism is treated as entirely outside
the scope of this volume.
PART FOUR
APPENDICES

APPENDIX A

THE POSITION OF THE EXTERIOR


CASTES
The substance of this Appendix has been taken with some
abridgement, but with little other change, from Appendix I to
my Report on the Census of India, 1931. No attempt has been
made here to deal with events that have taken place since 1931,
but the position of these castes had probably altered very little
between that year and the outbreak of war with Japan in 1941.
Since that war came to an end, change has probably been more
rapid, but from what it is possible to learn from correspondents
in India and from articles in Indian periodicals, the rate
of change is a great deal slower than many Indians appreciate,
and in the more remote areas, where communications are slow
and difficult, it has probably been of a negligible quality in
spite of new laws and of genuine attempts to help the exterior
castes, particularly on the part of the Central Government. At
the same time it has been a very remarkable and striking cir-
cumstance that the new Constitution for India should have
been piloted through the assembly by a person who belonged
himself to one of the exterior castes, and a circumstance that
cannot be without real and weighty significance for the future.

The term *
exterior *
for the Hindu castes hitherto known as
*
depressedwas originally suggested by the Census Superinten-
'

dent for Assam ana was adopted in the report ^ as the most
satisfactory alternative to the unfortunate and depressing label
'Depressed Classes'. It has been criticized as being the same
term as outcaste only of five instead of two syllables, and it

must be admitted that 'exterior' is but old 'out' writ large.


At the same time it is here submitted that outcaste, with an e,
has not unnaturally attracted to its connotation the implica-
tions of the quite differently derived outcast, with no e. Out-
caste correctly interpreted seems to mean no more than one who
is outside the caste system and is therefore not admitted to
Hindu society, but since in practice the exterior castes also con-
tained those who had been cast out from the Hindu social body
1 i.e. Report on the Census of India, 1931.
POSITION OF THE EXTERIOR CASTES 193

forsome breach of caste rules, outcaste and outcast were in


'

'

some cases synonymous, and the derogatory implications of


obliquity attaching to the latter term have unjustly coloured
the former, a taint which is not conveyed by the substitution of
the word exterior

which may connote exclusion but not
*
extrusion. The term a-varna, without varna or outside the
four varnas (vide pp. 64 sqq.), is sometimes used and aptly
expresses the same idea.
The instructions of the Government of India for the taking
of the 1931 census concluded with the following enjoinder:

The Government of India also desire that attention should
be paid to the collection of information conducive to a better
knowledge of the backward and depressed classes and of the
problem involved in their present and future welfare.'
In that connexion the following instructions were issued to the
various Superintendents of Census Operations [1931] in India:

For this purpose it will be necessary to have a list of castes
to be included in depressed classes and all provinces are asked
to frame a list applicable to the province. There are very
great difficulties in framing a list of this kind and there are
insuperable difficulties in framing a list of depressed classes
which will be applicable to India as a whole.'
A subsequent instruction ran as follows:
*
I have explained depressed castes as castes, contact with
whom entails purification on the part
of high-caste Hindus.
It is not intended that the term should have any reference to
occupation as such but to those castes which by reason of
their traditional position in Hindu society are denied access
to temples, for instance, or have to use separate wells or are
not allowed to sit inside a school house but have to remain
outside, or which suffer similar social disabilities. These dis-
abilities vary in different parts of India, being much more
severe in the south of India than elsewhere. At the same time
the castes which belong to this class are generally known and
can in most parts of India be listed for a definite area, though
perhaps the lists for India as a whole will not coincide.
The question of the preparation of lists for each province was
discussed at a meeting of the Superintendents of Census Opera-
tions in January 1931 before the census took place. It was
agreed that each province should make a list of castes who
suffered disability on account of their low social position and
on account of being debarred from temples, schools or wells.
No specific definition of depressed castes was framed and no
more precise instructions were issued to the Superintendents of
Census Operations, because it was realized that conditions
varied so much from province to province and from district to
13
194 APPENDIX A
district even, within some provinces, that it w^ould be unwise
to tie downthe Superintendent o Census Operations with too
meticulous instructions. The general method of proceeding
prescribed was that of local inquiry into what castes were held
to be depressed, and why, and the framing of a list accordingly.
It was decided that Muslims and Christians should be excluded
from the term depressed class' and that, generally speaking,
hill and forest tribes, who had not become Hindu but whose
religion was returned as Tribal, should also be excluded, and
in the numbers of the exterior castes given below these prin-
ciples have been followed. A
note on the depressed and back-
ward classes in Assam submitted to the Franchise Committee by
the Superintendent of Census Operations for that province
affords a very clear example of the way in which these
principles were intended to oe applied and have been applied
by Superintendents of Census Operations, and an extract from
it is given towards the end of this appendix.
Both for social and for political reasons it is obviously neces-
sary to know the number of these classes not only in India as a
whole but also in different provinces. The matter is of impor-
tance not only with refei'ence to their representation in the body
politic, but also with reference to any social work that is to be
done towards raising them from their present backward position
to one more nearly comparable with that of more advanced
social groups.
The Census Commissioner in 1921 {Census of India, 1921,
voL I, part I, para. 193) gave what he describes as minimum
numbers of the depressed classes in various provinces, making
a total of 52,680,000. This figure, he states, must be taken as
a low estimate, since it does not include all those who should
have been included, and he says: *We may confidently place
the numbers of these depressed classes, all of whom are con-
sidered impure, at something between 55 and 60 millions in
India proper. Of the 52|- million for which the Census Com-
missioner gave actual figures, less than million were to be
found in British India. This figure agrees fairly well with the
42 million odd given as the figure of de}>res.sed classes by the
Franchise Committee of 1919. It is also not greatly at variance
with the 44^ million estimated by the Nair Central Committee
of 1929 as the figure of depressed classes in British India, but it
varies very considerably from the Hartog Committees figure of
approximately 30 million. Clearly it is time that some more
definite figures were obtained than the estimates hitherto
employed. There however, a considerable nitmber of
are,
difficulties in arriving at a determined figure.
The definition to be used in arriving at the figure of depressed
classes is a very difficult matter. The following possible tests
are to be considered:
ig6 APPENDIX A
on the position of being the one to leave the road. If, how-
A
ever, possible for A and his friends^ by boycotting B and
it is

his friends for certain purposes, to bring pressure on B to dis-


regard his legal rights and to conform to A's religious prejudices
and leave the road whenever A is seen at a distance, clearly B
has in practice no freedom of action in the matter of the road,
whether his religious scruples are involved or not. This ques-
tion of the use of roads has been taken as an illustration, but
in point of fact the restriction of the use of roads is one which
seems to be generally disappearing and has possibly disappeared
to such an extent that the question may be ignored as far as
British India is concerned. The use of wells, however, is
another matter, and the disability of the exterior castes varies
from not being allowed to approach the village well at all to
the position, common in Bengal, in which persons of certain
castes may not draw water themselves but must await someone
of a clean caste, who draws water for them at the well. The
question of schools is another very real problem for the exterior
castes, since in many parts of India if tney sat inside the school
they would be made to suffer in some other way by the higher
castes using the school; and whereas the acquisition of reading
and writing, at least, may be taken for granted in the case of
the children of any Brahman, and of other castes as well, it is
an exception in the case of the exterior castes, the presence of
whose children is disliked in the school by their social superiors
and whose children, if they read at all, must sit outside in the
sun and dust. It is often argued that untouchability is merely
dependent on the occupation, so that an untouchable person
or caste abandoning an unclean occupation becomes touchable.
This may be true in a literal sense, but it is not true morally,
since members of exterior castes who may have abandoned their
traditional calling for two or three generations are still liable
to be treated as outside the pale of decent society, and their
presence is apt to be regarded as an offence by members of
interior castes, while they would not be ordinarily admitted to
social functions on a footing of eauality.
For purposes therefore of deciaing what persons are to be
included in the numbers of the exterior castes, it has been necess-
ary for each province to deal with the problem in its own way
and to produce its own list. It is not possible to say generally
that such and such a caste is exterior to Hindu society and to
apply that dictum to the whole of India. It may be possible
to do so in the case of certain castes, such as those of Dom and
Bhangi, but it certainly is not the case that a caste which is
depressed in one part of India is depressed everywhere. Conse-
;quently, each provincial superintendent has had to draw up
his own hst, taking into account the various factors enumerated
above, and to reckon as depressed only those castes or classes
POSITION OF THE EXTERIOR CdSTES 197

who definitely suffer from serious social and political handicap


on account of their degraded position in the Hindu social
scheme. So much is this the position that, in the Central
Provinces for instance, the castes to be treated as depressed for
purposes of census figures have varied from district to district,
and no list is possible at all which is applicable to the whole
of the province for all the castes concerned. Many castes and
tribes who would be included by some, at any rate, of the tests
mentioned above have been excluded from the list on the
ground that they suffer no tangible disability as a result of
their inferiority in the Hindu system. Many primitive tribes,
for instance, are in an ambiguous position by reason of their
not really being Hindus at all. Such tribes, when they come
within the Hindu system, often become automatically depressed,
largely on account of the fact that they have no prejudices
against occupations and food which are taboo to interior
Hindus. On the other hand, the socially superior individuals
of these identical tribes are very frequently able to get them-
selves incorporated into the Hindu system as Rajputs or Ksha-
triyas, though their fellow-tribesmen may remain exterior. In
some cases, however, a complete tribe has succeeded in estab-
lishing its claim to a more or less equivocal twice-born status.
In these circumstances, therefore, non-Hinduized hill and forest
tribes have been excluded from the total of the exterior castes,
as, until they reach the stage of incorporation in regular Hindu
society, they do not really suffer by their potential position in
that scheme. Similarly, criminal tribes have not been included
unless their condition be such that, even if they ceased to be
criminal by habit or profession and lived as peaceable and law-
abiding citizens, they would still be depressed on account of
their social position as distinct from their occupational stigma.
Again, there are numbers of castes which, though they are
regarded by interior Hindus as ceremonially polluting and as
such that from their hands water cannot be taken, have in many
cases such strong caste organizations and include so many indi-
viduals of substance and education that they have built up for
themselves a strong position which obviates the need of any
special measures for their social, political or religious protec-
tion. It is these considerations which have caused the Census
Superintendent of Cochin State to exclude the Iruvan from the
number of those depressed, while the Shaha, Teli and Maliishya
of Bengal and Assam would themselves protest at any such
inclusion.
The note on the Depressed and Backward Classes in Assam
by the Census Superintendent for that province, which forms
an appendix to the Assam Census Report [1931], affords an
excellent example of the facts and considerations taken into
account in determining what constitutes a depressed caste. In
igS APPENDIX A

the case of Assam, the numbers of the exterior castes are so


high in comparison with the total number of interior Hindus,
or rather the number of interior Hindus is so comparatively
small in a province in which the number of backward classes
and hill tribes is high, that the disabilities attaching to depres-
sion are slight compared with those in most provinces, but the
be taken into account are the same everywhere. The
factors to
non-Hindus and the hill and forest tribes included by the
Assam Census Superintendent as backward classes have been
excluded from the figures given in the table opposite, which
amount for all India to just a little over 50 millions.
These then being the numbers of the depressed classes in
India, some estimate is necessary of their positiofi at the present
time [1933], and it has already been pointed out that their dis-
abilities can be roughly divided into two categories. First, that
under which they are barred from public utilities, such as the
use of roads and tanks, and secondly, their religious disabilities
which debar them from the use of temples, burning grounds,
mats and some other institutions. In addition to the above,
but arising out of the second of these, there are the disabilities
involved in relatioirs with private individuals, such as the ser-
vices of barbers and the admission to tea-shops, hotels, or
theatres owned by private individuals. A resolution of the
Anti-Untouchability Conference of 1929 in Madras regretted

to note that in Restaurants, Coffee Hotels, Hair-Dressing


Saloons, Water Pandals, etc., notices are hung prominently ex-
cluding the untouchables , and the Census Superintendent for
the Central Provinces writes: The fact that a sweeper may sit

beside a high-class Hindu in a railway carriage or a motor-


lorry without any question of his right to do so has not yet
made it any more easy in the interior for a touring officer to
persuade cartmen of some castes to carry his sweeper from
camp to camp. In fact, in certain districts it is always essential
to employ at least one cartman of humble caste for thi.s
purpose.
Theoretically perhaps the admission to Hindu temples would
be enough, once it is conceded, to remove all the other dis-
abilities, for tire temple is not merely a religious institution but
is also in many ways a social one, for the term must be taken to
include such buildings as namghars in Assam, which, as temples
do in some parts of India, serve as a village hall or a town hall
for the public generally. A temple also contains a school, so
that the absence of the right of entering the temple may debar
an individual from the possibility of attending the school. In
some Hindu temples, such as the Parbati Temple in Poona,
the Kalarama Temple at Nasik, the Hajo Temple in Assam,
it has been in the past the practice to admit non-Hindus,
such
POSITION OF THE EXTERIOR CASTES 199

Percentage of
exterior castes on Percentage
Total Total of exterior
Total
Province or State exterior castes who
population Hindu castes Hindu Total are
popula- popula- literate
tion tion

India 350,529,557 239,195.140 50,195,770 21 14 1.9


Provinces 271.431,549 177,727,988 39,064,009 22 14 1.6

1. Ajmer Merwara 560,292 434,509 76,816 18 14 2.2


2. Andamans and Nicobars 29,463 7,618 512 7 2 ?
3. Assam 8,622,251 4,931,760 1,829,009 37 21 3.1
4. Baluchistan (Districts &
Administered Territories) 463.508 41,432 5,702 14 1 6.9
5. Bengal 50,114,002 21,570,407 6.899,809 32 14 5.0
6. Bihar and Orissa 37,677,576 31,011,474 5,744,393 19 15 0.6
7. Bombay 21,854,866 16,621,221 1,750,424 11 8 2.8
8. Burma 14,647,497 570,953 No return
of caste
9. Central Provinces & Berar 15,507,723 13,338,223 2,818,346 21 18 1,5
10. Coorg 163,327 146,007 24,803 17 15 1.5
11. Delhi 636,246 399,863 72,883 18 11 1.4
12. Madras 46,740,107 41.277,370 7,234,104 18 15 1.5
North-West Frontier
13.
Province 2,425,076 142,977 5,468 4 - 3.6
14. Punjab 23,580,852 6,328,588 1,279.459 20 5 0.8
15. X/nited Provinces of
Agra and Oudh 48,408,763 j
40,905,586 11,322,281 28 23 0.5

States and Agencies 79,098,088 61,467,152 11,131,761 18 14 8.1


16. Assam States 625,606 272.890 1,421 12.9
17. Baluchistan States 405,109 12,249 20' '

?
18. Baroda State 2,443,007 2,152,071 203,043 9 8 10.3
19, Bengal States 973,336 641,662 30,822 5 3 ?
20. Bihar arid Orissa States 4,652,007 4,194,878 631,864 15 14 1.0
21. Bombay States 4,468,396 3,921,088 348,574 9 8 2.8
22. Central India Agency 6,632,790 5,852,204 797,902 17 12 0.3
23. Central Provinces States 2,483,214 1,788,401 252,732 14 10 0.5
24. Gwalior State 3,523,070 3,271,576 678,119 21 19 ?
25. Hyderabad State 14,436,148 12,176,727 2,473,230 20 17 0.6
26. Jammu &
Kashmir State 3,646,243 736,222 170,928 23 5 0.5
27. Madras States Agency 6,754,484 4,323,150 1,960,370 45 29 13.8
Cochin State 1,205,016 780,484 125,339 16 10 4.8
Travancore State 5,095,973 3,134,888 1,769,735 56 85 14.9
Other Madras States 453,495 407,778 65,296 16 14 8.5
28. Mysore State 6,557,302 6,015,880 1,000,326 17 15 1.4
29. North-West Frontier
Province (Agencies and
Tribal Areas) 46,451 13,651 542 4 1
30. Punjab States 437,787 383,883 94,347 25 221
0.3
31. Punjab States Agency 4.472,218 1,887,249 392,999 21 9)
32. Rajputana Agency 11,225,712 9,578,805 1,565,409 16 14 0.4
33. Sikkim State 109,808 47,074 2,029 4 2 ?
34. United Provinces States 1,206,070 950,724 208,864 22 17 i).2
35. Western India States
Agency 3,999,250 3.245,768 318,220 8 1.9
200 APPENDIX A
and Muslims, to a point
as Christians to which the untouchable
Hindus were never admitted at alL Itis, however, not quite

certain how farunder present conditions the actual right of


admission to a temple would, if conceded, remove the social
disabilities of the depressed castes, since it might have the effect
of merely driving the higher castes to shun the temples to which
the untouchables were admitted ^ In any case the right varies
much in different places. An Iruva or a Tiya in Malabar must
stay 325 feet from the curtain wall of the temple of Guruvayur,
for instance, and this wall encloses a space 350 feet square
with the temple in the centre. Yet the Iruva have not been
treated, in Cochin State at any rate, as a depressed caste at all,
since, though deprived there likewise of temple entry, they are
otherwise well-to-do and not ill-educated.
As regards the civil rather than the religious disabilities to
which the depressed classes are at present subjected; the first
one mentioned above is the right to use the public roads; up
to how recently the use of public roads has been debarred in
certain cases may be gathered from the fact that the untouch-
ables of Travancore made an organized attempt in 1924 to
obtain the use of roads which skirted the temple at Vaikom.
These roads were public roads, maintained by the State for
the use of everybody, but, on account of their proximity to
the temple building, the untouchables were not allowed to
use certain sections which skirted the temple too closely. Ulti-
mately, as a result of satyagraha, the temple compound was
enlarged and the ban on the roads was removed, the roads
having been realigned so that their users were no longer within
the polluting distance of the temple. In 1926, and again in
1930, a similar case occurred in Travancore in connexion with
Sachindram Temple, which is one of the richest in the State.
Here again the depressed classes wanted the right to use a road
maintained by public funds and belonging to the State. In
the matter of the use of roads generally, however, the depressed
castes are no longer in the position in which some of them were
when the Mangalore District Gazetteer was written, when the
Ande Koraga had to carry a spittoon round their necks as being
so highly polluting that they could not be allowed to expecto-
rate on the public road. It has recently been reported however
that a caste has been found in the Tamilnad, the very sight
of which is polluting, so that its unfortunate members are
compelled to follow nocturnal habits, leaving their dens after
dark and scuttling home at the false dawn like the badger,
hyena, or aard-vark. The Hindu of 24 December 1932 writes
of them as follows 'In this [Tinnevelly] district there is a class
:

of unseeables called Purada Vannans. They are not allowed to


2 This in fact is reported to have been the result
in Malabar, where
most well-to-do households have a private shrine.
POSITION OF THE EXTERIOR CASTES 201

come out during day-time because their sight is considered to


be pollution. Sdme of these people, who wash the clothes
of other exterior castes, working between midnight and
daybreak, were with difficulty persuaded to leave their houses
to interview Mr A. V. Thakkar, who described them as coming
only after repeated persuasion and thexi with
their whole
bodies shaking and trembling
The case of wells has already been alluded to, and is a far
more widespread and real grievance than any which may still
survive in regard to the use of roads. Generally speaking, if
the exterior castes have succeeded in asserting their right to use
public wells, the higher castes have given them up. Here again
the difficulty about the use of wells will be found to be most
prevalent in the drier parts of India where water is scarce. In
Assam this difficulty is not worth mentioning, and in Bengal,
where it exists, it is usually got over by water being drawn for
the exterior castes by some interior-caste Hindu. This arrange-
ment is obviously open to certain drawbacks, but it is not so
serious as the custom in many parts of upper India and
southern India which prohibits the exterior castes from having
water at all from the well which is used by interior Hindus.
The same applies to the use of dh&ramshdlas and of public
burning ghats and burial grounds, in regard to all of which the
position of the exterior castes is much the same as it is in
regard to the use of wells.
With regard to schools, the Director of Public Instruction for
Bombay, reporting for the year 1928-9, remarks that admission
to schools was not refused to the children of the depressed
classes in the schools under the control of any local body except
the District Local Board at Ratnagiri, the District Local Board
of the Nasik District (some schools only), and in the
Ahmedabad and Surat districts. In the latter districts the
students generally used to sit in temples, dharamsfialas or
private houses and the pupils of the depressed classes^ were
objected to, but the objection was withdrawn on warning being
given that in this way the schools would lose any grants that
they enjoyed from the Government. How far this withdrawal
of objection was effective is doubtful, since at any rate in one
case the depressed classes at Surat had to withdraw their children
from the school as a result of the indirect pressure exerted on
them by higher castes. Similarly, in some cases the depressed
caste pupils find it better to sit outside the schools, as, if they sit
inside, they are boycotted and compelled to leave the school.
Thus at Kaira in April 1931 some Dheds took their seats with
other Hindu boys in the municipal school. Kaira had been one
of the centres of Mr Gandhi's activities and no objection was
raised on that day, but when the parents of the interior-caste
3 Cf. p. 81, n. 31.
205 }
APPENDIX A
boys heard of it they threatened a boycott of the school and of

the teacher ifthe Dhed boys were allowed to sit with theirs, or
even to occupy the same classroom, with the result that the next
day the Dheds were refused admission to the school premises.
On the other hand, in Sind and in the central and southern divi-
sions of Bombay it was reported that there were no primary
schools managed by the local authorities which refusea admis-
sion to the children of the depressed castes. Similarly, in Assam
no inconvenience appears to have been experienced by the ex-
terior castes in the matter of school attendance. On the other
hand, in many parts of India the inconvenience is greater than
it is in Bombay. Very few of the exterior castes attend schools
nominally accessible to them in Negapatam, Kumbakonam,
Tinnevelly, Cocanada, Bezwada and Narsapur and other towns
in southern India, though in Madras, Madura, Sivaganga and
some other towns a number of children of the exterior castes
attend schools which are not especially reserved for them and
which are not boycotted by the higher castes. In most parts of
southern India it is necessary to have special schools for the
exterior castes, since it is not yet possible to induce the higher
castes to study in their company. In July 1931, when it was
decided to admit exterior castes into all the aided schools, a
number of schools had to close, and from some other schools the
higher caste children were withdrawn. Similarly, in Baroda State
the abolition of separate schools in November 1931 is reported
to have caused great resentment among the caste Hindus,
who in some cases withdrew their children from schools and in
others destroyed the crops of the exterior castes or poured
kerosene oil into the wells used by them. In Bengal the Rural
Primary Education Bill, passed in 1930, appears to have been
opposed by members representing the caste Hindus, and it is
alleged that this opposition was aimed at depriving the non-
caste Hindus, and also the poorer Muslims, of the benefits of
literacy. In Cochin State, on the other hand, much has been
done to open all educational institutions to the exterior castes,
though this has involved in some cases the removal of the
school to another site, while cases have occurred of some ill-
treatment of the castes now admitted to the schools. Generally
speaking, however, during the last decade [1921-31] the exterior
castes at school in Cochin increased in number from some 1,500
to some 14,000 and out of 700 recognized schools only three
were still reserved to the higher castes in 1931, and a Protector of
the backward and depressed classes had recently been appointed.
Cochin, however, is probably in many ways exceptionaL
In regard to the matter of the right to enter Hindu temples,
the exterior castes were advised by Mr Gandhi not to attempt to
gain entry, as God resided in their breasts. A
temple, however,
as has been pointed out, is more than a purely religious insti-
POSITION OF THE EXTERIOR CASTES 203

tution, and the


right of temple entry is by some regarded as
the key position with regard to the removal of untouchability.
The claim to enter temples is not opposed only by high-caste
Hindus; even Justice, the organ of the anti-Brahman Party,
writes as follows:
'
For many centuries these peoples, most of whom until
recently were Animists, were content to worship at. their own
shrines, and to try to force themselves into Hindu temples is
not ... to make themselves popular. Nor can we think that
any grave wrong is done by their continued exclusion they. . .

would be better occupied in improving their own condition


than in a violent attempt to assert rights which no one had
heard of till a few years ago.'

There on the other hand, a definite movement among


is,

many more advanced Hindus to remove the ban on temple


entry,- a movement which has since the census (of 1931) been
given much impetus by Mr Gandhi's fast and the negotiations
following it for the admission of untouchables to Hindu temples
generally, and it may be mentioned as an instance of this that
eight temples of a Telugu community in Bombay were reported
to have been opened to untouchables in February 1930. On
the other hand, recent cases of attempts to obtain entry by
exterior castes had led to violence. Some 2,000 untouchables
collected outside the Kala Ramtemple at Nasik on 3 March
1930, and a meeting which attempted to bring about a settle-
ment was stoned by the orthodox. Ultimately some 150,000
Mahars and Chamars were reported to have collected at Nasik
and the temple had to be closed for about a month to keep
them from entering it. The admission of caste Hindus by a
private passage ended in violence in which the orthodox were
the aggressors, and which was extended to Mahar villages in
the neighbourhood, where the exterior castes were violently
attacked by caste Hindus, their wells polluted and in some cases
their houses burned. A
similar attempt to force an entrance a
year later had similar results, and in December 1931 the same
sort of situation arose on an attempt of the exterior castes to
bathe in the Ramakund, the sacred pool at Nasik, and again
in April over the dragging of the rath. Similarly at Singanallur
in south India the question of entry to a temple led to a free
fight in 1930, and at some other places also. In Nagpur a
temple was voluntarilv opened to untouchables, and in the
Dacca Division in Bengal a satyagraha of 9 months' duration
ended in the temple doors being forced open by a band of high-
^ A Independence Day
Congress
pledge runs:
'

We know that the
distinction between Caste Hindus and Harijans [i.e. the exterior castes]
must be abolished and a Brahman political leader has actually declared:
We want intermarriage to make the people biologically one
04 APPENDIX A

caste women who sympathized with the satyagrahis. This was


in May, but it was reported in October that the temple was
deserted by high-caste Hindus. One wonders therefore how
far the higher castes are likely to make use of a pan-Hindu
temple recently endowed at Ratnagiri with a view to providing
a common place of worship for all castes.
The prohibition against exterior castes entering the Hindu
temples naturally raises the question whether they can really
be called Hindu at all. Generally speaking, the answer must
be that they are definitely Hindus; they worship the same
deities and, though they are refused entry to the temples, boxes
are placed outside, at the limits to which they can approach, to
receive their offerings. The degree of Hinduism does, however,
vary considerably; tnus the Mahar and Chamar in general are
very decidedly Hindu, whereas the Chuhra of the Punjab is
very doubtfully so, taking a Hindu tone when living by a
Hindu village, a Muslim tone in a Muslim village, and that of
a Sikh in a Sikh village. Possibly the Chuhra should really be
described as tribal by religion, and the possibly connected
Chodhra of the Bombay Presidency are definitely regarded as a
forest tribe rather than a caste. In any case, in the Punjab
the question has been partly solved at this census by the Chuhra
returning himself as an Ad-Dharmi, that is to say, a follower
of the original religion. For him it means the customs imme-
morially observed by his caste or tribe, but of course the term
might mean very different things to different peoples. In the
Punjab Its use as distinct from Hindu (in other provinces the
exterior castes have frequently returned themsdves as Adi-
Hindu, Adi-Dravida, etc., with a similar implication to that of
Ad-Dharmi) is probably a political expedient in order to obtain
more effective representation on the provincial legislative body,
and, in spite of the occasional use of the word Hindu' with
reference to caste Hindus and excluding the exterior castes, it
must be held that these castes generally are Hindu by religion
even if they are not Hindu socially, hence the expression ex-

terior caste is suitably applied to castes who follow the Hindu


'

religion, but are not admitted to Hindu society.


This social, bar tends to foster conversion to the Sikh faith,
to Islam, or to Christianity, though even after conversion the
social stigma does not vanish at once. This is hardly to be
expected. The Mazhbi Sikhs are looked down upon by the
Sikhs who are not Mazhbi. The southern Indian Christians
distinguish between the castes of their converts in their seating
fi
A reviewer of the first edition of this book, writing in the Journal of
the Royal Asiatic Society, drew attention to experience of caste in the
.

armed services in the war of 1939-45. He stated that while Hindus of


varying castes were willing to feed with each other and with adherents of
other religions, they were not prepared to do so with men recruited from
POSITION OF THE EXTERIOR CASTES 05

accommodation in churches, and the dislike of the exterior


castes does not imm ediately disappear when they turn Muslim.
At the same time, once they are converted it does not take them
very long before they can rise in the social scale, and in the
case of Indian Christians in southern India a period of about
three generations often, perhaps usually, sees them accepted as
the equals of previous converts of most castes.

There are other points in which the exterior castes suffer


socially.Thus exception is taken to their wearing the orna-
ments usually worn by higher castes, and in some cases they are
not allowed to wear gold ornaments at all. Cases are on record
in which Chamars, for instance, have been beaten for dressing
like Rajputs, and the mounting of an exterior caste bridegroom
upon a horse for his bridal procession has led to a boycott of the
caste in question by the higher caste neighbours. In December
1930 the Kallar in Ramnad propounded eight prohibitions, the
disregard of which led to the use of violence by the Kallar
against the exterior castes, whose huts were fired, whose grana-
ries and property were destroyed, and whose livestock was looted.
These eight prohibitions were as. follows:
Vi) that the Adi-Dravidas shall not wear ornament of gold
and silver;
(ii) that the males should not be allowed to wear their clothes
below their knees or above the hips;
(iii) that their males should not wear coats or shirts or
baniyans;
*(iv) no Adi-Dravida shall be allowed to have his hair
cropped;
(v) that the Adi-Dravidas should not use other than earthen-
ware vessels in their homes;
their women shall not be allowed to cover the upper
(vi)
portion of their bodies by clothes or ravukvais or thavanies;
Vvii) their women shall not be allowed to use flowers or
saffron paste; and
*(viii) the men shall not use umbrellas for protection against
sun and rain nor should they wear sandals.
In June 1931, the eight prohibitions not having been satis-
factorily observed by the exterior castes in question, the Kallar
met together and framed eleven prohibitions, which went still
further than the original eight, and an attempt to enforce these
led to more violence. These eleven prohibitions were:
the exterior castes. He went on to say that while such persons were
often enlisted in ordinary units formerly, even the exigencies of the recent
war have permitted only of their being taken into special formations
And from this he infers that untouchability seems likely to endure as

the vital criterion It had however long been the practice to enrol
Mazhbi Sikhs, who are Chuhra by extraction, in separate pioneer units.
ijo6 APPENDIX A
*
1. The Adi-Dravidas and Devendrakula Velalars should not
wear clothes below their knees.
*
2. The men and women of the above said depressed classes
should not wear gold jewels.

3. Their women should carry water only in mud pots and
not in copper or brass vessels. They should use straw only to
carry the water pots and no cloths should be used for that
purpose.
*
4. Their children should not read and get themselves literate
or educated.
'
5. The children should be asked only to tend the cattle of
the Mirasdars.
6. Their men and women should work as slaves of the
Mirasdars in their respective Pannais.

7. They should not cultivate the land either on waram or
lease from the Mirasdars.

They must sell away their own lands to Mirasdars of the
8.
village at very cheap rates, and if they don't do so, no water
will be allowed to them to irrigate their lands. Even if some-
thing is grown by the help of rain-water, the crops should be
robbed away, when they are ripe for harvest,
*
9. They must work as coolies from 7 a.m. to 6 p.m. under
the Mirasdars and their wages shall be for men Re. 0-4-0 per
day and for women Re. 0-2-0 per day.
10. The above said communities should not use Indian
Music (Melam, etc.), in their marriages and other celebrations.
11. They must stop their habit of going on a horse in
procession before tying the Thali thread in Marriage, and they
must use their house doors as a palanquin for the marriage
processions and no vehicle should be used by them, for any
purpose.

Similarly, in Bengal the use by a Namasudra bridegroom of


a palanquin in a marriage procession led to a disturbance, and
other instances could probably be quoted from other parts of
India,
It is not suggested that the eight or eleven injunctions of the
Kallar would be taken very seriously by anyone but themselves,
but they are quoted here to indicate the attitude often adopted
towards the exterior castes.
Economically the exterior castes in eastern India are generally
speaking self-supporting and by no means desperately poor.
In western India their position often is that of scavenger or
village menial, for whose service there is a certain amount of
necessity which is recompensed by a traditional provision of
certain village lands or by other methods of payment in kind.
The trouble frequently is that the number of these menials has
become superfluous and they none the less expect to be fed by
POSITION OF THE EXTERIOR CASTES 507

the village for unwanted services, and if they do not get so fed
they steal the grain, no doubt regarding it as theirs by right.
In southern India again the exterior castes are generally derived
from various classes of cultivating serfs who until recently were
tied to the soil. In northern India their economic position
varies a good deal, since leather-workers for instance in industrial
towns find a ready livelihood, whereas the exterior castes in the
agricultural areas, where they can only obtain the worst land and
the worst wells, are very often extremely poverty-stricken.
The origin of the position of the exterior castes is partly racial,
partly religious, and partly a matter of social custom. There
can be little doubt but the idea of untouchability originates in
taboo. Reminiscences of such a taboo are still to be seen in
Burma, where grave-digging is a profession involving a social
stigma of a kind which will not permit of association with
persons of other professions. A comparison of this custom with
the position of those who dig graves for the hill tribes in the
surrounding areas leaves little doubt but that the repulsion
originates in the fear of some sort of death infection, and the
underlying idea is not that the person himself is polluted by
unclean work but that his mere association with death may infect
others, with whom the grave-digger comes into contact, with the
probability of dying. The treatment of washermen all over
India as a depressed caste is almost certainly traceable to a
similar taboo, the objection perhaps arising from an association
with the menstrual clothes of women and consequently an infec-
tion which, in the first instance, is magical, though it later comes
to be regarded merely as a matter of personal cleanliness. The
untouchability which has originated in taboo has undoubtedly
been accentuated by differences of race and the racial antipathies
which seem common to every branch of the human family and
have reinforced the magical taboo. An instance of this sort of
thing may again be found in Burma, which, as there is no caste
there in the Indian sense, is useful as providing indications
of the process of caste formation. Here pagoda slaves, the status
of whom is hereditary, are looked down upon by other classes.
In general, pagoda slaves have been recruited from non-Burmese
races. Large numbers of Arakanese, Talaings, Manipuris and
Siamese have been settled in various parts of Burma by various
conquering monarchs as slaves of different pagodas. This racial
element is probably to be traced again in the prohibition of
the wearing of ornaments by certain castes. Thus in the Ao tribe
in the Naga Hills of Assam one of the sub tribal groups, which
is apparently of different origin to the rest of the tribe, is not
allowed to wear ivory on both arms. Similar restrictions are
found in the Laccadive Islands. The same element probably
appears in the practice of hypergamy, and one of the first steps
which any sectiqri of an exterior caste takes in order to raise its
208 APPENDIX A
scMzial standing is to deny to pther sections the right to marry its
daughters though continuing for the time being to take wives
from among them. Thus have the Haliya Kaibartta severed
themselves from the Jaliya Kaibartta and established themselves
as a separate caste, no longer depressed, and even so a section
of the Namasudra is now attempting to segregate itself. Indeed,
as between different sections of the exterior castes prejudice is
just as strong as between the interior castes and the exterior
castes genermly. The Mahar in Bombay have objected to
sharing their counsels and conferences with Chamars, and
Mahar and Chamar have unanimously spurned the Bhangi,
Similarly, inMadras Pallans have objected to being classed with
Chakli whom they regard as no less inferior to them than the
Paraiyan, though all alike are untouchables to a good-caste
Hindu. Religion, of course, with its apparatus of holy vessels,
sacred animals and sacrosanct priests, has also contributed to the
creation of the idea of untouchability; and society in general
by its natural dislike of certain unclean occupations and by its
very proper antipathy to criminal professions lias done much to
depress and stereotype the position of the exterior castes.
In 1916, and again in 1920, the Government of India called
for a report as to the moral and material conditions of the
depressed classes and for proposals for their amelioration from
The records of the consequent reports
all local governnie^nts-
will be found under Proceedings of the Government of India
in the Home Department, nos. 130-1 of July 1916, 329-41 of
August 1920.
During the decade 1921-31 a good deal was done, particularly
in Madras, to benefit the exterior castes in various ways. Not
only did the Madras Government appoint a Commissioner of
Labour entrusted with the task of encouraging the education
of the depressed classes and of looking after their economic
interests, but many private societies were also at work. Besides
a number of Christian Missions, the Depressed Classes Union,
the Poor School Society, the Social Service League, the Andhra
Deena Seva Sangam, and the Depressed Classes Mission of
Mangalore are some of the institutions working for this object.
The Arcot Mission started an agricultural school, and Local
Boards were also spending money towards the same end. In
addition to this much was done by the Self-Respect movement
and the Depressed Classes Conference. The actual steps taken
by the Government of Madras included, besides the appointment
of a Commissioner of Labour, the insistence on the right of
admission of exterior caste pupils into all publicly managed
schools, the refusal of grants-in-aid to privately managed schools
which did not admit exterior caste pupils, the removal of
publicly managed schools from places inaccessible to exterior
castes, the opening of special schools and hostels for the exterior
POSITION OF THE EXTERIORCASTES iog

castes, the fees, and the provision of scholar-


remission o their
ships and of
special facilities for the training of exterior caste
teachers. On
the economic side the Government of Madras
took steps towards the provision of house sites for the relief
of congestion and for purposes of freeing the' exterior castes
from oppression by their landlords, the organization and run-
ning of co-operative societies, the provision of drinking water by
constructing new wells and repairing old, the provision of burial
grounds and sanitary requirements for the exterior castes, the
assignment of land for cultivation both by reservation for and
free assignment to exterior castes of lands not classed as valu-
able, that is, lands not affected by irrigation projects, and by
reservation for exterior castes of lands classed as valuable and
assignment to them on payment of market value in easy instal-
ments. As a result of these measures some progress was
made. In 1920 there were 150,000 pupils of exterior castes in
schools; in 1930 there were 230,000. Over 100,000 of these were
in Christian Mission schools and many of the remainder in
schools maintained by the societies mentioned above. Most of
these schools are only primary schools, there being only about
2,700 pupils in educational institutions above the primary stage,

of whom about fifty only were in colleges. Out of 230,000 read-


ing in schools in 1930 about 7,500 were girls. Of the 230,000
only 16,000, however, were reading in ordinary schools not
reserved for exterior castes, while over 70,000 pupils who did
not belong to the exterior castes were reading in schools specially
maintained for those castes. As compared with the rest of the
population, while about 6 per cent of the population of Madras
was going to school when the census (1931) took place, only
about 4 per cent of the exterior castes were at school, and of
that total less than 1 per cent was beyond the primary stage.
Economically very considerable sums had been spent by the
Madras Government during the past ten years in financing and
organizing co-operative societies for the acquisition of land for
house sites and for the granting of agricultural loans and for
purposes of flood relief, rural credit and collective bargaining.
In 1920 there were over 14,600 members of exterior castes in
about iOO co-operative societies; as a result of the work done
during the decade there were in 1931 about 2,000 such societies
consisting mainly of exterior castes, and during recent years new
societies have been registered at the rate of over lOO a year.
Over 55,000 house sites have been provided and some 300,000
acres of land have been assigned to exterior caste cultivators.
But at the time of the 1931 census there were still six districts
in Madras untouched by the activities of the Commissioner of
Labour. In Bombay, apart from the work done by private
institutions among which the Servants of India Society is very
'

"

14
aio appendix a
prominent, little has yet been done. In November 1928 a Com-
mittee was appointed by the Government of Bombay to inquire
into the condition of the depressed castes and aboriginal tribes
in the Presidency. As a result of the recommendations of that
Committee an officer was appointed in 1931, charged with the
duty of watching over the progress of, and of upholding the
rights of, the backward classes. In addition to that, a Resolution
of the Government of Bombay directed the recruitment of
members of the depressed classes in the police. It is perhaps
significant that the Committee found it necessary to include in
their recommendations an addition to the Government Servants'
Conduct Rules framed with a view to enforcing polite treatment
of the depressed classes by the officials of Government. Many
of the Indian States have also recognized the necessity for special
measures for the uplift of exterior castes, in particular, Baroda,
Cochin and Travancore; while in Gwalior and Jaipur action
has been taken more particularly in regard to the criminal tribes,
of which there are considerable numbers in those States. In
Jaipur areas of land have been allotted to the criminal tribes
and special schools have been opened for the education of their
children, and the same has been done in Gwalior.
The occasion of the 1931 census, coming as it did at a time
when political reforms appeared imminent, complicated the
already plentiful difficulties in the way of getting an exact return
of numbers of exterior castes. A number of conflicting forces
were at work, as, apart from the natural desire of individuals
of exterior castes to raise their own social status by making
themselves out to be something other than they were recognized
to be by their neighbours, a definite movement was set afoot
by the Hindu Mahasabha for the return of ail Hindus as Hindu
simply, with no qualifications of caste or sect. In 1928 the Hindu
Mahasabha itself passed a resolution declaring that the so-called
untouchables had equal rights with other Hindus to study in
schools, and to use wells and roads and temples, and the same
resolution called on priests, barbers, and washermen to afford
their services to untouchables. At the 1931 census, however,
political considerations probably outweighed all else, and many
eflforts were made to induce untouchables to record themselves
as Hindus and nothing else. The exterior castes themselves
*
'

were, however, generally alive to the fact that their interests


required their numbers to be definitely known; at the same
time they were not unconscious of the fact that it might be
advantageous to them to represent as many castes as possible as
being depressed in order to swell their numbers and importance.
Consequently, while in the Punjab the All-India Shradhanand
Bepressed Classes Mission was calling on the exterior castes to
return themselves as Arya Hindus instead of, for instance, Achuts
or Dalits, the exterior castesVown leaders were calling on their
POSITION OF THE EXTERIOR CASTES nit

followers to return themselves as Ad-Dharmi by religion and

not Hindu at all. In other provinces the associations of the


exterior castes were representing as depressed castes which
*
'

very doubtfully fall into that category and many of which have
been excluded on scrutiny. Generally speaking, however, it is
believed that the figures for the exterior classes obtained at the
1931 census have been accurate on the whole, and the methods
adopted in diflEerent provinces to determine what constitutes an
exterior caste have already been referred to, while extracts from
some of the Provincial Reports will be found below.

ASSAM .

The treatment of the subject by the Census Superintendent


of Assam has already been mentioned as typical of the method
adopted in all provinces at this census. He writes as follows:
'''Hindu Exterior Castes.*' This, as I have confessed, is an
expression to which I plead guilty. I am by no means proud
of it and it is open to many criticisms. I have, however, asked
many Indian gentlemen to give me a better one but they have
not succeeded. The expression, as it stands, connotes castes
which are Hindu castes but which are outside something and
that is really what I mean to imply.
*
What are they outside? The answer is that they are outside
the social pale of Hindu Society; that they are below the
salt that they are on the other side of a barrier which pre-
vents them from moving upwards.

But before going further I must define what I mean by
Hindu Exterior Castes

By this expression I mean castes recognized definitely as
Hindu castes whose water is not acceptable and who, in additioh,
are so deficient as castes in education, wealth and influence or,
for some reason connected with their traditional occupations,
are so looked down upon, that there seems little hope of their
being allowed by Hindu Society to acquire any further social

privileges within at any rate the next decade.

By the use of the word exterior caste " I certainly do not
intend to imply that such a caste can never raise itself to a higher
level. On the contrary, I intend to imply that this can happen,
as it actually has happened in the past, and that an exterior
caste may in the course of time possibly become what I may call
an "interior one.
Itis impossible to lay down any simple test to distinguish
members of the Hindu exterior castes in Assam from others.
The main test to distinguish clean castes from unclean
castes is whether the water of the caste is accepted by members
of the upper castes. A caste whose water is acceptable is known
i.e. 'of the original (aboriginal?) religion
1
SIS APPENDIX A
in the Surma Valley as jalchaV\ and a caste whose water is

not accepted may be conveniently defined as jahachal In


Assam the words "'pani chale and '' pani na chaie'' are in
vogue. But we cannot apply this simple test alone in order to
find out which castes are exterior and which are not. It is true
that all exterior castes are jal-achal, but it is not true that ail
jal-achal castes are exterior. For example, in the Sylhet district
the Shahas are technically a jal-achal caste, but they are a
very wealthy and influential community who are treated with
considerable respect in society, and by a peculiar social
_

convention are permitted to purchase their brides from the


higher castes.
The Shahas are, in fact, a good example of a caste which
though technically unclean have by their own efforts raised them-
selves to a position in which the upper classes simply cannot
afford to ignore them.

I have not therefore classed the Shahas as an Exterior
Caste One of my friendly critics, Babu K. C. Dutta, Extra-
Assistant Commissioner, has taken up this point:
You do not class , he writes, the Shahas as an exterior
caste simply because of their education and the wealth and
influence they command. Yet they are not
jal-chal, they are
not allowed entrance into the Thakurghar of the cleaner castes,
and the disability is not likely to be relaxed in the next decade.
I do not concede for a moment that the disabilities that bar the
Shahas are any more pronounced in the case of the less influen-
tial and uncultured castes. These exterior castes are suffering
from want of education, wealth, culture, in fact, all that con-
tributes to social influence. As soon as they have achieved these,
their position will be akin to that of the Shahas neither clean
nor depressed.

The Shahas are, of course, exterior to the
extent that they
fall on the other
side of the great line which divides Hindu

Society the jahchal line but though this line is still of great
importance, other things must be taken into consideration.
If, for example, the Patnis and Namasudras could shake off
the tradition which associates them with occupations regarded as
low (most of them are now cultivators, but tradition associates
them with fishing and boat-plying) and could acquire, as a caste,
a reputation for wealth and culture, they would, I admit, be in
much the same position as the Shahas are today.

This process will, however, take many generations, and in the
nieantime they are, in my opinion, clearly suffering from greater
disabilitiesthan the Shahas, the disabilities being the very
absence of those factors which have made the Shahas a respected
caste.
Thus, while the jahchal line is a useful line of division
between the upper and the lower castes, it is not of much use
POSITION OF THE EXTERIOR CASTES ni$

as a test for determining the Exterior Castes


Nor does the
test o temple entry afford us much assistance.
'
Generally speaking, in the Surma Valley all castes which are
jahachaV* are not allowed into the actual Thakurghar of
temples in which the higher castes worship and are not allowed
to assist in the ceremonies by bringing tulsi and flowers with
which to decorate the idol. Jahachal castes are, however,
allowed darshan and are permitted to come into the
compound of the temple. In the Assam Valley where the
Namghar'* generally takes the place of the temple the same
principle holds good, but a distinction is made between different
classes of jahachal castes. Nadiyals and Banias,*^ for example,
are not allowed at all into the Namghars of higher castes,
whereas Katonis and Suts are allowed in some districts to enter
the part not regarded as particularly holy, i.e. they are not
allowed to enter the Monikhut
*
The whole matter being, therefore, so indeterminate, how,
it may be asked, can I possibly venture to say, with any degree
of certainty, what castes are exterior? The only possible method
was to find out by. local inquiry in each district the general
social position of all castes which might be thought to come
under the definition of **
exterior castes This is the method
I adopted. ...

Caste in the Assam Valley is not, as elsewhere, chiefly a func-
tional division; it is really a racial division and functional castes
are very few.

Probably for that reason Hinduism in that valley is tolerant
towards the tribal communities which have not yet been
completely absorbed into its organism. It must, in fact, be
extremely difficult for an Assamese Hindu to despise at heart
a man whose Hinduism is open to considerable doubt, but who
considers that he is just as good a man as any Koch or Kalita.
In fact, people like the Deoris consider that they are much
better and dont care who knows it. Nobody can be depressed
who hits you with a big stick if you attempt to show your
contempt for him.
*As a result the only castes in the Assam Valley which can be
called exterior are castes which are either traditionally associated
with some degrading occupation (such as selling fish) or whose
traditional origin is associated with a- bar sinister. About some
castes in this valley there is, however, no possible doubt. Let us
deal with these first.
There is, I consider, no doubt that in the Assam Valley the

caste which at this census has adopted the name of Bania and

^ i.e. Brittial-Bania
*
The term * Brittial-Bania * was itself a recently
adopted term, a substitute probably for Dom or Hari. The use of the
terra Bania (Baniya) is a mere usurpation from the name of the trading

castes of Rajputana and elsewhere tn northern and western India.


APPENDIX A

which at previous censuses was styled Brittial-Bania is an exterior


caste. Some of the leading men of this community have in fact
informed me that their position in society is hopeless and have
asked to be classed as a depressed caste.
That this caste is caste is also the unanimous
an exterior
opinion of responsible officers whom I have consulted.
all
*
Nor is there any doubt about the large class which has now

adopted the caste name of Kaibartta and which was previously
known as Nadiyal; nor about the Charals of Lower Assam who
now call themselves Namasudras with their offshoot the Hiras.
The general opinion about all these castes is unanimous.
'
Thus the Census Officer, Dhubri, reports:
Namasudras or Nadiyals or Jaluas or Charals or Kaibarttas
or Dorns are considered untouchable by caste Hindus who
neither admit them into their places of worship nor take water
touched by them.
and the Census Officer, Jorhat:

Among the indigenous Assamese castes the following are


depressed:
(1) Kaibarttas or. Nadiyals or Dorns.
(2) Brittial-Banias.
(3) Hiras or Charals (found in Lower Assam only).
*
From Darrang comes further evidence:
That Dorns, Nadiyals, Nairiasudras, Charals, Hiras and
Brittial-Banias are depressed is admitted by the members of the
community themselves who were consulted , writes the Census
Officer.


In fact, in every district of the Assam Valley the opinion is
unanimous that the Brittial-Banias and the Kaibarttas (which
name may be taken to include Charals, Nadiyals, Hiras and all
the other names which from time to time have been applied to
various branches of this family) are the most exterior castes in
the whole of the Assam Valley.

These castes are socially outside the pale , and though the
Brittial-Banias have worked hard to improve their position and
have a considerable number of educated men amongst them
they appear to be as far off as ever from any sort of social
recognition.
Ancient custom and practice have ordained that members of
these castes are to be treated as practically untouchable. It is
true that the former necessity of taking a bath if touched by a
member of one of these castes has fallen into disuse, but a
Brahman officer of about 30 years of age has informed me that
when he was a small boy he had to take a bath if, by accident,
he was touched by one of the hated Dorns.

POSITION OF THE EXTERIOR CASTES 515



these castes came others whose position seemed to me
Above
for a long time to be extremely doubtful. They are the Naths
or Jugis (known in Upper Assam as Katonis) and the Suts who
are also commonly called Borias.
A
careful study of the position of these castes has, however,

convinced me that I would be wrong to class them as exterior.


. .The truth about the Naths and the Suts appears to be that
.

they are exterior castes who have made considerable ejflEorts to


raise themselves socially and that their efforts are beginning to
bear fruit. There is a bar against them, but they are beginning
to break through it: they are really superior exterior castes"
who are moving upwards. There is, in fact, hope for them.
As one Assamese officer has briefly expressed it:
All
Assamese low castes have a chance of rising in the social
scale except the Dorns and Haris whose case is hopeless.
The Suts and Naths do not themselves desire to be classed

as depressed or exterior and in fact strongly object to it. . .

Again:
*

The Education Department in 1915 provided three general


dining halls and four separate rooms and the understanding
has always been that the upper caste boys are to dine in the
general dining halls and the depressed or backward class boys
are to use the above four rooms. . For the last two years,
. .

there has been a very strong movement for admitting these


remaining backward caste boys into the general ^dining hall
there is an overwhelming majority in their favour, only a small
minority of about five or six still being in the opposition.

Professor Sen Gupta has subsequently informed me that the


movement referred to in the last paragraph of his note resulted
in August 1931 in the abolition of the remaining restriction and

that as a tentative measure ^permission was given to the stud-
ents who had formerly dined in the four rooms set apart to
dine in the general dining hall. The Professor thinks it prob-
able that the concession will continue.
This last piece of information, I confess, surprised me. If
the concession is allowed to continue, it will reflect great credit
on the liberal spirit of Professor Sen Guptas mess and may, in
time, have a far-reaching effect on social custom in the Assam
Valley. It does not, however, change my opinion that the
Kaibarttas and Banias are definitely exterior castes
On
the whole I feel fairly confident that my classification of
^

exterior castes in the Assam Valley is correct. ...


I only wish I felt as confident about the exterior castes in the

Surma Valley.
Conditions in that Valley are very different from those in
the Assam Valley. Sylhet is linguistically and ethnologically
connected with Bengal, and the inclusion of this large district
oi6 APPENDIX A
in Assam was originally merely a measure of administrative
convenience.

Sylhet and Cachar (which is largely populated by people of
Sylhetti origin) are therefore essentially Bengali in their culture.
Hence there are many more functional castes than in the Assam
Valley where, as I have already explained, caste is largely racial.
Moreover, the presence of a large upper-caste zamindar class in

Sylhet the arbitrators of social usagehas not tended to en-
courage any relaxation in the treatment of the lower orders of
society. Take, for example, the case of a M.A. of the Sylhet
. . .

Mali caste occupying a good Government post. Many respons-


ible Hindu officers have informed me definitely that if such a
person came alone to see them in their paternal homes a chair
would not be offered to the guest. A jol-chauki (small wooden
stool) might be offered. Even Muslims treat these low castes
in the same way. A Muslim Sub-Divisional Officer tells
me that if a Dhubi friend of his occupying a good position in
Government service were to come to his house to see him " I
would not offer him a chair. I would simply say 'sit down'
and the Dhubi would not take a chair.
Surely if the upper-caste Hindus wished to help the lower
castes the least they might do is to treat the educated men
among the lower castes with the same courtesy as they would
extend to an educated Muslim.
The following opinion of a responsible Hindu Government

officer is of interest: Some low-caste men by their submissive


attitude hearts of the upper-caste men and get partial
win the
admission into Others claim as of right and get refused.
society.
A lot depends on the man himself. If he claims too much, he
gets badly snubbed. In places where orthodoxy is strong he
will not, in any case, get fair treatment. In the towns treatment
is more liberal than m
the villages. In the villages orthodoxy
still prevails. If a Patni is the tenant of a Zemindar, he will
never get fair social treatment however highly educated he is.
Orthodoxy is strongest amongst the Zemindars. But if the low-
caste man is an executive officer such as a Sub-Deputy Collector
or an Extra Assistant Commissioner, he would be given better
treatment than a non-executive officer, e.g. than a Deputy
Inspector of Schools. The above remarks refer, of course, only
to private social intercourse. On all social and public occasions
the educated Mali or Patni is simply nowhere. He has (if
he goes to the ceremony) to sit along with his other caste-men
outside the house on the mat provided for their caste, while the
higher castes sit inside the house.
8 Washerman.
8 The Sylhet
Mali *
caste here referred to is the caste more accurately
called I Bhuinmali which is much lower in the social scalethan the
'
Mali ' gardener caste.
FOSITION OF THE EXTERIOR CASTES
Add the following facts:
(1) Members of castes like Mali, Patni, Muchi, etc., are not
allowed to enter into the temples set up by the higher castes;
(2) The upper castes will not take water or food touched by
them; and one begins to realize the dreadful sense of mingled
inferiority and hatred which an educated member of one of
these exterior castes must feel in most cases towards the higher
castes.

The exterior castes themselves are, however, guilty of similar
treatment to each other, and an exterior caste which considers
itself to be on a higher social level than another exterior caste
adopts exactly the same attitude as the higher castes do towards
the exterior castes. A
case which recently happened in Sunam-
ganj illustrates this point. The local ferryman there (a Patni
by caste) was prosecuted for refusing to row a Muchi across the
river. His defence was that, according to social custom, a Patni
could not row for a Muchi and that it has always been the
practice, if a Muchi wanted to cross the river, for the paddle
to be given to him so that he could row himself across.

After careful consideration and analysis of the evidence I
have collected, I have come to the conclusion that the following
are the main castes in the Surma Valley which should be classed
as exterior. They are arranged in alphabetical order:
1. Dhupi or Dhobi. 6. Mahara.
2. Dugla or Dholi. 7. Mali (Bhuinmali).
3. Jhalo and Malo. 8. Namasudra.
4. Yogi (Jugi) (Nath). 9. Patni.
5. Kaibartta (Jalliya). 10. Sutradhar.

There is general consensus of evidence that these castes are


exterior throughout the Surma Valley, though the position of


some of them seems to be much more hopeful in Cachar than
it is in Sylhet. I have no doubt that this is due largely to the
absence of a Zemindari class in Cachar which is a temporarily
settled district. ... I have been told by several officers that the
Maharas are b. jahchal caste, and that people of this caste were

made clean '' by some ancient Raja of Sylhet the idea being
that the Maharas used to carry the Rajas palanquin and, as the
Raja wanted to smoke in it, he had to have as palanquin-bearers
people who could attend to his smoking requirements. On the
other hand, some officers completely deny this story and say that
the Maharas are not a jal-chal caste . the Maharas are princi-
. .

pally to be found in South Sylhet. The Sub-Divisional Officer


(a Muslim) reported as follows:
I have made close and careful enquiries and there is a
general consensus of opinion that the Mahars are not jahchal
and are a depressed class. The story that Raja Subid Narayan
made them jal-chal for smoking requirements only, seems to be
*i8 4PPENDIX A

true. If the Maharas are at all jal-chal, they are jal-chal only in
the sense that a man of higher caste can smoke a huka filled with
water by a Mahara. There is not a single graduate among the
Maharas in this sub-division and not even a single matriculate
can be found. The Deputy Inspector of Schools reports that the
only educated Maharas he has met in the whole sub-division are
three persons working as Vernacular teachers in Primary and
Middle English Schools. So the Maharas are depressed both
socially and educationally. . .


One gentleman from Karimganj ^himself a Nath ^has,
indeed, no hesitation in including his community among the
exterior castes. He writes as follows;
So far as my knowledge goes, amongst the Hindus inhabit-
ing this sub-division the Patnis, Jogis (Naths), Namasudras,
Malis, Dhubis and Duglas are to be properly included in the
list of depressed classes. The reasons of depression regarding
each of these communities are almost the same, namely:
(1) The members of these communities are not allowed by
the so-called high-caste Hindus to enter the temple; even their
shadow defiles the image in the temple.
(2) The high-class Hindus never take any food and water
touched or shadowed by these people.
(3) Brahmins of caste Hindus never agree to officiate as
priests in ceremonies performed by these people, even if they
request them.
"(4) Some of these communities are not allowed to have the
same barber who works amongst the high-class Hindus to work
for them.
In conclusion I beg to say that these are but few amongst
the many disadvantages from which these people suffer.

Previous census reports show that for the last forty years the
Naths have been endeavouring to raise their social position by
giving up widow remarriage and refusing food prepared by
other castes In spite, however, of these efforts the Nath com-
munity of the Surma Valley is still very much looked down upon
and I must, I consider, class them as an exterior caste ".

This classification may seem peculiar inasmuch as I have not
classed the Naths (Katonis) of the Assam Valley as exterior.
How far the Assam Naths are connected with the Sylhet Naths
is a matter intp which I intend to inquire further, but I doubt
whether there is any very close connexion. In any case, a
sufficient explanation of this difference in treatment would
appear to be that Hinduism in Sylhet is not so tolerant as it
is m
the Assam Valley. Even in the Murarichand College caste
restrictions seem to be much more closely observed than in die
Cotton College. I have received a note on the system of messing
i" i.e. Brahmans who officiate as prits for caste Hindus.
POSITION OF THE EXTERIOR CASTES sig

in that college, and it appears that even the Sahas [Shahas] are
not allowed to take their meals in the general dining hall reserved
for the upper-caste students. In fact, the jal-chal line is strictly

observed there at least nominally and the students who do
not belong to the upper castes have their meals served to them
either in their own room or in those set apart in the main
block or in two out-houses provided for the purpose

BALUCHISTAN.
Of the Chuhras the Census Superintendent (1931) of Baluchi-
stan writes:
The Chuhras censused in Baluchistan have returned them-
selves as belonging to the religious groups named below:
Caste or Tribe Keligion
Chuhra Hindu Balmiki
Chuhra Hindu Lai Begi
Chuhra Musalman Lai Begi
Chuhra Musalman Balashai
Chuhra Sikh Mazhabi
Chuhra Chuhra
Although these persons without exception are not allowed to
drink from wells belonging to real Hindus, Muslims or Sikhs,
and are not permitted to enter their places of worship, I include
them in the figures for the various religions to which they claim
to belong, giving separately the numbers (males and females) of
these untouchables in a footnote in each case.'

MADRAS.
The following extracts are taken from the Census Report
(1931) for Madras, where the Census Superintendent gives as
the total number of depressed classes 7,300,000 in round figures,
or 16| per cent of the population of that province. He goes on:

For reasons already given this figure cannot be taken as an
absolute tale of those to whom the peculiar disabilities summed
up in the broad term depressed" attach. There are many
Christian converts on whom disabilities press no whit lighter
than in the untouchable communities they owned before. These
are not included, for personal and local and sectarian variations
enter too largely allocation to be possible. There
are other bodies the difficulties oi whose life are hardly less than
those of any Adi-Dravida, but to whom the technical stigma of
untouchability does not apply. Such do not figure in the census
list. The census total therefore can fairly be termed only an
approximation. It is, however, a good approximation and, as
an indication of the general dimensions of untouchability and
through it of the depressed classes.problem, is absolutely reliable.
stsio APPENDIX A
Whether its approximation is above or below would depend on

the point from which approached. If this is viewed primarily


as the existence of heavy social disabilities, the igure 7,300,000
is a minimum, if it considers strict personal polluting power it
is a maximum. The general dimensions of seven millions are
beyond contest,

7,300,000 figure and the discussions above refer of course
The
only to persons enumerated within the province on census
night. ... A
third of the emigrants belong to the depressed
classes and consequently, were the natural population to be
considered, eight millions would have to be taken as the round
figure for them instead of seven. ...
... It cannot be said that the social disabilities under which
*

these communities labour are in sight of extinction despite the


growth of tolerance and the inevitable effect of the development
of communications and of urban life. Distinguished individual
effort ... is by no .means rare, but it remains individual. Com-
munities cannot yet be said even to have altered appreciably in
outlook, I came across in a Telugu delta district a subordinate
officer of the Labour Department occupying the dak bungalow,
an unusual thing for such officers, who ordinarily put up with
some casteman in the village. His castemen however shied off
him, because of his employment, which brought him into con-
stant association with the depressed classes. This man was of
no notably exalted caste, but a Telaga. It is probable that
resentment at special consideration shown to the depressed
classes in land assignment and other directions is reflected also
in such an attitude, the resentment that the rising of the under-
dog never fails to arouse in those who have kept him down, a
feature not peculiar to India. It remains however an indication
of the true position in the rural areas where the depressed classes
^
are most represented. ...
*
A
peculiar refinement of the untouchability theory was dis-
tance pollution. This set out certain castes as polluting not
merely on contact but by mere approach. The Nayadis were
the backmarkers in this handicapping system and were practi-
cally denied ordinary use of public ways. When the system was
in full force Nayadi process must have borne a strong resem-
blance to that of a malefactor for whom a warrant is out and
whose one object is to avoid close contact with his fellow men.
A lifetime so spent can hardly produce elevation of thought or
desire, and a community whose chief aspiration is to avoid
notice cannot contribute much to national life. This remark-
able development of the superiority theory was practically con-
fined to the west coast and of late years has greatly weakened
even there, probably more because of development of commu-
nications and increase of population than from any conscious
realization that there is in such a system something hardly
POSITIOjY of the exterior castes SSI

compatible with claims to culture and advancement. It is


probably becoming evident that a person of such rare texture
that a presence sixty feet away pollutes him had better park
himself on a desert island or develop a less fragile purity. The
train began the breakdown of this preposterous system; the
bus may complete it.^^
Contact pollution on the other hand existed and exists all
over the presidency.
An instance of the modification and at the same time of the
persistence of discrimination is afforded by the river and canal
ferries of the Telugu delta districts. According to petitions
quoted in a Government Order of 1919 a member of the
depressed classes might have to wait for hours before being taken
over as he and a Brahman would never be taken together and
the Brahman always had priority. In the bigger boats plying
on the two rivers and larger canals there is now no such prefer-
ence; any person waiting is admitted on board. Depressed
classes, however, have to keep to a different end of the boat from
Brahmans. In the cross-river ferries the disappearance is not
so complete and an influential Brahman would be taken over
in preference to a crowd of depressed classes of prior arrival.
In other cases either the Brahman or the depressed class person
would hold back to avoid travelling with the other. The extent
to which prejudice and preference have scope varies from village
to village and with the importance of the Brahman or enlighten-
ment of the depressed. It has been noticed that Christians of
depressed class origin make no bones about getting into the
boat whether a Brahman is waiting or not. The ferryman
occasionally too has prejudices.

On the small canal ferries Brahman precedence is still the
general rule, but where the traffic is considerable and the bala-
kats big, as at Nidadavole before the bridge was built, conditions
resembled those on the river boats.

The same petition complained that in certain municipalities
depressed classes were denied the use of water taps reserved
for higher castes, despite circumstances of proximity and conve-
nience. Such restriction if it ever existed as an official practice
no longer does.
Despite their lowly status, these communities play a large
and important part in the life of the presidency. It is they who
furnish the backbone of agricultural labour in the chief rice-
growing districts. In one form or another they have been the
A recent observer, Mr Eric Miller, reports that distance pollution is
now seen only in ritual situations, and he mentions the case of a particular
Nayar who insisted that a man of an exterior caste should remain at a
distance when bringing him a formal offering on a ritual occasion, but
had no hesitation in approaching a man of the same caste who had cut
his hand and needed assistance in attending to the wound. This was
in 1948-9,
n2 APPENDIX A
victims of an agrestic serfdom wherever they have been. This
generally took (and still takes) the form of compulsory advances
from their employers which could never be repaid in full and
thus tied the borrower to the soil. This was most noticeable
in Tanjore, but a parallel system of advances produced the
same effect in South Kanara. It must be laid to the credit of
Ceylon and other estates that they have done more to raise the
self-respect of the South Indian depressed classes worker than
any other single circumstance. It is possible for the same reason
that emigration is opposed in certain quarters. The Madras
Government appointed an ofi&cer as Commissioner of Labour
and among his particular functions is the attending to the needs
of depressed classes. The decade has seen much expenditure
on provision of wells for them, of schools, and a most important
feature, the buying of house sites for them mainly in the delta
areas. A notable example of a breakaway from caste traditions
is in the Nambudri who was schoolmaster in a depressed classes
school in Malabar. The Nayadi colony of Olavakkot formed
to house members of possibly the most contemned community
in Malabar has been able to develop its activities more than
it anticipated. Recently, however, some difficulties have arisen
through a boycott by other castes of a school which received
some Nayadi pupils.'
APPENDIX B

HINDUISM IN ITS RELATION TO


PRIMITIVE RELIGIONS
IN INDIAN
Though derived no doubt from multiple sources, the Hindu
religion may fairly be said to have taken its final form as the
result of the impact of the social ascendancy of the Indo-
European invaders of the second millennium b.c. on pre-existing
religious institutions .2 The. first occupants of India were
probably Negritos, and elements of their belief, perhaps includ-
ing the reverence for the pipal tree and possibly a primitive
phallic fertility cult, both of which are found in the Andaman
Islands, may have been perpetuated by the proto-Australoids,
who were the next comers and possibly contributed the totemic
theory, or at least the basis thereof. Later elements were prob-
ably of Mediterranean and Iranian origin, and may have con-
tributed a phallic and a megalithic culture and the life-essence
theory, but the relative positions and identities of the Dravidian-
speaking, Mediterranean, Armenoid, Eurasiatic, proto-Austra-
loid, Kolarian and Mon-Khmer or Austro-Asiatic peoples are
difficult to determine and there is little material from which to
draw a conclusion; some would identify the proto-Australoid
and the Kolarian-speaking racial elements. If the latter elements
be distinct from the proto-Australoid, it would be convenient to
suppose that the Kolarians came after them with a life-essence
theory and the Mediterraneans still later to develop it into one
of reincarnation, while bringing in the worship of the Great
Mother, but it is conceivable that the Mediterraneans brought
both the theory and development and the Kolarian came
its
later as a barbarian invader, though no doubt already in posses-
sion of the soul-matter philosophy. At any rate, the hill tribes of
Assam, Burma, and Indo-China appear to contain an element
of Caucasian stock which penetrated to the south-east of Asia
before the southern migration of Mongolians of the Pareoean
branch, and the soul-matter theory must have arisen very early
in the history of the human race. Both Kolarian and Medi-
terranean must have been followed by religious elements from
Asia Minor, brought visi Mesopotamia by traders and settlers
from the west, which no doubt superseded a fertility and soul-
matter cult by one of personified deities, sacrificial propitiation
and a formalized worship, again with phallic elements and such
This appendix is little but a reprint of the greater part of 176 in ch. xi
1

of my
Report on the Census of India, 1931. Such changes as there are are
due to minor excisions, to a few alterations in wording or in punctuation,
and to the addition of references. 2 vide supra, ch. i.
S24 APPENDIX B
institutions as that of the devadasi,^ together with astronomical
lore and cults of the heavenly bodies and priestly institutions
which formed the basis of modern Hinduism; the final form
of which was no doubt determined by the successful coniiict of
this proto-Hinduism on the religious side with the imported
religion of Iranian and Aryan invaders, to whom, however,
it had to concede much socially, resulting in tire socio-religious
position of the priestly order so familiar in India.
The generally accepted view of the Hindu religion, or society,
used to regard it as originating in Aryan invaders of about
1500 B.c. who came in with a higher civilization and a fairer
skin to find the great peninsula inhabited by dark-skinned bar-
barians on whom they imposed the religion of the vedas. This
view can no longer be maintained, and the doubts cast on it
appear to be confirmed by discoveries including that of a figure
of Shiva among the remains at Mohenjodaro, while Sir
John Marshall has clearly shown that the pre-Aryan religion
of the Indus Valley involved a cult of the bull, and of the

snake typical Mediterranean cults, to be found in Crete and
also of phallic symbols, including ring and baetylic stones,
which are probably all part of the soul-fertility cult which is
associated throughout India with menhirs, dolmens, and a
megalithic culture generally; indeed, Heine-Geldern connects
the megalithic Mycenean theatre with India and so with the
Far East and the Pacific Islands. It has been pointed out with
some aptness that in modem Hinduism only those elements of
vedic rites have survived which are essentially social, such as
the marriage ceremonies; the argument being that though
society was or aimed at being Aryan, its religion is older than
that of the so-called Aryan invasion. The god of the Rigvedic
Indo-Europeans is Indra, the thunder god, who fills in later
developments an entirely minor role, apparently bein^ absorbed
into the Hindu pantheon, just as the minor gods of primitive
tribes have been, retaining, however, his personal identity by
virtue of a social prestige or privilege which other tribal gods
have lost in the process of assimilation. The historical Hindu
religion first appears not in the Punjab, which must be regarded
as the area most completely occupied by the Indo-European
invaders, but to the east of it in the Brahmarshidesha,
where stable fusion between these Indo-European invaders
and the previous inhabitants probably took place. When
alien cultures and religions fuse to form a new culture or
reli^on, it will not be found that this fusion takes place where
the intrusive culture is strong enough to predominate. It will
rather appear, away from the centre where the intrusion is
strongest, in some area where the previous culture was strong
enough to resist complete suppression and make its influence
3 supruy Tpy. sqq.
HINDUISM AND PRIMITIVE RELIGIONS
felt on the new one. Thus it is that the efflorescence of Greek
culture took place not in Sparta, where was the purest blood of
the northern invaders, but in Athens, where the grasshopper-
wearing inhabitants regarded themselves as autochthones and
where there was probably effective fusion between the fair-
haired northerner and the dark-haired Pelasgian. '
Similarly,
there is some reason to believe that Rome grew from a fusion
between the ancient Etruscans and later invaders, whether the
latter came from the east or from the north. In the same way
it is suggested that Hindu religion and society finally took form
and flourished as a result of the impact of the invading Indo-
European on the indigenous religion that he found in India. It
is quite clear that the previous inhabitants of India lived in
cities and had a high civilization, probably of western Asiatic
origin, and it is significant that Hinduism is remarkable for the
similarity of many of its tenets and practices to those of Asia
Minor and Mesopotamia. The indigenous religion of any
country inevitably starts with an advantage over that of an
invading people, since it is the priest of the country who knows
how to approaclx the gods of the soil and propitiate them, and
for that reason there is always a tendency for a local religion to
establish its ascendancy over an intrusive one. This appears to
have been the case in India, where the important position of
Shiva, Vishnu, and Kali, as compared to the unimportant one
which Indra now holds, signalizes the triumph of the older gods.
The religious history of pre-vedic India was probably similar
and parallel to that of the eastern Mediterranean and of Asia
Minor. Professor Tucci points out^ that though the moon does
not appear to have been an independent divinity, ancient lunar
cults have been assimilated by Devi in the forms of Durga, Kali
and Tripurasundari. The cult of snakes, and the worship of a
mother goddess, were probably brought in by earlier invaders
of Mediterranean or of Armenoid race, speaking no doubt a
Dravidian language, whose religion must also be associated with
fertility cults, phallic symbolism, the devadasi cult, and probably
human sacrifice. Recent discoveries in Crete have revealed a
remarkable snake cult associated with the symbol of the double
axe. With Mesopotamia, too, we must perhaps associate a moon
god and sun goddess, whose sex was changed with a change from
matrilineal to patrilineal descent perhaps under the influence
of the Rigvedic invaders. It is worth pointing out that the
deification and worship of kings, very typical of the Hindu atti-
tude to kingship, is stated by Langdon to be characteristic
of Sumerian religion in contrast to Semitic. It would also
appear not characteristic of the religion of the Rigveda, but on
^ In a note on *
Traces of Lunar Cult in India *
in Rivista degii Studi
Orientali, vol. xn (1930), fasc. iv, quoted in The Indian Antiquary iot
January 1931}. p. 17. 1931, p. 367.
15 ,.
n6 APPENDIX B
the contrary to be connected with the beliefs in the external
soul and in life-essence discussed below, inasmuch as the king
contains or represents the life principle of the community he
rules. Like the cult of the snake, the transmigration of souls
too appears to be a doctrine in no way typical of northern reli-
gions, in which the dead live on underground, and Fustel de
Coulanges has pointed out ^ that it is not a feature of any nor-
thern religion though it has survived and been incorporated in
them from the more ancient religions of Greece and Italy..
Ancestor worship again is very strong in India, and this, too,
would appear foreign to northern European reiigicm, and indeed
it is almost impossible that nomads should be ancestor wor-
shippers, and the Aryan invasion, so-called, was probably an
invasion of steppe-dwelling tribes, pastoral in habit and still
nomadic. Cremation they may have brought in, and if so, they
gave it a social cachet which is still leading to its gradual
adoption by tribes which have previously practised burial or
exposure, but it seems much more likely that the Rigvedic
Aryans buried their dead and adopted cremation from the
inhabitants whom they conquered. The eighth book of the
Rigveda contains the following words addressed to the dead:
*
I place this barrier (of stones) for the living that no other . . .

may go beyond it. May they live a hundred numerous autumns,


keeping death at a distance by this hill. Enter the mother
. . .

earth. Earth, let his breath rise upward (easily); oppress


. . .

him not. Even as a mother covers her son with the end
. . .

of her cloth, so do ye, earth, cover him may these homes


. . , . . .

for all time be his asylum. I heap up earth above thee,' etc.t
This passage seems very clearly to indicate burial in a tumulus,
and the word translated barrier is stated in a note to be

'

paridhi, which may mean part of a circle of stones. It is true


that the tenth hymn of the Rigveda clearly refers to cremation,
but the author above quoted reprds it, rightly no doubt, as
the later passage, and suggests that ritual exigencies involved
the dislocation of the verses and their fusion for ceremonial
purposes in the Yajurveda and the Sutras, the reference to
inhumation being then interpreted as indicating the burial of
the ashes. It appears, however, quite clear that the hymn quoted
above can only refer to the inhumation of the Vody, and that
this practice as well as that of cremation was in vtse at the time
of the Rigveda, while cremation is not mentioned until the tenth
hymn, admittedly a much later composition than the earlier
ones; cremation also seems definitely to have been the practice
in the Indus valley of the Mohenjodaro period, and therefore
LaCite Antique, pp. 7 sq.
7 Rajendi^alala Mitra, Indo-Aryans, xii p, 133.
Mr H.
J. E. Peake thought that it' refers originally to a wooden fence,
which may have been translated into stone Uttr (personal communication).
HINDUISM AND PRIMITIVE RELIGIONS
the more likely one of the two to have been adopted as an alter-
native by the Rigvedic Aryans at their period of fusion with the
pre-existing population. The Aryan sanctity of fire seems likely
to have been incompatible with cremation, and it will be re-
membered that Herodotus^ taxes Cambyses with impiety for
having had the body of Amasis burned, for the Persians regard
fire as a god and therefore to burn the dead is on no account
allowed . for they say that it is not right to offer to a god the
. .

corpse of a man \ It may be noticed that Wilson remarks of


the Rajputs of Kachh and Kathiawar that they encouraged their
concubines to commit s&tJ in preference to their wives, and he
actually gives as the reason that sati was a custom of low castes
and therefore derogatory to Rajputs; and it may be noted also
that Rajputs in Gujarat who forbid widow remarriage are called
*
Vanka, crooked', and those who allow it Padhra, straight

With Asia Minor or Mesopotamia again we must associate


astronomy and the worship of the heavenly bodies, which form
an important part of Hindu culture, and in particular the cult
of the moon god. Sun worship appears to be less important
in the Rigveda than at a later date, when the Bhavishya Purana
is largely devoted to a cult of the sun. It is, however, possible
that it was the influence of Rigvedic invaders which changed
the sex of the sun in India from female to male and gave rise to
the sun-descended nobility as distinct from the moon-descended.
In Rigveda X
Soma the moon is represented as male and as
marrying Surya the daughter of the Sun. The name of the
latter suggests that the Sun himself, Surya, was originally the
female that married the Moon, and that there has been a change
of sex associated in so many parts of the world with variations
of the Phaethon legend. Similarly, again, the existing holy
places of the Hindus are outside Brahmavartta where one might
have expected to find them, if it were really the fact that the
religion actually arose in that area, while to find them elsewhere
is consistent with a view that they are places regarded with
devotion by the religions which preceded the invasion. This
view is sometimes emphasized by the existence of Hindu shrines
where priests and custodians are not Brahmans but some pseudo-
Brahman or Sudra caste, e.g. the Malis who arc the officiating
priests of some Orissa temples and probably the Panda Brahmans
of the same region.ts It is doubtless significant that sacrifice of
cattle was detested by the public' though enjoined by the
earlier vedas, the inference from which is that the reverence paid
to cattle predates the Rigvedic invasion, and Buddhism and
9 m, i6.
^^Infanticide in Western India, p, 74.
Campbell, JBofn&fly Gazetteer, IX, i, p. itj H.
CL also Wise, op. dt., p. a7o.
5|8 appendix B
Jainism, the latter o which contains extremely ancient ceremo-
nial survivals, may represent a reaction towards the pre-vedic
religion to which the majority of the inhabitants of northern
India were attached and which was modified but not destroyed
by contact with the invaders. The first prohibition of cow-
killing seems to be found in the comparatively late Atharvaveda
and to be applied specially, if not exclusively, to Brahmans, while
elsewhere we learn that the cow, although a fit offering for Mitra
and Varuna, should not be sacrificed because such sacrifice is
opposed to public feeling, a clear indication of the contrast
between the religions of the socially superior Aryan invaders
and the cattle-cherishing inhabitants who formed the bulk of
the population. Let not the Rajanya desire to eat the inedible
*

cow of a Brahman (Atharvaveda, v.18.1). In southern India

the cow is as much revered in those areas of the presidency


with the lightest tincture of Brahmanism as in those more
affected which may be taken to indicate that reverence for the
*

cow in India is older than the vedic religion In any case the
sanctity of the cow is foreign to the Rigveda and appears far
more suggestive of the religions of Asia Minor, Egypt, and Crete
than of the Indo-European invaders who came from the steppes
of the north-west to conquer northern India in the strength
of their horses and of their iron. Indra, moreover, appears as
the author of sacrifice, and in the Yajurveda it seems still to be
Indra and Varuna who are the principal recipients of sacrificed
cattle. Nor is it possible to accept Sir John Marshalls anti-
thesis between the worship of the bull and the worship of the
cow. Both are surely different aspects of the same reverence for
cattle which characterizes the pre-equine civilizations of the
Mediterranean basin, and in India are pre-' Aryan in origin.
The vedas after all enjoined gaumedha, and the Black Yajur-
veda lays down an elaborate list of deities to whom bulls, oxen
and cows can be appropriately sacrificed, Vishnu, Shiva, and
Kali, the great gods of Hinduism, are not Rigvedic deities at
all. Sakti is probably a cult derived from the Great Mother
goddess of Asia Minor, and the cult of Shiva is inevitably asso-
ciated with it, the two being bound up with the phallic religion
of southern Asia and of the eastern Mediterranean. It is prob-
ably significant that the word Ungam is definitely of non-Aryan
origin, as Przyluski has demonstrated, while the word puja is
also believed to be a non-Sanskritic loan word. With the
worship of Shiva, too, is to be associated the snake cult of which
there are so many survivals in southern India and which appears
to have been at an early date in definite opposition to Brah-
manistic Hinduism, the conflict between the two being indi-
cated, for instance, by Krishnas exploits against serpents, by
the destruction of serpents at the burning of the forests of
13 Yeatts, Report on the Census of Madras, 1931, p. 3ifo.
HINDUISM AND PRIMITIVE RELIGIONS stsig

Khandava and the slaughter of serpents in the Mahabharata.


Vishnu, apparently a post-Rigvedic god, is perhaps the fruit of
the reaction of what we may call proto-Hinduism to the Rig-
vedic invaders, as also the present ascendancy of male over
female conceptions of the deity, and Przyluski ascribes a
Dravidian origin to the name Vishnu *
At the same time
Vishnu would seem to have some associations with religious
beliefs which must be regarded as represented chiefly in beliefs
yet surviving among primitive tribes. Indra apparently is him-
self declared in the Mahabharata to be guilty of brahmanicide
in killing Vrtra and Namuci who were Danavas, though the
Rigveda praises him for the same deed. Pargiter adduces con-
siderable evidence to show that the true Brahman families were
of pre-Rigvedic origin, and that the Aryan kings of Madhya-
desha were their own priests and in the earliest times had no
Brahmans. Strabo remarks that it is recorded that The

Indians worshipped Zeus Ombrios, the river Ganges, and the


indigenous gods,^^ and as Zeus Ombrios is clearly Indra, the
thunder god, the suggestion that the other gods worshipped
are of indigenous origin is probably very near to the truth, and
the traditional view that the Hindu religion is a growth entirely
subsecment to the Rigveda, or rather to the Rigvedic invasions,
is no longer tenable. R.ai Bahadur Ramaprasad Chanda, in a
paper on the Non-vedic Elements in Brahmanism

has made
a number of points which indicate the continued existence of
the pre-Rigyeaic religions alongside of or in opposition to the
orthodox Hinduism of the Brahmarshidesha. Quoting Kuma-
rila and Medhatithi he points out that the Smartas include non-
vedic elements; that thus of the four orders named in the
Dharmasutra of Gautama (Student, Householder, Ascetic and
Hermit) only that of householder (grihastha) is prescribed in
the veaas; that the Upanishads were not originally recognized
as part of the vedic canon at all and had their origin outside
vedic Hinduism; that the Yatis destroyed by Indra are probably
the forerunners of the Yatis of the Upanishads and the Smritis
and that the latter order were organized on a pre-vedic model;
that the Pancharatra and Pasupata systems were condemned by
Kumarila as non-vedic and that the Vaishnavite and Saivite sects
are derived respectively from .those two systems; that contact
with Pasupatas, Saivas, Jainas, etc., involved purification; that
by the time of the Mahabharata, however, Pancharatra and
Pasupata are placed on a footing of recognized and orthodox
religious authority. His general conclusion is that the cults of
Vishnu, Shiva, and Sakti originated among a people of different

ethnic origins from the midlandic Aryans. The point to be


Archiv Orientdlni, tv, p.
i? (August
15 '
tou5 egchonous daimonas% bk. xv, p. 718.
1* MS. (unpublished?) in the authors possession.
31 JO APPE^jyiX B
emphasized here is not so much Chanda's precise conclusions as
the evidence he adduces of the survival of pre-vedic religion
alongside and inside the later forms of Hinduism, and of their
gradual absorption and acceptance as a recognized part of it,
which has perhaps since developed into the position of their
forming the most important part of it.
If the view be accepted that the Hindu religion has its origin
in pre-vedic times and that in its later form it is the result of
the reaction by the religion of the country to the intrusive
beliefs of the northern invaders, many features of Hinduism
will become at once more comprehensible, while the very strik-
ing difference between the religion of the Rigveda and that of
the Dharmashastras will seem natural. It will, however, be still
necessary to look westward for the source of Hindu religion;
though its spread in India was possibly in the nature of a peace-
ful infiltration, along the trade routes from Asia Minor, of
beliefs and practices which associated themselves with those
already followed by the indigenous inhabitants. This would
explain Hinduism's amalgamation with and absorption of local
cults and its excessive multiformity, and is, moreover, in entire
accordance with the manner in which it still spreads at the
present day, absorbing tribal religions by virtue of its social
prestige, by identification of local gods with its own, by the
experimental resort to Hindu priests, and by the social promo-
tion of pagan chiefs who are provided with suitable mythological
pedigrees. Into the early Hindu beliefs spread in this manner
the religion of the Rigveda has been imposed and absorbed.
Features survive curiously in out-of-the-way parts beyond the
pale of Hinduism itself. Thus the horse sacrifice has become a
tertility rite among the non-Hindu Garos of the Assam hills and
appears likewise in the wilds of Sumatra, though it has failed
to establish itself in orthodox Hinduism. Similarly, though the
sacrifice of cattle is anathema to the true Hindu, the Taittiriya
Brahmana recommended a whole series of animal slaughtering,
including both bulls and cows, to be performed at the Pancha-
saradiya Sava and at the Asvamedha, a series extremely sugges-
tive of the scales laid down for successive feasts for the acquire-
ment of social merit at such ceremonies as the Terhengi of the
Angami Nagas, while the Grihya Sutra enjoins the Sulagava
spitted cow'ceremony (corresponding roughly to the Angami
Sekrengi, as the Panchasaradiya docs to the Terhengi) at which
^e beast was killed, as today by Serna Nagas, with a pointed
stake, and its death was accompanied by the erection of a
wooden post with a round top mortised on to it and by the
distribution of the animal's flesh.^^
Viewed in this light it is not difiicirit to titidtettand the claim
15 Vide ^
Mim, op. ck., x, pp. 169 sqq. Wamm, The Serna
p.ssg-
HINDUISM AND PRIMITIVE RELIGIONS Si^i

of certain politicians that the term Hinduism should cover all


religions having their origin in India, even though we hold that
the original impulse came from the Mediterranean or Asia
Minor, since Jainism, Buddhism, and Sikhism are all offshoots
of the same root. The claim is less logical when applied to
tribal religions which have not yet reached the stage of accepting
Brahmans as priests or of attaching any sanctity to the cow or of
worshipping in Hindu temples in their own villages. An occa-
sional visit to Hindu temples away from home is not quite a
safe test, since many such shrines undoubtedly occupy sites dedi-
cated originally to more ancient indigenous deities and subse-
quently Hinduized, and in any case it is typical of primitive
religions, to propitiate the gods of any locality they may visit.
Admittedly, however, the line is hard to draw between Hinduism
and tribal religions. The inclusion of the latter within the
Hindu fold is easy, and wherever hill or forest tribes live in
permanent daily contact with Hindus their religion rapidly
assimilates itself to that of their neighbours though the old
method of thinking is unchanged. Thus it is that religious
or quasi-religious beliefs and practices among Hindus appear
very frequently to be based on the principles of magic, mana
or other ideas common in primitive religion. The very word
brahma itself seems to have possibly connoted originally super-
natural /power or influence of the nature of mana a view

appareiitly supported by the Atharv^aveda and these beliefs and
practi^res survive and operate with all their primitive qualities
alongside the loftiest heights -of asceticism and philosophy. If
it appears that the latter aspect of the Hindu religion is lost
sight of in these pages, it is because their purpose is rather to
articulate the fragments of the more primitive and material
philosophy that preceded it, than to emphasize what is already
of long-standing and undisputed recognition.
We may therefore expect to find very ancient and primitive
beliefs continuing under the guise of Hinduism, or even consti-
tuting its major content. The sanctity of the fig-tree, for in-
stance, is possibly to be associated with the beliefs of the Negrito
inhabitants who appear to have formed the earliest population
of India. It is probably on account of its milk-like sap that the
associated with fertility cults in Africa, Italy, and New
fi^cus is
Guinea as well as in Assam and in southern India, and it is
generally also connected with the spirits of the dead. This cult
appears to be shared by the Andamanese who are an approxi-
mately pure Negrito race and perhaps the only race still surviv-
ing in the world comparatively unmixed in blood. At any rate
they and their beliefs have probably been isolated for some five
thousand years at least. Similarly, though the probability is that
this element of Hinduism is due to some pre-Aryan immigrant

cult from the direction of Asia Minor, the possibility that some
SJS APPENDIX B
tribal and totemic taboo has acted as a contributory factor in the
religious sanctity attaching to cattle cannot be entirely over-
looked. Thus the flesh of cattle is tabooed by certain clans
among primitive tribes of Assam and Indonesia who do not
appear to have come even remotely under the influence of
JElmduism, while on the other hand the cow is regarded as com-
pletely tabooed by the Shins of Chilas, who are described by
Leitner as a Hindu tribe with nowadays a veneer of Islam, the
highest caste in Dardistan, and really Brahmans themselves,
though expelled from India or from Kashmir by Brahmans.
Not only do they taboo the flesh of the cow but also its milk, and
they only touch a calf at the end of a prong.^
Pargiters view of the original conception of brahma as akin
to that of mana has already been mentioned, and the view
seems naturally to associate itself with the views on soul-matter,
or life itself, as a transferable and material substance, which are
so familiar in Indonesia and further India, but which are
actually common enough in India itself. It is on this theory
of the indestructibility and transferability of life-matter that
the underlying principle of head-hunting is based in Assam;
in other parts of the same cultural area it has been manifest in
human sacrifice or in cannibalism, the latter perhaps being its
most primitive manifestation and the former its most developed.
That the principle is still strong in India may be inferred from
a number of recent instances, several of which are given later.
In the form of head-hunting this theory involves that which
regards the head as the particular seat of the soul, and this
belief is apparent in India proper in the sanctity which attaches
to the headf or to the hair, as also in many cases where the (soul-
impregnated?) hair does duty for the individual. Thus, in the
case of the Naga who dies far from home, a portion of his hair
is brought back by his companions to be attached to the head
of the wooden effigy, which is then the subject of the usual
funeral ceremonies, and one may compare the way in which
the head-hunter so often substitutes the nair of his dead or even
of his living, and unwitting, victim for the head he cannot carry
off. The Ujli Minas when unsuccessful in dacoity will only
shave at home and after propitiating their goddess. Probably
they fear that they may be suffering from a loss of life-essence
as Samson did when his hair was cut. Conversely, a Korku
woman of the Central Provinces tries to obtain as a cure for
barrenness a hair from the mother of a large family which she
buries under her bathing stone. The same theory may per-
haps be the origin of the familiar caste mark placed in front

of the forehead just between the eyebrows. The Angami Naga


Leitner, Dardistan in 1866. 1886 and 1893, Appendices vi, p. 1 and vm,
p. 15 ; cf. Biddulph, Tribes of the Hindoo Koosh, p, 113.
also supra, pp. 163-4. Russell, op. cit., iii, p. 563.
BIMDUISM AND PRIMITIVE RELIGIONS 233

tribe regard particular place as the special seat of the soul,


tliis

conceived of as a diminutive human shape, which it is necessary


to guard from the infectious influence of strangers by means
of disinfectants. This is done by attaching to that particular
spot on the forehead a small' fragment of the leaf of the worm-

wood an effective disinfectant of spiritual influences, like other
aromatic plants. It seems likely that it is in a practice of this
sort and as a protection against danger to the soul that the use
of the so-called caste mark may have first originated.21
Fertility cults have already been mentioned. These, like
head-hunting and human are intimately associated
sacrifice,
with, agriculture, and the line is hard to draw at the point at
which a purely magical fertility rite begins to develop into
ceremonial of a genuinely religious nature. At any rate a point
is easily reached by the former which is correlated to the latter,
and the tribal cult ceases to be purely tribal and is identified
with some definitely religious festival so that the magical, cere-
monial and devotional aspects become merged. Magical ferti-
lity rites, originally regarded as necessary to ensure the processes
of nature, are thus conserved and crystallized and continue to
be accepted as a natural feature in the ceremony, when the
reason of their being there is forgotten. So, too, features of
such rites which in the beginning are natural and inevitable,
since they are regarded as essentially necessary to make the rite
effective, and for this reason are performed without any sense
of impropriety or obscenity, become, when they cease to be
essential to the ceremony, effectively indecent, but are not recog-
nized as such as long as the traditional form of the ceremony
continues to be unquestioned. In this form ceremonies and
practices survive long after the conditions of society in which
they originated have changed. Thus rites essentially priapic
survived at Isemia near Naples at any rate into the nineteenth
century actually under aegis of the Church,^^ and it is only on
I
contact with and under criticism from some external source that
familiar and therefore unquestioned practices are seen in a new
and critical light. That this process is now taking place in
Hindu society is sufficiently obvious. The Cochin Government
prohibited the singing of obscene songs, etc., at the Holi festival,
and advanced Hindu opinion would probably welcome a similar
prohibition in many other parts of India, where already the
festival seems to be generally celebrated with less excess than
used to be the case. The Government of Mysore abolished the
institution of devadasis, and here again an influential element
^
in Hindu society was loud in its approval.
21 See Hutton, The An^ami Nagas, p.
179, and cf. the Mexican custom
of sticking a small black disc of cloth called chiqueadores on to the forehead
as a cure for headache (Antiquity, March 1936, p. 97).
DnlsLUte, Des Divinitds Gdneratrices, -pi^. sqq.
m 'APPENDIX B
It is perhaps symptomatic of the tendency to reform, and to-
wards even more drastic change in Hinduism, that it should
prove possible, as apparently it has in Bombay, to constitute
an Anti-Priestcraft Association the professed purpose of which
*

is to combat all religious and social beliefs and customs and


*

institutions which cannot stand the test of reason and members


of which are reported to entertain a frankly boishevist attitude
towards ail religions and to advocate the destruction of all tem-
ples, churches and mosques. On the other hand, little effective
has been accomplished in the way of removing untouchability
in its real sense. It is often said that the concisions of modern
life have broken down the idea that contact with certain castes
involves pollution and this is true just to that extent to which
the use of conveniences such as trams, buses and trains necessi-
tates a relaxation of the rule that certain castes pollute by touch
and still more that they can pollute by mere proximity. Fur-
ther, there is a tendency, obviously consequent on the necessity
of the relaxation referred to above, to relax the rule of pollution
by touch in the case of members of untouchable castes who do
not pursue untouchable avocations. This does not necessarily
involve any real abandonment of the attitude of caste Hindus
to what the Census Superintendent of Assam [1931] conveniently
described as the Exterior Castes
'
The water they touch, is
still undrinkable, food they touch becomes impure, and they
are not admitted to places of worship or to restaurants, nor
will the ordinary barbers serve them. Indeed, the most that
seems to have been yet accomplished is the occasional staging of
inter-caste meals, gestures which appear so far to have had
little practical effect on the general attitude of the caste to the
outcaste Hindu, though it is only fair to add that in Assam and
Bengal, at any rate, students and schoolboys have shown a
manifestly more tolerant attitude than their elders. Moreover, it
must not be forgotten that the attempts of the depressed classes
to obtain the right of entry to temples are perhaps sometimes
as much inspired by social motives as by religious ones, and
produce an antagonistic reaction which might be absent if
religion alone were involved.
The tribal religions, as has been indicated already, represent,
as it were, surplus material not yet built into the temple of
Hinduism. How similar this surplus is to the material already
used will appear in many ways and may be noticed, to start
with, in the cults of the dead. The Hindu rites of the shradh
provide for the creation of a new body to house the soul of the
deceased and, though theoretically renewed every year to main-
tain it, they are usually as a matter of fact gradually abandoned
with the lapse of time. In the tribal religions this cult of the
dead is seen in a precisely parallel form, but at a very much
more matter-of-fact and materialistic stage of the development
HINDUISM AND PRIMITIVE RELIGIONS 535

of the idea. Thus in Mysore the Hasala caste redeem the soul
with a pig from the magician who has caused the death and
domicile it in a pot where it is supplied with food and water.^s
The Nicobarese and some Naga tribes fashion wooden figures
on which the skull of the deceased is placed in order that the
soul may leave it and enter the wooden figure.^^ It is for a

time kept supplied with all worldly necessities. A similar prac-


tice must formerly have obtained among the Garo of Assam,
but it has disappeared, and in the wooden figures now used the
pegs that held the skull in place have become unrecognizable,
surviving apparently as a sort of a pair of ornamental horns,^
though an obsolete grave figure in the Indian Museum in
Calcutta has a pair of horns much more nearly approximating
to the trans-frontier Konyak Naga type. Farther west and south
the Sawara of the Ganjam Agency tracts use a similar but more
conventionalized wooden figure to accommodate the soul of their
cremated dead during the interval between death and cremation
and the time for the erection of a stone or stones for the souls
of the dead during the year past, which is done annually about
the time of the transplantation of the crop.^ Still farther west
the Kunbi of the Central Provinces 27 make an image of their
dead in brass, which is kept until the superfluity of such images
necessitates their deposit in the water of some sacred river. In
the north the statues of the dead made by the head-hunting
Kafirs of the western Himalayas had probably a similar purpose,
while in western India may be seen the chattries of deceased
Hindu Maratha rulers in which the recumbent bull and a lingam
face the waxen images of the dead prince and his wife, which
latter are piously supplied with food and other requisites and
are entertained with music and have their clothes changed regu-
larly once a week. In the very south of India the Mala-Arayan
of Travancore make a metal effigy of their dead, which is kept
in a miniature stone cist covered with a capstone (like the
tattooed skull of a Konyak Naga in the north-east) and erected
on high ground. The image is brought out annually and feasted
and worshipped with tulsi leaves on its head.^s
23 Iyer, Mysore Tribes mid Cestes, in, p. 306.
2-* Bonington, Ossuary Practices in the Nicobars % in Man, xxxii, 133
'

(May 1935) ; Hutton, Two Tours East of the Naga Hills % in M.A S,B,, xi,

no. 1 (1929).
Playfair, The Garos, p. 113.
26 Hutton, Census of India, 193I, i, pt. iii b, p. 4. Dr Elwin writes in
1950 that he can find no trace among the Sawara of the use of such a conven-
tional wooden figure, but I saw one myself in March 1931, arid it was
explained to me as set down here.
27 Russell, op. cit., IV,
p. 39.
28 Mateer, Native Life in Tramneore, L. A. K. Iyer, Travancore
p. 75 ;
Castes and Tribes, i, pp. 179-90.
jje appendix s
Allusion 29has already been made to the theory of soul-matter
as a fertilizer of the aops and a producer of life generally, a
theory which appears to pervade magico-religious thought and
practice throughout Indonesia and south-west Asia and survives
in strength in farther India. The collective disposal of the
villages dead at the time of sowing is clearly associated in some
Naga tribes with their aspect as crop fertilizers, while the Oraons
of Chota Nagpur again, if the paddy has sprouted, inter their
dead temporarily to cremate them the following year before it
sprouts. The connexion between the souls of the dead and the
fertilization of the ground is reflected again in their very frequent
association with water. It is hardly necessary to call to mind
the value set by Hindus upon the immersion of their dead in
the Ganges, but there are a number of parallel beliefs in more
or less primitive tribes which do not seem to owe their existence
to Hindu influence but rather to share their origin with the
ingredients of that religious system. Thus the Meithei practice
of disposing of the frontal bone of the deceased in the Ganges
appears, at first sight, to be the result of their Hinduization, and
no doubt their choice of the Ganges is such a result, but their
neighbours the Kacharis, when yet un-Hinduized, used to con-
sign their frontal bones to the Kopili river after the harvest,
while the Rengma Naga make a pool for water at the grave
of any notable man that the rain, and rice, may be plentiful;
and at least one other Naga tribe pours water on a grave to
cause rain, while the Palaung of eastern Burma fetch a bier
pole from a grave and put it m
a stream for the same purpose.
The Santal again have the practice, at any rate under certain
circumstances, of consigning a piece of bone from the head and
another from the breast of the dead to the waters of the Damodar
river.99 The Panwar mourner, *1 besides throwing into the
Narbada the bones of the dead, throws in with them some of
his own hair also, thus perhaps vicariously accompanying the
soul. The Kunbi practice of consigning to the Ganges the
brass images of their ancestors has already been mentioned, and
the Bishnoi Brahmans of Sind were described by Tod as burying
their dead at their thresholds and raising over them small altars
on which they place an image of Shiva (sc., a lingam?) and a
jar of water.82 Though nothing is said of any fertilizing effect,
these various practices would seem not unconnected. It is even
possible that there may be some similar association with the
dead or their avenging spirits in the ordeal by water described
in a communication from Warren Hastings's as practised by
Hindus in tlie Ganges, the accused man submerging himself and
29 See above on p. S5* and, for an instance
of such a specific theory,
p. S50, below. so Dalton, Ethnology
of Bengal, p. Ji8.
SI Russell, op. cit., rv, Rajasthan, vni, ii (vol. in, p. 1*97).
p. 348.
ss On the Trial by Ordeal among the Hindus ,
by Ali Ibrahim Khan
in Asiatick Researches, 1, p. 390.
A. Obsolete type of Garo Grave Figure (Kima) from specimens now in the
Indian Museum, Calcutta.
B. Statue of dead man with skull receptacle on head. Ukha Village,
Konyak Naga.
C. Soul-figure of a woman at Chongvi, with horns for keeping her skull
in place (Konyak Naga tribe).
D. So-called
Devil-Scarer * in the Nicobars, with horns projecting from
his top hat.
E. Statue of Nicobarese with skull of deceased as head. The wooden body
serves as a cupboard for his bones (Teressa Is., 1930).
v

yem ,

me
JC^fjrjsfAo

ap

Sef >f.

^ASbkoJ

fem&ie,

G
,

sfApS^*
msie

B^
A tOocmorpe

effigies,

0eoerel%$^

jnemofjai

deM<f

Ih^
Kafir
HINDUISM AND PRIMITIVE RELIGIONS 3539

holding to the foot of a Brahman, a form of ritual parallels to


which are practised by a number of primitive tribes from western
India to Vizagapatam and from the Central Provinces via Assam
and Burma to Indo-China. Even in the Rigveda itself there
seems to be some notion of the departure of the souls of the
dead to the waters or into vegetation, in spite of the more defi-
nite and prevailing idea of a house of the dead, and in the later
vedic rite of garbhadhana to promote conception the husband
infused grass in water and then poured the water down his wife's
nostril, which looks as though it were intended that soul-matter
from the grass should enter the woman to cause her to conceive.
The idea of soul-matter as a fertilizing agent is probably also
responsible for distinctive treatment of the bodies of those who
die by bad deaths and are therefore probably either unfertile
*
'

or unsuitable or likely to lead to the reproduction of bad results.


It is thus that we find everywhere special treatment accorded to
the bodies of women who die in child-birth, while other forms
of death are treated differently, i.e. as bad or otherwise, by differ-
ent tribes. In the case of persons killed by wild beasts the
idea is perhaps that the soul-stuff of the dead is absorbed by
the wild animal, and this is illustrated by the widespread belief
that the soul of the dead rides on the tiger, as told, for instance,
by the Annamites of Cambodia and the Baigas of Central India.
The idea that the soul of a person killed enters the killer is
found elsewhere, e.g. in Australia.
The same doctrine of soul-matter is probably the principle
which underlies head-hunting, human sacrifice, and cannibalism.
The first of these has been recognized in India only in Kafiristan
and in Assam, though it is reported by Colonel Cole of the
Bhils^ that they were at one time accustomed to bring back
the heads of their enemies and hang them up in trees; and the
Kondhs circulated a head, hair and fingers as a signal for a
rising in 1882, and of four heads taken in that rising one at any
rate was affixed as a trophy to a tamarind tree near Billat
'
'

village; again, a Brahui clan explain their name of Sarparra


as meaning decapitator

and have been identified with
Strabos Saraparae, that is Decapitators

who lived near the
Guranii and the Medes, and were savage intractable mountainy
'

men' who slash round the legs and cut off the heads of
strangers The theory, however, on which it is based, that
Report on the Census of the Rajputana Agency, 1931, p. i5.
35 Thurston, Castes and Tribes, m,
pp. 411 sq.
36 Bray, Report on the Census of Baluchistan, igii, R. Hughes-
p. 186 ;

BuHer in Risley, Ethnographic Appendices, p. 68.


3 y Strabo, xi,
531. The generally preferred rediding, periskuthistas for
periskelistas, is translated * scalp in die Scythian manner , but this is hardly
compatible with decapitation ; either would seem to be ihapax legomenon,
and the former reading is much more likely the emendation of a scholiast
who knew his Herodotus, but was as unacquainted with what is a common
APPENDIX M
the soul-matter is specially located iu the head, may be detected
elsewhere. Thus the Andamanese attach special importance 'to
the jaw and to the skull in mourning, and at any rate one
case is recorded of their carrying off the cranium of a victim
killed in warfare.^^ The Newars of Nepal apparently show
traces of the separate treatment of the head in disposing of
their dead, a feature frequently associated with head-hunting
in Assam, Indonesia, and Oceania, Melanesia in particular, .and
undoubtedly based on the same idea. Head-hunting as a neces-
sary preliminary to marriage, as it is in most if not all genuine
head-hunting tribes, is to be explained by the idea that unless a
man has taken heads he has no surplus soul-matter about him
to beget offspring,-*^ Probably the same notion is to be seen in
the Chang Naga practice of naming a child after a village raided
by his father. Thus the chief of Yongemdi was named Long-
khong and his brother Ongli after the Ao villages Lungkhung
and Ungr which, were successfully raided by their father with
considerable slaughter about the time that they came respectively
into this world. The same notion is clearly present in the cases
that come to light from time to time in India of murder as a
remedy for barrenness in women. Thus in October 1929 the
High Court of the Punjab had to deal with a case in which a
girl, desperately anxious to bear her husband a son, killed a
child, cut off its hands and feet and bathed herself standing upon
them. It cannot be doubted but that the idea was that the life
of the dead child should become the life of a fresh child in her
womb. Again early in 1930 a case occurred in Gujarat of the
murder by means of sulphuric acid of a 12 months old boy by a
girl of 20 who had no diild, and it was part of the prosecution
case that it was a comparatively common practice on the part
of barren women to attempt to quicken themselves by burning
marks on children in the street, a practice no doubt ultimately
feature of head-hunting practice as with the unusual word coined by
Strabo to describe it ; Casaubon indeed explains the name of the tribe as
a facetious metaphor from Persian trousers {sarabara^.* jodhpurs ) dimin-
*

ishing at the knee % A Naga head-hunter who does not actually remove
and suspend from the village head-tree the foot and leg of his victim will
frequently slash the legs in order to entitle him to wear the embroidered
gaiters of a warrior who has taken his enemy's legs, while some tribes on
the north bank of the Brahmaputra are reported to cut off the hands and
feet of their enemies (though they do not decapitate), probably to hamper
possible attempts of the ghc^t to pursue and harm them.
3 Special importance often attaches to the lower jaw among
head-hunters
in both Asia and Africa, and sometimes to the tongue, which is regarded as
the seat of life.
39 portman. Relations with the Andamanese, i, p. i8a.
^0 Northey and Morris, The Gurkhas, p. tSg.
Conversely I have known a gardener in the Simla Hills refrain from
setting grafts himself since to do so would prejudice his chance of begetting
children, clearly on account of the loss of life-matter; instead he called
in an old man past the breeding age to do it for him.
BiNDUIS-M AND PRIMITIVE RELIGIONS $41

derived from one which involved, the taking of the child's life,
and comparable with that referred to below of branding children
offered to the Syrian goddess.'^^ And if head-hunting is rare in
India, human sacrifice, on the other hand, has been widespread
and has clearly been ultimately based on the same conception
of the necessity or at any rate the desirability of releasing soul-
matter to fertilize the earth. No doubt it was later interpreted
as the placation or propitiation of an earth deity, but this must
be regarded as a sophisticated justification of a practice the true
meaning of which had become obscure or been forgotten. The
Kondhs are described as having performed their meriah sacri-
fices to the earth mother, but the details of the ceremony and
the practice of distributing fragments of the sacrificial meat in
their fields and granaries show a very patent connexion with
the disposal of enemy flesh by head-hunting Nagas and the
underlying idea is undoubtedly the same. In one form of the
sacrifice the victim was squeezed to death in a cleft in a
green tree, and in another the tears caused by his sufferings
Drought rain in proportion to their profusion. Similarly, the
Wa of Burma definitely associate their head-hunting with the
sowing of the crop,^^ while the successful Kafir head-hunter was
greeted, on his return from the foray with his trophy, by a
shower of grain.-* In Kulu the transplanting of the rice is
accompanied by the sacrifice of a rough dough image of a man
to the house god.^ So again the Dasehra festival, now asso-
ciated by all Hindus with the killing of Ravana by Rama,
coincides throughout most of India with the sowing of the winter
crops, in particular with that of millet, a more ancient staple in
south Asia than rice, as well as with that of wheat. It is this
festival that is associated in western India with the worship of
weapons of war (and it is still regarded there as a proper day
on which to go forth and loot), while it is then that human
sacrifices used to be performed in eastern India; and it is still
on this festival that the gupta puja, the hidden rite, to ensure
the prosperity of the person, house or family, would be resorted
to it ever. The association of human sacrifice with the^ prosper-
ity of the individual and with the success of the State in war
seems clear enough, and its association with crops may be in-
ferred with equal safety. It was probably some such association
of soul-substance with fertility, and perhaps with some notion

p. 54, infra*
^3 Macpherson, Memorials of Service in India, pp. 113 sqq.; Thurston,
Castes and Tribes, ni, 369 sqq., and see Reclus, Primitive Polk, pp. 304-35.
Scott and Hardiman, Gazetteer of Upper Burma and the Shan States,
I, pp. 500 sq. Robertson, kdfirs of the Hindu-Kush, p, 153.
Grooke, Religion and Folklore, p. 853, and ct. Pant, Social Economy of
the Himalayans, p, 111, where he says that the women transplantinr rice
offer * a red powder ... to heaven ' and apply it to their foreheads. Ochre
has definitely some ceremonial association with fertility, probably being
associated with bloodand semen.
IS
APPENDIX B.

of a higher fertility value attaching to Europeans, that led


Oraons to remove portions of the body of a recently buried
European not so very long ago. The evidence at the trial made
it clear that special value attached to European bones for magi-
cal purposes.^^ The location of tiie soul in the head and the
confusion of the soul with the shadow are illustrated in the
Himalayan cure for fever in which the patient stands in the
sun and a bone filled with grain is buried in the spot where
the head of the shadow falls*'* The idea of soul-matter as
a fertilizer again is probably at the bottom of human sacrifice
as a cure for illness, as in the case of a Santal of Dhaiibad who
in 1931 garlanded and then beheaded his infant son in order to
effect a cure of his own maladies.
True cannibalism is only traditional in India but vestiges of
ceremonial cannibalism survive in many places or have done
until recently. Thus Portman records the practice of Anda-
manese homicides, who drank of the blood and ate of the raw
fat of the victim and of the flesh of his breast, the latter appa-
rently cooked. In the north-west the Kafirs used to eat a piece
of the heart and drink some of the blood of the enemies they
killed. In the north-east the Lushei of Assam used to taste the
liver and lick from the spear-head the blood of the first victim
slain in war.i The nearly related Thado eats his first meal after
taking life with hands aeliberately imbrued with his enemys
blood and still uncleansed therefrom, a custom practised as lately
as 1919, and the same custom obtains in a decayed form among
the independent Sernas of the Assam frontier, where a returning
head-taker must eat at least a morsel of food before entering his
village; the insistence is on the meal now, but it is eaten with
still ensanguined fingers, as he may not cleanse himself until
after he has taken this meal. The intention in all these cases
is certainly to transfer to the slayer the soul-matter or life-matter

of the slain, just as the soul-matter of so many kings has been


transferred to their successors by their murder. Similar in
^

principle is the practice of anointing at his succession the intru-


sive Rajput ruler with the blood of the indigenous Bhil who
regards me right to give it as a precious privilege, even though
the giver is believed always to die within the year, his soul-
matter exhausted no doubt in providing for the fertility and
prosperity of the State (compare Forsyth, Highlands of Central
India, p. 145). It was no doubt a similar idea to that of the
Kafir and the Thado which inspired the action of Nana Pharari,
a notorious dacoit of Nasik in Bombay Presidency, who stabbed
a personal enemy in July 1930, pulled out his knife from
See Indian Antiquary, December ig^g.
Crooke, Religion and Folklore, p.
Relations with the Andamanese, i, pp. gi, 115.
fioLeimer, op. cit., pp* 14, 53, 61,
Lewin, Wild Races of South-Eastern India, p. ^69.
HINDUISM AND PRIMITIVE RELIGIONS S43

the victim and applied its gory blade to his forehead and to
his tongue. The action of applying the blood to the forehead
offers a very close parallel to that of the Angami Naga who never
drinks liquor without -applying on the tips of his fingers a drop
to his forehead for the bea^t of the materiai soui resident
within. The drinking of human blood and the tasting of human
flesh are common in Indonesia and Oceania, and it is likely
that it has at one time been more prevalent in India than it is
now. It was reported of the Wa
of Burma by Sir George Scott ^2
that probably human flesh was eaten on special occasions, poss-
ibly at the harvest festival. The Wa
are also credited by the
Shans with eating their dead relatives the Batak of
Sumatra, and this, a practice probably arising rather from a
belief in reincarnation than directly from that in life-, or soul-
matter, has also been reported of some tribes in India. Thus
Gardiner reports it of a tribe in the Himalayas, who eat the

heart and liver of their own dead'. The Kafirs too, he says,
occasionally eat the ashes of these organs, mixed with herbs, as

a sign of affection Herodotus mentions the Callatians as


an Indian tribe known by JIarius to pra,ai^ th ;
, ,

attributes the same custom to the Padaei, for whom an identi-


fication has been suggested with the Birhor of Chota Nagpur
whom Dalton states to have admitted its former practice by
their tribe.^ The same custom has been likewise attributed to
the Lobas of north-east Assam (called Mishu Ting Ba by the
Tibetans) in particular, to the hill tribes of Chhattisgarh and
of the Amarkantak tableland,^ as to the hill tribes of Assam
in general, and to some of the transfrontier Kachins of north-
west Burma, though in these cases there 4s no well-authenticated
evidence. A possible survival of the same practice is to be found
in the Kharia custom in the Central Provinces of catching, and
eating, communally, a fish on the third day after a person's
death, the fish being a common vehicle of the soul as is noted
elsewhere in this chapter. All this points to some pre-existence
of this practice and to a clear cultural link between the more
primitive tribes of India and those of the Indian archipelago.
Scott and Hardiman, op. cit., i, p. 497.
p. 498. The same authority says that
S3 Scott and Hardiman, op. cit., i,
the Shans eat human flesh at certain tattooing ceremonies (ibid., p. 357).
3 ^ Edgeworth, Abstract of a Journal kept by Mr
Gardiner during his
Travels in Central Asia in J.d.S.B., no. 3 of 1853, p. 303?.
S3 III, 38. 36
99.
37 Ethnology of Bengal, pp. 158, sq.
38 Duff-Sutherland-Dunbar, A tors and Galongs, p. 5.
3 Rowney, Wild Tribes of India, Wilson, Indian Caste, i, p. 358.
p. 7 ;
60 Russell, op. cit., m,
p. 450 ; Roy, The Kharias, i, p. ^95 sq.; Crooke,
Religion and Folklore, -p. 228 sq. (and cf. p. )?44). Br Verrier Elwin tells
me that this custom is not confined to the Kharia tribe, but known and
practised all over the Central Province*.
;

S44 APPENDIX U
Pcriiaps tile cFucIfist forni in whicli tiic doctrine
of soul
substance appears is the vujgar but widely credited
superstition
which attributes to the European the prattice of
catching fai
black boys and hanging them by the iieels over a slow
to fire
distilhorn a puncture in the skull the seven drops
of vifil
essence which impart to sahibs in general tiicir
eneiVy i
sports and their activity of mind and body.
Curiousiv enoudi
this life-essence, this momiyat, seems to
have startetl as' bitumen
simply, - or some similar mineral substance,
and to have been
used as a quite legitimate metlicine, then
to have become a
spurious substitute in the form of resin, the
supiJosed virtues of
which were later attributed by confusion to
the embalmed bodies
from which this resin was most readily obtained.
body a fourth transfer has taken place
From the dead
and superstition now
imputes the virtue of the medicine to
its distillation from
the
living body in the form of

blTn
hgo
f its life-essence.
strangers to be beaten in Saharanpur
on the suspicion that they were
not long
'Fins belief cau ed

manufacturers if this elixir


mentions that this belief gave some
trouble during
r 1896 in Bombay, and is
active.
dearly sdi!
Possibly it has some bearing on
the reluctance felt in
India to feraain in a hospital.
Here again it is by Hindus or
^ that this superstition is generally
and the Kabuli trader who bringf held
momiyai for sa e from
or virtue. Ihe theoiy is also to be seen in the composition
philosophy AghoriDanth
and in occasional cases of cannibalism
to light m that^come
the criminal courts. Thus
in September 1931 two
Warently a Rarhi Brahman ascetic and
ff K
Mahabrahman, the othe^a
were accused in Bankura of having
dug up the
iiaving there cooked and
r V eaten part of iv the
Brahman admitted having eaten a
littlc^of the Iieart as he
religion to do so
narMv^l '. It is perhaps

attached in India
to hi rSrSuetToi 3 is

he nexT^generaS L 1 reduplicated to
primftive tribes
me SitvTf
the penalty of fS;,. ?o m
failure to marry is extinction
at the hands of a
demon who bars the pathway of the
her of Hmdu cartes, as also
dead, while aj^ona a
among dte Toda of
mm
InS"' iSerf
* Crooke, Religion and Folklore,
pp m sq.; CJordon, '
MomiySi ', in

Nazaror^e
Nazaroff, The White m "aL'p.f'll*
Lady^ fn
White' LI?-' of regasus,
Pegasus, ch. via;
cb. viii
y? S(J.: Huntingford, ^oiiyai f In
iPP- '

!vy A, red..,
HINDUISM AND P R*I M 1 T 1 V E RELIGIONS
the corpse of a person dying unmarried is married before crema-
tion as a necessary qualification for future happiness.^ The
doctrine of soul-substance as a fertilizer is naturally not less
applicable to animals than to human beings, and it is therefore
not surprising to find the Mala of southern India and Ahir at
the Gaidaur festival causing their cattle the young in particu-
lar to trample a pig to cleath, after which, according to the
ancient custom, the corpse of the pig is eaten by the Aliir who
thus share in the transfer of the porcine life-substance to their
cattle. In the case of the Korava, who have a similar practice,
an instance of the substitution of a human being for the more
usual pig is actually on record,*^ the unfoitiniate victim having
been buried to his neck before the cattle were driven over him.
Involved again in the belief in soul-matter probably are the
practices of erecting megalithic monuments ana wooden images
of the dead. The two practices are not completely separable,
as both appear primarily to be intended to afford a texmporary
dwelling lor the soul pending its operation as a fertiK/er of the
crop. The megalithic monument appears very often as merely
a permanent substitute for the impermanent wooden statue
which can be given greater resemblance to the human body.
Thus the. wooden statues of the dead put up by the Angami
Nagas of Assam are in some villages destroyeci after the harvest
and the others have a small stone erected behind them to do
duty when they have perished. In other villages again a mans
\youngest son, on succeeding as he does to his fathers dwelling-
liouse, must put up a monolith for his deceased parents, an act
corresponding to tnat of some other villages in which the mono-
liths are erected during their lifetime by specially prosperous
persons to enhance the prosperity of the community as a wnole.
The significance of the latter monolith is quite definitely phallic,
and ancient specimens still exist elsewhere in Assam whose
form puts this beyond dispute, both as solid menhir and as
hollow monolith which contained the ashes of the dead; and
there is no doubt but the association is here again with the soul-
matter as a fertility agent, and an echo of the doctrine is perhaps
to be found in the Vijayanagara legend of the head of the hero
Ramanatha which, when returned to Kummata, became united
with the lingam of Shiva, It may be noted that ICampila, the
defender of Rumraata, appeared, or at least his troops did, under
Thurston, ibid,, pp, 106 sq.; Omens mid Siiperstitions, p, 51 Castes ana
Tribes, I, pp, 5Q sq.i v, p. 197, vi, pp, m ; Iyer, Coc/iin
;

and
Malabar Marriage Gomnussion,
Tribes
Castes, n, p, 19S (quoting the report of the
which his own inquiries did not confirm); Wilson. Indian Caste, n, pp. 75,
77, Thurston, Castes and Tribes, m, pp, 463 sq,
s Hutton, Angami Nagas, pp. 13(1, 533 ; Monoliths at Dimapur, pp. 58,
Erection of Monoliths, pp. sq,; Use of Stone, p, 85.
67 ;

Mills and Hutton, Monoliths of North Cfldmr ; Hutton, Monoliths at


Dimapuri Monoliths at Jamuguri ; Signifimnee of Head^hunting,
246 APPENDIX B
the guise of head-hunter, while dolmens have actually been
used as Saivite temples; in the Naga Hills the cult of head-
hunting is, like that of the dead, associated with menhirs and
dolmens and other symbols of fertility like the milky ficus or
euphorbia treesJ^ The remains of this megalithic culture in
India are widespread. though in most places completely decadent,
and they generally show sporadically very similar traits. Thus
the disposal of the dead of the community during the sowing
season has already been mentioned, and in farther India it is,
or used to be, associated with a certain amount of mummifica-
tion of the dead to make them keep, some tribes, e.g. the Ao
Naga, smoke-drying the body, others, e.g. the Khasi, embalming
it in honey. The use of a soul-figure is merely a different
method of obtaining the same end, and by the Konyak Naga
of Assam a wooden figure is provided to house the soul until
the head can be separated and disposed of in a phallic stone
cistJ2 Soul-figures, probably of similar purpose in the first
instance, are made in earth by the Handi Jogi of Mysore.'^ The
Nicobarese, like some Naga, place the skull of the dead on a
wooden body, and at the census of 1931 such a figure was to be
seen on the island of Teressa, the skull doing duty as a head
incongruously surmounted by an old top hat, the treasured
headgear of the dead manJ^ It is impossible not to see the
same idea underlying the waxen figures of deceased Hindu
princes already referred to, accompanied as they are by a stone
fingam. It seems therefore not unlikely that the carved stones
erected to the memory of Rajput dead of both sexes have
a similar origin, and table stones may occasionally be seen in
Rajputana erected to mark the site of a sati, recalling the fact
that the dolmen is used as a memorial of the dead by the Munda
of Chota Nagpur, while by some Assam tribes the upright mono-
lith and recumbent dolmen are used to correspond respectively
to the male and to the female sexes. A
reference to voL ix of
the Bombay Gazetteer definitely confirms the supposition that
the Rajput memorial stone has the same origin as the Naga or
Khasi menhir. Unhewn stones (khatra) or carved stones are
raised, we learn, by most classes of Hindus in Gujarat for
deceased persons, sometimes for all, but more particularly for
those that have died a violent death or been remarkable for
70 Journal of the Mythic Society, Tanusry 1930.
71 Hutton, Two Tours Mast of the Nam liilk, pp. 6, m, 39, 41* ;

Mills, Lhota Nagas, pp. stB, 108, iJig.


72 Hutton,
The Disposal of the Dead at Wakching % in Mm, xxvii, 44*
73 It. A. Iyer, Mysore Tribes and Cosies, iii*
p. 500*
7^ Bonington, loc. dt.
7s xhe Census Superintendent of Madras
inay be conveniently quoted here
on similar practices in South India i
*
Pudams or shrines exist to which no priests or temples are attached
and the pfevailing worship is in fact a kmd of goblin propitiation, the
goblins being usually the spirits ol persons who died a violent death.
HlAfDUISM AND PRIMITIVE RELIGIONS 547 1

!|

These stones are sometiraes placed in a shrine,


their holiness.
j

sometimes under a pipal tree. Until the shrine is set up the


spirit of the deceased
is dangerous. We
may perhaps infer that
tne reason why such stones are set up, particularly for persons ,
j

whose spirits are likely to be dangerous, is in order to appease i

||

them by providing an abiding place for them and so conforming i;

to an ancient custom otherwise liable to be forgone. The I,

carved stones take the ordinary form of a Rajputana or Kathia-


war memorial stone or may be replaced by *a bust of black 1

marble, brick or wood, but it is significant that they are wor- jj

shipped by the newly married, or by a bridegroom on the way jj

to fetch the bride, recalling the ceremony performed by childless 1 1


!

Konyak Nagas on their phallic stone skull cists, and the practice I
ij

of barren women to strip themselves at dawn and embrace naked :


j

the stone slab carved with the image of Hanuman.'^ At certain 1 1

temples in southern India barren women are or were seated i|

astride a particular stone to get offspring, success depending 1

upon the experience of an orgasm while in this position. In !|

the Punjab hills again (in Chamba, for instance), monoliths, |i|

or wooden substitutes, are put up for the dead, with feasting jii

on a great scale, and for the sake of acquiring merit. rough A j

effigy of the deceased is usually carved on the stone, and the Hi

wooden substitute sometimes has a hole and a spout for water, j |

when it is set up in the stream beside which it would otherwise |

be placed. Sometimes a circular stone is placed on the top of j

the monolith (recalling the former Khasi practice), and in the ;


i

case of ruling families (e.g. of Mandi and Suket, where the


practice is confined to royalty) the wife and concubines of the
deceased are also represented on the stones, as they are by Assam
hill tribes. The Mala-Arayan of Travancore make a metal j!

Animal sacrifice and frequent admixture of human blood are common- : ij

places in their ceremonies. One such shrine in Tinnevelly district is to


the spirit of a European killed in the Travancore wars and the offerings H!;

made are of articles considered peculiarly acceptable to one of his race: I


li

bread, fowls, cheroots and brandy. The spirit of Muhammad even is said i j:

to inform one granite pillar in Tinnevelly where daily puja*" is done if


by Hindu votaries. Vows are made to it by Hindus who flock to seek cure !|!

of disease, rain, and other boons, panja ahd cheroots are the form the j
offerings take, these being considered peculiarly attractive to Muhamma-

dans. . .. In effect the real religic|i of th^ presidency, in the south, at


any rate, is directed rather towards shrines ind saints than towards deities.*
(Yeatts, op. cit., p. gao.)
One may compare with these practices that of embracing a certain
pillar in the church of the Virgin at Orcival ih Auvergne or sitting in the [j

chair of St fiacre in the church of the village of that name, the stone seat [
ji

of which, like the pillar at Orcival, had the power of rendering barren
women fertile. It was necessary, however, that there should be no garment
between the stone and the sitters body. Similarly in the chapel of St H
Antoine de Paule at S^agossa there was a tombstone on whi^ barren
women lay in order to become fruitful. (Dulaure, Des Dwinim GMra- rj

irtc^Sy p. 151.) .
APPENDiK B
effigy of the dead and put it into a miniature cist of stone verti-
cals and a capstone which is erected on high ground and wor-
shipped annually. Even the method of transporting megaliths
seeim to have lett traces in western India. The Naga of Assam,
like the people of Nias in the Indian archipelago, transport
megaliths on wooden sledges made from the forked trunks of
trees which are dragged by very large numbers of men pulling
on cane or creeper ropes. In the case of the Naga, at any rate,
an essential implication of the ceremony is the infection of the
village with the prosperity of the celebrant, and when the cere-
mony is performed with wooden instead of stone emblems these
emblems are formally dragged all round the village with this
express purpose.'^ Very suggestive of a degenerate form of this
ceremony is the village festival (ghasbavji) in Rajputana at
which the god, consisting of a large waterworn boulder,
is dragged round the village on a sled made out of a forked tree
trunk. Probably of similar origin also is the general veneration
paid to stones throughout India, particularly of course to those
of queer or unusual shape. Crooke gives a very large number
of examples which it is unnecessary to recapitulate, but it
may be recorded here that a suit was argued in the Calcutta
High Court on 2S April 1929 about a stone about 5 ft. stjuare,
apparently of black slate or marble from Jaipur Qeypor^, which
*
changed hands for Rs. 10,000 (75(^ as being very efficacious
in the matter of getting a son'. 'The suit arose because the
stone failed to function, and it was stated in evidence that num-
bers of even quite well-educated Hindus believed in the efficacy
of stones of this kind when used with the correct rites, and that
so much as a lakh might be paid for such a stone. It was also
mentioned that sitting on a stone is an essential feature in many
Jain rites.
It may here be urged that the reverence and superstition
paid to stones in general, Is not. as Crooke suggests, a vague
superstition which develops into the use of memorial stones,
but on the contrary is the degenerated remnant of the life-
essence fertility Cult. The use of a stone or mere pebble as a
pretyasila, 'stone of the disembodied spirit', by Hindus in
western India seems a definite instance of this process of decay.
The stone is picked up by the chief mourner at the place where
the corpse was put down and is anointed with oil (recalling the
baetyls of Naga fertility cults ). a crow Is induced to eat corn
scattered about it and the pebble is then thrown into running
water or kept among the household deities. Similarly, the
Koraati caste of Mysore invokes the soul of the deceased to
enter a pebble temporarily, while the ashes of the deceased
n Hutton. Carved Mcmollffia M Biinapur and an Angami Naga Cere-
mony (j Meaning and Method of the Erection of Monolithli by the Angami
Nagas , in wi. Migion and Fotklore, eh, wi.
n Hutton. Thi Sm Nagat. p. 154.
HINDUISM AND PRIMITIVE RELIGIONS S49

are cast into a sacred, river or some other water. It


seems therefore much more likely that a veneration of stones
on account of the indwelling soul has been transferred to
peculiar stones in general on account of similar possibilities,
than that a vague reverence of stones should exist without
reason and be elaborated into a coherent creed which really
turns on a belief in the power of soul-matter to promote the
fertility of nature. An extension of the same idea would seem
to appear in the practice of requiring spirit or soul-matter of
the dead to impart permanence or, to use our own metaphor,
'life' to buildings. This practice is referred to as that of
foundation 'sacrifices', but the idea of sacrifice, if present at
all, is dearly later than the practice itself, which is essentially
of the same origin in India, where at least a belief in the
necessity survives, and in Oceania, where in Fiji and in the
Marquesas, for instance, human sacrifices were required to
invest a building or a canoe with the necessary mam. This
belief is still so prevalent in India as to be the cause of a good
deal of disturbance from time to time. In 1922 the Deputy
Commissioner of Dibrugarh had to issue a notice reassuring
people against mur katas (=' head-cutters ) coming to kill
or behead male children for some unknown purpose, and to
threaten with prosecution anyone assisting to spread the
rumour. In May 1923 a riot took place on Cinnamara tea
estate on account of a scare about the kidnapping of children
for the foundation of a new bridge. In June 1924 a rumour in
Calcutta that the Port Commissioners were seeking for children
to bury in the foundation of the new Kidderpore dockyard led
to Punjabi taxi-drivers being killed in the belief that they were
agents decoying children for this purpose, and in the same
month strangers in Patna were maltreated in connexion with a
similar scare about a bridge in Bihar. In 1926 in North
Lakhimpur there was a rumour that mhos (i.e, head-takers?)
were trying to obtain children's heads for the construction of a
railway bridge, In the same year the district engineer construct-
ing a recent bridge on the Sibsagar Road-Khowang Railway had
trouble from a kidnapping scare, which was so much intensified
. when a retaining wall burst that schoolboys asked leave in order
to stay at home for fear of being kidnapped on their way to
school. Several petitions were filed in the S.D.O.'s office, also
against suspected kidnappers, while villagers insisted on visiting
and counting tea-garden coolies during the night in order to
assure themselves that they were not concealing victims in their
lines, and a number of assaults took place, one serious. In July
of the same year there was a scare in Cachar that a human
saaifice was required for the oil-borlngi to promote the flow of
oil in the Burmah Oil Companys wells. A
strange coolie was
* L, K. A. Iyer, Myiora Tribts md Ct, i, pp. 87> sq-
S50 APPENDIX B
roughly handled on the 12th of that month in Cachar and a
stranger from Sylhet was beaten and confined on the same sus-
picion on the lOth. In January 1929 a rumour in Bombay that
children were being kidnapped for the construction of a bridge
in Baroda led to an attack on Pathans and on a Greek engineer,
to the death of a Hindu carpenter, and ultimately during two
days' time to the killing of seventeen Pathans, while two other
Muslims and three Hindus were killed in the consequent affrays,
very many other persons being injured.' In the subsequent riots
149 people were killed and the clamage to property came to at
least 5 lakhs of rupees (37,500). In May 1929, in Bihar again
this time, a man was beaten to death in the belief that he was
seeking to kidnap a child as a foundation sacrifice for a bridge

at Jamshedpur, while the floods in the Surma valley of the same


year led to a scare that children were being kidnapped for the
reconstruction of certain embankments in the Manipur State.
In certain tribes of the Naga Hills in Assam a live chicken is
still placed in the hole dug for the main house post and the
post stepped on to it, while it is asserted that certain transfron-
tier tribes use a human being thus, as formerly in Fiji, to hold

up the post '. The same idea is to be seen in a practice attri-


buted by Mr S. N. Roy^ to Bengali boat-builders who have
had to dispute their dues with a customer; a few drops of the
shipwrights blood are plugged into a cavity in the planking,
when the boat acquires a malevolent vitality, drowns its crew
and continues an independent existence as an aquatic phantom.
A sworn, used in sacrifice acquires a similar vitality from the
blood of the victims it decapitates, and, like boats, it is furnished
with eyes painted in vermilion.
It has already been indicated that the soul is often conceived
of in the tribal religions of India as having the form of a mani-
kin and being located in the head, and though this conception is
apparently at conflict with the theory of soul-matter or a mate-
rial life-substance, it is held concurrently with it without any
consciousness of inconsistency. There is nothing remarkable in
this, but the question arises whether these two conceptions are
of different origin or may have come into being from the same
source, and it is here submitted that the conception of the soul
as a manikin is merely the effect, on a rather vague conception
as to the nature of life, of that tendency towards anthropomor-
phism which is inevitable when man is to conceive of a material
with some of his own attributes and no known shape, a tendency
which is apparent in the conception of the deity in the great
majority of religions. The Karen of Burma have a doctrine
of a material substance which is the cause of life, or rather
which is actually life elf, a sort of ectoplasm of life, which
'

1 Mon in India, ix, no. 4, p. *7*.


ManbMi Karen Poffple of Mutma, p. SMM.
HINDUISM AND PRIMITIVE RELIGIONS 251

leaves the dying body to enter the herbs of the field or the seeds
of the earth, and which then passes through grass into cattle
and through grain, or indirectly through the meat of grazing
animals, into man and passes through the seminal fluid to
generate fresh life, almost as if it consisted of carbohydrates.^
This doctrine may perhaps be regarded cs having arisen very
early in the history of mankind as a natural result of speculation
as to the cause of the change that takes place at death and as
to the nature of that which has left the body. Speculation of
some kind would be inevitable, if only as a result of the natural
curiosity required by any animal, human or otherwise, to adapt
itself to environment in the struggle to survive. Abstract ideas
come late in the development of a language and presumably
therefore late in the development of thought, and hence the
necessity, before any philosophic idea of life can be framed, of
regarding it as a material substance and of thinking of that
substance as taking some form. The idea then of a manikin
living Dehind the forehead whose movements are registered on
an infants fontanelle is a not unnatural symptom of the deve-
lopment, or degeneration, of the life-substance theory and may
perhaps be traced in that doctrine of the Vedanta school of
Hindu philosophy which regards the soul as encased in a series
of sheaths, the interior of which accompany the soul on it?
migration while the exterior constitute the material body; and
the location of the soul in the head is illustrated by the Hindu
belief that it escapes through the crevice of Brahma *, through
*

which ascetics can project their soul (and so die) at will, while
for less holy persons it is necessary to fracture the skull with a
conch shell to let out the soul. The saldgramd held to the
aperture perhaps served the same purpose as the pretyasila
mentioned above. One is also reminded of the story in the
Aitareya Brahmana^ of how the gods killed a man for their
sacrifice, but the part in him fit for an offering went out, leaving
him deformed, and entered a horse, and so on through an ox
(which turned to a gayal when the fit part left it), a sheep and
a goat and entered the earth, where the gods surrounded it so
that no escape was possible, when it turned into rice. Father
Schmidt considers that in India the materialistic Sankhyan
philosophy most certainly arose from matrilineal animism,
and he suggests that the spiritual philosophies of classical Greece
may have had at least in part a similar source. Having ^t as
far as materializing our life principle, its conception as similar
in feature to the human body is inescapable, and the way is clear
*3 It seems certain that the strange provision in the Brahmanical code
which makes the Telis an untouchable caste (but not the Tilis, who only
sell and do not press oil) is due to their practice of destroying the seed
in the pursuit of their occupation without provision for its transfer to
another living organism, a suggestion confirmed, I think, by Manti, iv, 85.
** Vide itajendralaia Mitra, Indo-Aryans, 11, p. 76.
'

S5a appendix B
for the doctrine of reincarnation. The doctrine reached is logi-
cally irreconcilable with the theory with which we started, but
experience shows that the two can be held simultaneously, at
any rate by primitive man, without any consciousness of incon-
sistency. There are probably, however, steps by the way which-
contribute to this belief the idea of the soul coming back as
an insect is one, and one to which the conception of the soul
as able to leave the body and flit about at night, derived in
pan no doubt if not entirely from the phenomena of dreams,
has contributed. As instances the Lhota Naga of Assam, the
Kunbi of Bombay, and the Kami and Bhuiya of Bengal may
all be quoted as watching for an insect after a persons death,
or the Ahir and Gond who go to a river and bring back an
insect or a fish as containing the soul and sometimes, the case m
of the Gond at least, eat it to ensure its rebirth.* Another con-
tributory observation to the insect notion is perhaps the
mysterious way in which large numbers of in.sccts appear from
nowhere in particular, as if caused by superfluity as it were of
life-substance, an idea which would have been comprehensible
-nough to the ancient world, which regarded, for instance, insect
life as spontaneously engendered in dung dropped under a wax-
ing moon, ctc.,8 which regarded the Nile floods as pouring
soul into the sods so as to fashion live creatures from the very
soil, bringing us back again to that fertility cult which asso-
ciates soul with water. Similarly, Diodorus Siculus says,
moisture venerates creatures from heat, as from a seminal prin-
ciple aiiw a little farther on, they say that about Thebes in

Egypt, after the overflowing of the river Nile, the earth thereby
being covered with mud and slime, many places putrify through
the heat of the sun, and thence are bred multitudes of mice. It
is certain, therefore, that out of the earth animals are gene-
. . ,

rated ,* These ideas of life in dung and water seem to have


been combined in the practice which so disgusted the Abbe
Dubois at the temple of Nanjanagud in Mysore some 150 years
ago. Barren women and their husbands were described by him
as drinking out of the temple sewer from hands soiled by setting
aside a portion of the ordure to be examined a few days later
whether insects or vermin were engendO'ed in it, which was
regarded as a favourable prognostic for the women ,*8 What-
ever be its origin, however, a vague belief in reincarnation is
common to most of the tribal religions in India and is generally
associated more or less with some degree of ancestor worship.
ssCrooke, BeUgion and Folklore, SSi Eoy, Mill Mhufym^ p, MO$
I

Rlsley,

Tribes and Castes, i, p. sgs ,UMU op, cit, II* p. tS, iii,,p,
Vide Pliay, Nat, Hist1,. f-k. xvui, ch. xxx;
ixlii oditof *1 ooto,
*^[Nilus] . . . glaebis ett infuniul mirms, m ipaqm hmm
viulk
f^ngaiPoBiponiui Mela, i, ss

mtory, 1,1. Fiopit 0/ indk, ppvjp iq* cli, ill)*
EiNBUISM AND PRIMITIVE RELIGIONS
a tendency ^to which is everywhere apparent. This reincarna-
tion belief is to t . seen very clearly m
the ancient Brahmanic
theory that after the birth of a son the sexual relationship of
husband and wife should end, since the son is the father's self
and the father's wife has become his mother also. It is stated
of the Kochhar subcaste of the Khatri, a trading caste of the
Punjab, that funeral rites are performed for a father in the
fifth month of his wife's first pregnancy, which points to the
same idea. The Bishnoi of Hissar bury an infant at the thres-
hold that its soul may re-enter the mother and be born again.^
Among the Bhuiya every child is regarded as a reincarnation of
some deceased relative,^ while the Mikir, it may be noticed,
believe in reincarnation except for the souls of those who have
been killed by tigers. Among the Lushei the reincarnated soul
sometimes appears as a hornet, sometimes as dew, and in the
latter form the belief is hardly distinguishable from the Karen
theory of life-substance. So again it is a common practice with
the tribes mentioned that while a dead grandfather's name or
that of another ancestor must be given to a child, the name of
a living ancestor shall not be given, as, if it were, either he or
the new-born child would die. The practice was perhaps similar
in ancient Indian society, as in old lists of kings it is common
to find a grandson named after the grandfather. This practice
seems, however, to have changed, as the name of any ancestor
living or dead is reported now to be avoided by Hindus.
The association of reincarnation with the soul-fertility cult is
perb .ps confirmed to some extent by Malcolm's record^ of the
practice of jumping off certain high rocks in central India in
order to be reborn in a royal house. Forsyth also records 5 the
account of an eyewitness, Caplain Douglas, Political Assistant in
Nimar, of a scene at Omkar, a shrine of Shiva on Mandhatta
island in the Narbada, at which a young man leaped off a rock
90 ft. high to his death in 1822, and mentions a later case of
an old woman who hesitated and was pushed over. In the case
of this rock apparently if the jumper survived he was killed by
a priestess with a dagger, but in the case of another of these
'

rocks if a man survived the fall he was made Raja, and the asso-
ciation between the soul and the fertility of the land impinges
on that between the fertility of the land and the king as the
living receptacle or embodiment of the life-spirit, and one which
must not be allowed to grow old. One is reminded, however,
by this habit of jumping off a cliff to royal incarnation, of a
number of similar practices associated in each case with the
Golden Bough, iv, p. 189, and cf. Crooke, Northern India,
pp. Jtos sq. Kaul, Report on the Census of the Punjab, 1911, i, p. 1199.
Roy, Hill Bhuiyas, p. 186.
93 Shakespear, Kuki Clans, pp. 6439.
94 Memoir of Central India, ii, p. aio,
highlands oj Central India, pp. 180 sqq.
554
APPENDiX B

fertility of the soiL Thus rope-sliding (beduart) in the


Himalayas would appear to be a definite survival of a similar
form of voluntary or involuntary human sacrifice. The slider,
an acrobat or dancer (beda) by caste, is worshipped as Mahadeo,
bathed in milk, dressed in new clothes, and carried round the
village fields before the ceremony, which is resorted to when
harvests have been bad.^ That this is a survi\'al of human
sacrifice is clearly indicated by the fact that both the rope used
and the hair of the slider are distributed as fertility charms,
while the slider himself becomes infertile, for his fields go barren
and the seed he sows fails to burgeon. He is, in fact, spiritually
dead and his life-matter has been distributed to liis neighbours.
We are recalled to the Kondh rneriah whose parent was con^
soled by his neighbours in words which Macpiierson has
recorded: Your child has died that all the world may live.'
'

It is worth remarking that the hereditary caste of rope-sliders


is the Nat caste, and that the women of that caste are associated
in many parts of India with dancing and {uostitution. Both
are probably closely connected with fertility rites, and it may
be that the professions of tight-rope walker, acrobat, daxtcer, and
prostitute take their origin in services performed primarily for
the benefit of the crops. It is said that in former days the slider
who fell off the rope was cut to pieces at the bottom, and the
rite as a whole suggests a chastened form of human sacrifice in
which it was essential that the victim should fall from a great
height. It recalls insistently the ceremony described as practised
at the temple of the Syrian Goddess where from a very lofty
porch between two gigantic phalli the animal or sometimes
apparently human victims were hurled to the ground, and where
worshippers let their children down in a sack after branding
them^ as devoted to the deity. In the north-east corner of
India the Angami Naga still hurls from the roof of a house his
sacrificial victim, a puppy dog invested with the symbolic attri-
butes of a man, while a calf released below is literally torn to
pieces by the crowd. Meanwhile asShes representing Houds and
cotton seeds representing hail are thrown by the priest from the
roof, clearly showing that not only is a distribution of the
victims life-essence involved but a fall of fertilizing water is
probably intended to be magically ensured by a fall of the victim
Tod {Annals, XI, ch. iv) says * love of offspring *
is the motive, but he
does not explain, and one is almost inclined to suspect him of a subtle pun.
llRose, Tribes and Castes, i, pp. 345 sqq,
Lilcr^ the Maithil Brahman mentioned by Wilson {Indian Caste, n, p.
194) who was outcasted his family because he recovered after his funeral
ceremonies had actually been performed in, expectation of his decease (cf.
supra, pp. it>9=:'io), and like the Bhil referred to above (p. ^^42) who died
within a year of the use of his blood for anointing his lord as king.
Macpherson, op. cit., p. 1X5.
1 See Purchas his Pilgrimage, bk. i, ch. 15.
HINDUISM AND PRIMI TI VE R ELI GIO NS 255

from above.2 The practice of hook-swinging', i.e. suspending


a devotee on hooks inserted through the flesh under the muscles
of the back and rotating him at the end of a horizontal pole
mounted high on a vertical bearing, a practice at one time
widely used in Bengal and in most of southern India, was un-
doubtedly another manifestation of the same fertility cult.^
One phenomenon of primitive religion which cannot be
ignored when writing of India is totemism, traces of which are
shown by primitive tribes in all parts of India and by not a
few castes that have reached or retained a high social position.
From the Bhils in the west to the Wa of eastern Burma; from
the Kanets of the Simla hills to members of Telugu castes of
southern India, clear traces of totemism are found to survive,
and it is needless here to go into details already sufficiently well
known and recorded. It may be enough to recall the long-

tailed Ranas of Saurashtra Jethwa Rajputs who claim descent


from Hanuman, and the ruler of the Malabar coast whose death
involves abstention from fishing lest the soul-inhabited fish be
captured, to show that totemistic ideas are not entirely confined
to primitive tribes and to castes low in the social scale. What
is called totemism is no doubt remarkable for the extreme variety
of its manifestations, ranging from the sacramental consumption
of the totem or its use for magical augmentation of the food
supply, to the merest peg for exogamy to hang upon, and it is in
the latter form that it is commonest in India, where it has gene-
rally decayed into a mere totemistic clan name. There are,
however, traces of taboos and beliefs essential to it at an earlier
stage. Some tribes having what appear to be tofemistic clan
names no longer regard them as such. Thus the Thevoma clan
of the Angami Nagas and the Awomi of the Serna would appear
in both cases to bear names translatable as Pigmen
but no
such meaning is ascribed by the Angamis, who explain the name
as a human patronymic, while the corresponding Serna clan
gives an adventitious explanation of an ancestor who was bitten
by a pig; on the other hand, the Ao Naga neighbours of these
tribes have a dog clan which still claims canine characteristics,
e.g. speed of foot and doglike features;^ moreover, they taboo
the dog as food or rather used to taboo it until recently they
found it desirable to break the taboo in order to benefit by the
medicinal virtues of dog-flesh. Another Ao clan again, the
2 Hutton, Monoliths at Dimapur, p. 69.
3 A full account with references to the older reports will be found in
an article by Prebendary J. H. Powell in Folk-Lore, voL xxv, pp. 147 sqq.
^ In the Nicobars, where descent is claimed from a dog and a woman,
the dress of man is said to be intended to simulate a doggy appearance,
consisting of a fillet round the head with two ends sticking up from the
knot on the forehead to resemble dogs ears, while the private parts are
concealed in a blue bag with a long red point to it, and the waist-band
is arranged to fall down behind in a tail.
,56 APPENDIX S
Wozakumr, Hornbill people, claim descent from a woman
i.e.

who conceived as a result of a feather dropping in her lap from


a hombiU SlWg overhead, and it is taboo for them to kill a
hornbill or eWh to see a dead one. It may likewise be noted
that the Hindu Chasa of Orissa regard the injuring of their clan
totem as punished by leprosy, an affliction which the Brahman
regards as the punishment for killing a cow.
Into totemism the belief in life-matter seems hardly to enter,
perhaps, but the idea of the external soul, which seems to be a
devfflopment of the life-matter belief, does appear as connected
with totemism in India much as it is in parts of Africa. As a
possible source of the connexion between totemism and the
external soul the phenomena of birth may be suggested. The
afterbirth is well Known to be intimately associated with the
idea of the external soul and is regarded in some cases as actually
containing it. Thus among the Baganda it is buried under a
plantain tree, and a woman may conceive if pollinated, as it
were, by the dropping on her of a plantain flower. In the
case of the chief, however, the placenta is carefully preserved
and brought to him to handle and return to safe keeping on
State occasions. Here we are reminded of the placenta standard
on the palette of Narmer, No
doubt as the abode of the
kings vital essence the proximity of such a standard in time of
danger would be useful, and if the standard be the abode of
the external soul we have the explanation both of the association
of totems with standards and the apparent paradox of taking
into danger a very highly prized emblem the capture of which
is regarded by the enemy as of great significance.^ One may
recall the Fairy Banner of the McLeods, which not only brinp
vietpry but causes a cow to drop her calf or a pregnant woman
to give birth on the sight of it. It may be inferred that its
quickening property is due to its being the seat of external soul-
matter. If the placenta be thus regarded as the location of
the external soul, a possible origin of totemic belief at once
Frazer, Totemism and Exogamy, u, p, ^oj.
Seligman and Murray, Note on an Early Egyptian Standard, in
Man, XI, 97 .

r The earliest Roman standards are said to have consisted of a bundle


of hay on a pole. Can it have been held that such a bundle of hay would
be, like a growth of mistletoe, a convenient hiding place for external soul-
matter, just as the Kayan leaves the leafy tops of trees unpruned in order
to aSora a refuge for the spirits of vegetation when clearing the fields?
It might be that the same idea was present in the bandaged pole which
on an Egyptian temple represented the god, the hieroglyphic for which
resembles a flag though it is described as a hatchet and stated to be in
effect a bandaged pole with a loose flap projecting.
* Strasser (ne Mongolian Horde, p. 104) relates that the Tashai Lama
of Urga in Mongolia pressed his standards to the slit throats of his victims,
'

saturating them in their spurting life blood. One is tempted to see in


this an attempt to imbue the standards with soul-matter which would no
doubt contribute towards their victorious progress.
HINDUISM AUD PRIMITIVE RELIGIONS 3*57

suggests itself in the possibility of the placenta being devoured


by some scavenging animal or bird, or being aissociated with the
tree on which it is placed for security or with some plant which
springs up on the spot where it is buried. It is necessary, of
course, to postulate a subsequent transfer of the totem from the
individual to the exogamous clan descended from him or her,
but such an origin would perhaps account for a purely social
form of totemism in which there was no sacramental element
and no magical food production. It would also account for a
soul-transmigration form of fotemism.
It is wortn while in this connexion to draw attention to the
case of the plantain tree as deriving Baganda soul-matter from
a placenta at its root in Africa and as being used as the equiva-
lent of a human being in Assam and Oceania. Thus a
plantain tree is in many parts of the Naga Hills an euphemism
'

for a slave for decapitation. The same equation appears in


Micronesia, Fiji, Polynesia, Madagascar and nearer home among
the Palaungs of Burma, with whom the tree is also an emblem
of fertility, while it is frequently a plantain tree which is used
in India in the mock marriages sometimes performed for elder
children to enable their juniors to be married before them. A
close parallel to the Baganda theory of conception from a plan-
tain ilower is to be found in India in the Muslim belief that a
woman may conceive if the flowers of a rose tree or jasmine
which is growing from the tomb of a dead saint should fall upon
her. One may also call attention to the existence in the case
of the Ho
tribe of a clan whose peculiar totem is the hole of a
mouse or rat, a totem immediately explicable on the placenta
theory when one is informed that the clan is descended from a
person whose placenta at his birth was buried in a rat hole.^
The same placenta theory perhaps also appears in the Kora story
reported by Risley from Bengal to account for the fruit of a
certain tree being taboo to the tribe, since their ancestor once
accidentally ate a human placenta which had been exposed in
a tree of that particular variety. The existence of clans with a
placenta totem has also been reported from the Agariya and
Gond tribes of the Central Provinces.
Another origin of totems has been suggested as likely to be
found in food restrictions. We
should be inclined rather to put
it the other way round and regard it as perhaps to be found in
peculiarities of diet. The discovery and search for forms of
vegetable food must have held a very important place in early

J. G. S. Turner, Cemu5 0/ Nigeria, 1931, vi, p. 7:


Cf. also
It is
said that some of the Yoruba eat the placenta. ... In the Cameroons it is
usually buried under a plantain tree and the jffuit is henceforth taken by
the native doctor and later by the child." -

10 Majumdar, Tribe m Transition, p. o6.


Tribes and Castes, i, p. 507. .

12 V. Elwin, The Agaria,


p. 76 sq.; Hivaie, The Pardhans, p. 40.
17 .
sg8 APPENDIX B

domestic economy. Under any conditions in which food was


scarce and its collection uncertain and laborious, as is probably
frequently if not normally the case in a pre-agricultural stage
of existence, there must have been a tendency to conceal as long
as possible the source of some hitherto unknown supply of food
lest that supply be exhausted by other gatherers. Experiment
in strange vegetables is dangerous, particularly in the tropics.
It is therefore suggested that to the discovery and communication
to the kindred and concealment from other clans of new forms
of vegetable food must be ascribed the importance of certain
wild vegetables in clan ceremonial among the Naga tribes of
Assam. The test of whether a clan in one tribe is to be identi-
fied or not with a clan in another tribe speaking a totally
different language often depends on the vegetable used in certain
ceremonies. If the identical plant is used, the clans would be
regarded as related, and clansmen of one tribe going on a trad-
ing expedition into the territory of another will feel secure in a
house of a related clan, whereas otherwise they would lie down,
if at all, in fear of their slumbers being sublimated by the
sharp dao of decapitation. It is true that the wild vegetables
used in these clan ceremonies are not always regular articles of
food, but it is the writers impression that they are always edible.
Here surely is another possible source of ideas leading to the
adoption of a vegetable totem by a given clan.
It is not argued that all totems are accounted for directly by
any of the ideas suggested above. Pigs tripe, for instance, will
not quite fit, though we might perhaps suppose a fragment to
have got left in a pot in which an after-birth was hung up;
buffalo dung is harder still, unless dung be regarded {vide supra,
252 sq.) as a source of spontaneous life. It is, however,
S
.

ely from every point of view that totemism in general has


received accretions from a number of sources, and that while it
may have originally started with, say, Frazers theory of concep-
tion in ignorance of the fact of paternity, it has been encouraged
and perpetuated by the ideas of life-matter, a separable soul,
transmigration and probably other connected ideas, and that
a number of these have contributed to totemism as still found
in India.
Magic, when limited to purely imitative or sympathetic magic,
is rather nemer the domain of science than of religion. There
is nothing religious at all about the effort of an Ao Naga to
influence the rice by planting a root or two in earth put in the
hollow top of a bamboo, and so raised above the rest of the field
which is thus induced to grow high; in the rather inconsiderate
Kuki plan of putting a bug into the bundle of the departing
guest m order that the rest of the vermin may leave the house
likewise; or in the custom of giving a Prabhu bride a grindstone
to hold which she gives to her husband saying take the baby .
HINDUISM AND PRIMITIVE RELIGIONS 359

When, however, the efficacy o such magic depends not on the


practice but on the practitioner, we may suspect that the idea
of soul-matter is present and that it is, often at any rate, the
superfluity of this material that enables the magician to make
his magic successful. Here again one may perhaps see the
reaction of the simple belief that like produces like to an inde-
pendent belief in the existence of soul-matter. However that
may be, a belief in magic both white and black pervades all
the more ignorant classes in India and is frequently responsible
for serious crime; nor is it always eliminated by culture and
education, as witness the comparatively educated persons fre-
quently victimized by rogues who profess to be able to double
currency notes miraculously. Thus, to give a single instance,
in May 1931, a well-to-do merchant of Indore imprudently
handed over Rs. 2,800 to one Pandit Sri Krishna who claimed
to have a marvellous process of doubling notes. Ignorant
villagers are much more easily imposed on, as in the case of a
village near Multan in the Punjab which about the same time
parted with Rs. 15,000 in cash and ornaments to a Muslim
fakir who first called down a few rupees from heaven to inspire
confidence in his piety and miraculous powers, and then pro-
fessed to be able to turn silver ornaments into gold or one rupee
into three. A belief in magic again, for it can hardly be describ-
ed as anything else, even if it involve the theory of the impreg-
nation of matter with soul-essence, appears in the practice,
reported on good authority in Rangoon, of a director of an
international trading corporation who, when ill, had sewn into
the seat of his pyjama trousers by his Catholic wife a pious
fragment of the holy St Theresa's petticoat. What the effect on
a male Naga would be of wearing a piece of any woman's petti-
coat, however holy, we hesitate to set down in print, but pre-
sumably Herr Direktor experienced benefit.
Often a belief in witchcraft leads to the murder of the reputed
witch.^ In 1928 in Bihar, for instance, nine cases of murder
were ascribed to witchcraft, and in 1931 in the Yarpur mahalla
of Patna a small Lohar girl was murdered in retaliation for the
supposed enchantments of her mother, while an aboriginal *

woman suspected of being a witch was killed near Ghatsila. In


Faizabad, in 1927, a man was killed on the advice of a medicine
man as being the cause of another's prolonged dysentery, and
in Budaun a Chamar, who was suspected of having bewitched
an idiot of good family, was pegged out and periodically
belaboured while the bewitched one was watched for improve-
ment in his condition. Ultimately, as there was none,
the Chamar succumbed. In July 1920 a mob in the Nizam's
13 See also El win, Maria Murder and Suicide, ch. v, and C. R. Hemeon,
*
Short Notes of some Remarkable Crimes in the Central Provinces and
Berar in Man in India, September 1943.

560 APPENDIX M
Dominions killed a woman who was believed to have brought
cholera on the village, a belief arising from the hysterical state-
ment of a possessed woman into whom had entered the spirit
of the goddess who was being worshipped at the time by the
village. In November 1930 in Gonda m
the United Provinces
a wizard was murdered by his own pupil in the belief that he,
the wizard, had caused an evil spirit to destroy his pupils wife
and would cause it also to destroy him himself; and as the pupil
was one of ten years' standing, this instance testifies to the
wizard's belief in his own system.
On the other hand the witches themselves likewise commit
murders for their own ends, and to that extent anyhow justify
their persecution. A
boy was sacrificed in Bhagalpur in 1928,
for the purpose of exorcising evil spirits from a possessed woman,
and on the eve of the Dasehra festival of 1930 two sorcerers of
Sambalpur, described by the High Court as men of standing
*

sacrificed a boy for some nefarious purpose of their own. This


case, however, may possibly have been one rather of a homicide
of the kind alluded to above as occasioned by the belief in the
need for soul-matter. The use of human sacrifice in order to
exorcise spirits is probably unusual, as it is commoner to treat
the body of the possessed by more direct methods. Indeed, a
Hindu girl was beaten to death in Lahore in November 1929,
in the attempt to cure her of possession, and this apparently
at a shrine frequented by persons in order to experience posses-
sion by the deity. In view, however, of the extent of illiteracy
and of the population concerned, the amount of violent crime
actti^lly due to a belief in witchcraft appears to be unexpectedly
small, though naturally apt to increase with the appearance of
cahimiii^s or epidemics, which are ascribable to the malevolence
of witches. In Chota Nagpur there are professional witch
doctors called spkha whose business it is to indicate the witch
responsible for apy calamity or epidemic that has occurred.
Personal magic, however, is not the only form in which magic
appears. Tribal magic, in which the community combines,
usually at some fest^ival, in rites or dances intended to secure
fertility or prosperity, is a normal feature of tribal religion.
Such festivals or rites are usually associated with tlie agricultural
year and may involve sexual licence which is probably intended
to have a magical effect on the fertility of the crop and of the
community itself, and no better instance of such a festival can
be quoted than the Holi, which has survived as a Hindu festival
throughout India.^^ It is tempting, if possibly fanciful, to
Unless be the marriage festival annually celebrated in June by the
it
lord of a feudal manor in Normandy, who after participating in the bridal
festivities of his serfs picked out the couples who appeared to him to be
the most amorous, and caused them to consummate their marriages in
the boughs of trees or in the waters of the local river. Dulaure speaks of
this practice as one of tyrannical jesting, but the interpretation here assumed
HINDUISM AND PRIMITIVE RELIGIONS 6i

trace the qrigin of the widespread belief in the magic effect of


coition on the fertility of the soil, of animals and of people in
general to a period in, human history when the relation between
cause and effect in the begetting of children was not yet fully
comprehended, but when the two were already seen to have
some association, so that what was really the cause of the con-
ception of otie particular child was regarded as merely the cause
of parturition in general.^^ Such a stage in the process of
deduction from observed facts, if its existence be credible, would
account for the common practice of assisting the fertility of the
crops by the act of sexual coition. A reference has already
been made to the probable connexion between fertility cults
and the practice of sacred prostitution. An explanation may
here be offered of the peculiar part played by strangers in this
cult. In the sacred prostitution of both Babylon and Byblos
it is clear that the dedicated woman gave herself to strangers;
similarly, it was commerce with strangers which was so necessary
to the fertility of the fields of Kamul and so contributive to
the prosperity of the people of Caindu. Yule mentions the
custom as reported of the Hazaras and of other peoples, includ-
ing even the Nayars. It seems not unlikely that the underlying
motive is the acquisition from the stranger of additional life-
matter not already inherent in the soil or its inhabitants. The
soul-matter of any given place may be regarded as limited in
extent and the transfer from one individual to another merely
redistributes but does not increase, whereas the reception of
soul-material from a stranger is additional to that already in
circulation and will naturally therefore increase fertility. This
hypothesis likewise offers a possible explanation of the custom
mentioned by Gait^s of the Todas, who are reported to call in
a person from another village to deflower a girl about to attain
puberty, who otherwise finds it difficult to marry, and perhaps
also of the talikettu ceremony in south India generally. Appro-
val of commerce with strangers, which is perhaps associated
always with the fertility of the soil, \s reported of the Uighiirs,
Hazaras, Chukchis, Koryaks and (by a tenth-century Arab
traveller) of some Turks. Dancing likewise has^ probably a
magical origin, and it certainly has a magical aspect as when it
is perhaps more likely. Fertility rites survived in Europe in spite of the
Church: *
fernmes prostitutes . .
les qui suivaient la Gout
. . ttaient
. .

*
tenues, tant que le mots de mat durait, de faire le lit du rot des ribands
(Dulaure, op. cit, pp. Si% sq a8o); and sometimes even undei its auspices,
as at Isernia.
have heard a Sema chief of great tribal authority and experience,
I
Inato of Lumitsami, affirm thatit was ridiculous to suppose that pregnancy

would result from coition on one occasion only, which indicates that even
now the relation between cause and effect in this particular is not com-
pletely grasped in that tribe. Vide supra^ pp. 163 sqq.
u' Marco Polo, i, p. ai* n.; 11, p. 56 n.
18 Census of India, 1911, i, p. aoo. ^^ Vlde Ynh, Marco Polo, loc. cit
j6,
appendix B
involves leaping up in the air to encourage the growth of paddy,
and Russell has acutely suggested that acrobatic displays have
originated in die same idea. Similarly, animal dancing, such
as that for instance of the Gonds and Bhatras of the Central
Provinces, probably originates in an attempt to increase, or
perhaps merely to assemble, by magic the wild animals on which
the community partly depends for its food supply. Even the
20 may have
highly mimetic dancing of the Juang of Orissa
originated in that way. When, however, the spring hunting is
considered, it is apparent that the soul-matter cult is again
prominent. The Aheria of the Rajput in western India, the
hunting festival of the Halvakki Vakkals in Kanara, the Jur
Sital of Bihar, the spring hunt of the Chota Nagpur tribes, and
of the Bhatras, Gonds and Gadabas of the Central Provinces
and Orissa, the Sekrengi hunt of the Angami Nagas and the
corresponding festivals of other Assam tribes are all designed to
secure prosperity through the coming year, and inasmuch as
all manner of living things are destroyed they are probably
intended (the Sekrengi certainly is) to collect a supply of life-
essence and are to be regarded in much the same light as the
spring man-hunt of the Wa of Burma.
It would be impossible here to go into all the aspects of the
tribal religion in India, but enough has perhaps been said to
show that the beliefs held are not mere vague imaginings of
superstitious and untaught minds, 'amorphous as they were
described in the Census report of 1911, but the debris of a real
religious system, a definite philosophy, to the one-time wide-
spread prevalence of which the manifold survivals in Hinduism
testify, linking together geographically the Austro-Asiatic and
Australoid cultures of the forest-clad hills where the isolated
remains of the original religion still hold out in an unassimilated
form. It is probably this philosophy of life-matter which
accounts for the fact that in so many parts of the world, e.g. in
India and southern and eastern Europe, Greece and Italy in
larticular, the real religion of the people is hagiolatry. It is
{ess the orthodox gods of the religion who are worshipped than
rfirines and holy places, generally tombs particularly associated
with some deceased saint or hero likely to have been rich in
soul-stuff, the benefit of which may be obtained at the grave,
originally no doubt in the form of material emanation. Be
tihat as it may, showing traces in Europe on the one hand and
stretching down into Australasia on the other, this creed must
have been in its time a great religion, not so great perhaps in
altruism, but great in extent and in constituting a very definite
rung in that poor ladder up which our race still tries to climb
in its effort to ascertain the unknowable, to scale the ramparts
of infinity.
PART FIVE

BIBLIOGRAPHY
Reference willbe found in the text or notes to the following
periodicals and works. Both the periodicals and books are
sometimes referred to by the abbreviated letters or titles indi-
cated, and in the case of the books the date of the first publica-
tion is given in this list wherever possible, while the edition
used is also indicated where that is other than the original.
Where the book has not been published in London the place of
publication is given.

PERIODICALS
Antiquity Antiquity, a Quarterly Review of Archaeology, Gloucester.
A. R. The Asiatic Revietv,
B. S, 0 ,A,S, Bulletin of the School of Oriental and African Studies,
University of London.
The Ceylon Journal of Science, Colombo.
Folk-Lore Folk-Lore, a Quarterly Review of Myth, Tradition, Insti-
tution and Custom, being the Transactions of the Folk-
Lore Society.
I, A. The Indian Antiquary, Bombay.
J, A,O.S.
'

Journal of the American Oriental Society. Newhaven, Conn.


J,A,S.B, (and J.R.A.S.B,) The Journal of the {Royal) Asiatic Society
of Bengal. Calcutta.
M,A.S.B. Memoirs of the above Society, Calt:utta.

J.R.A.L The Journal of the Royal Anthropological Institute,


J.R,A.S. The Journal of the Royal Asiatic Society,
The Journal of the Mythic Society, Bangalore.
Man Man, a Monthly Record of Anthropological Science,
published by the K.A.I.
Man in India Man in India, a Quarterly Record of Anthropological Science
with Special Reference to India, Ranchi.

AUTHORS
Abbott, J. The Keys of Power, a Study of Indian Ritual and Belief, 19311.

Achan, P. a, a Hebrew Inscription from Chennamangalam


* in LA,
for July 1930. Bombay, 1930.
Aiyanoar, M. S, Tamil Studies, Madras, 1914.
Aiyafpan, Anthropology of
a. the Nayadis, (Madras Govt. Museum
Bulletins, N.S., voL n, no. 4,) Madras, 1937.
Ansaki* G. Muslim Caste in India* in Wiener FdlkerhundUche
*

Mi^dlwngen, Jahrgang, Nr. pp. r65-70, Vienna, 1954.


- Muslim Caste in Uttar Pradesh (a -study of mlmte contact)., Lucknow,
564 BIBLIOGRAPBT
Bailey, F. G. Caste and the Economic Frontier. Manchester. 1957,
Baines, A. Ethnography (Castes and Tribes). With a list of the more
*

important works on Indian ethnography by Dr W. Siegling % in


Grundriss der Indo^Arischen Philologie und AUertumskunde, ii Band,
5 Heft. Strassburg, igu.
Banerjee, a. R. See Sarkar, S. S,

Banerji-Sastri, a. Mother-Goddess Cult in Magadha


* *
in The Searchlight
(Anniversary Number, 1929). Patna, 1930.
Barbosa, Duarte. A Description of the Coasts of East Africa and Malabar
in the Beginning of the Sixteenth Century. (Hakluyt Society.) 1866.
Bartlett, F. C. Anthropology in Reconstruction 1943. (Huxley Memo-
rial Lecture.)
Barton, W. India's North-West Frontier, 1939.
Bernier, F. Voyages de Frangois Bernier, Docteur en MMedne de la
FacuUd de Montpellier, Contenant la Description des Etats du Grand
Mogol, etc. Amsterdam, 1709.
Bhattacharjee, Papia. See Sarkar, S. S.
Bhattacharya, J. N. Hindu Castes and Sects. An Exposition of the Origin
of the Hindu Caste System and the Searing of me Sects towards each
other and towards other Religious Systems. Calcutta, 1896.
Biddulph, J. Tribes of the Hindoo Koosh, Calcutta, 1880.
Blunt, E. A. H. Report on the Census of the United Prounnces of Agra
and Oudh, 1911 (vol. xv, pt. i, of the Census of India, 19x1). Allahabad,
191a (cited as Census of the U.P.).
The Caste System of Northern India, 1931 (cited as Caste System).
Bonington, C. J. Ossuary Practices In the Nicobars, in Man, xxxn,
133. May 193a.
Bonnerjea, B. Possible Origin of the Caste System in India

in LA.,
March, April, May 1931. Bombay, 1931.
Bor, N. L. '
The Daflas and their Oaths % in J.A.S.B,, ii, i, xgsO,
Bougle, C. Bssais sur le Rigime des Castes {Travaux de FAnnde Socio-
logique). Paris, 1908.
Bradley-Birt, F. B. The -Story of an Indian Upland. 1905.
Bray, D. Report on the Census of Baluchistan^ xgix (voL xv o! The Census
of India, xgii). Calcutta, 1913.
Briffault, R. The Mothers, A Study of the Origins of Sentiments and
Institutions. 19^7.
Brough, J. The Early Brahminical System of Gotra and Fravari. A
Translation of the
Gotra-pravara-manjarl of JPurusommapandita.

Cambridge, 1953.
Buchanan, F. Journey through Mysore, Canara and Malabar, xSoy. (The
edition used is that of 1870 published in Madras? a$ the original edition
is in three volumes and that of 1870 in two only, diapter references are
given as well as pages.)
An Account of Assam with some notices concerning The Neighbouring
Territories. By Francis Hamilton, M.D., F.R.S. Edited by Dr S. R.
Bhuyan. Gauhati, 1940.
BilHLER, G. The
Sacred Laws of the Aryas as taught in the Schools of
Apastamba, Gautama, VSsishtha and BaudhEyana (vol, ii in The Sacred
Books of the East, edited by Max Mhllcr)* Oxford. 1879.* (Cited as
n
Sacred Laws.)
BIBLIOGRAPHY *65

BiiHLER, G. The Laws of Manu, translated with extracts from seven com-
mentaries. (Ibid. vol. XXV.) Oxford, x886.
BURNELJL, A. C. See Yule.
Burton, R. F.Sindh, and the Races that inhabit the Valley of the Indus;
with Notices of the Topography and History of the Province. 1851.

Campbell, J. M. Gujarat Population: Hindus (Gazetteer 0/ the Bombay


Presidency, vol. ix). Bombay, 1901.
Cafpieri, M. Le Caste degli intoccabili
'
in India in Rivista di
Antropologia, vol. xxxv. Rome, 1947.
Carre, TAbbe. The Travels of the Abbi Carrd in India and the Near East,
i6j2-i6y4. Translated by Lady Fawcett and edited by Sir Charles
Fawcett with the assistance of Sir Richard Burn. Hakluyt Society, und
Series, nos. xcv-xcvii. 1947-8.
Chakravartti, M. R. See Sarkar, S. S.

Chanda, R. The Indo-Aryan Races/ A Study of the Origin of Indo^Aryan


People and Institutions. Rajshahi, 1916.
Non-Vedic Elements in Brahmanism. Unpublished (?) MS. in author'S
possession.
Cole, B. L. Rajputana Classes compiled under the orders of the Govern-
ment of India (Handbooks for the Indian Army, 1921). Simla, ig2.
Report on the Census of the Rajputana Agency, 1931 (vol. xxvn of the
Census of India, 1931). Meerut, 193.
CoLEBROOKE, H, T. ^numeration of Indian Classes, in vol. v of Asiatick
Researches, 1788. (The edition used is the 4th edition, London, 1807.)
CoRNFORD, F. M. Plato's Cosmology. The Timaeus of Plato translated with
a running commentary. 1937.
Crooke, W. Religion and Folklore of Northern India. Oxford, 1936.
(Crooke's An
Introduction to the Papular Religion and Folk-lore of
Northern India was published in 1894 and was followed by a second
edition called Popular Religion and Folklore, etc., the edition used here
being the third.)
Tribes and Castes of the North-West Provinces and Oudh. Calcutta,
.1896 (cited as Tribes and Castes).
Things Indian. Being Discursive Notes on Various Subjects connected
with India, 1906.
Natives of Northern India, in * The Native Races of the British

Empire *
scries.

(Editor).
(cited as
1907 (cited as Northern India),
Tod*8 Annals and Antiquities of Rajasthan.
Tod's AnnaU, etc.).
Oxford, loao

Dahlmann, j. Das Altindische Volkstum und seine Bedeutung filr die


Gesellschaftskunde. Kdln, 1899.
Dalton, E. T. Descriptive Ethnology of Bengal. Calcutta, 187a.
Dames, M. L. The Maloch Race (Asiatic Society Monographs, vol. rv).
1904-
Dawson, W. R. The Bridle of Pegasus. Studies in Magic, Mythology and
Folklore. 1930.
Day, F. The Land of the Permauls, or Cochin, its Past and its Present.
Madras, 1863.
Desai, G. H. a Glossary of Castes, Tribes and Races in the Baroda State.
Bombay,
266 BISLIOGRdPHT
Diix, S. Roman Society in the Last Century of the Western Empire, 1898.
(The edition used is that of 1899.)

Diodorus Siculus. (The edition used is The Historical Library' of Diodorus


the Sicilian, translated by G. Booth, Esq., 1814.)
Dowson, J. d Classical Dictionary of Hindu M'ythology and Rdmon,
Geography, History and Literature, (Trlibners Oriental Series, vof. vi,
1B79.) (The edition used is the 4th edition of 1903.)
Drury, R. See Oliver.
Dube, S, C. The Kamar, Lucknow, 1951.
Dubois, J. A, A Description of the People
of India, 1817. (The transla-
tion used the Madras reprint of 1879, but chapter as well as paginal
is
references are given.)

Dulaure, j. a. Des Divinitds Odniratfices ou du CuUe du Phallus chez les


Anciens et les Modernes, Paris, 1825. (The edition used is the reprint
of 1885.)

Dumont, L. Hierarchy and Marriage Alliance in South Indian Kinship, in


Occasional Papers of the Royal Anthropological Institute, ,1957.
Dunbar, G. D.-S,- *
Abors and Galongs % in M,A,SS,, vol. v, extra no.
Calcutta, 1916.
Duncan, J. Historical Remarks on the Coast of Malabar, with some
'

Description of the Manners of its Inhabitants in vol. v of Asiaikk


Researches, Calcutta, j^88 and after. (The edition used is the 4th
edition printed in 8vo in Ixmdon, 1807.)
Dutt, N. K, Origin and Growth of Caste in India, i. Calcutta, 1931.

Edgeworth, M. P,
Abstract of a Journal kept by Gardiner during Mr

his travels in Central Asia ^with a Note and Introduction % in J,A,S,B,,
no. 3 of 1853.
Ehrenfels, 0 , R, Mother^right in India, Hyderabad, Deccan, 1941.
Ellis, R. H. A Short Account of the Laccadwe Islands mi Minkoy,
Madras, 1924 (cited as Laccadive Islands),
Exxis, W. History of Madagascar, 1838.
Elwin, V. The Baiga, 1939.
~The Agaria, Bombay, 1942.
Maria Murder and Suicide*
Bombay, 1943.
The Muria and their GhoiuL Bombay, 1947.
Enthoven, R. E. The Tribes and Castes of Bomhay, Bombay, i9io
(cited as Tribes and Castes),

Review of Dutt*s Origin and Growth of Caste in India, in f,M,A,S,,


Jan. 1932.
Evans, F, B, See innes,
Evans, I. H. N. '
Kempunan *, in Men, xx, 38 , 1910.

FitzGerald, B, V. Review of Caste in India in B,S, 0 ,d,S vol. xii, pt. 1,


P 1947^
Fleure, H* j. See Peake.
Forbes, A, R. Rds Mala! or Hindoo Annah of the Protdnm of Goozerat,
sn Western India, 1856.
lOrais, J.. Oriental Memoirs. j8a. (fbe edition uwd to ihM of 1834.)
BJBLIOGRAPBT 67

Forsyth, J. The Highlands


of Central India, Notes on their Forests and
Wild Tribes, Natural History, and Sports, 1876. (The edition used
is that of 1889.)

Frazer, J. Totemism and Exogamy, A Treatise on Certain Early Forms


G.
and Society, 1910.
of Superstition
The Golden Bough (3rd edition). 1911-15.
Fuchs, S. The Gond and Bhumia of Eastern Mandla, Bombay, i960.
VON FUrer-Haimendorf, Ch. The Raj Gonds of Adilabad. 1948.
Furnivaix, j. S, Netherlands India, A Study of Plural Economy,
Cambridge, 1939.
FusTEL DE CouLANGES, N. D. La Citd antique, Paris, 1864. (The edition
used is the igsg edition in the Librairie Hachette,)

Gait, E. A. Report on the Census of Bernal, Bihar and Orissa, and Sikkim
(vol. VI of Census of India, 1901). Calcutta, 1903,
Report on the Census of India, 1911, vol. i, pt. i. Calcutta, 1913.
See also Risky,
Gardiner, See Edgeworth.
Genesis, Book of. Authorized Version.
Ghosh, A. The
Original Inhabitants of the United Provinces: a Study in
Ethnology, in vol. xi of Allahabad University Studies. 1935.
Ghurye, G, S. Caste and Race in India, i93,
Gilbert, W. H., Jr, Peoples of India (Smithsonian Institution War Back-
ground Studies, no. i8). Washington, 1944.
Giuffrida-Rucgeri, V. First Outlines of a Systematic Anthropology of Asia.
Calcutta, i9i.
Gordon, D. H, MomiySi *, in Man, xxix,

160 (Dec. igag), xxxni, 163
(Sept. 1933), XXXIV, 83 (April 1934).
Green, M. M. Ibo Village Affairs, 1947.
Guha, B. S. *
An Outline of the Racial Ethnology of India in An Outline
of the Field Sciences of India, Calcutta, 1937.
Gordon, P, R, T. The Khasis. 1907. (The edition used is that of 1914.)

Hamilton, A. New Account of the East Indies, Edinburgh,


Hamilton, F.aw uchanan, q.Vv
Hamilton, W. The East India Gazetteer, 1815.
Handy, E. S, C, Polynesian Religion (Bernice P. Bishop Museum Bulletin
no. 34). Honolulu, i97,
HARDm.AN, J. P. See Scott,
Hatch, W, J. The Land Pirates of India, i98,
Haughton, G. C. Dictionary, Bengali and Sanskrit, 1833.
Himeon, C, R, Short Notes of some Remarkable Crimes in the Central
*

Provinces and Berar % in Man in India, xxiii, no. 3. Ranchii 1943.


Herodotus. Historiai, (The edition used is that edited by Blakesleyi 1854,)
Hill, S, C, *
Origin of the Caste System in India *, in I,A March-October
1930. Bombay, 1930,
HiRA Lal. See Russell.
'


.
_

Hivale,S. The^Pariham $f the Upper Narbada Valley, Bombay, 194$.


HoBfiNw-joisoH/see yule.

268 BIBLIOGRAPHY
Hocart, a. M. India and the Pacific % in the Ceylon Journal of Science^
*

Section G. Archaeology, Ethnology, etc., voL i, pt. t, Colombo, 1925!


The Temple of the Tooth in Kandy, 1930.
*
The Basis of Caste in India *, in Acta Orientaliat xiv. Leyden, 1936.
Kings and Councillors, Cairo, 1936.
'
The Estates of the Realm in Thakaundrove, Fiji % in B,S, 0 ,A.S,,
vol. IX, pt. 2, 1937 - 9 *

Les Castes (Annales du Mus^e Gnimet). Park, 1938.^


Hodson, T. C.
The Genna among the Tribes of Assam *, in J,R.A.L,
vol, XXXVI, 1906.
The Meitheis, 1908.
The Naga Tribes of Manipur, 1911.

Meithei Literature *, in Folk-Lore, vol. xxiii, 1912.
Holdich, T. H. India, (The Regions of the World, edited by H, J.
Mackinder.) 1904.
Hoixand, T. H. *
The Coorgs and Yeruvas, an ethnological contrast % in
J.A,S,B,, vol. Lxx, pt. iii, no. 2, 1901.
Hopkins, E. W, The Social and Military Position of the Ruling Caste in
Ancient India as represented by the Sanskrit Epic; with an Appendix
on the Status of Woman, (From J,A, 0 ,$,, voL xm). New Haven,
Connecticut, 1889.
Hunter, W. W. Annals of Rural Bengal, 1868, (The edition used is the

7th edition published in 1897.)


Orissa, or The Vicissitudes of
British Rule, 1872.
an Indian Province under Native and

Huntingfori), G. W. B. Letter on
Momiyai *
in Man, xxxiv, 22 (January),
1934 -

HunoN, J. A Popular Accountof the Thugs and Dacoits, the Hereditary


Garotters and Gang-Robbers of India, 1857.
Hutton, J. H. The Angami Nagas, 1921.
The Serna Nagas, 1921.
Carved Monoliths at Bimapur and an Angami Naga Ceremony

*,

in J,R,A,L, vol. ui, 1922.

Meaning and Method of the Erection of Monoliths by the Angami


Nagas.' Ibid,
*
Carved Monoliths at Jamuguri in Assam In J,R,A,L, vol. nix, 1923.


The Use of Stone in the Naga Hills *, in J,R,A,L, voL tvi, 1926.
The Disposal of the Dead at Wakching in Man, xxvii, 44, 1927.
The Significance of Head-hunting in Assam', J,R,AJ voL Lviii,

- 1928.
Diaries of Two Tours in the Unadministered Area East of the
* Naga
Hills ', in M,A,S,B., vol. xi, no. 1. Calcutta, 1929. *
Report on the Census of India, 1931 (vol. i of the Census, of India,
1931), Delhi, 1933.
- A Primitive Philosophy of Life (Frazer Lecture). Oxford, 1938.
See also Mills.

1 The original English version has since been publi^ei


(1930) as Caste;
a Comparative Study. With a preface by Lord Eaglan.
BIBLIOGRAPHY 69
Ibbetson, D. Panjab Castes, Being a reprint of the Chapter on * The Races,
Castes and ribes of the People in the Report on the Census of the
*

Punjab published in 1883 by the late Sir Denzil Ibbetson, K,C,S.L


Lahore, 1916.
INNES, C. A. Gazetteer of the Malabar and Anjengo Districts. Edited by
F. B. Evans. Madras, 1908.
Iyer, L. A. K. The Travancore Tribes and Castes. Trivandrum, 1937-41.
Iyer, L. K. A. Lectures on Ethnography. Calcutta, 1925.
The Cochin Tribes and Castes. Madras, 1909-1^.
Anthropology of the Syrian Christians. Ernakulam,
Iyer, L. K. A. and Nanjundayya, H. V. The Mysore Tribes and Castes
Bangalore, 1958-35.

Jackson, A. M. T. *
Note on the History of the Caste System ', in
J.A.S.B., vol. m, no. 7, for July. Calcutta, 1907.
JEVONS, F. B. (Editor). Plutarch's Romane Questions. Translated a.d.
1603 by Philemon Holland (Biblioth^ue de Carabas, vii). 1895.
Jolly, J. The Institutes of Vishnu (vol. vii in The Sacred Books of the
East, edited by Max Muller). Oxford, 1880.

Kapadia, K. M. Hindu Kinship. Bombay, 1947.


Karandikar, S. V. Hindu Exogamy. Bombay, 1959.
Karve, I. Kinship Organization in India. Poona, 1953.
Changing India; Aspects of Caste Society. Bombay, 1961.
Hindu Society: An Interpretation. Poona, 1961,
Kaul, H. Report on the Census of the Punjab, 1911 (vol. xiv of the
Census of India, 1911). Lahore, 1915.
Keith, A. B. Chap, iv, * The Age of the Rigveda in The Cambridge
History of India, vol. i '
Ancient India *, edited by E. J. Rapson.
Cambridge, 1955.
Ketkar, S. V. History of Caste in India. Ithaca, N.Y., 1909.
Essay on Hinduism, Its Formation and Future. 1911.
Khan, A. I. On the Trial by Ordeal among the Hindus, communicated
by Warren Hasting to Asiatick Researches, vol. i. Calcutta, 1788.
(The edition used is the 5th, London, 1806.)
Kirkpatrick, W. '
Primitive Exogamy and the Caste System in Proceed-
ings, A.S.B., vol. viii, no. 3. Calcutta, 1915.
Kitts, E. J. A Compendium of the castes and tribes found in India. . . .

Bombay, 1885.

Lacroix, P. Mmirs, Usages et Costumes au Moyen Age et it VEpoque de la


Renaissance. Paris, 1871,
Langdon, S.
A Sumerian hymn to Ishtar (Innini) and the deified Ishme-
Bagan % in J.R.A.S., 1931.
Leach, E. R. (Editor). Aspects of Caste in South India, Ceylon and North-
West Pakistan. Cambridge, i960.
Leitner, G. W. Dardistan in 1866, 1886 and 1893 being An Account of
the History, Religions, Customs, Legends, Fables, and Songs of Gilgit,
Chil^, Kandid (Gabridl), Yasin, Chitrdl, Hunza, Nagyr and other
Parts of the Hindukush, as also a supplement to the second edition
of The Hunza and Nagyr Handbook (printed by the Government of
370 BIBLiOGRdBMT
India in and an epitome of Part III o! the Author^s The
1889),
Languages and Races of Dardistan, Woking, N,B. (c. 1895)*
Lewin, T, H. Wild Races of South-eastern India. 1870.
Linton, R. The Tanala, A Bill Tribe of Madagascar (Field Museum
Publication 317). Chicago, 1933.

Little, K. L. *
The Psychological Background of White-Coloured Contacts
in Britain % in The SocioTogicat Review, ?o!. xxxv. 1944.
Logan, W. Malabar. Madras, 1887.
Love, H. D. Vestiges of Old Madras, 1640-1800, in Indian Records Series,
19 ^ 3 -
Lowie, R. H. The Origin of the State. New York, igty.
Luard, C. E. Ethnographical Survey of the Central India Agency.
Lucknow, 1909.
Lucian. The Works of Lucian from the Greek, by Thomas Francklin. 1781.

McCrindle, J. W, Ancient India as described by Megasthenes and Arrian.


1877. (The edition used is the Calcutta reprint of 1936.)

Macdonell, a. a. India's Past. Oxford, 1937.


MacGregor, G. L. See Mead.
Macpherson, W. (Editor). Memorials of Service in India. From the
correspondence of the late Major Samuel Charters Macpherson^ C.B.,
Political Agent at Gwalior during the Mutiny, and formerly employed
in the suppression of hunian sacrifices in Orissa. 1865.
Maine, H. S. Ancient Law; its Connection with the Early History of
Society and its Relation to Modern Ideas. 1861. (The edition used
is the eighth, 1880.)
Majumdar, D. N. *
Pseudo-P,ajputs *, in Man in India, vol. vi, no. 3.
Ranchi, 1936.

A Tribe in Transition, A Study in Culture Pattern. Calcutta, 1937.



Some Aspects of the Cultural Life of the Khasas of the cis- Himalayan
Region m J.R.A.S.B., Letters, vol. vi, no. i, Calcutta, 1940.
Malcolm, J. A Memoir of Central India, including Malwa, and Adjoining
Provinces. With the History, and Copious Illustrations, of the Past
and Present Conditions of that Country. 1833.

Marett, R. R. and Penniman, T. K. (Editors). Spencer's Last Journey,


being the Journal of an Expedition to Tierra del Fmgo
by Sir Baldwin
Spencer. Oxford, 1931.
Marriott, McK, (Editor). Village India. Chicago, 1933,
Caste Ranking and Community Structure in Five Regions of India and
Pakistan (Deccan College Monograph Series, 33.) Poona, i960,
Marshall, H. I. The Karen People of Burma: A Study in Anthropology
and Ethnology. (Ohio State University Bulletin, xxvi.) Columbus,
Ohio, U.S.A., 1933.
Marshall, J. Mohenjo-daro and the Indus CmUzaUon. 1931.
Martin, M. The
History, Antiquities, Topography and Statistics vf Eastern
India. (Three volumes.) 1838.
Mateer, S. The Land of Charity*: a descriptive account of Travancore

and its People with especial reference to missionary labour, 1871.


Native Life in Travancore. 1883.
Mayer, A. C. Caste and' Kinship in Central India, i960.
BIBLIQGRAPHT
Mead, P. J.and MacGregor, G. L. Report on the Census of Bombay,
1911. (VoL VII Vi the Census of India, 1911.) Bombay, 1913.
Meek, G. K. A Sudanese Kingdom, 1931,
Mela, P. Pomponii Melae de Chorographia libri tres. (The edition
is that edited by G. Parthey and published in Berlin, 1867.)

Menon, C. a. Report on the Census of Cochin State, 1911 (vol. xvxii of


the Census of India, 1911). Ernakulam, 1913.
Mills, J. P. The Lhota Nagas,
The Ao Nagas, igad.
Mills, J. P. and Hutton, J. H. "
Ancient Monoliths of North Cachar
in vol. XXV, no. i (ig^g). Calcutta, 1930.
Minchin, C. F. Las B^la (vol. viii, Baluchistan District Gazetteers). Allaha-
bad, 1907.
Mitra, R. IndOAryans: contributions towards the elucidation of their
Ancient and Mediaeval History, Calcutta, 1881.
Molony, j. C. Report on the Census of Madras, 1911 (vol. xii of the
Census of India, Madras, igia.
A Book of South India. 1926.
Morris, G. J. See Northey.
Murphy, J. Lamps of Anthropology. Manchester, 1943.
Murray, M. A. See Seligman.

Nanjundayya, H. V. See Iyer, L. K. A.


Nazaroff, P. S. '
The White Lady % in Blackwood' s Magazine, August
1931-
Nesfield, j, C. Brief View of the Caste System of the North-West Provinces
and Oudh. Allahabad, 1885.
Newell, W. H. The Brahman and Caste Isogamy in North India in

J.R.A.L vol. Lxxxv, pp. 101-10, 1955.


Northey, W. B. and Morris, C. J. The Gurkhas, their Manners, Customs
and Country. 1928.

Oldenberg, H. Zur Geschichte des indischen Kastenwesens


*
in Zeitschrift
der Deutschen Morgenldndischen Gesellschaft, vol. li, pt. 2. Leipzig,
1897.
Oldham. C. E. A. W. Review of Caste and Race in India by G, S. Ghurye,
in Man, xxxii, 316 . 1932.
Oliver, F. (Editor). Madagascar; or Robert Drury's Journal, during
Fifteen Years' Captivity on that Island, and a Further Description of
Madagascar by the Abbi Alexis Rochon, 1890.
O'Malley, L. S. S. Indian Caste Customs. 1933.
Popular Hinduism. 1935.
(Editor). Modern India and the West. 1941.
Oppert, G. On the Original Inhabitants of Bharatavarsa or India. West-
minster, 1893.

Pant, S. D. The Social Economy of the Himalayans Based on a survey


in the Kumaon Himalayas. 1935.
Pargiier, F. E. Ancient Indian Historical Tradition. 1922.
Parry, N. E. The Lakhers, 1932.
BIBLIOGR iPBY
Pate, H. R. Tinnevelly, (Madras District Gazetteers.) Madras, 1917.
Peake, H. and Fueure, H. J. Merchant Venturers in Bronze. 1931.
Peate, I. C. The Welsh House. Liverpool, 1944.
Penniman, T. K. See Marett.
PiOLET, J.-B. Madagascar et les Hova. Paris, 1895.
Pirr-RivERS, G. H. Lane-Fox. The Clash of Culture and the Contact
of Races. 1927.
Playfair, A. The Caros. 1909.
Pliny. C. Plinii Secundi Haturalis Historia. (The edition used is that
published by Hackius in Leyden and Rotterdam in 1669.)
Plutarch. The Roman
Questions. (The edition used is the translation
by H. Oxford, 1924.)
J. Rose.
Polo, M. The Book of Ser Marco Polo the Venetian concerning the
Kingdoms and Marvels of the East, translated and edited by C<3 Sir .

Henry Yule, 1871. (The edition used is the third edition of 1921,
revised by Henri Cordier.)
Portman, M. V. A History of our Relations with the Andamanese compiled
from Histories and Travels, and from the Records of the Government
of India. Calcutta, 1899.
Hook-swinging

Powell, J. H. **
in India % in Folk-Lore, voL xxv
1914.
Przyluski, T. *
Le nom de dieu Visnu et la l<gende de Kfsna *, in Archiv
Orientdlni, iv, 2. Prague, 1932.
Ptolemy. KLAUDIOU PTOLEMAIOU ALEXANDREOS peri ms GEO-
GRAPHIAS biblid okto. (The edition used is that of P. Bertius pub-
lished in Amsterdam in
^
1618.)
Purchas, S. Purchas his Pilgrimage, or Relations of the World, etc. (The
edition used is the fourSi, 1626.)

Rajkhowa, B. Short Accounts of Assam. Dibrugarh, 1915.


Rakshit, H. K. See Sarkar.
Rangachari, K. See Thurston.
Rao, H. Indian Caste System. Bangalore, 1931.
Rapson, E, j. (Editor). The Cambridge History of India. VoL i, Ancient
India. Cambridge, 1922.
Ray, Gautamsankar. See Sarkar, S. S.
Reclus, Primitive Folk* Studies in Comparative Ethnology: (in the
E.
Contemporary Science Series edited by Havelock Ellis an<f published
by Walter ^
Scott). N.B.
Rivmwtrr, M.-E. Cours de Droit dgyptien. (Ecole du Louvre.) Paris,
1884 -
Rice, S. The Origin of Caste % in A.R., voL xxv.

1929.
Hindu Customs and their Origins. 1937.
Richter, E. Manual of Coorg. Mangalore, 1870.
Risiey H. H. Tribes and Castes of Bengal---Ethnographk Glossary.
Calcutta, 1891. (Cited as Tribes and Castes.)
- Tribes and Castes of Bengal^-Ahthropormric Data. Czlmm, 1801.
(Cited by title m full.)
^
BIBLIOGRAPHY *73
RiSLEY, H, H. The People of India. 1908. (The edition used here is that
of 1915 edited by Crooke,)
RiSLEY, H. H. and Gait, E. A. Report on the Census of India, 1901, voL
I, pt. i. (Vol. I of the Census of India, 1901.) Calcutta, 1903.
Rivers, W* H. R. The Todas, 1906.

The Origin of Hypergamy in the Journal of the Bihar and Orissa
Research Society, March igsi- Patna,. 1951.
Robertson, G. S. The Kdfirs of the Hindu-Kush. 1896.
Rochon, a. See Oliver.
Rodriguez, E. A. The Hindoo Castes, The History of the Bramihical
Castes, containing a minute description of the origin, ceremonies,
idolatry, manners, customs, etc. Of the Forty Two sects of Bramins
of the British Indian Empire: deduced from authentic manuscripts after
particular investigations and inquiries. 1846.
Roscoe, Immigrants and their Influence in the Lake Region of Central
J.
*

Africa The Frazer Lectures,



in by divers hands, edited by
Warren R. Dawson. 193a.
Rose, H. A. A Glossary of the Tribes and Castes of the Punjab and North-
West Frontier Province, Lahore, 1911-19. (Cited as Tribes and Castes.)

Caste \ in the Encyclopaedia Britannica, 14th edition, vol. iv, igug.
(Cited as Caste.)
Rowe, G. S. See Williams.
Rowney, H. B. The Wild Tribes of India, 1885.
Roy, Sachin. Aspects of Padam-Minyqng Culture, Shillong, 1960.
Roy, S. C. The Oraons of Chota Nagpur. Ranchi, 1915.
The Hill Bhuiyas of Orissa. Ranchi, 1935.
The Kharids, Ranchi, 1937.
*
Caste, Ra and Religion in India in Man in India, vols. xrv,
vol. xvu, no. 4 xvni, nos. s and 3. Ranchi, 1934, 1937, ^ 93 ^-
; ;

Roy S. N. Supposed Animation of Inanimate Objects % in Man in India,


*

vol. IX, no. 4. Ranchi, igag.


Russeix, R. V. and Lal, H. The Tribes and Castes of the Central
Provinces of India. 1916.
Ryan, Bruce. Caste in Modern Ceylon. New- Jersey, 1953.

Sarkar, B. K., assisted by H. K. Raeshit. Folk Element in Hindu Culture:


a contribution to socw-religious studies in Hindu folk-institutions. 1917.
Sarxar, S. S.; Ray, Gautamsankar; Chakravartti, M. R.; Banerjee, A. R., &
Bhattacharjee, Papia. a Physical Survey of the Kadar of Kerala.
Memoir No. 6, 1959, of the Anthropological Survey of the Government
of India. Calcutta, 1961.
Sayce. a. H. * The Aryan Problem ^fiifty years later % in Antiquity, I, ii

(June, 19R7), p. 05n.


ScHOEBEL, C. UHistoire des Origines et du Ddveloppement des Castes de
l lndsi (Extrails, nos. 8 et 18, du Bulletin de la SociStS Acadimique
IndorChinoise, s s^rie, t. ii et m, Novembre i88 et Janvier 1883.)
Paris, 1884.
Scott, J. G. and Hardiman, J. P. Gazetteer of Upper Burma and the Shan
States, Part I. Rangoon, 1900.
[Scon, J. G,] Shway Yoe. The Burman: His Life and Notions^ i88a.

18
BIBLiOGRdFMr
Sedgwick, L. J. Report on the Census of Bombay Presidency (voL viii of
Census of India, ign). Bombay, igm.
Seligman, C. G. and Murray, M. A.
Note on an Early Egyptian Stan-
dard *; in Man, xi, 97. 1911.

Senary, E. Les Castes dans Vlnde. Paris, 1896.1 (The edition used here
is that of 1927.)

ShakespEar, J. Lushei Kuki Clans, 1912.


Shaw, W. Notes on the Thadou Kukts, vol xxiv, 1928, no. i.)
Calcutta, 1929.
Sherring, M. a. Hindu Tribes and Castes, Calcutta, 1881.

Slater, G. The Dravidian Element in Indian Culture, 19*4.

Sleeman, j. L. Thug, or A Million Murders, N.D. (c. 1939, but compiled


from General W. H. Sleeman's notes of 1839).
Sleeman, W. H. Rambles and Recollections of an Indian Officml, 1S44.
(The edition used is that edited by V. A. Smith. Constable, 1893.)
SoNNERAT, M. Voyages aux Indes Orientales et d la Chine, Paris, 1782,
Spencer, B. See Marett.
Srinivas, M. N. Religion and Society among the Coorgs of South India,
Oxford, 1952.
Caste in Modern India and Other Essays, Bombay, 1962.

Steele, A. Law and Custom of Hindoo Castes within the Dekhun Protnnees
subject to the Presidency of Bombay, 1868. (This edition, published
in London, is called a * new edition an earlier one seems to have
been published in Bombay in 1827.)
Stevenson, H. N. C. Status Evaluation in the Hindu Caste System
in J.R.AJ,, vol. Lxxxiv, 1954.
Stevenson, S. The Rites of the Twice-Bom (in *
The Religious Quest
of India *
Series). Oxford, 1920.
Strabo. Geographika, (The edition used is that published by Wolters,
Amsterdam, 1707.)
Strasser, R. The Mongolian Horde: translated from the German by
*
R.T.G.* with an introduction by Sir M. Sadler. 1930.

Thoothi, N. a. The Vaishnavas of Gujarat / being a study in methods: of


investigation of social phenomena, 1935.
Thurston, E. Ethnographic Notes in Southern India, Madras, 1906.
(assisted by K. Rangachari). Castes and Tribes of Southern India,
Madras, 1909.
Omens and Superstitions of Southern India, 1912.
Tod, James.' Annals and Antiquities of Rajasthan, 1829-32. (The edition
used is that of 1920, Oxford, edited by Crooke.)
Turner, A. C. Report on the Census of the United Provinces of Agra and
Oudh, 1931. (vol, xviH of the Census of India, 1931.) Allahabad, 1933.
Turner, J. G. S. Medical Census, Southern Provinces, (Census of iVigerie,
1931, vol. VI.) 1932,

Webb, A. W. T. These Ten Years, (Rajputanaf, voL xxiv, pt. i of the


Census of India, 1941.) Bombay, 1941.
1 Translated into English by Sir Denison Ross in 1930 as Caste in India,
BI B LIO GRA PH 75

"'bom/'i; o !^t'

emong the Cannibals, labouss
and edited by G. S. Koive.)
WILSON, J. History of the suppression
of infanticide in Western India
under the Government of Bombay. Bombay,
Indian Caste* 1877.
1855.

J
on the B(Kes, Castes and Trades
of Eastern Benml i88

'e 'Se.l SSafS^sg.'"-


"
Yule, H* See Polo.
Yule, H. and Burnell, A. C. Hobson-Jobson. 1886. (The
^ eaition used
edition
IS that of 1903, edited by Crooke.)

ZiNKiN, Taya. Caste Today. 196s.


"

GLOSSARY
Abor Assam tribe now known as Adi (al?ar*i *
uncivilized
or * untamed *)
AgSriya A primitive tribe of central India whose principal
occupation is smelting and forging iron.
Agarwala An important mercantile caste of upper India with
traditions of descent from a Naga (snake) ancestress.
Many are Jains by religion (see Crooke, Tribes and
Castes, i, pp. 13 sqq.).
Aghoripanthi A mendicant sect whose philosophic tenets lead them
to feed on human corpses and excrement..
Agnikula Division of the RSjputs regarded as descended from
Fire (agni), as distinct from those descended from the
sun and moon, probably because descended from
foreign tribes * reborn '
as Rajputs after a symbolic
purification by fire.

Ahar A caste of graziers and herdsmen in Rohllkhand.


AhTr A caste primarily of graziers and cowherds, also of
cultivators, widely distributed in northern and central
India.
Ahom Shan tribe from which Assam takes its name.
Ambalavisi A caste, or perhaps rather a group of castes, of temple
servants, having many subdivisions, in Malabar (see
Thurston, Castes and Tribes, i p. 30). The term
Ambalavdsi is also reported to be used sometimes as
a synonym of NambJssan, the name for a particular
caste, or subcaste, of Ambalavasi.

Anibatlan A Tamil barber caste


And Koraga See Koraga,
Angami A Naga tribe of Assam.
anuldmH Sec p. 55.
Ao A Naga (q.v.) tribe of Assam.
Apastamba An ancient writer on ritual and law.
Apa Tani Assam hill tribe north of the Brahmaputra River.
Aratta .An ancient Punjab tribe with matrilineal inheritance.
Arora A trading caste of the Punjab and north-west India
with some affinity to the Khatri, in regard to whom
they stand in some sort of inferior relationship.
dsram Religious order; hermitage; school or college.
Asura A divine being hostile to the Aryan gods; a pre- Aryan
people with magical powers and a knowledge of the
working of metals.
Asvamedha Horse-sacrifice; one of the great vedic rites.
Atharvaveda The fourth veda, characterized by a large number of
spells and incantations.
Avesta The collective sacred writings of the Zoroastrians.
Awfin A Muslim landed and cultivating caste or tribe,
claiming Arab extraction, in the Punjab.
Ayar See Idaiyan,
Ayyar (Aiyar) An honorific title of Tanoul Brahmans.
GLOSSARt 77
Bibhan (ssBhuInhar) A landowning and cultivating
caste of
upper India, particularly of Bihar; they claim to be
of Brahman extraction, but it is not unlikely that
they are really of Rajput derivation. See Risley,
Tribes and Castes, i, s.v. BSbhan.
Badaga A cultivating tribe (or caste) of the Nilgiri Hills; fire-
walkers.
BSLgdi A caste of fishermen, palanquin-bearers, and field
labourers in central and western Bengal, and in Bihar.
Baidya A caste of physicians in Bengal (sVaid).

Baiga A primitive tribe of central India.


BSiti (Baoti) A small caste engaged in the manufacture of lime from
shells, also in the making of mats, and of music.
Known otherwise as ChunSri, ChuniyS (from the
word for lime), or DhOli (drum), or Dugia.
Balija An important I'elugu trading caste.

Balti A Himalayan tribe.

BalQichi (Balocb) A tribe of Baluchistan (see pp. 40 sqq.).

Banajiga A Kanarese trading caste corresponding to the Telugu


*
Balija

Baniya A term applied to traders, moneylenders and shop-


keepers, and often used for vanous trading castes,
particularly those of Rajputana and western India.
Banjlra A caste of carriers, traders, and cattle-herders more or
less nomadic. Also known as *
Vanjfira , Lambadi

Labana (see p. at).

Barhai A caste of carpenters in upper India.


Barui A pan growers in Bengal.
caste of

ImatH A girl dedicated to a god after the custom of some


Kanarese castes, who, when they have no male
children, marry a daughter to the deity. She lives in
her fathers home and her children take his name and
no stigma attaches to her or to them.
Bauri A caste of field labourers in Bengal and Bihar.
Bedar A Kanarese fighting caste of southern India (*B6ya).
BSkanwala Pork-butchers, a subcaste of Khatik (
< bacon).

Ben 4 Israel A Jewish caste of Bombay (see p. 19).

Bgri Cbctti A caste of traders tn south India (see Chetti).

Bhaugi Caste of sweepers and scavengers over all Ihdia except


the south.
Bhar A tribe of the United Provinces and Bihar, formerly
of some political and social Importance, now reduced
to a caste of low status,

BharbOiija A caste of grain -parchers in upper India,


Bhat A upper India of hereditary bards, genealo-
caste in
ists, and heralds, probably of mixed Brahman and
fLajput extraction, but claiming Brahman descent.
They used to have the reputation of compelling the
performance of any promise made to them, or of
extorting money, merely by threatening to kill them-
selves or some member of the family case of non- m
compliance. They had also the reputation of abiding

18
78
GLOSS ART
by kind (cf. ChSran '). There are some
threats of this '

Muslim Bhats following a mixture of Muslim and


Hindu practices.
Bhitia A trading caste of western India.

Bhatra A forest or hill tribe (or caste) in Orissa.

Bhll An important tribe of west central India noted as


bowmen and bandits.
Bhmia A tribe in western India of mixed Bhll and Rajput
extraction.

Bhoksa A tribe in the Himalayan foothills of the United


Provinces, and claiming Rajput extraction (see Crooke,
Tribes and Castes, n, pp. 35 sqq.).
Bhotiya A Mongoloid tribe of the lower Himalayas claiming
an origin from Bhutwal in Nepal and now trans-
formed into a Hinduized caste (see Crooke, Tribes
and Castes, 11, p. 61).
Bhulnhir See Babhan,
BhuinmSli A cultivating, palanquin-bearing and menial caste of
eastern Bengal, probably representing an aboriginal
tribe: m
Sjdhet it now calls itself * Mali
Bhuiya A Kolarian tribe of Chota Nagpur and Bengal.
Bhumi Belonging to the soil, from die Sanskrit bhu^* earth %
and thus sometimes used in the names or titles of
castes or tribes claiming to be the original cultivators
of the soil or to be the lords thereof.
Bind A caste of cultivators and field labourers in the
United Provinces, probably of Rolarian tribal origin.
biradari See p. 99.
Birhor A quasi-noraadic Kolarian forest tribe of Chota
Nagpur and Orissa.

Bishnoi A caste of sectarian origin in Rajputana and adjoin-


ing areas. Though spoken of as Bishnoi Brahmans
by Tod, they are, in fact, derived from various castes,
particularly (according to Rose, Tribes and Castes,
n, s.v. Bishnoi) from Jats and Khatris.
B5hra A trading caste of western India now predominantly
Muslim (see Enthoven, Tribes and Castes, t, p. 19*7,
s.v. Bohora

Russell, op. cit., n, p. 345 j Ibbetson,
op. cit., p. S45).
Bondo Barja A primitive Kolarian tribe of Orissa.
Br 3ihxnan A vama comprising a large number of castes whose
traditional occupation is that of priests and who stand
as a rule at the apex of Hindu society (see pp. 64-7).
They are divided roughly into two great classes said
to be of five orders each, the Panch Dravida and the
Ranch Gaur, roughly south and north of the Vindhya
Hills respectively, but there are other classes no
longer recognized as belonging to either of these two
groups and in some cases regarded as degraded.
brahmanas Religious works of the post-vedic period consisting in
theological treatises explaining the vedas. They ex-
pound the sacrificial ceremonial in minute detail with
speculations as to origin, etc. The brahmanas prob-
ably belong to about the seventh century b.c.
GLOSS ART
Braliraarshidgsha
The country of the holy sages i.e. the upper
region between the Jumna and Ganges rivers.
Brahmavartta
The holy land \ i.e. the country about the river
Saraswati, in the neighbourhood of Sirhind and
Ambala in the Punjab south-east of tlie Sutlej River.
Brahui A Muslim tribe of Baluchistan speaking a Dravidian
language (see pp. 41, 151).
Brittiai Baniya An exterior caste of Assam (see pp, 113, 313 sq.).

Brokpa A Himalayan tribe of the Hindu Kush,


Burusho A Himalayan tribe of the Hindu Kush speaking a
language so far not identified as belonging to any
recorded linguistic family.

Cliakkiliyan A Tamil caste of leather- workers corresponding to


Chamar in the north, and Madiga in Telingana.

Chakli See Chakkiliyan.


Chakml Tribe of Chittagong Hill Tracts in East Pakistan.
Chaliyan A caste of cotton-weavers now found in Malabar.
Chamar Skinners, tanners, and workers in leather; northern
and central India generally.

ChamSrgaur A subcaste of Rajput with both Hindu and Muslim


branches.
Chandal An exterior caste, traditionally the lowest in Hindu
society and sprung from the union of a Brahman
woman with a Sudra man; specifically a caste of culti-
vators, boatmen and fishermen in JBengal (=Nama-
sudra).
Chang A /Naga tribe of Assam.
ch&pati See p. 73.
CharSl =Chanral or Chandal, q.v.

CUaran A caste Gujarat of hereditary bards, genealogists,


in
and heralds claiming Rajput origin, and possibly of
mixed Rajput and Brahman extraction, with customs
similar to those of Bhat (q.v.), like whom they had
the reputation that they would keep their word or die.
chm A cultivating caste of Orissa.
Chattarkhai A caste in Orissa derived from persons who lost their
caste in 1886 by eating in relief kitchens (chatira)
during the famine (see Risley, Tribes and Castes, i,
p. 196).
chattri A platform, generally of stone, covered by a domed
roof but with open sidesa sort of loggia.
CliaubS =Chaturvedi, so named because they follow the

fourth, Atharva, veda a caste of Kanyakubja or
Kanaujya Brahman in upper India.
Chenchu A I'elugu-speaking forest tribe, mostly food-collecting
and non -cultivating, of southern India.
Chero A of cultivators of tribal origin and Kolarian
caste
affinitiesin the Mirzapur district of the United Pro-
vinces and in adjoining areas (see Crooke, Tribes and
Castes, II, p. 241).
Cheraman A South Kanara and
caste of agrestic serfs in part of
in southern Malabar corresponding to, or identical
a

u8o GLOSS ART


with,Pulayan (q.v.) in northern Malabar, Cochin,
and Travancore*
Chetti A caste of bankers, brokers, shopkeepers, money-
lenders, etc., in southern India; corresponds to Baniya.

Chin A group of hill tribes in Burma and Assam.


ChitpSvan A caste of Brahman from the Konkan in western India
distinguished for tlieir administrative capacity and
their fairness of complexion.

ChSdhra A more or less nomadic hill tribe or caste of Bombay.

ChQhra A caste (or tribe?) of scavengers in the Punjab and


north-west India (see pp. 38-9, tip).

jDafla An Assam tribe on the north bank of the Brahma-


putra,

Daitya A race of demons or titans that fought against die


gods; representing, perhaps, the indigenous inhabi-
tants of India who fought against the Aryan-speaking
invaders.

dao A bill or broad-bladcd hacker with a single cutting


edge.
Dard Himalayan tribe of Dardistan.

darga The shrine of a Muslim saint (properly royal


court ').

Darzi Tailor caste in India generally.


D 3 syu A race of dark-coloured evil bemgs, enemies of gods
and men; probably representing aboriginal inhabi-
tants of India who fought against the Aryan -speaking
invaders, and/or perhaps against more civilized in-
habitants who preceded those invaders as immigrants.
Debri (or Deori Chutiya) A Assam credited with magical powers.
tribe in

Dgshasth A caste of Maharashtra Brahman coming originally


from the dlsha^ i.e, the country above the Sahya
Ghats in western India.
D5va The race of gods, no doubt representing in effect the
Aryan-speaking invaders of India who fought against
the Daityas, q.v.
ddvadSsi Girl dedicated to the service of a deity, and generally
to temple prostitution.

Dgvanga A caste of weavers in the Carnatic country.


DSvSndrakula A fancy name for the Pallan (q.vj caste, who, it Is
Veliaian thereby claimed, were created by Devendra, the king
of the gods, to labour for the Velllla.

Dhangar A shepherd caste of western India.


Bhanuk A caste of field labourers in upper India,
dMr&msUld Rest-house for wayfarers (lit. * pious edifice *).

Dharhi (or Dhadhi) A caste of musicians or panegyrists.

dh&rm& Duty; conformity to the laws; justice; piety. See p. 1^4.


dharmasUstra See sdstra.
Dhed An exterior caste of field labourers and private ser-
vants in western India,
OLOSSART
Dh6bi A widespread caste of washermen standing low in the
social scale on account of their occupation -(ssDhubi,
Dhupi),
DhCH See Haiti.
DogrS An inhabitant of the Dugar tract in the Himalayan
state of Jammu.
DQm A widespread caste of scavengers, musicians, and some-
times weavers, traders, or even moneylenders; possibly
representing an aboriginal tribe of some influence and
power (ssDCmra, D5mbu). See pp. 54-5.
DOmbo sBdm.
Dugla See Haiti.
DumSl An aCTicultural caste of Orissa, claiming affinities to
the Goala.

Emadan A Malabar hill and forest tribe of very low social


status and apparently having unusually pronounced
brachyccphaly. Sec also footnote 27 on p. 80.

FermghI A Christian caste of Hengal claiming descent from Por-


tuguese pirates (Feringhi < Frank '
i.e. European ).

GadabS A primitive Kolarian tribe of Orissa.


Gaddi A ca^te of graziers in the Punjab, partly Muslim,
partly Hindu (see Rose, Tribes and Castes^ n, pp. a5K,
sq.)

GarhwSli Ah inhabitant of GarhwSl in the Himalayas.


Garo A matrilineal hill tribe of Assam.
gaumedha The sacrifice of cattle.
Gautama The founder of the NySya school of philosophy and
author of a dharmasSstra or law book.
GayawSl A subcaste of mendicant Hrahiman maintaining them-
(or Gayali) selves on the offerings of pilgrims in the city of Gaya.
Ghanchl A caste of oil-pressers and sellers in western India
(corresponds to Teli).
Ghasiya A caste of musicians, fishermen, and artificers in brass,
of low status in upper central India.
ghee (or ghi) Clarified butter.
GhbshI A pastoral caste, commonly but not always Muslim,
acting as milkmen in the Punjab.
Gbaia A caste of cattle-breeders and milkmen in northern
India generally.
Golapurab An agricultural caste, probably of Brahman extrac-
tion, in the district of Agra.

Golla A Telugu caste of herdsmen, cattle-brecden, and


milkmen in southern India.
G6nd A widespread tribe in central India.
Gbpa A caste of herdsmen, cattle-breeders, and milkmen in
Bengal (ssGfSala, Ahir, etc.).
gdtra An exogamous group, descended from a common
ancestor, inside the endogamous caste (cc pp. 47
sqq. and index s.v, gdirn).
JiSs
GL a SSdRt
Gfijar A grazier caste of the Punjab and eorth-west India
generally; their origin is attributed to the White Huns
or Ephthalites who came into India in the third
century a.u. (see p. 38), Many are Muslims, e.g, in
Kashmir,
Gujarati Inhabitant of GujarSt, formerly a part of the Bombay
Presidency, and the language of that area.

guru Spiritual adviser.

Halalkhor A subcaste of Bhangi, q.v.


subcaste of Agarwala < haldi =
'

Haldiya A turmeric % the eat-


ing of which is taboo to them.
Haliya Kaibartta Kaibartta of the plough (=Ghasi Kaibartta or
Mahishya).
Halvakki Vakkal A cultivating caste of Kanara with quasi- totcmistic
clans.

Halwai A caste of confectioners in northern India generally.

Handi J 5gi A caste of raendicanus, pig-breeders, quacks and snake-


charmers in southern India.
HSrT A scavenger caste of eastern India (=:Haddi).
HasSla A caste (or tribe) of agrestic serfs and collectors of
forest produce on the Kanara border of Mysore State.

HazSra A Himalayan tribe of the Hindu Kush.


Hira A subcaste of ChandSl (q.v.) Apparently makers of
pots and of brass wire, vide F. Hamiltons Account
of Aesam, pp, 55, 60, and Martin, Eastern India,
in, pp. 676, 679.

Ho A Kolarian tribe of east central India.


Holeya A Kanarese caste of field labourers and village ser-
vants, formerly agrestic serfs (corresponding to
Paraiyan).

Holt The Spring festival. See Crooke, Things Indian, s.v.


Festivals and Hobson-Jobson s.v. Hooly.
Huligamma A southern Indian goddess, with a temple in
Dharwar District, who is worshipped by eunuchs or
by male priests in the dress of women (see Fawcett,
'
On Basivis % Journ. AnthropoL Soc, of Bombay, n,
P- 343 )-
huqqa Tobacco-pipe in which the smoke is drawn through
a water-container.
Hur Quasi-religious follower (bandit) of the Pir Pagaro in
Sind (see p. 40).
hypergamy The system under which women from a group of
lower status are married to mm
in a higher group.
hypogamy The reverse of hypergamy.

Idaiyan A Tamil caste of herdsmen, shepherds, cattle-breeders,


and milkmen in southern India (=Ayar, Golla).
Idayar See Idaiyan.
Ilavan Sec Izhaw "1.

Iluvan See Izhavan,


Iravan See Izhavan*
GLOSSARY S83
Imla A tribe of the Nilgiri Hills and related caste of the
same name in the Madras plains.
Iruvan See Izhavan.
Izhavan (=Iravan, Ilavan, etc;) A toddy-drawing caste of the
Malabar coast nearly related to the Tiyan, and believ-
ed to have come from Ceylon, or from Indonesia by
way of Ceylon. They are matrilineal in Travancore
and southern Cochin, but patrilineal in northern
Cochin and southern Malabar,

JaiswSl A trading caste of Rajputana,


Jaliya Kaibartta Kaibartta (q.v.) of the neta fishing caste of Bengal.
Jalua A term applied in Bengal to fishing castes in general,
but in. Noakhali district apparently to a specific caste
with endogamous groups (see Risley, Tribes and
Castes, 1, p. 340).
Janappan A Telugu caste which make gunny-bags of hemp fibre
and act as hawkers and cattle-brokers.
Jat An agricultural caste of upper India (see pp. 36 sqq.).
jatakas A collection of the birth stories of Buddhas that pre-
ceded Gautama.
Jhalo Probably a synonym for Malo (q.v.); in any case it
refers to a caste of fishermen in Bengal.
Joiaha A caste ofMuslim weavers widespread in upper India
(see Crooke, Tribes and Castes, m, p. 69).
Juang A primitive Kolarian tribe in the Orissa hills.

Jgi A caste of rearers of silkworms and weavers of silk in


Assam.

KabTrpanthi A sectarian caste drawn largely from weavers; the


followers of the reformer Kablr* who flourished at the
end of the fifteenth century; it has Muslim as well as
Hindu practices.

Kabuli A native of Kabul; an Afghan or Pathln.


KachSri An Assam tribe, now mainly Hinduized (see Kbchh).
KSchhl A casteopium-growers and market-gardeners in
of
upper India (=Sahnai).
Kachin A group of hill tribes in Burma and on its northern
borders.
KMar A food-collecting (non-cultivating) forest tribe of
southern India,
Kadva KunbI A subcaste of Kunbi practising periodic marriage
(see p. 19).

KSfir A Himalayan tribe in the neighbourhood of Gilgit


(see J). 43).
Kah5r A caste of fishermen, porters, and domestic servants in
upper India. Kahars also commonly act as well-
sinkers and cultivators of water-nuts.

Kaibartta Originally apparently a tribe, now a caste or castes of


Bengal and adjoining areas, their occupation being
fishing or cultivation i(?=Kewat, see -pp. p, ao8),

KaikSlan A Tamil caste of weavers.


a 84 GLOSSdRT
Ka:iit A cultivating caste of Assam (?= Kalta of Orissa).
Kalian A cultivating and predatory Tamil caste notable for
their efficient agriculture, ejcpert thieving, robbery,
lengthened ear lobes, use of the Indian
cattle-lifting,
boomerang, and practice of a sort of bull-fighting
(see Thurston, Castes and Tribes, iii, pp. 53 sqq.).

Kalta (or Koim) A cultivating caste of Orissa (?BBKaliti of Assam).

Kalwar A caste of distillers and liquor-sellers in northern


India (ssKalSi).
Kami The blacksmith caste of Nepal; found also in adjoin-
ing areas.
Kamma A numerous Tclugu caste, so-called after an ear-
jewel of the goddess Lakshmi, consisting of cultivators
claiming Kshatriya origin; sometimes divided into five
groups according to the manner in which they carry
water, viz, on the head, on the hip, by the hand
only, by two persons carrying between them, or on a
pack-bullock. Besides gotra they have a number of
exogamous septs, some with quasi-totcmic names (see
Thurston, Castes and Tribes, ni, pp. sqq.) Cf, infra
s.v. '
OkkalTga \
Kammalan See Panchala.
Kammavan A Telugu owners and cultivators in
caste of peasant
Tinncvdly practise a form of cross-
district; they
cousin marriage in which a man must marry his
fathers sistcr^s daughter, or his mothers sisters
daughter, or his own sisters daughter.
KanbI See Kunbi.
Kant A caste of peasants in the outer Himalayas formerly
practising fraternal polyandry.
Kaniyan A caste of astrologers on the Malabar coast (wKani-
san).

Kanjar A gipsy caste of criminals and/or mat-makers in upper


ana central India.
Kansara See KasSra.
Kapu A very nuracrous caste of south India primarily culti-
vators; insome districts they rank next to Brahmans
and claim a jRaJput origin* They use the title

Rcddi %
Karalar Rulers of Clouds; applied to a tribe of cultiva-
tors and hunters in South Arcot district, to the
Vellala caste m
Malabar, and to the Malayali of
the Shevaroy hills in Salem district in southern India.
Karan A writer caste of Orissa (corresponding to KSyastha).
k&rmU Sec p. 1^5.
Karuppan A South Indian god associated especially with the
Kallar caste.
KasSra A caste of brass-founders and coppersmiths in nor-
thern India generally ( Ktstr).
Katlcarf A more or less nomadic forest tribe of Bombay.
Katooi A cane of weavers in Assam (also known as Nith).
Kayattba' (Klyaith)^ A writer caste of upper and of eastern India (ranks
next to Brahman in Bengal),
QIOSSARY ,85
KSwat A fishing and cultivating caste of upper India (corrfe-
sp- ads to Kaibartta).

Khandait A quasi-military cultivating and landowning caste of


Orissa.

KhangSr A caste of watchmen in the Central Provinces.


Kharis A Kolarian tribe of Chota Nagpur and central India.
KhSrvS A seafaring and salt-making caste of western India.
KharwSr A cultivating caste of tribal origin and Kolarian
affinities in Bihar, Chota Nagpur and adjoining areas.

Khas A tribe (or caste) of Nepal.

KhSsi A matrilineal tribe of Assam speaking a Mon-Khmer


language.
Khatl A caste of Cartwrights and carpenters in the Punjab.
Khatik A caste of labourers, butchers, and vegetable sellers,
etc., in northern India.

Khatxi A trading caste of the Punjab and north-west India.


Kho A Himalayan tribe of the Gilgit area.
Kh6ja The name Muslim trading castes of western
of two
India; the Khoja of the Punjab are Sunni and are
largely derived from the Hindu caste Khatri; the
Khoja of Sind, who are Shiah, are similarly largely
derived from the LohSna caste of Hindus of mat
province.
KisSn A cultivating caste of upper India, probably allied to
KSchhi, Koiri, and Kurmi, q.v.
K6chh A tribe of Assam and northern Bengal; in latter area
generally Hinduized as the caste of Rajbansi, but in
Assam K6chh is used as a caste label for several

tribes when Hinduized, particularly KachSris, LSlungs,


and Mikirs.
Kodaga (or Koraga) The warrior caste or tribe of Coorg.
Koiri (or Koeri) A cultivating and market-gardening caste of upper
India.
K61 A Kolarian tribe of central India and Orissa.
Kolarian Possessing a language or culture associated with a
group of peoples of which the principal entities are
the Munda, Kdl, and Santal trioes.
KSlI A caste of cultivators and labourers in western India
possibly of Kolarian affinities.

Kolta = Kalta, q.v.


Kdmati A Telugu trading caste.
Kond (or Kondh) A hill tribe of Orissa, noted fox human saaifice and
infanticide, nearly allied to Gond (spelled commonly
Khond, also Khaund, Kandh, etc.). See pp. 25-6.
Konkan The coastal area of Maharashtra.

Konyak A group of Naga tribes in Assam.


Kora A caste earthworkers of Kolarian extraction in
of
Chota Nagpur and Bengal.
Koraga An exterior caste of basket-makers and labourers in
South Kanara district. The women wear petticoats of
leaves. The Andepot ) Koraga had to wear a
('

small pot suspended to a string round the neck in


which to spit so as to avoid polluting the roads.
m GLOSS ART
Kopga See Kodaga.

Korava A nomad Tamil caste (or tribe| of fortune-teliens,


quacks, and thieves operating particularly on railways*

Korku A hill or forest tribe of central India.

Korwa A hill or forest tribe of central India.

K5 ta A tribe, or low caste, of the Nilgirls; musicians and


artisans*

mtui Vdiaia A small and exclusive subcaste of Vellaia (q.v.) living


within a fort (^ottat) in the Tinnevelly district of
southcm India.
Ksbatriya A mtm; the ruling and warrior class of the vedas,
now represented by Rajputs, and by other castes
claiming Kshatriya origin (see pp. 64-7),
Kudumi A caste of servants, firework-makers., and (formerly)
soldiers in Travancore.

Kuki A group of hill trib in Assam and Burma*


Kulin Sec pp. 53 sqq. of this volume.
KurahSr The potter caste; all India except the south* The
KumhSr have many endogamous subcastes.
Kunbi A cultivating caste of western India (* Kurmi ; also
aKanbT).
Kurava A Malabar caste of agrestic serfs*

Kuricchan A caste (or tribe) in Malabar which subsists by hunt-


ing and by shifting cultivation.
Kurmi A widely distributed cultivating caste of northem
India (= Kunbi).
Kiiruba A caste of shepherds, weavers, and stone-masons in
southern India, probably identical originally with
Kurumba or Kuruman (see Thurston, Castes and
Tribes, iv, pp. 155 sq.)

Kusavan A Tamil caste of potters (corresponds to Kumhir).

LSlbegi A scavenging caste of upper India with customs partly


(or Lal-Begi) Hindu, partly Musli.m (see Crooke, Tribes and Castes,
I, s.v, Bhangi).

LambMi See BanjSra.


Lapcha A hill tribe of the Himalayas, Bengal*

Let A subcaste of BSgdi {q*v.) in the Murshidabad a.nd.


Birbhum districts.

Lhota; A Naga tribe of Assam.


Ungam The symbol of the God Shiva; a stone of cylindrical
or phallic shape.
LingSyat A numerous of sectarian origin derived from
caste
different from Brahman downwards, in the
strata,

west of India, and frequently in rivalry with the


Okkallga, q.v. {see Enthoven, Tribes and Castes, 11,
PP' 343 2md also I4.K.A. Iyer, Mysore Tribes and
Castes, jy, pp. 81 sqq.)
GLOSS ART
LodhS A caste of cultivators and labourers in the United
Provinces of Agra and Oudh, probably identical with
Lodhi, a caste of landowners, probably of Rajput
affinities in central India (see Russell, op. cit., iv,
p. 112, and Rose, Tribes and Castes, in, p. 35).
LOhSna A trading caste in Sind.
L^har A blacksmith caste, widespread in India north of the
Godavari River.
Loi A tribe of Manipur State.

Ldk Sabha The lower house of legislature.


Lori (or Ltiri) A nomad tribe of Baluchistan, tinkers and musicians,
probably = Dom. Luri perhaps < Lur, = Luristan;
or else Luristan < Liiri,

Lusbei A Kuki-Chin tribe of Assam.

Madiga .A Telugu caste of leather-workers, corresponding to


ChamSr and Chakkiliyan in northern India and the
Tamilnad respectively,

Magh A Buddhist tribe of Bengal; of Burmese, or rather


Arakanese, affinities.

MahabrShman A caste of Brahman, the function of which


is to receive
the funeral offerings of clothes, jewellery, etc., which
belonged to the dead man, and by the wearing of
which the soul of the dead is provided with necessities
and luxuries in the next world.
Mahar A clan of Huna extraction in Sind. Also a subcaste of
Kahar (q.v., and p. 90),
Mah5r An exterior caste of field labourers and village servants
in west and central India.
Mahara Used in Sylhet probably for the caste of Kahar (q.v.).

Mahishya A one of the Sanskrit commentators


caste, according to
on the Code of Manu, sprung from a Brahman father
and a Kshatriya mother, or, according to others, of a
Kshatriya father and a Vaishya mother, and ranking
with Kayastha (see Wilson, Indian Caste, i, pp. 55,
65). Such a caste, however, is no longer known, at
any rate by that name, and the name has been adopt-
ed by the ChSsi or HSliya Kaibartta as a status-raising
designation for tliemselves (see p. 113). See Gait,
Census of Bengal, 1901, p. 380.
MShli A caste of labourers, porters, and basket-makers in
central India, having affinities with the Santal, Ho,
and Munda tribes (=:Mahili).
mahwd The flowering tree Bassia latifolia (Roxb.), wild, or
Bassia longifolia, cultivated, from the blossom of
which liquor is distilled.
Maithil Brahman A Brahman caste originally associated with the ancient
kingdom of Mithila rougnly corresponding to Bihar.
Majhwar A forest tribe (or caste) apparently of mixed origin in
upper central India.
mahkathdyam Descent in the male line, women being normally
excluded from inheritance,
MSIa A Telugu caste of field labourers and village servants
(corresponding to Pulayan).
i88 GLOSSARf
Mala-Arayan A tribe of Travancore who make miniature
hill
dolmens to house their dead, who are represented by
a small brass image or a smooth stone.
Mala-Pantaram Food-collecting (non-cultivating) forest tribe of Tra-
vancore.
Malavtan A primitive tribe of southern India (Vedtn).
MalaySlam That branch of the Dravidian tongue spoken on the
Malabar coast (Kerala) from Cannanore southwards to
Cape Comorin.
MalaySli A Tamil caste inhabiting the hills {malai) in Salem
district; probably derived from the Vellsila caste
(q.v.).

Malayan A patrilineal caste of Malabar which cures disease by


expelling the demon it by dancing in
responsible for
masks and disguises (erSnSn; maByan means hill-
man *).

Malar (Mai) A hill tribe of Bihar and Bengal.


Mali A caste of domestic and market gardeners in all India
except the south.
Mali (In Sylhct)Bhuinm5li, q.v.

Malo A caste of boatmen and fishermen in Bengal,


m&na An Oceanic concept of mysterious impersonal power
attaching to indiviauals, objects, places, or even words,
and set in motion as a rule by ritual; the positive
aspect oi the supernatural as opposed to taboo, the
negative aspect involving the avoidance of such power
as dangerous.

Mannan A Malabar caste of washermen, *=Vannan but follows


the mammakkathSyam system; not to be confused
with the hill tribe of Travancore called MannSn.
mantra A text of the vedas used as a magic formula.
Manu, Code of A digest of existing laws and creeds probably of about
the third century B.c. but based upon earlier works.
Maratha A warlike community of west^n India (see pp, ig-so).

Maravan A cultivating, marauding, caste of the


cattle-lifting
extreme south of India having obvious cultural
affinities to the Kalian whose practice of bull-fighting
they share as well as the use of boomerangs (see
Thurston, Castes and Tribes, v, pp. sqq-)

Marayan (or Maran) A caste of drummers, temple servants, and barbers


serving Nayars and higher castes; their functions and
status vary to north and south of the Malabar coast
(see Thurston, Castes and Tribes, v, pp. 5 sqq.).

Maria A subtribe of Gond in east central India.


marumakkath&yam System of descent and inheritance in the female line,
so that a mans heir is his sister or his sister^s child.

Marwari Strictly speaking an inhabitant of Mirwir in Raj-


putana, loosely used for BaniySi hrom that region.
Maurya The name of the dynasty, and the empire associated
with that dynasty, founded late in the fourth century
B.c. by Chandragupta. It reached its zenith in the
third century unaer the great Buddhist emperor
Asoka, and broke up early in the second.
GLOSSdRt 389

Mazhbi A section of the Sikh community socially of inferior


status as having been recruited from low or exterior
Hindu castes.

Meithei A tribe, the ruling people of the Manipur State (see


p. 116).
MelakkSran A name given to two castes of musicians in south
India, the one Telugu, the other Tamil (see Thurston,
Castes and Tribes, v. p. 59 ).

MSman A Muslim section of the L5h5na (q.v.) caste in Sind.

Mgo A Raiputana and upper India following mixed


tribe in
Hindu and Islamic practices. When predominantly
Muslim they are called Mew5ti, but Meo when
regarded as Hindu.
Mer, Merat A tribe of Merwara in Rajputana. The Muslims call
themselves Merat
the Hindu
Mer (alsossMair)
generally use the title Rawat.
meriah Person reserved, often hereditarily, for sacrifice, to
promote the fertility of the soil, by the Kondhs of the
Orissa hill tracts; loosely used also for the rite of
such a sacrifice.
MSwafarSsh Fruiterer; a subcaste of Khatik.
Mikir A hill tribe of Assam.
Mina A tribe in Rajputana, probably of common
marauding
origin with Meo (see Crooke, Tribes and Castes, s.v.
Meo).
ntirasdar Proprietor.
A Muslim caste of genealogists and musicians in the
Punjab.
Mishmi An Assam tribe north of the Brahmaputra River.
M5chi See Much!.
Momln (i) A caste of weavers (=:Jolah5) in northern India;
(ii) == Meman.
Mopiah (MSppila, A fanatical Muslim tribe or caste of western India,
Map ilia) traditionally of partly Arab extraction. They are
matrilineal in north Malabar but patrilineal in the
south of it. (See Hobson-Jobson s.v. Mopiah.)
Motasu A section of the KSpu (q.v.) caste.

Mm A Buddhist tribe in the Chittagong Hill Tracts of


Fast Pakistan.
Much! A caste of workers in leather all over India except the
soutli; commonly =3 ChamSr (also M5chi).
Mukkuvan A SMalabar caste of fishermen, litter-bearers, and
boatKpen.
Munda A Kof^rian hill tribe of Chota Nagpur. They put up
raegalirhic monuments to their dead.

muni *
A holy sage, a pious and learned person, endowed
with mote or less of a divine nature, or having
'
obtained \t by rigid abstraction and mortification
(Haughton p. The title is applied to rishis
and others.
MusShar A caste (or primitive tribe) of Kolarian affinities in
Uttar Pradesh and Bihar.
Musalli Muslim of inferior status socially on account of the
low status of the Hindu caste of his extraction.
^ ,

ago Q LOSS A
NadiySl An boatmen and fishermen in Assam.
exterior caste of

Naga A group of hill Assam and Burma (see p.


tribes in
183). Naga %Naga (i.e. neriga naked 7 ) peopite.
Niga Literally' snake a ptc^Arym people in India, with
whom the worship of snakes, particularly the cobra,
is associated.

NSgar Brahman A casta of Gujarati Brahman associated in origin witir


half a d02:en towns in tlie north-east of Gujarat
(
< naguftown),
Nai A barber caste all over northern India (=*Nhavi).
NamasQdra A numerous exterior caste of cultivators and boatmen
in Bengal, formerly known as Chand^l (q.v.), Risley
(Tribes and Cmtcs, i, p. 181) regards them as repre-
senting an aboriginal tribe and possibly allied to the
MSler of the Rajmahal hills; their own traditions
assign them an origin from Brahmans or other caste
Hindus degraded for one reason or another,
Nambadri A caste of Brahman in Malabar in which it has been
the practice for the eldest son to make a norma!
endogaraous marriage while the younger .sons contract
with Nayar women unions governed by the raatri-
lineal inheritance laws of the latter.

nSmghar Communal house of meeting and worship in Assam.


NlnagStri A caste of Brahman in Garhwal (see p. 154).

Nat A caste of singers, dancers, acrobats, and professional


criminals (the feminine is natni or noli) who generally
move about without a permanent territorial location
(sec F. Hamilton, op. cit,, p. 55, Martin, op. cit., in,
p. 676, and in particular Crooke, Tribes arid Castes
IV, pp. 56 sqq.).

nat A spirit or godling in Burma.


Nath Jugi or Katoni (qq.v.).
Nattukbttai A subcaste of Chetti in Madura district who are
moneylenders and traders, with distinctive customs of
their own (see Thurston, Castes and Tribes v, s.v.
Nlttukhttai),
nattuvan A name for teachers of dancing-girls whose per-
formances they accompany as musicians, I'hey may
belong to any of several castes, Kaikdlan, D^vSnga,
Occhan, for instance, among them.
Nayidi An exterior caste (or tribe) of the Malabar coast,
hunten and beggars by profession, polluting to
Brahmans at nearly too yards distance (sec pp. 80-1).
Nayar An important aristocratic matrilineal caste of the
Malabar coast whose traditional occupation is fighting
in the case of the higher subcastes, but the caste as a
whole includes oil-sellers, potters* herdsmen, barbers,
washermen, and others by no means military.
Newar A tribe of Nepal claiming to have been in occupation
before the Gurkhas, and largely Buddhist in religion.
Nishada An ancient caste derived traditionally from a Brahman
and a Sudra woman; a dark-skinned people settled in
the country before the Rigvcdic invaders (see Wilson,
Indian Caste, i, pp. 55, ii6, i6n, 36, 241, 418).
^

G LOSS ARt
Occhasi A caste of temple priests at village shrines in .southern
India.
Odde 5= Odh, q.v,

Odh A migratory caste of earthworkers ( = Orh, Vaddar,


etc.) probably of Kolarian origin from Orissa.

Ojlia A caste or subcaste of medicine-men and magicians;


tiiere are Ojhas of Brahman as of other castes, and
there is a subtribe of Gond of this name.
Okkailga A group of agricultural castes in Mysore and
Eanara. Like the Kamma, they have a tradition of
escape from persecution across a river which parted
its waters for them as the Red Sea for the Israelites.
Women practised mutilation of the hand at the ear-
boring of children, like the Morasu section of the
Kapu. Cf. supra p. ii (see L. K. A. Iyer, Mysore
Tnbes and Castes iv, pp. S5 sqq.).
Oraon A hill tribe of Chota Nagpur using a Dravidian
language.
OswSl A trading caste from Mewar in Rajputana. This caste
isvery largely Jain by religion, but Hindu and Jain
Oswais recognize each other as caste-brethren.

paita (Telugu) Outer end of sari draped over breasts and


shoulder.
Paliyan A primitive tribe of southern India subsisting mainly
on hunting and food-collecting.
Pallan An exterior Tamil caste of agricultural labourers.
Palli An agricultural caste of southern India claiming
Kshatriya origin (= Vanniyan). They are spoken of
as painters by Carrd (op. cit., ii, p. 595), and in mo-
dem times have assumed the title Mudaiiyar, used by
the chief painters ' of the Malabar (Tamil) caste in

1680 (Madras Records, quoted in a footnote to above


by the Editors).
< Sanskrit parna leaf
*
pdtn the plant Piper betel which
is chewed with dried areca nut and lime as a masti-
catory. According to Sir Thomas Roe it bites in the

mouth, accords rheums, cooles the head, strengthens


the teeth
pUn bari An enclosed and shaded garden in which the creeper
Piper betel is grown in rows on sticks.

miiM See Malayan.


Fanchala A gtoup of five (punch) artisan castes goldsmiths,
braziers, carpenters, stonemasons, and blacksmiths in

southern India.
p&nckSy&t See above pp. 99 sq.
Pandaiaii Appertaining to Fandya.
PSadya An ancient kingdom of southern India %vhich had its
capital at Madura.
Panlkkan A castO' of barbers and, weavers in the Madura'
Tinnevelly of Madras..
districts

Faniyaa A hill tribe of Malabar and iheN'%iris.


PantS-rans See Mala-Fantlram.
m GLOSS ART
Panta Rcddi A subdivision of ESpu (qv.).

PanwSr One of die Agnikula {* fire-bom ') Rijput dans


(xsPramara, PonwSlr).
Parahiya A small hill tribe (or caste) in Minapur district of
Uttar Pradesh.
Paraiyaa An exterior Tamil caste of field labourers and village
servants whose caste name has given us the word
* pariah
Parasurlxna (i.c, RSma with the axe \)

A
Brahman hero of the
epic MShUbh&rata who cleared the earth of Kshatriyas
twenty-one times and gave it to the Brahmans.
Parava A caste of south Kanara which cures disease by
exorcising devils by dancing, and which also makes
baskets and umbrellas.

Parja A tribe with a fColarian language in Orissa; see also


Bondo.
Parwir (or Parwil) A mercantile caste of Rajputana.
PSsi A caste of toddy-drawers in northern India who tap
the tafi palm for liquor.

PathSn A Muslim tribe of the North-West Frontier. See p. 42.


Patni An exterior caste of fishermen and basket- makers in
eastern Bengal and in Assam, probably nearly allied
to BOm.
Peravadar A mentioned in early Tamil literature as living
caste
on the coast by, fishing; probably = Para van, a fishing,
pearl-diving, and sea-gomg community on the extreme
south-east coast of India.

Pirali (or PIr-Ali) A subcaste of BrShman in Bengal reputed to have


become segregated from the Rarhi Brahman caste as
a consequence of smelling a Muslim meal.
Prabhu The writer caste in the west of India; it claims a
Kshatriya origin and can show much in support of its
claim (see Enthoven, Tribes and Castes, m, p. ^35).
prdtiWmU Sec p. 55.
prUArH See p. 58.
pretyasilS See p. 248.
pUjS A Hindu religious ceremony; hence almwt any kind
of rite. See Hobson-Jobsm, s.v. Fooja*
Pulayan A caste of agrestic serfs in the north of Malabar dis-
trict, in Cochin, and in Travancore, corresponding to
the Tamil Paraiyan (q.v.).
PiirSda VatinSn A caste of Tamil washermen who wash the clothes of
exterior castes, and of whom the very sight is pollut-
ing.

purSnas Sacred books of the Hindus recording the pan-


theistic form of
this religion which followed the vedic
period; the earliest go back to at any rate the fourth
century b.c. and the famous epics, the Ramayana and
the Mahabharata, were succeeded by purdnas in the
form of a dialogue between an inquirer and an expo-
nent in verse, ancient and different
books have distinguishing names, e.g. Fadma^PurMna,
GLOSS ART 993
the purana which contains an account of the period
when the world was a golden lotus {padma); Agni-
Purana, the one which was communicated by the deity
of fire (Agni) to the vedic sage .Vasishtha. The
Bhavishya Purana, nominally a book of prophecies, is
actually more concerned with rites and ceremonies.

Rabha A tribe of Garo and Kachari affinities in Assam.^

Rajanya The aristocratic and ruling class of the Aryan invaders


of India circa 1500 B.c. from which the Kshatriyas
and Rajputs are derived.
Rajbansi A formed from the Kochh
caste (q.v.) tribe, in nor-
thern Bengal and Assam.
Rajgar A mason; a subcaste of Khatik with the occupation of
mason.
Rajput An widespread in western, northern,
aristocratic caste,
and central India, whose traditional functions are
fighting and ruling. They represent the ancient
Kshatriya varna, and rank next to the Brahmans
socially (see pp. 33, 54-5, 66). See also s.v. Agnikula.

Rarhi A Brahman caste of Bengal.


rath A wheeled car used in particular for carrying an idol
in procession.
Rathor A clan of Rajputs, q.v.
Ravulo A caste of temple servants in Orissa.
Reddi See Kapu.
Rengma A Naga tribe of Assam.
Rigveda The original compilation of vedic hymns to nature
deities. The
Yajurvedas, Black and White, and the
Samaveda were derived from it, and are chiefly con-
cerned with sacrificial ceremonial.
rishi An inspired poet or sage; in particular the inspired
persons to whom the hymns of the vedas were reveal-
ed, and the ancestors of those Brahmanical gotra in
the pr&vdrd of which they are named (see p. 59).

sabha An association, a council.

Sadgdp A caste of cultivators in Bengal (=Ch5s5).


Saka A northern people Scythians

who invaded India
about the second century b.c.
Sakaldwipi A caste of Brahmans in Bihar who arc reported to be
divided into a number of exogamous units known as
pur, marriage being allowed within the gotra but not
within the pur,
SSkta A worshipper of Sakti or the female energy of the wife
of the deity, the deity generally being Shiva,
sUkti See p. 189.
Sakuntaia The heroine of a play by Kllidasa written in the fifth
century a.d.
salUgramU A found in certain rivers, having mystic
pebble,
and commonly containing the fossil known as
virtues,
an ammonite. See Hobson-Jobson s.v. Saligram. It
is particularly associated with Vishnu.

10
m GLOSS ARY
SSmaveda See Rigveda.

Sansiya A vagrant criminal tribe of Rajpiitana and adjoining


provinces, dependent for a livelihood on theft and
robbery.
Santa! A Koiarian tribe of Chota Nagpur.,, Biluir and Bengal.
s&pindS, See page 6o, n. 59.
Sarak A small caste of cultivators and weavers in eastern
India; an archaic community originating from farther
west and originally Jain, biit now mostly Hinduired,
though some are Buddhists (see Risiey, Tribes and
Castes, XI, p. S36 ; Gait, Censtis of Bengal, p. 427).

Saraswat .A caste of Gaur Brahmans primarily in the Punjab


and Sind which take their name from the river Saras-
wati, formerly a tributary of ilu* Indus but now lost
in the deserts north of Rajpiuana. They are reported
to marry within the gdirk bur. to retain certain exo-
garaous clans.
sdstra A rule, or treatise. Any book recognized as having
divine authority, but particularly the ancient books
of Hindu law.'
sastri Person learned in the religious law.

Satapatha The Brahmana Hundred Paths, the most


of a
Brahmana important of the brdhmams, q.v.
salt Act of immolation of a living widow on her deceased
husbands funeral pyre.

Satnami (i) The name founded at the end of the


of a sect
seventeenth century by Jag-jiwan Das, a Rajput, and
taking its name from its invocation of the God of *

Truth *, Satyanmna. (ii) A sect with similar tenets


founded in the early nineteenth century by Rae Das
and virtually confined to the Chamar caste.
satyagraha
Non-violent resistance.
satyagraki Person practising non-violent resistance.

Sawara (Also spelled Savar, Sora, Saora, etc.) A Koiarian


tribe of Orissa (-also Sahariya, central India).
Serna A Naga tribe of Assam.
SSniyan A weaver caste of southern Iiuiia (see Thurston,
Castes and Txibes^ vi, pp. ;j()i sq.).
ShagirdpeshS Caste of domestic servants in Orissa.
Shaha A caste of distillers, liquor-sellers, and shoi>kceners in
Bengal (=ssSunri).
Shaiva A Hindu sect worshipping Shiva as the greatest god.
ShSnan A Tamil toddy-drawing caste. Corresponds to rfyan.
Shiah Vide infra s.v, Sunni.
smarta 5= Appertaining to the smriti,
i.e. to * what has been
.
remembered that traditional lore, thus in-
is, to
.
eluding at its widest the epics {MMkabhdmta and .

Ramdyana), purdnas-, and sutras, but in particular the


dharmasdstras, i.e. the law books or "Codes, particularly
the Code of Manu, and the works of other sages who
recorded the * recollections of what they had received
*

from a divine source. A Smarta Brahman is a student


of the smritis, and the term is particularly applied to
GLOSSARY m
the followers of Sankaracharya, a Hindu philosopher
of the ninth century a.d., prominent in the extirpa-
tion of Buddhism. Most Smarta Brahmans are
Shaivas.
Sombatta Rope-maker; a subcaste of Khatik.
Sonar Goldsmith caste, distributed all over India except in
the south.
Soni Goldsmith caste of western India see Enthoven,
Tribes and Castes of Bombay. Distinct from Sonar.
sradh Obsequies and sacrifices performed for the manes of
deceased ancestors.
Sudra The varna of indigenous castes not entitled to the
initiationceremony of rebirth (see pp. 64-7).
Sunni, Shi ah These are the two principal Muslim sects. The Sunni
accept the authority of all the successors of Muham-
mad, whereas the Shiah regaid the fiist three Abu

Bakr, Omar, and Osman as interlopers, and Ali, the
Prophets son-in-law, as the first true Khalifa.
Sunri A caste of liquor-sellers and distillers in Bengal
(=Shaha).
Sut A cultivating caste of Assam; said to have a pratiloma
descent from a Brahman mother and a low-caste
father.

Suta An ancient caste of charioteers and bards, no longer


existing; also pratiloma (see p. 151).

Sut5r A carpenter caste; found all over India except the


south.
Sutradhar The carpenter caste of Bengal ( < Sanskrit sutra-2L
thread, because a thread is used to mark straight lines
to work by).

sUtras Religious books of Hinduism consisting of concise


treatises which condense for practical purposes the
earlier hrdhmanas, which are theological treatises ex-
plaining the vedas. The sutras were produced between
about 500 B.c. and the beginning of the Christian era.

Tags A land-holding caste of Rohilkhand, probably of


Brahman extraction, and having Muslim as well as
Hindu branches (sec Crooke, Tribes and Castes, s.v.
Taga).
talikettu The ceremony of the tying on of the tali, or neck
ornament emblematic of the married state, by the
bridegroom, on to the bride, or nubile girl, in sou-
thern India. (In many castes the tali must be tied
before the girl reaches puberty, and often the tali-
tier is not the person who will ultimately be the girls
husband.)
tantras The name given to a numerous class of religious and
magical works, generally of later date than the
puranas, and devoted in particular to the ciilt of the
sdktis (see Sakta); rule , ritual.

Tawi-if A dancing girls and prostitutes both Hindu


caste of
and Muslim; the men of the caste, who obtain wives
from outside it, act as pimps and musicians (see
Crooke, Tribes and Castes, iv,*p. 364).
296 GlOSSARf
Telaga A cultivating and military caste in Telingana*

Teli A caste of oil-seed and oil merchants in


crushers
upper and eastern India (see
pp* 75, 89).
Telingana The Andhra Desha, or country inhabited by speakers
of the Telugu tongue*
Thado A Kuki tribe in Assam.
thakurghar Place of worship (Hindu).
Than tap ulayan A primitive tribe of southern India.
Thathera Caste of braziers in upper India.
Tili A caste of oil merchants in Bengal (
= Tli, but see
p. 89).
Tirhut Brahman A synonym for Maithil Brahman, q.v.
Tiyan A toddy-drawing caste of tire Malabar coast, matri-
lineal in the north of Malabar district, patrilineal
in the south of it; corresponds to ShSnSin, and
Izhavan, qq-v,
Toda A matrilineal polyandrous tribe of the Nilgiri Hills;
buffalo herdsmen and dairymen.

.Turi A caste of cultivators, bamboo-workers, and basket-


makers, originally derived from a forest tribe of
Kolarian affinities in Chota Nagpur.

Ulladan A forest tribe of hunters and food-collectors in Tra-


vancore.
Up&nishads Treatises forming the concluding portions of the
veda, and called therefore tlie Vedanta, on which is
based much of the later Indian philosophy.
Uppara A Telugu caste of salt and saltpetre workers; also
well and tank diggers and earth workers generally.
Uppara A Kanarese caste of salt workers corresponding to the
Telugu Uppara.
Uppiliyan A caste of salt workers in Malabar; corresponds to
Uppara.
Urali A caste of agricultural labourers in Trichinopoly and
Madura districts of southern India* There is a hill
tribe of the same name in Travancore, According to
Thurston (Castes and Tribes vn, p. S4t) urali = * ruler
of a village

Vaishnavi Hindu sect worshipping Vishnu instead of Shiva as


the greatest god.

Vaishya The third varna of the Indo- European invaders of


the Rigvedic age; Vaishya, < msh^rnt people, repre-
sented the third estate of the invading community, i.e.
farmers and perhaps traders, as distinct from the
rulers and warriors (Eshatriya), and priests (BrSh-
mana) on the one hand, and the Sudra or servile
indigenes on the other. The term is now generally
used for the superior mercantile castes associated in
particular with Rajputana and the rest of upper India
(see pp. 65 sqq.).

Vania =:Baniy&, q.v.


GLOSSARY m
VSniyan A Tamil caste of oil-presscrs, corresponding to Teli
in northern India.
VannSn A Tamil caste of washermen, corresponding to Dhdbi
(q.v.) in nortliern India.

Vanniyan See Palli.


varna See pp. 65 sqq.
Vedan A Tamil labouring and hunting caste, formerly
soldiers, corresponding to Bedar.
vedas Books of Divine Knowledge forming the earliest
tvritten records of Hinduism; see Rigveda and
Atharvaveda.
Velan Apparently an abbreviation of Vellala and used as a
title by several castes in southern India.

Vellala A Tamil caste of cultivators, very strong numerically.


Vettuvan A Tamil agricultural and hunting caste.

Wa A primitive tribe of the Shan States in eastern Burma.

Yadava Anciently a tiibe of Rajputs. The term is now used


collectively by cattle-grazing castes to cover a number
of groups so employed; it has been chosen perhaps
on account of some resemblance to the word Idaiyan
and its use implies a claim to Rajput extraction.
Yajurveda See Rig\eda.
y&ti Holy men and wandering mendicants, some of whom
were destroyed by Indra who gave them away to be
devoured by wolves (saldvrikas) (see Chanda, Memoirs
of the Archaeological Survey of India, no. 41, pp. 32,
33)-
Yeruva A primitive tribe or low caste of Coorg.
IN DEX
N.B.
^Tlie numbers appearing in brackets after an authors
name refer to different works by him in the order in which
they appear under his name in the Bibliography, the reference
to which precedes them. Where one work only has been used,
the bibliographical reference is given in its paginal order.

Abarimon, 183 animal dancing and fertility, 262


Abbott, J., 189, if
63 Annamites, 239
Abor, ^7, 183 Ansari, Gh., xi, 263
aboriginal tribes, 3; in Assam, 57, Anthropological Survey, 7
223; in central India, 22, 23; in
Anti-Priestcraft Association ,
234
Orissa, 25; in southern India, 8, 9 anuloma, 55, 149, 150, 151
Achan, P. A., 14, 263 Ao, see Tribes, Naga
Achela Muni, i57n. Apastamba, 118
acrobatic displays, origin of, 262 Arab, descent, 17; invasion, 6, 39;
Aden, 18 Rind, 40; traders, 13, 15
Adya (goddess), 154 Arabia, 19
Afghan, 42 43, see also Pathans Arakanese, 207
Afghanistan, 38, 42 aren, 184, 185
Africa, functional groups in, 142, 143; Armenia, 5
head-hunters in, 24on.; plantain Armenoid, 223
tree in, 257 Artemis, isan., 163
Agastya, muni, 15711; gotra, 56, 59 Arthasastra, i26n.
Aghoripanth philosophy, 244 Aryan, and cattle sacrifice, 228; and
Agnikula, 157, 276 Hinduism, 224; immi^ants, 188;
Agra, 34, 56 invasion, 177-80; language, 6; Rig-
agriculture, 171; of Badaga, 16; and vedic, 13, 153, 155, 157, 158, j8o;
head-hunting, 28, 232, 241; and speakers (Aryans), 5
human sacrifice, 241; shifting culti- Asia, central, 4, 6
vation, 9, 22, 28; and soul-matter Asia Minor, 5, 223; cultural affinity
and water, 236 sqq.; terracing, 28 with southern India, 152; religious
agua, 109 impulse from, 231; sacred prostitu-
Aheria festival, 262 tion in, 163, 164, 165; trade routes
Ahmedabad (Ahmadabad), 18 from, 230
Aitareya Bralimana, 251 Asia, south-east, 191, 223
Aiyangar, M. S., 68, 78, 263 asram, 244, 276
Aiyappan, A., So, 81, 83, 263 Assam, 3, 4, 6, 27, 30, 63, 77, 82, 86,
Akali, see Sikhs 113, 116, 160, 223; agriculture in,
.Algeria, 174 n.5 28; caste in Assam valley, 211-15;
Ali, 39 exterior castes in, 197, 198, 211, 212,
amputation, of finger joint, 11, 291 213, 214, 215; Caucasian stock in
ancestor worship, 28, 226; and re- hills, 223; cult of dead in, 235;
incarnation, 252 cultures in, 28, 29; disposal of dead
Andaman Islands, 2, 8, 223 in, 245, 246; foundation sacrifice in,

Andamanese, 44 8, 249; geography of, 27; head-hunting


Andhra country (Andhradesa), 5, 10, in, 28, 239; immigrants into, 30;
11, iBt 11 life-essence in, 262; magic in, 258;
goo INDEX
Assam (cont,) Baluchi, 37, 39, 411

Mongolian element in, 57, 2gi and Baluchistan, 24, 40; nomadism in, 41;

Orissa, schools in, 215; soul- exterior castes in, 219


27;
matter, 232, 245; use of stone in, Ba- Nando, X39

245-8
Banerjee, A. R., 7
Abor, 27, 183, Banerji-Sastri, A., 152, 264
tribes in, 3, 28, 183;
Ahom, 27; Apa Tani, 28; barbers, 72, 187
276;
Dafia, 27; Garo, 29, 30, 230, 235; Barbosa, Duarte. 78, 160, 264
Kachari, 30, 114, 236; Khasi, 29, barga, 57

246; Kuki, 28, 145, 258; Loi, 49,


Baroda State, business widi Spain, 18
Liishei, 242, 253, 287; Bartlett, F. G., 130, 264
139;
Meithei, 116. 236, 289; Mikir, 253; Barton, W-, 43, 264
Mishmi, 27; Naga, 3, 28, 63, 86, Baruch, Book of, 165
159, 183, 207, 230, 232, 235, 240, basmnj 162, 165, 1167; see also Dcva-

242, 246-8, 252, 254, 255, 256, 257, dasis

258; Shan, 27; Thado, 146, 242 basil, sweet, 105

Assamese, 30, 109, 183 Bastar State, 23


Asura, 153, 188, 276 Batak, 186, 243
Asvarnedha, 230, 276 Ba-i6v!, 139
Atharvaveda, 154, 156, 228, 231, 276 Baudhyana, 149
Australian marriage class system, 22, beduart^ 254

63 Bellary, 10; dm.mlasis in, 162


Australoid (race), 3, 16, 31; and Benares, pandits of, 33
Austro-Asiatic cultures, 262; and Bengal, 5. 13, 18, 34, 27, 30, 31, 3*,

Yeruvas, 16; see also Prcto- Austra- 49- 5 *> 53. 61, n6; brachycephaly
loid in, 33; Brahmans in, 72; child
Austro-Asiatic speaking people 223; marriage in, 159; Education Bill in,
and Australoid cultures, 262 202; joking relationship in, 63;
Austric (language family), 4 King Balia! Sen of, 94; Mongolian
avarm, 193 in, 32; Muslim cultivators in, 31;
Avesta, 155 Muslims and Hindus in, 31; and
Awadh (Oudb), 33, 34 Sylhet, 215; wells in, 82
Bengal, Bay of, 4, 28, 29
Bengali, the, $it literature, 31;
Babylonia, 153; sacred prostitution in, Muslim# 31
261 Bentham, Jeremy, 107
bachelors' hall (dormitory), 23, 24 Bcrar, 17
'
bad death 146, 239 Bernier, F*, r66, 264
Badi-ul-Jamal, 151 Best, Eldon, quoted, 1S6
Baganda, 256, 257 Bhagavad Gita, 124, 189
Baiga, 22, 23, 62; physical similarity bhaiband, 99
with Oond. 23 Bhatrava, 11
Bailey, F. G., xi, 264 Bhattacharjee, F 7, 273
Baines, A.* 35, 264 Bhattacharya, J, N., 49, 54, 56, 122,
Balkh; 155 264
Baioch, 39, 40, 41; see also Baluchi Bhatti (sept), 36
Balochi (language), 41 Bhaviijhya Purana, 227
Baltis, 43 BhawSni, goddess Thugs, 127
INDEX 301
Bhil, 19, sti, stz, 35, 39, 55 classification of, 75; brachycephaly
Bhili, XI in, 30; castes and gotras of, 55 sq.;
Bhrigu, 56, 59, 153 and caste origin, 170; and caste
Bhutan, 43 system, 49, 175 sq.; aiid
125, 126,
Bhutanese, 37 cattle sacrifice, 22S; and birth pollu-
Bibi N5ni, 151 tion, 83, 84; colour and varna, 65,
Biddulph, J., 43, J23ifn., ^64 66; and distance pollution, 78,
79;
Bihar, xi, 33, 34, 5, 55, 57, 66, 7a, diversity in composition of caste of,
107 33 65, 67; divisions of, 48, 56; and
Bikaner, ruler of, 36 Dravidian -speaking civilization, 153;
Bikrampur pargana, 31 duties of, 149; and exterior castes,
bina (marriage), i6t 195 immigrants, 136, 151; in-
biradari, 99, 108 fluence on caste, 49, 92; Kulinism
birth, symbolic, 108 of, 53;and Kuricchan, 78; language,
blacksmiths, 142 87; and marriage, 58, 62, 154;
blood feuds, 42 migration, 116; mourning, 84; Mu-
bloody, significance of adjective, lagn. ghal conquest and, 95; occupations,
blow-gun, 9 33; pollution, 71-84, 126, 180,
Blunt, E. A. H,, 264; (1) 161, 171 sq. 184-9; position in caste, 49; power
(origin of caste), 109; (2) 52, 55, 71, of, 92; and purification, 106-8;
72, 7411., 75 (classification of caste), racial composition, 33; and Rajanya
77, 78, 89, 90, 96n 98, 99, 101, warriors, 65, 156; Right-hand and
102, 104, 105, 106, 107, 108, 121, Left-hand castes and, 67, 69; in
ixx Rigvedic age, 35, 65; post-Rig-
Bombay, 18 sq., 202, 203; census and vedic, 156; and rishi,^ 59; and
cross-cousin marriage, 62 sattva guna, 182; and secular
Bondo, 25 rulers, 93, 94; and Shins, 232; and
bones and fertility, 242 Sudra, 93; and varna, 48, 65, 66,
Bonington, C, J., 235, 246, 264 149
Bonnerjea, B., 178, 264 Castes, etc.: Ayyar, 62; Bishnoi,
boomerang, 1520. 253 ^78; Chaubg, 33;
Bor, N. L., 106, 264 Deshasth, 62; Gayawal, 33;
Bosci P. N., quoted, 123 Kachela, 119; Mahabrahman, 79,
Bose, S. C., quoted, 96 244, 287; Maithil, 94, 2540., 287;
Bougie, C., 113, 116, 117, 121, 122, Modh, 119; Nagar, 73, 290;
123, 124, 264 *
Nai *, 113; Nambudri, 14, 15,
bow, 9; and arrow, symbolic, 16 i6, 79, 80, 94, 109, 154, 158, 222;
brachycephaly (or broad-headedness), Nanagotri, 154; Panda, 22*7; Pir-
5, 15; in Baluchistan, 41; in Bengal, Ali, 31, 100, 292; Rarhi, 53, 54,
30; in Goorg, 16; in high and low 244, 293; Sakaldwipi, 56, 293;
castes, 30; in Konkan, 16 Saraswat, 52, 36, 294; Tamil, Telu-
Bradley-Birt, F. B., 24, 264 gu, 11; Tirhut, 73; Viswakarma %
BrahmS, 186; crevice of, 251 113; Vyasokta, 65
brahma, 231, 232 Brahmanical code, 190; and matrili-
Brahman, 11, 20, 31, 47, 52, 55, 56, ncal system, 154; scriptures and
113; 122, 153, 167, 182, 186, 187, development of caste, 49
j88, 189, 221, 227, 228, 229, 236, Brahmaputra River, 27, 28, 29, 3#
244, 256, 277J in Assam, 27; Blunts Brahmarshidesha, 224, 229, 278
:

50s
INDEX
Brahui, 4, S 9 ^19 Cape Comorin, 10, 13
Carloss, Lt, 151
brass, vessels of, 1401 soul-figures, ^35,
Carnatic, the, 10
2SQ
Bray, D., 4 > ^ 39
Carr^, TAbb^, 70, 265, 291

bride price and hypergamy, 54 n* Casaiibon, I., 24011.

bridges and human sacrifice, ^49 sq. Caste


Briffault, R., sgn., S64 aboriginal tribes and, 197
Brough, xvi, 59, 60, 61, ^64 acceptance into, 49
J.,
*
Brown Race % 3 analogous institutions, 46, 133; m
Buchanan, F., S64; (1) 68, 78, 950., Africa, iscj, 143, 143; In Burma,
(s,) 582, ago 143-7, ^07; in Ceylon, 136, 137,
98, 167;
Buddhism, 146, 148, 183, srsy, in 138; in Egypt, 140-3; in Fiji, 138;

Bihar, 33; in Burma, 144; in in Japan, 14.7-8; in Madagascar,


Japan, 147 sq,; Hindu reaction 137; in medieval Europe, 139 sq.;
against, 36; in Lahaul and Spiti, 44 in the Roman Empire, 134, 135;
Buddhist, $s, 870., 148; temples in in America, 136, 1740,
Ceylon, 137 caste control: councils, 8, 99*
buffalo, of Toda, 16 organisation of, 99, iix>, loi, 103;

bugs, evacuation of, ^58 procedure of, 105


Biihler, G., 564^ ^*65; (i) 149, ^51; (2) Caste Ciitclierry, 97
9a 94 118, 135, 9 H Christians and, 2, 47, 121, 304 sq.
building, caste restrictions in, 87 clothing and, 70* 85, 86, 205, 306
bull, cult of, 12, 153, 334, 338 definition of, 47-50; of Butt, 49; of
bull-baiting, 153 Ketkar, 48; of Risley, 47; of
Bundi, 36 Senart, 49; (suggested) 5,0

Burghers, of Ceylon, 136 dharmadhikari and, 98


Burma, 3, 37; Buddhism in, 143; Dhartm Sabha and, 97
grave-diggers in, 145, 307; head- disadvantages of, 133, 124
hunting in, 241; pagoda slaves and division of labour and, 134, 371

taboo in, 144, 307; seven outcaste divisions of, and subcaste, definition

classes in, 144, 145; system analo- of, 48; marriage between, 48-54
*
gous to caste in, 143, 145, 146, 207
exterior castes (or untouchabies*

Burnell, A, C., see Bobson-Jobson or *


depressed castes *)

Burton, R, F., 40, 365 amelioration of, ix, ,x, .208, 3'09,

Byblos, 166, 361 310


Co-operative Societies, 309
definition of, 191, 194, 195, 196,
Cachar, 316, 317 197; in Assam, 211-18
Calcutta, 33 disabilities, educational, social and
Cambay, 18 religious, of: 193-333
Cambridge University Press, xiii distriWtion of, 199
Gameroons, 3570. economic position of, 306, 307
Campbell, Sir J 159, 337, 365 H.indui8m of, 304
Canada, 134; half-breeds in, 136 numbers of, 130, 1.98,
'

.199; in

Canara 10 Madras, 319
cannibalism, 187, 339, 343, 343, 344; outcastes and, in Burma, 144-7,
ceremonial, 243; funerary, 35, 343 307; in Japan, 147, 14S
INDEX 303
Caste {conU) 173; Slater, 177; Westermarck,
pollution by, ^V-83; through 174
water, 71, 75?, ^11, aist; of wells, ornaments and, 85 sq., 305, ao6
196, oi Pali literature and, 179
famine and, 186 panchayat, 90, 98-109, x7a; head-
fluidity of, 50, laa, las; fissipaxous men of, 101, xoa; in low and
nature of, 50, 51; isolating high castes, loa; procedure, 105,
tendency of, 116, 117, 118 xo6; punishment by, X06-9
functional, lai, 170-7; villages, 138, pollution, 71-88; by childbirth and
3139, 184 by death, 83, 84, 85; court house
functions of: 111 sqq.; economic, 89, and, 8; distance, 79-83, 195 sq.,
xsij genetic function, 131, 135; a:?o;food and, 184-9, ^^75 gtave-
as a guild. 114; and the indivi^ digging and, 507; and menstrua-
dual, IIS, list; and medieval tion, 83, lag; Muslims and
guilds, 140, 171, 180; and plural Christians, 8a; relaxation of, a34;
society, 1x5, 1x9, 1^7, ia8, 133, rock, 80; by sight, 80, 81, lai,
iS4 2is a social unit, 1, a; ia6, aoo sq.; temples and, 8a, 83,
as a stabilizer, 44, 45, lao, lai; xa8 sq.; untouchability, 77-83, xai,
and State, 115 130, 180, 195, 196, ai8, aig,
genetics and, 131 -a aao, a34; from wells, 81 sq., 193,
geographical aspect and, 178, 190 195, 196, aoi
goira, 47, 55-6a, 17a, 173 position of women and, lag
inheritance and, 88 purification after pollution, 78, 85,
48, 64
jati, 107, 108, 109, 176, 180, a 14; by
magic and, 177 panchgavya, 88, 108
mark, asa sq. race and caste: and colour pre-
marriage and, 47-70, 85, i58-6a; judice, i7a-5; conquest and
endogamy, 47, 5a; hypergamy, colour bar, and racial inbreeding,
53-5, 173; matrilineal system, 13, i35> racial taboo, ao7
14, 70, 150-63, 168 sq., a5 religion and, 114, lai sq.; and taboo,
occupation and, 33, 51, 11 1> 116; mana and soul-stuff, 177, i8a-7,
groups, X16, 179; villages, 139, 189
184, 185 restrictions: building, 87; commen-
omnivalencc of, 90, 9x sality, 184; on dress, etc., 85, ao5

origin of, 46, 190; traditional, 149 sq.; on food, etc,, 7a-8; on
sqq.; mrna and, 48, 64-7, 149, language, 86; on marriage, 47-64;
171, I7 i 3 ; theories of: Blunt, on occupation, 87-90, 170-a; re-
171, 17a; Bonnerjea, 178; Chanda, laxation from, a6, on the use
17a; Colebrooke, 180; Bahlmann, of roads, 195, aoo, aoi; on schools,

171; Dubois, 170; Butt, 175, 177; 195, 196, aoi, aoa; on smoking,
Ghurye, 175, 178; Gilbert, 178; 76 sq.;, on spittings 81, aoo; on
Hayavadana Kao, 175; Hocart, temple entry, 8a, i98-ao4; on

176; Ibbetson, 175 sq.; Jackson, water, 71, 7a, 8a, 196, aoi, an
iBo; Ketkar, 180; Linton, 181; sqq.
Murphy, 180; Nesfield, 176; Rigvedic invaders and, 180, 188, 189
Oldenberg, 179; Rice, 178; Risley, ritual and, 176, 177
173: Roy, 175, 180; Senart, 17a, sapinda, 60, 6i
m INBSX
Caste (cont.) Ben-i -Israel, 19
secular control, 93-7, u8, 181; and Beri, 68
religious jurisdiction, 97, 98 Bhangt, 34, ao8, A77
Sikhs and, 117 sq., 204 Bhar, 34, 173?, 577
social status and, 139, 149, 150, 151; Bliarbunja, 7s, 75, 577
of castes and subcastes, 51 sqq,; Bhat, 3 S. 34, 37, 577
and education, 113, 114; and Bhatia, 18, 39, 57S
movement between, 49, 64, 65; Bhoksa, io, 578
and new castes, 90, n6, i6an,; Bhuinmali, si5, 517, sx8, ^78
and occupation, 34, 51, ii; social Bind, x6i, 578
positionand food taboo, 75-8 Bohra, x8, 578
system, 6, 44, 45; and culture
1, Boria = Sut, q.v.
patterns, lu, 114, 1x5; defined, Brahman, castes and subcastes,
50; development of, 189, 190, 191 main index s.v. Brahman
Theodosian code and, 134, 135 Brittial Baniya, 113, ,13, S14, ai^
totemism and,
7759., i57n., 178 Chakkiliyan, 68, 69, 86,
379
travelling and, 88
Chakli, S08, ^79
tribes and, ii6 Chaliyan, 68,
Castes Chamar, 33. 34, j0, 70, yg_ gg_
Agariya, 76, 116, ^76
203, mi, ..208, 259, 279
Agarwala, 18, 276
35, 77, Chamargaur, 105, 279
Aghoripanchi, 244, ^76
Chandal (or Chandala), 30, 86, 112,
Ahar, 50, 376
136, 145, X^X* rSo^ 2'ygj
34, 38, 50, 60, 1x3, *45, 55:?, Namasudra
^76 Charal (or Namasudra), 214
Ambalavasi, 48, 83, 576
CbSran, 34, 35, 181, 279
Ambattan, 68, 187, ^76
subcastes: Gujar Kachela, , 1x9
Ande Koraga, 81, aoo, 385
Chasa, 48, 49, 256,
Arora, 38, 376 279
Chattarkhai, 186, 279
Ayar, 179
Chaubf, 33, 279
B 5 bhan (or Bhuinhar),
33, 56* *77 Chero, x6x, 279
Badaga, ifi, 116
Cheniman, 78, 79, 80, 83, 279
Badi Banjara % 90
Chetti, 12, 57, 68, 28a
Bagdi, 31, 87, 577
subcaste: Let, 87
Chitpavan, 20, 2S0
Baidya, 30, 49, 157, Chodbra, 21, 155, 204, 280
J177
Baiti, ^77
*
Chogan *, 86 and see Izhavan
Balija, Chuhra, 38, 39, xx6,
^{77 204, 219, 2S0
Banajiga, 68, j?77 Barzi, 280
Baniya, ^0, 35, 189, :^x3n,, 5x77
subcaste: Gujar, 119
subcastes: Gujar, 1x9; Modh, 119 Beori, 2x3, 280
Banjara or Lambadi, x, 90. JJ77
Beshasth, 62, 2S0
Baoti, :?77 Bevanga, 68, 280
Barhai, ^77 Bhangar, 19, 20, 280
BarUi. 87, 188 Bhanuk, 52, 280
Bauri, 31, 640,,
*77 subcastcs: Badhik, Hazarl,
latha^
Beda, 254
riya, Laungbawa, Taihal,' 52
Bedar, 68, ir77
Bhlrhi, 121, 2^
BEX
Castf'i (rtwif.) jalna, 214 , 3 %
DIud, S 4 W, tHD Jaoappaiii, 68, 283
Idi|ai, Siiiali U' J**C 17 , 38, 39, tyt
Dhaht (ur niuipi), :I4 ity, 136, till. JiKib, tiy, 283
21 S, Jtilaha, iti, f%
l)U\ii [M Du:;la). ti^. *uH. si Jiigi Naih), S15, 217, tt8, 283
(or

Dagta, t*"'! Kabirpanthi, u8, 283


Doin (Of Domtiri). 34, 3f*. 77, 7S, .Kacbari, 30, 114, 236

n:i, uB, lijrt,, ti|. Kacbhi, 75, 28$


Diinuil* 57, 34^1 Kahar, 72, 75, 76, 90, 283
Ernaclain *%Si sabcasies: Dhiniar, Mahat, Kam-
exierior c;wcej, src? Appetli?c A, kar, 90, i6tn.

pp, iptsipp; fir puMiitiii Etkr Kaibartta, |i, 51, 52, 113, 2,14., 283;
caste s)su*iii hft undrr Caste see alsG Mahisbya
Fernigjtec, 5 K Kalkolaii;, %, 167, 283
Ciitlili, Kalita, 27, 284
Ciayawal, 5:^* sHi Kalhui.., it, 127, 152, tor,, go6 284,
Glutiiiiii. ii*;p iHs, Hi tSS
suhrasicv; Moilli, Huraii, up; Kalta, t6, 101, 284
Aliiiiailah.itii aiitl Chaiypatieri. Kalwar, 34, 55, 76
lii Kami, 252, 284
Ghasiya tSi, fBi Kaaikar, 90
Goala. i. 50* 57, 61 73 it|, til Kamma, 10, 284
subcasies: Satmylia. 57. Bi; Nao- Kammilan (or Panchaia), 12, 68,

niulia, 61 tQt
Golapurab, 56, tSt Kamma van, 63, 284
Golla, 50, 0S, tSi Kanbi, see Kunbi
Gopa, 50, 113, til Kanet, 43, 44, 255
Goriya, i6o. Kaniyan, 80, 284
Glial Nat, 90 Kanjar, 161, 162, 284
Gujar, 38, mh ui. sBt; .as siil> Kansara, 181, 284
castt\ 119 siibcastc: Champaneri, 181
llaliya Kaihartta, 5. 115* soS 'sSt Kapu (Reddi), 10, ii

llsilvakki Vakkal, tilt, tSsr Karalar, 179


Halwai, 7, 74, 75, 8 Karan, 48, 49, 190
llandi Jogi, 346, tSt Kascra, 89, 284

I lari, 115, i5, t8* Katoni, 213, 210, 284


Kayasth, 27, 31, 49, 65, 73, 76, 84,
llasala, *35, aSt
89, 113, i72.' igcs
Hira* I4, tSt
Holeya, 68, tSS, s8t Kewat, 52, 77, 285
Iciaiyan, 68, tSt Khandait, 26, 285
50, 179,
Iriila,
Khangar, 57, 285
f>a,'s8s
Iluvan, Irovar, subcastes: Dahat, Karbal, Mirdha,
Izhavan (Ilavaa,
Iravan, Izhiivan), 15, 80, 8t, 86, Rai, 57

100, 101, too, 583 Kbarva, 18, 285


197,
Jaiswal, 318
Kharwar, 161, 285

Jaliya Kaibama, ,51, :


p, U3
Khas, 136, 285
Kbatik, 34* 5^> ^01, 285
t8s
5^8 iMBBK
Cawnporc, 5 Cochin State, 8, 9, 10. 14, 160; control.

Central Asia, 58 in. 94; schools in, loi


Central Provinces, 17, H* Cole. B. L., 265; (i) 37; (*) 239
Colebrcx>ke, H, T., 6511., 180, 265
log, 197, m* ^43
Ceylon, u, 15, 1S7, 14* Coleman, George, his judgement at
Ciiakravartti, M. R. 7, *73 Chinglepui quoted, 68
Chamba, *47 colour bar and caste, 136, 174
Cbanda, R., 265, 297; (i) X 7 *J () conception, beliefs m to, 256, 257
29 SO Cornford, F, M., 141, 265
chapati, 73 Congo, 142
Chattisghar plateau, 2$ Conjeeverara, ti
Chatushringi, 165 conquest, effect of, nq, 120
child marriage and hypergamy, 159 Conti, N., quoted, 74
childbirth and pollution, 83 co-operative societies, 209
China, 30, 160, 166 C<K>rg. 5, 15, 16, 2S5
Chindu dance, u, 70 Coromandel coast, 10
Chittagong Hill Tracts, 30, $2 cotton spinning, 18

Chota Nagpur, 24; immigrants to court lumses anti pollution, 82


Assam, 30; racial affinity to Central cow, Brahman and Egyptian attitude
India, 25 sq. to, 140; sanctity of, 228, 232

Christianity: conversion to, in Assam, cremation, ste dead, disposal of


30; and exterior castes, 204 $q, Crete, 5, 153, 225, 228
Christians: in castes, 2, 14, isti; and criminal tribes, 12, 104. 127, 197 md
census, 194; and taboos, 185; tem- see Nat, Kan jar
ples and, 200 Crooke, W., 265; (i) 241, 242, 243,
Chukchi, 261 *44, *48. *5*: (*) S 5 . S 7 . 56. 66, u8,
circumcision, 41 *76, *78, *83, *86, *89, *90, *95; (3)
Clans 14. 74. 85. *8*! (4) 33. 34. 37, *53;
Buddhist in Sind, 39 (5) 55
exogamous, 54 55
Gond, 23
gotra and, 55 Dacca, 31
MahSr, 39 Dahlmann., J., 171, 173, 265
Naga: Daitya, 153, 280
Awomi of Serna, 235 Dalton, E. T., 25, i8|n*,
dog of Aos, 255 262, *65
hornbill, Wozakumr, 256 Dames, M. L., 41, 265
Thevoma of Angami, 253 Daniodar River, 256
Rajput; DSnavas, 153, 229
Chandel, 105 dancing, of Chindu, it, 70; curing
Chauhan, Kachwaha, Parihar, sickness by, 15: and fertility, *61 sq.
Ponwar (Panwar), Rathor, Ses- dao, 96. 280
odiya, Solanki, Tonwar, Yadava Dardistan, 43, 232
(or Jadu), 36 Dasehra festival, 101, 165, 241
Jethwa, 255 Dasyu, 155, 280
Kachwaha, 105 Davids, Rhys, quoted, 117
Rajput in Sind, 39 Davis, Allison, quoted. 17411.
totemic, 11, 20, 21, 255399- Dawson, W. R., 244, 265
INDEX S09
Day, F., 8o, 86, 87, 265 dolmens, 13; see also stone'
dead,
bad death 146, 539; 'caste Dowson, J., 640., 67, 266
differences in disposal of, ist, 168, Dravidian, see Languages
170; pollution from, 84, 85 Drury, R., see Oliver
cults connected with; megaliths, Dubd, S. C., xi, 266
monoliths, and carved stones, 13, Dubois, J. A., 83, 85, 86, 88, 126,
5, 28, 535, 545, a46-9; wooden 125 sq., 127, 164, 170, 252, 266
memorial figures, 43, 33a, 235, Dulaure, J. A., 233, 2470., 26on., 266
237, 238, 245, 246; see also stone Dumont, L., xi, 266
disposal of: burial, 12, 226; burial Dunbar, G. D.-S.-, 243, 266
erect in snow, 43; cremation, 25, Duncan, J., 79, 266
226, 227, 236, 244; embalming, Durkheim, xii, 124
246; smoke-drying, 246; eating,
Dutt, N. K., 49, i26n., 175, 177, 366
24 sq., 243
disposal of by; Andamanese, 8; Bir-
hor, 25; Kadar, 8; Kammalan, 12;
economic position and exterior castes,
Kuruba, 12; Malapantaram, 8;
206 sqq., 221
Munda, 24; Newar, 240; Parsi, i8;
Edgeworth, M. P., 243, 266
Persians and Rigvedic Aryans,
15X1 i77i caste system in, 140-2;
226, 227; Red and Black Kafirs,
cow, sanctity of in, 228; cult of the
43 238, 239, 243; Right-hand and
bull in, 153
Left-hand castes, 168; Sawara, 25
Ehrenfels, O. R., 35, 157, ss66
Deccan, 17
Ekavira, 165
dedication of girls and boys to tem-
and matrilineal system, Ellis, R. H., 7on., 86, 266
ples, 162-6;
Ellis, W., 137, 266
165 sq.
depressed castes, see Caste, exterior Elwin, V., 266; quoted, 115, 235n.,
Desai, G., quoted, iSgn. 5430.; (1) 22, 63, 104; (2) 257; (3)
Desai, G. H., 20, 265 mr M) H
devadasis, 162-7, 224, 225, 261, 280 emigration, see migration
Devi Mai (or Mata), 1520. Emmet, Ven. P. B., quoted, 69
dhdrdmsdld, 201, 280 endogamy, and sex ratio, 131 sq.; in
dh&rmd, 124, 190, 280 suocastes, 51; suggested origin of,
dhdrmddhikSri, 98 178; see also Caste, marriage and
Dharma Sab ha, 97 Enthoven, R. . 244, 266; (1) 19, 20,
Dharmashastra, 230 26, 37, 118, 278, 286, 292, 295;'
'

Bharraasutra, 229 117


()
Dharmeswar, 125, 190 Ephthahtes, 39 ,
: 1

Dharwar, 10, 282 ,


^

Eta, outcastes of Japan, 147


Bib S., 135, 266 eunuchs, 165 ^

Diodorus, 141, 252, 266


Euphorbia, 24.6
division of labour and caste, 124,
Eurasians, 136
170 sq. ,
.

Evans, F, B., see Innes


divorce, symbolism in, 108
Evans, 1 . FI. N., 187, 266
dog, 8; descent from, 2550.; medicinal
"

exogamy, and gotra, 54 sqq.; of Rajput


use of, 255'
long-headedness), clans, 54, 55; and the sept, 58; and
dolichocephaly
sex ratio, 131; and totemism,
15, 16, 5 L 33; in Baluchistan, 41 255;

m :
$10 INDEX
exogamy (conu) Furnivall, S
J. 115,
in tribal India, 57; sec also Caste, iS4> 67
marriage and Ffirer-Haimendorf, Ch. von, xi, ,^67
exterior castes, see Caste, exterior Fustel de Coiilanges, N. B.,
u6, *67

famine and caste, 181, i86 Gahlot Rajputs (Chamars), 1x3


female deities, 151-4, 163 sqq., sas Gaidaur festival, 245


fertility, and animal dancing, etc*,
Gait, E, A., 159, 267; (1) 49.
65,
cults and agriculture, $$, a6o, a6x;
^^7. S 94 ; () i59 S 62
cult in ancient India, 533; cult and Gandhi, M. K., x:oo, 201, 202,
stones, ^45 sqq.;
203
and Ficus, X46; Ganges River, 5, 27, 72,
109, 236;
and head-hunting, Hindu upper valley of, 34; virtue of
water
ceremony, a6; and Hinduism,
*34; of, 72, 74, 82, 108, X09 sq.
and horse sacrifice, *30; and human Ganges Valley, Indus culture in, 153
sacrifice, ^41 sqq.; and king, 54^, garbhadhana, 239
and monoliths,
S 5 S; 8; and ochre, gardening, ol Badaga, 16
a4in.; and plantain tree, S57; and Gardiner, see Edgeworth
prostitution, 166, 254, a6i; and rope- Gardner, B. B. and M. R quoted,
sliding, ^54; and snake cult, i^st, i74n.
1541 soul-fertility cult, aag; and soul- Garhwal, 35
substance, S36, ^39, if4o,
^43, i?47, Garo, 29; horse sacrifice among,
a48; and stone, j?46, 247; 230;
and inheritance, marriage, affinity 'to
strangers, afii; and tribal religion, Kacharis, 29, 30
a6o; and water, 536, 547, ^49, gaumedha, 228
Fick, quoted, 179
Gautama, 56, 149, 151
fig-tree (or Ficus), sanctity of,
iras, Gaya, 33
^46 Genesis, Book of, 142
Fiji, 138, 187; and analogy with caste, genetics and caste, 131, 132
138, 139; foundation sacrifice in, Gennep, A. van, quoted,
5o; human 137
sacrifice in, 549; Kings gens, 173
Carpenters in, 139; plantain tree in,
ghar baitfma and ghar^jawai,
ifii, 162
ghasbavfi, 248
finger, joint of, amputated, 11, jjgi ghi, purificatory nature
of, 74
fire-walking, 15517, Ghosh, A., 267
FitzGerald, S. V., i74n.
Ghurye, G. S>, 175, 178, 180, 267
Fleure, H. 5ee Peake
J., Gilbert, W*
Forbes, A* K., 73, 81,
Forbes, J., Sm., irfiS
m gipsies, Lori,
H*,
41;
jr*, 120,
castes,
i78sq.,'2S7
51, end sec
Nat s.v. Castes
Forsyth, J., 49n., 553. afiy
Oiuffrida-Ruggeri, V., 70.,
foundation sacrifices,
267
^49, aso Ooa, 6, 13, 18, 74n.; Council
Frazer, J. G., of, 47
jjfiy; (1) ^56; (ir)
74, 164, Godavari River, 4, 5, Map B
165, j?53
Gond, 17, 22, 23; exogamy of,
frontal bone, ^^36 57; and
Kondh, 26; see also Tribes,
Fuchs, S., xi, ^n., s.v.
io8n., Gond; MSria, 23, 25
functional groups and caste,
115, u6 Gonda, Dr J., quoted, 1260.
sq., U7. uS, 133.5, 138, 139. 170, Gondrani cave-dwellings,
in 151
Gordon, D. H., 244, 267
INDEX 3U
Gosains and caste control, 97 sq. Hardinian, P., see
J. Scott (1)
gotra (got) 47, 55.6, 131, 173; analog- Hastings, Warren, 97, 23(5, 269
ous terms for, 56, 57, 58; and Hatch, W.^ J., 127, 267
exogamy, 55, 131; gotrakara rishis, head, deformation of, 41; importante
56# 59 J
157 J and marriage, of among Andamanese, 240; sanctity
59-6*; meaning of, 55; Muslim, ui; of, 23s
pravara, 58, 59; Rajput, 55 head-hunting, and agriculture, 233,
Agastya, 56, 59; Angirasa, 56; Atri, 241; in Assam and Kafiristan, 239,
56; Bahlim, lai (Muslim); Bha- 241; beliefs connected with, 28; and
radvaja (or Angirasa), 56; Bhrigu, marriage, by Bhils, 239; and
240;
56. 591 Oautama or Angirasa, 56; Medes, 239; Naga, a8, 240; and
Kasyapa, 55. 56, 59; Khilji. in prosperity, 241, 244,
245; by Red
(Muslim); Shaikh, 1*1 (Muslim); and Black Kafirs, 43, 241
Vasishtha, 56, 59 headmen of panchayats, loi, 102
arsa-gotra and lauki ha- gotra, Hemeon, C. R,, 259, 267
59
Govind Singh, 37 Herat, 155
grave-digging taboo, 145, 147, jro7 Herodotus. 24, 140, 163, 166, 227,
Great Mother goddess, cult '
of, 151, 43. S67
Hill, S. C., 120, 124, 267
Grecian
tripods, 43 Himalayas, 1, i5an.,
4, 7, 183, 235,
Greece, culture of, **5, *51; religion 254; foothills of, 35, 154
of, 5*5 Hindu religion, 2, 18, 33, 89, 114,
Greeks, invasion by, 6; ostracism of, ii 5 117* i6, 182, 224, 225,
non.; tripods, 43 226-31; and fertility cults, 233; and
Guha, B. S., 3, 6, *43, 1780., *67 primitive religions, App. B, 223
guild organizations, 12; and castes, and stone erection, 246
sqq.;
140, 171, 17*, 180; and marriage, Hindus in Manipur, 29
134 sq* Hippolytus, temple of, 163
Gujarat (Bombay), 36; caste taboo in, Hira Lai, see Russell
188; burial practices of Hindus in, Hivale, S., 257, 267
246; kula in, 58 Hobson-Jobson, 10, 47, 267, 275, 282,
Gujarat (Punjab), 38 292
gupta puju, *41 Hocart, A. M., 268; (1) 187; (2) 137,
Gurdon, P. R. T*, 29, *67 176; (3) 176; 138;
(4) (5) 139; (6)
guru, 76, *8*; and caste control, 98 6f>, 69, 138, 176
Gwalior, 36, 95 Hodson, T. C., 268; (1) 183; (2) 30,
L89I (3) 183, 184
Holdich, T. H., 152, 268
hagiolatry, *62 Holland, T. H., 16, 268
hair, of 3; form among
Australoid, honey, collection of, 8
Negritos and Naga, 3n.; natural
hook-swinging ,255
partings and whorls of, 13; sacrifice Hopkins, E. W., 127, 268
of, 163, 164; sanctity of, 232 Howrah, 31
Hamilton, A., i66, 267 hpungyi, 145
Hamilton, F., = Buchanan, q.v. (2) Hughes-Buller, R., 239
Hamilton, W., 267 Huligamma, 165, 282
Handy, E. S. C,, 187, 188, 267 human sacrifice, and bridges, 249,
Hanuman, 247 250; as a cure, 242; and head-
INDEX
Indus valley, cremation, 336; culture,
human sacrifice {cont,)
Koiidh, $6 and see 153; pre-vedic civilization 'of, 155,
hunting, s$s;
meriak; in Marquesas and Fiji, i?49;
iBo, 334
infanticide, 36
reason for, and agriculture, 539,
inheritance of property, caste differ-
^41; and rope-sliding, 254
ence, 9, 63, 88; Garo, 39; Khasi, 39;
Huns, and Gujars, 38; invasion by,
Kottai Vellala, 13; matrilincai, 9,
6 36; and Rajputs, 115 sq., 157; in
,

Sind, 39
n, i3, !5 sq., *9, 154, 160, 161 and
Hunter. W. W., mi *41 *7 see Marunmkkathayam; Nambudri,
(0 ()

Huntingford, G. W. B., *44, *68 14; Toda, 16


initiation ceremony, 58, 64
huqqa, 76 sq,, *8*
Innes> C. A,, 8i, 83, 84, $5, 369
Hutton, J., 104, *68
Hutton, J, H., *68; (1) *33 *451 (*) insects and soul, *5*,, 353
institutions analogous to caste,
S50, *48; (3) *45, *46. *551 (4) HB* 133
sqq,; see alsG Caste; analogous
*48; ( 5 ) *451 (8) *451 ( 7 ) *48; (8)
86, institutions
*451 (9) *35> *46; (10) 8, 9, 81,
113, 130, 19s, **3n., *35; (u) 186;
intermarriage, 13, 48, 49, 131, '137

(is) *45
invasions, ,5, 6, 13, 15; brachycephalic,

Hutu, 14* 15; hypergamy and, 1$^, 136; Mus-


Hyderabad, 17 lim, 39; into pre-vedic India, 333;

hypergamy, 53, *07, *8a: and of Rajanya warriors, 35; of Rigve-


54,
dic Aryans, 155, 3*4; Saka, Kushan,
caste, 173, 174; and child marriage,
159; and invasion, 135, 136, 149; Hun, 35, 157
and racial restrictions and prohi- Iphino, tomb of, 163
bitions, *07, *08; and sapinda, 61 Iran, 5; cattle-keepers' migration, 16

hypogamy, 53, 55, 64; and bride Iranian, plateau, 5; stock, source of
price, 55 Baloch, 40
Iraq, 157
Isernia, 333, 36in.

Ibbetson, D., 37, 38, 39, 89, 94, 93, Islam, 39, 183; and exterior castes,

*69, 378 304 sq., *19; see also Castes, s.v,


176,
Ibo,
Muslim
143
Israaiii sects, 1740,
illmn, 58
isogamy, 33
immigration, see migration
Iyer, L. A. K., , 335, 369
Indo-Aryan, 5, *1, *3, *7, 33, 156,
Iyer, L, K. A., 3%; (1) fit; (t)' ififi,
18*
Indo-China, 3*3
*451 (S) ( 4) *SS. * 46 .
*49
Indo-European, 6, 14, 189; invaders
and Red Kafirs, 43; social system,
150, 3*4
Jackson, A. M. T., 93, 118, 119, 180,
Indonesia, and caste, 137, 187, 191;
iSx, 369
and cattle sanctity, 333; soul-matter,
Jadeswami, worship of, 17
belief in, 333, *43
Jagannath,'car of, tfi; pilgrimage to, 37
Indore, state control of castes in, Jainism, 148, ttS,' 331

94 sq. Jains, i8, ttg; ^stones in rites of, '348


Indra, 334, 335, 338, 339 Jaipur, 36; (ejeyporc), 348"
'

Indus River, 155, Map B Jaisalmer, 36


INDEX m
Japan, outcastes in, x, 147, 148 Kathiawar, 18, 227; memorial stones
jat (or jati)t 48, 64 in, 247
Jatakas, is 6 sq.^ 139, 184 Kaui, H., 97, 253, 269
Jeremy, Epistle of, 163 Kausika Muni, i57n.
Jerusalem, 14 Kautilya, i26n.
Jevons, F. B., 14011., Si6g Kayan, 256n,
Jews, Ben-i-Israel, 19; in castes, a; Keith, A. B., 65n., 269
White and Black, 14, 15 Kerala, 10, 13
jirgah (council of elders), 4a kesh, 38
johar, 37 Kesha Pingala, 15711.
joking relationship, 63 Ketkar, S. V., 269; (1) 48, 126, i8oj
Jolly. J., 149, 151, adg (2) 9
Joseph, 14J1 Keysser, 186
Jukun, 139 Khan, A. I., 236, 269
Jumna (Jamuna) River, 5, 33, 155, Khandava, 229
Map B Khandoba, 20, 165
Jur Sital (spring hunt), a6a Khas of Nepal, 136
Khasi, 29; disposal of dead, 246;
monoliths, 246, 247
Ks, the five, 38 khelf 58
Kabul, 155 Khilji, gotra, 121
kachcha food, 74, 75, 76. Khoja, Muslim Khatri or LohSna
Kachh, 18, 181; sati in, aay caste, 18, 38, 285
kachh, 38 Kirkpatrick, W., 51, 269
Kadi district, 19 kirpan, 38
Kafir, Red and Black, 43, 335, a4i, Kishengarh, 36
543, 543; memorials, agS Kitts, E. J., 269
Kafiristan, head-hunting in, agp, 241 Kodaga, 15, 16
Kai, i86 Kol, 24
Kali, 167, 225, aaS Kolarian, 4, 21, 23, 24, 25, 161, 223
Kalingapatam, 10 Kond^ (Kondh, Khond), allied to

Kampila, 245 Gond, 25, 26, 239, 241


Kanara, 10, 13, 17, 97 Konkan, 17, 285
Kanarese, country, 13, 167; language, Kopili River, 236
10, 68, 70; Right-hand and Left- Koryak, 261
hand castes, 70, 166 sqq. Kotah, 36
kmga, 38 kdvilf 57
Kangra valley, 136 '
^

Krishna, 26, 153, 228


Kannikan, '9 Kruijt, A. C., 186
Kapadia, K. M., xi, 269 Kshatriya, 19, 26, 38, 49, 64, 65, 66,
Kapila, 1570. 67, 83, 84, 85> 92, 94, 113, 116, 117,
Mra^ 98 149, 151, 156, 157, 180, 197, 284,

Karandikar, S. V,, 58, 269 286, 287, 292, 293; and Brahmans,
doctrine of, 121, 125, 146, 182 151, 156, 157; gotras ot 55J T^ajanya,

Karnataka, lo 65, 293; duties of, 149; varna, 66, 67

Karuppan, i52n., 284 kul, 58

Karv6, Smt. I, yiii, txi, 269 kula in Gujarat* 58


Eassites, 155: Kulinism, 53, 54

m INDEX
Kulu, ail law, customary and caste, n6, 124,

Rumaon, 35 125, iSg, 190; rule of, 128

Rumania, 229 Leach, E. R., viii, 269


Rummata, 245 Left- and Right-hand castes, see

kurij 58 Castes, Right-hand

Rurram valley, 42 Leitner, G. W., 232, 242, 2%


Rushans, 6, leprosy, 145, 146
55
Lewin, T. H., 242, 270
life-matter (or -essence), and children,
Laccadive Islands, pi ornaments in, 89
distribution of belief in,
restriction of, 70, 86, 207 185 sqq.; and fertilization of crops,
Lacroix, P., 140, 269 28, and hagiolatry, 262; and
261;
Ladakh, 43 mana, 187; and rope-sliding, 254;
Lahaul, 43, 44 and sacrifice, 254; and stones, 247;
Langdon, S., 225, 269 theory of, 223, 224, 232; theory of
Languages Karen, 250, 251: and toteinisra, 255,
diversity of and number of, 4
257; transferability of, 232
restrictions as to use of, by castes,
lingum^ 228, 236, 286
86 sq.
Lingayat, n8, 286
Aryan, 6, 224, 228
Linton, R., 137, 181, 270
''of Assam tribes, 28
Lion Gate of Puri, 27
ASsSamese, 27
litigation, cause of, 128
Austric, 4
Little, K. L., 174, 270
Balochi, 41
Bhili (Indo- Aryan), 21
Logan, W., 87, 270
Brahui, 4, 41
Lok Sabha, x, 287

Bravidian, 4, 10, 21, 22, 23, 24, 41, Love, H. D., 68n., 270

158, 223, 225 Lowie, R. H., 142, 143, 270


Halbi, 22 Luard, C. E., 57, 270
Hindi, 22 Lucian, X63, 270
Indo-Ary7.n, 5, 21, 23, 27
Indo-European, 6, 14
Ranarese, 10, 70 McCrindle, J. W*, 154, 159, 270
Kodagu, 10 Macdonell, A A., 640., 156, 270
Kolarian, 4, 21, 23, 24, 25, 161, 223 MacGregor, 0 , L., see Mead
Malayalam, to, 70 Mackenzie, F. S., quoted, i88n*
J,
Malayo-Polynesian, 187 Maclcane, Dr, quoted by Oppert, 68
Malto, 4 McLeod, Fairy Banner of, 256
Marathi, 20; Halbi dialect of, 22
Maepherson, W, and $. C,,. 241, 254,
Mon-Khmer, 4, 223
270
Pali (lierature), 179
Madagascar, and caste, 137; and plan-
Pashto, 42
rain tree, 257
Sanskrit, 43
Madras, to, 16, 116, 158; detmdmis in
Shan, 27
Tamil, 5, 10, 70, 187; 162; exterior castes in, 202-9, 219-22
(lit.) 178
Telugu, 5, 10, 11, 70 Madrasi, 31
.


Tibeto-Burraese, 29 Madura, 8, 57, 154
Tulu, 10 Mafuhi of New Guinea, 186
INDEX S15
magic, and caste, 177, 178; for food marriage, of Agariya, 63; Australian
supply, 6o, 6i; and soul-matter, system, 33, 63; of Baiga, 33, 63, 63;
^59; sympathetic, *58; tribal, j6o and after-life, 159, 160, 344; bina,
MShabharata, 154, a9* 9, ^94 161; of Brahmans, 63; child marri-
mUhUjanam, 67 age, 159; consanguinity and, 59
Maharashtra, 10, ao sqq.; cross-cousin, 9, 61, 63; and
Mahftvlra, 148 endogamy, 47, 53, 58, 160; and
Mahishmati, matrilineal system in, exogamy, 47, 54, 57, 58; of Garo,
154 39; ghar baithna, 163; ghar^jawai,
mahwa, 34, 187 161; in Goriya custom, i63n.; and
Maine, H. S., 154, *70 gdtra 55 sqq.; and hyper-
system,
Majumdar, B. N., a7oj (i) 55, 67; (a) gamy, 53 sqq., 159; of Kadva Kunbi,
357: (S) 44 and to flowers, 19, 30; of Khkngars,
makkathayam system, 70, 160, 163, 187 57; Khasi, 39;and Kulinism, 53, 54;
Malcran coast, 4a, 15a on Malabar coast, 63; and maru^
Malabar, 5, 10, 13, 14, 15, 79, 80, 8a, makkathayam system, see Maru~
83* 84, 86, ui, 158, 159, aoo; raste makkathayam) in matrilineal society,
division in, 166; child marriage 63, 65; and pagoda slaves, 144; and
absent in, 159; inheritance in, 88; pravara, 58; restrictions and re-
king of, 74, 555; language, 86; incarnation, 63; and sambandham,
marriage in, 6a; pollution in, 83, 160; of subcastes, 51 sqq.; see also
84, aao, snn. Caste, restrictions
MalSn mountains, 15a Marriott, McK., viii, xi, 370
Malaya, pygmy tribes of, a Marshall, H. I., 186, 350, 370
Malayllam language, 10, 70, a88 Marshall, J., 157, 370
Malcolm, J., 34, a53, syo Martin, M., 370, 383, 390
malik, 4a marumakkathayam system, 70, x6o,
Malto, 4 161, 163, 166, 167, 388
mana, st4g, a88; and brahma, a3a; and Masai, 143
caste, and soul-stuff, 18a, 187; and Mateer, S., 370; (1) 80, 87; (3) 81,
taboo, 177, i8a, 184-9 8s. *35
Mangalore District. 81 Mathura, 33
Manipur, ag, ago; Maharaja of, ag, 49, matrilineal system, 9, 13, 13, 39, 150,

95t IS9 151 153s 154s 161, 163, 165-9; and


Manu, Code of Law of, 64, 89, ga, culture of Rajputs, 35; and dedica-

93, 94, 118, 134 , H9 150^ tion, 165; inheritance under, 13, 14,

m
Maoris, 136, 188; and tapu, 186
39, 88;
basavi
and ghatjawai, 161; see also

Maratha, 19, ao; Brahmans, 48; cou- Maury a, 153


sin marriage, 6a; influence in India, Mayer, A. C., xi, 370
ai sq.; rulers, chattries of, 335; Mead, P. J., 40, 103, 371

rulers, power of, 95 Mecca, 40


Marathi, ao Medes, 339
Marco Bolo, 166, 373 Mediterranean, 5, u. 151; affinities
Marduk, 165 with Dravidian India, 163 sqq., 165;
Marett, R. Rl, ayo civilization, 338, 331; cults, 163, 165,

Markandeya Purana, 134 333, 334, 338

Marquesas, human sacriflcc in, 349 Meek, C. K., 139, 371


316 INDEX
megaliths, see stone Morris,- C, J., 340, 371

Megara, 163 mother goddas, 153, 338


Megastiienes, 154, 159, iSi; and see Mughal Empire, 6 30 ,

McCrmdle mul, 57
Mela, Pomponius, if 5a, ^71 muni, 65
Melanesians, 3, 7 Mura (Men*, Bokhara), 155
Meman (or Momin), 40, aSg Murphy, J., 180, 371
memorial figures, see stone Murray, M. A., 356, 374
menhir, see stone Muslims, 37, '69; and caste, 174; castes,
Menon, C, A., 94, 7)i 89 4 o w. (for names see
menstruation and pollution, 83, lap Casta); in .Bengal, 3,1; burial by,
Mer, Merat, 35, st^g 18; conception belief of, 357; and
meriah sacrifices, a6, 241 , *54, sSg entry into temples, 300; invasion of,
Mesopotamia, 5, 157, 165, 3 6, 159; and pollution* 83 sq., 186
Me war, 18 mutilation (ol fingers), 11, 391; of
Mexico, half-breeds in, 136; chiquea- genitals, 565
dorest Mylitta, temple of, 163
Micronesia and plantain tree, jt57 Mysore, 5, 10, 335
Midgan, 141?

migration, 5, 6; to Assam, 30; and


caste formation, 119, x8r, 19x1 of NSga descent of Erishna, 153
cattle-keepers from Iran, 16; of Naga Hills, 3* 183,
clans and castes, 35, 36; periodic, Naga tribes, 38, 39, 183 sqq*, 390;
42; of Odh, Jf6; of Pathans, 4a; of disposal of dead by, 335; food
Rlgvedic Aryans, 155; of Tamils, 15 taboos of, 1S5, 186; and hair sanc-
Miller, E., quoted, sain. tity, 333; head-hunting and men-
Mills, J. P. 571; (1) (a) 106, 186; hirs of, memorial stones of,
340-8;
( 3 ) 245 345; Mongolian element in, 39;
Minakshi, 154 Sekrengi hunt of, 163; transport of
Minchin, C. F., 151, ^71 megaliths by, 348; for individual
mirasdars, 206, 89 tribes see Tribes
mitakshara, ^64 Nagel, Dr, i$i
Mitra, aa8 Naina Bevi, and snake cult* ir^tn,., 154
Mitra, R.> sinG, 530, 351, 37X Nambudri, see B.rahiiian, CiMes
Modi, J. J,, quoted, 189m Nana, 153
Mohenjodaro, 153, 156, 334
NSni, Bibi, 151
Molony, J. C., 371; (i) 70, 94, 990.,
Nanjanagud, 353
loon., 109; (3) 76, i04n, X33, i$g
Nanjundayya, H. V*, see Iyer, L. E. A*
momiyai, 344
Narbada River, 154, 336, 353
'

' '

Mongolian, Mongoloid, etc,, 4, 6, 35,


nats, '146
33, 43, 333; clement in Assam, 371
nature worship, 6
in Kuki-Chin tribes, 39; sub-
Nazaroff, P, S., 344, 371
Mongoloid, 4
Mon-iChmer, 4, 333 Negrito, 3, 3, 7, 333; and sanctity of
monoliths, see stone fig-tree, 331

monuinents in Rajputana, 35 negro, 7


moon worship, 335, 337 negroes, social posi'tion of, 119, 136,
Morasu, 11, &gy 174; and pollution, 17411,
lUDEX 317
Nepal, 43, r.40; state control of caste outcastes or exterior castes or depres-
in , 94 sed castes, see Caste, exterior
Nesfield, J. C., 170, 171, Sip outcasting, 106-9
Nestorian missionaries, 14 Pacific, 187, 191
New Guinea, 'ij, 186, 187 Padma Parana, 93
Newell, W. H., vii, 571 pagoda slaves, and caste, 144-6, 307
Nias, people of, S48 Pakistan, in;, gin.
Nicobars and the Nicpbarese, 4, 535, pakka iood, 73-6, 79
j? 46, ^55n. pal, 58
Nilgixi Hills, jo, 16 Palni Hills, 9
Nisaya, 155 Pancharatra, 339
Nishada, 180, ago pUnchSy^t, see caste control
nomadic tribes, 6, 8, i, 34, 41, 43 panchgavya, 88, 108; ox panchamrata,
North American Indians, 136 108
Northey, W. B., 340, 371 Pandaian kingdom, children in, 159
North-West Frontier, 43 pandits of Benares, 33
nose form, (nasal index) 16, 175 Pandyan dynasty, 154
nukha, 38 pangat, 58
Pant, S. D., 34in., 371
Papuans, 3
oaths, J06, 336 sq. Parasurama, 67, 157, 393
occupational villages, 184 Parbati, 19
Oceania, 343; human sacrifice in, 349 Parekh, quoted, loon.
ochre, and fertility, 34in. Pargiter, F. E., 156, 339, 333, 37V
Odumbara (and Dorn), 34 Parry, N. E., 146, 371
Oldenberg, H., 179 sq., 371 Parsis, 18, 31; priests, 189
Oldham, C. E. A.' W., 168, 371 Pashto, 42
Oliver, P., i37n., 371 Pasupata, 339
Olver, A., 169 Pate, H. R., 373; and Tinnevelly
OMalley, L. S. S., Pathans, 39, 41, 43, 43; character of,
371; (1) 54, 63,
64, 71, 8in., 90, 94-uo passim, 43; skull deformation by, 41; tribal
system, etc. of, 43
186; (3) 34n.
Omkar, 353 patrilineal succession, 5, 14, 39; and
Baloch clans, 41; inheritance by,
Oppert, G., 68, 69, 371
88; and matrilineal system, 150,
Oraons, 33, 34, 336, 343? bachelors
157 165
dormitory of, 34; language of, 33;
Peake, H. J. E., 155; quoted, 336n.,
tattoo of, 36
373'
ordeals, 106, 336 sq.
Peate, I. C 55, 373
Orissa, 33, 37, 53, 57, 186; Hills, 35;
pellet-bow, 9
plains, 36; and Assam, 37 Penniman, T. K., see Marett
ornaments, and Aos, 307; and exte- Persia, 18
rior castes, 305, 306; and, Laccadives, Persians, 6
307; restriction on, 85 sq. petticoat, and pollution, 104: of St
Osu, 143 Theresa, 359
Oswal, 18, 35 phaid, 58
Oudh (Awadh), 33; see also United Philippines, 3
Provinces pilgrims, 36, 37
3x8 INDEX
Fintian, 1540. Pudukottai, 179
Piolet, J.-B., 137, 7 pajd, aa8, 141, aga
pipe (tobacco), 76 Puixeat Lake, 10
Fir Pagaro, 40 punishments, 106-9
pirs in Sind, 40 Punjab, 6, 34, 37, 44, 51, 56, 66, 116;
Fitt-Rivers, G. H. Lane-Fox, 131, si^s division of population in, 37; exte-
placenta and soul, *56, *57 rior castes in, ao4; Rigvedic Aryans
plantain tree and soul-matter, ^57 in, X 54 , 155
Plato, 141 Punjabi Muhammadan, 39
Playfair, A., stg, 835, aya pur, 56
Pliny, 154, m* 183, asa, aya purana, aga; Agni, xxo; Bhavishya,
plural societies, 115, 1x9 sq lay sq., aay; Harivansha, 65; Padmt, 93

Plutarch,
m xion., i4on., aya
Purchas*
pufdnh, 158
S., 154, 17a

pollution, see Caste purification from pollution, see Caste


Polo, Marco, x66, aya
polyandry, in Lahaul and Spiti, 441 of
Kanets, 44 racial inbreeding, 131, 135 sq.; see
Polynesia and plantain tree, a57 also s.v. hypergamy
Ponnani River, 8x Rajanya, warriors, 35, 65, 157, 593
population, diversity of, 1 sqq., 44, 45, Rajas and cante control, 93-7
xig sq.; and sex ratio, 131 Rajkhowa, B., xog, aya
Portman, M. V., a4o, a4a, aya Rajmahal Hills, 3, 4, 54, Map B
Portuguese, 6; pirates, 3a; edict of Rajputana, x8, ax, 35; fish-eating in,
1736, 74n. 77; ghasbavfi ceremony in, 548; me-
Powell, J. H., 355, aya morial stones of, 3f46; monuments
pTdtilnm&, 55, 149, X50, X5i, 157 xn, 35; state control of caste in, 94
prd'Ard, 58, 59; and gdtra, 60 Rakshit, H. K., see Sarkar, B, K,
pre- Aryan culture, 156, 177 sq. Ramanatha, ^45
pre-Dravidian, 3, 4, xo, x8a Rangachari, K., see Thurston (a)
pretyasila, a48, asi
Rao, H,, 175, up
pre-vedic Hinduism, X5S-7, aa6-so Rapson, E, J., 650., 135, 156, tig,
priests, in Ancient Egypt, 140; Brah- ays
mans as, see Brahman rath, t03,
primogeniture, 14 Ray, G., 7, 173
property, descent of, 9, 1$, ag, 88,
Rcclus, I., t4i, tyt
x6o sqq.
reincarnation, 63; and ancestor wor-
prostitution (sacred), in Asia Minor,
ship, t5, t53; and fertility,
i6a-6; in Egypt, 163; and fertility,
and soul, 550 sqq.
a6i; in India, 16a; Mediterranean,
religion, phallic, of southern Asia and
163; origin of, 166
eastern Mediterranean, a8; pre-
proto- Australoid, 3, 4, 8, 17, 155, aa3
proto-Mongoloid, Rigvedic, and Hinduism, ttS sq.;

Provence, hull-haiting in, 15a in pre-vedic India, 180, ^24; see also

proverbs on cooking, 73 Appendix B and Caste


Przyluski, i 53 i Revillout, M.-E.., 141, 272
J., 34,
Ptolemy, 183, aya Rice, S., 178, 185, 272
pudam, a46n. Richter, E., 16, 272
INDEX
Right- and Left-hand castes, see sacrifice, of cattle, 228, 230; of dog,
Castes, Right-hand 254; of hair, 163, 164; of horse, 230;
Rigveda, 5, 6, 153-8 passim, 55-3o human, 26, 232, 239, 241, 242, 249,
passim, S93; and souls of dead, 1139 250, 252, 253, 254
Rigvedic age, 35; Aryans, 13, 153, 155, St Thomas, 14
157, 158; burial, Sij6; classes, 64, Saka, 6, 35, 39. 157, 293
179; Hinduism, 168; Indo-Euro- Sakra, 153
peans, 524; invasion, 65, 155, 188; sH^kti, 189, 293
society, 48, 64 Sakti worshippers, 70, 77, 228, 229, 293
rishi, 56, 58, 59, 65, i9, 153, 157, Sakuntala, 123, 293
^*95 salagrama, 251, 293
Risiey, H. H 175, 572, 273; (1) 49, salt, 74n.

51-7 passim, 61, 73, 76, 87, 257, Samaveda, 154, 293
90i 594;* (2) 31; (3) 57, sambandham, 160
157. 39l (4) 54 i34 Sambhar, 36
ns Sandur, 10
Rivers, W. H. R., 273; (1) 160, 244; Sanskrit, xiii, 43
Santal Parganas, 24, 76, 236, 242
() 54
Robertson, G. S., 43, 241, 273
sapinda, 60, 61, 294
Rochon, A., 137, 271 Sarkar, B. K., 154, 273
Rodriguez, E. A., 273 Sarkar, S. S., 7, 273
Roe, Sir T., quoted, 291 sastri, 95, 294
Roman Empire, analogies to caste in, sati,35, 294; Rajput, 227, 246
Satpura Range, 22, Map B
134 sq,
Rome, 225 satra, 98
rope-sliding (human sacrifice), sattva guna, 182
254
Roscoe, Saurashtra, 255
J., 71, 139, 273
Sayce, A. H., 157, 273
Rose, H. A., 273: (1) 37, 38, 39, 43, 44,
scapegoat, 109
49 53 64, 66, 1520., 158, 254, 78.
Scheil, V., 1570.
281, 287; (2) 1670.
Schniidt, Father, quoted, 251
Rowe, G. S., see Williams
Schoebel, C., 67, 273
Rowlands, Martin, xiii
schools, exterior castes and, 201, 202,
Rowney, B,, 243, 273
208 sqq., 215, 222
Roy, Sachin, 1830., 273
Scott, J. G., 273; (1) 241, 243; (2) 145
Roy, S. C., 273; (1) 24; (2) 252, 253;
Scythians, 6, and see Saka
(3) 243; (4) i75
secular rulers, power of, 93-7
iBg Sedgwick, L. J., 131, 274
Roy, S. N., 250, ,273 Sekrefigi hunt and fertility, 262
Ruanda, 142, 141 Seligman, C. G., 256, 274
Russell, R. V., 20, 37, 57, 63, 71, 73, Senart, E., 49, 50, 150, 168, 172 sq.,
74, 232, 235, 243, 244, 252, 262, 273, 185, 274
278, 287 Sen Gupta, Professor, quoted, 215
Ryan, B., xi, i38n., 17611., 273 seniority of sons, 88
Sephardim, 15
septs, and exogamy, 56, 58; of Pathans,

saMa* 97, 9^1 ioo> 293 42; totcmistic, 11

sacred thread, 58, 64, 85 Servants of India Society, 269


t

S50
INDEX
shadow and soul, and head, 239 sqq,.; and human
Shaiva sect, 7*7
sacrifice, 239-42; and magic, 258,
259; and megaliths, 245 sqq.; and
Shakespear, J., S53, 274
Shamash, 165 plantain tree, 257; and totemism,

Shanar (Shanan), see Castes, s.v. 256-8; and water, 236

Shanan soul-stuff, see soul-niatter

Shaw, W., 146, 274 South America, 136


Sherring, M. A., 115, 1S3# *74 Spencer, B., 290., 270
Shiah, 38, 295 Spiti, 44
Shiva, 19, 524, 225, 229, 236, 245, 253; spring hunt, 24; and fertility, 262
marriage to, 154; and Ramanatha, sradh 6on., 234
and snake cult, 228 Srinivas, M. N., vii, viii, x, xi, 274
545;
Siamese, 207 standards and life-matter, 256
Sikhism, 231 Statesrmnt The (Calcutta), quoted, 86

Sikhs, Akali, 37; as carpenters, 123; Steele, A., 62, 84, 88, 274

Keshadhari, 118; Mazhbi, 204; Stevenson, H, N. C., xi, 274


Sahejdari, 117 Stevenson, S., 73, 75, 14011., 186, 274
Sikkim, 43 stone, in Assam, 245; cult, 223; dol-

Sind, 6 , 39, 42 mens, 13, 245, 246; and fertility,

sin-eating, J09 247 sqq.; graves lined with, 12, 170;


Singapore, 18 Khasi, etc., 246. 247; megaliths, 224,
Sirohi, 36 235, 245-8; memorial, 24, 25, 245,
skull deformation, 41 246; monoliths of Sawara, 25; monu-
Slater, G., 153, i 77 ^74 ments, 15211.; of Munda, 24: Naga,
slaves, pagoda, 144, 145, 207 28; and prosperity, 245; and salt,
Sleenian, J. L,, 127, 274 35, 246; and soul, 235; transport of,
Sleeman, W. H,, 99, 274 248; veneration for, 248, 249
smoking restrictions, 76, 77 Strabo, 141, 164, 229, 259, 274
snake worship, 152, 154, 225, 228 strangers and fertility, 261
sokha, 260 Strasser, R,, 256n., 274
Somali, 142 subcastes, 51 sqq., 116 sq., iBr; see also
Sonnerat, M,, 68n., 123, 274 Castes
soul, in foundations, 249; and dreams, Sudra, 64, 65, 71, So, 84, 85, 87, 2,
252; and fertility of crops, 26, 74, 93, 122, 125, 149, 150, 151, 295
223, 236; and fish, 255; and the Sughda (Sogdiana), 155
head, 232, 250; Kondh belief in, 25; Sukgaxm, 230
and mma, 186, 187; as a manikin, Sumatra, horse sacrifice in, 230
233, 251; seat of, 232, 253; and sun-worship, 227
shadow, 242; and sradh rites, 234; Sunni, 38, 295
-substance and momiyai, 244; and Surat, Gulf of, 18
taboo, 185; and wooden figures, 25, Surma Valley, exterior
'

castes in,

*35 215 sqq.


soul-fertility cult, and reincarnation, sutras, ,151, 22S, 295

252; and king, 253 suttee, sec smi


soul-matter, philosophy, 184-8, 223, Sylhei, 31, 215-19
232, 236, 239sqqi; and blood, 242, Syria, Syrian, ,5; Christians, 14;
243; -essence and haglolatry, 262; Church, 14; .Goddess, 254
INBMX
taboo, II, 187-90; on building Tod, 36, 236, 253 sq., 274, 278
J.,
materials, etc., 87; in liurma, 145, Toda inheritance, physical type, etc.,
146; and caste, 115, 18a; on food, i6; numbers, 44
77, 185; Indonesia and Pacific,
in Tomal, 142
186; and intermarriage, 14a, 186; in tooth-chipping, 8, 9
Japan, 147; and maha, 186; and Toradja, 186
totemism, *55; and untouchability, totemism, u, 21, 23, 78, i57n., 178,
S07 255-8; origin of, 257, 258; of rishis,
Tagore, 31; family and Pir-Ali Brah- 1570.; in standards, 256
mans, 100 totems, and food restrictions, 77 sq.,
Talaings, aoy 257 58
talikettu ceremony, 166, a6i, ags Tower of Silence, 18
Talmud, 131 transmigration, 226, 239 sqq., 250 sqq.
tamaoat 186 transvestism, 165
Tamil, n, la sq.; caste divisions. Travancore, 8, 9, lo, 200, 247
Right-hand and Left-hand, 67-70, travelling, caste restrictions on, 88
167; country, see Tamilnad; langu- tree houses, 9
age, 5, io, 68, 70, 187; literature, 178 tribal religions and Hinduism, 223
Tamilnad, 5, 10, la, 13 sqq., 262
Tanila, 137 Tribes
tapu (of Maori), 186 Abor, 27, 183, 276
tattooing: Gonds, 3; Kondh women, Afridi, 42
s6 ;Oraons, a6 Agariya, 62, 276
Telingana, 10, and see Andhra Ahom, 27, 276
Telugu, 5, 10, 11, 13;Brahman, 11; Andamanese, 2 sq., 7, 44, 231, 240,
caste divisions of Right- and Left- M2
hand, 67-70, 167; -speaking area, 10, Ande Koraga, 81, 285
70; and Tamil, u; upper class, 11 Apa Tani, 28
temples, and exterior castes, 8, 83, Aratta, 154
198, soo, SOI -4, 210 Assam hill tribes, 27 sqq., 183; see
Thakkar, A. V., oi also Naga, etc.

Thar desert, 7 Awan, 39, 276


that, 58 Badaga, 16, 116
Thebes (Egypt), 141 Baiga, 22, 23, 62, 63, 239
Theodosian code and caste, 134, 135 Baloch, 39, 40, 41, 277
Thoothi, N. A., 58, 274 Balti, 43

Thugs and Thuggee, 127 Bhatra, 262


Thurston, E,, 274; (i) 13, 62, 160, 244, Bhil, 19, 21, 35, 116, i55,*239, 242,

245; (2) n-13, 26, 57, 68, 69, 85, *78


109, 1670., i88n., 239, 241, 245, 276, Bhotiya, 106
2S4, 286, 28B, 289, 290, 294, 296; (3) Bhuiya, 24, 252, 253
7 83. 245 Birhor, 24, 25, 243
Tibet, 27 Bondo, 25
Tibeto- Burmese languages, 27, 28, 29 Brahui, 4, 24, 41, 151, 239
tiger-flesh,
77 . .
Brokpa, 43
tight-rope walking and fertility, 254 Burusho, 43, 279
Tinnevelly, 81, 100, 102, 200, 202, Chakma, 32
246-711. Gbauhan, 36
,

Tribes (conL) Mahsud, 42


Chencliti, 17 Mak-Arayan, 9, 235, 247, 288
Chin, 28, 145, x8o Maia-JPantaram, 8
Chodhra, 21, 155, 504, 280 Malavetan, 9
Chuhra* 39, 116, 204, 219, 8o Maler, 24
crimina! tribes, position of, 197 Manipuri, 207; se also Meithci and
Dafla, 27, io6n. main index, s.?. Manipur
Dard, 43 Meithci, 116, 139, 236
ErnSdan, 80 Mco, 35, 2$9'
Gadaba, 25, ^ 26a
Mex, 35, 289
Garhwali, 43 Merat, 35. 289
Garo, 9, 30, 230, 235, 237 Mikir, 253, 289
Ghikai, 42 Mina, 35, 94, 232, 289
Gond, 17, 22, 23, 25, 26, 57, 114# Mishmi, 27
161, 252, 262; Maria, 23, 25, 2d Mohmand, 42
Gurkha, 43 Moplah (Mipplla), i, 289
Hasala, 235, 282 Mru, 32, 289
Hazara, 261 Munda, 24, 246
Ho. 24, 257 Muthuvar, 9
Juang, 24 Naga, 290 {see aUo main index);
Kachari, 30, 114, 236 Angami, 30., 230, 232 59., 243,
Kachin, 243
Kadar, 3, 8, 9, 83
45 * m
62
Ao, 86, 106, 185, 207, 240, 246,
Kafir, Red and Black, 43, 235*, 241, 255. *5^
242, 243 Chang, 240
Kannikar, 9 Kachha Naga, gn.
Karen, 186, 250, 253 Konyak, 30., 235, 246, 247
Katkari, 21 Lhota, 159, 252
Kharia, 24, 1890., 243, 285 Rengma, 236
Khasi, 29, 246, 247, 285 Serna, 230, 242, 255
Kho, 43 Nayadi, 80, 83, 220, 222, 290
Kochh, 30, 116, 213, 285 Ncwar, 240, 290
Kodaga, 15, 16, 285 Odumbara, 34
Kol, 24, 285 Otakzai, 42
Kondh, 25, 26, 239, 241, 254, 285 Oraon, 24, 26, 236, 242
Kora, 257, 285 Falaung, 236, 257
Korava, 127, 245, 286 Paliyan, 8, 116
Korku, 23, 232 Panwar, 236
Korwa, 23, 286 Parja, see Bondo
Kuki, 28, 29, 145, 258, 286; Thado, Paihan, 42, 43 .

146, 242 Rabha, 116, 293


lakher, 14611. Rind, 40
Laptha, 32 Sansiya, 104, 294
Loba, 243 SantSl, 24, 236, 294
Loi, 49, 139 Sawara, 25, 235, 294.
Lori, 41, 287 Shan, 27,, 24s
Lushei, 242, 253, 287 Shin, 232
Magh, 32, 287 Thantapulayan, 9
iNDBK m
Tribes (cont.) varnasankara, 17a
Toda, i6, 44, *44, #96 Varuna, aa8
Ujli Mina, aga Vasistha, 118, 149, 156
Ulladan, 9 vedas, 65, ga, 154 sqq., aassq., aa8-
Urali, 9, age SU agy
Wa, a4i, a4s, 855, 68, *97 Verelst,H., quoted, ,96
Yeruva, 16 Vilayur pollution rock, 80
Yusufeai, 4a villages, functional or occupational,
Tribes, Lost Ten, 19 139^ 184, 185
Trichinopoli, (Ta on map), 108, 109 Vindhya Range, aa, Map B
Tucci, Professor, quoted, 148, aas Vishnu, a6, aas, aaS^ aag; Institutes of,
tuisi, 105 151; name and origin of, 153;
Tulu, 10 Purana, 153
Turks, a6i Viswakarma, la, 7a, 113, 17a
Turner, A. C., 35, 158, 161 sq., i6an., Viswamitra, 156

Turner,
m G. S., 557, 574
Vyasa, 65, 156
Vyasokta, 65
J.
Tussi, or Tutsi, 14a
Twa, 14a
*
twice-born {dvtja), 57 sq., 64, laa Warneck, 186; quoted, 187
*
Twin Brethren
^ 157 Warren Hastings, a69 97, asC,
twins, seniority of, 88 water, and pollution, 71, 7a, 76, 77,
ail, aia, ai7, ai8, and s.v. wells;
and fertility, a36, asg, a47, a49,
Udaipur, 36 a6on.
Uighur, a6i Webb, A. W. T., las, a74
Umia Mata, 19 wells and pollution, 81, 195, 196, aoi,
United Provinces, 51, 5a, 66, 78, 101, aio
116; bina in, 161 Westermarck, E., 131, 174, a75
United States, negroes in, 119, 136, Western Ghats, 5, 10, 13
174 Williams, T., 160, a75
Unja, 19 Wilson, J a75; (i) aa7; (a) 54n., .56,
unseeables
*
81, ia6, aga 65, egn., 80, 90, 93, 100, ia4, ia6n.,
untouchables, see exterior castes i 49 > HSf 2540., a87, ago
up&nitydnd ceremony, 58, 64 Wise, J., 51, 5a, 7in., 74, 1580., i67n.,
Upanishads, 156, aag, age aa7n., a75
witchcraft, ai, as, asg; and sacrifice,

a6o
Vatshnava, 77, agb women, position of, among the Ban-
Vaishya, 64, 65, 66, 84, 85, ga, 113, jara, ai; among the Bhils, ai; and
laa, 149* ^9 caste, lag; on Malabar Coast,* 14;
Vajrasuchi, 65, isyn. among the Marathas, ao; panchayat
Vamsadhara Eiver, 10 of, loa
vamu, 48, 56, 64, 65, .66, 67, 84, 85, wooden *
monoliths *,
347; post in

93, 118, lag, 137, 149, 150, 151, 168, Sulagava ceremony, aso; sledges, a48
171, 17a, 176, 180, 18a; connotation wooden soul figures, 43, asa, a35, 237,
of, 64-7; and marriage, 66, 149-51; 238, 246, 247
and Rigvedic society, 65 sq. Woodthorpe, R. G., 238
SH INDEX
(iCTclopment of, 155, 156 Ycniiba,
writing.
Wukari, Yule, H. 475; (t) t6i; (t) see B.ohson-
139

Yajurveda, 154, 156, saO, ^93; and


cattle worship, 228
Zanzibar, iH
(oC Bifd N3ni), 152
Yati, 229, 297
Yeatts, M. W. M., 228, 24611. sq., 275 Ziiilciii, T., liii, 275
Zoroastrians. 18
Yebir, 142

Social
' ' LIBRARY
^ /ce No
C> a,iKo..
V Fwk. ^o..
.. ,

y I
>

X. ^Ikkabui

You might also like