Articles On Caste and Class in Early India 2016

On the Translation of the Term Varna
Author(s): Thomas R. Trautmann

Source: Journal of the Economic and Social History of the Orient, Vol. 7, No. 2 (Jul., 1964), pp.
196-201
Published by: BRILL
Stable URL: http://www.jstor.org/stable/3596240
Accessed: 03-03-2015 07:32 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
BRILL is collaborating with JSTOR to digitize, preserve and extend access to Journal of the Economic and Social History of
the Orient.
http://www.jstor.org
This content downloaded from 14.139.62.113 on Tue, 03 Mar 2015 07:32:15 UTC
All use subject to JSTOR Terms and Conditions
196 MISCELLANEA
Al-Madkhal ild tanmiyatal-a'mdl bi-tahsinan-nyiydt,and also to a lesser extent in

the book of Abi Bakr titled Kitdb al-IHawddith
wa'l-Bida',. The themes
of the books of at-Turkumani and Ibn al-HIdjdj overlap. The style and attitude of
at-Turt.ii
at-Turkumani are, due to his orthodox faith, more passionate than with Ibn al-
IHIdjdj.At-Turkumini deals with his subjects more thoroughly and in the eagerness
of his presentation repeats himself often, whereas Ibn al-HIdjdj due to his calmer
approach, expresses himself more concisely and to the point. All the same, at-
Turkumdni always has more evidence for his arguments at hand, since he quotes
from legal books, biographical and literary works, and also from the Safi books,
the Diwans of the poets. It is noticeable that at-Turkumdniagain and again brings
quotations from the old and new testament, and puts them beside those from the
Qur'In. Ibn al-HIdjdj did not mention some of the bid'as, for instancethe Futuwwa
association, which at-Turkumani dealt with, but on the other hand gave more
complete information about the bid'ain the sphere of craftsmanship,manufacturing
and business life. Such bid'as remind us of the themes of the Islamic works about
the hiyal and the bisba.
The 3 above named books are essential sources for the understandingof the life
and strivings of Islamic society in the middle ages. These sources have been by no
means sufficiently used for research. The problems of the bid'a are little discussed
nowadays in the Islamic world, since the problems are no longer acute due the con-
tinuous progress of modernization.*
Subhi LABIB
(Hamburg)
*) This paperwas readat the z6thInternationalCongressof Orientalistsin New Delhi,

1964.
ON THE TRANSLATION OF THE TERM VARNA
No one who has been even slightly acquaintedwith mattersIndian can be unaware
of the four-fold division of Indian society into brdhmanas,ksatriyas, vaivyasand
jfidras,surely one of the oldest and most persistentideas of Indian social philosophy.
Each of these groups is referredto as a varya,a word which, in its basic and general
sense, signifies "colour". The purpose of this paperis to provide a suitabletranslation
for varnain its technical sense, that is, as it is applied to society.
We heartily endorse J. H. Hutton's view that varnais "a term which is often
confused with caste (jiti, jit) though it is far from having the same meaning". 1)
Other scholars have taken the opposite view. For example, A. A. MacDonell has
argued that varnais caste in the sense of a group in which membershipis hereditary,
which is endogamous, to which is ascribed an hereditary occupation and which
observes rules restrictive of contact and commensality with other such groups. He
feels that Pili literature affords special evidence confirming this thesis, apparently
because Buddhism could be more objective about a system which was essentially
brahmanic."Here," he says, "we find the termjdti, literally 'birth', used to express
'caste' like the Sanskritvarna.A man is describedas a Brahminor a Ksatriyabyjdti,
I) J. H. Hutton, Castein India (3rd ed.), Oxford, I961, p. 64.
MISCELLANEA 197
or to be byjdti a Chandila or Nisdda, etc., as belonging to the despised part of the

population." 1)
Prof. MacDonell's first sentence is misleading;_iti is also a Sanskrit word, as
reference to his own PracticalSanskritDictionarywill show 2). Indeed, in Sanskrit
literature, a distinction between the two words is scrupulously observed. Against
MacDonell's description of the way in which the words are used in Pdli literature
three objections may be raised. First, to describe a man as brdhmana byjdti is to say
he is such by birth. Varna(like caste, be it admitted) is a status conferred upon the
individual by virtue of his birth to parentsof the same group;jdti is here simply used
in its non-technical sense. Second, while there is some ambiguity in the use of the
word jdti, there is none in the case of the word varna. Varna refers exclusively to the
four social groups named above 3). It would be absurd to speak, for example, of the
Lohar, Dhobi or Lingdyat varnas.Finally, the fact that Buddhism esteemed deeds
more highly than birth, acquired status more highly than ascribed status, to employ
the anthropologist's categories, made it less sensitive to the distinction between
varnaand caste than the brdhmana pundits, in whose interests it was, by contrast, to
maintain the two institutions.
Before settling upon a translation for varnatwo questions must be dealt with:
What is vartnaand what is the relation of varnato caste? To the first question a simple
answer may be given. The varnzas are large and somewhat theoretical categories of
people which are believed to be of divine institution and the sum of which make
up the total of civilized, that is to say, Indian, society. We already have indicated
their hereditaryand functional aspects; it is their sacred characterwe wish to stress
here. The answer to the second question is more difficult. Nowadays many castes
claim to belong to a given varna,and it would at first glance appearthat the relation-
ship between the two terms should be that of genus to species, exactly that ot the
class of quadrilateralsto a square. Obviously if each of the units A1, A2 and A3 are
endogamous, the A-group as a whole is endogamous. The same goes for hereditary
membershipand rules restrictingcommensality.However, the genus-speciesrelation-
ship breaks down over the matter of function or occupation, even when one takes
into account not the actual occupations indulged in by a given caste but the occu-
pation ascribedto it by tradition. For we should expect the functions of the various
castes making up a varnato be specializationsof the function of the varna;to give a
hypothetical illustration, we should expect the ksatriyavarnato be composed of
artillery,cavalry, and infantry castes. But this is not the case-far from it. It is com-
mon knowledge that certain castes of the others
brahmahavarnapursue farming,
commerce, still others cooking, in defianceof the idstrasand in conformity with the
venerable usages of their particularcaste, and that in Kerala a jtdra caste, the Nairs,
were lords of the earth both literally as agriculturalistsand figurativelyas kings and
nobles, up to the time of the British Raj. So we see that the genus-species relation-
ship doesn't account for all the facts. And it is only fitting that this should be the
case on reflection,for we are dealing with social institutions, not geometry.
I) A. A. MacDonell, "The Early History of Caste" in The American Historical Review,

vol. xix, no. 2, January, 1914, p. 238.
2) MacDonell, A PracticalSanskritDictionary,Oxford, 1924.
3) With the exception of early references to the Arya and Disa varnas,which don't affect
our argument.
198 MISCELLANEA
The question inevitably leads us to the dim past and the labyrinthineproblem of
the origin of caste, a problem which we are not competent to consider in detail.
However, there are a few facts of a general nature regarding caste and varnain
ancient India which are more or less widely recognized and which are pertinent to
our theme. The first of these is that varnais anterior to jdti; the four-varnasystem
could and did exist without caste. This might suggest that caste in some way or
other evolved out of varna.Manu has stated that the castes are the result of irregular
liasons between members of different varnas: what might be called the "mongrel
theory" of the origin of caste 1). Now the second fact: In ancient Persia there existed
the four piftras, a four-fold, functional and sacred division of society analogous to
the varnas,which failed to give rise to caste. Thus the relationship of varnato caste
is not a necessary one; whether or not varnza indeed is the ground and origin of
caste, it clearly need not have given rise to caste. As to the mongrel theory, it would
scarcelybe novel to say that it is an unconvincing attempt to make dogma fit the
empirical situation. Indeed, emile Senart said this long ago and even MacDonell
is forced to admit it 2). Senart went further and said that the whole varnasystem
was an artificial creation of the pundits, having very little bearing on reality. We
agree, but would specify that varnais a sacred concept, which explains why it is
so long-lived. Castes are born and wither away, fuse together or split into smaller
castes, but the same four varnasmentionedin the Rg Vedaendure.The relation, then
of varnato caste is that of the sacred and enduring to the empiricaland ephemeral.
It is not surprising that one who holds, as MacDonell does, that the varnasare
castes would call them the "four main original castes"3). But other writers who are
more or less sensitive to the distinction betweenjdti and varnahave been known to
indulge in similarlymisleading periphrasis.Thus Senart provisionally speaks of the
"quatre castes primitives", the "trois hautes castes" (dviavarndh)and the "quatre
castes" before deciding that the varnasare classes4). Kosambi has come up with the
"four 'original' castes", the "four primaryclass-castes",the "four-casteclass system"
and an even greater mystery, a "four-caste division into classes"5). The dangers of
translating varna in some such way should be apparent: It is implied that varnais
jdti or a kind ofjdti.
"Class" is by far and away the most acceptable translation for varnaso far put
forward. It suggests the economic nature of the groups, representingas they do the
priesthood, nobility, bourgeoisie, and bondsmen. Yet for that very reason "class"
is unsatisfactory: In Indian thought the varnasare not essentially economic but
sacred, that is, immutableand of divine creation, as we have said above. To translate
varnaas "class" is to choose a term much too objective, scientific and modern to
represent adequately the notion of varna.We should like to suggest the use of the
word "order" or "estate".
J. Huizinga remarksthat the synonymous words "estate" and "order" had many
i) Mdnavadharmaidstra x. 6-40.
2) I. Senart, Les castesdansl'Inde (nouvelle ed.), Paris, 1927, pt. 2, passim; MacDonell,
The Early History of Caste, p. 236.
3) The Early History of Caste, p. 234.
4) Op. cit., e.g. pp. 5, 131.
to the Studyof IndianHistory, Bombay, 1956, e.g. pp. 104,
5) D. D. Kosambi, Introduction
239, 154, and 141 respectively.
MISCELLANEA 199
meanings in medieval Europe and goes on to say, "That which, in medievalthought,

establishes unity in the very dissimilarmeanings of the word, is the conviction that
every one of these groupings representsa divine institution, an element of the orga-
nism of Creationemanatingfrom the will of God, constituting an actual entity, and
being, at bottom, as venerable as the angelic hierarchy1)." An extension of the idea
of the divine institution of the orders is that of the greater sanctity of the higher
orders as compared with the lower. "Now, if the degrees of the social edifice are
conceived as the lower steps of the throne of the Eternal, the value assigned to each
order will not depend on its utility, but on its sanctity-that is to say, its proximity
to the highest place 2)." Another featureof the idea of the three estates, to which the
regeneratevarnascorrespond,is that the third estate, the commons, ratherarbitrarily
lumps together people of diverse occupation and rank. We quote Prof. Huizinga
again: "No distinction in principle was made, in the third estate, between rich and
poor citizens, nor between townsmen and country-people. The figure of the poor
peasant alternates indiscriminatelywith that of the wealthy burgher, but a sound
definition of the economic and political functions of these differentclasses does not
take shape 3)."
These features of the medieval European notion of estate or order correspond to
the Indian notion of varna.The divine creation of the four varnasas well as the idea
of the differing degrees of sanctity adhering to them are made explicit in the myth
of their creation from the mouth, arms, thighs, and feet of Purusaor Brahmd4), and
there can be no doubt that the vai.yaand .fdra varnasconsisted and still consist of the
most diverse economic groups and castes, including cultivators, artisans,merchants
and even kings, so as to be all but meaningless categories from a sociological point
of view. It must, ot course, be admittedthat the Europeanidea of the "three estates"
came to have strong political connotations, being linked with the development of
representativegovernment. Such a connotation vartnahas never had, and for that
reason the use of the word order is perhapspreferable.
But that medieval European thought provides us with a concept corresponding
to an Indian concept should not surprise us, for there are other correspondences
between the two cultures with respect to social thought. The derivation of the four
varnasfrom the body of Purusa reminds one of the medieval European idea of the
"body politic", the idea which John of Salisbury carried to its furthest possible
extreme in his Policraticus,where he drew the following analogy, summed up by
Gierke: "The servants of Religion are the Soul of the Body..., the prince is the
head, the senate the heart, the court the sides, officers and judges are the eyes, ears,
and tongue, the executive officialsare the unarmedand the army is the armed hand,
the financial department is belly and intestines, landfolk, handicraftsmenand the
like are the feet, so that the State exceeds the centipede numerositate
pedum;the pro-
i) J. Huizinga, The Waningof the Middle Ages (Eng. ed.), London, 1924, pp. 47-8. 0.
Gierkeassertsthat,whendealingwith the originsof groups,the medievalmindalwayshad
recourseto the idea of divine creationratherthan that of naturalgrowth, "in accordance
with its general view of the universe" (Political Theoriesof theMiddleAges, F. M. Maitland,
trans., Cambridge, I900, p. 29).
2) Huizinga, op. cit., p. 48.
3) Ibid., p. 49.
4) Rg Veda x. 90o and Manavadharmaiastra i. 31.
20z MISCELLANEA
tection of the folk is the shoeing; the distress of these feet is the State's gout ... 1).
The Indian arthaidstraliteratureoften speaks of the seven angasor limbs of the state
(rdjya), namely, the ruler (svdmin),minister (amdtya), land with its population (rds.traor
janapada), fort or capital (durga), treasury (kola), army (dayda) and ally (mitra). The
idea that each element of the state and each order of society had a relationship to
the whole analogous to the relationshipbetween the organs of the body to the body
as a whole is a major theme of traditionalIndian social thought as it is of medieval
European 2).
Side by side with, and often contradictingthe myth of the derivation of the four
varnas from the body of Purusa, the Parvan of the Mahdbhdrataand certain of
•dnti(golden age) men lived in naturalharmony,
the Purdnasdeclarethat in the K.rtayuga
without kings, property or social distinctions. It is stated in the BhdgavataPurdna
(ix. 14-48) that in that age there was only one Veda, the word "Om"; one god,
Nardyana;one Agni and one varna,and that the triple Veda, and presumablyalso
the four came about with the coming of the Tretayuga,i.e. with the decline
of humanvarnas,
nature. Bhrgu in the Mahdbhdrata (2z.i81.i ff.) categorically states that
there is no such thing as varna,that all the word was once brahmanicand that men
became separatedinto the four varnasaccordinglyas they fell away from the study of
the Vedas and clean occupations. Mahdbharata12.5 9.1 ff. although silent on the lack
of social distinctions during the golden age, says that men lived according to dharma
and protected one another, and that there was neither king nor coercion (daynda).
But error set in and human minds became clouded; anarchyshowed its malevolent
aspect. The gods, suffering from the dearth of sacrificial nourishment, petitioned
Brahmd,who composed a treatise on kdma,artha,dharma,and moksafor the use of
men. Prthu, a descendantof Visnu, took an oath from the rsis to uphold the Vedas,
protect brdhmanas and prevent the mixture of varnas,becoming the first earthly king.
The medieval European concept of the state of nature is strikingly similar to the
Indian theory of the and the origin of kingship, property and the estates.
This concept may beK.rtayuga
traced back to Seneca and his "policy theory of the state",
according to which when vice set in mankind fell from its primal innocence and
founded the state to keep debased human nature in check. St. Irenaeus casts this
theme into a Christianmould: men, living in the state of nature, turned away from
God and fell a-fightingas a belated result of Adam's sin ; God thereforeset some men
over the rest to rule. "We have here," says A. J. Carlyle,"an explicit statementthat
the institution of government has been made necessary by sin, while it is also a
divinely appointed remedy for sin 3)." This view of government, which enjoyed
great authority in the Europe of the Middle Ages, is singularly Indian in spirit. In
references to anthropomorphism in
I) Op. cit., pp. 131-2, n. 76. (Nn. 76-97 give copious
medieval European thought.)
z) Vide P. V. Kane, "Seven Afigas of Rdjya" (esp. pp. 17-19), chapt. 2 in History of
Dharmaidstra,vol. iii, Poona, 1946. The Sukraniti(i. 61-2) compares the seven angas to the
parts of the body as follows: the king is the head, the ministers the eyes, the ally the ear, the
treasurythe mouth, the army the mind, the fort the hands and the territory, and population
the feet. Alas, Dr. Lallanji Gopal has proved the ?ukranitito have been written in the last
century: "The Sukraniti-a Nineteenth-century Text", in the Bulletinof the Schoolof Oriental
and African Studies(Univ. of London), vol. xxv, pt. 3, 1962.
3) "St. Augustine and the City of God", chapt. 2 in TheSocialandPolitical Ideasof Some
Great MedievalThinkers(F. J. C. Hearnshaw, ed.), London, 1923, PP. 45-6.
MISCELLANEA 201
Christendom the good monarch, by firm rule, good example and the protection of
the church, creates the conditions favorable to Christian virtue and hence salvation;
so too, the ideal Indian rdjan, through the exercise of danda and liberality towards
brdhmanas,causes his subjects to practice dharma and so attain to high heavens or
even final release.
Thus if the notions of on the one hand and estate or order on the other
correspond very closely it is only because of the broader similarities between tra-
varn.a
ditional Indian and medieval European thought : the great importance attached to
the principle of hierarchy, the scholastic tendency to spin out involved analogies,
the preference for the idea of divine creation over that of natural growth when
speculating on the origin of things, and the idea of a fall from primordial innocence.
Thomas R. TRAUTMANN
AN INSCRIBED WEIGHT FROM MATHURA

The discussion of ancient Indian systems of weights and measures has frequently
revolved around two problems, the reconstruction of the indigenous systems of
which accounts, or partialaccounts, are preservedin such texts as the Arthaidstraor
the ManavaDharmaidstra,and the determiningof the actual weight systems to which
specimens--mainly coins--may be assigned. The difficulties attendant upon both
attempts are too well known to requirerepetition. While the weights of many coins
of various classes have been published, very few actual weights, least of all inscribed
weights, have been so recorded. The purpose of this note is to publish for the first
time the weights of two specimens from Mathurd,one having a numeral inscribed
upon it. 1) It is to be hoped that other such inscribed weights, of which I believe
there may be more than one in existence, may also find their way into publication.
i. The first specimen is a sphere of highly polished brown and grey-brown jasper
or chert, bearing the museum registrationnumber 3948. Its condition is nearly per-
fect, the surfacehaving a high gloss. The diameteris 7 cm. The registrationnumber
is markedin thick red paint, and there is also another single stroke of red pigment
of uncertainage. The weight of the specimen in 437.3 gm. There is no further in-
formation regarding the provenance, save that it comes from Mathurd,and that it
was not the product of regular excavation.
2. The second specimen is of a similar red-brown stone, probably chert. It is
spherical,though slightly flattenedon two faces, and is less highly polished than the
first. It bears the registration number 2757. There is also an incised sign engraved
upon the surface of the stone, measuring about 3 x 2.6 cm. (Fig. i). Immediately
beside this sign is another indistinct mark which appearsto be a naturalflaw in the
stone (but which could possibly be a furtherstroke belonging to the sign). The weight
I) Both specimens are in the Curzon Museum, Mathurd. I wish to acknowledge my

gratitude to Shri Muhamad Zaheer, Officer for Archaeology and CulturalAffairs, Govern-
ment of Uttar Pradesh, for his kindness in facilitating my visit to Mathurd in December
I961; also to Shri V. P. Mathur and members of the Museum, particularly to Shri Brij
Murari Lal Saksena who weighed the specimens for me. All weights in this paper are in
Metric grammes, etc., unless otherwise stated.
JESHO VII 14
Bhandarkar Oriental Research Institute
A GLANCE AT THE WORD JĀTI IN THE VEDIC LITTERATURE

Author(s): Uma Chakravarti
Source: Annals of the Bhandarkar Oriental Research Institute, Vol. 86 (2005), pp. 127-130
Published by: Bhandarkar Oriental Research Institute
Accessed: 03-03-2015 07:25 UTC
Bhandarkar Oriental Research Institute is collaborating with JSTOR to digitize, preserve and extend access to Annals of the
Bhandarkar Oriental Research Institute.
A GLANCE AT THE WORD JATI IN THE VEDIC LITTERATURE
By
Uma Chakravarti
The wordjãti formedthe rootjan 'to be born' or 'cause to be born'

means 'birth', 'production', 'family', 'kind', 'species' etc. Indo-Eropean
gen-tiis the source of Sanskritjãti. In Latin it is gens (kind) and in Avestan
zato (birth).In thisconnectionAvestanfra-zaintis'progency',New Persian
nizãd (birth)may also be referredto.1
The two words : Varna and jãti have been used in theorderto signify
and the like along withtheir
classification,social order,status,stratification
othermeanings related to phonetics.Varna division was initiallybased on
the dark complexion of the aboriginal non-Aryan enemies who were
conquered by the fair-complexionedAryans.2Later along withthe progress
of time,the word varna signifiedthe Vedic people who had been stratified
in fourclasses: brãhmana, ksatriya,vaišya and südra. Except forone mantra
(X. 90. 12), belongingto laterVedic period the words vaišya and südra do
not ocçur in the Rgveda. The word vis, however, occurs many a time in
differentcase-endings and compounds. The four varnas are mentioned,
probably,forthefirsttime3in theMaitrãyaniSamhitã (IV. 4. 6); catvãro vai
purusãh brahmano rãjanyo vaišyah südrah.
We intendtohave a look now atthejãti as itoccursintheVedic literature.

It does not exist at all in the Samhitã literature.It occures in the Aitareya
Brãhmana II. 39 forthefirsttimein thewhole of theVedic literature:'Having
recitedthe silentpraise he recitesthe Puroruc; thushe propagatesthe seed
l MayrhofenVol. 1, 1956,pp.427-8.
2 UmaChakravarty: "A glanceat theword'varna'in theVedicLiterature." Annals
BhandarkarOrientedResearch Pune,Vol. LXXXIV,1-11.
Institute,
3 Accordingto L. vonSchroeder: compared to theTaittriya
Samhitã,Kãthakaandthe
Samhitãs
Maitrãyani areclosertoeachother andhealsoconsidersthesetwoSamhtãs
aretheearliest
literatureoftheyajusperiod,(ed.MatrãyaniSamhitã,1923,p. XVII).
Keith,however,findsall thesethree
workscontemporaneous. (HOS 18,p. XCVI).Our
viewaccordsSchroeder 's.
128 Annals BORI, LXXXVÍ( 2005 )
when developed; first there is development, then birth.' 4 Šatapatha

Brãhmana5 mentionsjãti in the sense of kinsmen.Through the world of
ritualthis Brãhmana bringsus to the real life and tells us about marriage
restrictionsamong kinsmenat thattime : Thus the separation(of theeater
and the eaten) is effectedin one and the same act; and hence fromone and
thesame man springbothenjoyer(the husband),and theenjoyed (the wife):
for now kinsfolk (jãtyãh) live sporting and rejoicing together,saying,
inthefourth(or) thirdman(i.e. generation)we unite.And thisso in accordance
withthat(separationof spoons.)' (Eggeling).
Kãtyãyana Srautasütra(XIV. 2. 32) uses jãti meaning'species' while

instructing the ritualiststo collect animals of different
species needed to be
givenaway as gift(<dakasinã) in theVãjapeya sacrifice.Theyarecows, horses,
goats, buffaloes etc. Each species should be a group of seventeen : jãter
jãtesca saptadasaganapüranät.6 The same ŠrautasútraXV. 4. 7 tells us that
the countries are named after the name of the lineage of the king like
Kuravah, pãftcãlahetc. Apastamba Dharmasütra(II. 6. 1), uses the word
jãti in the sense of family/ lineage ( gotra) etc. and not in the sense of
'caste' characterized by commensalityand conniibiality(cf. The Oxford
Historyof India, impression2001, p. 60). The purportof the textis : If a
stranger approaches a teacher for knowledge and apeais to him with,
all modesty: "O lord, I am here, caste on me your auspicious glacness,
be,propitiatedand teach me." Then, hadrthe teacherdoubts in the lineage
(jãti) and conduct (<ãcãra ) of the person thenhe should makingFire (god)
thewitnessequire: 'O noble (onel) to whichgotrado you belong and how is
yourconduct(ãcãra)!1
4 Aitareya
Brãhmana 11.39:tüsnim šamsam sastvãretastadvikrtam prajanayativikrtirvã
'
agre thajãtih*AlsoseeLãtyãyana Srautasütra X. 1.9,Äpastamba SrautasütraII. 2.4,
Taittiriya
Ãranyaka VI. 5.1 fortheuseofthewordjãti.
5 ŠatapathaBrãhmana 1.8.3.6:tadvãetatsamãnaèvakarman vyãkrijatetasmãd usamãnãd
evapurusãd attãcãdyasca jãyetèidamhicaturthe purusetrtivesangacchãmahe iti/vid
evamdivyamãnã jãtyããsataetasmãd u tatII
6 Alsosee Brhaddevata V. 146;Lätyäyanasrautasütra X.1.9.
7 Äpastamba Dharmasütra 11.6.1: avijnätapürvo yo dharmãrtham adhyayanãrtham
' ' caksusãpašya, sivenamanasã
ãgacched/upasidetupapannosmibhagavan, maitrena
anugrhãna ,prasidamãm adhyãpaya ititasya
jãtyãcãrasamsaye satiagnim upasamãdhãya
yatrakvacidagnimityãdyan yadupadadhyãd ityantam krtvãtatsannidhu jãtimãcãram
ca prcchet,'kim si kim ãcãras cãsi'1
gotro' saumya,
UmaCharra varti : A Glanceat theWorldJãtiintheVedicLiterature 129
Baudhäyana Dharmasütra(1. 1. 10) uses thewordjãti as thepopularly

undershatood:
avratãnãm amantrãnãmjãtimãtropajivinãm
sahasrašca sametãnãmparisat tãm na vidyate
A surveyof the numberof use of the two words varna (= brightness,

complexion, classificationetc.) and jãti (= related to phonetics,part of a
componded word, lineage (gotra) group of people, kinsmen,caste [?] etc.)
mayhelp us in gettinga pictureof thechronologicallyearlieruse of theword
varna thanjãti in the Vedic literature:
VARNA No. of ufce JATI No. of use
Rgveda 23 JATI nil

Later Samhitãs 97 JATI nil
Brähmanas
Aranyakas 63 JATI 3
Upanisads
Vedaňgas 98 JÃTI 99
It is no denyingthe fact thatthe social conditionsas reflectedin the

Sütraliterature, especially,in theDharmasütras,vastlychangedas a resultof
the multifirious typesof mixed marriagestakenplace withinthe varnas and
also outsideitalong withotherreasons.The law-makersstratified thechildern
of these marriages: anuloma (hypergamy)and hypogamy)8, giving them
differentclassified names.9 The word Jãti in the sense of 'caste' as we
understandthe word today,however,does not exist in the Vedic literature.
Evernthewordsanulomaandpratiloma10denotingthemixedmarriages
in relatonto thesocial statusofthemarriagepartnersdo notoccurin thewhole
of the Vedic literatureexceptingthe Sütra literature.
8 Tambiah, S. J.'FromVarnatoCastethrough MixedUnions'inJackGoody(ed.)Char-

acterofKinship, CUP, 1973,p. 194.
9 Baudhayana DharmasütraI. 8.7; 1.9. 1-3;1.9.6-8;GautamaDharmasütra IV. 16-17
etal.
10 Thesetwowordshowever, havebeenusedmanya timeinliteral sensei.e. in'regular
course'and'reverse'.
130 Annals BORI, LXXXVI( 2005 )
References
Press,Poona.(Eng.tr.)
Anandashram
Brãhmana,
Aitareya A. B. Keith,H O S 25.
Dharmasütra
Apastamba ed. MahadevaShastri,
Govt,ofMysore.
Dharmasütra
Baudhayana ed. Narendra Delhi:Vidyanidhi
Kr,Acharya, 1999.
Prakashan,
H O S 5,6.
Brhaddevatã
Samhitã
Maitrãyani vols.I & II (ed.)L. VonXchroeder, 1923.
OttoHarrassowitz,
Leipzing:
Manfred,
Mayrhofer, 1956,Kurzgefasstes Etymologisches vol.I CarlsWinter:
Woerterbuch
Universitaets
verlag.
Institute
vol.I (ed.)Research
Brãhmana,
Šatapatha Scientific
ofAncient NewDelhi,
Studies,
1967.
Samhitã
Taitirlya Press,Poona(Eng.tr.)A. B. Keith,HOS 17,18.
Anandashrama
Social Scientist
Social Stratification in Ancient India: Some Reflections

Author(s): Vivekanand Jha
Source: Social Scientist, Vol. 19, No. 3/4 (Mar. - Apr., 1991), pp. 19-40
Published by: Social Scientist
Accessed: 03-03-2015 07:00 UTC
Social Scientist is collaborating with JSTOR to digitize, preserve and extend access to Social Scientist.
VIVEKANAND JHA*
in AncientIndia:
Social Stratification
Some Reflections**
WiththeIndus scriptstillundeciphered,1 in spiteofnearlysixtyyears

of excavationsand substantialevidencerelatingto variedaspectsof
life,muchof thereconstruction ofthesocialorganization oftheBronze
Age Indus Valley or Harappa Culture,2coveringpartsof Punjab,
Haryana, Sindh, Baluchistan,Gujaratand fringesof westernUttar
Pradesh,3 during the third-secondmillenniaBC,4is hypothetical.
Though its exact relationshipwith the pre-existing culturesin the
Indo-Pakistansubcontinent has not been established so far,5its
emergenceas a result of a West Asian stimulus is being widely
discounted now,6 and it is regarded sui generis,7an indigenous
developmentamongpeoplesofmixedoriginand diverseracialtypes,8
who had residedin theIndusValleyforcenturies.
Scholarshave distinguished Early,Matureand Late periodswithin
the chronological frameworkof the Harappa Culture,9with
pronouncedrural traitsin its Earlyand Late periods,10and a high
level of urbanismin its Matureperiod.11Awide agrarianbase with
surplusfoodgrains producedby thepeasantsin thecountryside being
stored in the granariesat Harappa and Mohenjo-daroto feed the
sizeable non-food-producing urbanpopulation;12 and extensivelong-
distancetradewithWest Asia,13flourishing tradewithBaluchistan,
Afghanistanand the CentralAsian region,14 and a lively internal
trade,both regionaland inter-regional, characterizedthisperiod.15
We also comeacrossdevelopedcopperand bronzetechnology; a highly
professionalblade industry;adequate use of objectsof silver,gold,
preciousstonesand faience;crafts likebead making,shell working and
* IndianCouncilofHistoricalResearch,
New Delhi.
** Thisis a slightly Addressto theAncientIndia
modifiedversionof thePresidential
sectionof the 51stsessionof the Indian HistoryCongressheld in the University
of
Calcuttain December1990.
SocialScientist,
Vol.19,Nos. 34, March-April
20 SOCIAL SCIENTIST
ivorycarving;wheel-turned, well-firedpottery,generallyplain,but
also oftenprovided witha slip over whichare painteddesignsin
black pigment;hand-modelledterracottasof gracefulhuman and
animal figurines;bricklayingand masonryon a vast scale; and the
manufactureof exquisiteseals, cottontextiles,boats,carts,etc.16A
large numberof full-time city-basedspecialistartisansproducinga
varietyof articlesof high artisticmeritforthe relativelyaffluent
privilegedstrataas well as forexport,theirruralcounterparts also
engagin1 in craftsrelatingto stone,clay, shell,bone, metalsand
textiles;1and a substantialworkforcecomprising wood-cutters, fuel
burners,grain-pounders,carters,streetand drain cleaners,waste
removersand slaves18are among the otherdistinctfeaturesof the
Matureurbanphase oftheHarappaCulture.
The overwhelmingimpressionis thatof a highlycomplexsocio-
economicstructure withthe cityholdinga centraland commanding
positionvis-a-visthe countrysidewhichit dominatedand exploited,
and a definitestratification along class lines withinthe cityitself
withthe privilegedrulingeliteenjoyingunequal wealth,powerand
prestigein relationto themassofcommonpeople.'9Thereareno doubt
seriousdifferences of opinionregardingtheactualcomposition of the
rulingclass. To take onlya few examples,V. GordonChilde thinks
that a 'ruler' dwelt in the citadels at Harappa and Mohenjo-daro;
includesamong the rulingclass princes,priests,merchants, officials
and scribes;and maintainsthatsuperstitions musthave played an
enormousrolein consolidating and maintaining socialinstitutions
and
economicarrangements.20 StuartPiggott,A.L. Basham,D.D. Kosambi,
Bridgetand RaymondAllchin,Ildiko Puskas and IrfanHabib are
inclinedto agree.Thus Piggottspeaksof a stateruledby priestkings,
wieldingautocraticand absolutepower,controlling productionand
distribution,and levyingtollsand customs.21 Bashamrefers toa single
centralizedtheocratic stateand continuity of government throughout
the lifeof the civilisation.22Kosambi pinpointsthe curiouslyweak
mechanismof violenceand theuse of religionas an ideologyby the
dominantprieststo extracttribute fromthetraders(who wereallowed
freedom to amass considerable wealth on their own) and to
appropriatesocial surplusand maintaintheclass structure.23 Bridget
and RaymondAllchinattestthepresenceof priestkingsor a priestly
oligarchy who controlled the religious life, economy and civil
government and functioned as administratorsas well.24Puskaslocates
supreme power in the priests'hands.25IrfanHabib underlinesa
combination ofgods,superstitions and priestsbindingtherulersand the
ruled alike in an awesome dread of change.26R.S. Sharma,on the
otherhand,excludesthepriestscompletely fromthecategoryofrulers
and gives the pride of place to the traders.27 Withoutcategorically
refutingthe likelihoodof priestswieldingpower,K. Antonova,G.
Bongard-Levinand G. Kotovskyalso visualize the possibilityof
SOCIAL STRATIFICATIONIN ANCIENTINDIA 21
power in the Harappan cities being of a republican,oligarchic

variety.28There is disagreementabout the actual positionof urban
craftsmen as well. While Puskas29and R.S. Sharma30have no doubt
regardingurban craftsmen being a part of the rulingclass, Childe
associatesthemwiththemodesturbandwellingsof thelowerstrata,31
even thoughtheyweretoa largeextentproducing'forthemarket'.32 In
a recent article Massimo Vidale expresses the opinion that craft
productionwas under thepoliticalcontrolof the urbanelitesof the
Harappa Culture.33Antonova,Bongard-Levin and Kotovskyfindthe
presenceof both impoverishedand prosperousartisanswithinthe
precinctsof the cities.34The class characterof the MatureHarappa
Cultureis, however,generally recognized.35
Stretchingthe evidence to make out the existenceof caste and
untouchability as well in the Harappa Culture,on the otherhand,
does notseemto have adequatebasis. IravatiKarve,a sociologist, first
referred to theprobability of 'somethingverylike castes'at Harappa
and a streetexclusivelyoccupied by a 'caste-likegroup' whichhad
specialized in pounding rice there.36She also loosely spoke of
untouchability as a characteristic of the caste structurefromtop to
bottom.37 Followingher,S.C. Malik,anothersociologist, imaginedthe
'roots' of caste and the 'perpetuationof caste status by birth'in
Harappan society.38 'Casteclass patterns', in his opinion,developedin
thesocio-economic organization at Harappa and theincomingAryans
adoptedthemin theprocessofbeingIndianized.39 ThatMalikis notat
all seriousabouttheuse oftheterm'class'heremaybe gaugedfromhis
referenceto 'the emergenceof complex socio-economicclasses'
comprising'therichand thepoor' in theHarappa Culturealongwith
theclarificationthat'thisis notin thesenseofclassconsciousness or an
interclassstruggle'.40 And Malik'sreference to castein theHarappan
contextevoked fromA. Ghosh,a muchmoreperceptive, matureand
balanced scholar,the apt commentthatsuch hurlingof institutions
fromthe knownto theunknownto suggesttheiroriginand bringing
themdown fromtheunknownto theknownto provetheirpersistence
does not carry conviction.41Suvira Jaiswal,too, questioned the
proprietyof Malik's tagging caste with the existence of class
differences reflectedin thesettlement patternof theHarappancities,
which survivedeven afterthe citiesthemselveshad disappeared.42
The concentration ofvariouscrafts in specificquartersor streetsbeinga
normalfeatureof theOrientaltownsup to thepresentday has been
underscoredby several scholars.43Significantly in his editedwork,
Determinants of Social Statusin India,4 whichis presumedto reflect
'a multi-facetedtrans-disciplinary approachto thestructure of society
from ancient history to contemporary times in the Indian
subcontinent',45neitherMalikin his Introductory parer'Determinants
of Social Status in India: Problems and Issues',F6nor any other
contributor makesa singlereference to castein theHarappancontext.
22 SOCIAL SCIENTIST
That Malik's view of Historyas a disciplineis singularlynarrow-

almostmyopic-is evidentfromhis observation in theIntroduction to
anotheredited work,Indian Civilization:The FirstPhase-Problems
of A Sourcebook,47 that historical generalizations are normally
arrivedat by scrutinizing individualfactsthatare not amenableto
any rational systematicanalysis and that historiansformulate
generalpropositions, rathertentatively,citinga fewinstances.48
Among the serious and eminenthistoriansRomila Thapar has
supportedthehypothesisput forwardby Karveand Malikregarding
casteand untouchability in someofherwritings. Avowedlytaking'the
help of social anthropology'-nothistoricalevidence-she notesthe
probabilityof caste as a pre-Aryansystemand part of the social
in theHarappa Culture,wherea smallgroupmayhave
stratification
preserveditselfthroughstrictendogamousmarriage,claimedritual
purity and higher status, and also exercised authorityin a
hierarchicalsocialset-upbased on a divisionoflabourwiththenotion
of pollutionattachingto certaingroupsof menialworkers.49 She even
talksof 'servicerelationships'on thepatternof thehereditary jajmani
systemand findsin castetheanswerto thevexedquestionas to who
was in authorityand how that authoritywas maintainedin the
Harappa Culture.50The detailcd reconstruction in the absence of
corresponding tangibledata in thematerialfindsto supportitcan only
be regardedas exceedinglyspeculative,foralthoughthe notionof
ritualimpuritymay be exemplifiedby the GreatBath at Mohenjo-
daro and theexistenceof separatequartersforgrain-pounders (a non-
polluting vocation in itself)at Harappa is admitted,evidence of
divisionoflabourand occupationalgroupson thebasisofritualpurity
and impurity is wantingand thereis no traceofthecontinuity ofcaste
and untouchabilityin either the Late Harappa period or their
transitionto the Rigvedicperiod.51The materialculturein thepost-
Harappan Vedic period(c. 1500-500BC), forwhichwe have copious
literaryevidence in the Samhitas,Brahmanasand Upanisads (they
representa sortof transition fromprehistory to history),is glaringly
differentand comparatively muchless advanced.Thislongperiod,too,
can be divided intotwo distinctphases-Rigvedic(c. 1500-1000BC)
and Later Vedic (c. 1000-500 BC). The Rigveda, comprisingten
mandalas(books),is widelyacceptedas containing twobroadstrataof
historicallayers,theearlierrepresented by BooksII-VII and thelater
by BooksI and VIII-X.
The hymnsportraythe Aryansas firstand foremosta warlike
people driving horse-drawnchariotsand using weapons of ayas
(copperor bronze)effectively againsttheirnon-Aryan foesin theland
of the seven rivers.Professionalfighters organizedin separatetribal
groups,theycontinuouslyfoughtinternecine wars as well. Bootyor
spoilsof war (lotra)52formedan important meansof theirlivelihood.
Theyalso engagedin primarily pastoral53and subsidiaryagricultural
pursuits,theiragriculture beingsubsistence-oriented. Usingox-yoked

plough (langala,sira) with wooden ploughshare(phala) and such
other implementsas hoe (khanitra),axe (parasu, vrikna),sickle
(datra,srini),etc., theyproducedyava (barley).54Leatherstrapfor
the plough (varatra), furrow(sita, sunu), termsforfield (ksetra,
urvara) and ploughman(kinasa) are mentioned.Since most of the
references to agriculture are foundin theadmittedly lateBooksI and
VIII-X and a hymnin Book IV (57.1-8), which is consideredan
interpolation,agrarianeconomyobviouslystabilizedtowardstheclose
of theRigvedicperiod.
The absence of evidenceforthe sale, transfer, mortgageor giftof
land or its disposal in any otherway makesindividualownershipof
land at thisstagedoubtful.The kin-basedtribe,Rigvedicvisorjana,55
whosememberswerenormallyon themoveand temporarily dweltat
one particularplace, appears to have collectivelyowned both the
cattleand the land and worked togetherin the fields.Iron being
unknownat thisstageand tillingthevirginlandbeingbyall meansan
arduoustask,thewholetribetoiledto producebarley.
Scholarsdifferseriouslyabout theuse of hiredextra-tribal labour,
especially slaves, in these limited agriculturaloperations. R.S.
Sharmafirmly and consistently arguesin his writingsthatslaveryin
theRigvedawas purelydomestic,thatslaves,mostlywomencaptured
in wars,wereused forreplenishing thedepletingAryanranksthrough
begettingchildrenand forhouseholdchores,and thatthe Rigveda
does not have an termforwage or wage-earners.56 A.A. Macdonell
and A.B. Keiths and Basham, on the otherhand, assertthatthe
Dasas werein manycases reducedto slaveryand hencetheworddasa
has the sense of 'slave' in severalpassages of theRigveda.P.V. Kane
in his monumentalHistoryof Dhrmasastra cites a Rigvedicpassage
referring to thegiftofa hundreddasasin thesenseof slaves.59In Dev
Raj Chanana's opinion the Aryansmust have knowndebt slavery,
slaveryas a resultof defeatin gamblingand war slaverybeforetheir
adventintothiscountry and quitea fewDasas ma havebeenenslaved
and almost any servicedemanded fromthem.60Kosambiholds the
view thatjustas cattlewereherdedin commonand fieldsweretilled
in common,thedasa was also used as commontribalproperty in the
Rigveda.61Romila Thapar presupposesthe employment of non-kin
dasa labourin thehouseholdingeconomyof theRigveda.62 According
to IrfanHabib, the dasas workedlike cattleon the fieldand tended
the herds.63R.N. Nandi,too,atteststheuse ofnon-kindasa labourers
in barleyfields.64
Only a fewcraftswere practisedin thepost-urban ruralmilieuof
the Rigveda,which mentionsthe karmara(metalsmith), the taksan
or tastri(carpenter),the carmamna(tanner)and thevaya (weaver).
The karmarasmeltedthe metal ore in fire(hence the designation
dhmatri or smelter)and madehouseholdutensils, toolsand weaponsof
24 SOCIAL SCIENTIST
copperor bronze.The taksanor tastribuiltchariotsforwar,65prepared

ladles and bedsteadsof wood and engaged
vessels,buckets,sacrificial
in carving.The carmamnatannedhide (carman) fromwhich were
manufactured bowstrings,slings,reinsandbags.Cottonbeingunknown,
thevaya (weaver) preparedwoollencloth.Commonwordsforthese
artisansin severalIndo-EuropeanlanguagessuggestthattheRigvedic
Aryansdid notlearnthesecraftson theIndiansoil.The Rigvedicword
forpotter(kulala)has, however,no parallelsin otherIndo-European
languages; this may indicate the adoption of local traditionsin
pottery.
The productsof craftsmen's labourand skilldo notappearto have
beenmeantforsale or collection in theformoftaxesand in viewofthe
unavailabilityof surplusto supportthemtheyprobablyengagedin
food productionas well. The evidenceof a Rigvedichymncomposer
callinghimselfa poet,his fathera physici.anand his mothera grinder
of corn66suggeststhatthe indispensabledivisionof labourhad not
advanced beyond a point and specialization had not become
hereditary. In view of theusefulnessof theirwork,theartisanswere
respectedmembersof theAryanvis. Characterizing theirrelationship
withothermembersof thetribeas jajmani67is perhapstoo bold.The
economywas notyetfullyor even primarily agrarian;craftsmen and
peasantswerenottwocompartmentalized categories;and theirmutual
relationshipin a semi-sedentary set-upwas nothereditary, subsisting
fromgeneration togeneration.
Also thesurplusproducedin thispredominantly pastoralRigvedic
economy was not substantialenough to underminethe broadl
egalitariantribalstructureor to lead to thedevelopmentof classes.28
The Rigvedictribalchiefhad hardlyany regularor fixedsourceof
incomein theformofcerealor cattleon whichhe could flourish along
withhis priests.The termbalioccursin theRigvedain thesenseofa
voluntaryoffering or presentfromtribesmen to theirchief.Occasional
exactionof tributesfromtheconqueredpeopleand spoilsofwarwere
theothersourcesofhisincome,butsinceresourceswerenotadequateto
maintaina regulararmy,he had to sharethesein periodiccommunal
sacrificeswith membersof the tribewho formedthe militia.The
institutionof mutual giftsalso checked the growthof economic
disparityand therewas no leisuredclass livingoffthe surplusof
others.Differentiation withinthe tribehad, however,begunand the
tribalchiefsand priestswerenotonlyclaimingand enjoyingsuperior
ranks,theyalso receivedthemajorshareof the spoils of war in the
formof slaves, animals, weapons and ornamentsand enjoyed a
somewhatbettereconomicposition.
The termvarna occursin theRigveda a numberof timesand is
initially used to distinguish Arya from Dasa and Dasyu. The
differencemay initiallyhave been both ethnicand cultural.69The
wordsbrahmanaand ksatriya occurfifteenand ninetimesrespectively
in thetext,oftenin thesenseofopenand fluidfunctional categories,

but
the word varna is never used in connectionwith them.Neither
kingshipnorpriesthoodappearsto have dependedmerelyon birthor
heredityat thisstageand thereis no evidenceofrestrictions as regards
partakingof food or marriage.The sudra occurs only once in the
purusasukta70 in Book X and thatalso along withthebrahmana,the
rajanya and the vaisya (the two lattertoo occur forthe firsttime
here). Here the sudra is statedto have sprungfromthe feetof the
Purusaor PrimevalBeing,unlikethebrahmanawho sprangfromthe
mouth,the rajanyawho sprangfromthe arms and the vaisya who
sprangfromthethighs.Significantly thewordvarnais notused in this
context.Also thereis a consensusamongscholarsthatBookX belongsto
thelateststratumof theRigvedaand virtuallysynchronizes withthe
Later Vedic texts.Evidentlyvarnain theRigvedadid not have the
sense it came to acquire later. The fourfoldvarna system had
definitelynotbeenbroughtto Indiabyanygroupor wave ofAryans.It
was an indigenousdevelopmentand was nota realityin theRigvedic
period.71Certainlycontactwithpeoplesis nottabooedand thereis not
evena semblanceofuntouchability in thetext.72
The LaterVedic period(c. 1000-500BC), information pertainingto
which is based not only on the post-Rigvedic texts,but also on the
archaeologicalfindsof thePaintedGreyWare culturesynchronizing
both in time and regionwithit,73furnishesevidence of all-round
materialprogress.Victoriesin wars and penetration to new areas in
the east broughtlarge tractsof UttarPradesh,northBihar,partsof
Rajasthan,besides Punjab and Haryana, under the political and
culturalsway of the Aryans.Forestswereextensively burntand land
was cleared for cultivationto meet the needs of an expanding
population.Economybecameprimarily agrarian.Besidesbarley,rice,
wheat,millet,lentils,severalkinds of pulses, sesamumand linseed
were produced.References to theuse of largeand heavyploughs,to
whichsix,eight,twelveand even twenty-four oxen wereyoked,and
paviravanior pavirava in the sense of metal ploughshareoccur in
Later Vedic literature.The lone evidenceof iron ploughsharefrom
Jakherain Etah districtof UttarPradeshhas, however,been assigned
by R.S. Sharma74to theclose of theperiod.It seemsthathardwood
(khadira,udumbara)ploughshareswere used fordeep diggingto get
betteryield. Intensivecultivation,applicationof manure (karisa,
sakritand sakan),practice of irrigationand betterknowledgeof
seasonscontributed to increasein production and sufficientsurplusto
make possibletheemergenceof classes.Ownershipof land devolved
fromthe tribeto familiesunderpatriarchal heads and theprocessof
disintegration oftheAryantribesbegan.
Crafts,too, proliferated.The rathakara, distinctfromthe taksan,
appearedas a professional craftsman forthefirsttime.The profession
ofkarmara becameenormously important, forhe beganto manufacture
26 SOCIAL SCIENTIST
iron(syamaor krisnaayas)artifacts on a modestscale about900 BC or

a littlelater.75Althoughtheearlyuse of ironwas largelyconfinedto
weapons of war such as spearheads,arrowheads,etc.,foundat the
excavated Painted GreyWare sites,the role of ironaxe in clearing
forestsfor cultivationin thickvegetationareas of the middle and
lower Ganga basin in the seventhcenturyBC is incontrovertible.76
Ayastapa(heaterof metal,ironor bronze)and kosakari are mentioned
in the texts;bellows appear to have been used at Atranjikhera; and
twofumacesforsmelting ironand forging objectsfromitwerefoundat
Sunerivilla e in Jhunjhunu districtof Rajasthantowardstheclose of
theperiod. Increasein ksatriyaor rajanyapowerand transformation
of the Rigvedictribalchiefsinto relativelystrongmonarchsruling
overthefirstterritorialkingdomsof theperiodmayhave beendue to
theirexclusivepossessionand use ofironweapons.78Extensiveuse of
bows and arrowsnot only in wars but also in huntingled to the
development of the specialized crafts of the bowmaker
(dhanvakrit/dhanvakara), arrowmaker(isukrita,isukara)and the
makerof bowstring(jyakara).The jeweller(manikara)and workerin
gold (hiranyakara),too, make theirappearance. The textsfurnish
moredetailsaboutleatherwork(carmanya). Predominance of women
in weaving,dyeing,embroidery, basketmakingand thornworkingis
reflectedin terms like vayitri,rajayitri,pesakari,bidalakariand
kantakikari respectively. Washing, too, had given rise to a
professionalcategoryin which both men (malaga) and women
(vasahpalpuli)participated.Some of the craftsmen and craftswomen
apparentlybelonged to non-Aryansegments.The rathakara,the
taksan and the karmara were treated with utmost esteem and
considerationowing to the immensevalue of theircraftsforwars,
agricultureand general social comfortsin a predominantly rural
setting.The kingvisitedtheirhousesto perform certainceremonies in
course of sacrificesto show themrespectand ensuretheirsupport.
Thereis definitely no traceof any stigmaattaching to anycraft.Some
of thesecraftsmayhave tendedto becomehereditary, thoughthereis
no textualreference to occupationaljatisat thisstage.
Division of labour and specializationof functions evidentlymade
definiteheadwayduringtheperiod.Agriculture becametheprimary
concernof the vaisyas,themostnumerousof the fourvarnaswhich
developed during the period. Cattle-rearingwas a secondary
occupationforthem.Some craftsmen, too,mayhave belongedto this
category.The periodalso saw therise of thefourthvarnaof sudras
fromthe conquered aboriginesand the defeatedand dispossessed
sections of the Aryans. Although occasionally wealthy cattle-
owners,79thesudraswereby and largeless well-off thanthevaisyas
and engagedin theserviceoftheupperclasses.References tothesudra
being dedicatedto toil (tapase)in thepurusamedha (symbolichuman
sacrifice)in the Vajasaneyi Samhita,80 TaittiriyaBrahmana,81 and
SatapathaBrahmana82 pointto his belongingto theclass of labourers.

Accordingto Kosambi,thenew organization ofsocietymadeavailable
for the firsttime a supply of labour whose surplus was easily
expropriated and theplace of slave was takenby thesudra.83In Irfan
Habib's opinion,theuse of heavyploughswithoutironploughshares
implied as its inevitable corollary the employmentof servile
labourers.84 The brahmana as a professional priest and the
kastriya/rajanyaas warrior/rulerhad also emerged as specific
varnas,leadingto theformation of thefourfold varnasystemduring
the period.
Reflecting the emerging social stratificationas a result of
progressivedivisionof labour,specializationof functions and growth
of surplus,the varna systemwas fromthe beginninghierarchical.
Withbirthand hereditybecomingincreasingly important factorsin
this divisionof labour and specializationof functions, jati (literally
caste) also developed duringthe period,the termfirstoccurringin
Yaska's Nirukta85and beingapplied to a womanof the sudracaste
Varnawas in essenceexploitative
(sudra-jatiya). in natureand content.
Thereare crudestatements to theeffectthatthevaisyaand thesudra
are to be exploitedforthe advantage of the rulingclass with the
brahmana priest's active cooperation and help. The Aitareya
Brahmana characterizesa vaisya as anyasya balikrit,'tributaryto
another',anyasyadya,'to be eaten or lived upon by another',and
yathakamajyeya,'to be oppressedat will', and a sudra as anyasya
presya,'to be expelled at will',and yathakamavadhya, 'to be slain at
will'.86Sacrificesare consciouslydesignedto help rulersovercome
internalconflictsand to makethevaisyaand thesudrasubmissive.87
Brahmana-ksatriyaclaims and counter-claims to supremacy
notwithstanding, theirdistancefromthevaisyaand thesudrain the
emergingclass structure was growing.The former twojoinedhandsto
repress and exploit the vaisya and the sudra. In fact,brahmana-
ksatriyacollaborationis regardedas indispensableand vitalfortheir
mutualwell-beingand prosperity in severaltexts.The crucialroleof
thebrahmanas,withmoreor less completecontroloverritualsand the
Vedic lore, in theoreticalformulationsfacilitatingthe process of
tribal disintegrationand class formationis transparent.Their
invaluablesupportto buttressthetemporalauthority entitledthemto
giftsfromtheruler,visamatta (eaterof thevis) becomesone of whose
epithets.Taxes collected in kind throughkinsmen(sajata) of the
monarchbecamenow theprimarysourceofextraction oftheavailable
surplus. Varna division thus approximatedto class division.88
Productivityin the pre-ironagriculturephase, however,not being
high,the materialbasis of this class divisionwas weak. The tribal
bonds were,therefore, notcompletely sundered,89 and thevaisyasnot
onlyformedpartof thetribalmilitia,butalso receivedan honourable
place in rituals,a SatapathaBrahmanapassage even ordainingthat
28 SOCIAL SCIENTIST
the ksatraand the vis should eat fromthe same vessel.90A full-
fledgedclass societyand statewithsubstantialappropriablesurplus,
regular taxation, army, administrativeapparatus and monetary
economydevelopedonlywhentheuse ofironin agriculture and crafts
becamecommonin thepost-Vedic period.
III
Few historianshave written morecomprehensively and adequatelyon
the problem of social stratificationin ancient India than my
distinguished teacher, Professor R.S. Sharma. D.D. Kosambi's
pioneeringstudies and brilliantinsightstouch the core of several
themeshandledby himin his booksand numerousarticles.91 Romila
Thapar is full of freshideas and her writingsshow a remarkable
awareness of the latesttrendsand developmentsin disciplineslike
sociologyand social anthropology. B.N.S Yadava's masterly use of a
wide range of originalsources in his book Societyand Culturein
NorthernIndia in the TwelfthCentury92and articles is worth
emulationbyeveryyoungresearcher in Indianhistory. B.P.Mazumdar,
Suvira Jaiswal,R.N. Nandi and a host of otherhistoriansincluding
thosefromthesouthhaveenrichedourunderstanding ofcasteand class
in theancientIndiancontext.Attempts to understandthepatternsof
social developmentin different regionsof thecountryin thepastand
regional studies of the problem of social stratification are truly
commendable, thoughthereis need and scope formuchmoreworkin
this area. Among the medieval Indian historiansno one covered
variousbranchesof ancientIndian historyin as muchdetail as the
present Chairman of the Indian Council of HistoricalResearch,
ProfessorIrfanHabib.The Anthropological SurveyofIndia has under
its 'People of India' project in course of 1985-90 compiled and
computerized thelatestdata on 4,384communities in all theStatesand
Union Territoriesof India including 426 Scheduled Tribes,443
ScheduledCastes (quitea fewof thesewereneitherin thepastnorat
presentare regardeduntouchables in different partsofthecountry) and
1,051BackwardClasses in 120volumeswhichwillproveinvaluableto
researchersin history.Dr K.S. Singh,thepresentDirector-General of
the Survey,and the scholarswho have assisted him in this major
academic endeavour deserve all compliments.Historiansin this
countryneed to look up more carefullythe good work done by
sociologistslike M.N. Srinivas,Andre Beteilleand G.S. Ghuryeas
well as theirWesterncounterparts to have a fewusefulinsightsfor
theirown researchesin history.I have myselfwrittena fewlengthy
articleson someuntouchable groupsand thedespicablephenomenon of
untouchability in theancientperiod.I shall notattemptto coverthe
entiregamutof social stratificationin post-Vedictimesup to AD 1200
in this briefarticleand I shall draw your attentionto only a few
aspectsof thisproblemin a generalway.
One majordevelopment in theLaterVedicperiodwas thebeginning

of the process of assimilation of forest-dwellingtribes on the
periphery of theimmigrant Aryansettlements. Quitea fewsuchtribes
are mentionedin the texts. Among these are the Nisadas,93the
Candalas, the Paulkasas, the Andhrasand the Kiratas.It is not that
all thesetribesand manyotherswho are referred to in latertextscame
intoequllyclose contactwiththeAryans.Butsomeof thosewho did
and had poor material backgroundfared badly in the unequal
encounter.In facttheywere among the firstpeoples who became
tabooed and were subsequentlydamned as untouchables.It is a
historicalfact that when untouchability firstappeared in the full-
fledgedclass and caste societyof thepre-Mauryan post-Vedictimes,
theywere thefirstvictimsto be relegatedto therituallylowestsocial
position.I have in mindthe well-known-rather notorious-casesof
the Candalas, Mritapas,Matangasand laterSvapakas,Dombas and
others.These were the originalinhabitantsof the countrywho are
known to have belonged to the Munda-speakingProto-Australoid
ethnictype.94One theoryabout theoriginofcasteand untouchability
is that these were pre-Aryaninstitutions95which the Aryans
themselves imbibedfromthem.Thisis simplynottrue.
Sinmlarly thereis no basis to suggestthatthe caste systemin our
countryfirstoriginatedamongthe Dravidiansin thesouthand then
percolatedto thenorth.96 The southdevelopedthephenomenon only
as a result of the impact of northIndian cultural and political
contact.97 Fromwhatwe knowof theHarappansand theAryansit is
clear thattheycannotbe equated so faras theircontribution to caste
and untouchability in thiscountryis concerned.Thereis no positive
evidenceforuntouchability at Harappa and theAryansdid notbring
the institution but developed it on Indian soil a fewcenturiesafter
theiradvent.
Ideologyand forcewere both systematically employedto slowly
develop caste and untouchabilityin this country.The notion of
pollutionin relationto certainsocialgroupswas firstelaboratedin the
Dharmasutrasof Apastamba,Gautama, Baudhayanaand Vasistha,
thenin the Smritisof Manti,Visnu,Yajnavalkya,Narada, Brihaspati
and Katyayana,and still later in the early medieval Dharmasastra
and Nibandhatexts.98 Evenseculartexts(byancientIndianstandards)
like theArthasastra of Kautilyafellin line.Detailedrulesand norms
wereprescribedregardingmarriage, food,associationand contactand
thosewho violatedthem-unlessof coursetheywere materially and
politicallystrong-werein forserioustrouble.One carefullook at the
institutionof outcastes(patita-they differedfromthe untouchables
in not being permanentor hereditary-inDharmasastraliterature
would show that they were to be no less severelypunished for
violation of prescribed norms and intimate contact with the
untouchablesthan the untouchablesegmentstill theyrelentedand
30 SOCIAL SCIENTIST
observedthe ritualsof penitenceand redemption. Theywere to lose

inheritance;even theirwife and childrenwere expectedto disown
them;and ofcourseforthesocietytheywouldsimplyceasetomatter.
The king's danda (coerciveauthority)was to be applied forthe
observanceof caste rules99and several inscriptions bear out royal
claims to followthe Dharmasastradirectivein thisregard.In their
own way theepicsand thePuranaslentsupportto castenorms,whose
essence lay in its institutionalizedinequality.100The brahmanasand
the ksatriyasand all those who could by virtueof theirpower,
resourcesand positionjointheelitegroupand had no rolein primary
productiveactivitiesbenefitedfromthe system;for its apparent
rigiditynotwithstanding, caste always retainedtherequisiteamount
of flexibilityand an attitude of accommodation.Many foreign
invadinghordeswereassimilatedas highercastegroups.Even when
indigenoustribesbrokeup,thebestamongthemcouldbe accommodated
as priestsor evenas rulers.
The Dharmasastrawritersemployednew theoreticalconceptsto
explainthe social phenomenon.One such conceptwas thetheoryof
varnasamkara,l?lwhich was used to explain the statusof several
emerginggroupsand theuntouchability ofsectionsliketheCandalas,
the latterbeing simply regardedas the lowest pratilomacaste-
offspring ofa hypogamousunionbetweenthefourth varnaofsudraand
a brahmanawoman.
The notionofjatyupakarsa(upward mobilityof a caste) owingto
marriagein a highervarnaor pursuitofan occupationprescribed fora
highervarna continuouslyforfiveto seven generations102 does not
appear to have been valid withrespectto the Candala. Downward
mobility(jatyapakarsa)was, however,possible in the case of other
theoretically pratilomacategoriesthroughmarriagein a lowervarna
orpursuitofan occupationprescribed fora lowervarnacontinuously for
five to seven generations. As M.N. Srinivas points out, the
untouchablesdifferfromtheotherlow castesin that,unlikethelatter,
the formerhave no means of pushingthemselvesup in the caste
hierarchy and even Sanskritizationdoes nothelp.
It is significantthatthe Sanskrittermasprisyaforuntouchability
was firstused in theVisnusmriti,103 a textofthethirdcentury AD, and
the phenomenonexistedforlong withtermslikeanta,antya,antyaja,
antyayoni,antyavasayin,apapatra,abhisasta,etc.
The classicalvarnatheorydid nothave any place fora fifth varna,
thoughin his commentary to the Brahmasutra, I. 4.12,Samkaracarya
(earlyeighthcentury)shows awarenessof a school of thinkers who
regardedthe Nisada as a fifthvarnaand the Samba Purana,66. 10
(sixth-eighth centuries)mentionsthe fifthvarna.Untouchability was
evidentlyconsideredan integralpartof thevarnasystem.
It is not true that Buddhismtriedto confrontthe caste system
squarelyor soughtto destroyit.104 Caste was denounced;brahmanical
superioritywas challenged;divinesanctionbehindit was questioned;

it was notpermitted withintheSamgha(Orderofmonks);but Buddha
did notseekto weed itout fromthesociety.Castewas partand parcel
of theprevailingmodeofproductionwhichbenefited thehavesat the
cost of the have-notsand Lord Buddha was perceptiveenough to
broadlyaccepts this social reality.Mahaviraand Jainism, too, went
along a similarline.
Beef-eatinghad nothingto do withtheoriginof untouchability.105
It was not prohibitedin the Dharmasastratextsuntil the early
medievalperiod.
Bhaktisucceededin relaxingtherigoursof casteto someextent.106
Lokayata,Tantraand theSahajiyaswereopenlyhostileto caste and
did notdeterminate againstthelow orderbut theyfailedto dislodge
castefromitsentrenched positionin organizedsociety.107
That the practiceof untouchability was immediatelyconnected
withexcessiveand abnormalnotionsofpurityand pollutioncannotbe
denied, but then this is also true that caste did not develop in
primitivesocieties where these notionsare found. Varna in India
provided a frameworkfor theirgrowthand systematization and
projectedthroughthem the dominantmaterialrelationsin ritual
terms.The ideologyofpurity/pollution was surelyused to assignlow
position, segregate and hereditarilyexploit a large segmentof
population.
That therewas periodicallystiffresistanceto caste oppressionis
reflectedin theaccountsof theKali age in theepics and thePuranas,
which show the discomfiture of the upper castes and an unusual
aggressivenesson thepartofthelowerorders, ? butthetempodoesnot
appear to have been sustainedand continuousenoughto disruptthe
system.
NOTES AND REFERENCES
1. Despite numerous attemptsand more than fortyclaims to success, the

decipherment oftheIndusscript(theunilingual mostlyon sealsand
inscriptions,
some on amulettabletsor evenas scratches on potsherds withnevermorethan
twentyand usually not more than ten symbols,are too short)remainsan
unresolvedissue and does not shed lighton the available archaeological
material. In 'The Study of Society in AncientIndia: A Reorientationof
Perspectives',
Presidential
AddresstotheAncientIndiasectionofthe31stsession
of the IndianHistoryCongress,RomilaThapararguedthatthedecipherment
mustconformto a grammatical and linguisticsystemand the readingof the
mustmakesensein termsofthecontext
inscriptions oftheculture;see Proceedings
oftheIndianHistoryCongress (hereafterPIHC), Varanasi,1969,p. 26. According
to Asko Parpola,who withhis Finnishcolleague,KimmoKoskenniemi, has on
thebasis ofcomputer-aided analysisoftheIndusscriptproducedan impressive
concordanceof theknowninscriptions, A Concordanceto theTextsin theIndus
Script(University
of Helsinki,1982),and withSimoParpola,Seppo Koskenniemi
and PenttiAalto as co-authorswrittenTheDecipherment of theProto-Dravidian
Inscriptionsof the Indus Civilization (hereafter Decipherment; The
ScandinavianInstituteof Asian Studies,Copenhagen,1969),theIndusscipt is
32 SOCIAL SCIENTIST
fullof challenging problems;see 'InterpretingtheIndusScript',in B.B. Lal and

S.P. Gupta (eds.), Frontiers of theIndus Civilization, Sir MortimerWheeler
Commemoration Volume(hereafter IndianArchaeological
Frontiers), Societyand
Books & Books,New Delhi,1984,p. 191. IravathamMahadevan,authorof the
IndusScript:Texts,Concordance and Tables(Archaeological Surveyof India,New
Delhi, 1977),regardstheIndus scriptas one of thesevenpictographic scripts
developedin theancientOrientduringtheBronzeAge (c. 3000-1500BC), the
othersix beingSumerian,Egyptian, Proto-Elamite,Cretan,Hittiteand Chinese,
generallywritten fromtheright, onlyabout7 percentbeingwritten fromtheleft,
having425?25distinctive signsand beingan independent invention; see WhatDo
We Know AbouttheIndusScript?NetiNeti('Not ThisNor That')',Presidential
Addressto theHistoricalArchaeology, Epigraphyand Numismatics sectionof
the49thsessionof theIndianHistoryCongress,PIHC,Dharwad,1988,pp. 600,
604-5, 614. On the basis of his structuraland analyticalstudyof the script,
Mahadevan, like Parpola and his team and a group of Soviet philologists,
ethnologists and mathematicians (G.V. Alekseev,M.A. Probst,A.M. Kondratov,
I.K. Fedorova,B.Ya.Volcokand N.V. Gurov)led by Yu. V. Knorozov,who too
have used thecomputerto bringout 7heSovietDecipherment oftheIndusValley
Script:Translation and Critique(hereafterSoviet Decipherment), edited by
Arlene R.K. Zide and Kamil V. Zvelebil (Mouton,The Hague/Paris,1976),
maintainsthatthelanguagewas Proto-Dravidian and refutesS.R. Rao's theory
put forwardin his TheDecipherment oftheIndusScript(Asia PublishingHouse,
Bombay,1982) thatthelanguagewas an archaicbranchof theOld Indo-Aryan
and thescriptevolvedin twostages,theearlyormaturescriptcomprising 62 basic
signsduring2500-1900BC and thelatescriptcontaining only20 basicsignsduring
1900-1200BC, thechange-over beingfroma logographic-syllabic to a syllabic-
alphabeticscript;see Mahadevan'sreviewarticlein VivekanandJha(ed.), The
IndianHistoricalReview(hereafter IHR), Vol. VIII, Nos 1-2, IndianCouncilof
HistoricalResearch, New Delhi,July1981and January 1982,pp. 59-60,64-66.
2. Culture,observesV. GordonChildein Social Ewlution(Watts& Co., London,
1951,p. 26),is an organicwhole,nota mechanical aggregate oftraits. Ofthetwo
mostimportant and bestknownsitesoftheIndusValley(bothin Pakistannow),
Harappa in Punjab,thoughsmallerin size thanMohenjo-daro in Sindh,being
discoveredin 1921,one yearearlierthanMohenjo-daro, gave its nameto this
culture.There are stillwide gaps in the archaeologicalmaterial,notesA.H.
Dani; see RecentArchaeological in Pakistan,
D)iscoveries Unesco,Paris,and the
CentreforEastAsianCulturalStudies,Tokyo,1988,p. 1.
3. In thepost-Independence periodIndianarchaeologists haveidentified morethan
700 sitesof thiscultureinsidethecountry and excavatedto a varyingdegreeas
manyas 40 of them;see B.K. Thapar,RecentArchaeological Discoveries in India,
Unesco,Paris,and theCentreforEastAsianCulturalStudies,Tokyo,1985,p. 52.
4. SirJohnMarshallfirstestimatedthedurationoftheHarappaCulturefrom3250
to 2750 BC; see Mohenjo-daro and theIndus Civilization(hereafterMohenjo-
daro),Vol. I, Arthur London,1931,pp. 104,106.SirMortimer
Probsthain, Wheeler
dated this culturein 2500-1500BC; see 'Harappa, 1946: The Defencesand
CemeteryR 37', Ancient India,Bulletinof theArchaeological Surveyof India,
No. 3, New Delhi, January1947,p. 82. In The Indus Civilization (3rd edn.,
CambridgeUniversity Press,1968,p. 125; firstpublishedin 1953),however,
Wheelermodifiedhis positionand postulatedthefounding of thenuclearcities
some timebefore2400 BC and theirenduringin someshape in theeighteenth
centuryBC, thesetimebrackets notfittingcloselyand mechanically to theIndus
townsand villagesofall sizes and in all locations.AskoParpolaetal havedated
theHarappa Culturein c. 2500-c.1800BC (Decipherment, p. 3), whileKnorozov
etal have dated its outerlimitsin c. 2200-c. 1750 BC (Soviet Decipherment,
Preface,p. 5). D.P. Agrawalplottedsometwodozenradiocarbon dates,including
those forKot Diji, Kalibanganand Lothal,and, based on uncalibrated dates,
concludedc. 2300-1750BC as themaximum datebracket ofthisculture, thoughat
the individualsitesthedurationof thisculturemighthave been stillsmaller
('Harappa Culture:New EvidenceforA ShorterChronology', Science, Vol. 143,
No. 3609,AmericanAssociationfortheAdvancement of Science,Washington,
1964,pp. 950-51);cf.'HarappanChronology: A Re-examination oftheEvidence',
in D. Sen and A.K. Ghosh (eds.), Studies in Prehistory, Firma K.L.

Mukhopadhyay, Calcutta,1966,pp. 139,147.
5. A. Ghosh (ed.), An Encyclopaedia of Indian Archaeology(hereafter
Encyclopaedia), Vol. I, ICHR and MunshiramManoharlal,New Delhi, 1989(a
posthumously publishedtwo-volume monumental workwhichcontains themajor
findingsof Indianarchaeology in prehistory, protohistory and ancienthistorical
periodduringthelastone 150yearsand encompasses information availableup to
1978),p. 75.
6. A.L. Basham,The Wonder ThatWasIndia(hereafter Wonder), 3rd revisededn.,
Sidgwick& Jackson, London,1967;firstpublishedin 1954,p. 15; GlynDaniel's
Preface to Childe, Man Makes Himself,Watts& Co., London, 1965; first
publishedin 1936,p. xii.
7. A. Ghosh,TheCityin EarlyHistorical India(hereafter City),IndianInstitute of
AdvancedStudy,Shimla,1973,p. 2.
8. Childe, New Lighton theMost AncientEast (hereafter New Light),reprint,
Routledge& KeganPaul,London,1963,first publishedunderthistitlein 1934and
as The Most AncientEast in 1928,p. 175;Man Makes Himself, p. 169; What
Happenedin History(hereafter WhatHappened),reprint,Penguin,1972,first
publishedin 1942,p. 132; StuartPiggott,Prehistoric India,Harmondsworth,
Middlesex,1950,p. 140; A.D. Pusalker,'The Indus ValleyCivilization', in R.C.
Majumdar(ed.), TheVedicAge,BharatiyaVidya Bhavan,Bombay,1951,pp.
176,196;Wheeler,TheIndusCivilization, p. 136;Bridgetand RaymondAllchin,
The BirthofIndianCivilization:Indiaand PakistanBefore 500 B.C (hereafter
Birth),Penguin,1968,p. 126; RaymondAllchin,'The Legacy of the Indus
Civilization', in Gregory L. Possehl (ed.), Harappan Civilization: A
ContemporaryPerspective(hereafterHarappan Civilization), American
InstituteofIndianStudiesand Oxford& IBH Publishing Co.,New Delhi,1982,p.
332; K. Antonova,G. Bongard-Levin and G. Kotovsky, A History ofIndia,BookI,
ProgressPublishers,Moscow, 1979,pp. 14-16; V.K. Thakur,Urbanization in
AncientIndia,AbhinavPublications, New Delhi,1981,pp. 26-27;Dani, Recent
Archaeological Discoveriesin Pakistan, pp. 2, 55.
9. A. Ghoshplaces theEarlyand MatureHarappaperiodswithinc. 2700-1900BC
and theLate Harappa periodwithinc. 1700-1000BC; see Encyclopaedia, Vol. I,
pp. 87,90. Themostrecentradiocarbon dates(withoutMASCAcorrection) arec.
2900-2100BC forthe EarlyHarappa period,c. 2200-1800BC fortheMature
Harappa period and c. 1800-1300BC fortheLate Harappa period.Applying
MASCA correction, theEarlyand MatureHarappaperiodsextendfromc. 3200to
2200BC and fromc. 2700to 2100BC respectively; see K.S. Ramchandran, 'Dating
theIndusCivilization'in B.B.Lal and S. P. Gupta(ed.),Frontiers, p. 539.
10. GregoryL. Possehl (ed.), AncientCitiesoftheIndus(hereafter AncientCities),
Vikas PublishingHouse, New Delhi,1979,pp. x, 4748, 287. ThoughtheEarly
and Late periodswere substantially rural,accordingto JimG. Shaffer, urban
centreswerenotabsentduringtheLateHarappaperiod;see 'HarappanCulture:A
Reconsideration', in Possehl(ed.), HarappanCivilization, p. 49.
11. The new cities are spatiallylarger and can accommodatea much denser
populationthanthe agricultural villagesthathave been absorbedin themor
thatstillsubsistbesidethemand urbanismon a vasterscale thanon theNile or
theEuphratessignified progressin termsof organicand culturalevolution;see
Childe,Man MakesHimself, pp. 14-15,142. The Indus Civilizationmarkedthe
zenithofthefirstperiodofurbanization duringtheBronzeAge,maintains Dani;
see RecentArchaeological Discoveries in Pakistan, p. 1.
12. VivekanandJha,'AgriculturalLabour and Village Artisansin EarlyNorth
Indian History(up to c. 500 BC)' (hereafter 'Agricultural Labourand Village
Artisans'),SocialScienceProbings, Vol. I, No. 4, People'sPublishing House,New
Delhi, 1984,pp. 544 46 (the articlewas firstpresentedat the IndianHistory
Congresssymposiumat its 45thsessionheld at AnnamalaiUniversity, Tamil
Nadu, in 1984); cf. Childe, WhatHappened, p. 135;Progressand Archaeology,
Watts& Co., London,1944,p. 49; Man MakesHimself, p. 131; EmestMackay,
EarlyIndusCivilization, 2nd enlargedand revisededn.,Luzac,London,1948,first
publishedin 1935,pp. 132-33;Piggott,Prehistoric India,pp. 153,155;Wheeler,
The Indus Civilization, pp. 72, 84; Basham,Wonder, p. 18; D.D. Kosambi,An
34 SOCIAL SCIENTIST
Introduction to theStudyofIndianHistory (hereafter Introduction), Popular

Book Depot, Bombay,1956,pp. 55, 62; TheCultureand Civilisation ofAncient
Indiain Historical Outline(hereafter Culture),Routledge& KeganPaul, London,
1965,p. 54; Bridgetand RaymondAllchin,Birth, p. 126;R.S.Sharma,'Stagesin
AncientIndian EconomyI: BronzeAge Urbanismto Iron-basedAgriculture'
(Sectionon 'Urban Experiment, c. 2600-1500BC'), Perspectives in Socialand
EconomicHistory of Early India (hereafterPerspectives), Munshiram
Manoharlal,New Delhi,1983,pp. 105-6;Ancient India,3rdrevisededn.,NCERT,
New Delhi, 1990,p. 48, firstpublishedin 1977;IrfanHabib, 'The Peasantin
IndianHistory', GeneralPresident's Addressto the43rdIndianHistoryCongress
session,PIHC,Kurukshetra, 1982,p. 6.
Somescholarshaveenvisionedtheexistence ofa peasantsegment in thetowns
also; cf.Childe,Progress and Archaeology, p. 49; idem,'The UrbanRevolution',
reprinted in Possehl(ed.), AncientCities,p. 15; RobertMcAdams,'TheNatural
HistoryofUrbanism', loc.cit.,p. 20.
13. Shereen Ratnagar,Encounters: The Westerly Tradeof theHarappaCivilization,
OxfordUniversity Press,1981,p. 232; Childe,New Light, p. 186;Man Makes
Himself,p. 150; WhatHappened,p. 134; Piggott,Prehistoric India,pp. 13340;
Wheeler,TheIndusCivilization, pp. 81-82; Kosambi,Introduction, pp. 55-57;
Culture,pp. 59-60;R.S. Sharma,Perspectives, p. 107;AncientIndia,pp. 49-50;
Bridgetand RaymondAllchin,The Riseof Civilization in Indiaand Pakistan
(hereafter Rise),pp. 219-20;A. Ghosh(ed.), Encyclopaedia, Vol. I, p. 85.
14. Dilip K. Chakrabarti, TheExternalTradeof theIndusCivilization, Munshiram
Manoharlal,New Delhi, 1990,pp. 166, 169; Basham,Wonder, pp. 18-19;R.S.
Sharma,Perspectives, p. 107;AncientIndia,p. 49; A. Ghosh(ed.),Encyclopaedia,
Vol. I, p. 85.
15. Kosambi,Introduction, pp. 57-58;Bridgetand Raymond Allchin,Birth, p. 129fn.;
S.P. Gupta,'InternalTradeof theHarappans',in B.B.Lal and S.P. Gupta(ed.),
Frontiers, pp. 417,424.
16. Bridgetand RaymondAllchin,Rise,pp. 193-95,197,199,201-2;A. Ghosh(ed.),
Encyclopaedia, Vol. I, pp. 83-84.
17. VivekanandJha,'Agricultural Labourand VillageArtisans', op. cit.,p. 547.
18. Amongthescholarswho attesttheexistenceof slavesare Mackay,EarlyIndus
Civilization,p. 39; Piggott,Prehistoric India,pp. 169-70;D.H. Gordon,The
Prehistoric Background of Indian Culture,BhulabhaiMemorialInstituteand
N.M. TripthiPvt.Ltd,Bombay,1958,pp. 71,74; Basham,Wonder, p. 18;Kosambi,
Introduction, pp. 55,62,and Culture,p. 54; Dev Raj Chanana,Slaveryin Ancient
India as Depictedin Pali and Sanskrit Texts,People's PublishingHouse, New
Delhi, 1960,p. 170; and Antonova,Bongard-Levin and Kotovsky, A History of
India,Book I, pp. 20, 23. UnliketheSovietscholarV.V. Struveand theGerman
scholar WalterRuben,however,who regardedthe Harappa Cultureas a
specimenof a slave-basedsocial formation, R.S. Sharmaexpressesdoubts
regarding slave labourbeinga significant component oftheHarappaneconomy;
see Perspectives, pp. 108-9;cf.Antonova, Bongard-Levin and Kotovsky, A History
of India, Book I, p. 23; VivekanandJha,'AncientIndian PoliticalHistory:
Possibilitiesand Pitfalls',Social Scientist, Vol. XVIII,Nos 1-2 (200-201),New
Delhi,January-February 1990,p. 41 (in a slightlymodifiedformtheartidehas
appearedin IHR, Vol. XIV,Nos 1-2,July1987and January 1988,publishedin
December1990,see pp. 101-2,and an earlierversionwas firstpresentedat the
50thIndianHistoryCongresssessionat Gorakhpur University in 1989).
Eventhosescholarswhodo notconsidertheevidenceregarding slaveryin the
Harappa Cultureas irrefutable, forexample,G.K. Rai (Involuntary Labourin
AncientIndia,ChaitanyaPublishingHouse, Allahabad,1981,p. 46) and Uma
Chakravarti('Of Dasas and Karmakaras:ServileLabourin AncientIndia',in
Utsa Patnaikand ManjariDingwaney(eds.), ChainsofServitude: Bondageand
Slaveryin India,SangamBooks,Madras,and OrientLongman,1985,pp. 4243),
acknowledgethe existenceof 'a sortof organisedlabour,witha measureof
compulsion neverfaraway'or 'regimented dependent labour'(see Rai,op. cit.,pp.
4647), or 'a sectionwhichlaboured'in thisstratified societyand residedin
barrack-likequartersnear the granaries(see Uma Chakravarti,loc. cit.).
Wheeler,too,had referred muchearlierto servileand semi-servile labouras a
familiarelementof all ancientpolitiesincludingthe Harappan (The Indus

Civilization,p. 54).
19. Y.M. Chitalwala.'The Problemof Class Structure in theIndusCivilization', in
B.B. Lal and S.P. Gupta (eds.), Frontiers, p. 211. In his opinion,theso-called
massacreat Mohenjo-daro mayhavebeentheresultofan internal conflict
rather
than an all-outexternalinvasion(ibid., p. 215). Possehl,too, findssociety
internally differentiatedand structurallyspecialized ('Archaeological
Terminology and theHarappanCivilization', ibid.,p. 30),and envisagespeasant
revoltsas one of the possible factorsof the declineof the Harappa Culture
(Ancient Cities,p. 288).Antonova, Bongard-Levin and Kotovsky do notacceptthe
view thattheHarappansocietywas pre-classin character and findunmistakable
evidenceof class stratification here (A HistoryofIndia,Book 1,p. 22). Amita
Ray,too,has no doubtaboutthehierarchical structure of theHarappanurban
societybased upontheruleofthefewovermany('HarappanArtand Life:Sketch
of A Social Analysis', hereafter 'Harappan Art', in Debiprasad
Chattopadhyaya(ed.), Historyand Society:Essaysin Honourof Niharranjan
Ray,hereafter Historyand Society, K.P. Bagchi& Co., Calcutta,1978,pp. 117,
119, 129). Cass societies,accordingto Childe,subsumea smallminority that
annexes,concentrates and accumulates thesocialsurplusand themasseswho at
bestretainjustas muchoftheproductoftheirlabouras is requiredfordomestic
consumption (SocialEvolution,p. 37).
20. Childe, New Light,p. 174;Man Makes Himself, pp. 130, 142; Progressand
Archaeology, p. 22.
21. Piggott,Prehistoric India,pp. 136,153.
22. Basham,Wonder, pp. 15-16.
23. Kosambi, Introduction, pp. 58-59,61-62;Culture, pp. 64,70.
24. Bridgetand RaymondAUchin, Birth, p. 137;Rise,p. 182.
25. Puskas, 'Societyand Religionin the Indus Valley Civilisation',in Bridget
Allchin(ed.), SouthAsianArchaeology, CambridgeUniversity Press,1981,p.
163.
26. IrfanHabib,GeneralPresident's Address,op. cit.,p. 6.
27. R.S. Sharma,Perspectives, p. 106;Ancient India,p. 50.
28. Antonova, Bongard-Levinand Kotovsky,A HistoryofIndia,Book I, p. 53.
Pusalker,too,regardsit a 'democratic bourgeois'polity;see 'The IndusValley
in Majumdar(ed.), TheVedicAge,p. 173.
Civilization',
29. Puskas,op. cit.,p. 163.
30. R.S. Sharma,Perspectives, p. 106.
31. Childe, New Light,p. 175; WhatHappened, p. 135.
32. Childe, WhatHappened, p. 134.
33. J.M.Kenoyer,'SpecializedProducersand UrbanElites:On the Role of Craft
Industriesin MatureHarappanUrbanContexts', in Kenoyer(ed.),Old Problems
and New Perspectivesin the Archaeology of South Asia (hereafterN e w
Perspectives),
University ofWisconsin, Madison,1989,p. 180.
34. Antonova,Bongard-Levin and Kotovsky, A History ofIndia,BookI, pp. 21,23.
35. Therich-poor divideamounting toclassdivisionis indicatednotonlybythewell-
to-do sectionsinhabitinglarge houses and the poor takingshelterin tiny
dwellings,butalso bytheburialpractices showingthewealthyas beinginterred
withjewelleryand decoratedvesselsand thepoor withaccoutrements on a far
moremodestscale (ibid.,p. 23). Accordingto AmitaRay,themoreconventional
and sophisticated theintaglioseals,thebeardedbustsand thedancing
potteries,
Harappanmale seemto reflect theethosand psycheof thedominant minority,
while theterracotta femalefigurines and animals,thevegetaldecorationsand
narrativepaintings,theMohenjo-darogirland themale torsoreflecttheethos
and psycheoftheworking communities ('HarappanArt',op. cit.,p. 129).
36. IravatiKarve,Hindu Society-An Interpretation, firstpublishedin 1961,2nd
edn., DeshmukhPrakashan,Poona, 1968,pp. 54-64,and Introduction by W.
NormanBrown, p. vi.
37. IravatiKarve,KinshipOrganization in India,firstpublishedin 1953,3rd edn.,
Asia Publishing House,Bombay, 1968,p. 7.
36 SOCIAL SCIENTIST
38. S.C. Malik,Indian Civilization:The Formative Period-A Studyof Archaeology

as Anthropology, firstpublishedin 1968,reprint, IndianInstitute of Advanced
Study,Shimla,and MotilalBanarsidass, Delhi,1987,p. 107.
39. S.C. Malik, Understanding IndianCivilization: A Framewrkof Enquiry, Indian
InstituteofAdvancedStudy,Shimla,1975,p. 76.
40. S.C. Malik, 'Harappan Social and PoliticalLife',in B.B. Lal and S.P. Gupta
(eds.),Frontiers,pp. 204,208.In a recentarticle, 'An EnquiryintotheConceptsof
Technology,Surplusand Social Stratification' (in Dilip K. Chakrabarti (ed.),
Man and Environment, Vol. XII, IndianSocietyforPrehistoric and Quaternary
Studies,Deccan College,Pune,1988,pp. 1-16),MakkhanLal, too,has madean
elaborate,clumsyand facileattempt todenyanylinkoftheappearanceofsurplus
and advance in technologywiththe emergenceof social stratification in the
ancientperiod. His broadsideon Childe in the names of all the supposed
celebritieson the otherside of the fencethatMakhanLal could thinkof is
especiallymisplacedbecauseChildealso maintainedthat'man'sprogressfrom
savageryto civilizationis intimately bound up withthe advanceof abstract
thinking'(The Aryans:A StudyofIndo-European Origins,hereafterAryans,
KeganPaul,Trench,Trubner& Co., London,1926,p. 3), and thirty-seven years
beforeKosambipublishedhis famousarticle,'CombinedMethodsin Indology'
(Indo-Iranian journal,Vol. VI, Nos 3-4, Mouton,The Hague,1963,pp. 177-202),
Childe advocateda coordination of literaryevidencewitharchaeologicaland
anthropological data (Preface,p. xii)forthereconstruction ofhistory.
41. A. Ghosh,City,p. 84.
42. Suvira Jaiswal,'Caste in the Socio-EconomicFrameworkof Early India',
PresidentialAddressto the AncientIndia sectionof the38thIndianHistory
Congresssession,PIHC,Bhubaneswar, 1977,pp. 27-28;idem,'Studiesin theEarly
IndianSocial History:Trendsand Possibilities', IHR, Vol. VI,Nos 1-2,July1979
andJanuary 1980,pp. 11-12.
43. Cf. Piggott,Prehistoric India,p. 170.
44. Indian Instituteof AdvancedStudy,Shimla,and MotilalBanarsidass,Delhi,
1986.
45. Preface, p. ix.
46. Pp. 1-27.
47. IndianInstitute ofAdvancedStudy,Shimla,1971.
48. P. xv.
49. RomilaThapar,Presidential Address,op. cit.,pp. 21,36 fn.;idem,ThePastand
Prejudice,National Book Trust,New Delhi, 1975,p. 29. In his Forewordto
StephenFuchs, TheChildren ofHari: A Studyof theNimarBalahisin Madhya
Pradesh, India (Verlag Herold, Vienna, and The New Order Book Co.,
Ahmedabad,1949,p. viii),Christoph von Furer-Haimendorf no doubttentatively
putsforwardthehypothesis of an urbanoriginof untouchability, butdoes not
consideritpossibletoascribethegrowth ofthissocialphenomenon toanydefinite
periodin Indianhistory. Evenso, thehypothesis is notwell-founded in evidence,
fornotonlyhas themovement fromvillagestotownsinrecenttimesbeena factor
in lesseningtherigoursof casteand untouchability, the'secondurbanization' in
theGanga basin in thesixthcenturyBC borrowednothingfromtheHarappa
Culture(A. Ghosh,City,pp. 2, 30),and untouchability firstappearsprominently
in post-Vedictextsbroadlyreflecting a ruralsetting. Thestudyofthesociallyand
culturally lowestsegmentof theNimarBalahisby Fuchshimself does noteven
remotely indicatethattheiruntouchability was at anystagedue to theirhabitat
inan urbanmilieu.
50. RomilaThapar,Presidential Address,loc. cit.;cf.J.M.Kenoyer, 'Socio-Economic
Structureof the Indus Civilizationas Reflectedin SpecializedCraftsand the
Questionof RitualSegregation', in Kenoyer(ed.),NewPerspectives, p. 189.In her
recentstudy,FromLineageto State:SocialFormations in theFirstMillennium BC
in theGangaValley(hereafter Lineage)(OxfordUniversity Press,1984,p.53),
withoutspecifically mentioning caste,she refersto thepossibility of Harappan
societybeingruledby an aristocracy claimingpowerthrough ritualand religion
and thenotionofpurityapd pollutionprevailing there.
In herA History ofIndza(Vol. I, firstpublishedin 196,reprint, Penguin,1986,
p. 37),however,RomilaThapardoes notrecognisecasteas a phenomenon ofthe
HarappaCultureand categorically statesthattherewas no consciousness ofcaste

evenwhen theRigvedicAryansfirstcame to India.In herAncient IndianSocial
History:SomeInterpretations (hereafter Interpretations) (OrientLongman,1978)
also, wherethe PresidentialAddresshas been reprinted in pages 211-39,she
maintainsin thechapter'Societyand Law in theHinduand BuddhistTraditions'
(pp. 26-39) thattheorderofcastesemergedfroma divinesourcein theHindu
tradition(p. 29) and in the chapter'Ethics,Religionand Social Protestin the
FirstMillennium BC in Northern India'(pp. 40-62)sheassignsthegrowth ofcaste
to thisperiod(p. 47). Althoughin her GeneralPresident's Addressto the44th
IndianHistoryCongresssessionshespeaksofcontinuities betweentheHarappan
and post-Harappansocieties,includingtheVedic (PIHC, Burdwan,1983,pp. 4,
18)-she had expressed her inabilityto identifyany 'specificallyAryan
elementsin thevarietyofpost-FHarappan culturesin theIndusand Gangavalleys
in herVaranasiPresidential Addressin 1969,PIHC,pp. 16-17-sherefrains from
mentioning castein theHarappancontext.
51. Vivekanand Jha, 'Candala and the Origin of Untouchability'(hereafter
Candala), IHR, Vol. XIII,Nos 1-2,July1986and January 1987,pp. 33-34fn.A
shorterversionof thearticlewas firstpresentedat theInternational Seminaron
'New History'organisedby theCentreforHistoricalStudies,Jawaharlal Nehru
University, in collaboration withtheIndia International Centre,New Delhi,in
February1988; and a more elaborateversionwas presentedat the National
Seminaron 'Untouchability in AncientIndia' organisedby the Department of
AncientIndian History,Cultureand Archaeology, BanarasHindu University,
Varanasi,in March1989.
TheRigvedicAryansareno longerregardedas theimmediate successors ofthe
MatureHarappaCultureor theinvadinghordeswhichdestroyed it,as had been
initiallysuggestedby Childe (New Light,pp. 187-88)and Wheeler('Harappa,
1946:The Defencesand Cemetery R 37',Ancient India,No. 3, January 1947,p. 82)
and enthusiastically supportedlater,amongothers, by Kosambi(Introduction, pp.
6869; Culture,pp. 55,71). Waybackin 1931SirJohnMarshallthought thatthe
HarappaCulturecouldhavebeenbut'a mereshadowofitsformer self'whenthe
Indo-Aryans enteredPunjababoutthemiddleofthesecondmillennium BC,and
since no evidenceof a large-scalearmedconfrontation had been foundat the
excavatedsites,theywerenot its destroyers (Mohenjo-daro, Vol. I, pp. 110-12).
Amongtheadherents tothisviewareBasham(Wonder, pp. 24,28);AskoParpola
et al (Decipherment, p. 5); K.A. NilakantaSastri(Aryansand Dravidians, P.C.
Manaktala & Sons, Bombay,1979,pp. 8-9); GeorgeF. Dales ('The Mythical
Massacre at Mohenjo-daro', in Possehl (ed.), AncientCities,p. 294); Vishnu-
Mittre ('The Harappan Civilizationand the Need of A New Approach',in
Possehl (ed.), HarappanCivilization, p. 37); RomilaThapar(Interpretations,
pp. 18,152-53.262);Antonova, Bongard-Levin and Kotovsky (A History ofIndia,
Book I, pp. 27-29); R.S. Sharma (Perspectives, p. 110); and A. Ghosh (ed.,
Encyclopaedia,Vol. I, p. 89). As Dani pointsout,theRigvedadoes notspeakofa
greatstateagainstwhichtheAryansfoughtand theiropponentsmayhavebeen
thosewho lived in small territorial zones like thatof Taxila; seeTheHistoric
CityofTaxila,Unesco,Paris; and the CentreforEast Asian CulturalStudies,
Tokyo,1986,p. 35.
52. EmileBenveniste, a well-known authority in comparative linguistics,has found
wordscorresponding to Sanskritlotrain severalIndo-European languages;see
Indo-European Language andSociety, Faber& Faber,London,1973,pp. 135-36.
53. R.S.Sharmahas counted176references togau (cattle)as against21 references to
agriculturalactivitiesin the Rigveda.Cattlenotonlyprovidedmilk,meatand
hide,butas theprimary sourceofenergywerealso used in ploughing fieldsand
drawingcarts.Battleswerefoughtforthesake ofcattle,whichalso formedthe
mediumofexchangeand weretheverymeasureofwealthoftheRigvedicAryans;
see 'Formsof Property in theEarlyPortionsof theRigveda', Essaysin Honourof
Professor S.C. Sarkar,People's PublishingHouse, New Delhi, 1976,p. 40; cf.
MaterialCultureand Social Formations in AncientIndia (hereafter Material
Culture),Munshiram Manoharlal,New Delhi,1987,pp. 38-39;Sudrasin Ancient
India (hereafterSudras),firstpublishedin 1958,3rd edn.,MotilalBanarsidass,
Delhi,1990,pp. 10,12,21,29; Kosambi,Introduction, p. 77.
38 SOCIAL SCIENTIST
54. No othergrainis specified.

55. The traditional idea ofvisbeinga subdivisionofjana has beendisputedby R.N.
Nandi, in whose opinionjana signifiedtheearlierwanderinggroupwhilevis
markedthebeginning ofhouseholdlife;see 'Anthropology and theStudyofthe
Veda', ReviewArticleon RomilaThapar'sLineage, IHR, Vol. XIII,Nos. 1-2,pp.
155-56.RomilaThapar'shypotheses regardingvisand lineagepresented in her
bookon LineagehavebeendisputedbyA.M. Shah,R.N.Nandiand R.S.Sharma.
56. R.S. Sharma,'Conflict,Distributionand Differentiation in RigvedicSociety'
IHR, Vol. IV,No. 1,July1977,pp. 3,4, 11;Perspectives, pp. 28,113.
57. VedicIndexofNamesand Subjects(hereafter VedicIndex),Vol. I, firstpublished
in 1912,reprint, MotilalBanarsidass, Delhi,1967,p. 357.
58. Basham,Wonder, p. 32.
59. VIII. 56.3; Kane,Historyof Dharmasastra (hereafter Dharmasastra), Vol. II, Pt.
1, firstpublishedin 1941,2nd edn., BhandarkarOrientalResearchInstitute,
Poona,1974,pp. 180-81.
60. Chanana,Slavery, pp. 19-20.
61. Kosambi,Introduction, pp. 92-93.
62. RomilaThapar,Lineage, pp. 3940.
64. R.N. Nandi,'Anthropology and theRigveda', op. cit.,pp. 162-64.
65. In Nandi's opinionhorse-yoked chariotswere also used forgathering fruits,
honeycomb, somaplantandgame;ibid.,p. 164.
66. Rigveda,IX. 112.3.
67. JaimalRai, TheRural-Urban Economy and SocialChangein AncientIndia (300
BC-300 AD), Bharatiya VidyaPrakshan,Varanasi,1974,pp. 99-100.
68. R.S. Sharma,Sudras,p, 30; Perspectives, pp. 27-28; 'Conflict, Distribution and
Differentiation in RigvedicSociety',op. cit.,p. 11.
69. R.S.Sharmarefers to thedifference ofcolourand physiognomy as wellas cultural
differences betweentheAryansand theirenemies;see Sudras,pp. 14-16.Basham
stressesthereligious, socialand culturaldifferences betweentheAryansand the
non-Aryans, butconcedesthattheracialconnotation ofaryahad notbecomequite
meaninglessin theRigvedicstage;see 'Aryanand Non-Aryan in India',in M.M.
Deshpandeand P.E. Hook (eds.), Aryanand Non-Aryan in India,University of
Michigan,1979,pp. 4-5.
70. Rigveda,X. 90. 12.
71. The view expressedin VedicIndex(Vol.II, p. 250),byMacdonelland Keiththat
thecastesystemwas alreadywell on itsway towardsgeneralacceptancein the
Rigvedais not correct.Childerightly denies theexistenceof castein thistext
(Aryans,p. 32). IrfanHabib correctly pointsout thatit is futileto expectthe
social institutionlikecasteto existbeforetheproducersin societywereable to
provide a 'surplus'and the varna initiallypresagedverylittleof the caste
systemthatwas to grow later;see Casteand Moneyin IndianHistory, D.D.
KosambiMemorialLecture, 1985,University ofBombay, 1987,pp. 4-5.
72. VivekanandJha,'Stages in the Historyof Untouchables'(hereafter 'Stages'),
IHR, Vol. II, No. 1,July1975,p. 14;'Candala',IHR,Vol.XIII,Nos. 1-2,p. 1.
73. R.S. Sharma,'The LaterVedic Phase and the PaintedGreyWare Culture',in
DebiprasadChattopadhyaya (ed.), History and Society, pp. 131 43.
74. R.S. Sharma,MaterialCulture, p. 60.
75. Ibid.,p. 58; 'Oass Formation and ItsMaterialBasisin theUpperGangeticBasin
(c. 1000-500BC)' (hereafter 'Class Formation'), IHR, Vol. II, No. 1, p. 2; B.P.
Mazumdar,'ChangingProfile ofEconomicHistory', Presidential Address,Ancient
Indiasection, 37thIndianHistory Congresssession,PIHC,Calicut,1976,p. 38.
76. R.S.Sharma,'Class Formation', op. cit.,pp. 2-3; 'Problems ofSocialFormations in.
EarlyIndia',GeneralPresident's Address,36thIndianHistoryCongresssession,
PIHC, Aligarh,1975,p. 5.
77. R.S. Sharma,MaterialCulture, p. 59.
78. R.S. Sharma,'Class Formation', op. cit.,p.7.
79. Bahu-pasu,PancavimsaBrahmana, VI. 1. II.
80. VajasaneyiSamhita,XXX.5.
81. TaittiriyaBrahmana, III. 4. 1. 1.
82. SatapathaBrahmana, XIII,6. 2. 10.
83. Kosambi,Introduction, p. 91;Culture, p. 24.

85. )(1. 13. The Niruktais a commentary on theNighantu,a Vedicglossaryin five
chaptersand is pre-Paninean.
86. AitareyaBrahmana, VII. 29.
87. Satapatha Brahmana, VI. 4. 1. 13.
88. Claude Meillassoux,'Are ThereCastes in India?',Economy and Society, Vol. Il,
No. 1,London,February 1975,pp. 89-111.
89. Basham,Wonder, p. 41.
90. Satapatha Brahmana, IV. 3.3. 15.
91. It is unfortunate thata compilation ofall hisartidesexceptthoseon coins(Indian
Numismatics, (ed.) B.D. Chattopadhyaya, OrientLongman,Delhi,1981)is yetto
be published.
92. CentralBookDepot,Allahabad,1973.
93. See myartide,'FromTribeto Untouchable: TheCase ofNisadas',in R.S.Sharma
and VivekanandJha(eds.), IndianSociety:HistoricalProbings(In Memory of
D.D. Kosambi), ICHR and People'sPublishing House,New Delhi,1974,pp. 67-84.
94. A.L. Basham,'Aryanand Non-Aryan', in Deshpandeand Hook (eds.),Aryanand
Non-Aryan in India,p. 2.
95. Cf. Sudhakar Chattopadhyaya, Social Life in AncientIndia, Academic
Publishers, Calcutta,1965,p. 151.
96. This hypothesis was put forward by N.K. Dutt,OriginofCasteinIndia,Vol.I, c.
2000-300BC,London,1931,pp. 106-7.
97. K.A.N. Sastri, Aryans and Dravidians,pp. 48-82. In his book entitled
Untouchability: A Historical Study(Koodal Publishers, Madurai,1979,pp. 127,
13241, 144),K.R. Hanumanthan has shownthatuntouchability in thesouthhas
a distinctlylater origin than in the northand the earliestreferencesto
untouchability can be foundin Acarakkovai (fourth/fifth centuryAD) which
shows theDharmasastra influence.
98. The obligatorynatureof expiatoryritesand penances,relativelysimpleor
complex,and the strongsocial sanctionbehindthemis provedby elaborate
provisionsregarding theirstrictenforcement in thesetexts.
99. The preservation of thevarnaorderis ordainedas theprimary responsibility of
themonarchin theDharmasutras ofGautama(XI.9-10) and Vasistha(XIX.7-8),
theArthasastra of Kautilya(1. 3. 14-17),theSmritisof Manu (VII. 35), Visnu
(III. 1, 33), Narada (XII. 113) and Yajnavalkya(I. 363).
100. The Ramayana, forexample,portrays Ramakillinga sudranamedSambukawho
in violationof Dharmasastranorms was practisingpenance which had
purportedly resultedin thedeathofa brahmana's son (UttaraKanda,LXXIII.2-
LXXVI.15,GitaPressEdition.
101. See myarticle'Varnasamkara in theDharmasutras: Theoryand Practice', Journal
oftheEconomic and SocialHistory oftheOrient, Vol. XIII,Pt.3, Leiden,1970,pp.
280, 287-8; G.C. Pande, Foundations ofIndianCulture,Vol. II, Dimensions of
AncientIndianSocialHistory, Delhi,1984,p. 229; S.J.Tambia,'FromVarnato
Caste ThroughMixed Unions' in JackGoody (ed.), Characterof Kinship,
Cambridge, 1973,pp. 218,223-24.Thesubstantial increasein thenumber ofmixed
castes in Manu (55 accordingto P.V. Kane, Dharmasastra, Vol. II, Pt. I, p. 59)
reflects thegrowingfusionand assimilation ofvariouselementswiththeAryan
population.Varnasamkara was presumedto be causednotonlybymarriage with
women unfitfor marriageor promiscuityamong the varnas,but also by
relinquishing one'sobligatory duties(Manusmriti, X. 24).
102. GautamaDharmasutra, IV. 22-24;Manusmriti, I. 96.
103. Visnusmriti,X. 37-38. Katyayana (AD 400-60) also uses the termasprisyatwice
in thesenseofuntouchables (verses,433,783).
104. A usefulrecentstudybased on theBuddhistCanonis thatby Uma Chakravarti,
The SocialDimensions ofEarlyBuddhism, OxfordUniversityPress,New Delhi,
1987. It is significantthat like the Upanisads, Dharmasastra and other
brahmanicaltexts,Buddha also expressedfullfaithin thetheoryof highand
low birthsand materialpositionbeingconnectedwithactionin previousbirth.
B.R.Ambedkardid notcarefully go intoall theevidencewhilepropoundingthe
theorythatBuddhismin a way effectively counteredcasteand untouchability
in
40 SOCIAL SCIENTIST
thecountry.Fora detailedexposition ofmyviewsin thisregard,see myarticles

'Stages',IHR, Vol. II, No. 1, pp. 21-23,and 'Candala',IHR, Vol. XIII,Nos 1-2,
pp. 24-31.
105. B.R. Ambedkarmade untenable claims in this regard in his book The
Untouchables (Delhi, 1948,pp. 103, 155,159); cf.VivekanandJha,'Stages',op.
cit.,p. 31.
106. See myarticle'SocialContentof theBhagaradgita', IHR, Vol. XI,Nos. 1-2,July
1984 and January1985,pp. 1-44 (firstpresentedat theVIIthWorldSanskrit
Conference heldat theKernInstitute, Leiden,in August1987).ThisIHR volume
was releasedin 1988.I havenotfoundanyevidenceoftheimpactoffeudalism on
bhakti(devotion)in thistext.
107. DebiprasadChttopadhyaya has in his book Lokayata:
A Studyin Ancient Indian
Materialism (People's PublishingHouse, New Delhi,1959),and B.N.S. Yadava
has in his studySociety in NorthernIndiaprovidedevidenceforthis.
108. R.S. Sharma,'The Kali Age: A Periodof Social Crisis'in S.N.Mukherjee (ed.),
History and Thought(Essays in Honourof A.L. Basham), Subarnarekha,
Calcutta,1982,pp. 186-203;B.N.S. Yadava, 'The Accountsof theKali Age and
theSocial Transition fromAntiquity to theMiddleAges',IHR,Vol. V, Nos. 1-2,
July1978and January 1979,pp. 31-63;Vivekanand Jha,'Candala',op. cit.,pp. 21-
23.
Towards a Historical Sociology of Stratification in Ancient India: Evidence from Buddhist
Sources
Source: Economic and Political Weekly, Vol. 20, No. 9 (Mar. 2, 1985), pp. 356-360
Published by: Economic and Political Weekly
Accessed: 03-03-2015 07:44 UTC
Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political
Weekly.
SPECIAL ARTICLES
Towards a Historical Sociology of Stratification

in Ancient India
Evidence from Buddhist Sources,
Uma Chakravarti
Sociological ana!yses of stratification in India have in recent years raised new issues relating to the caste system,
whether
particularly its existence at the level of the infrastructure.But in the existing state of knowledge the issue of
caste is infrastructureor supeir-structurecannot be subjected to a meaningful analysis because of the inadequacies
caste
of the data used by scholars wvhohave relied primarily on secondary sources. Before one can argue that
of
constituted infrastructure, or was part of the superstructlure,we need to have a full-fledged diachronic study
and
caste which combines Indology with history and anthropology: there is need for both chronological clarity
social
the rigorous use of sources. There is also a need to show the relation between caste categories and other
and economic categories over time.
This paper intervenes in the discussion by bringing in the evidence of Buddhist literatureand the earliest group
This
of inscriptions available for research, both of which have been totally ignored by scholars writing on caste.
categories
evidence is in our opinion the most reliableset of sourcesfor an empiricalanalysis of social and economic
in ancient India.
THE dominant perception of the system particular has been widely used by In- evidence from non-Brahmanical sources
of stratification prevailing in India is dologists and sociologists. These texts are and which already existed in the material
based almost exclusively on Brahmanical the bedrock of Brahmanical legal theory sphere or on the relations between those
sources, and this has led to the view that but what they reflect is just the ideology social categories which had already
ritual factors predominate in determining of caste; they are hardly concerned with crystallised. An analysis of such relations
social stratification as manifested through its real working. An exclusive reliance on will contribute to a better understanding
the caste system.2 The most important these texts will therefore quite naturally of the actual socio-economic conditions
systematic exposition of what may be resultboth in an incomplete pictureof an- of the time rather than confining the
described as the ideological view of caste cient society and in a misleading view of discussion to the realm of the theory of
is the work of Dumont.3 His argument it. A direct consequence of the heavy re- social order.
that the ritual domain of caste encom- liance on Brahmanical sources is that the It is in this context that we consider it
passes the political and the economic do- caste framework alone has traditionally important to bring in the evidence of the
mains, has had a great deal of influence been used for the study of Indian society. Buddhist texts, spanning a period of 500
on the contemporary scholarship on In recent years the class framework has years (5th century BC to 1st century BC),
caste.4 It has been suggested that this begun to be used to study contemporary and support it with the evidence of the
view has somehow found a place even in Indian society. However caste continues earliest set of inscriptions which has come
the writing of the Marxist anthropologist to be treatedas the primarylevel of reality down to us (2nd century BC to 2nd cen-
Godelier when he argues that caste con- in the context of ancient Indian society. tury AD). The regional spread of the texts
stitutes the infrastructurein Indian socie- Furtherit is sometimes treated as the only takes account mainly of easternUP, Bihar,
ty.5 But the suggestion that caste was the reality in social history. The use of the and central India, while the inscriptions
basis of the system of production is a caste frame of reference alone is clearly are scattered over central and western
misrepresentationof social reality,at least inadequatebecause it does not sufficiently India. The picture of stratification that
for the post-Vedic and pre-Christian account for the material basis of society. emerges from our work would thus be re-
period of Indian history,since most of the Caste did not always exist as the finished presentative at least of these areas. The
caste categories had no real empirical product that it is now made out to be and Buddhist texts are an extremely valuable
referents and were essentially theoretical sociologists in particularhave been hardly source since they are narrativein style and
constructs in the minds of the formulators concerned with a diachronic study of deal with people, events and places. They
of the Brahmanical texts. We would like cas -. We need to study caste beginning are therefore a definite improvement on
to show instead the existenceof categories from the time that it had not yet crystallis- the Brahmanical theoretical texts with
which were actually the basis of produced to a time when it had donrXso because their marked pre-occupation with the
tion relations and which in our view truly there was a long period in Indian history ideology of caste. Similarly, the inscrip-
comprised the infrastructure during this when caste was still embryonic-still in tional evidence is of fundamental import
period. the process of crystallisinlg.Using the caste because inscriptionshave to communicate
A major factor in the formulation of frame of reference for this period will in- with people. Consequently theoretical,
the dominant perception of caste is that evitably give a distorted view of society as non-existent categories cannot be used
not only is it based exclusively on it is ahistorical to assume that caste com- and only empirical categories are referred
Brahmanical sources but that it relies prised the basis of Indian society at all to. Finally,we need to stressthe point that
heavily for its theoreticalunderpinningon times in the past. Instead we could fruit- the picture of stratification that emerges
texts which date only from the first cen- fully focus on understandingthose aspects from the sources must be assessed in-
tury AD. The Dharmashastraliteraturein of social life about which we have dependent of the Brahmanical texts and
356zl No9.liticarl ch,'k19
.,an
356 t.1j .X. so 9. MarchX2, 1985
ECONOMIC AND POLITICAL WEEKIY March 2, 1985
not be used merely as a point of reference sudra whose specific function is to serve ding Brahmanicaland Jaina texts. An im-
as has invariably been done in the past. the three other varnas finds no parallel in portant point to note in the above men-
We propose to clarify some of the issues the Buddhist texts. In fact the sudras are tioned classification is the absence of the
relating to caste on the basis of the associated with hunting and such other vessa and sudda (Pali for vaisya and
material outlined above. This materialhas functions rather than with serving sudra) from the list of categories which
been inadequately used by historians and brahmanas in the Buddhist genesis myth is difficult to account for. Oldenberg has
virtually ignored by sociologists. Our which represents a significantly different drawn our attention to the fact that the
analysis will throw light on empirical version of the origin of the four social text gives no indication of the possibilities
social categories outside the caste groups. 13 of any other jati being consideied either
framework. It will also show that certain The Buddhist texts also indicate the ex- high or low.16 We shall return to this pro-
categories which have been treatedas hav- istence of the notion of high and low for blem later in the paper.
ing crystallised into castes were still in a various social categories. However, it is According to the Buddhist kula
fluid state in the pre-Christian era. important to point out that the system of classification the khattiyas, brahmanas
stratification as portrayed in the Pali and gahapatis are consideredhigh whereas
II canon depicts a social reality without other kulas such as the candala, vena,
A comprehensive reading of the early religious sanction unlike the Brahmanical nesada, ratthakaraand pukkusa are con-
Pali texts revealstwo major features:First, conception of hierarchy. The Buddhist sidered low.'7 In one reference the high
the use of three different terms to texts were merely reflecting ideas prevail- kulas of khattiya, brahmanaand gahapati
categorise people-varna, jati and kula; ing in the region in which early Buddhism are associated with white while the low
secondly, the use of two separate schemes was located, adopting some of these ideas kulas of candala, vena, nesada, ratthakara
of categorisation, one reflecting the ex- and adapting others to suit their and pukkusa are associated with black. 18
isting Brahmanicaldivisions of society in- philosophy. There is no attempt at por- Further,the high kulas or the unit of khat-
-to brahmanas, kshatriyas, vaisyas and trayingthe system of stratification as nor- tiya, brahmana, and gahapati are in-
sudras using the terms varna andjati; and mative. However, relativeimportance was variably associated with attributes which
another which is unique to non- associated with the different strata since are evaluated as high, such as wealth,
Brahmanical literature,categorising peo- striving for the high status position of a eminence, and learning.19In contrast, the
ple into kshatriyas, brahmanas and kshatriya, brahmana and gahapati candala, nesada, vena, ratthakara, and
gahapatis, using the term kula for the unit through a meritorious life was endorsed pukkusa kulas are described as low and
of categorisation.6 The fourfold as a creditable goal. it is stated that fools will be born into such
Brahynanical varna scheme is used only On the basis of the evidence of the Bud- kulas in their next life.20It may be noted
in the context of reporting the Buddha's dhist sources it appears that there was no that there is a correspondencebetween the
dialogues with the brahmanas and occa- complex system of ranking in the society Buddhist enumeration of the jati and the
sionally with kings;7it is most often used reflected in them. Instead a simpler two- kula categories except for the marked in-
in dismissing these varna divisions as ir- tier system of stratification is indicated. clusion of the gahapati among the high
relevant. It never occurs in discussions The linear order of castes of the kulas which is missing among the high
with the common people. The alternative Brahmanical texts where the four varnas jatis. This is significant for our argument
scheme of kshatriya, brahmana and are ranked one below the other is reduc- and will also be considered later in the
gahapati is invariably used in situations ed to two strata in the Buddhist literature. paper.
where the common people are present,8 A basic opposition between high and low The division of kamma and sippa into
and appears to represent a division of appears in the context of iati, kula, kam- low and high in the Vinaya text cited
society into the domains of power, ritual ma (work), and sippa (craft);thus thereare above throwsfurtherlight on the Buddhist
and the economy.9 highjatis and low jatis; high kulas and low view of stratification. Low kamma is
The Buddhist use of the fourfold divi- kulas; high kamma and low kamma; and defined as kotaka kammam (work of a
sion of society is however not a replica of high sippas and low sippas. The concep- storeroom keeper) and puppha chaddaka
the Brahmanicalsystem of differentiation; tion of high and low is quite explicit in kammam (work of a flower sweeper) and
the kshatriyas are invariably enumerated the classifications of jati and kula. A long are described as work which is disdained,
first with the brahmanas following passage in the Vinayatexts representsjati, disregarded, scorned and despised in the
next.10 In fact all through the Buddhist nama (name), gotta (exogamous clan region. High work is defined as kasi
texts a special tension is noticeable bet- unit), kamma, and sippa as being of two (agriculture), vanijya (trade) and gorak-
ween the kshatriyas and the brahmanas kinds: ukkatta (high) and hina (low).14 kha (cattle keeping) which is not disdain-
with the Buddha emphaticallydenyingthe While nama and gotta referto individuals, ed, disregarded, scorned and despised in
Brahmanical claim of superiority over all jati, kamma, and sippa refer to groups. the region. Similarly, the work of the
other social groups including the We shall confine ourselves here to the two nalakara sippam (craft of the basket
kshatriyas.1"Similarly, even while using groups being categorised as high and low. maker), nahapita sippam (craft of the
the fourfold scheme the Buddhists re- Thus ukkatta jati is defined as khattiya barber), pesakara sippam (craft of the
jected the Brahmanical arrangement of (Pali for kshatriya) and brahmana, while weaver) and cammakara sippam (craft of
categories in a hierarchy of services in hina jati is defined as candala, vena, the leather worker) are rated as low and
which the low automatically serve the nesada, ratthakara and pukkusa.'5 The are described as being despised in the area
high. The Buddha refutedthe brahmanas' latter categories are conventionally whereas mudda (computing) ganana (ac-
claim to superiority and their right to translated as lowcasteman, bamboo counting) and lekha (writing) are
draw service from others by pertinently plaiter, hunter, cartwright, and flower classified as high and regardedas not be-
pointing out that anyone who possessed sweeper or scavenger by Buddhist ing scorned in the area.22What is signifi-
wealth regardless of his place in the var- scholars, and the specific combination is cant in these referencesis the information
na scheme could employ others to work unique to the Buddhist texts. They do not that the classification of high and low
for him.12 The Brahmanical role for the appear in the same form in the correspon- work and occupations had a regional
357
March 2, 1985 ECONOMIC AND POLITICAL WEEKLY
dimension, which suggests that the Bud- rated as high whereasbasket-making,pot- Brahmanical Dharmasutra literature
dhist texts are reflecting an existing con- tery, leather work, weaving, and the skill where categories such as the pukkusas or
ception of high and low prevalent in the of a barber are considered low. Third, the nesadas are described as the progeny of
area in which the Buddhist texts were Buddhists exhibit some notion of racial mixed marriages.32In the early Buddhist
located. Since the additional factor of superioritysince aboriginalgroups like the literature it appears that the children of
what is disdained and scorned in the nica kulas or the hinajatis (candala, puk- mixed marriages do not form a new caste
region does not appear in the context of kusa, ratthakara, vena, and nesada) who but are absorbed into the social group of
jati one may conclude that the notion of were probably associated with a low one of the parents.33 But the most
high and low in the case of jati was not material culture are ranked as low. significant feature of society during this
related to a specific area but was more. We may now proceed to analyse the period in relation to the Brahmanical
widely prevalent. various strands of evidence on social system of stratification is tne absence of
It should be noted that the Buddhist stratification available in the Buddhist vaisyas and sudras as verifiable and ex-
texts are always consistent on the groups texts and attempt to locate an underlying isting categories. Both categories appear
that are described as high or low. Thus structurein the different schemes of rank- to be theoretical groups which are im-
young men of good families (kula puttas) ing. In this context an important feature possible to locate. Only the brahmanas
who are always from khattiya, brahmana of the term varna is that it appears only and the cshatriyas of the Brahmanical
and gahapati families are associated with in the context of abstract divisions of scheme are discernibleas living categories
agriculture, cattle keeping and trade, or society into various social categories. We in the Buddhist texts. Further, although
with computing, accounting, and writing, have no evidence of it being used in any in the theoretical discussions contained in
all of which are rated as high work or concrete situation. It seems to have re- the Brahmanical literature (which is also
skills.23 On the other hand, individual mained a theoretical concept without any reiterated by the brahmanas in their
barbers and potters are described as be- relevanceto actual application and no one discussions with others in the Buddhist
ing of low birth and the skills of the pot- is ever described as belonging to the khat- literature)the vaisyas are associated with
ter and barber are identified as low.24 tiya varna, vessa varna, or sudda varna. agriculture,cattle keeping, and trade, and
One of the low kulas, the pukkusa, is On the other hand the termsjati and kula the sudras with service, nowhere in the
described as flower sweepers. This had a appear in concrete situations. For exam- early Buddhist texts do we get people or
parallel in the low kamma of the puppha ple Buddha refers to himself as a Sakya groups occupied with agriculture, cattle
chadaka whose function is to sweep jati and others refer to him in the same keeping, or trade being referred to as
flowers.25Similarly, the low kula of vena, manner.26King Bimbisara is convinced vaisyas, or those with service being refer-
described as basket weavers,has a parallel from the appearance of the Buddha that red to as sudras. Insteadthe Buddhisttexts
among one of the low sippas in the craft he can only be of khattiyajati.27 Two associate agriculture with the gahapati;
of the nalakara or basket maker. It is thus brahmanas refer to themselves as being the cattle keeperis described as a gopaka;
possible to establish a correlation between brahmanas by jati and kula.28Elsewhere and the term vanijya is used for the
kula, kamma, and sippa in the Buddhist Buddha is described as belonging to a trader.34Another important category in
literature especially in the context of the leading khattiya kula.29To sum up, it can the Buddhist literaturewas the setthi but
categories that were ranked as high. be argued that although the varna divi- nono of these specific economic groups
Similarly,one can also draw a correlation sions constituted a purely conceptual are in any way linked with the vaisya.
between low kulas, low kamma, and low scheme which had no application,jati was Similarly, while there are no sudras there
sippa. We may represent the argument in both a conceptual and an actual scheme are innumerable references to dasas and
the manner indicated in the accompany- of categories based on ascribed status. karmakaras who are associated not with
ing Table. However,what really seemed to matter to service of the higher varnas but with pro-
Evaluating the evidence available in the the Buddhists were the kula divisions and viding labour for their master who are
Buddhist texts it appears reasonably clear it was this term ratherthan jati which was almost invariably gahapatis.
that the categorisation of kula, kainma used by them when they needed to in-
and sippa into high and low was based on dicate social stratification. We have A reasonable conclusion on the basis
certain principles. First, in the Buddhist numerous instances of good behaviour, of the evidence cited above is that the
system of ranking those who work for wisdom and a meritoriouisexistence be- significant factor for purposes of iden
themselves as owners and producers are ing rewardedwith rebirthin the high kulas tification in the society depicted in the
high, whereas those who work for others of khattiya, brahmnana,and gahapati;30 Buddhist literaturewere the occupational
are low. Thus the gahapati, vanijya, and the opposite characteristics on the other divisions among people. The function one
gopaka are high whereas the flower hand would result in rebirth in the low actually performed provided the basic
sweepersor the storeroomkeepersare low. kulas of candalas, nesadas, pukkusas, identity of different individuals. The Bud-
Second, among the skills there is a divi- venas, and ratthakaras' dhist texts clearly indicate that the
sion of high and low corresponding to An important dimension of the Bud- categories that can be located as having
manual and non-manual skills since dhist texts is the complete absence of any an existential reality were either the
accounting, writing, and computing are notion of the varnasamkaratheory of the various occupational divisions like barber,
TABLE
metal smith worker, potter, etc, or the
categories of kshatriya, brahmana and
Kula Kamma Sippa gahapati. It is not possible for uF to list
here the hundreds of examples ta3atare
High: Khattiya, brahmana kasi, vanijja, mudda, ganana
available to us in the Buddhist texts but
gahapati gorakkha lekha a few examples may be cited to substan-
Low: candala, pukkusa, kottkaka kammam nalakara, nahapita, tiate the point. Thus among those who
nesada, vena, ratthakara pupphachadaka kumbhakara, would be categorised as low was Cunda
cammakara,
who became famous as the person who
pesakara
provided the last meal to the Buddha. He
358
is referred to as a kammaraputta (son of references to the varna and jati schemes Although the evidence of the Buddhist
a metal smith worker).35Similarly, a cer- contained in the Buddhist and Jaina texts. texts is unambiguous in its representation
tain monk is referred to as having been In contrast the gahapati is an inherent of the gahapati as an economic category,
a nahapita (barber) before he joined the part of the kula scheme depicted in the scholars have been casual about assessing
sangha;36 Ghatikara and Dhaniya are Buddhist literature, which as we have this evidence. Almost all the analyses to
referred to as kumbhakaras (potters).37 pointed out before is favoured by the date havetried to fit the Buddhistevidence
This association with occupations is par- authors of the Buddhist texts when they into the Brahmanical frameworkof caste.
ticularly noticeable for those groups sought to indicate social stratification Attempts have thus been made to equate
which would normally be considered low generally, or depict empirical categories the gahapatis with the Brahmanical
in the society reflected in the Buddhist more specifically. The gahapatis as a category of the vaisyas,44but this is sim-
literature. Further it may also be pointed group were intrinsic to the economic do- ple reductionist exercise which militates
out that the specific categories listed as main and were closely associated with against the entire weight of evidence
low jatis and low kulas (such as the vena, agriculture.They werea high status group 3vailable in the Buddhist sources. The
nesada, etc) do not appear in the context recognised as such by the wider social Vaisyais a theoretical category associated
of providing the identity of individuals. milieu which included the king. Further, with agriculture,cattle keeping and trade,
However, as we have already suggested the gahapati was listed as an important and the sudra with service, in the
there was a relationship between the low attribute of the sovereignty of the king Brahmanical scheme which is based on
kulas, low kamma and low sippa and it and was also the primary tax payer.4'In ascribed status. The gahapati on the other
was the kamma or sippa which provided fact, it was the tax-paying characteristic hand is clearly a category in the system
the identity for the lower orders. that determined the inclusion of the of production; he is one who commands
With regard to the high categories the gahapati as one of the seven attributes of and hires the labour of the dasa-
status identification of groups which are sovereignty.42 More specifically the karmakara.The term brahmana-gahapati
empirically locatable in the Buddhist gahapatis were the owners and controllers signifying a category of brahmanas based
literaturecorrespond to the kula divisions of the primarymeans of productionin the on land, shows the need for moving
of khattiyas, brahmanas, and gahapati. form of land. It appears from the Bud- beyond the Brahmanical caste categories
The khattiyas are known by their clan dhist sources that while the bulk of them in order to explain existing reality. It
names and are largely confined to the were subsistence farmers some of them should also be noted that it was not the
gana-sanghas or oligarchical apolitical played a crucial role in the extension of ordinary brahmana who drew services
units in north-west India and the terai agriculture; in producing both for from the sudras in the Buddhist texts. It
region of UP and Bihar. Thus we have themselves and for the market; and a few was only the brahmnana-gahapatis
various people being identified as Sakyas, had even built up a certain amount of (alongwith other gahapatis) who drew ser-
Mallas, or Lichcchavis. Apart from the capital which was invested in trade.4 vice from the dasa-karmakaras in a rela-
clan kshatriyas there are only a few other Most importantly, these sections of the tionship that originated from the
kshatriyas associated with the royal gahapatis were the major employers of brahmnana-gahapatiscontrol over land
families of the monarchical kingdoms. labour. In the Buddhist literature the rather than from any notion of ritual
For example in the MahaparinibbanaSut- gahapatis stand in a direct economic rela- superiority of the brahmanas.
ta of the Digha Nikaya various clan tionship not with the suidrasbut with the
members as well as Ajatasattu the King dasa-karmakaras as their masters. The The inability of the caste frameworkto
of Magadha come forward and ask for a gahapatis were not a caste or a group depict the social realityof the period with
share of the Buddha's, ashes saying whose status was based on birth. In fact which we are concerned, and the failure
"mayam pi khattiya38 (we too are khat- the gahapati cut acrossother social groups of the Brahmanical scheme to accom-
tiyas).39 Further it is clear that the since the texts clearly use the term modate the gahapati has resulted in por-
brahmanas had definitely emerged as a brahrnana-gahapatifor brahmanas wvho traying not only a partial but a distorted
distinct social group who stronglyasserted were based on the land (that is for view of society. This conclusion is more
their brahmanical identity as being based brahmtanaswhose identity was based on than borne out by the inscriptional
on ascribed status. This was the thrust ascribed status but who performed the evidence of the period 200 BC to 200 AD.
of many of their discussionswith the Bud- functions of a gahapati). It is noteworthy What is striking about the considerable
dha. It is equally clear that the brahmanas chat in the Buddhist texts the only other body of inscriptions,referringto hundreds
of Buddhist literaturewere not necessari- social groups associated with land apart of individuals, is that the empirical
ly associated with the performance of from the gahapatis are the brahmanas. categories mentioned in them not only
ritualor with the pursuitof religiousgoals The term brahmana-gahapatiis therefore correspond to the social categories of the
at all. They were most often associated especially significant. The gahapatis were Buddhistliteraturebut also point to a near
with land and they ranked second, after undoubtedly an economic category, absolute similaritywith them. The inscrip-
the gahapatis in the amount of land that perhaps evan a class, whose position in tions make the same distinct use of terms
they controlled. Finally, the category of society was defined by their ownership such as gahapati, setthi-gahapati, setthi
the gahapati most certainly had an em- and control over land and the power this and vanijya;of dantakara(ivory worker),
pirical basis and innumerable people are gave them over the dasa-karmakaras kartizika (labourer) and halika
described as gahapatis in the Buddhist whose labour they used in order to work (ploughman).45 The individuals who
texts. The word gahavai which is variant the land. The failure to accommodate the feature in these inscriptions derived their
of Lheterm gahapati appears in a similar gahapati in the Brahmanical system of identity from the membership of an oc-
fashion in the Jaina texts also.41 stratification is, in our view, the greatest cupational group. They saw themselves
The argument contained in the prece- failureof the Brahmanicalmodel: it shows and others saw them as occupying posi-
ding paragraphs brings us to a cruicial up clearly the model's rigidity and utter tions in a system of production either as
aspect of the evidence available in the distance trom empirical reality. The gahapatis, or as membersof various skill-
Buddhist texts; the special category of the Brahmiianical model is weakest in explained and semi-skilled occupations. If these
gahapati The gahapati is a category ing the politico-economic domain. On the individuals were supposed to be vaisyas
which iE comlpletely mis.sing fromuthe other hand the inclusion of the gahapati or sudras according to Brahmanical
llrahmalnical varna anldjcati schemesboth in the system of stratification is the theory it was irrelevant to them. The
inl theu Brahinanlicalsoulrcesas ;'ell a1sill strength of the Buddhist scheme. Brahmanicalspectacles that we have used
359
March 2, 1985 ECONOMIC AND POLITICAL WEEKLY
upto now has resulted in obscuring our spicuously associated with agriculture.48 and History',New Left Reyiew,No 112,cited
understanding of social reality for too Most importantly it is clear from the con- in Dipankar Gupta "Caste",Infrastructure
long. During a period of approximately temporary evidence that the ranks of the and Superstructure',Economic and Political
Weekly, Vol 16, No 51, 1981, p 2093.
600 years (5th century BC to 2nd century grahast are open: anyone who comes to 6 DN, 1, p 80, 204; MN, II, pp 404-423.
AD) the primary identity of the people control land gets to being called a 7 MN, 11, pp 310-311;MN, I, p 375.
was based on their economic functions. grahast.49The men who work for him or 8 AN, II, p 89; MW, I, p 122; MN, J, p 70;
There is no evidence to suggest that there are bonded to him are called his jan in Cullavagga, p 255.
9 AN, III, pp 75-76.
was even a secondary identity based on north-easternBihar and correspondto the 10 DN, III, p 64; DN, 1, pp 80, 204; MN, II,
jati during the same period because the dasa-karinakaras of ancient times. pp 311-312.
Buddhist literatureand the inscriptions at It should t2 interestingto try and trace 11 DN, 1, p 86; DN, Ill, p 64; MN, 11, pp
least are not concerned with matching the the history ot ihe gahapati or the con- 442-448.
existing scheme of categorisation with the troller of the means of production 12 MN, 11, pp 311-312.
Brahmanical theoretical scheme.* This is especially land through time and to 13 DN, 111,pp 72-73.
in sharp contrast to the Brahmanical ap- analyse the relationship between such a 14 Pacittiya, pp 10-12; (translated by I B
proach which, especially through the category and the Brahmanical caste Horner as "TIheBook of Discipline", Ox-
ford University Press, 1940, III, pp 173-176.
Dharmasutra and Dharmasastra categories. But for this it is necessry to 15 Ibid.
literature,attempteddesperatelyto enforce first understand social reality without be- 16 H Oldenberg, 'On the History of the Indian
a congruence between their own concep- ing pre-conditioned by the Brahmanical Caste System', Indian Antiguary, 1920, Vol
tual frameworkand the existing social and framework. It is unfortunate that the XLIX, p 225.
17 AN,1j, p 89.
economic categories as is evident from the Brahmanical scheme of stratification 18 AN, 111,pp 94-95.
theory of the varnasamnkara (mixed came to be treated as sacred simply 19 MN9,I1, pp 281, 287; MN, 111,p 248.
unions). because it has succeeded in enforcing its 20 MN, 111,p 240.
view of stratification through the 21 Paciitmiya,p 11.
II1 Brahmanical texts (which were given 22 Ibid.
23 MN, 1, p 119.A notable feature of the Bud-
It should be evident from the preceding tremendous importance by the British dhist Texts is that agriculture is invariably
discussion that a significant aspect of jurists and subsequently by Tndian rated as a high occupation.
material and social life reflected in the scholars). This in turn has resulted in the 24 Pacittiya, p 421; MN, 11, p 272.
Buddhist literature and supported by the obsession with the pure and the impure, 25 Theragatha, Khuiddaka Nikaya, Pali
Publication Board, 1958, Vol 11, p 300.
inscriptional evidence, cannot be explain- and of the view that the ritual domain en- 26 Sultta Nipata, Khuddaka Nikaya, Pali
ed by the caste framework. It is possible compasses the economic and political do- Publication Board, 19S8, Vol 1, p 334.
that a system of stratificationindependent main.'!' While this may be valid for ex- 27 Ibid, p 330.
of the Brahmanical scheme existed not plaining Brahmanical theory it does not 28 DN, 1l1, p 63.
29 MN, II, p 430.
only in the pre-Christian era but has hold for an understanding of empirical 30 MN, I11, p 248.
always existed in India. The gahapati as categories in ancient India. The existing 31 MWN, Ill, p 240.
we have outlined earlier was an Indological, sociological and even 32 Baiudhavatia, 1.8.16; 1.9.17.
agriculturist; someone who controlled historical research has also ignored all 33 DN, 1, pp 84-86.
34 Mahmaagga, p 255; Mahavagga, pp 5-6;
land and employed labour. Laterevidence evidence which is contraryto Brahmanical SutttaNipata, Khuddaka Nikaya, Vol 1, p
points to the same features of the theory. Further, it is assumed that the 270.
gahapati. The Tantra-Vartika of Kumarila, evidence of the non-Brahnianical sources 35 DN, II p 98.
a text approximately of the 7th century representsmerely a deviant point of view 36 Mahav'agga, p 262:
37 MN, 11, p, 272; Parajika, p 51.
AD, describes the gahapati as the master and, therefore, is less valuable as well as 38 DN, ll, pp 126-127.
of the karmakara's, who labour on his marginal to the dominant picture of 39 'Dialoguesof the Buddha': Vol II, London,
land.46 In eastern Uttar 1Pradeshand stratification in India. This tilt has to be 1977, pp 189-190.
40 .1C Jain, "Life in Ancient India as Depicted
north-eastern Bihar the gahapati has a rectified: a more meaningful understan- in the Jain Canon", New Book Company,
counterpart even today in the person of ding of social reality,both for the past and 1947, p 143.
the grahast, an individual who has land for the present may then emerge. 41 DN. I, p 53.
and who employs labour. l The Grihast 42 D.N, Il, p 46. Other attributesof sovereign-
is mentioned in a similar context in the Nott's ty arc the Queen consort, the minister, the
priceless.gem, the royal elephant, the royal
mid-18th century records of the British 1 The main source for this paper is the Pali horse, and the wheel.
and at one point the possibilities of mak- Vinaya and SunttaPitaka (Nalanda. Pali 43 For a comprehensive discussion of the
ing a settlement with them was being Publication Board, 1958). The [iianav gahapali see Uma Chakravarti,"The Social
Pitaka consists of four volumes: Purajiku, Dimensions of Early Buddhism'"unpublish-
recommended by some British adminis- Pacittiya, Mahavagga, and Cullavagga.The ed PhD thesis, University of Delhi, 1981.
trators since the grahastas were con- Sutta Pitaka consists of four subdivisions; 44 R S Sharma, "Sudras in Ancient India",
they are: Digha .'Vikalya(in three volumes Mlotilal Banarasidas, 1980, pp 117, 153.
* It is extreimelysignificant that the Asokan abbreviated as DN); MajjhimnaNikava (in 45 H Ludders Berlin, "A List of Brahmi In-
inscriptionisalso do not use the term vaisya three volumes abbreviatedas MN); Sanw7Xsut- scriptions", Indological Book House. 1973.
ta ,Nikaya.(in four volumes abbreviated as 46 Tanilavartlikaof Ku1marila,Poona, 1910,
and suidra. Instead they use the term ibhya SN); and the Anguttara ANikaya(in four p 3185 cited in D R Chanana, "Slavery in
for a wealthy person (i e, those in control of volumes-abbreviated as AN). Ancient lndia", Peoples Publishing House,
the means of production) and the term dasa- 2 See for example A M Hocart, "Caste", 1960, p 132.
bliataka for labouring men. It may also be Methuen 1950, pp 17-20, 34-42; Celestin 47 R Guha, "4 Rule of Property for Bengal",
Bougle, "Essayson the Caste System",Cam- Mouton, 1963, pp 54-55.
noted that the categories mentioned as be- bridge University Press, 1976, p 29; Max 48 Private Communication frioii Anand
ing involved with each other in a relation- Weber,"Religion of India",Glencoe, 1958, Chakrayartiand Lal Bahadur Verma, bas-
ship of productionare neverthe brahtnanical pp 55-100. ed on field experiencefrom north Biharand
categories:in fact they are dr-voidof any caste 3 Louis Durthont, "Homo Hierar-chicus", Gorakhpur District.
Paladin, 1972, p 76. 49 Private Communication from Anand
implicatioils (Murti and Aiyangar, "Edicts 4 Ibid, pp 115-117. Chakravarti.
of Asoka". Madras, 1950, pp 14-15). 5 MauriceGodelier, 'Infrastructures,Soci.ties 50 Dumont, Paladin, 1972, p 81, pp 115-117.
360
Canonical Authority and Social Classification: Veda and "Varṇa" in Ancient Indian Texts
Author(s): Brian K. Smith
Source: History of Religions, Vol. 32, No. 2 (Nov., 1992), pp. 103-125
Published by: The University of Chicago Press
Accessed: 03-03-2015 07:36 UTC
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to History of Religions.
Brian K. Smith CANONICAL AUTHORITY
AND SOCIAL
CLASSIFICATION:
VEDA AND VARNA IN
ANCIENT INDIAN TEXTS
I
"Hinduism" is, according to most recent definitions of it, notoriously
difficult to define. Outsiders have waxed metaphoric in the presence of
this exotically protean religion: Hinduism is like an all-consuming
sponge, an excessively fecund and chaotic jungle, or "afemale presence
who is able, through her very amorphousness and absorptive powers, to
baffle and perhaps even threaten Western rationality, clearly a male in
this encounter."1 Insiders (i.e., "Hindus") have for their own reasons
also found it advantageous to believe, as Nehru did, that "Hinduism as
a faith is vague, amorphous, many-sided, all things to all men."2 Such
a religion can be portrayedas infinitely tolerant, universalistic in scope,
and even nonsectarian while its sociopolitical status as the dominant
and dominating religion of "secular" modern India is soft-pedaled.3
My thanks to Bruce Lincoln, Laurie Patton, and Katherine E. Fleming for their sug-
gestions on reading earlier drafts of this article.
1 Ronald
Inden, Imagining India (Oxford: Blackwell, 1990), p. 86. Inden's book in-
cludes a valuable survey of such constructions of Hinduism. See esp. chap. 3, "Hindu-
ism: The Mind of India," pp. 85-130.
2 JawaharlalNehru, The Discovery of India (London: Meridian, 1960), p. 63.
3 An Indian "senior civilian official," quoted by Mark Fineman in a story concerning
yet another wave of Hindu-Muslim riots published in the Los Angeles Times on Oct. 31,
1990, "insisted that the confrontation is at least as much a political as a religious is-
sue.... 'Hinduism by definition is secular. It embraces all religions,' the official contin-
ued. 'I just hope that what the country is witnessing now is not a redefinition of
Hinduism itself.'"
? 1992 by The University of Chicago. All rights reserved.

0018-2710/93/3202-0001 $01.00
104 Canonical Authority and Social Classification
Others have insisted on definitional clarity. Among those who have

attemptedto define Hinduism two features of the religion are often men-
tioned: (1) the claim that the Veda is absolutely authoritative and "sa-
cred" and (2) the importance of the caste system for structuring"Hindu"
doctrines, practices, and, of course, society. Definitions revolving
around the first feature focus on the canonical status of the Veda within
Hinduism;4 those revolving around the second construe the religion in
terms of its distinctive social theory-and most especially the notion
that one's duty (dharma) is calibrated to one's particular class (varna)
and stage of life (dsrama). Hinduism, under this latter definitional
option, is thus "as much a social system as a religion.... Its social
framework has from very early times been the caste system, and this
has... become... increasingly identified with Hinduism as such."5
It is not usually clear, however, what correlation there might be be-
tween these two criteria for "Hinduism." What do canon (Veda) and
caste (or at least the underpinnings of caste found in the varna system)
have to do with one another? How are these two supposedly crucial
qualities of Hinduism linked?
As we shall see in this article, one such connection is due to the fact
that the framework for the caste system is laid out in the Veda itself.
Caste thus derives at least part of its endurance and persuasiveness
from the fact that it has canonical legitimacy. Furthermore, both the
Veda and the varna scheme are traced back to the dawn of time. Canon
and a particular form of social classification are part of creation itself
according to Vedic cosmogonies. Canon and class are not only primor-
dial; they are also represented in the Veda as structurallyreduplicative
of a generalized cosmic pattern and are therefore both supposedly part
of the "naturalorder of things." Finally, because both Veda and varna
are predominantly regarded as divisible into three components, canon
and class are isomorphic. Thus, in addition to the legitimation the so-
cial structurereceives by being part of the content of the Veda, thefor-
mal authority of the structure of the canon (which is also the structure
of the universe as a whole) is lent to a vision of society also compris-
ing three principal parts.
4 For a
survey of the literature dealing with the Veda as canon, see Brian K. Smith,
Reflections on Resemblance, Ritual, and Religion (New York: Oxford University Press,
1989), esp. p. 13, n. 28. In that work I also have offered a canonical definition of the re-
ligion: "Hinduism is the religion of those humans who create, perpetuate, and transform
traditions with legitimizing reference to the authority of the Veda" (pp. 13-14). See also
Brian K. Smith, "Exorcising the Transcendent:Strategies for Defining Hinduism and Re-
ligion," History of Religions 27, no. 1 (August 1987): 32-55.
5 R. C.
Zaehner, Hinduism (New York: Oxford University Press, 1966), p. 8. Similar
definitions are surveyed in Smith, Reflections on Resemblance, Ritual, and Religion, pp. 9-
13. For the importanceof caste in Western discourse about India, now consult also Inden's
chapter entitled "India in Asia: The Caste Society," pp. 49-84 in his Imagining India.
History of Religions 105
Constituting a canon (a finite set of "texts" that are regarded as foun-

dational and absolutely authoritative),6 constructing a mechanism for
its transmission, and establishing the means for its infinite interpretabil-
ity (so that the canon will perpetually be "relevant")7generate the con-
ditions of possibility for what we call a religious tradition.8 The Veda
is the canon of Hinduism. Like other such works, the Veda was deemed
canonical retroactively by those later religious traditions that created
themselves through this very act of canon formation. The Veda, how-
ever, also represented itself (and was not only represented later) as the
summation of all truth, the unassailable wisdom of the ages. The
canonical status of the Veda was first established, self-referentially and
tautologically, in Vedic texts; the absolute truth and authority of the
Hindu canon was posited from its Vedic inception and reasserted in its
later reception.
The three Vedas, or "triple wisdom,"9 are declared, in the Veda itself,
equal to satya or "truth"(SB 9.5.1.18; see Appendix for abbreviations),
6 For the etymological history of the term "canon" (originally "rule"or "measure,"and
later "list"), see William A. Graham,Beyond the WrittenWord: Oral Aspects of Scripture
in the History of Religion (Cambridge: Cambridge University Press, 1987), pp. 52-53.
Grahamprefers the word "scripture"to denote a religiously authoritativetext: "A book is
only 'scripture'insofar as a group of persons perceive it to be sacred or holy, powerful and
portentous, possessed of an exalted authority, and in some fashion transcendent of, and
hence distinct from, all other speech and writing" (p. 5). "Canon," as I use the term here,
avoids the literary connotations of "scripture":nonliterate groups can and do have the oral
equivalent of a written canon (a set of myths, stories of origins, legends, histories, etc.),
not to mention the case of the Veda that was preserved only orally until recently.
7
"Where there is a canon, it is possible to predict the necessary occurrence of a herme-
neute, of an interpreter whose task it is continually to extend the domain of the closed
canon over everything that is known or everything that exists without altering the canon
in the process. ... [A] canon cannot exist without a tradition and an interpreter"(Jonathan
Z. Smith, "Sacred Persistence: Toward a Redescription of Canon," in Imagining Religion:
From Babylon to Jonestown [Chicago: University of Chicago Press, 1982], pp. 48-49).
8 "Canon" or "scripture"thus does not exist apart from the perception of it as such
by
a community and a tradition. One cannot but agree with William Grahamwhen he writes
that "neither form nor content can serve to distinguish or identify scripture as a general
phenomenon or category.... [F]rom the historian's perspective, the sacrality or holiness
of a book is not an a priori attribute of a text but one that is realized historically in the
life of communities who respond to it as something sacred or holy. A text becomes
'scripture'in active, subjective relationship to persons, and as part of a cumulative com-
munal tradition. No text, written or oral or both, is sacred or authoritative in isolation
from a community" (Beyond the Written Word, p. 5). On the other hand, religious tradi-
tions and communities come into existence only when they assign to themselves a point
of origins (to which they endlessly return) and absolute authority-i.e., a canon.
9 The three principal Vedas are the Rg, Yajur, and Sama. The fourth Veda, the
Atharva, attained its status relatively late in the scheme of things and became one ex-
ample of adding an inferior fourth to a prior triad. The other principal instance of this
phenomenon is, coincidentally enough, the addition of the Sudra or servant class to the
basement of the previously tripartite varna structure. For the much more common ten-
dency to place a "transcendent fourth" on top of an inferior triad in Hinduism (e.g.,
moksa to the three purusarthas), consult Troy Organ, "Three into Four in Hinduism,"
Ohio Journal of Religious Studies 1 (1973): 7-13.
or to vac, "speech" or "the word" also in the sense of "truth."Vac is the

"mother of the Vedas" (TB 2.8.8.5) and is divided into three forms that
are no other than the three Vedas.10 Alternatively, the Vedas are equated
with the brahman, the universal principle that is the ground and end of
all knowledge.1l The Vedas, it is said in the Veda, are "endless" like
great mountains while human knowledge of them is likened to mere
handfuls of dirt (TB 3.10.11.3-5).
Furthermore,the Veda frequently wrote itself into its own accounts
of the creation of the world; the canon is not only absolutely authorita-
tive but also primeval. In some texts, it is even claimed that the uni-
verse in its totality was originally encapsulated in the three Vedas and
was generated out of them.12 The structureof the cosmos as a whole is
thus patterned on the structureof the Veda.13 Conversely, the very tri-
partite form of the Veda is proof of the eternal verity of its contents
since it produces and reproduces the form of world.
The canonical Veda is also the sanctifying source of a social system
that has persisted in India for thousands of years and has become, as we
have indicated, emblematic of the Hindu tradition. The caste system is
founded on a division into three or four classes or varnas-Brahmin
priests, Kshatriya rulers and warriors, Vaishya commoners, and later
including Shlidras or servants. The social superiority of the Brahmin
class was ensured by the Brahmin authors of the canonical Veda.
The social system presented in the Veda is also structurallyreduplica-
tive of the tripartiteform of the canon that is in turn itself a mirrorimage
of the structureof the cosmos. Society becomes merely one expression
of a universe created in the image of the Veda. Because of the bandhus
10 For the three Vedas as the three forms of

vac, SB 6.5.3.4; SB 10.5.5.1,5; and PB
10.4.6,9 (with Sayana's commentary). The three (or four in the case of the Atharvavedic
GB, e.g., 1.5.28) Vedas are also equated with the sacrifice, another all-encompassing en-
tity in Vedic thought. See, e.g., JB 1.358; SB 1.1.4.3; 3.1.1.12; 4.6.7.13; and esp. 5.5.5.10
("The whole sacrifice is equivalent to that threefold Veda."). This chain of connections
also entails the linkage of speech (vac) and sacrifice, for which consult G. U. Thite, Sac-
rifice in the Brahmana-Texts (Poona, India: University of Poona, 1975), pp. 288-90.
l For example, SB 10.1.1.8; SB 10.2.4.6; JUB 4.25.2. For connections between vac
and the brahman, SB 2.1.4.10; AitA 1.1.1; BAU 4.1.2.
12 For example, SB 10.4.2.21-22, where it is said that in the beginning Prajaipati(the
Lord of Creatures) "surveyed all beings and perceived all beings in the triple wisdom.
For in that [Veda] is the essence (dtman) of all meters, of all hymns of praise, of all
breaths, and of all the gods. This indeed exists for it is immortal, and what is immortal
exists; and this [contains also] that which is mortal. Prajapatireflected to himself, 'Truly
all beings are in the triple wisdom.'"
13 This notion continues into recent
history. The nineteenth-century Hindu apologist
HaracandraTarkapancananacould declare that "if there is to be faith in a book, let it be
in the Veda, since it has prevailed on earth from the time of creation onward." Cited and
trans. in Richard Fox Young, Resistant Hinduism: Sanskrit Sources on Anti-Christian
Apologetics in Early Nineteenth-Century India (Vienna: Publications of the De Nobili
Research Library, 1981), p. 99.
or connections that govern Vedic philosophy,14 the two tripartitestruc-

tures of Vedas and varnas are regarded as transformationsof one an-
other, reduplicative manifestations of the fundamentaltriadic form.
Although direct equations between the Vedas and varnas were usu-
ally not drawn,15the absence of explicit connections equating the three
Vedas and the three social classes does not mean that such homologies
were not implied or even presupposed. The Vedas and the varnas share
mutual linkages to components of other realms-metaphysical, spatial,
temporal, ontological, theological, ritual, anatomical, zoological-that
are their analogues. Canonical and sociological classes can thus be in-
terrelated by tracking the connections they have in common, and one
can assume that these homologies were so well known to the Brahmin
theologians that they, just as we, could easily extrapolate from them to
conjoin scripture and society.16
In what follows, I delineate two ways in which Vedas and social
classes are implicitly represented in Vedic texts as homological trans-
formations of one another.
First, the three Vedas are often metonymically represented by their
essential kernels, the three vydhrtis or utterances: "The vydhrtis are
bhah, bhuvah, and svah," one text explains, "and they are the three Ve-
das. Bhih is the Rg Veda, bhuvah is the Yajur Veda, and svah the Sama
Veda" (AitA 1.3.2; cf. TU 1.5.3). Furthermore,the three vydhrtis are
also names for the three worlds of Vedic cosmology, earth (bhah), at-
mosphere (bhuvah), and sky (svah),17 and the worlds are regularly
14
For a survey of the Vedic philosophy of universal resemblance and interconnection,
see Smith, Reflections on Resemblance, Ritual, and Religion.
15
The only exception to this rule I have encountered is provided by TB 3.12.9.2:
"They say that the Vaishya class is born from the verses (rks, i.e., the Rg Veda). They
say the Yajur Veda is the womb of the Kshatriya. The Sama Veda is the procreatorof the
Brahmins." The Vedas associated here with Brahmins and the Vaishyas are inverted in
comparison to the usual homologies, as we shall see. The fact that the passage does in-
deed directly connect the varnas and the Vedas, however, is significant in itself. It dem-
onstrates that such bandhus were not only theoretically possible but actually articulated.
16 It is possible that direct connections between the Vedas and varnas were not drawn
for a reason. As Bruce Lincoln has pointed out in another context, "social stratification
can well be-and often is-expressed by implication alone.... In ways, that which is
unsaid can be far more powerful than that which is openly asserted, for by being left
mute it is placed beyond question or debate" (Lincoln, The Tyranny of Taxonomies,
Occasional Paper no. 1 of the Center for Humanistic Studies [Minneapolis: University of
Minnesota, 1985], pp. 16, 17).
17 In most Vedic taxonomies, the heaven of Vedic salvation and the
sky of Vedic cos-
mology were conceived differently, although the difference was often enough assumed
and implicit. That such a distinction was indeed made by some Vedic writers is borne
out by the following passage where, in a quadripartititecosmology, an explicit division
between "yonder world" or the sky, on the one hand, and heaven on the other is formu-
lated: "By nine [verses] the Maitravaruna [priest] carries him from this world to the
world of the atmosphere; by ten from the world of the atmosphere to yonder world, for
the world of the atmosphere is the longest; with nine from yonder world to the world of
associated with the three social classes: earth = Brahmins, atmosphere

= Kshatriyas, and sky = Vaishyas. If the Vedas = the three worlds, and
the social classes = the three worlds, then the Vedas = the social
classes.
The second mode of indirectly equating Vedas and society centers
around the various meters (chandases) in which the Veda was com-
posed. Particularmeters are, inter alia, explicitly connected to particu-
lar social classes, the "elemental powers" that are the essences of each
varna (i.e., the brahman, ksatra, and vis), and certain distinctive meta-
physical qualities (e.g., tejas or "fiery luster" for the Brahmin meter,
"power" or indriya for the meter of the Kshatriya class, and a certain
animal nature that is characteristic of Vaishyas). Furthermore, the
meters are connected to components of various realms that are also di-
rect analogues of the Vedas, for example, the three worlds. We are
therefore again led to logical, although unstated, equations of the so-
cial classes and the Vedas: if the meters = the varnas = the worlds =
the Vedas, then the varnas = the Vedas.
II
We may begin to fill out this algebraic skeleton of one corner of Vedic
homological thought by turning to some myths of origins.18 The follow-
ing cosmogonic tale has many repetitions and variants in Vedic texts.
In the beginning,Prajapatiwas the only one here.He desired,"MayI be, may

I reproduce."He toiled. He heatedup ascetic heat. Fromhim, from that one
who hadtoiled andheatedup, the threeworlds-earth, atmosphere,andsky-
were emitted.He heatedup these three worlds.Fromthose heated [worlds],
threelights (jyotis) were born:Agni the fire, he who purifieshere [Vayuthe
wind], and Suryathe sun. He heatedup these threelights. Fromthose heated
[lights],threeVedaswereborn:fromAgni, the Rg Veda;fromVayu,the Yajur
Veda;andfromSurya,the SamaVeda.He heatedup those threeVedas.From
those heated [Vedas], three essences (sukras)were born:bhih from the Rg
Veda, bhuvahfrom the Yajur Veda, and svah from the Sama Veda. [SB
11.5.8.1-3]19
heaven" (AitB 6.9). For heaven (svarga) as the "fourth"world, beyond the cosmological
"worldof the sky"(divi loka, svar, dyaus),see JanGonda,Loka:Worldand Heavenin
the Veda (Amsterdam: N. V. Noord-Hollandsche Uitgevers Maatschappij, 1966), p. 91;
and H. W. Bodewitz, "The Waters in Vedic Classifications," Indologica Taurinensia 10
(1982): 49, n. 27: "Heaven, regarded as the 'beyond' rather than as the sky of the day
time, was also described as boundless (ananta)."
18 I have
analyzed Vedic cosmogonies and the varna system from a different angle
elsewhere. See Brian K. Smith, "Classifying the Universe: Ancient Indian Cosmogonies
and the varna System," Contributions to Indian Sociology, n.s., 23, no. 2 (1989): 241-60.
19 Variants include SB 4.6.7.1-2; JB 1.357; AitB 5.34; KB 6.12; JUB 1.1.1-4; JUB
20.1.23.1-6; JUB 3.15.4-9; and PU 5.3-5.
Four different orders of things and beings, each order divided into
three parts, are here depicted as coeval: the three cosmological worlds
of earth, atmosphere, and sky (the spatial order); three naturalelements
or "lights" (fire, wind, and sun) that are identical to three deities (Agni,
Vayu, and Surya/Aditya);20 the three Vedas; and the three verbal es-
sences of the Vedas (bhih, bhuvah, and svah). The Vedas and their
verbal essences are thus situated within a primordial nexus of connec-
tions to other cosmological, natural, and superhumanrealms. The three
chains of associations21 that co-order the cosmological worlds, natural
elements/gods, scriptures, and sacred utterances are thus as follows:
1. Earth = fire/Agni = Rg Veda = bhuh
2. Atmosphere = wind/Vayu = Yajur Veda = bhuvah
3. Sky = sun/Sirya = Sama Veda = svah
A close variant of this text goes on to add the three principal sacrifi-
cial fires of the Vedic sacrificial cult-the centerpiece of ancient In-
dian religion-to the three associative chains.
The gods saidto Prajapati,"If thereshouldbe a calamityin oursacrificedue to

the verse (rk,i.e., the Rg Veda),or due to the formula(yajus,the YajurVeda),
or due to the chant(saman,the SamaVeda), or due to unknowncauses, or a
total miscarriage,whatis the reparation?" Prajapatisaid to the gods, "If there
is a calamityin yoursacrificedue to the verse,offerin the garhapatyafiresay-
ing 'bhih';if due to the formula,in the ignidhriyafire [in soma sacrifices]or,
in the case of havirsacrifices,in the daksinafire saying'bhuvah';if due to the
chant,in the dhavaniyafire saying 'svah';and if due to unknowncauses or a
total miscarriage,offer only in the dhavaniyafire saying all consecutively-
'bhuh,''bhuvah,''svah.'"[AitB 5.32; cf. 5.34; KB 6.12; ChU4.17.1-8]
And in yet another version of the establishment of these same basic

linkages, the appropriatepriests are added to the chains.
[Prajfpati]heatedup thesethreeworlds.Fromthis worldhe emittedAgni, from

the worldof theatmosphereVayu,fromthe sky Aditya.He heatedupthesethree
lights. FromAgni he emittedthe verses,fromVayuthe formulas,fromAditya
the chants.He heatedup the threefoldwisdom.He spreadout the sacrifice.He
recitedwiththe verse,he proceededwiththeformula,he chantedwiththechant.
20 Manu 1.23 also

connects Agni (fire), Vayu (wind), and Aditya (the sun) with the Rg
Veda, Yajur Veda, and Sama Veda, respectively. ChU 3.6-10, alternatively, associates
the Rg Veda with Agni and the Vasus, the Yajur Veda with Indraand the Rudras, and the
Sama Veda with Varuna and the Adityas (and goes on to connect Soma and the Maruts
with the Atharva Veda, and Brahma and the Sadhyas with the Upanishads).
21 Lincoln, in his article The Tyranny of Taxonomies,
prefers the term "module" for
what I will call "chains of associations" to refer to a set of elements (or "analogues")
capable of being homologized to one another.
Thenhe developedthe essence of fieryluster(tejas) for this threefoldwisdom,

for thehealingof theseVedas.He developedbhahfromthe verses,bhuvahfrom
the formulas,andsvah fromthe chants.... It is by meansof the verse thatthe
hotrpriestbecomeshotr,by the formulathatthe adhvaryupriestbecomesadh-
varyu,by the chantthatthe udgatrpriestbecomesudgatr.[KB 6.10,11]
Combining the components of these two texts to those established

above, the scheme now appears as follows:
1. Earth = fire/Agni = Rg Veda = bhuh = gdrhapatya fire = hotr
priest
2. Atmosphere = wind/Vayu = Yajur Veda = bhuvah = agnidhriya or
daksina fire = adhvaryu priest
3. Sky = sun/Sirya/Aditya = Sama Veda = svah = ahavaniya fire =
udgitr priest
Still other triads from other arenas fill out these three sets of homol-
ogies even further. At SB 12.3.4.7-11 three metaphysical qualities-
"light," "might," and "fame"-are depicted as the primary generative
categories. Components from the cosmological, theological, scriptural,
and bodily realms22 are then asserted as analogues:
This worldis light (bharga),the atmosphericworldis might(mahas),the sky

is fame (yasas), and what otherworldsthere are, that is everything(sarva).
Agni is light, Vayu is might,Aditya is fame, and what othergods thereare,
that is everything.The Rg Veda is light, the YajurVeda is might, the Sama
Veda is fame, and what other Vedas there are, that is everything.Speech is
light, breathis might, sight is fame, and what otherbreathsthereare, that is
everything.One shouldknow this: "I have put into myself all the worlds,and
into all the worlds I have put my self. I have put into myself all the gods
etc., ... all the Vedasetc., ... all the breathsetc." [SB 12.3.4.7-10]
In this text, light, might, and fame generate the three worlds (earth, at-
mosphere, and sky), the three naturalistic deities (Agni the fire, Vayu
the wind, and Aditya the sun), the three Vedas (Rg, Yajur, and Sama),
and three physical functions (speech, breath, and sight).23 The passage
may be compared to the following text from AitA 3.2.5 (cf. SanA 8.8),
where the categorical system proceeds from an analysis of speech (i.e.,
Sanskrit) into consonants, sibilants, and vowels:
22
KU 1.7 anomalously connects the Yajur Veda and the belly, the Sama Veda and the
head, and the Rg Veda and "form," while JUB 4.24.12 regards the various parts of the
right eye as analogues of the various Vedas.
23
Compare JUB 1.25.8-10 where Rg Veda = speech, Yajur Veda = mind, and Sama
Veda = breath.
And now for this secret teaching (upanisad)concerningall speech.... The

consonantsare the earth,the sibilantsthe atmosphere,andthe vowels the sky.
The consonantsare Agni (or fire), the sibilants Vayu (or air), the vowels
Aditya(or the sun). The consonantsare the Rg Veda, the sibilantsthe Yajur
Veda,the vowels the SamaVeda.The consonantsare the eye, the sibilantsthe
ear, the vowels the mind.The consonantsare the inhalation,the sibilantsthe
exhalation,the vowels the circulatorybreath.
Note here the addition of three breaths to the triads that include, as in
other passages already encountered, the three worlds, the three gods/
natural forces, and the three Vedas. Corresponding to the Rg, Yajur,
and Sama Vedas, however, are in this instance different bodily organs:
the eye, the ear, and the mind, respectively, as opposed to speech,
breath, and sight as one of the texts cited above argues.
The tripartite Veda is, in sum, depicted in many Vedic texts as cre-
ated in the beginning as part of the cosmos in which we live. The three
worlds, three natural elements, three deities (or types of deities), the
ritual components of three fires and three principal priests, three quali-
ties (light, might, and fame), three bodily parts or functions, three as-
pects or speech, three kinds of breaths-all are homologized to the
three Vedas and their verbal essences. We thus arrive at a composite
tripartite structure, here reconfigured to present the three Vedas first:
1. Rg Veda = earth = fire/Agni = bhah = garhapatya fire = hotr
priest = light = speech or eye = consonants = inhalation
2. Yajur Veda = atmosphere = wind/Vayu = bhuvah = dgnidhriya or
daksina fire = adhvaryu priest = might = breath or ear = sibilants =
exhalation
3. Sama Veda = sky = sun/Surya/Aditya = svah = ahavaniya fire =
udgatr priest = glory = sight or mind = vowels = circulatory breath
It will be noted that nowhere in the texts cited thus far are there
specific social attributions given to the Vedas or their analogues; the
social classes, in other words, are not mentioned in any of these cos-
mogonies. From other associations found elsewhere in the Veda, how-
ever, we may assume what the authors of these texts undoubtedly did.
Although unstated in the texts above, each chain of resemblances in-
cludes a social component too.
Light, might, and fame, for example, may be regarded as ideal quali-
ties of the three social classes (Brahmin priest, Kshatriya warrior, and
Vaishya commoner, respectively) or as transformationsof the three ele-
mental metaphysical powers that are the essences of the three Aryan so-
cial classes: brahman,ksatra, and vis. The three worlds are also regularly
associated with the three groups constituting society. The earth belongs
to the Brahmin varna, the atmosphere to the Kshatriyas, and the sky to
the Vaishyas.24 The deities included in the three chains are also varna
encoded: Agni and the Vasus are Brahmin deities; Vayu, Indra, and the
Rudras are Kshatriyas; and Aditya, Sirya, Varuna, and the Adityas are
Vaishya gods.25
Evidence of a similar kind comes from the ChU (3.1-5). That text
associates the Rg Veda with the east, the south is linked to the Yajur
Veda, and the west is connected to the Sama Veda.26 As we will see
below, the cardinal directions are regularly given varna attributions
with the east being the Brahmin direction, the south the Kshatriya
quarter, and the west (or north) belonging to the Vaishyas. The direc-
tions thus also serve as mediators linking Vedas and social classes: Rg
Veda = east = Brahmins; Yajur Veda = south = Kshatriyas; and Sama
Veda = west (or north) = Vaishyas.
The Rg Veda, we may conclude, is the Brahmin Veda, for typically
Brahmin components such as the earth, Agni, speech, and the east are
regularly associated with it. We can infer on the same grounds that the
Yajur Veda is that of the Kshatriyas, and the Sama Veda belongs with
the Vaishyas. These conclusions regarding the social correlates for each
of the three Vedas are corroborated when we isolate one triadic set of
the structure, the three vydhrtis or syllabic essences of the Vedas that
are also, as we have observed, the three worlds of Vedic cosmology.
At SB 2.1.4.11-13 these metonymical representativesof the Vedas are
directly correlatedwith the three varnas (portrayedin the form of neuter
elemental powers) and also to cosmological and ontological triads:
24 For
example, SB 2.1.4.11-13; SB 11.5.8.1-4; SB 12.3.4.7-11; KB 6.10,11; KB
22.1-3; JB 1.18; JUB 1.1.1-4; JUB 1.8.1 ff.; JUB 3.15.4 ff.; ChU 4.17.1-6. This scheme
is obviously hierarchical even though the spatial correlates for each of the varnas are
vertically inverted: the "lowest" world, earth, is associated with the "highest" of the so-
cial classes, the Brahmins; and the "highest" spatial world is connected with the "low-
est" of the three varnas. The earth, like the Brahmins, is from this perspective logically
prior, primary, and foremost; the other two worlds and the varnas that characterize them
are, concomitantly, presented as subsequent, secondary (and tertiary), and derivative.
The atmosphere is the cosmic realm that, because of its natural characteristics, suggests
the tempestuous warrior on the rampage. The countless stars in the sky, together with the
other planets, sun, and moon, perhaps suggested the great numbers that make up the
commoner class, the "masses" of ancient India.
25 For the varna assignments of the principal deities of Vedism, see, e.g., BAU
1.4.11-15.
26 The text goes on to connect the Atharva Veda (and the itihasa and purana litera-
ture) to the north, and the Upanishads are associated with the zenith. Compare TA 2.3,
which declares the Yajur Veda the head (= east) of the fire altar regarded as a person; Rg
Veda the right side (= south); Sama Veda the left side (= north); and Atharva Veda, the
lower part, the foundation (= west).
Prajapatigenerated this [world by saying] "bhuh," the atmosphere [by saying]

"bhuvah," and the heavens [by saying] "svah." As much as these worlds are,
so much is this all.... Prajapati generated the brahman power [by saying]
"bhih," the ksatra power [by saying] "bhuvah," and the power of the vis [by
saying] "svah." As much as the powers of the brahman, ksatra and vis are, so
much is this all.... Prajapati generated the Self (dtman) [by saying] "bhuh,"
the human race [by saying] "bhuvah," and the animals [by saying] "svah." As
much as these Self, human race, and animals are, so much is this all.
This text confirms what we have assumed above-that connections can

be drawn between the Vedas (since the three vyahrtis are equated with
the three Vedas) and the three varnas (which are here also correlated to
the vyahrtis). Because the vydhrtis are also the names for the worlds
that, as we have seen, are said to have varna attributes, the text formu-
lates the equation between the Vedas and the varnas through the
mediation of the three worlds: Vedas = worlds = social classes.
It should also be noted that we have explicitly stated here what in
other texts might seem less apparent: the establishment of a hierarchical
ranking for each of the three strings of associations. The hierarchy is in-
dicated by the order in which the chains are presented; the components
that constitute the chain that includes within it the Brahmins are invari-
ably first, those aligned with the Kshatriyas second, and those connected
to the Vaishyas third. But hierarchy is also enunciated in terms of the on-
tological entities that are located within each string. The Brahmin cate-
gory contains within it the all-encompassing dtman or cosmic Self, the
summation and essence of all beings. The Kshatriyas are linked to the
lower ontological class of humans, and the Vaishyas here, as elsewhere,
are associated with animals.27 So too, we may conclude, are the three
Vedas similarly hierarchically ranked in the eyes of the Vedic classifiers.
Rg, Yajur, and Sama correspond to the Brahmin (and the Self), Ksha-
triya (and humans), and Vaishya (and animals), and in that order.28
27 For the
Vaishyas as animals, consult Brian K. Smith, "Classifying Animals and Hu-
mans in Ancient India," Man (The Journal of the Royal Anthropological Institute), n.s.,
26 (September 1991): 323-41.
28 One wonders whether a
hierarchy of the Vedas is also being posited in a post-Vedic
text that works with slightly different ontological correlates (gods, men, and ancestors,
respectively): "One should never recite the recitations (rks) or formulas (yajurs) when
there is the sound of chants (sdmans)," says Manu, for "the Rg Veda is known to be sa-
cred to the gods, the Yajur Veda to men, and the Sama Veda to the ancestors. Therefore
the sound of the latter is impure (aSuci)" (Manu 4.123-4). Elsewhere, however, one en-
counters passages where the Sama Veda, and not the Rg Veda, is exalted as the "highest
Veda." See, e.g., SB 12.8.3.23, where "the saman" is said to be "the essence of all the
Vedas." Alternatively, a Kshatriya connection for the Sama Veda is suggested at SB
13.4.3.14; SB 14.3.1.10; AitB 3.23; and esp. SB 12.8.3.23: "He then sings a saman. The
saman is the ksatra power. With ksatra he thus sprinkles him [i.e., consecrates a king].
III
The Veda is also analyzable into the meters (chandases) in which the
Vedic verses (rks), formulas (yajurs), and chants (sdmans) are com-
posed. The meters are even given the same primordial standing as the
three Vedas themselves: "From that sacrifice in which everything was
offered," one reads in the famous creation hymn, "the verses [i.e., the Rg
Veda] and chants [the Sama Veda] were born, the meters were born from
it, and from it the formulas [the Yajur Veda] were born"(RV 10.90.8).29
Linkages between certain meters and the social classes are regularly
forged in Vedic texts, most notably in those places where the ritual
mantras are modified according to the class of the sacrificer.30Each of
the meters is supposed to embody a power or quality that is particu-
larly characteristic of the inborn and ritually actualized traits of one or
another of the three varnas. In one rite that entails taking the sacrificial
fire forward from one fireplace to another (see AitB 1.28), a gdyatri
verse (a triplet consisting of eight syllables in each verse) is recited if
the sacrificer is a Brahmin, for "the Brahmin is connected with the
gdyatri. The giyatri is fiery luster (tejas) and the splendor of the brah-
man (brahmavarcasas), and with those he makes him prosper." If the
sacrificer is a Kshatriya, a different verse in the tristubh meter (a quar-
tet of verses each containing eleven syllables) is used, for "the Ksha-
triya is connected with the tristubh. The tristubh is force (ojas), power
(indriya), and virility (virya); truly thus with force, power, and virility
he makes him prosper." Alternatively, in the case of a Vaishya sac-
rificer the verse is composed in the jagati meter (a quartet with each
verse comprising twelve syllables), for "the Vaishya is connected with
the jagati and animals are connected with the jagati. Truly thus with
animals he makes him prosper."
In the initiation or upanayana described in the Grhya Sutras,31 the
sdvitri verse (RV 3.62.10: "We contemplate the excellent glory of the
divine Savitr; may he inspire our intellect!") was imparted to the boy
And the saman is imperial rule. With imperial rule he thus brings him to imperial rule."
These kinds of "inconsistencies" have led scholars, past and present, to ignore the over-
whelming number of "consistencies" in the connections drawn in the Brahmanas, some
of which are traced in this article.
29 The hymn from which this citation is taken, the famous "Purusa Sukta,"
provides
the best-known example of the Vedic claim that the Vedas and varnas were created to-
gether at the beginning of time. Following the verse already cited we read: "His mouth
became the Brahmin, his arms were made into the Kshatriya, his thighs the commoners,
and from his feet the Shidras were born" (RV 10.90.12).
30 The phenomenon is in general called mantra uha, the details for which are de-
scribed in S. C. Chakrabarti'sThe Paribhasas in the Srautasutras (Calcutta: Sanskrit
Pustak Bhandar, 1980), pp. 132-36; 154-65.
31 For the following, see also Brian K. Smith, "Ritual, Knowledge, and Being: Initia-
tion and Veda Study in Ancient India," Numen 33, no. 1 (1986): 65-89.
to inaugurate his period of Veda study. The verse is to be composed in

different meters for members of the different classes. Brahmins were to
learn the verse in the giyatri meter, Kshatriyas in the tristubh, and
Vaishyas in the jagati.32 The adjustment was not only in order to
match the boy's varna to the meter that bore the proper power. The syl-
labic composition of the meters (eight syllables for each line in the
gdyatri meter, eleven in the tristubh, and twelve in the jagati) was also
reduplicative of the respective ages for initiation of boys from different
classes (eight for the Brahmin, eleven for the Kshatriya, and twelve for
the Vaishya).33
Other examples of the ritual uses of the meters "according to varna"
(yathavarna) could be cited.34 In all instances, the meters were thought
to embody and instill certain properties that were characteristic of the
ontology of members of the different social classes. Varna-encoded
properties were ritually injected into the appropriateperson through the
metrical medium.
The Brahmin meter, the gdyatri, was supposed to hold within it typ-
ically Brahmin traits. The text cited above that connects this meter to
the powers of fiery luminosity (tejas) and the splendor of the brahman
(brahmavarcasas) is not alone in making such a claim.35 Other texts
regard the gdyatri as the representative of speech or the mouth (TS
5.4.10.4, 7.2.8.1; KB 11.2; JUB 4.8.1), of light (KB 17.6), of the sac-
rifice (SB 4.2.4.20), of perfection (SanA 2.15), or as the bearer of the
elemental metaphysical power invigorating the Brahmin class called
the brahman.36
Furthermore,the gayatri, like other Brahmin entities, is regarded as
the primary and foremost member of its realm. When the priest puts a
kindling stick on the fire with a giyatri verse, "He thereby kindles the
32 SGS
2.5.4-7; MGS 1.22.13; PGS 2.3.7-9; BDhS 1.2.3.11. According to other texts
(VGS 5.26; KGS 41.20; cf. MGS 1.2.3), wholly different Vedic verses, each in the
appropriate meter, were to be imparted to initiates of different classes. See also R. K.
Mookerji, Ancient Indian Educational (Brahmanical and Buddhist) (London: Macmillan,
1947), p. 182; and P. V. Kane, History of DharmaaSstra, vol. 2, pt. 1 (Poona, India:
BhandarkarOriental Research Institute, 1974), pp. 300-304.
33 The difference between these ages and the last ages possible for
performing the up-
anayana for members each class (also eight, eleven, and twelve, respectively) numero-
logically strengthened the bond between the varnas and the meters.
4For the correlation of the three meters and the three social classes when the adh-
varyu priest puts firesticks into the fire within the ritual of setting up the sacred fires for
a new ahitagni, consult the texts brought together in R. N. Dandekar, ed., Srautakosa
(English Section), vol. 1, pt. 1 (Poona, India: Vaidika Samsodhana Mandala, 1958),
pp. 16, 20, and 24. For similar mantra adjustments in the agnicayana ritual, see, e.g.,
TS 5.1.4.5; TS 5.2.2.4.
35 See also KB 17.2; KB 17.9; TS 5.1.4.5; AitA 1.1.3; SB 4.1.1.14; PB 6.9.25; PB
12.1.2.
36 For the equation of the gayatri and the brahman, AitB 3.5; AitB 3.34; AitB 7.23;
KB 7.10; AitA 1.1.3; SB 4.1.1.14; JUB 1.1.8; JUB 1.6.6; JUB 1.33.11; JUB 1.34.2.
gayatri; the giyatri, when kindled, kindles the other meters; and the
meters, when kindled, carry the sacrifice to the gods" (SB 1.3.4.6). At
PB 8.4.2-4, the tristubh and jagati are said to have been created from
the primordial gayatri. The Brahmin meter, like other Brahmin compo-
nents of the universe, is prior to and generative of others.
The Kshatriya meter, the tristubh, is the meter of force (ojas), power
(indriya), and virility (virya), as we have observed above. Elsewhere
similar attributes such as physical strength (bala) as well as the ele-
mental ksatra power itself are said to be inherent in that meter.37 The
eleven-syllabled verses of the tristubh are also homologized to Indra's
great weapon, the thunderbolt or vajra, and thus replicate the coercive
force of that cosmic armament within the ritual.38
The jagati is frequently associated with the Vaishyas (TB 1.1.9.7;
TA 4.11.1-2) and with animals who are, in turn, connected to the com-
moner class.39 An etymological basis for the correlation is also some-
times encountered: "He offers [oblations] with jagati verses, for
animals are mobile (jagata). By means of the jagati he thus obtains
animals for him."40In at least one text (SB 8.3.3.4), the Vaishya meter
is connected to both animals and food, the latter also a typical designa-
tion of the commoners vis-a-vis the higher social "eaters."41The meter
of the third varna is also said to be weaker than the other two, just as
Vaishyas are supposedly weaker (although numerically larger) than the
Brahmin and Kshatriya elites: "The gayatri and the tristubh are the
strongest among the meters. In that these are on either side and the ja-
gati is in the middle [in this chant], thereby, he encompasses the ani-
mals with the strongest of the meters" (PB 20.16.8).
Creation stories for the three meters explain why each belongs to
one or another of the social classes and the day parts (morning, mid-
day, and evening) correlative to each of those classes. One myth tells
how the gdyatri flew to heaven and procured the soma:
37 TS 5.4.1.5; AitB 3.5; AitB 4.3; AitB 6.21; AitB 7.23; AitB 8.2; AitA 1.1.3; KB
7.10; KB 8.7; KB 10.5; KB 11.2; KB 16.1; KB 16.2; KB 17.2; KB 17.9; KB 18.6; KB
30.11; TB 3.3.9.8; PB 18.10.7; JUB 4.8.1.
38 SB 10.2.3.2; cf. SB 7.5.2.24; SB 8.5.1.10-11; SB 9.2.3.6; AitB 2.2; AitB 2.16.
39 AitA
1.1.3; KB 16.2; KB 17.2; KB 17.9; KB 18.6; SB 8.3.3.3; SB 13.1.3.8; SB
13.2.6.6; SB 13.6.2.5; TS 2.5.10.1; TS 6.1.6.2; TS 3.2.9.4; AitB 3.18; AitB 3.25; AitB
3.48; AitB 4.3; and PB 18.11.9-10. See also Brian K. Smith, "Classifying Animals and
Humans."
40 SB 12.8.3.13; cf. SB 1.2.2.2; SB 1.8.2.11; SB
41
3.4.1.13; TB 3.8.8.4; and KB 8.7.
See Brian K. Smith "Eaters, Food, and Social Hierarchy in Ancient India: A Di-
etary Guide to a Revolution of Values," Journal of the American Academy of Religion
58, no. 2 (Summer 1990): 201-29. For an exception to the association of the jagati and
these obviously Vaishya powers, see KB 11.2, where the meter is connected to the ordi-
narily Kshatriya virtues of physical strength and virility (bala and virya).
What she [the gayatri in the form of a bird] grabbed with her right foot became
the morning pressing [of the soma plant at the soma sacrifice]. The gayatri
made that her own home, and therefore they regard it as the most perfect of all
the pressings. He who knows this becomes foremost, the best; he attains pre-
eminence. And what she grasped with her left foot became the midday press-
ing. That crumbled off and thus did not match the former pressing. The gods
wanted to fix this, so in it they put the tristubh from the meters and Indra from
the deities. With that [in it] it became equal in strength to the first pressing. He
who knows this becomes successful with both pressings of equal strength and
equal in relationship. That which she grabbed with her mouth became the third
pressing. While flying she sucked out its sap. With its sap sucked out it did not
equal the two previous pressings. The gods wanted to fix this. They saw it in
domestic animals. When they pour in an admixture [of milk], and proceed with
the [offering of] butter and the animal [offering], with that it became of equal
strength with the previous two pressings. He who knows this becomes success-
ful with pressings of equal strength and equal in relationship. [AitB 3.27]
Hierarchy and inequality-with the Brahmin gayatri and the morning

pressing presented as "perfect," and the other two pressings as defec-
tive in various ways-are transformed into pressings of "equal
strength" and "similar quality" through supplementation: the tristubh
meter and the warrior god Indra beef up the midday, while different
symbols for animals bring the third pressing up to par. The Brahmin
components of morning and the gayatri meter are thus depicted as self-
contained, primary, and preeminent, the "womb" of the others.42
A variant also assumes original inequality but tells the story some-
what differently.
[Originally] the gayatri was composed of eight syllables, the tristubh of three,
the jagati of one. The eight-syllabled gayatri carried the morning pressing up-
ward. The three-syllabled tristubh was unable to carry the midday pressing up-
ward. The gayatri said to her, "I will come [to the midday pressing]. Let there
be something here for me too." "Okay," replied the tristubh. "Add these eight
syllables to me." "Alright," [the gayatrl said]. She added herself to her. Thus at
the midday [pressing] the last two [verses] of the introductory verse dedicated
to Indra Marutvatand the response belong to the gayatri. She (the tristubh) be-
came eleven-syllabled and carried up the midday pressing. The jagati had one
syllable and was unable to carry the third pressing upward. The gayatri said to
her, "I will come [to the evening pressing]. Let there be something here for me
42
Compare AitB 6.9: "He recites verses in the gayatri meter [for] the morning press-
ing is connected to the gayatri. He recites nine small [verses] at the morning pressing.
Seed is spurted into that which is small. He recites ten [verses] at the midday pressing.
When the seed which is spurted into that which is small reaches the woman's midsection
it becomes that which is most broad. He recites nine small [verses] at the third pressing
[for] children are born from that which is small."
too.""Okay,"repliedthejagati. "Addtheseeleven syllablesto me.""Alright,"

[thegayatrisaid]. She addedherselfto her.Thusat the third[pressing]the last
two [verses]of the introductory versededicatedto the VishvaDevasandthe re-
sponsebelong to the gayatri.She (thejagati) becametwelve-syllabledandcar-
ried up the thirdpressing.This is how the gayatribecameeight-syllabled,the
tristubheleven-syllabled,and thejagati twelve-syllabled.He who knows this
becomessuccessfulwith pressingsof equalstrengthand equalin relationship.
[AitB 3.28]
The Brahmin gayatri meter of eight syllables is once again represented

as perfect and self-sufficient, capable of "supporting" on its own the
morning pressing to which it is assigned. The Kshatriya and Vaishya
representatives are, as in the other version, originally smaller and thus
unable to carry out their functions. But in this account they are not
supplemented by other props; rather,they are infused with the gdyatri's
eight syllables to attain their proper syllabic strength, and thus all the
meters are made "of equal strength and of equal quality."43The myth,
in sum, belies its own overt message of equality among the meters, the
day parts, and (implicitly) the varnas by arguing that such equality was
achieved by the incorporation of the gayatri (i.e., Brahmin) component
into those representing the Kshatriyas and Vaishyas.
The connections made in the texts above between the morning and
the gayatri, midday and the tristubh, and evening and the jagati are ex-
tremely common.44 These three meters, which are sometimes said to
be the very "forms" (rupas) of the three parts of the day,45 are also
here and elsewhere routinely assigned to the Brahmins, Kshatriyas, and
Vaishyas. The linkages to the three parts of the day are, therefore, yet
another transformationof the varna of the meters.
The meters, no less than the social classes they signify, are clearly
structuredin a hierarchical fashion. The Brahmin meter, like the Brah-
min class, is the "first" of the meters-hierarchically and chronologi-
cally-and also the meter of the fewest syllables, just as the elite
Brahmin social class is numerically small: "The gayatri, while being
the smallest meter, is the meter yoked first on account of its strength
43 Yet another version of the story (SB 4.3.2.7-11) starts with the opposite premise:
the meters were originally equal. The conclusion, however, is identical to those reached
in variants that begin differently. Originally equal, the varna-encoded meters are soon
rendered unequal, only to be made again equal-but as slightly transformed versions of
the all-encompassing Brahmin meter. Here, as elsewhere, the gayatri is, mythically, the
meter not only of the morning soma pressing but of them all, for "all the soma pressings
are connected to the gayatri."
44 Other texts that draw
homologies between these three meters and the three parts of
the day include SB 4.1.1.15-18; SB 4.2.5.20; SB 4.5.3.5; SB 14.1.1.17; AitB 3.12; AitB
6.12; KB 14.3; TB 3.8.12.1-2; PB 7.4.6.
45 For example, TB 3.8.12.1-2; AitB 3.12; KB 14.3; SB 4.2.5.20; SB 4.5.3.5; PB
7.4.6.
(virya)" (SB 1.8.2.10). The "larger" Kshatriya and Vaishya meters, on

the other hand, are displayed as subsequent to and the inferior
offspring of the gayatri. Less, indeed, is more.
The gayatri, like the social class it represents, is simultaneously said
to encompass all the other meters. This notion is ritually put into play
according to SB 4.1.1.7-12, where the well-known code of morning =
gayatri, midday = tristubh, and evening = jagati is tapped in order to
efficiently and efficaciously collapse the soma sacrifice into a single
rite. All three of the day's ritual pressings of soma are symbolically
condensed into the morning pressing through manipulating the number
of syllables of the three meters.46
They complete the entire sacrifice at the morning pressing only.... He presses
[the soma] eight times. The gayatri consists of eight syllables and the morning
pressing is connected to the gayatri. Thus this [pressing of the soma eight
times] is made to be the morning pressing.... He then presses [the soma]
eleven times. The tristubh consists of eleven syllables and the midday pressing
is connected to the tristubh. Thus this [pressing of the soma eleven times] is
made to be the midday pressing.... He then presses [the soma] twelve times.
The jagati consists of twelve syllables and the evening pressing is connected to
the jagati. Thus this [pressing of the soma twelve times] is made to be the
evening pressing.
The varna codes for the three meters are exemplified somewhat
differently in the fragmentary tripartite classification scheme of the fol-
lowing text:
The animals have Vayu [the god of wind] as their leader, and Vayu is breath;
the animals are animated by means of breath. He [Vayu] departed from the
gods together with the animals. The gods prayed to him at the morning soma
pressing, but he did not return. They prayed to him at the midday soma press-
ing, but he did not return. They prayed to him at the afternoon soma press-
ing.... If he had returned at the morning soma pressing, the animals would be
among the Brahmins; for the gayatri is the morning soma pressing, and the
brahman is the gayatri. And if he had returned at the midday soma pressing,
the animals would be among the Kshatriyas; for the midday soma pressing
concerns Indra [the king of the gods], and the ksatra is Indra. And since he re-
turned at the evening soma pressing-the evening soma pressing concerns the
Vishva Devas ["All the Gods"], and this all is the Vishva Devas-therefore the
animals are everywhere here. [SB 4.4.1.15-16,18]
46 For some
different symbolic uses to which the syllabic content of the meters is put,
see AitB 6.2; SB 1.7.3.22-25; TB 3.2.7.4-5; KB 10.1; JUB 4.2.1-10; and ChU 3.16. In
other passages, however (e.g., KB 25.3; KB 26.8; AitB 6.21; AitB 6.30; PB 4.4.8; SanA
1.2; SB 1.8.2.13), the ritualists warn against recitations in meters inappropriate to the
time of the pressing.
The gayatri meter is here equated to the morning pressing of soma, the
Brahmin social class, and the metaphysical power of the brahman. The
midday pressing is connected to the Kshatriyas, but then, instead of fol-
lowing the order of the first series of associations (which would require
at this point the appropriatemeter), the text conjoins midday and Ksha-
triyas to the deity Indra before returning to the expected order and the
supplying the metaphysical power called the ksatra. The connections
issuing from the evening pressing are limited to the related divinity (the
Vishva Devas) and the animals (an ontological component).
The missing links can be identified, and the holes in the text filled, by
comparing the associations posited here to others made in other texts.
The first category presented here (morning = Brahmins = gdyatri = brah-
man) is the most complete in the text, but unlike the other two it fails
to mention the deity belonging to the series. Correlative to the Kshatriya
Indra and the Vaishya Vishva Devas is Agni for the Brahmin category,
as we will have occasion to observe below. The second set (midday =
Kshatriyas = Indra = ksatra) omits the meter of this category corre-
sponding to the gdyatri meter in the first set; the tristubh, which is ob-
viously called for given the other homologies, is here left implicit. The
third and least filled-out category (evening = Vishva Devas = animals)
neglects the appropriatesocial class, meter, and metaphysical power of
the chain. But, again, as we know from other regularly formulated ho-
mologies, the Vaishya, jagati, and power of the vis can be supplied. Fur-
thermore, as we have seen above ("Prajapati generated the Self [by
saying] 'bhah,'the human race [by saying] 'bhuvah,'and the animals [by
saying] 'svah'"), the ontological correlates for the Vaishya "animals"
are the Brahmin Self or atman and humans for the Kshatriyas. A recon-
structed tripartiteframework of the text would appear as follows:
1. Morning = Brahmins = Agni = gayatri meter = the brahman = the
Self
2. Midday = Kshatriyas = Indra = tristubh meter = the ksatra =
humans
3. Evening = Vaishyas = Vishva Devas = jagati meter = the vis =
animals
The linkage of the gayatri, tristubh, and jagati meters and three
varna-encoded deities (usually Agni, the divine priest; Indra, the
deified exemplar of the warrior; and the Vishva Devas, the "masses"
among the gods;47 or the three groups called the Vasus, Rudras, and
47 SB 11.5.9.7; KB 14.3; KB 14.4; KB 30.1; AitB 8.6. For other correlations between
the meters and individual gods, see SB 10.3.2.1-6 (gayatri = Agni, tristubh = Indra, and
jagati = Aditya); SB 5.4.1.3 ff.; TS 4.3.2.1 ff.; and TS 5.5.8.2 ff. (gayatri = Agni, tris-
tubh = Indra, and jagati = the Maruts); AitB 3.47 (gayatri = Anumati, tristubh = Raka,
Adityas48) is also well attested. This series of correlates between

meters and divinities can include within it the Brahmin morning, the
Kshatriya midday, and the Vaishya evening: "Prajapatiassigned to the
gods the sacrifice and the meters in portions. He allotted the gayatri at
the morning pressing to Agni and the Vasus, the tristubh to Indra and
the Rudras at the midday [pressing], the jagati to the Vishva Devas and
the Adityas at the third pressing" (AitB 3.13; cf. SB 14.1.1.15 ff.).
Numerological explanations are put forward to explain the particular
bonds between meters and deities. The connection between the three
meters (containing verses with eight, eleven, and twelve syllables, re-
spectively) and the Vasus, Rudras, and Adityas is explicated in one
passage that declares that there are thirty-three gods who drink soma,
"eight Vasus, eleven Rudras, twelve Adityas, Prajapati and the vasat
call" (AitB 2.18). Furthermore,in some texts the deities that are else-
where connected to the meters and the three parts of the day are con-
joined with the meters and the three worlds:49 meters (gayatri,
tristubh, andjagati), gods (Agni and/or the Vasus; Indra and/or the Ru-
dras or Maruts; and one or more of the Vaishya deities), space (earth,
atmosphere, and sky), and time (morning, noon, and evening) are thus
brought together in triadic equations.
In the realm of anatomy, the gdyatri is, unsurprisingly, assimilated to
the head (and "the head means excellence," comments SB 4.2.4.20) or
the mouth, the usual Brahmin body parts; the tristubh, like the
Kshatriya, was created from the chest or arms of the Cosmic Man; and
the Vaishya meter, the jagati, is connected to the hips (e.g., SB 8.6.2.6-
8; 10.3.2.1-6), belly, or penis. Indeed, in several texts (PB 6.1.6-11;
TS 7.1.1.4-6; JB 1.68-69), the meters and the social classes are pro-
duced from these distinctive body parts of the creator, as are deities
(Agni, Indra, and the Vishva Devas), seasons (spring, summer, and
rains), and animals (goat, horse, and cow).50 In yet other cosmogonies
jagati = Sinivali); KB 7.10 (tristubh = ksatra = Varuna;gayatri = brahman = Brhaspati);

and KB 10.5 (gayatri = Agni, tristubh = Soma = ksatra).
48 For example, SB 6.5.2.3-5; SB 12.3.4.1-6; SB 13.2.6.4 if.; JB
1.239; JB 1.283-84;
JUB 1.18.4-6; JUB 4.2.1-9; ChU 3.16.1-5. Compare KB 22.1 ff., where the deities of
the three chains are Agni and the Vasus, Indra and the Maruts, and Varuna and the Adit-
yas, respectively.
49 For example, KB 8.9 (Agni = gayatri = earth; Soma = tristubh =
atmosphere;
Vishnu = jagati = sky); KB 14.3 (with the respective deities being Agni, Vayu, and
Aditya); SadB 2.1.9 ff., JB 1.102, JB 1.270 (the gods here being Agni, Indra, and Suirya);
and KB 22.1-9 (Agni and the Vasus; Indra and the Maruts; and Varuna and the Adityas).
For an inversion of the usual pattern, see SB 14.3.1.4-8 (gayatri = sky, tristubh = atmos-
phere, jagati = earth).
50 These texts actually form quadripartite structures: "From his feet, from his firm
foundation, he emitted the twenty-one-versed (ekavimsa) hymn of praise; along with it
he emitted the anustubh among the meters, not a single one among the gods, the Shudra
(SB 5.4.1.3-5; 8.1.1-2; TS 4.3.2.1-3; 5.5.8.2-3), the expected cardinal

directions are added to the chains of connections (east, south, and west
for the Brahmin and giyatri, Kshatriya and tristubh, and Vaishya and
jagati, respectively).51
Vedas and social classes can thus also be indirectly related through
the mediation of the meters, just as we have seen they can be through
the mediation of the vydhrtis. The meters are explicitly equated to the
three social classes, the metaphysical qualities that are so often defini-
tive of each varna (e.g., fiery luster, force and power, and animality), as
well as to parts of the day, deities, worlds, parts of the body, seasons,
animals, and directions that are all class encoded. As a comparison of
table 1 (which maps the analogues of the three Vedas) and table 2 (sur-
veying the analogues of the three meters) demonstrates, the shared con-
nections are so numerous that we may surmise from these connections,
ourselves employing the homology typical of the Veda, that each of the
meters is an analogue of one of the three Vedas (gdyatri = Rg Veda,
tristubh = Yajur Veda, and jagati = Sama Veda). Since the meters are
given varna attributions and are implicitly connected to the Vedas, the
Vedas are thus in this way too analogues of the social classes.
IV
One of the questions that remains unaddressed,however, is why particu-
lar Vedas were correlatedto particularvarnas. For this I do not claim to
have definitive answers but feel compelled to offer some speculations.
The Rg Veda is often accorded a status above that of the other two
scriptures. Like the Brahmins in the social realm, the morning or
spring in temporal categorizations, and the eastern direction in spatial
structures, the Rg Veda is the "first"or "primary"member of its class.
Moreover, as the Veda of the hotr priest who is the "reciter" of verses
among men. Therefore the Shudra has abundantanimals but is unable to sacrifice, for he
has no deity which was emitted along with him. Therefore he does not rise above simply
the washing of feet, for from the feet he was emitted. Therefore the twenty-one-versed
among the hymns of praise is a firm foundation, for it was emitted from the firm founda-
tion" (PB 6.1.11; cf. TS 7.1.1.5-6; and JB 1.69, where the fourth chain is feet/firm foun-
dation = twenty-one-versed hymn of praise = anustubh meter = yajnayajniya chant = no
god = Shudra = sheep = washing feet).
51 These texts go on to add chains of homologies for the north and the zenith as well,
e.g., SB 5.4.1.6-7: "Ascend to the north! May the anustubh (meter) impel you, the vair-
dja chant, the twenty-one-versed hymn of praise, the autumn season, fruit (phala) the
power. Ascend to the zenith! May the pankti (meter) impel you, the Sdkvara and raivata
chants, the twenty-seven- and thirty-three-versed hymns of praise, the winter and cool
seasons, splendor (varcas) the power." For the equation of the gayatri and the east, tris-
tubh and the south, jagati and the west, anustubh and the north, and pankti and the ze-
nith, see also SB 1.2.5.6-7; SB 8.3.1.1,12; SB 8.3.2.9; SB 8.3.3.1; etc. For four-part
equations (leaving out pankti = zenith), consult TB 3.2.9.6-8. For six-part structuresthat
include meters and cardinal directions (together with zenith and nadir), see KB 22.1-
23.8 and MU 7.1-6.
TABLE 1
THE THREE VEDAS AND THEIR ANALOGUES
Rg Veda Yajur Veda Sama Veda
Social class Brahmin Kshatriya Vaishya

Elemental power Brahman Ksatra Vis
Quality Splendor Greatness Fame
Cosmological world Earth Atmosphere Sky
Season Spring Summer Rainy
Part of day Morning Midday Evening
Cardinal direction East South West
Natural element Fire Wind Sun
Deity Agni, Vasus Vayu, Indra, Rudras Sirya/Aditya,
Varuna, Adityas
Ontological entity Self Humans Animals
Body part Eye Ear Mind
Physical function Speech Breath Sight
Breath Inhalation Exhalation Circulatory
Sacrificial fire Garhapatya Agnidhriya/daksina Ahavaniya
Priest Hotr Adhvaryu Udgatr
Linguistic component Consonants Sibilants Vowels
TABLE 2
THE THREE METERS AND THEIR ANALOGUES
Gayatri Tristubh Jagati

Social class Brahmin Kshatriya Vaishya
Elemental power Brahman Ksatra Vis
Quality Fiery luster, Force, power, virility Animals, food
splendor of
the brahman
Cosmological world Earth Atmosphere Sky
Season Spring Summer Rainy
Part of day Morning Midday Evening
Cardinal direction East South West
Deity Agni, Vgsus Indra, Rudras, Maruts Varuna, Vishva
Devas, Aditya,
Adityas
Ontological entity Self Humans Animals
Body part Head, mouth Chest, arms Hips, belly, penis
Animal Goat Horse Cow
in the sacrifice, it is also connected to speech. Since the Veda as a

whole, as we have also noted, is sometimes said to be the summation
of all speech, the Rg Veda might have been regarded as the metonym-
nical placeholder for the Veda qua creative, sacred speech. As such it
stands in the same relation to the other Vedas as the Brahmin (the full-
est representative of the class of human being) does vis-a-vis the lesser
social classes.
The Yajur Veda, the Veda of the adhvaryu priest or the officiant who
is charged with many of the actual ritual maneuvers, is appropriately
classified with the Kshatriyas, the social class noted for physical activ-
ity. Other associations-with might, force, power, and virility; with
the turbulentrealm of the atmosphere; and with Vayu the wind and the
microcosmic anatomical equivalent, breath-are in keeping with the
"active" nature of both the Kshatriya varna and the Yajur Veda.
Finally, the connection between the Sama Veda and the Vaishyas
would seem to follow from the fact that both are characterized by mul-
tiplicity. The Sama Veda is used by the udgatr priest in the soma sac-
rifices in which he is employed accompanied by a group of supporting
chanters or singers. Just as the sky is the analogue of the vis in cosmol-
ogy (owing to the countless heavenly orbs), and just as the dappled or
spotted animal belongs to the Vaishya because of its multiple mark-
ings,52 so too, it would seem, do the Sama Veda and the multitude of
priests connected with it indicate a connection to the commoner class.
Furthermore,the third place given to the Sama Veda in these structures
may also be attributedto the fact that the Sama Veda is derivative: all
its chants are reworkings of the hymns of the Rg Veda.
More important than the specifics of the equations forged between
the Vedas and the social classes is the fact that such homologies are
made at all, albeit in a roundabout way. The canonical powers of the
Veda-as a supposedly authorless (apauruseya) text outside of the
realm of particular individual or social interests; as primordial and
eternal, and therefore not subject to the contingencies and quirks of
historical time; and as unquestionable, and therefore not subject to
contestation-all these canonical powers are brought to bear on the
hierarchical social order of the varnas.
The caste system (or at least caste in nuce) can thus be presentedas ca-
nonical: the authority of caste derives from the authority of Veda, but
more than that caste is made to appearas a social transformationor redu-
plication of canon. The legitimacy, or even indisputability,of the distinc-
tive social scheme of historical and contemporaryIndia rides piggyback
on the unquestionabletruth of the Veda, and both are part of the eternal
cosmic order of things. Any challenge to either the sociological or the
scripturalstructurecan be debunked as obviously "unnatural"and false.
University of California, Riverside

52
For such connections, see, e.g., TS 2.1.3.2-3; SB 5.1.3.3; SB 5.3.1.6; SB 5.5.2.9;
SB 8.7.3.21.
APPENDIX
ABBREVIATIONS
AitA Aitareya Aranyaka PB PanicavimsaBrahmana
AitB Aitareya Brahmana PGS ParaskaraGrhya Sutra
AV Atharva Veda PU Prasna Upanisad
BAU BrhadaranyakaUpanisad RV Rg Veda
BDhS Baudhayana Dharma Sitra SadB Sadvimsa Brahmana
ChU Chandogya Upanisad SanA SankhayanaAranyaka
GB Gopatha Brahmana SB SatapathaBrahmana
JB Jaiminiya Brahmana SGS SankhayanaGrhya Suitra
JUB Jaiminiya Upanisad Brahmana SV Sama Veda
KB Kausitaki Brahmana TA Taittiriya Aranyaka
KGS KathakaGrhya Sutra TB Taittiriya Brahmana
KU Kausitaki Upanisad TS Taittiriya Samhita
MGS Manava Grhya Sutra TU Taittiriya Upanisad
Manu Manu Smrti VGS Varaha Grhya Sutra
MU Maitrayani Upanisad YV Yajur Veda
Social Scientist
Varna Ideology and Social Change

Author(s): Suvira Jaiswal
Source: Social Scientist, Vol. 19, No. 3/4 (Mar. - Apr., 1991), pp. 41-48
Accessed: 03-03-2015 07:04 UTC
SUVIRA JAISWAL*
Varna Ideologyand Social Change**
The traditionalview of the varnas is thattheyare of divineorigin,

fixed and universal.The varnastratification pervades the entire
universel and is relatedto the threebasic qualitiesof nature,satya
(goodnessor purity), rajas (passion)and tamas(darknessor ignorance).
Thesequalitiesare inherent and inbornin everyobjector being.Hence,
in its conception varna stratificationis both functional and
hierarchical.Occupations have to be hereditaryas functionand
aptitudeare determinedby birthand these stand in a hierarchical
relationsanctifiedby religion.Anytransgression is sinfulas itmeansa
reversalof thenaturalorder,whichis to be maintainedby distancing
thevarnasfromeach otherthroughthepracticeofendogamy.Thusas
faras the ideologyof varnais concerned,social changeaffecting the
functions or positionof a varnais an aberration, a straying away from
thenormal,causingtheconfusion ofthevarnas.
The question arises as to what was the process leading to the
formationof this ideology,what was its relationshipto empirical
realityand how was it relatedto social change.It also involvesthe
problemof theinternalization of thisideologyby thosewho stoodat
thelowerrungsof thevarnaladder.
As to thebeginningsof thisideology,theorieswhichtraceit to the
peculiar genius of the Indo-Europeansor of pre-historic or proto-
historicnon-Aryans are highlyspeculative,based on a subjectiveand
not rarelymistakeninterpretation of data.2This is not theplace to
examine such theoriesin detail but mentionmay be made of two,
namely,theviews of GeorgesDumeziland thoseof D.D. Kosambias
bothhave madea deep impression on Indologistsand historians.
Dumezil argued3thatthetripartite socialorderof priests,warriors
and commonerswas characteristic of theIndo-Europeans, who had a
predilectiontowards a three-foldcategorization. Thistypified their
*CentreforHistoricalStudies,JawaharlalNehruUniversity, Delhi.
11Paperpresentedat thesymposium on 'Ideologyand SocialChange'at the51stsession
oftheIndianHistory
Congree,29 December1990.
Social Scientist,Vol. 19, Nos. 3-4, March-April1991
42 SOCIAL SCIENTIST
world-viewand the Indo-Europeanpantheonwas a projectionof

the tripartiteclass system,which later hardenedas varna in India
and pistra in Iran. A critiqueof this theorywas providedby John
Brough4who showed thata similarthreefold orderingof thesocial
orderintopriests,warriorsand cattle-herders maybe seen in Semitic
societiesportrayed in theOld Testament. Morerecently,
BruceLincoln5
has shown thatthe Nilotic tribesof East Africatoo have a similar
three-folddivision with priestsenjoyinga hierarchicalsuperiority
over thewarriors.His basic argumentthattherootsof theseparation
of thepriestand warriorelementslie in theecologyof cattle-keepers
is quite plausible,althoughit is evidentthatthisseparationis only
functionalat this stage and is not lineage-based or genealogically
determined.Even Emile Benveniste,6who agrees withDumezil in
tracingthe tripartite social divisionto Indo-Europeansclearlystates
thatthese were functionaldivisionsnot 'political'or 'genealogical';
thesewere not kin-based.In his view theIndo-Europeansocial units
werefamily, clan,tribeand country in themannerofconcentriccircles,
but therewas no uniformity in this regard;each group of Indo-
Europeansdeveloped theseinstitutions independently.
Nevertheless,
Benvenistelike Dumezil tracesthedichotomyof priestand warrior
and the notionof hierarchyof social ordersto the Indo-European
phase.
But the idea thatthose who deal withthe divine are superiorto
thosewho have temporaltasksis nothingunusualor typicalofanyone
culture.Varnaideologyis muchmorethana recognition ofthreesocial
categoriesto whicha fourth one, thatof thesudras,was added due to
thehistoricalcircumstance of theAryansconfronting and subjugating
the non-Aryansin vedic and post-vedictimes.The basic questionis:
whatled to theoriginofan endogamouscastestructure whichderived
its legitimacyfromthe vedic notionof a hierarchicalgroupingof
occupationalgroups,when thereis no traceof endogamouslineage-
clustersamong the Indo-Europeansor in the Rigveda.This is the
question which puzzled D.D. Kosambi. He pointed out7 thatthe
Yajurvedic four varnas were quite different fromthe fourclasses
mentionedin the ancientIranian sources,namely,the priest,the
charioteer, thetillerand theartisan.Endogamyis nowherementioned
in the Yasna and in ancientIran all the fourclasses were equally
honoured,but thiswas not thecase withtheancientIndianvarnaor
caste organization,whichhad bothendogamyas well as hierarchy.
Accordingto Kosambi an internalfourfoldcaste systemamongthe
Aryan tribes in India developed due to the assimilationof the
survivorsof the Harappa culturewiththe conqueringAryans.The
subjugationof the Harappan agrarianpopulation,identifiedas the
Dasas of the Rigveda by Kosambi, formed the nucleus of the
Dasa/sudra varna, but the adoption of the 'rituallysuperior'
priesthood of the Harappa culture by the Aryan tribes proved
VARNAIDEOLOGYAND SOCIAL CHANGE 43
catalyticin his opinion, for it separated the priesthoodfromthe

kshatriyaeliteand led to theformation ofan endogamousvarna-caste
system,'forotherwise', he remarks,'thereis no reasonfordemarcation
intoendogamouscastes.'8
It is curiousthata perceptivescholarlike D.D. Kosambifailedto
note thathis hypothesisregardingthe originof thebrahmanacaste
hardlyexplainsendogamy.In his detailedstudyof thebrahmanical
gotras he argues thatof the seven sages regardedas the primary
foundersof thebrahmanicalgotrasonlyVisvamitrawas 'theone real
indubitableAryan',therestwerenon-Aryans. In his view thestoryof
the miraculousbirthof Vasisthafroma jar and his description as the
'firstbrahmana'are unmistakable proofsofhis non-Aryan origin.Even
if we grantthis,9we findthattheVisvamitrasoccupya considerable
space in the gotra lists,hence the Aryanpriestlygroup was not
numerically insignificant.
However,whatis moreimportant is thefact
that endogamy does not demarcate the Visvamitrasfrom the
Vasisthas,ratherthetwocombineto formone endogamousvarna-caste
which is separatedfromthe kshatriya,vaishyaand shudravarnas.
Thusethnicisolationdoes notexplaintheemergence ofan endogamous
brahmana caste in the firstinstance.Of course, endogamydoes
facilitatetheabsorptionof tribalgroupsof diverseethnicoriginsin a
varna/castesystem,whichallows theirincorporation withouttheloss
of theirseparateidentity, a pointwhichhas becomecommonplacein
manysociologicalstudies;but as the exampleof thebrahmanacaste
formationshows, endogamous varna categoriesemergedwhen the
processof social and ethnicfusionwas goingon at all levels.Kosambi
himselfhas suggestedthata fewof the Dasa chiefseulogizedin the
Rigvedichymnsmayhave been survivorsoftheHarappaculturewho
wereadopted intotheAryantribes;and thereis nothingto showthat
such absorptionshad led to the formation of separateendogamous
castes withinthe folds of the kshatriyaor brahmanavarna in the
vedicperiod.In ouropinionendogamywas practisednottodemarcate
the Aryan and the non-Aryanbut to differentiate occupational
categoriesof varyingstatusand thushad a socio-economic basis. To
explaintheendogamyof thecaste societyin termsof racialor ethnic
exclusivenessis quitefallacious.Endogamyis botha manifestation and
a toolforperpetuating classand genderexploitation;'0 and onceitcame
to characterize thevarnasystem,it providedthesystembothstrength
and flexibility.For,thesystemcould go on expandingan hierarchical
societyby providingspace to alien groupsand drawingthemin its
vortexwithoutdoing itselfor the notionof hierarchyany serious
damage. Thus caste endogamyco-existedwith hypergamywhich
allowed limited mobilityin the favourof the upper castes. An
exploitativesystemwhich has the capacity to enroll the best of
whateverrankand originin itsown serviceis farmoreperniciousand
long-lasting thantheone whichis absolutelyclosedand static.
44 SOCIAL SCIENTIST
However,Kosambimade a veryimportant observationthat'casteis

class at theprimitive levelofproduction.'The divisionof societyinto
fourcategoriesand theseparationof thebrahmanaand thekshatriya
elitegroupsin an hierarchical mannerneednotobfuscatetheclassrole
of thissystem.It does notmean thatthevarna or caste systemwas
rootedin a religiousprinciple,the oppressionof the pure and the
impure,as is assumedby Louis Dumont.11 The separationmayhaveits
rootsin the religionof pastoraltribesas is arguedby BruceLincoln.
Cattle-keepers tendto develop two classes of specialists,thefighters,
who specialize in cattle-raids and thusincreasethestockof cattlein
possession of theirtribe,and the priestswho specialize in cattle-
sacrificeand offerto thegods theitemswhichare valuedmostin their
societyand in thisway obtaintheblessingsofthegods and ensurethe
increaseof the wealthand prosperityof theirtribethroughdivine
favour. It has been shown12that the crystallization of the vedic
priesthoodas the brahmanavarnawas a processsimultaneouswith
thegrowingcomplexity and elaborationof thevedicritualof sacrifice.
But the later vedic textswhile emphasizingthe separationof the
brahmanaand thekshatriyaalso speak of theirinterdependence and
oppositionto thelowervarnas.
Thus the varna divisions had a historicalorigin in the real
conditionsof existence;and theseconditionsgave riseto an ideology
whichlegitimizedexploitation.It was the ideologyof thedominant
classesand notexclusively ofthebrahmanas.As IrfanHabibremarked
in his presidentialaddressto theIndianHistoryCongressin 1982,the
karmatheory propagated by Buddhism and Jainismprovided a
powerfuljustificationto the caste doctrine,and the principleof
ahimsa or non-violencelegitimizedthe hatredof the land-owning
peasantrytowardsthe huntingtribes,who were forest-dwellers and
who must have come in conflictwith the agrarianexpansionists
covetingand appropriating the forestland. The earlyBuddhisttexts
denouncethe huntersand foresttribesas hinajatisand theAitareya
Brahmana speaksofsuchcommunities as nicyasand apacyaswho had
theirown chiefs.13 It should be kept in mind thatthe landowning
agriculturists of the sixth century B.C., the gahapatis of the
monarchical kingdoms and the khattiyaclans of thegana-rajyas
constituted the dominantclasses and theirsocio-economic statuswas
quite differentfromthat of the depressed peasantryof the early
medieval times.It is in the BuddhistNikayas of the thirdcentury
B.C. or thereaboutthatthe notionof the pure and the impurejatis
appearsforthefirsttime.Thus thevarnaideologyofthepureand the
impurecastesemergesas theideologyofthedonmnant classes;and itis
onlynaturalthatthebrahmanasbeingpriestsand ideologuesshould
becomeitschiefspokesmen.
We would liketo emphasizethatwe do notregardthevarnasystem
and its ideology as a priestlyinventionbut an expressionof the
VARNAIDEOLOGYAND SOCIALCHANGE 45
dominantmaterialrelationships prevailingin society.This ideology

pervaded even the Buddhistand the Jainworld-viewswith the
difference thatthesereligionscontestedthehierarchical positionand
the higherritualstatusof the brahmanasabove the kshatriyas. But
as we have arguedelsewhere,14 thevarnaideologyhelpedin securing
the structureddependenceof the landless labour in the formof
untouchablesor theso-calledmenialcastesand sustainedthenexusof
jajmanirelationshipbuiltaround the landownerin a pettymode of
production.Hence,even whenempiricalrealityhad transformed itself
fromthefourfold varnastratification
oflatervedictimesto a complex
jati structureof the Gupta and post-Guptatimes,the jati system
derivedits legitimation fromthe varnaideologyand theconceptof
the panchama (fifth)varna and the varnasamkara(birth of new
castes throughintermixingof varnas) were floated15in orderto
provide theoreticalexplanationof the new situation.One need not
regard these theoreticalexercisesas the 'diabolical designs' of the
'cunning'brahmanathinkers. People thinkand act in termsof how
theysee the worldand consciouslytryto findsolutionswithintheir
own frameofreference. However,thedominantclassis able to impose
itsownperceptions and itsownconsciousness on thewholesociety;and
it is theroleof theideologyto securetheacceptanceoftheexploited.
Nevertheless,how was this ideology internalizedby the lower
classes? Why were theynot able to perceiveits truecharacter? We
have remarkedearlierthatin actual practicethe systemallowed the
incorporation of alien and aboriginalpeoples into the caste society
without destroyingtheir kin-structure, customs, etc. While the
majorityof such groups being poor and backward inflatedthe
categoriesof the sudra varna, theirchieflyand priestlylineages
gained admittanceto the brahmanaor kshatriyavarnas.This has
been amply illustratedby the well-knownstudies of the Abhiras,
Gurjarasand theBhumijs.Thiskindofmobility added strength to one
systemto recruitnew membersto the rulingand dominantclasses
withoutany need fora revolutionary changeor conflict.Iftherewere
tensionsand resentment at thelowerlevels,thesecould nothavedone
anythingmore than increase the feelingof group-solidarity and
cohesiveness.The hierarchy of thevarna-jatiwas based on thepetty
modeofproduction and henceitcontinuedtoregulatesocialrelations.
However, varna ideologyoriginatedas an ideologyof hierarchy
legitimizingsocial inequality; and it played a seminal role in
transforming semi-pastoralcommunitiesinto stratified agricultural
communities and theemergenceof earlystates.Withthegrowthof a
morecomplexsocio-political formation in thecenturies
preceeding and
succeedingthe Christianera changedeconomicfactsled to certain
modificationsin the traditionalnotion of the varnas, and it is
interesting to notehow thevarnaideologyhas made fromtimeto time
certainadjustmentsin thosesphereswhichwere in conflict withthe
46 SOCIAL SCIENTIST
materialrealitywhileretainingits formalstaticappearanceowingto
its religiouscolouring.To elucidate,initiallythosewho wereengaged
in cattle-keepingand agriculturewere regardedas vaishyag.They
formeda partofthe'twice-born' community and as suchwerecloseto
the upper two varnas.The categoryof theshudrascomprisedof the
marginalpeoples reducedto domesticslaveryor landlessagricultural
labour providingservice to the upper threevarnas. But with the
greateravailabilityof surplus the gulf between those who were
engagedin manuallabourand thosewho wereable to appropriate the
fruitsof such labourby controlling themeansof production widened;
and thisresultedin the socio-economic degradationof peasantsand
primaryproducers.But as the early Buddhistsourcesindicate,the
well-to-dopeasant could investhis surplusin trade,whichbrought
himprestigeand prosperity. Henceforth, thosewho used theirown or
theirfamilies'labour in agricultureor craftscame to be knownas
shudrasand a vaishyawas one who was primarily a trader.We have
suggested elsewhere16that one of the reasons why the vaishya
communities adoptedor patronizedJainismwas thefactthatJainism
tookthedoctrineofahimsato itsextremeand denouncedagriculture as
it involved the killingof the insects.Thus Jainismcould help a
vaishya in raising his status above the depressed peasant by
emphasizinghis distancefrommanualagricultural activities.
Changes in the conceptionof what constitutesa 'vaishya' or a
'shudra' are inter-related and have a profoundsignificanceforthe
historyof the caste system.The originalfourvarnastratification
developed in the latervedic timesin the Westernhalfof theGanga
basinincludingtheDoab in theregionextending fromnearaboutDelhi
to Patna, the area which was known as Aryavarta.To thisdate this
area has communities assignableto all thefourvarnas.Butin theage
of the Buddha and of the early Pali texts(600 B.C.-300 B.C.) the
huntingand food-gathering tribeswere condemnedby the peasant
communitiesas hina-jatisor low castes. Their assimilationin the
expanding Aryan society as 'shudras' meant the increase and
diversification of the shudra varnawhich came to have depressed
communities at variouslevelsofdevelopment. Thisdivergence became
evenmorepronouncedin eastemand thesouthemIndia,whereAryan
culturemade significant inroadsin theGuptaand post-Gupta timesand
powerful land-owning peasantcommunities engagedinagriculture were
rankedas shudrasin accordancewiththenorthern notionsof varna
during this period. Hence, the Kalitas of Assam, the Kaibartasof
Bengaland theReddis and theVellalasof theSouthwereall dubbed
as shudras.Thisdevelopment led to a dilutionofthenotionof'shudra'
especiallyin these areas. These areas also had culturallybackward
like the Pallans, Pariyans and Madigas in the South and the
Namasudras,Doms, Aborsand Kaibartasin theEast.Such a complex
structurecould hardly be explained on the basis of the earlier
VARNAIDEOLOGYAND SOCIAL CHANGE 47
functionaltheoryof the originof the fourvarnas.The notionof the

panchama varna, althoughfloated,could nottakedeep rootsowingto
its lack of scripturalsanctionin the Purusha-suktahymn.But the
theoryof the originof numerouscastesof varyingstatusesfromthe
inter-mixing of the varnaswas foundmoreuseful;as in shifting the
emphasisfromtheoccupationalcriterion to thequestionofthedegree
of purity-impurity arisingout of the union of the originalfounder-
parents,itallowedmuchscopefortheplacementofthenewlyenrolled
communitiesin an hierarchicalmannercommensuratewith their
materialcondition.Thus,thosewho were deemedbornof approved
anulomaorderfollowingthe rules of hypergamy had a 'pure' status
and those who were the childrenof the disapprovedhypogamous
pratiloma unionswereimpure.It shouldbe notedthattheimpurity did
not arise fromimpureoccupations;rather,thosewho were bornof
impureunions,thatis thosewho had an impurebirthwerecondemned
to followimpureoccupations.Thus the theorybuttressedthecaste
patriarchiesand reinforced the hereditary natureof the hierarchical
social structure.It is only in a situationof such shiftin thevarna
theorythat the Reddy kings of Andhracould take pride in being
shudras having been born fromthe feet of Vishnu17and in the
Akkalapundi grant,the panegyristof Singaya Nayak could claim
thattheshudravarnawas purerthantheotherthreevarnas,as it was
bornalong withtheriverBhagirathi (fromthefeetof Vishnu)!
Modernindustryhas replacedthepettymode of productionwhich
favouredcraft-exclusiveness on a non-competitive basis. It has eroded
the notionsof hierarchyof castes, untouchability taboos on inter-
dining,etc.,at leastin theurbanareas whereitsimpactis felt.Butthe
prohibitionon inter-castemarriageis still practisedwidelyas this
elementof thecaste systemis notin conflict withthecapitalistmode
of production.On the contrary, endogamyalmostinvariablymeans
arrangedmarriageson considerations ofwealth,powerand statusand
as such is well-impregnated with the capitalistvalue system.As a
matterof factin some aspects the strengthof the caste has even
increased in moderntimes and caste ideology may be said to be
undergoinganothertransformation. It has been pointedout thatthe
of
traditionaljajmani type personalexchangerelationships between
thevariouscastesofa villageare now beingincreasingly replacedby
the contractual,pecuniaryand impersonalrelationshipsunder the
influenceof thecapitalistmarketforces,withtheresultthatin times
of adversityone has to depend on themembersof one's own casteto
providegroup-support. The presentday politicstoo allowstheeliteof
a caste to exploitthecaste-consciousnessof theircastemenin orderto
competewith the elite groups of othercastes and communities for
politicalpower. Thus caste ideologygains strength bothforpolitical
and economicreasons in spite of the factthatthereare increasing
differentiationsof wealthand statusof individualswithineach caste.
48 SOCIAL SCIENTIST
Inter-casterelationsare now increasingly markedby cleavages and

conflictsin place of the traditionalcooperationand a sense of
togethernessof all the communitiesin the countryside.Thus caste
upon
ideologyarisingthrougha historicalprocessis havingits'effect
history'througha dialecticalinteractionand has becomea material
force,a formidableimpediment to progress.Onlywhenwe realizeits
truecharactershallwe be able to overcomeit.
NOTES AND REFERENCES
1. The Brahdaranyaka Upanisadspeaks of the gods belongingto the brahma,

kshatra,vaishya and shudra varnas,1.4.11-15;R.E. Hume (trans.),Thirteen
PrincipalUpanisads, 2ndedn.,OUP, 1969,pp. 84-85.
2. For detailssee SuviraJaiswal,'Stratification in RigvedicSociety:Evidenceand
Paradigms',TheHistorical Review,Vol. XVI,1990(forthcoming).
3. C. Scott Littleton,The New ComparativeMythology:An Anthropological
Assessment of theTheories of GeorgesDumezil,Universityof CaliforniaPress,
Berkeley,1973.
4. JohnBrough,'TheTripartite Ideologyof theIndo-Europeans: An Experiment in
Method',Bulletinof theSchoolof Orientaland African Studies,Vol. XXII,1959,
pp. 69-85.
5. Bruce Lincoln,Priests,Warriors and Cattle:A Studyin theEcologyofReligions,
University of CaliforniaPress,Berkeley, 1981.
6. EmileBenveniste, Indo-European Language andSociety,London,1973.
7. D.D. Kosambi,Introduction to theStudyofIndianHistory,Popular Prakashan,
Bombay, pp. 94-95.
8. D.D. Kosambi,'On the Originof the BrahminGotras',Journal of theBombay
BranchofRoyalAsiaticSociety, Vol. XXVI,1950,p. 50. See Kosambi,Introduction
to theStudyofIndianHistory, pp. 94-95.
9. We findthisviewhighlyspeculative.Fora detailedcritiquesee theforthcoming
articleacceptedforpublicationin TheIndianHistorical Review, Vol. XVI.
10. SuviraJaiswal,'Studiesin EarlyIndianSocial History:Trendsand Possibilities',
TheIndianHistorical Review,Vol VI, July1979-January 1980,pp. 5-6; reprinted
in R.S. Sharma(ed.), SurveyofResearch in Economicand SocialHistory ofIndia,
AjantaPublications, Delhi,1986,pp. 43-44.
11. Homo Hierarchicus, Delhi, 1970.
12. K.K. Potdar,Sacrifice in theRgveda, Bombay,1953; Padma Misra,Evolution of
the BrahmanClass (in perspective VedicPeriod),Banaras Hindu University
SanskritSeries,Vol XIII.
13. AitareyaBrahmana, VIII, 38.3.
14. 'Studiesin EarlyIndianSocial History', IHR, VI, pp. 12,21.
15. Ibid., p. 40-41; also see Vivekananda Jha, 'Candala and the Origin of
Untouchability', IHR, XIII,1986-87,p. 13 fn.10.
16. IHR, VII, p. 39.
17. Epigraphic Indica,VIII, inscription no. 3, lines2-3.
Indian Sociological Society
VARNA AND JATI

Author(s): André Béteille
Source: Sociological Bulletin, Vol. 45, No. 1 (MARCH 1996), pp. 15-27
Published by: Indian Sociological Society
Accessed: 03-03-2015 07:32 UTC
Indian Sociological Society is collaborating with JSTOR to digitize, preserve and extend access to Sociological Bulletin.
VARNA AND JATI
Andre Beteille
I WOULD like to use the present occasion to discuss some important
changes taking place in the caste system in our time. The focus of
attention will be on caste as a system of representations, and I would like
to justify my approach by referring you to Durkheim whose view was
that social facts are things, but they are also, and at the same time,
representations.
The social morphology of caste continues to be one of its important
features. The division of Indian society into innumerable castes and
communities has been noted by the many Backward Classes
Commissions set up in independent India, and Mr Mandal's commission
listed as many as three thousand seven hundred and forty three. More
recently, the monumental People of India project undertaken by the

Anthropological Survey of India has drawn public attention to the
continuing significance of the divisions of caste and subcaste in
contemporary India. But I will dwell less on these divisions and

subdivisions themselves than on the ways in which they are perceived,
particularly among the intelligentsia whose role in contemporary Indian
society should not be discounted.

My argument is a simple one. In the classical literature of India, caste
was represented as varna and for two thousand years, when Hindus
wrote about it, they did so characteristically in the idiom of varna. This
is no longer the case and caste is now represented much more typically
as jati, or its equivalent in the regional language. This displacement of
varna by jati indicates much more than a simple linguistic shift. It
indicates a change of perception, a change in the meaning and legitimacy
of caste even among those who continue by the constraints
to abide
imposed by its morphology on marriage and other matters. This change
Andre Beteille is on the faculty of the Department of Sociology, University of Delhi,

Delhi 110 007. This is the text of the special lecture he delivered to an open session of
the XXII All India Sociological Conference held at Bhopal during December 16-18,
1995.
SOCIOLOGICAL BULLETIN, 45 (1) March 1996
16 Sociological Bulletin
has not as yet received the systematic attention from sociologists that is
its due.
It is a truism that caste is not merely a form of identity, arising from

birth in a particular group; it is also a matter of consciousness. It was
believed by many at the time of independence that with economic and
political development, with a secular Constitution and with the spread of

education and a scientific outlook, the consciousness of caste would
decline or disappear, at least from public view. It is quite evident that
caste consciousness has not disappeared, and many would even question
that it has declined. But what we have today is a somewhat different
kind of consciousness, with jati rather than varna in the foreground.
I do not wish to suggest that the consciousness of jati, or the idea of it
is a new one, only that it had a subordinate place in representations of
the caste system. Much of the reality of everyday life must have turned
around the divisions and subdivisions of jati. It may have been the case
even in the distant—as in the more recent—past that peasants, artisans and
labourers gave little thought to the larger scheme of things expressed by
the concept of varna. My attention today is on India's long intellectual
tradition. Within that tradition, the literati, those who reflected on what
we call caste today, and wrote about it, represented it in the idiom of
varna rather than jati. When their contemporary counterparts write and
speak about it, they make use of a different idiom.
I would like to turn now to the path-breaking essay by M. N. Srinivas
(1962: 63-69) called 'Varna and Caste' published forty years ago. It may
be noted that the title pre-empts, at least by implication, the term caste
for the designation of Jati. Srinivas was reacting against the Indological
representation of caste as varna which he felt gave a distorted view of

the Indian reality: 'The varaa-model has produced a wrong and distorted
image of caste. It is necessary for the sociologist to free himself from the
hold of the varaa-model if he wishes to understand the caste system'
(Ibid.: 66). My point is that Indians are in fact freeing themselves from
the hold of that model. The conditions under which this disengagement
is taking place were not discussed by Srinivas, but they merit serious
attention.
Varna and Jati 17
Srinivas's impatience with the vama-model was a response to the

dominance in Indian writing about society of what he called the
'book-view' which eager to replace with the 'field-view'.
he was He
pointed out, with great success, that the way people actually live is very
different from how they are supposed to live, and that sociologists
should concentrate on the former and not the latter. This was true of the
Indian village community, the Indian joint family and, of course, also of
caste. But then, people everywhere have some conception of how they
ought to live. Today in particular, they are acutely aware that they do not
always as they ought to do, and it would be a mistake for the
live
sociologist to ignore how people think they ought to live, and dwell only
on how they actually live. It is in this sense that I consider
representations to be an important part of the social reality.

No matter how we argue, we cannot turn our back on the book-view
of Indian society which may be regarded as a particular form of
collective representations. Of course, collective representations have to
be studied even where there is no book-view, as Durkheim did in his
work on the Australian
Aboriginals who had no book, hence no
book-view. But India is not just an aggregate of tribal and peasant
communities. It is and has been a major civilization in which the
book-view, or, rather, different and even competing book-views have

existed for two thousand years and more. The social reality on the
rarely changes without some change in collective

ground
representations; and when those change, the book-view also undergoes
change.
Just as the social the ground and its morphological
reality on
framework change, so also do
collective representations and the
authoritative texts in which they are encoded. The authoritative texts of

the past no longer enjoy their old authority today. Their influence has
declined, although it has by no means disappeared. Here it is useful to
remember that the late Professor P. V. Kane found a place for the
Constitution of India in his history of

monumental work on the
Dharmashastra. Nor am I talking only of the Dharmashastra, with or
without inclusion of the Constitution. Today, the book-view of Indian
society may be found in a variety of texts: legislative debates, judicial

decisions, political manifestos, essays, pamphlets and books of a great
these documents, but it figures
variety of types. Caste figures in many of
more commonly as jati than as varna, in contrast with the ancient and
medieval texts.
Sifting this vast and amorphous mass of material for convincing

evidence of a clear direction of change in the social perception of caste
is no easy task. It is
not something that can be accomplished
single-handedly by any individual scholar. I cannot say that I have even

made a proper beginning of a systematic enquiry. The only point I would
like to make very briefly here is that the really crucial evidence of the
shift in representations of caste will be found not in English but in the
Indian languages. I can claim some competence in only one of those
languages, namely Bengali, and I have been struck for some time that
Bengalis, particularly of the younger generation, hardly use the term
varna or (barna) in either speech or writing. Casual enquiries from those
whose mother tongue is Hindi seem to indicate that something similar is
happening there as well; beyond that, I am not able to even suggest

anything further.
The idiom of varna has no doubt been used extensively in the present
century in the process of upward social mobility described as

Sanskritization(Srinivas 1966). It is possible that the new opportunities
provided by censuses and ethnographic surveys since the end of the 19th
century may even have revived to some extent the language of varna
among groups aspiring to upward social mobility. But the impression is
that this trend reached its peak in the earlier part of the present century
and is now on the decline. When so many castes with manifestly
inappropriate antecedents claim that they are Kshatriyas, the category

itself is bound to become devalued.
Where sixty years ago a caste would
claim to be Kshatriya, today the same caste might prefer to be
designated as backward. This is not a change of small significance.
Srinivas seemed to suggest that there was an error in describing caste
as varna and that it should be described as jati. My view is that this is
not just a recognition of error, but also a response to change. A decade
before the publication of Srinivas's paper on varna and caste, the social
historian Niharranjan Ray (1945) published a book in Bengali entitled
Bangali Hindur Barnabhed, meaning caste among the Bengali Hindus.

Similarly, the anthropologist, N. K. Bose, who wrote much in Bengali
(1949a; 1949b; 1975), often for literary magazines, freely used the term
barna in describing caste. There is a repertoire of terms relating to varna
Varna and Jati 19
or barna: barnabyabastha, the affairs of caste; barnabinyas, the

arrangement of castes; asabarnabibaha, inter-caste marriage;
barnasankar, offspring of mixed unions; and so on. My impression is
that these terms are now far less commonly used among Bengalis than in
the thirties and forties.
What appears remarkable in retrospect is the continuing use of this
language in a social context that was making its categories obsolete.

Bose, in particular, was tireless in pointing to the forces that were
disrupting the design of traditional Hindu society. Both he and Ray were
well aware that the actual divisions of Bengali Hindu society did not fit
at all well into the traditional scheme of varnas: there were
only
Brahmins among the three upper varnas, the rest being in some sense or
other Shudras. Such has been the actual state of affairs for decades or
even centuries, yet the old language continued in use right until our own
time.
Much of Bose's description in fact related to such functional castes
and subcastes as Telis, Kumhars, Lohars, and so on, which he would
certainly recognize as jatis\ yet he commonly used the language of varna

to refer to such general features of their social arrangement as division of
labour, rules of marriage, and so on. One of the reasons in his case might
have been his interest in the distinction between tribe and caste, and in
what he called method of tribal absorption (Bose 1941). He
the Hindu
repeatedly argued that Hindu society had a distinct design, and that
non-Hindus, from both within and outside, had fitted themselves into it.
He continued to use the language of varna because of his interest in that
design even while he pointed out that it was being undermined by

internal and external pressures to an extent that had no precedent in the
country's history.
Srinivas (1962: 69) said at the end of his brief essay: 'Varna has
provided a common social language which holds good or is thought to

hold good, for India as a whole.' What I am arguing here is that it is this
that is
now, before our eyes, becoming obsolete and
language
anachronistic. When Bengalis speak or write about caste, they no longer
use barna as commonly as before, but jat in the spoken language, and
also jati in the written form. Their experience and perception of caste
has changed, and this change is expressed in the shift of vocabulary.
When I discuss the caste system with young, educated, upper-caste
Bengalis now, I am
struck by their lack of familiarity with the old
vocabulary pertaining to inter-caste marriage, hypergamy, offspring of

mixed unions and so on. Part of this is due to the reduced attention paid
in schools to the teaching of Sanskrit which was the basis of Bengali
grammar and etymology; another part is due to the obsolescence, or at

least the attenuation, of an old social code which upper-caste Bengali
children imbibed at home without conscious effort; and part of it is due
to a shift of attention brought about by the enlarged role of caste in
politics.
Both varna and jati are polysémie terms, and therefore it is natural
that there should be a large overlap of meaning between the two. Many
writers on the subject took colour to be the primary meaning of varna,
and sought its origin in the distinction between the light-skinned Aryas
and the dark-skinned indigenous population (Ghurye 1950; Srinivas
1962: 63-69). But Mrs Karve (1968: 50-52) rightly pointed out that the
term had other connotations in the early sacred literature and
grammatical works. It is best to adopt her suggestion to use the word

varna in the sense of hierarchical order, and to refer to the four varnas as
the four orders of society. This would be in conformity with the
European usage which spoke of the orders of society, or the three orders
or estates (Duby 1980). That usage continued in the English language
until late; Adam Smith, for instance, spoke of orders rather than classes.
It was only in the 19th century that the concept of order was displaced
by that of classes in response to major changes in society.

N. K. Bose drew attention to the varieties of categories to which the
concept of varna was applied. He wrote: 'The division into varna is not
confined to human society; it is widely known that even lands or temples
are classified into Brahmin, Kshatriya and so on' (1975: 91). Earlier he
dwelt in particular on the classification of temples into varnas (Bose
1964). 'In effect we may regard the varna system as a
He concluded,
particular method for dividing into classes various kinds of phenomena,

beginning with human society' (1975: 91). It was, in other words, the
pre-eminent scheme of social classification established by Hindu
cosmology.
Conceptually, the order of varnas is not only exclusive, it is also
exhaustive. The Dharmashastra
says Brahmin,
Kshatriya, Vaishya,
Shudra, these are the four varnas and there is no fifth; this means that in
Varna and Jati 21
principle all of mankind can be fitted into one or other of the four
varnas. According to Bose, this was regularly done in the past when
varna was an active principle of social classification. 'Whenever in
ancient India men came in contact with different communities, they tried
to find a place for them in one or another varna according to their
qualities and actions' (1975: 91). Varna has ceased to be an active

principle of social classification; it has been displaced by other
principles.
It is obvious
that varna did not cease to play an active part in the
arrangement and rearrangement of groups in society all at once. What is

now happening seems to be the culmination of a very long and tortuous
process. When the British established their empire in India, the new
rulers couldno longer be accommodated within the scheme of varnas:
here one might find significant differences between British India and the
princely states. The

process had started earlier, with the Islamic
conquest, although Muslim rulers adapted themselves to the Indian

social climate much better than their European successors. Nevertheless,
both Muslims and Europeans had other models of rulership, and where
their authority became established, the category of Kshatriyas inevitably
became emptied of some of its meaning.
Where large sections of the population became converted to Islam, as
in the Punjab and Bengal, it became difficult to fit those sections into the
scheme of the four varnas. To the extent that social divisions such as
those between Rajputs, Jats and Ahirs survived the conversion to Islam,
some continuity with the old forms of representation was maintained.
People recognized the gaps and inconsistencies, but still used the
language of varna in writing about caste.
II
The idea of jati is also an old one and has been used, along with that of
varna, for a very long time to refer to caste. But the connotations of the
two have perhaps always been a little different. The term jati refers more
to the units that constituted the system—the castes and communities—
than to the system viewed as a whole. It did not provide the kind of basis
for a universal social classification that varna did. Unlike the varnas, the
jatis were not thought of as being exhaustive in a formal sense. We have
noted that the Dharmashastra named the four varnas, and said that there
was no fifth. One cannot draw up a complete list of all the multifarious
jatis and declare categorically that none exists besides those listed. New
jatis could always be added on, but not new varnas.
Perhaps the term jati has been used more commonly than the term
varna for a very long time. It is also a polysémie term, and I am
suggesting that today it can be stretched to accommodate all kinds of
units that cannot be accommodated by varna. For instance, it would be
strange to describe the Muslims as a varna or a segment of a varna,

whereas it is common to hear them being described as a jati\ their
subdivisions, whether of the sect or the caste type, may also be described
as jatis. There being no fixed number of jatis, the word allows itself to
be used for denoting a group as well as a subdivision of it.
Whereas varna refers primarily to order and classification, the
primary reference of jati is to birth and the social identity ascribed by

birth. It is thought of as a natural kind whose members share a common
substance, although the sense of that may be weak or strong, depending
on how broadly the group is conceived. Jatis, unlike classes, are thought
of as organic divisions, self-generated and self-reproducing.
The term jatiis, if anything, even more elastic than its counterpart,
varna. It may refer to a very small group, such as a subcaste or a
sub-subcaste; it may refer also to the whole of humanity. Bengalis speak
commonly of the Sadgope or the Kayastha jati, but also of manabjati or
manushyajati. In current Bengali usage, the term may be applied to
Europeans, Germans, American Blacks, Muslims, Madrasis (meaning

south Indians) or Punjabis. The idea always is that the members of a jati
share some qualities in common which give then a distinctive identity
that is somehow present even when it is not visible. Men and women
may be referred to separately as jatis—strijati and purushjati—but not, so

far as I am aware, capitalists and workers.
Anyone who has tried to conduct an ethnographic census among
Bengalis in the Bengali language will know how frustrating it can be to

secure comparable information on caste. The entries under that column
frequently contain such items as Jain, Oriya, Sayyad, Sikh, Adivasi,

Santal, and so on, in addition to the names of castes as understood in the
sociological literature. Census-takers with tidy minds have always found
this to be a nightmare.
Varna and Jati 23
Every anthropologist has at one time or another been outwitted by his

informants, and I too have my own tale of woes. When during my
fieldwork in Burdwan, I asked my informants to which jat they
belonged, some of them naturally put the same question back to me. The
answer that I did not belong to any jat was rarely taken seriously.
Puzzled by my name, they would ask whether I was not in fact a
Bengali. When I pointed out that that had to do with my mother tongue,
not my caste, they would say, 'Ah, then you are a Christian.' If I denied
that, a sarcastic bystander might ask, 'Then I suppose you are a
Frenchman?' The point is that my informants—and indeed many of my
Bengali friends—believed that if I could not say that I was a Brahmin or

a Kayastha, I should admit to being a Bengali, if not that, at least to
being a Christian. Practically anything might serve; what does not serve
is not having any jat at all.
It is true that even today, the vast majority of Indians think of a
person without a jati as an anomaly; indeed, they suspect that such a
person probably has something to hide. At the same time, it must be

recognized that jati here includes other kinds of units besides those that
are listed and ranked as castesin works of ethnography. In this wider
sense, jatis are not always or necessarily ranked. Many upper caste
Bengali Hindus speak of Bengalis (meaning Bengali Hindus) and

Muslims (including Muslim Bengalis) or Christians (including Christian
Bengalis) as different jats, but that does not mean that they think that
Hindus, Muslims and Christians are unequally ranked. Similarly, when
they speak of Oriyajati and Telugujati, they think of them as different
rather than unequal.
At first sight, such units as Brahmins, Muslims,
Bagdis,
Sadgopes,
Oriyas and Santalsappear to be extremely heterogeneous. They cannot
be thought of as the differentiated parts of any kind of system based on
the division of religious functions. Hence they cannot be thought of as
varnas or fitted into the order of varnas. But such units are precisely the
ones that are increasingly competing with each other in the political
process. In Bengal certainly, and perhaps in other parts of the country as

well, when people think about caste today, they think less about religion
than about politics. Hence they find it more natural to represent caste as
jati than as varna.
In the sixties, some anthropologists argued that when castes compete
with each other in the political arena, they act in contravention of caste
principles (Leach 1960). One might perhaps say this about caste in the
sense of varna but hardly about caste in the sense of jati. The
competition for power between castes and between coalitions of castes

and communities is perhaps the most conspicuous feature of
contemporary Indian politics. Here it would be misleading to represent

the contending parties as varnas, but quite appropriate to describe them
as jatis. The order of varnas necessarily entails a hierarchy of ranks,
whereas the competition for power takes place between equals, or near
equals. What one caste lacks in ritual status, it may make up by strength
of numbers; where its members are wanting in educational attainments,
they may advance through superior capacity for organization.

Castes have become increasingly involved in politics, but they have
not ceased to be castes (Beteille 1969). Electoral politics increases the
consciousness of caste, and at the same time creates networks of
relations across caste (Srinivas The old cleavages
and Beteille 1964).
between castes are continuously redefined by the formation of new
coalitions among them. The sense of a common identity defined by birth
and of a shared substance among members of the same caste provides a
strong basis for the mobilization of electoral support (Kothari 1970). At
the same time, the momentum of democratic politics creates coalitions
between all kinds of groups, only some of which can be plausibly related
to the traditional order of varnas.
All these different
types of groups-castes, tribes, sects,
denominations, religious and linguistic minorities—may, depending on
context and situation, be designated as jatis. There is little sign of any
decline in their active participation in the competition for scarce
resources.Nor are they active only in politics. We shall presently see
how attachment to the group also provides a sense of economic security
to its individual members.
I have pointed out that all these various groups—those listed by Mr.
Mandai's Commission and those being catalogued by the
Anthropological Survey of India—may be called jatis, at least in the

Bengali language. But can they all be legitimately designated in the
English language as castes? There appears to be a problem of translation

here. The term caste answers only partly but not fully to what Bengalis
mean by jat or jati, which may refer also, according to context and
Varna and Jati 25
situation, to tribe, sect, and religious or linguistic minority. It is in this

light that we have to view the increasing use of such terms as ethnicity,
ethnic identity and ethnic group by sociologists and others to describe a
significant feature of contemporary Indian society and politics. I am not

suggesting that these are the most suitable terms, but they seem to
answer better than the term caste to the mixed bag of social groupings to
which I have been drawing attention.
Recently, Professor Srinivas has observed, 'In the future too caste
will remain important in Indian life. But it will be conceived more in
terms of ethnicity' (Padgaonkar 1993). That sums up very nicely what I
am now trying to say. Those who had feared that the organic unity of
society represented by the order of varnas would be disrupted by the

new economic and political forces have had their fears confirmed, for it
has become increasingly difficult to use varna as a standard of reference
for describing the relations between castes. But those who had hoped
that the new economic and political forces would lead to the demise of
caste also have had their hopes belied, for the collective identities
represented by the idea of jati have shown remarkable tenacity.

Writing in the twenties and thirties, Mahatma Gandhi (1962)
represented the moral order of Hindu society in terms of varna, and still
hoped that it could be revived and renovated for the benefit of all. It is
that memory, filtered no doubt through rose-tinted glasses, that lingered
in the minds of many Hindu intellectuals who wrote about caste until
almost the time of independence. Today, it is difficult to invoke even the
memory of a moral order in writing and speaking about caste.
When people now write about caste, they do not write about morality
but about politics, the two being viewed as widely different, if not
opposed in their nature. But is not loyalty to the community of one's

birth, whether viewed in terms of language, religion, caste, sect or tribe,
itself a moral fact in the sense given to the term by Durkheim? Here, the
matter is somewhat complex because while people might concede that
be legitimate, they
loyalty to language, religion, sect or even tribe may
seem less prepared to make the same concession for loyalty to caste. The
reason for this is that no matter how strong the pull of collective
to the
loyalties may be, our Constitution and our laws give primacy
to accommodate
rights of the individual. Those rights might be required
the claims of religion and culture to some extent, but it is difficult to see
why they should yield to the demands of caste which seems merely to
divide without providing anything beyond some undefined sense of
security to its individual members.

We should not underestimate the moral force of the sense of security
that attachment to caste and community gives to the individual in a
changing and uncertain world. It is to this that N. K. Bose drew attention

in one of his later writings, a brilliant essay on the motley assortment of
castes and communities that make up the city of Calcutta. He described
the various castes among the Bengali Hindus—Kayasthas, Kansaris,
Namshudras—living cheek-by-jowl with Oriyas, Sikhs, Urdu-speaking
Muslims, Bengali-speaking Muslims, Gujarati Baniyas, and many
others, all regarding themselves and regarded by others as so many
different jats. He drew attention to the economic compulsions that kept
them together, reinforcing in each a sense of its distinct identity:
Because there are not enough jobs to go around everyone clings as
closely as possible to the occupation with which his ethnic group is

identified and relies for economic support on those who speak his
. language, his co-religionists, on members of his own caste and on
fellow immigrants from the village or district from which he has
come (1965: 102).
The continuing strength of these collective identities is a reflection of

the failure of the institutions of civil society to take root and gather
strength in independent India. Civil society requires a variety of open

and secular institutions—schools, universities, hospitals, municipal
corporations, professional bodies and voluntary associations of many
different kinds—to mediate between the individual and the wider society
of which he is a part. At the time of independence, it was hoped that
these open and secular institutions would give shape and substance to
democracy in India and at the same time drive back the consciousness of
caste and community. They have failed to provide what was expected of
them, and it is no surprise that the older forms of collective identities
have not only held their ground but become increasingly assertive.
Varna and Jati 27
REFERENCES
Beteille, A. 1969. 'The Politics of "Non-Antagonistic" Strata', Contributions to Indian

Sociology, n.s. 3 (1): 17-31.
Bose, N. K. 1941. 'The Hindu Method of Tribal Absorption', Science and Culture, 7:
188-94.
. 1949a. Nabin O Prachin.Calcutta: Bengal Publishers.
. 1949b. Hindu Samajer Garan. Calcutta: Vishvabharati.
. Traditions in Indian Architecture. Calcutta, mimeo.
. 1965. 'Calcutta: A Premature Metropolis', Scientific American, 213 (3):
91-102.
1975. The Structure of Hindu Society (Hindu Samajer Ganan, 1949, translated
from the Bengali with an introduction and notes by A. Beteille). Delhi: Orient
Longman.
Gandhi, M. K. 1962. Varanashramadharma. Ahmedabad: Navajivan.
Ghurye, G. S. 1950. Caste and Class in India. Bombay: Popular.
Karve, Irawati. 1968. Hindu Society: An Interpretation. Poona: Deshmukh.
Kothari, R. (ed.). 1970. Caste in Indian Politics. Delhi: Orient Longman.
Leach, E. R. (ed.). 1960. Aspects of Caste in South India, Ceylon and North-west
Pakistan. Cambridge: Cambridge University Press.
Ray, Niharranjan. 1945. Bangali Hindur Barnabhed. Calcutta: Vishvabharati.
Srinivas, M. N. 1962. Caste in Modern India and Other Essays. Bombay: Asia.
. 1966. Social Change in Modern India. Berkeley: University of California
Press.
. and A. Béteille. 1964. 'Networks in Indian Social Structure', Man, 64 (212):
165-8.
Social Scientist
Caste: Ideology and Context

Author(s): Suvira Jaiswal
Source: Social Scientist, Vol. 25, No. 5/6 (May - Jun., 1997), pp. 3-12
Accessed: 03-03-2015 07:31 UTC
SUVIRA JAISWAL*
Caste: Ideologyand Context
Current definitionsof caste place its essenicein endogamy,1 which is

regarded as a residue fromthe past existingmore or less in a 'passive
way'. Its role in the occupational systemand social placement is not
denied, in fact its past links with the formof the 'labour process' are
duly underlined in Marxist historiography,2but since it is seen as
arising out of the incomplete fusion of tribal elements in the general
'Aryan', 'Vedic' or 'Sanskritic' society,3 the religious or ritual aspect
of caste, the custom of endogamy does acquire the primarilyrole in its
foundation making the play of political and economic factors in its
originias secondary. Such a perception makes caste a transhistoric
institution,a continuation of old clan and tribal units predating the
state.4 It is generallyheld that its influenceon political and economic
behaviour has declined with the development of capitalism and
industrializationin India and now caste operates largelyin personal or
religious matters only. Hence, it is being defined as a form of
differentiationwherein constituent units form not a continuous
hierarchybut 'discrete categories' observingritualized social practices
and endogamy, justified 'on the basis of putative biological
differences'.5
However, locating the basic, fundamental principle of the
institutionof caste in endogamynot only gives it a static characterbut
also provides gristto the mill of those who trace its originto the desire
of the conquering Aryans to keep the subjugated communities at a
distance. Thus despite the severe criticism6of making caste hierarchy
related to the proportional mixture of Aryan and non-Aryan blood,
racial theory continues to exercise a good deal of influence7 in the
academic as well as noni-academic popular circles. D.D. Kosambi
wrote that the Brahmana caste emerged as a result of 'initeraction
between the Aryan priesthood,and the rituallysuperior priesthood,of
the Indus culture'8 and the assimilation of the Aryan and non-Aryan
muLst have accelerated the formation of 'an internal, Aryan caste
system, essentially the separation of the Brahmin in function and
disciplinefromthe Ksatriya and settingof both above the householder
Vaisya, after dasas had been conquered . . . for otherwisethere is no
reasoni for demarcating inlto endogamous caste'.9 We have critiqued
TProfessorof Historv,Jawaharlal Nehru University,New Delhi

Social Scienitist,Vol. 25, Nos. 5-6, May-June 1997
4 SOCIAL SCIENTIST
this hypothesis of Kosambi in greater detail elsewhere.10 For our

present purposes it is sufficient to point out that although the
assimilation of non-Aryan, aboriginal priests and ruling lineages as
well as commoners into various varna categories is a process which
has taken place throughoutthe ancient and medieval periods of Indian
history,there is nothingto show that such assimilationhad given rise
to separate endogamous units or jatis within the various varna
categories in the Vedic period. On the other hand, Kosambi rightly
holds that the subjugation of the dasas and changes in the means and
relations of production provided the material background for the
emergence of the 'four castes' fully developed within the tribe:
Brabanana, Ksatriya, Vaisya, Sudra, in Yajurvedic times and this was
a new class division witlin each Aryan tribe'.11 The description of
varnas as 'castes' here is significantand shall have to be discussed
later.
Nevertheless, Kosambi's remarks on caste are not without
ambiguity,which leads IrfanHabib to state that according to Kosambi
'castes did not arise out of any internaldivision of the varnas in the
original Vedic society, but from an external process altogether',
through the fusion of tribal elements into a general society. This
assumes that in Kosambi's perception varnas were 'estates', 'order' or
'class' and not 'castes', a view contradicted by the passages quoted
above. Habib adds12 that tribes 'are often rigorouslyendogamous'
and may have practised endogamy when they enteredinto the general
society.
But the question whether caste endogamy should be viewed as a
continuation of tribal endogamous customs is far from settled.
Scholars such as Max Weber13 and Louis Dumontt4 have contrasted
tribe and caste on the ground that the internalstructureof the tribal
world is exogamous and of caste society endogamous. In our opinion,
Dumont's assertion that 'endogamy is a corollary of hierarchy,rather
than a primaryprinciple'15 of caste deservesserious consideration,of
course, with the caveat that for us caste hierarchyis not just a religious
principle but a system of exploitation of gender and weaker
communities.16
Kosambi held that caste is 'class at the primitive level of
production'17 and originatedin later Vedic times as varna divisions.
However, sociologists generallymake a clear distinctionbetween the
zvarna and jati regardingonly the latter as castes and the formeras
'estates' 'order' or status system'.This is done on the ground, that (a)
v'arna are broad categoriesand the real effectivesocial units today are
castes; (b) varnas are only four but castes are numerous, (c) anid
wlhereas varna hierarchyis clear, there is a lack of clarity in the
hierarchyof castes, particularlyin the middle regions. However, this
CASTE: IDEOLOGY AND CONTEXT 5
view of caste organization may be a more or less accurate description

of its morpllologyin moderntimes,but this does not mean that such a
complex caste organization could not have its beginnings in the
simpler varna organizationi.This is not to argue in favour of a linear
development; nor to account for the proliferationof jatis within the
varna structurein terms of fission of primary;categories throughthe
mixed (varna samkara) marriages in the manniervisualized by the
Dharmasastra writers; but to assert that whereas varnas emerged in
smnall-scalelater Vedic societies as tribes disintegratedwith the growth
of economic disparities,the jatis; emerged in post-Vedic times when
the varna-based society began to expand spacially and
dem-ographically.Both the processes of internal fragmentationand
incorporation of tribal communiitiesproduced the complex structure
of jatis withinthe varna framework.The patternof the structuringof
the wider society was provided by the principles of the varna
ideology, which was concerned not only with the grading of statuses
but with the perpetuation of hierarchy through insistence on the
1hereditary nature of occupations and rankingsand devaluation of the
status of women in-builtin the notions of hypergamyand endogamy,
the two being two sides of the same coin.1 8
However, interpretingvarna as 'category' or 'estate' and jati as
caste, Dumont distinguishesthe two on the ground that although there
is 'homology' and a certain degree of osmosis between- the two
systems,tlheirprinciplesare differenit, as in the formerstress is laid on
funlctioniratherthan birth but heredityis more importantin the caste
system.1 9 We may point out that the importance of the factor of
heredityin the varna scheme is obvious fromits linkage with theoryof
gunas, the threebasic qualities namelysatya (goodness or purity)rajas
(passion or glory?)20 and tamas (darkness, ignorance). Th-e
Bbagavadgita clearlysays that owing to theirnatural,inborn, qualities
(gunas) the four varnas have been assigned differentfunctions.21
Dumont argues that the fact that ruling dynasties of differentorigin
could acquire ksatriyastatus in later times shows that functionrather
than hereditywas importantfor the varna categorization.Btuthe does
niottake into account the fact that the ascription of ksatriyastatus to
people of differentorigiinswas itself dependent on their achieving
putative hereditarylinks with the ancient ksatriya lineages or heroes
through invented genealogies and those rulers who did no seek SuLCh
validation continued to be regarded as mlembersof lower varnas, for
example, Harsavardhana of Kanauj and Reddi kings of Andhra. The
hereditarynatuLreof var;za iden-tificationi is inidicatedby refer-enicesto
Ksatriya merclhanits (vaniks) in the records of the early of
cenituLies the
Clhristiainera in the Andlhi-acountry22 anid in a Gurjai-a-Pratilhara
inlscrition23 century fromithe Doalb regioniof IJ.P.
of the tenitlh
6 SOCIAL SCIENTIST
The main plank of the entire edifice of Homo Hierarchicus as

constructed by Louis Dumont rests on the assumption that the
separation of.the 'hierarchicalstatus', and the 'secular power', that is,
the brabsa and the ksatra elemenitsin the varnza schenie and the
subordinationof the latterto the formeris unique to the way Hindus
perceive things; and this is what makes caste a 'state of min-d',or
'systemiiof ideas aind values' based on a religious principle. For,
although Dumont insists on distinguishling between varna hierarchy
and caste lierarchy, he regards this feature of the v'arina system as
crucial for th-e slhaping of the relationship betweeni (religious)
'hierarchy' and (secular) 'power' in the caste system.24 However,
i-athertlhaniattributingit to the,idiosyncraticpsychiologicaldisposition
of the Homiio hierarchicus imagined as the 'otlher', the perfect
couniterfoilof Homo economicus, or Homo aequalis, as Dumont
describes the westerniindividual, one should examninethe material
context in wlich tlis ideology evolved. In his study25 of the East
African Nilotes and the Proto-Indo-IraniansBruce Lincoln has drawni
our attention to the remarkablesimilaritiesbetween the two cultuires
wlhiclhwere far apart in time and space with nio obvious links. The
most important of these is the dichotomizationiof the priest anid
warriorcategories; and he traces this traitto the ecology of the cattle-
keepers. Earlier, conitestingthe theory of Dumezil regarding the
typicalityof tripartitedivision among the Indo-EuropeanlsJohn Brough
had shown-26 that the three-foldoccupational division consistinigof
priests,warriors and com-nmoners existinigamong the pastoral Semetic
societies depicted in the Old Testament too. We hiaveargued27 that
the cattle-keepers tenided to develop two kinds of specialists, the
priests, who claimiled to provide protection and increase in cattle
wealtlhof the tribe by developing rituals of offeringsto gods of th-e
most valued itemsof their society,particularlyanimals, and in this way
secuLreddivinieblessiings,and warriors,who defendedthe tribe and its
cattle-wealtlh anid sulpplemented it through cattle-i-aids. It is not
Sliprising that those uwhowere regardedas capable of mediatingwith
the gods anid had control over rituals were considered superior and
'pUrer' than the latter category. It is this context which explains the
separation of the brabmza and the ksattrafrom the vis in the Vedic
society. In the process of class formatioinin the disintegratingVedic
tribes, botlh brabmiia and ksatra are seen as competing as well as
cooperatingwith each other against the vaisya commonerand defeated
sudra, and gradually hardening into castes. Multiple factors, wlich
include the processes of fission as well as fusion, lead to the
emiiergence of the jatis and widening of the varna categories. India
being a countrywitlha long anidcontinuous hiistory for more than two
thousand years has also carried the burdeniof its tradition, wlhich
nievertlhelesshas been continuously modified and reshaped in the
CASTE: IDEOLOGY AND CONTEXT 7
changed circumstances, without making a complete break from the

past. It requires a thorough study of the historyof various regions of
India to grasp the factors which contributed to the specificities of
regional social formations under the overall umbrella of the varna
ideology.29 The expansion of the varna systemto differentregions of
the subcontinent did not mean replication of the varna categories
everywherein a uniformmanner but securingthe subordinationof the
deprived classes in favour of dominant groups at the ideological level
following the basic principlesof the varna system.Thus in the south
where specific historical conditions did not favour crystallizationof
the ksatriya and vaisya categories, the land owning agriculturist
community of the Vellalas was ranked as sudra. Nevertheless, it
functionedas a status categoryat the regional and subregionallevels. It
is pointed out that communitiesquite distinctfromthe Vellalas have
attemptedand succeeded in gaining the Vellala status; and there is a
Tamil saying that a low-caste person, when he becomes prosperous,
tries to pass off as a Vellala.30 We have suggested elsewhere that the
perception of agricultureas the functionprimarilyof the sudras (and
not the vaisyas) and categorization of the dominant agriculturist
communitiesof the south as 'sudra' led to drastic modificationsin the
concept of the 'sudra' and the varna theory.These changes ultimately
culminatedin the shiftto the claims for superiorstatus being made on
the basis of the purityof birthand not function.It is significantthat the
bralimanical law-books impute the 'impurity' of social groups to a
hereditaryfactor, their descent from parents or ancestors wlho had
committed varna-samkara, that is, had married or had sexual union
in the pratiloma, disapproved order. They were condemned for thleir
imnpurebirth to carry out impure tasks, their impuritydid not arise
from their impure occuipations at least theoretically. Thus in
contradictionto what Dumont asserts: 'it is specialization in impure
tasks, in practice and theory, which leads to the attribution of a
massive and permanentimpurityto some categories of people' (p. 85),
brahamanical ideologues imagined the origin of impurity quite
differently.The realityof exploitativerelations,regional variations and
sharp differencesof status had to be encompassed within the divinely
ordained varia divisions.
More recently,George L. Hart (11)31 has argued that the notion of
impurityof certainsocial groups is to be traced to anicientTamils, who
conceived the divine in terms of dangerous, malevolent spirits
manifestingthemselves in death and anarchy. These qualities were
extended to those social groups whose functionwas to mediate with
and control the sacred, and contact with such persons came to be
looked upon as dangerous and polluting. This, in his view, accounts
for the low, untouchable status of the Paraiyans and the Velans
(officiatingpriests for the worship of the god Murugan). We have
8 SOCIAI. SCIENTIST
cr-itiquecdthese views elsewhere.32 The point whiclh deserves

explanationiis thlatwhy those who had the 'power' to harm others fell
in social status anidthose who were vulnerablecame to be regardedas
'sLperior' aind high caste? Dumont is rightin statingthat the notion of
impurityin thiecaste systemis differentfromthe notion of the sacred
beilg dangerous, whichiconcept prevails in some tribal societies. He
draws attentionto the reversalof the roles, 'the tribal chief is taboo,
i.e. dangerous for the common people,. whereas the Brahmnanis
vulnerable to pollution by anl inferior'.33 And if it was merely a
questioniof avoiding contaminationfromthose who were close to the
'danigerous sacred', it is strange that the ancient Tamil king who
according to Hart was the 'central enmbodiment of sacred powers' did
not sufferan)ytaboo or inherethe quality of pollution.
In fact the veryidea of wllat is pollutingseems to have evolved over
a lonigspani of time in. whiclhmaterial factorssuclhas depression of a
communityengaged in a particular type of mnanuallabour played a
crucial role. The case of leather and leatherworkers(the carmakarcas,
imoderin clhamar caste) provides a classic example. The later Vedic
texts34 describe milk, honiey,clarified butter (ghee), etc., filled in
leather bags for sacrificial use without any trace of thlenlotionthat
leather was a polluting substance Enough evidence has been
produced to show that the professioni of leather work was not
consider-ed low in later Vedic times.35 The traditional Hindu's
aversion to leatlherand leather worker need not be traced to earlier
tlhanithe earl) nmedievalperiod when smnrtis such as thloseof Angiras
anid Atri begin to mentionihiimamong the antyajas (lowest castes).36
OuEr argumetntreceives further support from two records of the
thirteenitlhcelntuLryincluded in the Lekhapaddhalti37, a collection of
miiodeldocumzenits comnpiledin the sixteenthor seventeenthcenturyin
westerniIndia for the use of the scribe. These speak of the same female
slaves being obliged to do both 'clean' as well as 'UnClean'work, such
as, cooking food, fetchingdrinkingwater as well as cleaning the drainls
and tlhrowingaway human excreta, etc., a situation which would be
untlilnkablein modern HinduLhomes.
In ouI- opilnion,the concept of inniateand relative puLrity of social
groups emerged as the ostensible explanation for the excessive
differenitiationisin a hierarchical society becaulse of the crystallizatioll
of a large brahmana caste at the apex, whlichfeatUre,ais we lhaveseeni,
was embedded in the ecologyof the Vedic cattle-keepers. It contilnued
to derive sustenanicefromthiesignificantrole of the barahmanas in thie
later power struLcturebotlh as ideologues and active participants,and it
is nlotsurprisingthat the world-viewof the priestlvcaste should show a
pre-occupation witlh the nlotiolns of the pure and the impur-ein
personial as well as social splherestracingtlhemii to Vedic roots. Tlis is
not a '"Voluntarlist' or 'determaiinistic'
tlhesisbut anl assertion of the point
CASTE: IDEOLOGY AND) CONTEXT 9
that social environment,which includes material as well as ideological

factors,plays a crucial role in shaping the way onie comprehendsand
tries to mould tlle world. Thus, when killingof cattle for sacrificeor
food became unpopular in the material milieu of the sixth centuryBC
and 'non-killing' was raised to the level of a sacred principle, the
priestly caste could not remain unaffectedby it.38 Vegetarianism
became the norm for the brahmana to be emulated by others who
wislhedto upgrade their status through a process of sanskritization.
Nevertheless,it was not merelyideological propaganda which led to
the wide acceptance of the varna system.It is contendedthat the caste
systemworked in the favour of the dominiantcommunitiesby securing
the structured dependence39 of menial and 'untouchable' castes. This
provided cheap labour to the land-based communities in the
countryside,reducing the cost of generatingsurplus and made more
revenuieavailable to the state.40 The congruence between caste and
class was well marked in pre-colonialtimes,but ran-king was largelyin
termsof the relative 'purity' of social groups witlheach region having
itS OWnl specifichierarchybased on the same general principles.
ThUs the functioning of the principle of 'purity' in the varnca
structurewas aimed at justifyingand strengthening the domin-anceof
th-osewho occupied the top rung of the ladder. Once the rules of the
game changed and numbers rather than the purity of birtlhbecome
critical for securing political space in the conitemiiporary
democratic
set-up, the caste systembegan-ito adjust itselfsuitably,undergoingyet
another metamorphosis.Dumont describes it as 'substantialization' of
castes in which the spirit of competition replaces the relationship of
interdependence and there is 'a transition from a structure to the
juxtaposition of substances'.40 It is interestingth-atDumont does not
deny that there h-as been a weakening of the concept of hierarchy
based on the opposition between pure and impure wlich in his view
had constitutedthe 'ideological core of the system',but he argues that
these changes reemainiincomplete as they affect only 'the politico-
econiomic domain which is encomiipassed in an overall religious
settinig.41 In his view recentmodificationsonly show tlle plasticityof
the system, which continiues to existg as the overall framework
remains unchanged. However, in spite of Dumont's assertion that
'calstes are still present anid ulltouchlability still effective',42 it will be
difficuLltfor aniyonlethe presenitday caste system is based on the
concept of ulltouclhability or the opposition of the pure aind impuLre.
Dumoniothiimiselfadmits that the rigiddity of the caste system operates
now^largely at the level coniubliuLm.43In other-words, the 'religioUs
setting'of the systemis nlOw provided niot by anlyniotionof hierarchy
of puritybLutby the pr-acticeof endogarmy.This is a CUliOUS colclusion
to .11riVe .at particularly in view of Dumonit's cearlier asseritioIn that
endogaaiy is a 'corollary of hierarchy'anid not the genieralprinciple.
10 SOCIAI SCIENTIST
The difficultywith Dumont's view on caste is that not only it is 'a

bralhmianical view of caste' as Berreman remarked,44 and
overvalorizes the brahmanical theory-which in our opinion was an
inventionto explain and justifythe depressed condition of the lower
classes and niot an adumbration of its causative principles-it
completely overlooks the gender issue, the subordination of women,
whiclh is the critical constitutive element of this form of social
stratfication.Caste emergedas a systemof exploitationnot only of the
less privileged anid backward conmmunitiesbut also of women.
DLumont criticises45 the oft-quoted remark of Marx that modern
industryand the railways will dissolve the Indian caste system,which
rests on the hiereditary, division of labour, by pointingout that castes
still exist. But Marx was referring to the politico-economicaspects of
the system,which have been eroded to a large extent even according
to Dumont. Dumont, lhowever, regards these developments as
belongingto the secondaryaspect of caste, its primaryaspect is located
in the continuation of fragmentedidentities through endogamous
boundaries. However, as we have argued elsewhere,46 the evolution
of the caste systemis to be linked not only with the building of class
society but also with stabilization of patriarchy an-d genider
exploitation. Only those aspects of caste which are in conflictwith the
capitalist mode of productionhave weakened and are in the process of
disappearinigwith the impact of industrialization.But endogamywhich
involves the 'gifting' (kanyadana) of the bride in a manner whiclh
reiniforcescaste relationships is quite in harmony with capitalist
notions of private property tranisactions. Only unrestrained
participationof women in social productionon an equal footingwith
men could create conditions for a free mixing of sexes and erode the
patterniof arranged marriages, whlichare naturally delimited in the
traditionlal channels. The problem of caste cannot be tackled by
divesting it of its social context.
NOTES
I. AnldreBeteille, 1991. 'The Reproduction of InequLality;Occupation, Caste and
Family'. Contributionsto Indian Sociology (N.S.) Vol. 25, pp. 3-28.
2. Irfan Habib, 1995. Essays in Indian History: Towards a Marxist Perceptioni,pp.
161-189. Tulika, New Delhi.
3. Ibid., p. 165.
4. Morton Klass, 1.980. Caste: The Emergence of South Asian Social System,
Instituteforthe Studyof lIuman Issues, P'hiladelphia.
S. Dipankar Gupta, 1984. 'Continuous Hierarchiesand Discrete Castes', Economic
and Political Weekly,Vol. 19, No. 46. reprintedin Social Stratiticationed. by
Dipankar Gupta, 1991. pp. 110-141. OUP. Delhi.
6. See Louis Dumonit, 1972. Homo Hierarchicus. p. 64f. Paladiin,London; Suvira
Jaiswal, 1979-80. 'Studies in Early In.diani Social History: Trenids and
Possibilities' Indian Historical Review (henceforthIHR), VI nos 1-2, pp 2; 7-8.
CASTE: IDEOLOGY ANI) CONTEXT 11
7. Th-us Dipankar Gupta traces the rigid rules of endogamv and the rationale of
Inatural superiority', whiclh govern the jatis to the Arvan conquest and
subjugation of local indigenous communities,which provided the metaphor of
colour. Idem, 1980. 'From Varna to Jati: The Indian Castc System, from the
Asiatic to the Fcudal Mode of Production', journal of ColitemlporaryAsia, X,
pp. 249-271.
8. D.D. Kosambi, 1956. Introductionto the Stuidyof Indian History,p. 96. Popular
Prakashan, Bombay.
9. Idem, 1950. 'On the Origin of BrahminGotras', journal of the BonmbayBranichof
the Royal Asiatic Society,XXVI, p.50.
10. Suvira Jaiswal, 1989-90. 'Stratification in Rgvedic Socicty: Evidenice antfd
Para47digm17s,IHR, XVI, inos 1-2, pp. 1-34. A rcader interestedin this theme may
also referto idem, 1993-94. 'Mystifyingthe Aryans',IHR, XX, nos 1-2, pp. 219-
228.
11. D.D. Kosambi, 1956. p. 94.
12. IrfanHabib, op. cit.
13. Max Weber, 1968. The Religioniof Inidia,pp. 30-33. Translated and edited by
Hans H. Gerth and Don Martindale, The Free Press, New York.
14. Louis DuLmont,op. cit., p. 154.
15. Ibid., p. 156.
16. Suvira Jaiswal, 1977, 'Caste in the Socio-Economic Framework of Earlv India',
PresidentialAddress, Section I, Proceedingsof the IndianiHistory Conzgress,38th
session, Bhubaneswar, Idem, Dec. 1991. 'Semitizing Hindusim: Changing
Paradigms of Brahmanical Integrationi',Social Scienitist,XIX, no. 12, pp. 20-32;
idem Caste: Origi;i, Funiictioni
aiid Dynamn ics of Chanzge,Manlohar, Delhi, 1998.
17. D.D. Kosambi, 1965. Ciultuireaiid Civilizationiof AnicienitlIndia in Historical
Ouitli;ie,p. 50. Routledge alnd Kegan Paul, London.
18. The Satapatha Brahmana clearly states that a ksatriya is born of a ksatriya,a
vaisva froma vaisva and a sudra froma sudra (XIV. 4.2.27; 1I. 1.4.4.) Howvever,
initiallv the second Varna was designated as rajanya, meaning kinsmen of the
rala indicatingcrystalizationof the rulinglineages.
19. Louis Dumont, op. cit., pp. 106-114.
20. rajas is generally tranislatedas the 'quality of ptassion'. However, accordinigto
Monier-William's Sanskrit-Englisb Dictioniary ralas is at places equated with
teljas, meaninig'glory' or 'authoritv' (S.V. Rajas). According to V.S. Apte the
qualitv of ralas is 'the cause of great activity seen in creatures' (V.S. Apte,
San1skrit-Euiglish Dictionary, Vol. III, S.V., taais). Perhaps in the context of the
gunas the termishould be translatedas 'glory' or 'cnergy'.
21. kar,naiii pravibhaktani svabhava prabbairqinahi., Bhagvadgitti,XVIII. 41.
22. K Gopalachari 1941. Early Historyof the Anidhira Country,p. 91. Madras; Suvira
jaiswal, 1977, pp. 15-16.
2 3. Epigrapbia Indclica, XIX, no. 6, pp. 52-4.
24. L. Dumont, op. cit., pp. 105-6.
2S. Bruce Lincoln, 1981. Priests, Warriors and Caittle:A Study in the Ecology of
Religions, UnJiversityof CaliforniaPrcss,Berkelev.
2 6. John Brough, 1959. 'The Tripartite Ideology of the Indo-Eutropeans: an
Experiment in Method', Bulletiniof the School of Orienitaland AfricanStudlies,
XXII, pp. 69-85.
2 7. Suvira Jaiswal, 1989-90. 'Stratification in Rgvedic Society: Evidence and
Paradigms', in IHR, XVI, nos 1-2, p 15f. Idem, March-April 1991. 'Varna
Ideology and Social Change' in Social Scientist,XIX, nos 3-4, pp. 41-2.
28. For changes in the concept of 'vaisya' and 'sudra' and shift of emphasis from
functionito the puritv of the brahmana varna see, Idem, 1979-80 and March-
April 1991 (footnotes6 anid27).
12 SOCIAL SCIENTIST
29. Idem, Caste: Origin, Function atndDynamics of Change, Delhi 1998, (in press),
Introduction.
30. S. Arasaratnam, 1981. 'Social History of a Dominant Caste Society: The
Vellalar of North Ceylon (Sri Lanka) in the EighteenthCentury',in The Indian
Economic and Social History Review, XVIII, nos 3-4, pp. 377-91.
31. George L. Hart, 1975. The Poems of Ancient Tamil: Their Milieu and Their
Sanskrit Counterparts,Berkley,California.
32. S. Jaiswal, 1979-80. Section V.
33. L. Dumont, op. cit., p. 87.
34. Pancavinmsabrah"mana,XVI. 13.13., Baudhayana Srauta Sutra, XV. 16.
35. R.S. Sharma, 1980. Sudras in Ancient India, Motilal Banarasidass, Delhi. 2nd
ed., p. 53.
36. P.V. Kane, 1941. Histor) of Dharmasastra, Vol II part I, p. 70. Bhandarkar
Oriental Research Institute,Poona.
37. Pushpa Prasad, 1988-89. 'Female Slavery in Thirteenth Century Gujarat:
Documents in the Lekhpaddhati,' IHR, XV, no 1-2, pp. 270-75.
38. Suvira Jaiswal, 1981. 'Origin and Development of Vaisnavism', 2nd enlarged
edition, pp. 123-129. Munshiram Manoharlal, Delhi.
39. Idem, 1979-80, p. 21.
40. Louis Dumont, op. cit., pp. 274-5; 378.
41. Ibid., p. 275.
42. Ibid., p. 265.
43. Ibid., p. 378.
44. G.D. Berreman,1971. Contributionsto Indian Sociology, New Series,V. 16 f.
45. Op. cit., p. 265.
46. Suvira Jaiswal, 1977; 1979-80.
Social Scientist
Caste, Gender and Ideology in the Making of India

Author(s): Suvita Jaiswal
Source: Social Scientist, Vol. 36, No. 1/2 (Jan. - Feb., 2008), pp. 3-39
Accessed: 03-03-2015 07:02 UTC
Caste, Gender and Ideology
in the Making of India
CO
c
<
CO
I am deeply aware of the honour that the Executive Committee of the p_
Indian History Congress has done me by electing me the General President
of the 68th Session of this august body. I wish to express my most
sincere thanks. Thirty-six years ago Iwas privileged to preside over the
Ancient Indian Section of the India History Congress held at
Bhubaneswar. On that occasion I had pointed out1 that for a
restructuring of Indian society on the principles of equity and social
justice it was necessary to have a scientific understanding of the factors
which have given rise and continue to nurture a highly exploitative system
of social stratification in the form of caste. Although annihilation of caste
discrimination has been one of the main priorities of national agenda
since Independence, it cannot be gainsaid that caste continues to impact
in amajor way not only the sphere of personal relations but also various
? and economic ?
aspects of public arena legal, political including
access to land, water resources,2 etc. Hence it may not be inopportune
to reflect upon its historical roots and inner dynamics in order to have a
better understanding of the reasons of its tenacity and stranglehold.
The post-modern and neo-colonial critiques of caste visualize it as a
relatively modern phenomenon, a product of the British colonial rule,
traceable not to the ancient, the so-called 'Hindu' period of Indian history
or to the Purusa-s?kta of the Rgveda and theManusmrti but to the British
Census Reports.3 Hierarchy-cum-interdependence, occupational
specialization, endogamy and commensal restrictions, which the earlier

Indologists and sociologists regarded as the defining features of the
caste system, were from this point of view, Orientalist-Colonial readings
motivated by the desire to systematize diverse forms of local social
realities and identities into a holistic theory 'essentializing' it as a unique
culture. Despite this
criticism of 'essentialization', endogamy is
recognized implicitly or explicitly4 by these scholars as the essence of
the caste system, without which the formation of 'discrete categories' or
the allegedly modern process of 'ethnicisation' or 'substantialization' of
caste ? which according toNichdas B. Dirks5 make it 'aworthy synonym
of community in the best sense' ? would not have taken place. The
question arises: is endogamy then an irreducible transhistoric
phenomenon embedded in the psyche of the Indian people or it evolved
through a historical process and has continued to survive through the
3
Social Scientist
0 ages in a favourable material environment ?

3 In the first decade of the twentieth century the British census
^ commissioner Sir HerbertJiisley ascribed the origin of an endogamous
2 caste structure to the desire of the Aryan conquerors to retain their racial
purity from contamination of the blood of defeated aborigines. A
^
1 hierarchical gradation of people born of mixed unions is said to have
? evolved in proportion to the admixture of aboriginal blood in them with
the brahmanas at the top representing the purest of Aryan blood. Later,
_?^ '
rs the principle of endogamy was strengthened, perpetuated and extended
~
to all ranks of society by the fiction that people who speak a different
o
language, dwell in a different district, worship different gods, eat different
55 food, observe different social customs, follow a different profession, or
^ practise the same profession in a slightly different way, must be so
> unmistakably aliens by blood that intermarriage with them is a thing not
to be thought of'.6 The thesis of Aryan invasion and racial origin of
caste is no longer subscribed, but the impact7 of Risley's ideas may be
still seen in explanations offered for the prevalence of endogamy. Thus,
according to one view the caste system may be defined as a form of
differentiation in which the constituent units justify endogamy 'on the
basis of putative biological differences which are semaphored by the
realization of multiple social practices.'8 Risleys' notion that castes
(i.e.j?tis) considered each other 'aliens by blood' was an elaboration of
his theory of racial origin of the system. It is curious that although the
theory of racial origin is no longer accepted, the perception that there
exists a 'mythical notion of biological differences'9 in the ideologies of
all castes is retained, and it is argued that because of this notion castes
value the principle of endogamy very highly'.10 However, Imay point
out that the idea that there are inborn, biological differences which
distinguish one varna or j?ti11 from the other is a typically brahmanical
concept invented to justify the hereditary nature of varnas,12 and it
cannot be regarded as part of all caste ideologies, particularly of the
subaltern castes, which in their origin myths almost invariably trace their
descent from a brahmana or a ksatriyas ancestor. It is not biological
difference but the fact that the original ancestor was cheated or had
violated some rule inadvertently which is regarded as the reason for the
present predicament of his descendants.13 Moreover, rise of a new
endogamous unit through the processes of fusion and fission owing to
the adoption of some new technological, professional, religious or cultural
practice, and emergence of a new social group in medieval times is a
feature well known to historians and sociologists, but this cannot explain
the origin of caste endogamy.
Nevertheless, the view that caste endogamy is a residue of the tribal

g
past of communities integrating with the expanding Aryan society is quite ^
common,14 perhaps because history does provide many instances of ^
tribal groups being transformed into endogamous castes. But this only ~
shows that as a rule assimilation into a caste society could take place
~~
?|
only on a group or community basis with the new entrants retaining their
distinctive identities, for the general society was already fragmented into
social groups differentiated on hereditary principles. Attributing the
origin of endogamous customs to incomplete fusion of tribal elements
would imply that caste identities were biologically constructed, but in
my view caste society was not a biological but social construct.
I have argued elsewhere15 that caste endogamy was not a borrowing
or survival of aboriginal practice. It evolved and consolidated in the
process of regulating hierarchical subordination of social groups and
reproduction of patriarchy. It is not possible to agree with Dumont's
strongly idealistic view16 of caste which makes it 'above all a system of
ideas and values' embedded in the Indian mind, the Homo Hierarchicus,
presenting a perfect contrast to the Western Homo-Aequalis.17 In my
view18 hierarchy, defined as separation and superiority of the pure over
the impure, of the priest (hrahma) over the warrior-ruler (Jcsatra), which
forms the keystone of Dumont's model, derives from the material context
of the ecology of cattle - keeping tribes, among whom two groups of
specialists emerge, one claiming to mediate with gods through
specialization over rituals and thus increase the cattle -wealth of the
tribe and ensure success in tribal wars, the other of warriors who provide
protection and increase the wealth of the tribe through cattle raids. Both
groups, initially functional, claim and are able to acquire privileged
positions. Caste ideology evolves gradually in consonance with changing
material conditions and is not a mental invention unrelated to its material
roots. Nevertheless, I agree with Dumont that 'endogamy is a corollary
of hierarchy, rather than a primary principle',19 although for me caste
hierarchy is not simply a matter of superiority of the pure over the impure
but a form of exploitation which evolved in the process of enforcing
subjection of women and weaker social groups.
II
The beginnings of the twin processes may be seen in the Rgveda.
D.D.Kosambi in a perceptive article 'Urvasi and Pur?ravas'20 analysed a
number of Rgvedic hymns containing traces of a matriarchal culture,
which was suppressed and superimposed by the Aryan patriarchy. In his
view, the conflict and transition is reflected in the earliest stratum, the
Social Scientist
matriarchal came from a pre-Aryan

elements culture and early Rgvedic
?
^ society was from a combination
formed of the conquered pre-Aryans and
their Aryan conquerors. Kosambi identifies the pre-Aryans as survivors
^
E of the Harappa culture, not the elite trading or ruling classes but the
?5 'women with their cults... .either as wives or slaves, which would account
? for all the traces of their cults'. He adds that in any case Aryan means a
? particular manner of life and speech and not a race.21 The thesis of
j^
direct confrontation and conquest of the Harappans by the Rgvedic
^ Aryans is now generally discounted,22 but there are strong grounds to
~
believe23 that pre-vedic elements were accommodated in the later
sections of the Rgveda^ particularly in Book VIII, which is supposedly
^
io authored by the sage Kanva and his lineage.
_ Although it is plausible that certain external matriarchal components
> crept into Rgvedic narratives through absorption of pre-Aryan elements
the?Ap?l? S?kta (RV VIII. 91), which is a female puberty spell,24 is a
case in point ? not all traces of women's autonomy and subjectivity
need to be attributed to external sources. There is a general tendency to
force interpretations suited to patriarchy even when hymns suggest more
equitable gender relations25 owing to the presumption that patriarchy
was among the warring 'bronze-age pastoral invaders'.
deep-rooted
Generally speaking the Rgvedic poet does reflect a patriarchal attitude

and speaks contemptuously of his adversaries as having been deprived
of their manliness (e.g. RVX.48.12). Nevertheless, systematic displacement
of women from ?rauta rituals and appropriation of their role by male
priests in later vedic texts has been pointed out by a number of scholars,26
and it is difficult to explain away all traces of earlier practices as derived
from a non-Aryan/pre-vedic source. The apologists of vedic tradition
have argued that marginalization or exclusion of women is only one side
of the story, but in fact women were central to Rgvedic concerns, the
desire for progeny, material wealth, etc. Vedic yajfta ritual required the
presence of the wife of the sacrificer too, and much of the Hindu women's
deep identification with religion, her 'positive self- image... .stemmed from
the Rgvedic Age'.27 Such an essentialist, ahistorical approach, apart
from creating a homogenized category of 'Hindu Women' regardless of
their position in terms of caste, marital status, etc., does not take into
account the material environment which may have induced women to
internalize the patriarchal values of the brahmanical culture. Moreover,
one cannot ignore the fact that the stereotyping of women as sensual
creatures lacking in wisdom and self-control and acting as temptresses
to reluctant males, who are devoted to higher moral and ascetic goals,
has its beginnings in the hymns of the Rgveda, as illustrated in the
dialogue between Yama and Yami (RV X. 10) and Agastya and Lopamudra
?
(I. 179). ?
po
III ?"
However, the consolidation of patriarchy and a hierarchically

^~~
differentiated society comprising four varna divisions is an ongoing
process in later vedic texts, and the two developments were closely linked.
Sedentary agriculture and availability of servile labour of the defeated
and enslaved D?sa and ??dra tribes made it possible for the men and
women of the vedic elite lineages to withdraw from manual labour and be
contemptuous of those, who had to serve others and perform physical,
filthy tasks. They were categorized into a distinct ??dra varna in later
vedic times. The Purusa- s?kta hymn, which is undoubtedly a late
insertion in the Rgveda,2S ascribes the lowest position to the s?dra, but
it has an organic conception of society and traces his origin from the feet
of the same Cosmic Being whose mouth, arms and thighs produce
brahmana, raj any a and vaisya respectively. But several later vedic texts
attribute divine origin to only three upper varnas. The TaittirTya
Brahmana states that the ??dras sprang from Asuras or demons.29 The
same text says at another place that the ??dra sprang from 'untruth' or
'non-existence' (asat).30 Emphasis on the 'otherness' and evil character
of the ??dra may have been partly due to ethnic prejudice but also because
of his dependence on slavish manual work and marginal location. It is
generally held that the ??dra varna arose out of defeated D?sa, ??dra and
other aboriginal non-Aryan tribes reduced to various degrees of
servitude. The ??dra tribe31 is not mentioned in the Rgveda^but it speaks
of the capture and enslavement of a large number of D?sa men and
women32 with the result that the tribal name 'D?sa' became a signifier of
'slave'. Rgvedic chieftains made liberal gifts of male and female slaves
to priests and composers of the hymns and, as Kosambi has argued,33 the
assignment of slave-labour to the priestly and warrior lineages by the
tribal chieftain played a catalytic role in the growth of social
differentiation within Aryan tribes and emergence of a class structure in
the form of four original castes, i.e., the varnas. Kosambi explains that
the subjugation of the D?sa, ??dra and other tribes gave rise to a
generalized form of servitude in the form of the ??dra varna and not chattel
slavery, for the tribal influence was still strong and individual property
had not developed sufficiently yet. Internal differentiations among the
vedic tribes emerged with the priestly and warrior lineages uniting to
exploit both the 'Aryan peasant (vai?ya) and non-Aryan helot (??dra)'.
The cooperation and interdependence of the priestly and ruling
Social Scientist
groups and their exploitation and subjugation of the vai?ya and ??dra
g
producers is a well-known feature of the later vedic epoch. What deserves
?3
^ note is the fact that the varna ideology from its very inception plays a
g political and not just religious role in the hierarchical structuring of social
"S relations. The varna system, and later its expanded version the j?ti
i the class structure of early India and as such was a
system,34 regulated
powerful instrument functioning in the interest of the ruling classes. Its
g
? strong links with contemporary political powers and politics have been
rs maintained, as we shall see, throughout its long history.
~~
Itmay be argued that elaboration of vedic sacrifices into rituals of
o made it a of and the
great complexity preserve specialist lineages,
;g transmission of expertise to one's descendants and disciples finally gave
Zl rise to the br?hmana varna.35 The vedic ritual specialists played a crucial
> role in providing religious justification for the superior claims of the
vedic chieftain and his raj any a kinsmen over the vis commoners reducing
the need for the use of force, which undoubtedly underpinned such
claims.36 As tribal structure disintegrated and socio-economic disparities
grew the raj any as emerged as a separate privileged group and formed
alliances with similar groups of other tribes giving rise to the ksatriya
varna. Thus, the oligarchical lineages of the gana-r?jyas of the age of
the Buddha claimed to belong to the ksatriya varna although their lineages
were described as j?tis, to wit, a?kya j?ti, Licchavi j?ti, Jft?trika j?ti and so
on. But these were not j?tis in the modern sense of the term constituting
separate endogamous units. They practised endogamy within the ksatriya
varna marrying across their own j?ti boundaries.37 The term 'j?ti' was
used in a literal sense birth in a particular group, hence, we
to emphasize
have also references and ucea j?ti, birth in a low or high
to hlnaj?ti
social group. But the operation of the principle of heredity in establishing
the identities of the br?hmana and r?j any a categories is clearly indicated
in the Satapatha br?hmana3* and early upanisads such as the
Ch?ndogya. The latter text links it to the doctrine of transmigration and
karma. It is said that those who have pleased the gods with their pleasant
conduct enter a 'pleasant womb'. They are born either as a br?hmana, or
a k?atriya or a vaisya. But those whose conduct has been evil enter a
'stinking womb' such as that of a bitch, a pig or a C?ndala.39 Thus birth in
higher varnas was considered the fruit of meritorious acts performed in
the previous life.
IV
In an illuminating lecture40 published posthumously late Professor A. L.
Basham meticulously examined the upanisadic passages which revealed
8
the gradual evolution of the doctrine of rebirth from inchoate speculations

<?
into a well-developed ideology. He shows that this doctrine was first <.
adumbrated by the brahmana and the ksatriya intellectuals in an age, which
^
was characterized by great material progress but also disintegration of
^
tribal social life and the rise of varna divisions. The developments created
|j~~~
a sense of insecurity and pessimism among many thinkers, who opted
out of society and became ascetics and wanderers trying to discover
'the ultimate meaning of existence' To this, one might add that the
emergence of wide socio-economic disparities must have been an
important cause of disillusionment. It is said in the Ch?ndogya
Upanisad*1 that in this world one's greatness depends on cattle, horses,
elephants, gold, female slaves, fields and houses. The doctrine of karma,
?
rebirth and moksa the idea that "the individual soul keeps on passing
from one body to another to reap the fruits of the good or bad deeds, and
this is an unending cycle from which one can find release only through a
realization of the impersonal brahma or Truth ? developed in this
environment. The significant point is that it was an ideology which
germinated in the elite circles. Not being rooted in early vedic thought it
was taught initially as a secret knowledge discussed by a few, but later
became the basicprinciple of the ideological explanation of the cosmos
{samsara) preached by the br?hmanas, wandering ascetics and mystics
filtering down from them to lower orders. Basham is quite emphatic that
itwas not a borrowing of a pre-existing idea from non-Aryan, indigenous
peoples having animistic beliefs as was suggested by earlier scholars
but an invention of the upanisadic thinkers. It had its sceptics in the
form of C?rv?kas, Lok?yatikas and N?stikas but soon became the ideology
of the mainstream, and at the time of the rise of Jainism and Buddhism it
was accepted by everyone in the Gangetic valley.
However, the texts available to us are documents of the upper castes
and it is difficult to infer on their basis the extent to which this ideology
was internalized by the depressed groups. Max Weber wrote42 that the
inexorable logic of this doctrine reconciled the poor and the depressed
to their lot in the hope that through good conduct they could improve
their destiny in their next birth. But field studies conducted among the
'untouchable' castes by a number of sociologists43 show that although
the ideas of transmigration and karma ? that is, sins committed in
lives are the causes of misfortunes in the one ? is
previous present
accepted generally, the low status of their caste is not explained in this
fashion. Their origin myths ascribe their present degraded social ranking
to some historical accident or trickery of the high castes played on their
ancestors or genealogical founder. We shall be doing less than justice to
Social Scientist
the common sense of the exploited if we imagine that there would have
g
g been no resistance even in thought let alone practice.
^ Traces of resistance are not altogether lacking despite the nature of
2 our sources. These have been overlooked generally owing to the
"8 preconceived notions of the Orientalist and Nationalist historiographies.
i If colonial reconstructions the static, stagnant nature of
emphasized
Indian society immune to changes owing to a rigid caste structure rooted
g
^ in religious beliefs, the nationalists presented an idealized picture of
rs social harmony and contentment with castes engaged in their traditional
~
occupations without any social tensions or conflicts.
? R. S. Sharma's analysis of the Kali age crisis44 mentioned
However,
^ in theMah?bh?rata, R?m?yana and some early pur?nas clearly shows
ZL that the varna order and its ideology faced serious challenge in the
> early centuries of the Common Era from the lower orders, and although
generally the upsetting of the social order is attributed to the vai?yas
and the s?dras, some passages also speak of the antyas or untouchables
in this context. The earliest reference to a revolt by the menial labourers
pertains to the slaves of the a?kyas who had carried away 'married women,
unmarried girls and daughters-in-law of high families of their masters'.
Dr. Devaraj Chanana was of the view that the way Buddha reacted to this
incident suggests that it was not the only occurrence of its kind.45 One
may presume that since the slaves acted in a collective manner in
retaliation to their exploitation, they formed a collectivity, perhaps a
defeated and enslaved tribal population, but their integration as a
depressed caste within the varna framework cannot be taken for granted,
for the d?sa - kammakaras of the Pali sources constituted an economic
category. Although the varna categories had hardened into exclusive
hereditary statuses in the age of the Buddha, the j?ti structure within
varna framework was yet to develop. The category of untouchables grew
rather slowly,46 and the first untouchable groups seem to have been
food-gatherers and hunters living on the periphery of agrarian
settlements. In the listing of social groups in early Buddhist sources
they are mentioned separately and not as a part of the s?dra varna. The
J?taka tales47 depict C?ruj?las being engaged as musicians, night
watchmen, executioners, corpse-removers and sweepers removing
garbage the streets, but not in agricultural work. They were kept
from
away from Aryan homes as their sight and proximity was considered
polluting.48 However, in brahmanical perception they all formed part of
the s?dra varna; for P?nini, who is generally assigned to the fourth century
BCE, speaks of two groups of s?dras the 'excluded' and the 'unexcluded'
(s?dr?n?m aniravasit?n?m, Ast?dhy?yT, II.4.10), and, as Patafljali's
comment on it shows, the former included C?ndalas, Mrtapas, etc.

<?
<
pj
y
Contrary to ideological interpretations which attribute the origin of the {?

caste system to Indian psyche, a careful scrutiny of our sources shows
~~
|j
that condemnation of certain peoples as of despicable and impure birth
(j?ti) begins much earlier than the formulation of religious concepts
ascribing permanent impurity to certain occupations and practices. The
most polluting task according to Hindu notions of impurity is cleaning of
human excrement, latrines, etc., a task imposed upon the lowliest of the
untouchables known by different names in various regions as Bhangi,
Balmiki, Chuhra, Paki, Hadi, etc. Dumont argues that the Hindu belief in
the desecrating nature of organic activities makes the Hindu of good
caste temporarily impure and leads to attribution of massive and permanent
impurity to those categories of people who have specialization in impure
tasks, in practice or in theory.*9 In other words, even without being
actually engaged in 'polluting' occupations theoretically they are
associated with such tasks and hence regarded as permanently
impure.
However, we may point out that in brahmanical theory impurity does not
arise from specialization in impure tasks but from impure birth. Those
whose putative ancestors are deemed to have violated the varna norms
and contracted mixedpratiloma unions or marriages are condemned to
subsist on impure vocations, and their impurity is not removed even
when they are not engaged in occupations ascribed to them by
brahmanical tradition. Dumont completely ignores the instrumental nature
of the ideology of purity/impurity invented by the brahmanical ideologues
for justifying a system of class exploitation. This becomes evident from
the fact that from the early medieval down to late medieval times the
work of manual scavenging in the houses of the well-to-do peoples50
was done by domestic slaves who were of 'clean' castes51 and the same
person had to do other type of housework as well such as fetching
drinking water, grinding corn, cooking food, etc., without any prejudice,52
a situation unthinkable in modern-day conventional Hindu homes. This
shows that notions of pure / impure could be modified or elaborated
depending upon circumstances to suit the convenience of the exploiting
classes. The creation of a caste of manual scavengers is linked with the
growth of towns53 and closed dwellings without open spaces; and the
process seems to have been accelerated in the nineteenth century perhaps
aided to some extent by the Government of India Act V of 1843
abolishing slavery, despite the opposition of the landed aristocracy on
the ground that it was an ancient custom for slaves to do all manual
Social Scientist
? labour for respectable people.54 Moreover, army cantonments too required

?n such services and the municipalities and cantonments created official
?" posts of manual scavengers. It is rightly remarked55 that the British did
2 not invent the caste of manual scavengers, but they intervened to
'
?> institutionalize it ; and the technology of sanitation was structured to
deepen social prejudice in India'.
jL
? A similar inference can be drawn from an enquiry into the emergence
j^ of a caste of Chamars (cobblers). In the early Buddlist sources leather
~H4 work is regarded as h?na (low) sippa (craft or occupation) of low status
value;56 but leatherworker or carmak?ra does not figure in the list of
27 h?na j?tis enumerated in the Vinaya texts. The h?na j?tis or n?ca kulas
^o (low lineages) repeatedly mentioned57 are C?nd?la, Nes?da, Vena,
PO _
? Rathakara and Pukkusa, who with the possible of
exception
> Rathak?ra58 seem to have been aboriginal tribes living on the margins of
the Aryan settlements. Although tanning of hide is an ancient profession
and is known to the Rgveda59 and the later vedic texts, there is nothing
to show that leather or leatherworkers were considered polluting even in
later vedic times. We have references to leather bags filled with milk and
clarified butter (ghrta) for use in sacrificial ritual.60 According to
Vivekanand Jha61 the Carmak?ra, the Rajaka (washerman) and similar
craftsmen and manual workers appear as untouchables only in texts
datable between CE 600 and 1200. TheManusmrti refers to the mixed
caste of leather workers with three different names, Carm?vakartin,62
Dhigvana63 and K?r?vara,64 which is taken as indicative of the existence
of subcastes among leather workers;but the dating of these passages is
problematic. Professor Ram Sharan Sharma
suggests a time bracket of
CE 220-400 with later portions added in the fifth century and even later.65
Iwould like to point out that a Buddhist Prakrit inscription66 from
Amaravat? speaks of a cammak?ra (Carmak?ra) Vidhika, who describes
himself as the son of an up?jh?ya (up?dhy?ya, apparently a br?hmana
teacher) N?ga. He made the gift of a slab with a filled vase.
Paleographically the inscription is assigned to the early centuries of the
Common Era and it shows that in the Deccan leather work was still a
respectable profession. Pataftjali commenting upon P?nini's s?tra
mentioned above assigns C?ruj?las and Mrtapas the lowest position
placing the carpenters, washermen, blacksmiths and weavers above
them67 but does not speak of the Carmak?ra in this connection.
Apparently, defining of untouchability / impurity with reference to leather
work is a later development.
12
VI g
However, notwithstanding the evidence of Amaravati inscription, it is
^
not possible to accept the thesis of B. R. Ambedkar that the Chamar and ^__
other dalit communities of modern India had been originally Buddhists ~
and were degraded as untouchables by the brahmana law-givers, as they

~~
?|
continued to eat beef even after it was given up by the br?hmanas and
brahmanical communities.68 The brahmanical law-givers, poets and
playwrights continued to countenance the eating of animal flesh including
that of cow till the end of the first millennium CE,69 and in the sixth
century Var?hamihira especially recommended to the king to eat the flesh
of the bull, buffalo and other animals on ceremonial occasions.70 But
references to untouchable communities have been traced in sources
datable several centuries earlier, and crystallization of social groups
earning their living by leather work as specific castes of low status may
be seen in theManusmrti.11 In later smrtis the Carmak?ra is clearly an
untouchable.72 So leather work, beef-eating, or eating of carrion cannot
be regarded as having given rise to the phenomenon of untouchability,
although, later, condemnation of such practices was undoubtedly
used73 to relegate large sections of lower classes and aboriginal
communities to untouchable status. This is obvious in the case of present
day Chamars. This large caste spread over a vast area of northern India
seems to have been formed through assimilation of a number of tribes,
artisanal groups, local castes, etc. Only a small proportion of this caste
lives on leather work, the rest subsists on agricultural labour in rural
areas.74
However, following the theory of Dr. Ambedkar a number of Dalit

scholars visualize a glorious Buddhist past of the ex-untouchables. The
more cautious among them do not attribute the origin of untouchability
to the brahmanical ostracization of the Buddhists but argue that the
revival of Brahmanism under the Guptas and the persecution of the
Buddhists was largely responsible for the large increase in the number of
untouchable castes in Gupta and post-Gupta times.75 The thesis is
emphatically espoused by Gail Omvedt,76 who asserts that the 'defeat of
Buddhism in India' was the result of the alliances between the br?hmanas
and the kings and violent persecution of the Buddhists. She hypothesizes
that the C?ndalas were indigenous to Bengal, they had been speakers of
a proto-Mund? language, their name being strikingly similar to the
Mundari speaking Santhals. The C?ndalas had spread out from Bengal to
Central India and into the regions of the Gangetic plains where under
brahmanic hegemony they had been defeated and reduced to untouchable
status. In Bengal, communities like the Kaivartas and the C?ndalas had
Social Scientist
been supporters of Buddhism 'imbibing its equalitarian high tradition'.

g
g Later, many of them converted to Islam to avoid persecution and being
?- made untouchables. However, those who were unable to convert for
3 whatever reason were reduced to untouchable rank. Omvedt does not
"8 accept Basham's explanation that the decline of Buddhism may be
i attributed to the decadence of Buddhist monasteries, hold of the br?hmanas
on the performance of life- cycle rituals, reformed character of Brahmanism
g
?5 with the adoption of ahims? doctrine and its syncretistic attempts in
rsi making the Buddha the ninth incarnation of Visnu. She is also critical of
~~
Kosambi's view77 that the brahmanical individual priests were more suited
? to meet the needs of the self-contained villages of the agrarian economy
;j5 rather than the large Buddhist monasteries, which had become
Z uneconomic, dependent upon the patronage of higher classes and out of
> touch with the common people, and that these institutions were now
'mired in wealth and superstition'. Omvedt argues that Buddhist
monasteries were not any more unproductive or 'parasitical than the
Brahmanic priests living off innumerable gifts from believers'. In her view
Buddhism did not 'decline' but was 'defeated' and eliminated by the
br?hmanas in collusion with the kings.
This is not the place to go into the causes of the decline of Buddhism
? which is no
doubt an important question and needs a separate treatment
notwithstanding the meticulous work of R. C. Mitra78 ? but it needs to
be mentioned that the attitude of the early Buddhist and Jaina writers
towards the C?ndalas as well as all those who are dubbed as h?na j?tis
was no different from the brahmanical authors of the Dharmasutras. This
is amply shown by Richard Pick, Devaraj Chanana, Vivekanand Jha and
Uma Chakravarti in their studies cited above. The theory of karma, which
is taken for granted by Jainism and Buddhism could be used effectively
to rationalize discriminations on account
of birth in a family or caste,79
and as Professor Irfan Habib had argued80 in his General President's
Address to the Indian History Congress given in 1982, the principle of
ahims? could legitimize the hostility of the land-based peasantry towards
hunting tribes of the forests living on the borders of the agrarian
settlements and justify their ostracism providing the basis for
untouchablility. The hatred towards such communities is fully reflected
in the early Jaina and Buddhist sources.
The myth of a harmonious and conflict free golden age of India has
been exploded in a number of studies,81 and there is no doubt that there
is evidence of tensions, conflicts and a few instances of persecution of
Buddhists too in early medieval times.82 For instance, Bh?deva, a king of
Katy?ri dynasty ruling over the regions of Kumaun and Gadhwal in the
tenth century took pride in describing himself as a great enemy of

<?
Buddhist monks (paramabuddha>ramajuaripu) and a great patron of the ^.
br?hmanas.83 Several rulers of this region assumed the title of ^
which to D.C. Sircar should be ~
paramabrahmaptya^ according
translated as 'highly devoted to br?hmanas'.85 We have also sculptures
^~~
from Bihar and Orissa
depicting Buddhist deities trampling the
brahmanical gods under their feet,86 unmistakably a reflection of acute
hostility between Buddhism and Brahmanism.
However, it will be a mistake to think that sectarian conflicts,
denunciations were directed against Buddhists
and violence alone. The
Periya Pur?na speaks of the impalement of eight thousand Jainas at the
instance of the ?aiva saint N?n?samband?r; and a festival to commemorate
the 'gruesome event' is observed to this day in the Madura temple.87
Even if the story is the invention of a sectarian mind, it reflects extreme
hatred. The Tamil ?lv?r and N?yan?r saint-poets denounced Jainism and
Buddhism, but the attacks on the former were particularly vehement.
Attempts were made to appropriate the worship of Jina Rsabha too, while
imprecating the Jaina sramanas at the same time.88 According to a <aiva
hagiographical work god ?a?kara had incarnated himself as the
philosopher aankar?c?rya to destroy heretics, particularly the Jainas,
who were massacred and their books and temples destroyed.89 The
Basava Pur?na and the Pandit?r?dhya caritra too speak of the severe
persecution of the Jainas and destruction of their temples.90 The
inscriptions of the tenth-eleventh centuries testify to the persecution of
the Jainas in the South. It is believed that Tailapa II of the C?lukya
family, who overthrew the Rastrak?ta dynasty, persecuted the Jainas and
destroyed their shrines in the process.91 Later the Co?a armies overran
the C?lukya country causing extensive destruction of Jaina
temples.92 The underlying causes may have been political, but greed for
the wealth stored in religious institutions was no a
less motivating factor.
The R?jatarangininarrates the iconoclastic activities of several kings,
?a?karavarman,93 Kala?a95 and his son Harsa, who,
Ksemagupta,94
astonished at the amount of wealth stored in a deserted shrine, was
tempted to loot the rich temples of gods and appointed especially an
' 96
'officer for uprooting the gods literature of this period gives
Religious
expression to acrimonious theological
disputes among the Jainas,
Buddhists, ?aivas and Vaisnavas denouncing each other in strong terms.
If there are traces of conflicts,97 there are many examples of
syncretism98 and equal veneration of deities of different and opposing
religions,99 giving grist to the mill of those who wish to paint a picture of
tolerant, harmonious and homogeneous "Hindu" India; although it is
Social Scientist
possible to argue that in many cases these may have been attempts at
g
<3 reconciliation and resolution of conflicts. Nevertheless, the crucial
2> question is, faced with similar challenges why did Buddhism decline
^ whereas Jainism was able to survive and retain its social base?
"8 It seems to me that the answer lies not in the conspiracy theory of
i the br?hmana-king collusion100 but the way the two 'heterodox' religions
responded to the caste system. Both Jainism and Buddhism denounced
g
? the cult of vedic sacrifices and challenged the superior position of the
rsj br?hmanas in the varna system but did not reject the division of society
~
into varna categories.101 Enumeration of the fourfold division of society
-? is a regular feature of the early Buddhist texts.102 However, Buddhism
had a more liberal attitude towards the s?dras and untouchables and it
PO
allowed them admission into its monastic organization. Although the
> majority of the monks mentioned in the early Buddhist sources came
from the br?hmana and the ksatriya background, quite a few, such as
Up?li and Subhadda (barber), Canna (das?putra), Talaputa (nata),
Dhaniya (potter) and S?ti (fisherman) were born in n?cakulas.103 In the
Theragath? the monk Sun?ta speaks of his birth in a 'low family' of
sweepers (pukkus?).104 It has been argued that it was not possible for
the Buddha to bring about a radical change in society owing to the
limitations of the existing mode of production; but he tried to create an
egalitarian order of monks, which was open to all irrespective of rank or
varua, to even those who had been slaves.105 The J?taka stories tell us
of the Bodhisattvas born in the low families of potters and Can?alas.106
Buddhism as a religion retained its catholicity and criticism of the caste
system even in its later phases, despite the fact that the Buddhist kings
of early medieval period, like Dharmap?la and Vigrahap?la of the P?la
dynasty of Bengal, took credit in their inscriptions for reestablishing the
varn?srama dharma and stopping any deviation from it, apparently
because caste provided a useful mechanism for controlling and regulating
the economic and political resources. However, Buddhism seems to have
had a large following among the lower classes. Kum?rila Bhatta (8th
century) wrote that the teachings of the Buddha were followed by those
who belonged to the fourth varna, i.e., ??dras or by outcastes
(niravasitas)}01 Pur?nas denounced the Buddhists asp?sandins, who
were adept in argumentation and wilfully transgressed the duties arising
out of the distinctions of caste and order of life.108 The example of
R?hulabhadra, the disciple of Aryadeva, shows that even a s?dra monk
could rise to the position of the abbot of the N?land? monastery and
control immense amount of wealth.109 The Vajras?c? of Asvaghosa110
makes a trenchant criticism of the caste system and the selfishness of
16
the br?hmanas. In the Latakamelaka of ?ankhadhara a Buddhist monk

rejects the idea a
expressed by Digambara monk that anyone can become
polluted by the touch of somebody, who is of a 'dissimilar caste' (asadrsa
jat?sparsa) .lil
Contrary to the Buddhist attitude, the Jainas fully endorsed the caste
system. Ravisena in his Padma Pur?na, which is the Jaina version of
R?m?yana written in Sanskrit in 676 CE, credits Rsabhadeva for creating
the four varnas from different parts of his body and assigning them their
respective duties.112 The theme is further developed in the ?dipur?na113
of Jinasena (9th century) and ?d?svaracarita114 of Hemacandra (11th
century) with some variations. The Jaina texts condemn the intermixture
of varnas as strongly as do the brahmanical law-books.115 Somadeva
S?ri (10th century) in his NTtiv?ky?mrta recommends that everyone
should stick to one's hereditary occupation determined by his caste116
and restricts
religious initiation to the upper three varnas only.117 Jainism
also evolved rituals for its laity. Jaina domestic rituals were similar to
brahmanical ones officiated by the Jaina brahmana priests.118 Kosambi's
pithy comment on the survival of Jainism and decline of Buddhism is
typical of his deep understanding. 'Jainism survives in India to this day
for the same reasons that prevented its spread outside the country... it
soon came to terms with caste and ritual, as Buddhism did not'.119Caste
system ensured the structured dependence of the agricultural and
artisanal labour, which was to the great advantage of the land-owning
and ruling elite.120 Movements of protest could not be sustained for
long without basic changes inmaterial conditions.
VII
The upper class
contempt of manual
labour has been one of the basic
organizing principles of the varna-j?ti hierarchy. It is held121 that the
Buddha forbade the monks manual labour in order to free them from
worldly preoccupations. The prohibition could have been also under
the influence of the doctrine of ahims? (non-violence), as levelling the
soil, watering fields, gardens, etc. destroyed 'lives'.122 Hence, Buddhist
monasteries were gifted ?r?mikas (monastery-slaves) and to supervise
their work a monk was elected as ?r?mika-pessaka (supervisor of
?r?mikas).123 Manual work was avoided by the nuns too and domestic
work was declared an offence by np?cittiya rule.124 Such a negative
attitude distanced the monk-philosophers of Mah?y?na Buddhism from
physical work to such an extent that they developed a philosophy which
took into cognizance only the 'mental' nature of our experiences arguing
that everything is essentially no more than a 'mental construction
Social Scientist
0 (praj?aptimatraY125 Debiprasad Chattopadhyaya describes it as a

3 revolutionary philosophy passing into its opposite.126 Inmaterial terms
the effect of this ideological turn of Buddhism was to further devalue
?^
2 and depress those who
earned their living through physical labour; and
ultimately it gave rise to Sahajay?na form of Buddhism popular among
jO)
1 the lower castes.127 In Jainism too there was an excessive emphasis on
? non-killing of all forms of life leading to the prohibition of agricultural
and other types of manual activities from the very beginning. I have
J?^
~ <^J shown elsewhere128that in post-vedic times with the depression of the
peasantry, the well-to-do
members of the vis community became traders
and adopted Jainism in order to emphasize their disassociation from
2?
?d agriculture, with the result that in course of time only merchants and
? traders came to be known as vaisyas. The shift from the later vedic to the
> post-vedic connotation of the term is indicative of the decline in the
status of those communities which were engaged in the cultivation of
the soil and artisanal activities involving manual work. In Brahman ism,
as we have seen, the servitude of the s?dra was the foundation stone of
the varna system. Disdain towards him was extended to cover all the j?tis
subsisting ?n manual work and primary production in post-vedic times
and the attitude was further hardened with the adoption of the principles
of ahims?129 in neo-Brahmanism and its use in ascribing impurity to
menial occupations and communities. TheManusmrti not only ranks
hunters and gatherers as low-born outcastes but lays down that if a
br?hmana or ksatriya is unable to earn his livelihood by his own specific
vocation, he may earn his living by the vocation of a vai?ya by trading in
uncondemnable wealth-increasing articles but should not practise
agriculture. '(Some) declare that agriculture is something excellent (but)
that means of subsistence is blamed by the virtuous; (for) the wooden
(implement) with iron point injures the earth and (the beings) living in
the earth'.130
Theideology had serious implications for women. According to
Manu the householder (grhastha) has five 'slaughter houses' (pa?ca
s?n?)9 the fire-place (cul??), the grinding-stone, the broom, the pestle
and mortar and the water-pot. Using these he is bound with the fetters
of sin, so he should expiate by performing the five great sacrifices (pa?ca
mahayaj?as) daily.131 But these are the 'sinful' sites around which
revolves the life of a common housewife. It is not surprising that women
were regularly clubbed with the ??dras in brahmanical texts.132 The pain
and drudgery of a life around the 'pestle and mortar' is vividly expressed
is some songs of the Buddhist nuns included in the Ther?gatha.133
Manu's 'attitude to work presents a striking contrast to the Rgvedic poet,
18
who worshipfully invokes the mortar set to work in every house to give
<?
a clear loud sound like the drum134 of conquerors and compares the <.
mother-goddesses Usas (in plural) to women singing as they perform ^_
in the fields.135 In later medieval literature the ~
visa, apparently working
term visti means forced labour; but in the Rgvedic hymn it is a collective <i
?
activity with no trace of scorn.
VIII
Attribution of impurity to tasks involving manual labour gave religious
sanction to the exploitation of the working classes and helped in the
evolution of a brahmanical paradigm of social integration136 of diverse
communities into a highly stratified caste society with an ideological
tool with which to measure and justify the ranking of a particular social
segment. The role of br?hmanas in the spread of this ideology from its
home in the Gangetic valley to the various regions of the subcontinent is
duly stressed; and R.S. Sharma has laid particular emphasis on the
consquences of landgrant to br?hmanas in tribal areas in early medieval
times. However, br?hmanas alone were not the carriers of caste ideology
which was useful in the restructuring of tribes into an hierarchical society
legitimizing the claims of the tribal elites as superior status groups based
on heredity. Itmay be noted that in Sri Lanka caste system developed
under the influence of the Buddhist monks,137 who had carried with them
the theory o? karma and a notion of the functional hierarchy of social
groups based on birth, but as there were no br?hmanas or a caste of
priests, castes were not defined in terms of pure/ impure communities.
The Sri Lankan example shows that the opposition of the brahmana and
the untouchable, i.e., the 'pure' and the 'impure', is not the founding
principle of the caste system as assumed by Dumont. Rather, it is a
superimposition on a structure of rigid class differentiations; and castes
can exist without the help of the ideology of pollution.
Expansion of the caste society in various regions of India took place
through multiple processes;138 and a few studies139 have underlined the
role of tribal chieftains, who emulated the ksatriya model in order to
legitimize their political power and control over community- resources
and took initiative for the diffusion and broad acceptance of brahmanical
norms in Orissa in early medieval times. In some regions the dominant
ideology could have been disseminated through Jaina and orthodox ?aiva
monastic orders. It has been argued140 that in the backward tribal territory
of Rayalseema in south-western Andhra Pradesh the transition from tribe
to state took place in the sixth-seventh centuries CE, when this area was
exposed to outside influence owing to its strategic importance in the
Social Scientist
oo power struggle between the C?lukyas of B?d?mi and the Pallavas of Kanc?.
o The region did not attract any br?hmana settlements or agrah?ras but
?. was penetrated by the Jaina monks and the Saivas of the K?l?mukha
2 sect, who used the local vernacular to spread their message. They were
"8 patronized by the emerging local elite and chieftains, who also showed
UT preference to the local language in their inscriptions in order to assert
(d their separate ethnic identity and local roots vis-?-vis the C?lukyas and
? the Pallavas. These developments contributed not only to the growth of
^7 Telugu language and literature but also integrated this region with the
~"
pan-Indian culture through ideologies which were opposed to the caste
o system. While we agree with the broad generalizations of this argument,
^ we may point out that neither the K?l?mukhas- who are wrongly confused
Z with the K?p?likas141 nor the Jainas in these centuries were opposed to
> the caste system. In fact, the K?l?mukhas were thoroughly imbued with
the dominant ideology and many of them became the preceptors of kings
(r?jaguru) or family priests of the village headmen (g?vundas). They
actively promoted construction of temples, which apart from being places
of worship also imparted brahmanical education to people and received
grants for the purpose.142 It is not surprising that a Kalac?ri inscription
of the twelfth century praises Vimala?iva, the ?aiva r?jaguru of king
Jayasimha, as one whose counsel had made even more distant people
pay taxes.143 Nevertheless, the basic point that the initiative of the
br?hmana caste is not the essential condition for the spread of the
brahmanical ideology of caste is substantiated by the example of the
numerically large caste of Kallar located in the southern districts of
Tamilnadu. Hutton144 describes it as a cultivating and predatory Tamil
caste notable for their efficient agriculture, expert thieving, cattle lifting,
etc. Dumont did intensive field work among the Pramalai Kallar, a subcaste
of the Kallars, and published a monograph145 on them from a social
anthropological point of view. He writes that the Kallars are relatively
unaffected by brahmanic ideas and customs. They bury their dead, and
although they have warrior pretensions, they willingly allow themselves
to be classed among the s?dra.146 However, in the local hierarchy they
occupy the middle rung of the caste ladder. They seem to have migrated
from the Andhra country and founded the kingdom of Pudukkotai between
Tanjavur and Madurai in the last quarter of the seventeenth century.
Nicholas B.Dirks, who has made a detailed study of the kingdom of
Pudukkottai,147 compares them with the Rajputs of northern India and
cogently argues that the assumption of power led to a restructuring of
the Kallar caste which got divided into a number of subcastes graded
hierarchically not on the basis of their 'purity /impurity but with reference
20
to their relative proximity to the royal power and control over land'.148
<?
Pramalai was the royal subcaste.
Kallar It is endogamous and is also 5L
known as Tevar (from Sanskrit deva), originally a political designation
^
but now a general title. The domination of the Kallars in the areas ~
occupied by them manifested itself in the imposition of extremely |j~~

humiliating and discriminatory prohibitions upon the 'exterior' castes,
who worked on their fields as labourers living in 'serf-like' conditions.
Dumont describes these impositions upon untouchables as 'customary'
but as recorded by Hutton,149 the prohibitions had nothing to do with
the notions of pure/impure and were merely expressions of arrogant power.
Dirks speaks of the increasing trend towards patriarchy and adoption of
some of brahmanic practices by the royal family of Kallar, such as
seclusion of women, disapproval of widow-marriage, etc. But the field
work of Dumont shows150 that widow marriage prevails among the Kallar
and divorce is extremely easy. It could be based on nothing more than
the desire of one of the spouses, male or female. It follows that the
brahmanical influence was perhaps confined to the royal family and
sankritizing efforts had very limited impact on the caste. Dirks'
assertion151 that caste as a social system, whether in the political milieu
of pre-British period or in its 'increasingly brahmanical forms under
colonial rule' was a most pervasive form of oppression directed against
women is valid in general; but as far as the Kallar women are concerned
the impact of caste formation is yet to be worked out.
Nevertheless, the twin pillars that sustain the caste system are firstly,
subordination of women and secondly, its capacity to reinvent itself in
changing social formations in the service of the powerful and the
dominant. It has been shown that in spite of its apparent rigidity the
system was able to enroll new members and create new caste categories
at various levels. There was scope for political or economic mobility
through processes of fission and fusion as examplified by the formation
of the Rajput and Kayastha categories. In these processes control over
woman's sexuality was critical, endogamy as well as hypergamy was used
to create a distinct caste identity and raise its status. In the pre-British
period fission rather than fusion was adopted for upward mobility; but
the trend inmodern India has been towards fusion to form numerically
large caste identities by integrating subcastes and groups having parallel
positions.152 It is said that the Census Reports have played a crucial role
in this development by generating greater caste consciousness and an
awareness for the bargaining possibilities of larger sodalities. Whatever
the case may be, the change of trend underlies the real nature of this
form of social stratification, which is its capacity to reconstruct itself as
Social Scientist
oo an instrument of power for the new elite in a different political formation.

o
o It is being argued153 that caste should not be disavowed or sought to be
rs
erased, as a 'site of identity and power'; for it has possibilities for political
cd
2 mobilization that would transform the prevailing relations of state and
_?
<L>
U_ society in the favour of the oppressed; and it provides a^potent ideological
I tool for the assertion of Dalit identity. Such post-modernist arguments
Z5
are in fact arguments in favour of the status quo pleading for the
C
fd replacement of one set of power-elite with another without bringing about
rsl any revolutionary change in the politico-economic patterns of domination
and exploitation. These do not take into consideration the fact that caste
is no longer tied with occupation. Post-independence changes in the
political and economic set up and industrialization have had their impact
m
on the internal homogeneity of castes which have thrown up their own
> elite, who may use the ideology of caste for narrow political interests
without any radical transformation
effecting in the condition of their
caste in general. It is not without significance that the caste battles are
fought these days on issues of reservation in jobs and institutions of
higher education, but there is no strong movement around the questions
of land-reform and primary education which would transform the lives of
the Dalit masses. Dr. Ambedkar seems to have foreseen this possibility
when he criticized the view that abolition of subcastes should be the
first step towards caste-reform. He categorically wrote, 'abolition of
sub-castes will only help to strengthen the castes and make them more
powerful and therefore more mischievous'.154 He argued for the
annihilation of caste155 for which he thought the real remedy lies in inter
caste marriages. The fact that he taught the Dalit communities self- respect
and organized them for collective political action does not mean that he
wanted to nurture caste identities. For Ambedkar social and cultural
emancipation of women and men was as important as political and
economic It is unfortunate
empowerment. that in the unabashed pursuit
of political power today the holistic vision of Ambedkar is completely
forgotten; and the pernicious strength of caste and patriarchal mentalities
in our society is not seriously combated.156
ouvita Jaiswal is former Professor of History, Jawaharlal Nehru University, Delhi.
Notes
I am thankful to Professor Kunal Chakrabarti, Dr. Rakesh Batabyal and Dr.Ranjan
Anand for helping me in various ways in the preparation of this lecture.
1. 'Caste, in the socio-economic framework of Early India', Presidential Address,

Section I, Proceedings of the Indian History Congress, 38th session,
22 23 - 48.
Bhubaneswar, 1977, pp.
2. On the 19th and 20th June 2007 the TV Channel Aaj Tak showed a Dalit named LO
C
Ramlal of Tonk in Rajasthan badly beaten up .He was punished by the upper <
caste villagers for the crime of drinking water from a nearby borewell and
received multiple fractures. This is not an isolated case. Incidents of this ?2

iSt'
nature are still frequently reported from various parts of the country.
3. Ronald Inden, 'Orientalist constructions of India', Modern Asian Studies, Vol.
20 no. 3 (1986), pp.401-46. As Aijaz Ahmad remarks 'Colonialism is now held
responsible not only for its own cruelties but, conveniently enough, for ours
too' In Theory, Oxford University Press, (hereafter OUP) Delhi, 1994, p. 196
7. Inden has, however, in a later work, Imagining India, (Basil Blackwell,
1990, p. 82) has modified his position a little by linking the formation of
'modern form' of caste to 'the collapse of Hindu Kingship' in the thirteenth or
fourteenth century. For a devastating critique of Inden's book, Aijaz Ahmad,

'Between Orientalism and Historicism: Anthropological knowledge of India'.
Studies in History. Vol. 7 no. 1 (Jan. June 1991), pp. 135-63.
'
4. In her essay entitled The Changing Caste System in India' Pauline Kolanda
writes, 'the persistent feature of Indian society, its basic building block, is the
endogamous group', which has now become a 'segmentary one rather than an
organic one' Pauline Kolanda, Caste, Cult and Hierarchy: Essays on the Culture
of India. Folklore Institute, Meerut, 1981, p. 83. Dipankar Gupta speaks of
the castes as 'discrete categories', which 'value the principle of endogamy

very highly' without explaining that these discrete categories could not exist
without practising endogamy. Interrogating Caste: Understanding Hierarchy

and Difference in Indian Society. Penguin Book?, India, 2000. p.70. For a
detailed discussion. Suvira Jaiswal, Caste: Origin, Function and Dimensions of
change, (Hereafter, Caste) Manohar, Delhi, 1998, Introduction.
5. Nicholas B. Dirks, Castes of Mind; Colonialism and the making of Modern

India, Permanent Black, Delhi, 2002, paperback , 2006, pp.7-8.
6. H.H. Risley ,Census of India, Vol. I, p.l, 1901 quoted in Dirks ibid, p.222.
7. For references, S. Jaiswal, Caste, p.40-1.
8. Dipankar Gupta, 'Continuous Hierarchies and Discrete Castes', Economic and
Political Weekly, 19, no. 46 (17 November 1984), reprinted in Dipankar Gupta
(ed), Social Stratification, OUP, Delhi, 1991, p. 137, also see idem,
Interrogating Caste: Understanding Hierarchy and Difference, p. 70; 84.
9. Ibid., p.70. However, earlier Gupta ascribed the origin of the varnas to the
attempt of the Aryans to maintain their social distance from the indigenous
community, 'From Varna, to J?ti : The Indian Caste System from the Asiatic
to the Feudal Mode of Production', Journal of Contemporary Asia, Vol. X
23
Social Scientist
00 (1980), pp. 249-71 reprinted in K.L. Sharma (ed.), Social Inequality in India,
8 Profiles of Caste, Class, power and Social Mobility. Rawat Publications, Delhi,
1995, pp. 159 -91.
10. Ibid.
11. I have shown that the two terms are used interchangeably in early Indian texts.
Caste, pp. 42-3.

k
12. P.V. Kane quotes S?ta Samhit?, Siva M?h?tmya Khan?la, 12.51.52, which
states that 'a man belongs to a caste by birth and no action of his can alter that
fact, that several castes are like the species of animals and that caste attaches
O to the and not to the soul' vol.
Z body History of Dharmas?stra, II, pt ii,
(Bhandarkar Oriental Research Institute, Poona, 1941), p. 52. The Bhagavad

m
g?t? (XVIII. 41) clearly says that owing to their natural (inborn) qualities
> {guitas) the four varnas have been assigned different functions. For a detailed
discussion, S. Jaiswal, The Making of a Hegemonic Tradition: The cult of Rama

th
D?sarathJ, S.C. Misra Memorial Lecture, Indian History Congress, 67 session,
March 2007, p. 26-7 note 54. For the implication of the theory of gun as for
the hereditary nature of varna organization, idem, 'Caste, Ideology and
Context', Indologica Taurinensia, Vol. XXIII- XXIV (1997-98) p. 611.
13. See the origin myths current among Chamars, Dacca Chandals, Kayasthas,
Vaniyans, Bhangis, etc. cited by Dipankar Gupta to show that there were many
and not one caste ideologies. Interrogating Caste, pp. 73-7.
14. D.D. Kosambi, Introduction to the Study of Indian History, Popular Prakashan,
Bombay, 1956, p. 25; Morton Klass, Caste: The Emergence of the South Asian
Social System, Institute for the study of Human Issues, Philadelphia, 1980, p.
175; Irfan Habib, Essays in Indian History: Towards a Marxist Perception,
Tulika, New Delhi, 1995, p.165. S.M. Michael, (ed.), Dalits inModern India:
Vision and Values, Sage Publications, New Delhi (1999), 2nd ed. 2007,
Introduction, p. 17.
15. S.Jaiswal, Caste, p. 9f ; 157-8.
16. Louis Dumont, Homo Hierarchicus: The Caste, System and its Implications,
first published, 1966, complete revised English edition, OUP, Delhi, 1988.
17. S. Jaiswal, Caste, p. 34-8, 103 Note 54; 118 note 207; idem. 'Caste: Ideology
and Context', Indologica Taurinensia, vol. XXIII-XXIV (1997-98),pp. 611
5. Also see Gerald D. Berreman's excellent critique, 'The Brahmanical view of
caste', Contributions to Indian Sociology (n.s.) No. V (1970), pp. 16-25.
18. S. Jaiswal, 'varna Ideology and Social Change'. Social Scientist, vol. 19, nos.
3-4, March-April 1991, p. 41f.
24
19. Louis Dumont, op.cit., p. 113. LO

C
<
20. D.D. Kosambi, Myth and Reality: Studies in the Formation of Indian Culture,
Popular Prakashan, Bombay, 1962, pp. 42-81

Bf.
21. Ibid., p.68; 76.
pu
22. Romila Thapar, Ancient Indian Social History, Orient Longman, New Delhi,
1978, p. 18; R.S. Sharma, Material Culture and Social Formations in Ancient
India, Macmillan, Delhi, Ltd. 1983, p. 157; D.N. Jha, Ancient India inHistorical
outline, Manohar, Delhi, 1998. For the latest summing up of the archaeological
and linguistic arguments respectively, Jonathan Mark Kenoyer. 'Culture and
Societies of the Indus Tradition', and Madhav M. Deshpande, 'Aryan Origins:

Brief History of Linguistic Arguments' in Romila Thapar (ed.), India: Historical
Beginnings and the concept of the Aryan, National Book Trust, New Delhi,
2006, pp.41-97 and 98-156.
23. S. Jaiswal, Caste, pp. 136-7, 195; idem, 'Inventing a Culture of Patriarchy: An
Aspect Brahmanism', Mamidipudi Venkatarangaiya Memorial Lecture-14.
Proceedings of the Andhra Pradesh History Congress, XXVI Session, Anaparti,

2002, Appendix I, pp.i-xv.
24. Suvira Jaiswal, 'Reconstructing History from the Rgveda: A Paradigm Shift?'
Social Science Probings, vol. 18, no. 2 (Dec. 2006), pp. 15-6.
25. For example, see Hans-Peter Schmidt on RV .X. 27.12 which speaks of a
beautiful woman choosing her spouse among the suitors of her own free will
(svayam sa mitram vanute jane cit). According to Schmidt the hymn shows
the prevalence of bride-price and the girl goes to the highest bidder. Hans
Peter Schmidt, Some Women's Rites qnd Rights in the Veda, Bhandarkar,
Oriental Research Institute, Poona, 1987, pp.76-7. He translates panyas as
bride-price, but Monier-Williams equates it with paniyas meaning very

wonderful' {Sanskrit-English Dictionary, s.v. panyas and S?yana glosses it as
'one who is eulogised with praise' (panyas? stotrena). For a detailed discussion,
S. Jaiswal, 'Process of gendering in the Brahmanical Tradition, Prajn? Bh?rat?,
vol. XI, in Honour of Prof. Ram Sharan Sharma, K.P. Jayaswal Research
Institute, Patna, 2005, pp. 21-5.
26. Sadashiv Ambadas Dange, Sexual Symbolism from Vedic Ritual, Ajanta
Publications, Delhi, 1979, pp. 73-4; Fredrick M. Smith, 'Indra's curse, Varuna's
Noose, and the Suppression of Women in the Vedic ?rauta Ritual' in Julia
Leslie (ed.), Roles and Rituals for Hindu Women, Motila! Banarsidass, Delhi,
1992, pp. 17-45, Kumkum Roy, ?.mergence of Monarchy in North India: Eighth
to Fourth Centuries B.C. as reflected in the Brahmanical Tradition, OUP,
Delhi, 1994, , p.67.
25
Social Scientist
CO 27. Katherine K. Young, 'Hinduism' in Arvind Sharma (ed.), Women in World

o
o State University of New York
rN Religions, Press, Albany, 1987, p.64. M.N. Srinivas
speaks of the considerable empowerment of high caste women through their

meticulous observance of rules, of
2 purity-pollution performance periodical
-Q
(D rituals, etc., which are considered necessary for the material and spiritual
welfare of the household. The Changing Position of Indian Women, OUP,
=3 Delhi, 1976, pp. 17-8. Nevertheless, high-caste women's assertion or celebration

C of self-worth the performance can
through of Hindu rituals hardly be linked to
Rgvedic vision. Women were debarred from listening to the vedas. The
Brhann?radiya Pur?na says, 'A man who reads the vedas in the proximity of
women and sudras goes to hells successively during thousands of crores of
vO kalpas', XIV. 144, quoted in R.C. Hazra, Studies in the Upapur?nas, Vol. I,
Sanskrit College, Calcutta, 1958, p.325. The prohibition is restated in the
O
> R?macaritam?nasa of Tulas?d?sa (1.109.1), published by Hanuman Prasad
Poddar, Gita Press, Gorakhpur, Samvat 2050, thick print, p. 107).
28. For the remodelling of an ancient myth regarding the creation of the cosmos
through the original sacrifice of the primordial being to justify the four-fold
social differentiation, S. Jaiswal, Caste, pp. 135-6
29. Taittirlya brahmana I. 2.6.7 quoted by U.N. Ghoshal, A History of Indian

Political Ideas, OUP, Madras, reprint, 1966, p. 31.
30. Taittirlya brahmana III. 2.3.9 quoted in Jogiraj Basu, India of the Age of the
br?hmanas Sanskrit Pustak Bhandar, Calcutta, 1969, p. 12. Some other texts
speak of the birth of the s?dra from Evil, K?thaka Samhit?, XXXI.2;
Maitr?yanJ Samhit?, IV. 1.3 quoted in U.N.Ghoshal, op.cit.
31. For the name of the ??dra tribe becoming a generic term for the fourth varna,
R.S. Sharma, S?dras in Ancient India, Motilal Banarasidas, Delhi, second revised
edition, 1980, p.34f.
32. Rgveda, I. 126.3 speaks often chariots carrying vadh?s given to Kaksiv?n as
part of his daksin?. These were apparently women captured from defeated
alien tribes, presumably D?sas. Rgveda, VIII. 19.36 mentions a gift of fifty
vadh?s given to the composer of the hymn by king Trasadasyu. Griffith
translates the term as female slaves. For women of the D?sa tribes participating
in wars against Aryan enemies, S. Jaiswal,' Process of Gendering in the
Brahmanical Tradition', pp. 25-7. Also see Rgveda, VIII, V?lakhilya 8.3 ;X.
62. 10. In RV 1.92.3 the poet beseeches the Dawn goddess to grant him ample
wealth in the form of brave sons (suv?rah), horses and troops of slaves {d?sa
-
Pravargd).
33. Introduction to the Study of Indian History, p.93; 104.
26
34. S. Jaiswal, Caste, p. 167; 196-7. LO

C
<
35. For the rise of br?hmana varna, ibid., pp. 149-62.
36. R.S.Sharma, Aspects of Political Ideas and Institutions in Ancient India, third
revised edition, Motilal Banarsidas, Delhi, 1991, pp. 178-80. Rgveda, VIL 6.5 ?
?U
describes Agni as using force (balainirudhya) to make the vis give tribute to
Nahusa.
37. According to a Tibetan tradition the a?kyas and the Licchavis were branches
of the same tribe. The origin myths of both the groups attribute brother-sister
marriage to the founders; and the origin of the Koliyas of R?mag?ma too is
traced from a ?akya girl in such texts as Sumangalavil?sinT and Mah?vastu.
Quoted by S.N. Misra, Ancient Indian Republics. The upper India Publishing
House, Lucknow, 1976, p.46. ?uddhodhana, the father of the Buddha, is said to
have married two Koliyan princesses, M?y? and Mah?paj?pati Gotam?. Kosambi
discounts this tradition on the ground that the ?akyas were too proud to marry
outside their tribe. He cites the story of Pasenadi, the king of Kosala, who was
tricked into marrying V?sabhakhattiy?, the daughter of Mah?n?ma a?kya by a
slave girl named N?gamuntf?. However, this only shows that the ?akyans did
not want to displease Pasenadi, who had asked for the hand of a ?akya girl, but
at the same time they did not wish to give a girl of pure ?akya lineage in
marriage to him, who belonged to the lowly M?tanga-kula. The tribe of
M?tangas was later equated with Car?alas. In the DTgha Nik?ya, the Buddha
tells br?hmana Ambattha that the khattiyas (ksatriyas) are more rigid and refuse
to accept in their own group a man who is not pure by birth for seven generations
on the side of his both parents, but the br?hmanas accept sons born of partial
non- br?hmana origin on either side and allow them to participate in yaj?a,
sr?ddha, sth?llp?ka, etc. DTgha Nik?ya, vol. I, pp. 92-7 quoted in N. Wagle,
Society at the time of the Buddha, Popular Prakashan, Bombay, 1966, pp.
101-3.
However, the Licchavis and J?atrikas, both members of the Vajjian confederacy,
are known to have had marriage relations. Licchavi chief Cetaka's sister Trisal?
was married to Siddh?rtha of the J?atrikas, the father of the Jaina Tirtha?kara
Mah?vira. Cetaka's daughter Cellana was married to Bimbis?ra the king of
Magadha and Aj?tasatru was her son. Cetaka had several daughters whom he
gave in marriage to ksatriya rulers of the time. No doubt kings took wives from
other varnas too but the mother of the heir-apparent or claimant to the
throne had to be of ksatriya lineage as is shown by the story of Vi4u?labha, son
of V?sabha-khattiy?. Also see S. Jaiswal, Caste, p. 15; 27 note 83. Wagle speaks
of a?kyas, Licchavis, etc, as extended kin-groups which slowly ossified into
castes by the time of the Manusmrti. That ksatriyas too had become a caste
27
Social Scientist
CO like the br?hmanas is shown the fact that an from Andhra

o by inscription
o Pradesh to the centuries of the Common Era
CNl paleographically assigned early
speaks of ksatriya marchants. K. Gopalachari, Early History of the Andhra
2 Country, Madras, 1941, p.91. The Manusmrti, X. 43-4 speaks of ksatriya j?tis
-O
<L> in plural indicating the existence of a number ksatriya castes within the broad
LU
I varna category.
ZJ 38. II. Xi. 5. 7. 4. 4.

c ?atapatha brahmana, 1.4.4; 1; XII. 6; 4.4.7; XIII 19.1.-2,
edited by Ganga Prasad Upadhyaya, The Research Institute of Ancient Scientific
Studies, New Delhi; 1970, Translator , J. Eggeling, Sacred Books of East Series,
Vols. 12, 26, 41, 43, 44, Reprint, Motilal Banarsidass, Delhi, 1988.
39. Ch?ndogya Upanisad, V. 10.7, The Thirteen Principal Upanisads, Translated
by R. E. Hume, 2nd edn, OUP, Madras, 7th impression, 1968, p.233.
> 40. A. L. Basham. The Origin and Development and

of Classical Hinduism, edited
annotated by Kenneth Zysk, OUP, Delhi, 1990, Chapter Three.
41. Ch?ndogya Upanisad, VII. 24.2.
42. Max Waber, The Religion of India: The Sociology of Hinduism and Buddhism,
translated by Gerth and Martindale, Free Press, New York, 1958, p. 122.
43. Pauline Kolenda, "Religious anxiety and Hindu fate", in Religion in South
Asia, edited by E.B.Harper, pp.71-81, University of Washington Press, Seattle,
reprinted in P. Kolenda, Caste, Cult and Hierarchy, pp. 169-83; Joan P.
Mencher, 'The caste system Upside Down, or the Not -So- Mysterious East'.
Current Anthropology, Vol.15 no.4, December 1974.Eleanor Zelliot, From
Untouchable to Dalit: Essays on the Ambedkar Movement, Monohar, Delhi ,
1996, p. 74 note 5; Robert Deliege, 'The Myths of Origin of the Indian
Vol No 3 (September 533 -

Untouchables', Man, New series, 28, 1993), pp
549.
Following Jan Vansina, {Oral Tradition : A Study in Historical Methodology,

translated by H. M. Wright, Routledge and Kegan Paul, London 1965), one
may assume that these oral myths had been a part of the consciousness of the
oppressed castes for a long time. But these have to be distinguished from caste
histories written in response to the colonial documentation project. See V.
Geetha, 'Rewriting History in the Brahmin's Shadow : Caste and the Modern
Historical Imagination', Journal of Arts and Ideas, December 1993, nos. 25
26, pp. 127-37; Badri Narayan, Women Heroes and Dalit Assertion in North
India: Culture, Identity and Politics, Sage Publishers, New Delhi, 2006, p.l70f.
However, the Mahar saint Chokhamela (13th -14th centuries) accepted his
birth in the low caste as a consequence of his karma, Zelliot, op.cit., p.7.
28
44. R. S. Sharma, Early Medieval Indian Society: A study in Feudal ization, Orient LO
C
Longman, Hyderabad, 2001, pp.50-1; 53. <
3
45. Vinaya Pitaka, Vol. IV p. 181 quoted in Devaraj Chanana, Slavery in Ancient
India, Peoples publishing House, Delhi, 1960, reprint 1990, p. 62. Uma
Chakravarti remarks that this is one of the first written records which shows
that women were the obvious targets in case of antagonism between two social
groups. Social Dimensions of Early Buddhism, OUP, Delhi, 1981, p. 27. fn.
145.
46. Vivekanand Jha, 'Stages in the History of Untouchables' Indian Historical
Review (hereafter IHR), Vol II, no.i (July 1975), pp.14-31; idem, 'Caste,
Untouchablilty and Social Justice: Early North Indian Perspective', Social
Scientist, vol.25 nos 11-12 (Nov- Dec. 1997). p.24.
47. Idem, IHR, II, i, p.22.
48. Richard Fick, The Social Organization in North East India, translated by S. K.
Maitra, University of Calcutta, Calcutta, 1920 p.43f: For the evidence of
early Dharmasutras, Vivekanand Jha, 'Car?ala and the Origin of Untouchability',

nos 1986 - Jan
IHR, XIII, 1-2, (July \%i), pp.4-7.
49. L Dumont, op.cit, p.47. Italics ours. Similarly, the view that primitive notions
about accepting food from non- kin causing pollution are at the base of
untouchability cannot be sustained. Restrictions on interdining and acceptance

of cooked food from various categories of people have evolved gradually,
much later than the emergence of untouchability. S. Jaiswal, Caste, pp.86-7;
125-6 note 285.
50. Ordinary people of 'clean' (?uddha) castes must have used open spaces.
51. Hiroyuki Kotani, cites a document of 18th century (quoted from G.S. Sardesai
(ed.) Selections from the Peshva Daftar, Vol 43-92) that a female servant
employed in the house of a br?hmana family turned out to be of the Chambar

caste. Hence all those who had come in contact with her had to undergo
various degrees of purification. Another instance cited by him shows that a
female slave belonging to a family of Prabhu caste committed adultery with an
antyaja (ati-sudra), which fact made all the members of the Prabhu family
impure. H. Kotani, "Ati-s?dra castes in the Medieval Deccan.' In H. Kotani
(ed), Caste System, Untouchability and the Depressed, Monohar, Delhi ,1997,
pp 56 -7.
52. For fair looking Rajaputra (Rajput) girls sold into slavery and obliged to do all
kinds of pure and 'impure' work, Pushpa Prasad, 'Female slavery in Thirteenth
century Gujarat; IHR, XV nos. 1-2 (July 1988 - Jan 1989) pp.269-75, S. Jaiswal,
29
Social Scientist
oo Caste, pp.84-5. The N?radasmrti, written perhaps in the 4th century CE

o
o clearly specifies that while a hired servant {karmakara) is supposed to do pure
work only, slaves are to do all kinds of impure work. P.V. Kane, History of
Dharmas?stra, II, pt.i.p.l84f. Also see Prabhati Mukherjee, Beyond the Four
2
_Q varnas: The Untouchables in India, Indian Institute of Advanced
<L> Study, Shimla,
LL
I Revised ed. 2002, p.75.
Z3 53. This does not, however, mean that the institution of untouchability can be
c
fd to Harappa as was von Furer-Haimendorf.
traced culture done by Christoph
See S. Jaiswal, Caste, pp.78-9.
o 54. T.R. Sareen, 'Slavery in India Under British Rule, 1772-1843', IHR, XV nos.l
Z
2, (July 1988 & January 1989), pp.257-68.
55. Gita Ramaswamy, India Stinking: Manual Scavengers in Andhra Pradesh and
> their work, Navayana Publishing, Chennai, 2005, p.6.
56. Uma Chakravarti, op.cit., p. 104.
57. Narendra Wagle, Society at the Time of the Buddha, Popular Prakashan, Bombay,
1966, pp.119-20; 122-23.
58. On Rathak?ra, Vivekanand Jha, 'Status of the Rathakara in Early Indian
History', Journal of Indian History, vol.52, pt.i (Trivendrum, April 1974),

39-47. In the Arthas?stra of Kautilya (III. 7.35) He is described as a
pp.
vaisya. The Venas seem to have been a non-Aryan tribe of 'bamboo-workers'
or basket-makers.
59. VIII, 5.38 has carmamna. S?yana explains that it refers to armour
Rgveda,
made of leather.
60. S. Jaiswal, Caste, p.82.
61. IHR, II, i, p.19.
62. Manusmrti, IV. 218.
63. Ibid, X.15.49.
64. Ibid., X.36. Also repeated inMah?bh?rata (er. edn.), XIII, 48.26.
65. ??dras in Ancient India, 2nd revised edition, Motilal Banarsidass, Delhi, 1980,
p.330-333.
66. Epigraphia Indica, vol.X, L?der's List No. 1273.
67. Pata?jali on P?nini II. 4.10, Mah?bh?sya of Pata?jali, edited by F. Kielhorn
vol.1 (Mumbai, 1892), p.475.
68. See Vivekanand Jha, IHR, II, i, pp.21-31.
30
69. D.N. Jha, The Myth of the Holy Cow, Verso, London, 2001, chapt. 3. LO
C
<
70. Ajay Mitra Shastri, India as Seen in the Brhatsamhit? of Var?hamihira, Motilal
Banarsidass, Delhi, 1969, p.214.
71. Manusmrti cited above. It also approves of meat eating by declaring that the
CU
flesh of an animal killed by a dog, a carnivorous animal and a C?ndala is pure.
Can?alas, called Dasyu in this verse, were apparently hunters selling animal
flesh.
72. The Par?sarasmrti places the women of washermen (rajaki), leather-workers
(carmak?ri), hunters (lubdhakT) and bamboo-workers (venujivanl) in the same
category and lays down the rules for purification, if a woman of any of these
castes stays even unknowingly in the house of a member of any of the four
varnas. Par?sarasmrti, edited with the 'subodhini' Hindi commentary,
Daivajnav?caspati Sri V?sudeva, Chowkhamba Sanskrit series, Varanasi, 1968),
VI, 44-45.
73. The Vedavy?sasmrti (113) enumerates Carmak?ra, Bhata, Bhilla Rajaka,
Puskara, Nata, Varata, Meda, C?n4?la, D?sa, ?vapaca, and Kolika as antyajas
and states that on seeing one of these or any other beef-eater (gav?sanah) one
should wash one's own eyes and have a bath on speaking to them. Smriin?m
Samuccayah, Anandaarama-sanskrt-granth?vali, no.48, 2nd ed., 1929, p.357.

An example of the intern?lization of this ideology by the untouchable groups
themselves is provided by K.R. Hanumanthan, who informs that according to
Valahkai Caritram, as non-beaf-eaters the Pallas considered themselves superior

to the Paraiyas who ate beef, although both were untouchable castes.
'Evolution of Untouchability in Tamil Nadu, A.D. upto 1600', IHR, XXIII,
Nos.1-2, (July 1996 & January 1997), p.64.
74. Joan, P. Mencher, 'The Caste System Upside Down, or the Not-So-Mysterious
East, Current Anthropology, vol.15, No.4, (December 1974), p.472. Briggs

writes that the Chamar belongs to the great class of unskilled labour. 'He is a
grass-cutter, coolie, wood and bundle carrier, drudge, doer of odd jobs, maker
and repairer of thatch and of mud walls, field labourer, groom, house servant,
peon, brick maker and even a village watchman', G.W. Briggs, The Chamars,
OUP, London, 1920, p.56. Dumont has to concede in this case that 'those
who are most materially are at the same time seen as supremely
oppressed
impure'. Homo Hierarchicus, p. 180.
75. S.M. Dahiwale, 'The Broken Men Theory of Untouchability' in S.M. Dahiwale
(ed), Understanding Indian Society: The non-Brahmanic Perspective, Rawat
Publications, Jaipur and Delhi, 1st published 2005, reprint 2006, pp.86-103.
Dahiwale points out that V.R. Shinde was the first to indicate the Buddhist
31
Social Scientist
CO background of a few of the present day untouchable castes, such as the Pulayas
o
o of Kerala and some outcastes of Orissa. He also from P.C.Alaxander's
quotes
Buddhism in Kerala (Annamalai University, Annamalainagar, 1949), which
na
shows that the Nambuthari br?hmanas converted Buddhist viharas into Hindu
S
-O
CL) temples and destroyed the influence of Buddhism by using the weapon of'social
ostracism'.
However K.R. Hanumanthan finds in the Buddhist and Jaina works like
C
Manimekalai and ?c?rakovai some traces of the concept of pollution and
untouchability 'due to puristic and ahims? doctrines of these religions. Op.cit.,
p.65.
76. Gail Omvedt, Buddhism in India, Sage Publications, New Delhi, 2003, pp 149
CO
185
"?
> 77. D.D. Kosambi, 'The Decline of Buddhism in India' first in 1956 and
published
included in Exasperating Essays: Exercises in Dialectical Method, published by
R.P. Nene, Pune, 1986, pp 63-6. Also see idem, Introduction to the Study of
Indian History, pp 246-7; 261-3; 291-4.
78. R.C. Mitra, The Decline of Buddhism in India, Visva Bharati, Santi Niketan,
1954.
79. It is for this reason that in propounding Navayana Buddhism Ambedkar rejected
the doctrine of karma and rebirth altogether, as according to him it was
contradictory to the basic Buddhist teaching of anatt? (non-soul), G. Omvedt,
op. cit. pp.2-6.
Badri Narayan writes that in the word 'Dalit' itself there is an inherent denial
of karma, pollution and legitimized caste hierarchy'. Women, Heroes and Dalit
Assertion in North India, p 34.
80. 'The Peasant in Indian History', General President's address to the Indian
History Congress, Kurukshetra, 1982 {Proceedings of the Indian History
Congress, 43 session), p. 17.
81. Y. Gopala Reddy, 'Socio-Economic Tensions in the Cola Period' Journal of the
Oriental Institute, Baroda, vol. 29, nos. 1-2, (Sept-Dec 1979), pp 74-84; R.S.
Sharma, Early Medieval Indian Society, pp 214-34; Suvira Jaiswal, 'Social
Dimensions of the Cult of Rama', in Irfan Habib (ed), Religion in Indian
History, Tulika, Delhi, 2007, pp 75-84.
82. For traces of such conflicts in the story of Hayagr?va incarnation of Visnu,
Suvira Jaiswal, 'The Demon and the Deity: Conflict Syndrome in the Hayagr?va
Legend', Studies in History, vol. I, no. 1, new series (1985), pp 1-13.
32
83. Stone of Bh?deva dated 916 Vikram Samvat, Shiva Prasad en

Bageshwar Inscription c
Uttarakhand ke Abhilekh evam <
Dabaral, Mudr?, V?ra-g?th?-Prak?shan,
Gadhaval Samvat pp 68-9; 162. 5*
(Vikram 2047),
84. Ibid., P?ncjukesvara copper plate inscription of Padmatadeva, tenth century,

?
p. 71. &
85. D.C. Sircar, Indian Epigraphic Glossary, (Motilal Banarsidass, Delhi, 1966),
s.v., parama-brahmanya.
86. B.N. Sharma, 'Religious Tolerance and Intolerance in Indian Sculpture', in
B.R. Saksena (ed.), Umesh Misra Commemoration Volume, Allahabad, 1970,
pp 657-68; C.S. Pathak, (ed.) Nalanda, Past and Present, Silver Jubilee
Souvenir, Nalanda, 1977, pp 109-13.Also see B.N.S. Yadava, Society and Culture
in Northern India (Central Book Depot, Allahabad, 1973), p 346 for some
instances of the persecution of Buddhists.
87. K.A. Nilakanta Sastri, A History of South India 3 edition, University of
Madras, Madras, 1966, p.424; Friedham Hardy, The Religious Culture of India:
Power Love and Wisdom, Cambridge University Press, South Asia edition, Delhi,
1995, p 51. He cites many examples of religious intolerance and conflict (p

105 f). The literature of this period also reflects intense sectarian rivalry.
Haribhadra S?ri's Dh?rt?khy?na is a biting satire on Puranic myths.
88. Padmanabha .S. Jaini, 'Jina Rsabha as an avat?ra of Visnu', Bulletin of the
School of Oriental and African Studies, vol. XL, pt.2, 1977, pp 321-37.
89. ?a?kara Pr?durbh?va quoted by Wendy Deniger O'Flaherty, 'The Images of
the Heretic in Gupta Pur?nas' in Bardwell Smith, ed, Essays on Gupta Culture,
Motilal Banarsidass, Delhi 1983, p. 121.
90. N. Venkataramanayya in G. Yazdani, ed., The Early History of the Deccan,
OUP, 1960, Vol II, part IX, p.712.
91. S.R.Sharma, Jainism and Karnataka Culture, Dharwad University, Dharwad,
1940, p.25; A.D.Pusalker in R.C. Majumdar (ed.), The Age of Imperial Kanauj
: The History and Culture of the Indian People, Bharatiya Vidya Bhavan,
Bombay, 3rd ed. 1984, p.291.
92. GaWarwad inscription of AD 1071, Epigraphia Indica, XV, p.337; K.A.Nilakanta
Sastri in G. Yazdani (ed.), The Early History of the Deccan, Vol.I, Pt.Vi, p.443.
Also see Ablur inscription of 12 century in Epigraphia Indica, Vol.V, no.25E.
R.N.Nandi has shown that the Jaina temples had become like landlords organizing
charities only for the followers of the Jaina religion, excluding the non-Jainas.
Religious Institutions and Cults in the Deccan, Motilal Banarsidass, Delhi,
1973, p.76.
33
Social Scientist
CO 93. ?a?karavarman of to meet other

o plundered sixty-four temples gods among
o the of the household
things, expenses royal (grhyakrtya). R?jataranginJ
vv. 167-9; R.S. Pandit's translation, Sahitya Academi, 1990,
reprint p.200.
S ?4. had the Vih?ra burnt down as his D?mara

-O Ksemendragupta Jayendra enemy
<L>
Samgr?ma had taken refuge in it. He also robbed the brass statue of Buddha and
built a shrine of Ksema Gaurisvara in ?rinagara. Ibid., VI, vv. 171-3, p.243.
rd
D
C 95. Kalasa took the copper of the Sun-god known as T?mrasv?min
King away image
and also many brass statues from the vih?ras. Ibid., VII, vv. 696; p.319.
96. Ibid. VII. vv.1080-90, pp 351-2, Harsa, however, spared the images of
O
Z Ranasv?min and M?rtan4a along with two Buddha images, vv. 1096-8. In the
Prabandha Cint?mani_Ac?rya. Merutunga refers to king Ajayadeva's destruction
m
of the temples set up by his predecessor, presumably by his father Kum?rap?la,
O
> C.H. The Prabandha Cint?mani or Wishingstone of
Tawney (translator),
Narratives. The Asiatic Society, Calcutta, 1901, p. 151.
97. According to B. V. Krishna Rao (AHistory of the Early Dynasties of?ndhradesa,

Madras, 1942, pp 57-8) and M. Rama Rao (Iksv?kus of Vijayapuri, S.V.
University Tirupati, p. 35f), Iksv?ku king Virapurusadatta had renounced
?aivism and adopted Buddhism. A sculpture at Nagarjunakonda depicts him
trampling ?iva-linga under his foot showing his denunciation of ?aivism.
98. R.C. Mitra, Decline of Buddhism in India, pp 38f; 55f; 74 and elsewhere; KA
Nilakanta Sastri in G. Yazdani (ed.), The Early History of the Deccan, Vol. I,
pt.VI, p 438f.
99. The Hoyasala ruler Ball?ladeva is described in his inscriptions as the supporter
of all the four samayas, M?hesvara, Bauddha, Vaisnava and Arhat. R.C. Mitra,
op. cit., pi 14. Similarly an inscription of 1022 CE from Belur informs us that
Akk?devi, the elder sister of Jayasimha II of the C?lukyas of Kaly?ni, performed

all the dharmas mentioned in the ?gamas of Jaina, Buddha, Ananta (Visnu)
and Rudra. Indian Antiquary, vol. 18, (1889), pp 279-5.
100.It is wrong to hold that royal patronage of Buddhism ceased after the seventh
century and the P?la dynasty (750 CE -1161 CE) was the sole exception
(Omvedt, Op. ct, p. 172). In central India the Gahadav?la king Jayacandra of
Kanauj was a Buddhist and his preceptor was a Buddhist monk named ?rimitra.
Indian Historical Quarterly, Vol. V (1979), pp.14-29. His predecessor
Govindacandra, although himself a paramam?hesvara, granted villages to
Buddhist monks in the Jetavana - His two

living vih?ra. queens Kumaradev?
and Vasantadev? were Buddhists and the former had the famous Dharmacakra
Jinavih?ra constructed at Sarnath. N.N. Das Gupta in R.C. Majumdar (ed.),

Struggle for the Empire, Bharatiya Vidya Bhavan, Bombay, 2 ed. 1966), pp
34
422-3. In Orissa and the Deccan too several Buddhism in the en

kings patronized c
eleventh twelfth centuries.
<
101.In the Aga??a Suttanta of the DJgha Nik?ya the Buddha explains to his two
brahmana the origin of the universe as well as of the khattiya

to
young disciples
and br?hamanajnandalas (groups) and of the vessas and suddas. In its origin
the division is functional with the khattiyas occupying the first place, but later
these are assumed to be fixed or hereditary. DJgha Nik?ya (T.W. Rhys Davids
and J.E. Carpenter, eds. 3 vols, Pali Text. Society, London, 1890-1911), vol.
3, p. 93f, translation by TW Rhys Davids (3 vols Sacred Books of the Buddhists,
London 1899-1921), vol. 3; p. 88f
We do not have Jaina works of a comparable early date, but the varna divisions
are taken for granted in the ?c?ranga and Uttar?dhyayana s?tras.
102.N. Wagle, op. cit, p 125 f
103.Uma Chakravarti, op. cit., Appendix C.
104.A.K. Warder, Indian Buddhism, Motilal Banarsidass, Delhi, 1970, pp 232-3.
105 .Debiprasad Chattopadhyaya, Religion and Society, Stephanos Nirmalendu Ghose
Lectures, (1981) of Calcutta University, Ma-Le Publishers, Bangalore, 1987,

Lecture VII; Devaraj Chanana, op cit, p. 60f.
106.Jataka nos. 59; 179; 309; 497; 498 in Jatakas, Fausboll (ed), 7 vols, with
Index, London, 1877-97. Tr. Various hands under the editorship of E. B. Cowell,
7 vols with index, Cambridge, 1895 -1907.
107.P.V. Kane, History of Dharmas?stra, vol. V, pt ii (Bhandarkar Oriental Research
Institute, Poona, 1962), p 926; 1009-10.
lOS.Visnudharma Pur?na, chap. 25; Brhann?radiya Pur?na, 14-70; 186; 22.9
quoted in R.C. Hazra, Studies in the Upapur?nas, vol. I, pp 147-8; 325-7. The
Arthas?stra of Kautilya II.4.23 instructs that the dwelling place of the/? ?s andas
and Car?alas should be on the outskirts of the cremation ground. The Kaut illy a
Arthas?stra, edited by R.P. Kangle, Vol. I, University of Bombay, Bombay, 2nd
edition 1970) p.39.
109.Nalinakshadutt in the Classical Age, History and Culture of the Indian People,
vol.Ill (Bharatiya Vidya Bhavan, Bombay, 3rd edition, 1970), p.386.
110. Vajrasuc? of Asvaghosa, edited and translated by Sujit Kumar, Visvabharati,
Santiniketan, 1950. The identification of this A?vaghosa with the author of
the Buddhacarita is, however, doubtful.
Ill .Latakamelaka, Act II, quoted by B.N.S. Yadava, op. cit., p 8.
35
Social Scientist
CO \ 12.Padma 4 vv Prasad
o Purana, pt. I, chap. 86f, quoted by Ram Bhushan Singh,
o Jainism in Early Medieval Kamataka (c. AD 500-1200), Motilal Banarsidass,
Delhi, 1975, pp.75-6.
2 VIII. XV.
-O 113.?dipur?nam, 64; 6-12, quoted by Malini Adiga, The Making of
<D
LL. Southern Kamataka: Polity and Culture in the Early Medieval Period,
I Society,
?^ Orient Longman, Chennai, 2006, p. 259.
n3
ZJ
C 114.U.N.
(ti Ghoshal, op.cit., pp 457-62.
rs
11 5.Malini Adiga, op. cit.
to
O 116.Jyoti Prasad Jain, The Jaina Sources of the History of Ancient India (100 BC
Z AD 900), Munshiram Manoharlal, Delhi, 1964, p. 215.
m
Wl.Yasastilaka, pt II, Book viii, quoted by R.B.P. Singh, op. cit., p. 73. R.S.
"5
> Sharma quotes Dharmanand Kosambi (Bhagav?n Buddha, trans. From Marathi
into Hindi, Shripad Joshi, Delhi, 1956, p. 258) to point out that Jainism
forbids the initiation of untouchables (Jumgita) into monkhood. Material
culture and Social Formations in Ancient India, pp. 129-30.
11 8.R.B.P. Singh, op.cit., pp 74-82.
119.D.D. Kosambi, An Introduction to the Study of Indian History p. 155, According

to Sukumar Dutta Buddhism was instrumental in giving a death blow to the
pretensions of caste superiority of hereditary aristocracy and divinity of
kingship in Cambodia^Transactions of Indian Institute of Advanced Studies,
Simla, vol. I, 1965, p. 179f.
120.S. Jaiswal, Caste, pp 53-4, Irfan Habib, Essays in Indian History: Towards a
Marxist Perception, Tulika, New Delhi, 1995, pp 161-79.
121.Devaraj Chanana, op. cit. p 82-4.
122.Yi-Jing (I-Tsing), A Record of the Buddhist Religion as Practised in India and
the Malay Archipelago, translated by J. Takakasu, Oxford, 1896, p. 62 quoted
by Irfan Habib, (ed), Religion in Indian History, Introduction, p xxiii.
123.Devaraj Chanana, op. cit., p.83.
124. Vinaya Pitaka, IV. pp 300-1 quoted by LB. Horner, Women under Primitive
Buddhism, Motilal Banarsidass, Delhi, first edition, 1930, reprint, 1990, p.222.
Also see pp.233-4.
125.Friedham Hardy, The Religious Culture of India: Power, Love and Wisdom,
Cambridge University Press, South Asian Edition, Delhi, 1995, p.451.
126.Debi Prasad Chattopadhyaya, op. cit.
36
127.B.N.S. Yadava, op. cit., p. 380; Nupur Chaudhuri and Rajat Kanta Ray, 'Eros en
c
and History: Sahajiya Secrets and the Tantric Culture of Love' in Irfan Habib <
(ed), Religion in Indian History, p. 107.
128.For the shift in emphasis from relative purity of function to relative purity of to*
birth in the varna-j?ti organization, S. Jaiswal, Caste, pp 13-18; 71-7.
129.Suvira Jaiswal, Origin andDevelopment of Vaisnavism, 2nd revised and enlarged
edition, Munshiram Manoharlal, Delhi, 1981, pp. 123-129.
130.Manusmrti, X. 84. B?hler's translation. Irfan Habib writes that this provided
one more argument for treating all peasants as s?dras, Religion in Indian
History, p. xxiii.
131 .Ibid., III. 68-70. On the changing concept of the panea-mah?yaj?as, S. Jaiswal,
Caste, p. 121 note 239.
132.R.S. Sharma, Perspectives in Social and Economic History of Early India,

Munshiram Manoharlal, Delhi, 1983, pp 45-8.
133.C.A.F. Rhys Davids, Psalms of the Early Buddhists, vol. II, P.T.S., London,
1948, pp. 15; 25; Uma Chakravarti, Social Dimensions of Early Buddhism, p.
34.
134.Rgveda, I., 28.5.
135.Ibid., I. 92.3. Does this hymn represent some older matriarchal substratum?
Later, verse 3 of this hymn prays for the gift of 'troops of d?sas\ Compare
D.D. Kosambi, Myth and Reality, pp. 68-9.
136.S. Jaiswal, 'Semitizing Hinduism: Changing Paradigms of Brahmanical
Integration', Social Scientist, vol. 19, no. 12, December 1991, pp 20-32.
Reference is to the brahmanical ideology and not to the role of the brahmana
caste in particular. Also see S. Selvam, 'Sociology of India and Hinduism: Towards
a Method', in Dalits inModern India: Vision and Values, edited by S.M. Michael,
p.189.
137.Richard F. Gombrich, Buddhist Precept and Practice: Traditional Buddhism in
the rural Highlands of Ceylon, revised edn., OUP, Delhi, 1991, p. 345.
138.Brajadulal Chattopadhyaya speaks of the shaping of regional societies as
essentially a movement from within. He also draws attention to the fact that
the horizontal spread of the varna ideology 'drew widely dispersed and originally
outlying groups into a structure which allowed them in a large measure to
retain their original character'. The Making of Early Medieval India, OUP,
Delhi, 1994, pp.35; 203. Also see S. Jaiswal, Caste, pp.230-1.
139.Hermann Kulke, Kings and Cults: State Formation and Legitimation in India
37
Social Scientist
OO and South-East Bhairabi Prasad

o Asia, Manohar, Delhi, 1993, pp. 82-93; Sahu,
o as a Mirror
'The Past of the Present: The Case of Oriya Society' Social
Science Probings, vol. IX (1-4), March-December 1995, pp.8-23.
S 140.S. and of Vernacular

-O Nagaraju, 'Emergance of Regional Identity Beginnings
? Literature: A Case Study', Social Scientist, vol. 23, nos. 10-12, October
I
December 1995, pp 8-23.
r?
C 141 .The confusion has been traced to R?m?nuj?c?rya, who in his book Brahmas?tra
Bh?sya (IL 1. 37-42) identifies the K?l?mukhas with K?p?likas. But in fact
these two were quite distinct; and the K?l?mukha monks seem to have been in
O active competition with the Jaina monks. Lorenzen is quite that there
positive
Z
\0 is no reason why the K?l?mukhas should not be regarded as orthodox pandits.
'The K?l?mukha Background to V?raaaivism: Studies in Orientology' in S.K.
O
> Maity, Upendra Thakur and A.K. Narayan eds., Essays inMemory of Professor
A.L. Basham, Y. K. Publishers, Agra, 1988, p.279. Also see S.C. Nandimath
'?aivism' in R. R. Divakar and others, (eds.),Jiarnataka Through the Ages,
The Government of Mysore, Bangalore, 1960, pp 153-4, R.N. Nandi, Religious

Institutions and Cults in the Deccan, pp.85-90.
Nagaraju refers to the anti-caste feelings expressed in the poems of Mallik?rjuna
Parujit?r?dhya. But he belongs to the twelfth century and was a contemporary
of Basava, the founder of the Ling?yat movement, which was strongly anti
vedic and anti-caste at least in its origins.
142.The ninth century Maruru inscription of Arkalgud Taluk records a landgrant to
a K?l?mukha centre for vidy? d?na. B.R. Gopal et al., Epigraphia Carnatica,
1984, 8: Ag 28 quoted in Malini Adiga, op. cit., 308.
143.Jabalpur Stone Inscription of Jayasimha, verse 44, V.V. Mirashi, Corpus(

Inscriptionum Indicarum, Vol. IV pt. I, no. 64, Government Epigraphist for
India, Ootacamund, 1955, trans, p- 339.
144.J.H. Hutton, Caste in India: Its Nature, Function and Origins, Cambridge
University Press, Cambridge, 1946, p. 249.
145.Louis Dumont, A South Indian Subcaste: Social Organization and Religion of

Pramalai Kallar, OUP, Delhi, 1986.
146.Ibid., p.12. For the reinterpretation of the s?dra category in the context of
South Indian communities, S. Jaiswal, Caste, pp 70-71.
147.Nicholas B. Dirks, op. cit., pp 12-60.
14 8. Ibid.
149.Hutton, op. cit., p. 178-9 for the eight prohibitions propounded by the Kallar
38
of Ramnad in December 1930. Some of these were that the Adi-Dravi<Ja en

c
women shall not cover the of their shall <
(untouchable) upper portion bodies,
not use flowers or saffron the males shall not wear clothes above their 3"
paste;
hips or below their knees and so on. Non-compliance led the Kail ars to use
violence against the Adi-Dravidas, whose huts were burned, granaries and
properties destroyed and livestock looted.
1 50.A South Indian Subcaste, pp 218-23.
151.Nicholas B. Dirks, op. cit., p. 72.
152.For example^ the assumption of the 'Y?dava' title by Gwala, Ahir, Gope,
Sargope, and Ghasi castes and formation of the All-India Y?dava Mah?sabh?.
Not rarely the drive towards unification remains confined to political level,
social interactions still being regulated by traditional customs.
1 53.Nicholas B. Dirks, op. cit., pp 295-6, Kancha Illiah, 'BSP and Caste as Ideology',
Economic and Political Weekly, 29 (12), 1994, pp. 668-9; idem, 'Productive
Labour, Consciousness, and History: The Dalitbahujan Alternative', in S Amin
and D. Chakraborty, eds., Subaltern Studies, IX, OUP, Delhi 1996, pp 165
200.
154.B.R. Ambedkar, Annihilation of Caste: An Undelivered Speech, edited by Mulk
Raj Anand, Arnold Publications, New Delhi, 1990, pp 81-82.
155.This entire essay goes against Dirks' assertion that Ambedkar 'was convinced
that caste (or, rather untouchable) identities had to be fo-stered in order to
combat centuries of oppression' (op. cit., p. 278). Ambedkar was so disillusioned
with the caste system that in the end he along with a large number of his
followers converted to Buddhism.
156.On Dalit mentality, S. Jaiswal, 'Dalit Asmit aur Agenda j?ti vin?sa k?' in
Akhilesh, ed., Tadbhav, vol. 15, (Jan. 2007), pp. 27-40.
39
Social Scientist
Caste, Untouchability and Social Justice: Early North Indian Perspective

Author(s): Vivekanand Jha
Source: Social Scientist, Vol. 25, No. 11/12 (Nov. - Dec., 1997), pp. 19-30
Accessed: 03-03-2015 07:00 UTC
VIVEKANANDJHA'
andSocialJustice:
Caste,Untouchability
EarlyNorthIndianPerspective
Social justiceis a comparatively moderntermand connotesjust and fair

treatment to thepeopleconstituting a society.It presupposesa social order
whichis non-discriminatory and people-oriented, one in whichdisparity,
inequality andinequity do notcharacterize social,economicandotheraspects
oflifeandinstitutions. Thetermoccurswithdueemphasisinourconstitution.
ItspreambleproclaimsthesolemnresolveoftheIndianpeopletosecureto all
itscitizens"justice,social,economicandpolitical"andarticle38 commits the
Stateto takeappropriate effective stepsto usherin "a social orderinwhich
justice,social economicand political,shallinform all theinstitutions
ofthe
nationallife".A fewotherarticles, too,dealwiththevariousfacetsofjustice.
Alltheseshowan acuteawarenessofthefactthatjusticeis byitsnaturean
integralwhole,thatelements ofinjusticeare pervasivein Indiansocietyas a
partof its colonial and precolonialheritageand thatseriousefforts are
requiredto remedythe situationand bringabout social transformation.
Continuous changesinthecourseofourlonghistory notwithstanding, thefact
of continuity fromthe past is undeniableand certainlyearlyIndia has
contributed its share to the presentsituation.Withoutgoing into other
dimensions ofthetheme,I shalltryto showhowtheoriginand development
of caste and untouchability in earlynorthIndia has been instrumental in
perpetrating social injusticeto a largesegment oftheIndianpeople.
Caste maybe definedas a system of socialstratification
characterized by
hierarchy, heredity, pursuit ofoneora fewparticular occupations, inequality,
endogamy,restrictions as to takingfoodfromoutsiders, and thenotionof
purity andpollutionassociatedwithhierarchy. Notwithstanding theexistence
of a full-fledged class societyin the pre-Aryan maturephase of Harappa
culture,the availablearchaeologicalevidenceunaidedby writtenrecords
owingto thehitherto undeciphered scriptdoes not warrantthehypothesis
regarding theemergence ofcasteanduntouchability there.1The evolutionof
casteas a socialphenomenon has,therefore, to betracedthrough thestudyof
twoseminalterms,varnaandjati,varnabeinganteriorto jati and receiving
muchgreaterattention in theearliertextsthanjati. Frombeingusedto dis-
tinguish Aryafromtheethnically and culturallyseparateDasa and Dasyu in
theRigveda(c. 1SOOBc-c.100OBc),varna,literally meaningcolour,cameto
FormerDirector,
IndianCouncilofHistoricalResearch,New Delhi
SocialScientist,
Vol. 25, Nos. 11-12, November-December
1997
20 SOCIAL SCIENTST
be appliedto thefourhierarchically rankedoccupationalcategoriesof the

brahmanas,kshatriyas, vaishyasand shudrasduringthelaterVedic period.
Although thebrabmanaandtheksbatriya arementioned intheRigvedaa few
timesinthesenseoffunctional groupswhichhademerged fromtheAryanvish
orjana,meaning the
tribe, brabmanaas a priestcomposing andreciting hymns
and officiating at thesacrifice ofthekshatriya warriorchief(rajan),theterm
varnaisneverappliedtothebrabmanaorthekshatriya. We occasionally come
acrossa fewgenerations ofchiefsintheRigveda,buttheexamplesofa poet
describinghis fatheras a physicianand his motheras a corn-grinder, of
anotherpoet enquiringfromIndrawhetherhe would be made a sage, a
protector ofthepeople,a rulingchiefor an ownerofenduring wealth,and of
kshatriya princesDevapi and Devashravasofficiating as priestsat thesacri-
ficesof theirrulingyoungerbrothers Shantanuand Devavata respectively,
show that professions had not becomehereditary at this stage,that the
brahmanaand kshatriya rankswereopen and thatthesewerea matterof
achievement ratherthaninheritance.2 As a peopleoftenon themovein the
of
land thesevenrivers, theprimarily pastoralRigvedicAryanswereneither
practisingendogamynor observingany restrictions regardingfood from
others.
Thefirst definite indication ofthefour-tier hierarchicalinequality develop-
ingamongtheVedicAryans, thoughwithout theuseofthetermvarna,isfound
in the famouspurushasukta of the tenthmandala of the Rigveda,which
represents thebrabmanaas themouth,therajanyaas thearms,thevaishyaas
thethighsand theshudraas emanating fromthefeetofthedivinePurusha,3
whenhe was sacrificed bygodsforthesakeofcreationoftheuniverse.The
hymnis presumedto have accordeddivinesanctionto theemerging social
structure.It issignificant thatthetenthmandalabelongsto thelateststratum
oftheRigveda,synchronizing withsomeofthelaterVedictexts,and thatthe
the
rajanya, vaishyaand theshudraappearinthishymnforthefirstand last
timein theRigveda.4
In therelatively stablesubstantially agrariansettingof theupperGanga
basinin laterVedictimes(c. 1000 BC-C. 600 BC) theprocessofsocial differ-
entiationwentonsteadily andthefourvarnas,distinct andseparatefromeach
other,appearas a full-fledged social reality,
thebrabmanasas a specialized
classofpriests monopolizing thecomplexritualsandas scholarsandteachers,
thekshatriyas as warriorsand rulerscontrolling largerterritorialunitsand
materialresources as a resultofparticipation andvictories inongoingbattles,
thevaishyasas tribute-paying peasants,cattle-rearers,artisansand traders,
and theshudrasas domesticservants, agricultural
labourersand slaves.The
textsdo notleaveanyroomfordoubtregarding thedominantpositionofthe
brabmanasand thekshatriyas vis-a-visthevaishyasand theshudrasin the
increasingly inegalitarian milieuofthetimeswithmoresurplusavailablefor
unequal distribution.Despite a protractedkshatriyachallenge to the
Brahmanical claimsto primacy,bothcombinedwell againstthetwo lower
varnasand theAitareyaBrahmanasdescription of thevaishyaas anyasya
balikrita(a tributary to others),anyasyadya(one who is livedon byothers)
andyathakamajyeya (onewhocan be oppressedat will)and oftheshudraas
and SocialJustice
Caste,Untouchability 21
anyasyapreshya(a servantor messenger of others),kamotthapya (one who

can be madeto workat anytimeofthedayor night)and yatbakamavadbya
(one who can be beatenat will),6is indeedrevealing.
Towardstheend of thelaterVedic periodthevarnastendedto become
hereditary, endogamousandbirth-based, leadingtotheformation ofjatis.The
termjati is derivedfromtheSanskrit rootjan,meaningto be born,andis first
appliedby pre-Paninian Yaska in hisNiruktato a womanof the blackor
shudracaste(krishnajatiya); itis maintained thatthoughsexuallyenjoyable,
she shouldnotbe approachedafterthefirealtarhas beenlaid as thisis not
conductive toreligious merit.7Paninishowsacquaintance withjatiinthesense
ofcasteinhissutra,jatyantachcha bandbuni.8 Thatbirthwas slowlybecom-
ingan important factorofsocialranking and thetheory ofkarma(deed)and
punarianma(rebirth), whichprovedsuchan effective ideologyin theinter-
nalizationoftheinequitouscastesystembytheoppressedand theexploited
and was ardently championedbyBuddhismand Jainism as well,was taking
shapeduringthisperiodis borneout bytheChbandogyaUpanishadassign-
mentof purebirth(ramaniyayoni)to thebrabmana,thekshatriya and the
vaishyaandimpurebirth(kapuyayoni) totheChandala,thedogandtheboar,
andattribution of birthintheformer category togood deedsand inthelatter
categoryto evildeeds.9
The periodsaw thebeginning Qftheprocessofassimilation, acculturation
and integration oftheaboriginaltribesintotheexpandingAryannetworkat
variouslevels.ThustheAitareyaBrabmanadescribestheAndhras,Pundras,
Shabaras,Pulindasand Mutibasas antas(borderpeople)and theprogeny of
thedefiant accursedsonsofsageVishvamitra,'0 and refers to theforesttribes
andhunters as apachyasandnichyaswiththeirownchiefs;1 1therearecopious
references to the proximity of and interaction withthe largerand better
organized Nishadas;12 thededication ofthePaulkasatobibhatsa(loathsomeness
as a deity)inthesymbolichumansacrifice (purushamedha) intheVajasaneyi
Sambita13and the Taittriya Brahmana14showsthatthePaulkasaswerean
objectof spiteand revulsion; and theBrihadaranyaka Upanishadstatement
thatall distinctionsvanishinthespiritual realmwhereeventheChandalasand
thePaulkasaslose theirseparateidentities15 indicatesthatdisparities were
growinginthematerialworldandthesetwogroupsstoodat thelowestlevel
of theexistingsocial hierarchy. Caste was evidently in its formative stage
during thelaterVedicperiodandjatiwasimbibing manyofthetraitsofvarna.
Till theend of thelaterVedic period,however,interdining amongthefour
varnaswas notprohibited, inter-varnamarriages didtakeplace,andtherewas
no untouchability.
The post-Vedicperiod(600 BC-200BC) is markedbytheextensiveuse of
ironforproduction, enormousexpansionoftheeconomy,substantial risein
theavailablesurplusand accentuated economicinequality inthefull-fledged
classsocietyofthemiddleGangabasinandfurther east.Thisprovidedan ideal
localefortheemergence ofa morestratified societyand consolidation ofthe
varna-jatistructure. TheDharmasutras ofApastamba,Baudhayana, Gautama
and Vasishtha(600 BC-300BC) reflectthisclearlyin the relatively more
frequent use ofjati in thesenseof caste.The termoccurseighttimesin the
22 SOCIAL SCIENTIST
GautamaDharmasutra,16 six timesintheBaudhayanaDharmasutra,17 and

four times each in the VasishthaDharmasutra18and the Apastamba
Dbarmasutra.19 Varna,however, continues to bethemajortermfordesignat-
ingcasteandoccurstwenty-four timesintheBaudhayanaDharmasutra, twenty-
threetimesin theApastambaDharmasutra, fourteen timesin theGautama
Dharmasutra and twelve times in the Vasishtha Dharmasutra. The
Dharmasutras placethehierarchical socialpositionandoccupationalrolesof
thefourvarnasina legalsetting anddetailtheprivileges ofthefirst threetwice-
born(dvija) varnas,demarcating themclearlyfromtheshudras,who are
saddledwithnumerousand varieddisabilities. These includedobligatory
serviceto thetwice-born and physicaltoilas landlessagricultural labourers,
artisans,wage earnersand slaves,denialof initiationwithsacredthread
(upanayana),exclusionfromVedic studyand sacrifices or sacraments with
Vedicmantras,inequality beforelaw inmatters relating to inheritance,rates
ofinterest and criminaloffences, lackofaccessto judicialand highadminis-
trative
positions, andrestrictions as tocommensality, associationandmarriage
withsuperior varnas.Thesetextsalso drawa linebetween thefirsttwovarnas
andthevaishyasand thougheventhelatterare permitted to takeup armsto
preventthe mixtureof varnas,20 primarily entrusttheformerwiththe re-
sponsibility to maintainthevarnaorderwhichis regardedas sacrosanct.
The Dharmasutras are unanimousin prescribing sixfolddutiesof study,
sacrificing,givinggifts, teaching, sacrificing forothersand receiving giftsfor
thebrahmanasand participating in battles,protecting peopleand wielding
political,administrative and judicialauthority forthekshatriyas. Gautama,
however,permits a brabmanato takeup agriculture and tradeprovidedhe
doesnotdirectly engageinit(asvayamkrite);21 Vasishthaallowsthebrabmanas
unabletomaintain themselves through theirlawfuloccupations(ajivantab)to
adopt thekshatriya profession of armsand, failingin that,thevaishyaoc-
cupationsof agriculture and tradewithrestrictions on sellingcertaincom-
modities,and even directlytillingland to produce sesamum (svayam
krishyotpadya tilan)provided duecareis-taken Qftheoxen;22andBaudhayana
notonlyechoesVasishthato theextentofaccording to sucha brahmanaeven
permission to ploughthefield(karshisyat)whiletreating theoxen mildly,23
butmaintaining thatthestudyoftheVeda and practising agriculture impede
eachother(vedahkrishivinashaya krishirvedavinashini), ordainsthathewho
isableto attendto bothshoulddo so andhewhoisunableto do it shouldgive
up agriculture (shaktimanubhayam kuryadashaktastu krishim tyajet).24
De-
spitethe flexibility shown by the lawgiversowing to considerationsof
practical
constraints, thereisnodoubtthatsizeablesectionsofbrahmanas and
kshatriyas who couldaffordittendedto withdraw themselves fromprimary
productiveactivitiesand came to broadlyrepresent'status' and 'power'
respectively; theshudrassubstantially providedproductive manuallabour;
andthegap betweentheeliteandthemasseswidened,intensifying thenotion
ofthehighand thelow.
Thesocialfabricatthistimewas,however, ina tremendous fluxandseveral
professions, crafts and tribeswerecrystallizing as distinctentities.Neither
theirexistencecould be ignored,norcould theybe identified withthefour
and SocialJustice
existingvarnas.Thisgaverisetothetheory ofvarnasamkara ormixedcastes,

whichascribedtheiroriginto interbreeding amongthemembers ofthefour
varnasand also amongtheirprogeny fromanuloma(innaturalorderor with
a womanoflowervarna)andpratiloma(in inverted orderor witha woman
ofhigher varna)unions.25 Therelativelysuperior rating
ofanulomatopratiloma
the
wasdueto patriarchal natureofsociety. The Dharmasutras mention a total
oftwenty-four suchmixedcastesresulting frommiscegenation at thespecified
varna levels.Seriousdisagreement amongthe authorsabout the number,
names,classification anddetailsofderivation ofthesemixedcastes,however,
exposeglaringcontradictions in thisspeculative theoreticalexercise.26 That
themixedcastesconstitute a residualcategoryapartfromthefourvarnasis
borneoutbyBaudhayana'streatment oftheinhabitants ofAvanti,Magadha,
Surashtra, Dakshinapatha, Upavrit, Sindh and the Sauviras as
sankirnayonayah.27 The notionofvratya,subsuming theAryanoriginof a
groupanditssubsequent lossofstatusduetothenon-observance ofvarnanorms
suchas upanayana,is anotherconceptwhichwas used independently28 and
as partofthevarnasamkara theory29 toaccommodate theexterior groupsinto
themainstream.
The notionofritualpollutioninrelationto theshudrahad madea fleeting
appearancetowardstheend ofthelaterVedic periodwhentheuse of milk
milkedbytheshudrawas forbidden at thedailyagnihotra(oblationto fire)30
and a personconsecrated fora sacrifice(dikshita)was enjoinednotto speak
to theshudra.31 The shudrahad evenbeendesignated ayajniya,thatis,unfit
to perform a sacrifice.32
Physicalcontactwithor touchof theshudrawas,
however,notregardedas polluting. ThoughtheDharmasutras continuedto
adheretothisposition, thenotionofpollutionas an enduring feature ofsocial
life-an important characteristicofcaste-was institutionalized as untouch-
abilityin theselawbooks.Untouchability meantpermanent and hereditary
pollutionowingtophysical contactwitha sectionoftheIndianpeopleandthe
groupfirst forthepurposewastheChandala.TheDharmasutras
identified are
unanimousinholdingthetouchoftheChandalasas polluting and prescribe
bathwithclotheson as a meansof expiation.The Chandalas also cause
pollutionthrough proximity, sight,hearingandspeech,entailing correspond-
ingexpiations.Physicalassociationand commensaland connubialtieswith
theChandalasare completely prohibited and theirsegregation is legalized.
The Shvapakas and the Antyavasayins may be identifiedas two other
untouchable castesattheChandalalevel.Termssuchas anta,antya,antyayoni,
bahya,apapatra,etc.,signify thisnewsocialphenomenon anddistinguish the
untouchables fromtheshudras.Evidently closerintegrationoftheChandalas
insocietyinvolvedfurther depression inthestatusofthislaterVedictribe.33
The theoretical originoftheChandalasfromthemosthatedpratilomaunion
of shudramenwithbrabmanawomenreflected thisdisdain,thoughsuch
union on any considerablescale was unthinkablewithinthe varna-jati
structure and was nevera tangiblesocial reality.
The trendset bytheDharmasutras on theissueof social stratification is
substantially endorsed,elucidatedand elaboratedbytheGrihyasutras, the
Ashtadhyayi of Panini,theArthashastra of Kautilya,the Manusmriti, the
24 SOCIAL SCIENTIST
Mahabhashya ofPatanjali,theRamayanaandtheMahabharata.Although the

detailsvary,we have information about manynewgroupsbeingabsorbed
withinthevarna-jatiframework, theirchangingpositionin socialhierarchy
and perceptions about it,theirprescribedand actual functions, and their
fissionand fusion.Varnaandjati areusedcommonly inthetextsinthesense
ofcasteand thoughthejatisbeingmorenumerousthanthevarnasareoften
distinguished in earlyIndianliterature, thetwo termsare also used inter-
changeably. To takejustoneexample,Manu'sviewthathinahinanprasuyante
varnanpanchadashaiva tu34doesnotmeanthatthelow(varnas)producefifteen
low (varnas),butthatthelow (six pratilomajatisinan ascendingorder,the
Chandala,Kshttri, Ayogava,Vaidehaka,Magadha and Suta) produce(on
pratilomawivesor through pratilomaconnections) fifteen lowjatis.35What
is significant is thatuntouchability developedin stagesand thenumberof
reallyuntouchablecastesat the bottomof societygrewratherslowly.The
cumulative evidenceoftheBrahmanical textsup to AD200 doesnotadd more
thanthreeor foursuchcastesto theDharmasutra listofthree.Of thesethe
Pulkasas and the Medas had, like the Chandalas, an indigenoustribal
background and wereessentially huntersbyprofession. The steadyadvance
oftheorganizedsocietyand itsencroachment intoforestareasdepletedtheir
sourceofsubsistence andobligedthemtojointhedominant productive system
forthesakeofminimum economicsecurity.36Thereisevidencetoindicate that
theexpectationremainedlargelyunfulfilled. The Chandalasemergeas an
extremely poorsegregated groupenlistedto workat thecrematorium, hang
criminals, dragpeoplecommitting suicideon thepublicroad,whipadulterous
women,and even to defendnew settlements in the countryside. They,
however, continuedas hunters, fowlersand butchers as well,withdogs,cats,
pigsand cocksas theirproperty, wearingclothesand ornaments ofthedead
and theexecuted,and depending on othersfortheirfood.The Mritapasand
theMatangas,whichapptarin someof thesetexts,mayhave beenmerely
subgroupsoftheChandalasand areoftenspokenofas synonymous withthe
latter.
It is important to considerhow the heterodoxsectsof Buddhismand
Jainismrespondedto thedevelopment of caste and untouchability. Surely
theirapproachdiffered fromorthodoxBrahmanism inseveralrespects. Their
religiousordersadmittedpeoplewithoutconsiderations of wealth,rankor
socialoriginsand permitted themto riseto thehighestpositionon thebasis
ofvirtueand knowledge alone.Theydeniedthatthefour-varna structure had
anydivinesanctionbehindit,thatthebrahmanas, eventhoughtheyengaged
in worldlypursuits, wereentitledto thehighestsocial status,and thatthe
shudrasweremeantto servethethreehighervarnas.Fora change,theycon-
sistently placedthekshatriya abovethebrabmanawhileenumerating thefour
varnas.Buddhismregardedagriculture (kasi), trade(vanijja) and animal
husbandry(gorakkha)as highprofessions (kamma).While Buddhismac-
cordedan honourablepositionto thegahapatisirrespective oftheirvarying
varnabackground, gahavaisbelonging to themercantile community received
similarrespectinJainism. Evenso, bothBuddhismandJainism acceptedthe
reality offourhierarchical varnasandseveraljatisas wellas untouchability as
and Social Justice
an integral and inalienablepartofthecomplexsocio-economic formation in

thepost-Vedic phaseandshiedawayfromattempting anystructural changes
in theexistingsocialorder.Theirhavenof equalityand fraternity remained
confined to theirsanghaand evenherewomenremained subordinate to men.
Buddhismwas pragmaticenoughnotto alienateitsclientsand patronsby
admitting slaves,debtorsanddeserting soldiers.Amongthepeoplewhojoined
thesanghaor supportedBuddhismfromoutsidetheoverwhelming majority
belongedto highcastesand families.37 The distinction betweentheaffluent
and thedestitute, thehigh(ukkattha)and thelow (hina),is pronouncedand
explicit,andisappliedtojatis(castes),kulas(families), kammas(occupations)
and sippas (crafts)in bothBuddhismand Jainism.The occupationof the
flower-sweeper (puppachhaddaka)and the craftsof the basket-maker
(nalakara), leather-worker (chammakara),weaver (pesakara), potter
(kumbhakara) andbarber(nahapita)aredesignated as low,andtheChandalas,
Nesadas,Pukkusas,Venasand Rathakarasarestigmatized as despisedcastes
(hina jatis) in the Pali Canon. The Chandalas, who are also known as
in
Matangasand Panas,and theSovagasare equallydespised theJaintexts.
The professions ofhunters and fishermen, too,are ratedverylow.
The natureof untouchability practisedin relationto theChandalasand
theireconomicplightintheBuddhist Jatakas,whichareposterior to thePali
Canon and oftencoincidewiththeManusmriti, are as severeas had been
prescribed bythislawgiver whoisgenerally regarded as thesymboloftheugly
faceof Brahmanism eventhoughtheChandalasare foundpursuingseveral
callingsincluding thoseofa corpsecarrier, remover andcremator, hangman,
sweeper, nightguard, publicperformer andhunter. TheNesadaswhoengaged
inhunting and fishing(thereareoccasionalreferences to theirvillages,chiefs
and guilds),thePukkusaswho werehuntersand refuse-cleaners, theVenas
who were bamboo-workers and also possibly hunters, and the Rathakaras
who builtchariotsforwar,areportrayed as sharingthestatusand penuryof
the Chandalas.It is significant thatthe professions of the basket-maker,
leather-worker, potter,weaverandbarberhada better ratingincontemporary
Brahmanicalworks;the Gribyasutras of Baudhayana38 and Bharadvaja39
prescribedeven upanayanaforthe Rathakara;and _hecaste of bamboo-
workerand basket-maker, Buruda,cameto be lookeduponas untouchable
onlyintheearlymedievalSmritis.40 The emphasisthatbothBuddhismand
Jainism laidon non-violence evidently colouredtheirattitude and perception
towardsmanyofthesecasteswhichwereobligedbytheirwretched material
condition and needs of subsistence and livelihood to engage in activities
involving violence.In thecase oftheChandalastheprimitive dreadof dead
humanbodieswhichtheyhandledaggravated thedegreeofpollution.
The basicfactisthatcastehadcometo playan important roleineconomic
lifeanditdevelopedstrong rootsinoccupationsinthebirth-based hierarchical
class societyof earlyIndia and fewmovements of protestagainstinjustice
involvedinitsintrinsic inequality had powertochangethisgroundreality or
tooffer anyviableandenduring alternative.ThatBrahmanism, Buddhism and
Jainismideologically contributed to theirreversible processof thedevelop-
mentofcasteanduntouchability insteadofresisting it tobeacknowledged.
has
26 SOCIAL SCIENTST
Casteactuallysolidified withthehardening ofclass relationsin northIndia

between600 BC and AD 200, and untouchability, too, originating in pre-
Mauryantimes,41 gotaccentuated byAD 200. Sincemostofthecasteswhich
wereinitiallyreducedtothelevelofuntouchables werethosewhichhad little
shareinthedistribution ofwealth,powerorprestige, untouchabilityhasto be
viewedas theextreme manifestationoftheinstitutionalized
inequalityofboth
casteand class.The untouchables wereand remaina partoftheIndiancaste
and class structure.
Therehas beena livelydebateas to whether theuntouchables are partof
thefour-varna structure or theyconstitute
thefifthvarna.The realityabout
theirbeingdistinct fromtheshudrais incontrovertible. Withoutsubscribing
to thefive-varna theoryKautilyaexplicitly theChandalasfrom
distinguishes
theshudras.42 It is, however,equallytruethatalmosttheentirecorpusof
Brahmanical literature inearlyIndiahas vehemently deniedtheexistenceof
thefifthvarna.LiketheDharmasutras, Manu is categoricalin thisregard:
brabmanabkshatriyo varnadvijatayah
vaishyastrayo
chaturtha
ekajatistushudronastitupanchamah.43
(The brahmana,thekshatriyaand thevaishyaare thethreetwice-born
varnas;theonce-born
shudrais thefourth.
Thereis no fifth
[varna].
So istheAnushasana ParvaoftheMahabbarata:panchamonadhigamyate44
(thereis no fifth
[varnal.The onlywell-known exceptionis Shankaracharya
who inhiscommentary on theBrabmasutra45intheeighthor ninthcentury
describes theNishadasas thefifth (varna):nishadapanchama parigrihitah.
The
Samba Puranadatedbetweenthesixthand eighthcenturiesalso speaksof
panchamamsavarnikam,46whichmaybetranslated as: "varnacategoriesare
five".Theconfusion abouttheuntouchables forming partoftheshudravarna
or beingnon-shudras is reflected inthedata collectedbythePeopleofIndia
projectoftheAnthropological SurveyofIndiaduring1985-92 whichshows
that70.2 percentoftheScheduledCastesperceive themselvesas beingshudras
andonlya fewScheduledCastesinAndhraPradeshcallthemselves Panchama,
a termliterallymeaningthefifth varna.4'Asregardstheuntouchablesbeing
castelessinearlyIndiaor evennow thereis no evidence.Alberunifoundthe
untouchablesin northIndiaduringthetwelfth century dividedintotwelve
castesof unequalstatus48 and a carefulappraisalof theavailablematerial
showsthenumberofsuchcastesbytheendoftheearlymedievalperiodto be
aroundtwenty.49 According to the 1991 Census,thenumberof Scheduled
Castes,mostof them-thoughnot all-erstwhileuntouchables,in all the
Statesand UnionTerritories is 1,091and theirpopulationexcluding Jammu
and KashmirwherethisCensuscould not be heldis 13,82,20,000,thatis,
16.48 per centof the total population.This steadyswellingof numbers
underlines theelement ofcontinuity withchangeincasteand untouchability
(untouchabilityhas beenconstitutionally and legallyabolished)inthecourse
ofourlonghistory and also showshowearlyIndiais stillverymuchwithus.
Withabout40 percentoftheIndianpopulationstillbelowthepoverty line
accordingto therevisedPlanningCommission estimates,onlyfourteen paise
and SocialJustice 27
ina rupeespenton development reaching thetargeted sections,liberalization
andglobalization oftheeconomytending to accentuatetheexisting widegap
betweentheaffluent and theindigent, and thecontinuing maledomination
overwomen,Indiaislikely toremainundertheshacklesofbothcasteandclass
in foreseeable futureand the problemof social justicewill continueto be
important and daunting-almostintractable.
So faras earlyIndiaisconcerned, theexpansionofcasteanduntouchability
fromAD 200 to AD1200 was an uninterrupted andcontinuousprocess.New
groups,indigenousas well as alien,wereabsorbedat variouslevelsof the
socialstructure in differentpartsand ideologyprovedflexibleand receptive
enoughto effectively cope withdevelopingsituationsand historicalcross-
currents. The extentof social mobility, bothupwardand downward,was
certainly remarkable, andthesystem, insteadofcrackingup underpressures
all-round, gotreinforced andfirmly entrenched. Bytheearlymedievalperiod
the untouchablescomprisedtwo broad segments.The firstincludedthe
backwardtribalcommunities whoseoriginaloccupationsand thoseacquired
afterintegration withthemainstream werenotenoughtogivethemeconomic
security.Virtually landless,theyservedas thecheapest-source ofservilelabour
inthecountryside. Withprogressive improvement intheoverallconditionof
theshudrasand theirdisplacingthevaishyasas thebulkofpeasantryalong
withotherlandedclasses,theinterests oftheshudrasand theuntouchables
oftendiverged.The secondcategoryof untouchablesincludedseveralde-
pressedartisancasteswhohad remained at thelevelofshudrasuntilthesixth
century ADbutundertheimpactofcertainfeudaldevelopments inpost-Gupta
timessuch as the declineof tradeand commerce,decreasein commodity
production, break-up ofcraftguildsandincreasing immobility becameclosely
integrated with the in
villages a predominantly agrarianeconomydominated
bythelandedclasses.50Withtheirhereditary skilland low wages,oftenin
kind,theywereanothercheap and easilyavailablesourceof exploitation.
Neitherofthesecategoriesof untouchables was a homogeneous unit.Their
divisionintoseveralhierarchical castesundertheinfluence of thedominant
casteideologythwarted thedevelopment ofanybondofunityamongthemor
withthelandlessshudrasand debilitated themforanyfight forsocialjustice
ontheirown.Parashara,whocomposedhisSmriti duringtheperiod,reflected
thissocialchangeclearlywhenheprescribed a number ofexpiations especially
fortheshudrasin case of theirbeingpollutedbytheuntouchables.51 The
availabilityofcheap landlesslabourand artisanalproductson an enduring
basishas beenunderlined byIrfanHabib as themajorcontribution ofcaste
fromwhichtheMuslimlandedand rulingclassesbenefited as muchas their
Hinducounterparts. 52
In factsocialjusticeis byitsnaturecloselylinkedwithequalityat thelevel
ofmundaneexistence andthiswas boundtoremainelusivewithoutsomesort
oftangibleequalityand justiceat theeconomiclevel.Thisrequireda kindof
socio-economic transformation whichwas outsidetheagendaofthereligious
movementsin earlymedievalIndia such as some formsof Tantricism,
SahajayanaoftheSiddhas,theLingayatasect,etc.,whoseprotest againstcaste
inequity was virulent andreflected theanguishoftheoppressed, and inwhich
28 SOCIAL SCIENTIST
theuntouchables and lowercastesthemselves playeda significant

rolealong
withconscience-stricken brahmanaslikeSarahapada,Basavesvaraand oth-
ers.53
The liberalattitude displayedbythesereligious movements towardsthe
sociallydepressedsegments, however,definitelybrought much-needed relief
intheformofequalityatthereligious andspirituallevelwhichhadbeendenied
to thembyorthodoxBrahmanism. As regardstheliberalmovements within
Brahmanism itselfsuch as Vaishnavism, Shaivism,etc.,whichemphasized
equalitythrough bhakti(devotionto God),theycouldhardlyfarebetter. The
Bhagavadgita,composed around the third-second centuryBC, definitely
adopteda stancewhichwas firmly and unequivocally opposedto thatofthe
Dharmasutras and theManusmriti on theissueofuntouchability.Recogniz-
ingtheformidable andinescapablerealityofthehierarchicaland inegalitarian
caste structure whichcould not be uprootedor abolishedwholesale,and
seekingto reformthe spiritof its workingfromwithin,it presentedan
idealizedversionof varna54and declareditselfin favourof eradicationof
untouchability.55 Thediscerning Dharmashastra authorssensedtheominous
nature of thework and itsmessageand forcenturies ignoredboth.Thiswas
thelegacyofGandhi,whodrewhisinspiration fromtheGitaandwhoselong
and arduouscampaignto eradicateuntouchability and get justiceforthe
untouchables fromwithinthesystem couldnotbe ignoredand was therefore
firmly opposedbythevestedinterests and theSanataniHindus.Timeshave
changedwithself-assertion by theDalits themselves,56but theagenda for
social justicewill makeheadwayonlyifit goes hand in hand witha well-
consideredand adequateagendaofeconomictransformation and economic
justicewithfocuson theimpoverished massesandwilltohonestly implement it.
NOTES
1. Fordetails,see myPresidential Address,AncientIndiasection,"SocialStratification in
AncientIndia: Some Reflections", Proceedingsof the IndianHistoryCongress,S1st
session(University ofCalcutta,1990),pp. 23-26.
2. Cf.SuviraJaiswal, "Stratification
inRgvedicSociety: Evidence andParadigms", TheIndian
HistoricalReview (hereafter
IHR), Vol.XVI,Nos. 1-2 (July1989andJanuary 1990),pp.
18-19, 22.
3. brahmano'sya mukhamasidbabu rajanyahkritahurutadasyayadvaishyah padbbyam
shudroajayata,90.12.
4. Rajanyahasthesenseofa closekinsman ofrajan;vaishyaisderivedfromvish;andshudra
mayoriginally havebeena conqueredtribeofthatnamewhichoccursthrice inthissense
intheearliestportionoftheAtharvaveda (IV. 20.4; IV.20.8; V.11.3).
S. VII.29. The brahmana's material dependence on the king is indicated by
yathakamaprayapyah (one who can be removedat will) appliedto himin thistext,
XXXV.3.
6. Sayana'sinterpretationofvadhyab as kupitenasvaminatadyo bhavatiicchamanatikranya,
meaning'an angrymastercan beattheshudraifhiswillhas beentransgressed' seems
appropriate here.The Nirukta,too,translates vadhaas 'to kill'as wellas 'to hurt'.
7. agnimcbitvana ramamupeyat; ramaturamanayopeyate na dhatmayakrisbnajatiya, XII.
13. P.V. Kane datesYaska between800 Bcand 500 sc. (HistoryofDharmasastra, Vol.
II,Pt1,2ndedn.,Bhandarkar OrientalResearchInstitute, Poona,1974,p. xi). Yaskamay
haveflourished in theseventh century Bc.
8. Ashtadbyayi, V. 4.9. TheKashikaVritti, commentary onPanini'ssutrasbyVamanaand
Jayaditya (earlyseventhcentury), citesas examplesbrahmanajatiyah, kshatriyajatiyah
and SocialJustice 29
and vaishyajatiyah. Unlikethebrahmana,brabmadoes not have the senseof caste,

maintainsPanini(brahmo'jatau, VI. 4.171). V.S. AgrawalaassignsPaninito the fifth
century Bc., India as Knownto Panini,2nd revisededn. (BanarasHindu University,
Varanasi,1963),pp. 476-78.
9. V.10.7. The crudeequationoftheChandalawithanimalsis striking.
10. VII. 18,Vishvamitra is statedto haveadoptedShunahshepa as hissonand givenhimthe
firstrankamonghishundredsonswiththerightofprimogeniture; thefiftyoldersons
refused to acceptthisand incurred hiswrath.
11. VIII. 38.3; SuviraJaiswal, "VarnaIdeologyandSocialChange",SocialScientist, 214-15,
Vol. 19, Nos. 3-4, (March-April 1991),pp. 44, 48.
12. Vivekanand Jha,"FromTribetoUntouchable: TheCase ofNisadas"inR.S. Sharmaand
Vivekanand Jha,ed.,IndianSociety:HistoricalProbings(In MemoryofD.D. Kosambi)
(People'sPublishing House,New Delhi,1974),pp. 67-71.
13. XXX. 17.
14. III. 4.1.14.
15. IV.3.22.
16. VI. 20; X.1; X.50; XI. 20; XI. 29; XII.1; XVII. 1; XXI.4.
17. 1.1.1.16;1.4.7.6;11.2.3.29;11.3.5.4;11.4.7.3;II. 10.18.14.
18. 1.17;11.2;III.5; XIX.7.
19. II.1.2.3; II.3.6.1; II.-5.
11.1I0;II.5.11.1I1.
20. varnasamvarge, VasishthaDharmasutra,111.24;varnanamsamkare,Baudhayana
Dharmasutra, II. 2.4.18.
21. X.S.
22. 11.22,24-36.
23. 11.2.4.16,18-21.
24. 1.5.10.10.
25. The statusofwomanis notequalto thatofmaninan anulomamarriage, as is presumed
in TheEncyclopaedia Britannica, Vol. V (1973 reprint),
p. 25, no doesanUlomameana
woman'ssexualalliancewitha lowerrankingpartner, as is maintainedby The New
EncyclopaediaBritannica, Vol. II, 15thedn. (1985), p. 930.
26. Fordetails,-seemyArticle" Varnasamkara in theDharmasutras: Theoryand Practice",
Journalofthe Economican:dSocialHistory oftheOrient,Vol.XIII,Pt.III (Leiden,1970),
pp. 277-80.
27. 1.1.2.13.
28. WhileBaudhayanauses vratyain the senseof a faithful observerof the prescribed
Dharmashastra norms(111.3.7.13-15), Baudhayanaappliesthetermto thesons of an
uninitiated(avrata)twice-born
whoareexdudedfrom savitri(savitribbrashtan)(1.8.16.16).
29. varnasamkaradutpannanvratyanahurmanishinah, 1.9.17.15.
30. BaudhayanaShrautasutra, XXIV 31; ShankhayanaShrautasutra, 11.8.3;Apastamba
Shrautasutra, VI.3.12.
31. ShatapathaBrabmana,111.1. 1.10; ApastambaShrautasutra, XV.20.16.
32. Panchavimsha Brahmana, VI.1.1 1;KumkumRoy,TheEmergence ofMonarchy inNorth
India:Eighth-Fourth CenturiesBC. (OUP, Delhi,1994),p. 232.
33. Vivekanand Jha,"ChandalasandtheOriginofUntouchability", IHR,Vol. XIII,Nos. 1-
2 Uuly1986 andJanuary1987),p. 34.
34. X. 31. TheAnushasanaParvaoftheMahabbarata,CriticalEdn.,48.18, has theexpres-
sionhinahinatprasuyante varnabpanchadeshaiva tu (theselow varnasproduceon the
low fifteenjatis).
35. S.J.Tambiah,"FromVarna to Caste throughMixed Unions" in Jack Goody,ed.,
Character of Kinship(Cambridge, 1973),pp. 204-5.
36. N.K. Bose,CultureandSociety inIndia(AsiaPublishing House,Bombay,1967),pp. 207-
10.
37. Uma Chakravarti,
TheSocial DimensionsofEarlyBuddbism(OUP, Delhi,1987), pp.
122-49.
38. 11.5.6.
39. I.1.
30 SOCIAL SCIENTIST
40. Atrismriti, verse196; Angirassmriti, verse3; and Yamasmriti, verse33.

41. Panini'sreference to shudranamaniravasitanam (11.4.10),unsegregated or notexcluded
shudras,implying theexistence ofthecategory ofsegregated shudrasor untouchables, is
definitivein thisregard;cf. R.S. Sharma,Sudrasin AncientIndia, 3rd edn. (Motilal
Banarsidass,Delhi,1990),p. 324.
42.. shudrasadharmanova,anyantrachandalebbyab, Arthashastra, III.7.37.
43. X.4.
44. 47.18.
45. 1.4.12.
46. 66.10.
47. K.S. Singh,"ManuandContemporary IndianEthnography" inD.N.Jha,ed.,Societyand
IdeologyinIndia:EssaysinHonourofProfessor R.S. Sharma(Munshiram Manoharlal,
New Delhi,1996),p. 433.
48. EdwardC. Sachau,ed. and tr.,Alberuni'sIndia,Vol. I (London,1910), p. 101.
49. VivekanandJha,"StagesintheHistoryofUntouchables",IHR,Vol. II,No. 1 (July1975),
pp.28-31. Thepositionhasbeendiscussedindetailinmyunpublished Ph.D.thesis"Early
HistoryofUntouchables in India",PatnaUniversity, 1972.
50. R.S.Sharma,SocialChangesinEarlyMedieval India(circaAD 500-1200), TheFirstDevraj
ChananaMemorialLecture, 1969 (People'sPublishing House,New Delhi,1969),pp.6-
9; idem,IndianFeudalism, c.AD 300-1200,2ndedn.(Macmillan, Madras,1980),pp.43-
48, 53-55, 187-90,220-22; idem,UrbanDecay inIndia(c. 300-c. 1000) (Munshiram
Manoharlal,NewDelhi,1987),pp. 135-36,154-55,158,165-67,177,182,184; B.N.S.
Yadava, Societyand Culturein Northern India in theTwelfthCentury (CentralBook
Depot, Allahabad,1973), pp. 141, 147, 163-72, 267-75; idem, "Immobility and
SubjectionofIndianPeasantry in EarlyMedievalComplex",IHR, Vol. I, No. 1 (March
1974),pp. 18,20,22-24; idem,"TheAccountsoftheKali AgeandtheSocialTransition
fromAntiquityto theMiddleAges",IHR,Vol.V,Nos. 1-2 July1978andJanuary 1979),
pp. 31,46,50,54,61.
51. VI. 28-29, 36, 44-45; X.8,20; XI.2-3.
52. "Caste in IndianHistory"in Essaysin IndianHistory:TowardsA MarxistPerception
(Tulika,New Delhi,1995),pp. 166, 171-72;idem,"TheoriesofSocialChangeinSouth
Asia", TheJournalof Social Studies,No. 33 (Centreof Social Studies,University of
Dhaka,April1966),pp. 38-39.
53. B.N.S. Yadava findstheirvisionand approach'revolutionary', "ProblemofInteraction
betweenSocio-Economic ClassesintheEarlyMedievalComplex",IHR, Vol. IIi, No. 1
(July1976), pp. 57-58.
54. IV.13; XVIII. 41-44.
55. vidyavinayasampanne brahmanegavihastini
shunichaivashvapakecba panditahsamadarsbinah, V.18.
(Panditasor menof knowledgclook upon a learnedand politebrahmana,a cowy, an
elephant,a dog and a Shvapakaas alike).
cf.Vivekanand Jha,"SocialContentoftheBhagavadgita", IHR, Vol. XI, Nos. 1-2 (July
1984andJanuary 1985),pp.22-29; B.N.S.Yadava,"SocialContentoftheBhagavadgita:
A Note", IHR, Vol. XVIII,Nos. 1-2 (July1991 andJanuary 1992),pp. 200-1.
56. B.R. Ambedkar'ssingularcontribution in thisregardis a matterof recordand duly
acknowledged.
Conceptualising Brahmanical Patriarchy in Early India: Gender, Caste, Class and State
Source: Economic and Political Weekly, Vol. 28, No. 14 (Apr. 3, 1993), pp. 579-585
Published by: Economic and Political Weekly
Accessed: 03-03-2015 07:25 UTC
Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political
Weekly.
SPECIAL ARTICLES
ConceptualisingBralimaniLcalPatriarchy in
Early India
Gender, Caste, Class and State
Uma Chakravarti
Castehierarchyand genderhierarchyare the organisingprinciplesof the brahmanicalsocial orderand are
closely interconnected.This article exploresthe relationshipbetweencaste and gender,focusing on what is
possibly the centralfactorfor the subordinationof the uppercaste woman:the needfor effectivesexual con-
trol oversuch womento maintainnot only patrilinealsuccessionbut also castepurity,the institutionunique
to Hindu society.
STUDIES of women in early Indian history The task of exploring the connections bet- caste can be ensured without closely guar-
have tended to focus on what is broadly ween patriarchy and other structures within ding women who form the pivot for the en-
termed as the 'status of women', which in a historical context was pioneered by Gerda tire structure. As Yalman's informants
turn has led to a concentration of attention Lerner (1986) and her work is both theoreti- pointed out the honour and respectability
on a limited set of questions such as mar- cally and methodologically useful for of men is protectedand preservedthrough
riage law, property rights, and rights relating historians. In outlining the historical process their women. The appearanceof puberty
to religious practices, normally viewed as in- of the creation of patriarchy in the thus marksa profoundly'dangerous'situa-
dices of status. The limited focus has left a Mesopotamian region Lerner describes her tion and is the context for major rituals
major lacuna in our understanding of social growing awareness of the fact that crucial which indicates the important relationship
processes which have shaped men, women, to the organisation of early Mesopotamian between female purity and purity of caste.
and social institutions in early India. It is society was the total control of women's sex- It is in orderto stringentlyguard the purity
now time to move away from questions of uality by men of the dominant class. She had of castesthat veryearlyon pre-pubertymar-
'status' whether high or low, and to look in- been puzzled by her evidence wherein riages were recommended for the upper
stead at the structural framework of gender women seemed to have greatly differing castesespeciallybrahmanas[Yalman: 25-58].
relations, i e, to the nature and basis of the statuses, some holding high positions and Yalmanalso points out that caste blood is
subordination of women and its extent and enjoying economic independence but whose always bilateral, i e, its ritualiquality is
specific form in early Indian society. In this sexuality was controlled by men. This led her received from both parents. Thus ideally
context we may point out that although the to recognise that there was a need to look both parents must be of the same caste.
subordination of women is a common beyond economic questions and focus on the However,this cannot alwaysbe ensuredand
feature of almost all stages of history, and control over women's sexuality and the man- is the basis of grave anxiety in the texts.
s prevalent in large parts of the world, the ner in which reproduction was organised and The anxiety about polluting the ritual
xtent and form of that subordination has thus to look for the causes and effects of orderand the quality of the blood through
een conditioned by the social and cultural such sexual control [Lerner 1986: 8]. A women is best demonstratedin the horror
environment in which women have been similar exploration of the process of of miscegenyas we shall show. In the theore-
placed. establishing control over women's sexuality tical explanations for the proliferationof
The general subordination of women in a highly stratified and closed structure caste the most polluting and low castes are
assumed a particularly severe form in India could be useful in analysing the connections attributedto miscegeny,i e, the mixing of
through the powerful instrument of religious between caste, class, patriarchy, and the state castes ('varnasamkara').Most polluting are
traditions which have shaped social prac- in the brahmanical texts of early India. The those castes which are the products of
tices. A marked feature of Hindu society is structure that came into being has shaped reprehensibleunions between women of a
its legal sanction for an extreme expression ghe ideology of the upper castes and con- highercaste and men of a lower caste. The
of social stratification in which women and tinues to be the underpinning of beliefs and ideologues of the caste system had a p.r-
the lower castes have been subjected to practices extant todav. ticular horror of hypogamy-pratiloma-or
humiliating conditions of existence. Caste A possible starting point for an explora- against the grain as it was described-and
hierarchy and gender hierarchy are the tion of the historical evidence on the crucial reservedfor it the severestcondemnationand
organising principles of the brahmanical place of' control over women's sexuality the highest punishmentas will be evident.
social order and despite their close intercon- within the larger structure in which Violations continued to be punished until
nections neither scholars of the caste system brahmanical patriarchy was located thus recenttimes by drowningmother and child
nor feminist scholars have attempted to could be the practices and beliefs prevalent (Yalman: 52] and excommunication and
analyse the relationship between the two. I among the upper castes as studied by anthro- ritual death.
will explore here (very tentatively) the rela- pologists. An insightful essay by Nur Yalman The safeguardingof the caste structureis
tionship between caste and gender, focusing (1962) on the castes of Ceylon and Malabar' achievedthroughthe highlyrestrictedmove-
on what is possibly the central factor for the shows that the sexuality of women, more ment of women or even through female
subordination of the upper caste women: the than that of men, is the subject of social seclusion. Women are regarded as gate-
need for effective sexual control over such concern. Yalman argues that a fundamental ways-literally points of entrance into the
women to maintain not only patrilineal suc- principle of Hindu social organisation is to caste system. The lower caste male whose
cession (a requirement of all patriarchal construct a closed structure to preserve land, sexualityis a threatto-uppercaste purityhas
societies) but also caste purity, the institu- women, and 'ritual quality within it. The to be institutionallypreventedfrom having
tion unique to Hindu society. The purity of three are structurally linked and it is impossi- sexualaccess to women of the highercastes
women has a centrality in brahmanical ble to maintain all three without stringently so women must be carefully guarded
patriarchy, as we shall see, because the purity organising female sexuality. Indeed neither [Ganesh 1985:16;Das 1976:129-45].When
of caste is contingent upon it. land, nor ritual quality, i e, the purity of the structureto preventmiscegeny breaks
Economic and Political Weekly April 3, 1993
down the brahmanical texts consider that the Evidence from the cave paintings in cen- pa in any definitive manner on any aspect
whole elaborate edifice of social order that tral India thus suggests that in the hunting- includingon questions of gender. However
they built up has collapsed. The Kaliyuga gathering stage there was no rigid sexual the existenceof numerousmother goddess
of the future is just such a time when women division of labour as has sometimes been icons and the bronze statue of the dancing
of the high castes and men of the low castes postulated, i e, men hunt and women gather. girl could be interpretedas the continuedim-
will r*gress from their duties. The Bhagavad In the case of central India in the mesolithic portance of women's special relationship
Gita, the normative text par excellence of the period, it is likely that women participated with reproduction,and may also be seen as
Hindus, outlines the collapse of the social in the hunt apart from the all important task an acceptance of their sexuality. The
and moral order when there are leakages in of gathefing which in any case accounted for evidence is not enough to indicate whether
the closed structure of marriages. Families the major source of food in tropical climates. the sexuality was already under some kind
are broken, rites are forgotten, women are The role of women in the economy was thus of control, whether by men or by certain
defiled and from this corruption comes the equal if not more than that of men. Based categoriesof women. Better interpretation
mixing of castes [Gita 1: 41-44]. Thus whilc on modern anthropological data on tribal and analysisof evidencefrom Mesopotamia
advocating conformity all the detailing of societies it has been postulated that the most is possibleas the numerousclay inscriptions
norms for women in the brahmanical texts egalitarian societies are to be found among have been deciphered.Lerner'sstimulating
are a powerful admission of the power of hunting-gathering tribes which are charac- study of the creation of patriarchywould
non-conformist women, or all women who terised by interdependency [Lerner 1986: 29]. suggest that some form of community or
have the power to non-conform, to break the The relative status of men and women can clan control over women and their sexuali-
entire structure of Hindu orthodoxy. For, at the most be characterised as 'separate but ty wereaspectsof social organisationin the
when women are corrupted all is lost. In the equal'. archaic state and may have existed in the
brahmanical texts it is evident that the up- What is of major significance to this essay Harappanculture too.
per caste woman is the object of moral is that the important role of women in the In contrast to the Harappan culture the
panic. Through the recalcitrance of women hunting-gathering economy, which was Rig Vedicperiodis characterisedby the lack
the established property and status order can highly valued, was enhanced by the impor- of informationon materialculturein general
be subverted. To prevent such a contingen- tance attached to the reproductive role of but particularlyon anything that may have
cy women's sexual subordination was institu- women. Pregnant women, women in their had a bearingon women.The Rig Vedaitself
tionalised in the brahmanical law codes and nurturing roles as mothers, and women por- however does throw some light on the
enforced by the power of the state. At the trayed in the act of childbirth are sometimes ideology of the early Aryans. Rig Vedic
same time women's co-operatiodl in the depicted in the paintings and the last has Society witnesseda continuingstrugglebet-
system was secured by various means: been identified as the figure of a mother ween the Aryansand the 'indigenous'tribes
ideology, economic dependency on the male goddess. Similar evidence from other pre- who wereviewedwith particularhostilityby
head of the family, class privileges and historic cultures in the Mesopotamian region the Aryans for their dark skins, and their
veneration bestowed upon conforming and has been used to suggest the prevalence of racial'inferiority'As the Aryanssucceeded
dependent women of the upper classes, and a pervasive veneration of the mother god- in establishing their control over certain
finally the use of force when required. dess. It has also been argued that the first areas most of the men either fled or were
form of religious expression for men and killed; the conquerors then enslaved the
I women is the psychological bond between women of the subjugatedpeoples. Thus the
mother and child, and that the 'life giving first large group to be enslaved in early
The process of caste, class and gender mother' appeared to have power over 'life Indianhistorywerewomenas therearemore
stratification, the three elements in the and death'; thus men and women, observing frequentreferencesto 'dasis' than to 'dasas'
establishment of the social order in India this dramatic and mysterious power of the [Chakravarti1985:561;the evidence of thle
shaping the formation of brahmanical female turned to the veneration of the Rig Veda is in consonance with Lerner's
patriarchy,' took a considerable period of mother goddess [Lerner 1986: 39]. argument that all early conquering tribes
time to evolve into its complex structure. killed the defeated men and enslaved the
Going by existing archaeological studies, Female reproductive power in such a women,at leastin the firststage of conquest
hunting-gathering society is regarded as
which do not lend themselves easily to ques-
valuable because the very survival of the
[Lerner1986:78ff]. Forour purposethe Rig
tions of stratification, none of the elements Vedicevidenceis extremelysignificant as it
of stratification outlined above can be clear- community is dependent upon it. Prehistoric reflects an essential stratification within
ly traced in the evidence available to us. paintings at Kathotia, Bhimbetka and Khar- women, betweenwornenof the conquering
There are, however, some indications that in wai treat female sexuality as one aspect of tribesand womenof the subjugatedpeople.
prehistoric cultures women's role in produc- female existence. Thus women as Their roles and their place in society were
tion and in reproduction was regarded as reproducers are as evident as women's pro- very different. The Rig Veda for example
valuable. In a recent study of cave paintings ductive activities in the hunting-gathering describesthe Aryan women as ruling over
at Bhimbetka (circa 5000BC) it has been economy [Roy 1987: 71. Society in this phase bipeds and quadrupeds,i e, slaves and cat-
argued that women were engaged in gather- has been characterised by one scholar as tle [Rig Veda:IX 85.431.While the dasis' or
ing fruit and other wild produce and in hun- 'matristic' one in which women were not the enslavedwomen'slabour and sexuality
ting small game using baskets and small subjected to the authority of men, or of wereto be used, this was under the overall
nets. They combined their roles as mothers other women [Neumayer 1983: 211. There control of the men of the conqueringclans.
with their activities as gatherers during this would be little need in such a society for the Referencesto dasasas objectof 'dana'(gifts)
hunting-gathering stage of society. The pain- sexual'control of women by men. makeit evidentthat the recipientsarealways
tings include those of a woman with a basket Evidence from the Harappan civilisation men; often the rajanya,as the captors, gift
slung across her shoulders with two children has not been analysed from the gender point them to priests. The possession of women
in it and she also carries an animal on her of view but there is some indication of the slaves was clearly a major element in the
head; women carrying baskets and nets otten emergence of social stratification, with a primitive accumulation of wealth.
depicted as pregnant; a woman dragging a class of people who laboured and others Manyof the mythsof the Rig Veda reflect
deer by its antlers; and women engaged in who wielded power and occupied the citadels an explicit relationshipof women with sex-
catching fish [Roy 1987: 3-4]. In group hun- in tbe structures that have been excavated. uality. Frequentlythis is an aspect which is
ting scenes too the paintings include women. An understanding of how this society was specially associatedwith demoniac women
From the elaborate head-dress that they wear organised internally in terms of its economy or with apsaras.Whiledemoniacwomenare
it is possible to argue that their presence in and polity is still inconclusive as the ar- a threatto men and to their rituals,the ap-
the hunt might indicate both a symbolic and chaeological data is not yet complemented saras are free from male control and even
an actual participationin ensuringthe suc- by writtenevidence. It is thus not possible set stringent conditions for any long-term
cess of the hunt. to use the evidenceavailableto us on Harap- cohabitation with men. For other Aryan
580 Economic and Political Weekly April 3, 1993
women, the patriarchal family had establish- (i e, uy rc. '.-int, women powerless, by woman, a Sudra, a dog, and a crow are the
ed a certain degree of control over women. appropriating ai: the sources of their embodimentsof untruth,sin, and darkness
Their position in the pastoral economy, with strength) it appears that kingship or the state [XIV.1.1.31).The view that women's innate
the household playing an important part in was already associated with the control of nature was lascivious and evil was so per-
production, ensured the recognition of their women and was an instrument through vasive that it features even in Buddhist
presence in society especially in the perfor- which their subjugation was achieved. literature.2 A Jataka story states that
mance of rituals. But the custom of Niyoga women are a sex composed of wickedness
which was the privilege of affinal male II and guile; womankind holds truth for
kinsmen indicates that control over female falsehoodand falsehood for truth.They are
The shift to an agricultural economy and
sexualitywas firmly established. Niyoga
the second urbanisation (800BC- 600BC) u*able as the sand, and as cruel as the
combines the utilisation of the reproductive snake (Jalaka 1:551.Says -another Jataka
was marked by the emergence of caste and story, "Wrathfulare women, slanderousin-
potential of women but under rules laid
down by men to- further cultural norms class divisions. The brahmana was a foirve grates,the sowersof dissensions and strife."
to reckon with and patrilineal succession wis Their passions are insatiab!eas they act ac-
which privilege them. And it is noteworthy
that while there is no special value attached fairly well established within the larger con- cordingio theirinbornnature(Jataka1:309].
text of a defined family structure distinct Even the Ramrayanaassociates most
to chastity, the example of the maiden who
from the earlier structure. Some of these women with being essentiallyweak and sin-,
abandoned her child (indicating definite
elements are captured in-the Buddhist origin ful. Accordingto Kausalyawomen do not
notions of legitimate reproduction) reiterates
myth where the institution of caste, private care for a good family, good deeds, or
that patriarchal control over women was in- property, the family, and the archaic state
stitutionalised[Rig VedaIV.19.9;IV.30.16J. are represented as emerging simultaneously wisdom,and theirheartsareeverinconstant
The post-Vedic literature reflects a two- [11.39.236-2401.The sage Agastyastatesthat
from an earlier stage of primitive existence it has beena woman'snatureeversincecrea-
fold development of ideology. While Aryan
[Digha Nikaya Ill 80ff]. Thesq changes, i e, tion began to cling to a man only when he
women were being marginalised in terms of
the emergence of a fairly stratified society prospers,and desert him in difficulty; their
their original roles in the sacrifice their roles
and the collapse of tribal economy and poli- fickle natures are modelled on the flashes
in the productive system were also chang-
ing. The increasing dependence on ty in the post-Vedic period, especially with of lightning [111.3.61.Anasuya complains
the establishment of private control over that normally women do not know right
agriculture as the major source of food
laRd [Chakravarti 1987: 23ff), held and from wrong, and even though they are
shifted the scene of food production outside
transmitted within a patrilineal system, ac- dependenton their husbandsfor protection
the households to the fields; the labour of
the subjugated peoples including dasis was companied by the beginning also of patri- they wanderabout with their heartssubject
extracted to work the land and this enabled lineal succession to kingship, and the preser- only to theirown desires[11.117.261.All these
to be rest'ricted vation of caste purity meant that the sexual examplesareused by Tryambaka,the author
the Aryan woman's labour
behaviour of certain categories of women of the S:ridharmapaddhatito stress the in-
to the household. Thereafter the participa-
needed to be closely guarded. Wives in par- natewickednessof womenin a generalsense,
tion of a certain class of women in 'produc-
ticular required to be under male control and but there are more specific forms of the
tion' that was valued ceased. Such women
this view finds explicit mention in a later innate impurity and sinfulness of women
from then onward were associated only with
text, the Apasiamba Dharma Sulr (circa 6th whichcome closerto the-problemof sexuali-
reproduction. Whether these developments
took place with the compliance of Aryan century BC), which rules that a husband ty. According to Tryambaka'sversion of
women or not (the dasis of course would should ensure that no other man goes near Manu's 'Stripumdharma' (written for
his wife lest his seed get into her 111.6.13.7). women in the 18thcentury) women are in-
have had no active part to play in the crea-
tion of such a system), a degree of tension It is at this point that a sharp distinction nately promiscuous,fickle minded, lacking
between men and women may be discerned required to be made between motherhood in love,and unfaithfulto theirhusbandseven
even in the Rig Vedic literature where the and female sexuality with the latter being when closely guarded.One reason for their
relationship between the gods and goddesses channelised only into legitimate motherhood innate impurityis representedas stemming
is often depicted as hostile. There are within a tightly controlled structure of from the fact that women became recipients
references also to suggest that women must reproduction which ensured caste purity (by of the guilt of brahmicide,alongwith the
be rendered powerless by ensuring that they mating only with prescribed partners) and earth and trees, which was shifted upon
do not gain in strength and are obedient to patrilineal succession (by restricting mating them by Indrawhen he killed Vishwarupa
men and follow them [Roy 1987: 23-30]. only with one man). From then on female and they thus became impure [Leslie:251].
The need for monitoring women's sexuali- sexuality had to be 'managed' and therefore Menstruation,accordingto this myth, was
ty is also evident. It appears that women's a crucial question for us to pursue is "in associated with women's participation in
sexuality is viewed as a threat, particularly whose hands does the management of brahminmurder.It is a markof a woman's
in relation to the sacrifice. Thus Dirghajivi, female sexuality come to reside; further do innateimpurityand at the same time her in-
a demoness whose sexual appetite is women participate in this process of mana- nate sexuality [O'Flaherty 1976: 153ff].
represented as ghoulish, is described as be- gement?" The congenital fickleness of women's
ing tamed by a handsome man Sumitra who This was the general context in which natureis specially pertinentto the problem
thus neutralises the danger that she presents women's 'essential nature' came to be iden- of dealingwith the innatelyoverflowingand
to the sacrifice [Roy 31; O'Flaherty 1984: tified with their sexuality although it was not uncontrollablesexualityof women. Thus in
101-031. The earliest references to the need directly or explicitly associated as such. At the ancient texts it is repeatedlystated that
to specially guard wives is also evident dur- a general level the innate nature of women they can never be trusted; further the
ing this period. The Satapaiha Brahmana was represented as sinful. According to one Mahabharatastatesthat they are difficultto
,expresses the fear that the wife might go to text, women have been sinful right from the control. The cunning tricks of the demons
other men [SBI 3.1.21J. Most significantly beginning when the creator first made the are known to be unique to women
there is a very embryonic notion of ultimate five gross elements, the three worlds, and he [XIII.39.51.In another text they are linked
cont'rol over women's sexual behaviour be- gave shape to men and women [Leslie 1989: to kings and creepingvines in that they will
ing asserted by the king. The Salapatha 2481. Women are the edge of a razor, poison, embracewhateveris beside them. They are
Brahmana 1.15.20] states that the divine 'ra- snakes, and fire all rolled into one [Leslie adulterousby nature and are permanently
ja' Varuna seizes the woman who has 1989). At the time of creation the original on the look-outfor an opportunityto seduce
adulterous intercourse with men other than Manu allocated to women the habit of lying, men:accordingto a Jatakastory "Asgreedy
her husband. Read along with another state- sitting around and an indiscriminate love of cows seek pasture a new, women unsated
ment in the same text [XI.4.3.1ffl which ornaments, anger, meanness, treachery, and yearn for mate on mate" [Jataka 1:1551.
alludes to kingdom being obtained by bad conduct [Manu, IX 17).As early as the The notion ;that the essential nature of
deprivingthe goddessSri of all her qualities Sa/apaiha Brahmona we are told that a women is vestcd in their sexuality is dealt

581
with most explicitlyby Manu, the most pro- the dest-ructiveand demoniaclust of women adyantageof her husband'sabsenceto carry
minent ideologue of the brahmanical which is consideredto be their 'true'nature. on with all manner of men. Unfortunately
system. After ruling that women must be A femak ascetlc to whom Astavakrais sent for her two parrots, who are like the sons
closely guardedday and night, regardlessof in preparationfor marriage repeatedlyat- of the brahmana,have been left behind to
theirage, Manutells us why it is that women temptsto seducehim in spite of her advanc- keepwatchand reporton her so her miscon-
must be guarded.Buildingup fromthe need ed age. She tells Astavakrathat for women duct is communicatedto the brahmanaon
to guardagainsteven the most trifling 'evil' thereis no greaterdelightnor moredestruc- his return. Between themselvesthe parrots
actionsof womenManuarguesthat by care- tive urgethan sex, that evenveryold women observe that one might "carry a woman
fully guardingthe wife (the most important are consumed by sexual passion and that about in one's arms and yet she would not
categoryof women as far as the brahmani women'ssexualdesirecan neverbe overcome be safe".The elder of the two parrotsthen
ido1Ques wereconcerned)a man preserves in all the three worlds (Mahabhara:a points out that only "wifely love can curb
the pority of his offspring, his family, X111.20.59-60;64-67; 22-29; Leslie 1989: a woman'slust" and it was wifely love that
himselt, and his meahs of acquiring merit 2681.The AsatamantaJataka reiteratesthe was lacking in the case of the brahmana's
[IX.71.Developinghis argumentManu tells same messagethat even an old woman is a wife [Jaiaka 1. 309.
us that after conception by his wife, the sexual hazard [Jataka 1. no 611. The representationof an inordinate.sex-
husband becomes an embryo and is born This projectionof the fearof women'sun- uality in the case of women of the ruling
again of her;accordingto Manu that is the controlledsexualitywas the backdropto the clans, landholdinggroups, and the priestly
wifehoodof a wife [IX.7-91.In orderto keep obsession with creatingan effective system classessuggeststhat thesecategoriesare par-
his offspring 'pure'Manu enjoins the hus- of control and the need to guard them con- ticularlyconcerned with 'impulse'control.
band to carefully guard his wife lest his stantly; the moment the controls are relax- While legitimacyin termsof succession ex-
futureis deniedto him. It is women'snature ed, or cannot be effectively mounted, plains the referencesto womenof the king's
which requiresthem to be so thoroughly women'sinordinatesexualappetitewill lead familyand the landholdinggroupsthe need
restrained.Accordingto Manu their essen- them to adulterousliaisons. to maintaincaste purityexplains the obses-
tial nature will drive women into seeking A strikingaspectof the obsessiveneed for siori with brahmanawives.
satisfaction anywhere, anytimne,and with control over women in the narrativelite- An interestingfacet of women's 'innate'
anyone. He states that ratureof the Buddhistsis that it has a close nature ('strisvabhava')unlike the innate
Women do not care for beauty, nor is their link with women of the upper strata,parti- naturesof other subordinategroupslike the
attention fixed on age; thinking it is euiough cularly with the wives of kings and brah- sudraswas the representation of conflictbet-
that he is a man, they give themselves to the manas and occasionally with 'gahapatis' weenthe inherentnatureof womenand their
handsome and to the ugly. who wereamong the dominant sections of dharma. While the 'innate' nature of the
Through their passion for men, through society and were closely associated with lowercastes thAtof renderingserviceto the
their mutable temper, through their natural land. twice-born,was in harmonywith the dhar-
heartlessness, they become disloyal towards In the BandanamokkhaJatakathe king's ma prescribedfor them by the brahmanical
their husbands, however carefully they may wife lays strict conditions of fidelity upon law-givers,strisvabhava,women'sessential
be guarded [Manu IX.15]. her husband but herself displays uncon- natureas sexualbeings, was in conflict with
The most revealing statement that Manu trollable lust when the king is away at the their stridharmaof fidelity to the husband:
makes in the context of women's essential frontierFightingto put down disorder.Her their strisvabhavawas constantly enticing
nature points out: extraordinary appetite leads her to seek them away from their stridhanna. Signi-
Knowing their disposition, which the lord of satisfaction with a series of messengers,64 ficantly some myths explicitly suggest that
creatures laid on them at creation (i e, their in all, who come to enquireabout how she a 'demoniac'strisvabhavawas the maternal
reproductive power, their sexuality, their is faring in the king's absence Finally she heritageof women whereasthe stridharma,
essential nature) every man should most attempts to seduce the royal chaplain who the duty of women was their paternal
strenuously exert himself to guard them refutesher advances.When the king returns heritage, given to them by the brahmana
iManu IX.16J. she accuses the chaplain of having attemp- priests [Leslie 1989:266J.These references
The crucial place occupied by the wife in the ted to seduce her and of beating her when suggest that, the original attitude of
whole system of perpetuating the social she resistshis advances.The kingordersthat prehistoric societies to the reproductive
order and in enabling men to gain immor- the chaplain be beheaded, whereupon the powerof women, wheretheir sexualitywas
tality through their sons is explicitly ar- chaplain tells the king the truth and at the acceptedas an inherentpart of their being
ticulated by Manu: same time advises the king to fotgive the and had posed no problem had given way
The production of children, the nurture of variouserrants,the messengersas wellas the to a system tequirhqgstringent controls.
those born, and the daily life of men, of these queen. Seeking pardon for the messengers Women'ssexualitythus had now become a
matters the wife is visibly the cause. the chaplainsays "Menare not to blame for problem;theiressentialnatures,theirmater-
Offspring, the due performance of religious they wereconstrainedby the queen" In the nal power, thus had to be organised and
rites, faithful service and heavenly bliss for case of the queen the chaplain pleads "She orderedby paternalpower in the emerging
the ancestors and for oneself depend on the is not to blame for the passions of women class-basedsocieties to serve the new social
wife alone [Manu X.26-271. are insatiateand she does but act according and politicalarrangements by men
organis&;l
It was this recognition that men were depen- to her inborn nature"[Jafaka I. 2641. of the dominant classes.
dent upon women to perpetuate the social The innate wickednessof women is the women'sgeneralsubordinationwasessen-
and moral orddr of their making which led subjectof anotherstorywherethe good hus- tial in this stagebecauseit wasonly thenthat
them to confront the problem of women's band (who is a prince who has fallen on the mechanism of control upon women's
sexuality. Reproductive power was the one troubledtimes) performsthe most unimagi- sexuality could actually be effective. The
power tnat women still held in the new nable sacrifices to save his wife from star- mechanism of control operated through
structure of relations in which they were vation but who, at the first opportunity, threedevicesand at threedifferentlevels;the
subordinated and one way of dealing with abandons him for a common thief and at- first was through ideology, through the
it was to simultaneously exaggerate and treat tempts to murderthe husband by pushing stridharma,or pativratharma, internalised
as terribly dangerous women's 'innate' him down a precipice.The prince however by women who attemptedto live up to the
nature. Their uncontrolled sexuality was escapes and becomes the king; he then ex- ideal notion of womanhoodconstructedby
perceived as posing a threat and the narrative poses the evil nature of his wife saying the ideologuesof the society. In the case of
and normative literature of ancient India is "'womendeserveto die, they haveno truth"; Hindu society the design of the patriarchal
thus full of references to the wickedness of thereafterthe king rules death for both the caste-classstructurewas mappedout by the
women and of their 'insatiable' lust. sinners [Jalaka 11. 193]. brahmanas;pativratathe specificdhanrmna of
The story of Astavakra, narrated by Similarly in the Radha Jataka, the the Hinduwife then becaniethe ideologyby
Bhismato Yudhistra,graphicallyillustrates unguarded wife of a brahmana takes which women acceptedand even aspiredto
chastity and wifely fidelity as the highest her lord, dwells with him after death in mised that he and the younger wives would
expressionof their selfhood. heaven" and is called sadhvi, a chaste render her the honour she deserved. Sam-
This was the'lowestlevelof operationand woman, a faithful wife, by the virtuous bula and the king lived happily after that
one that requiredto controlas chastitycame (Manu IX.291.These internalisednormsare [Amore and Shinn 1981:34-36].
to be viewedas the means of salvation and the subject of much of the literatureon The Sambula story is an interesting
was therefore self-imposed. Pativrata, the women. variantof the Sita legendin the Rainayana.
ideological 'purdah'of the Hindu women A little known story where the focus on Thereareparallelsas wellas pointsof depar-
was thjusthe mask by which the hierarchical chastityis not explicitbut latent,or evenhid- ture. Both women accompany their hus-
and inegalitarian structure of the social den, indicates the value of such norms for bandsthroughtheirtravailsand both arethe
orderwas reproducedwith the complicityof womenas it enablesthe controlupon women object of an ogre's attention. Both have to
women. to be invisibilised.The story pertainsto an provethemselves,as theirchastity is suspect
It may be arguedthat the success of any extraordinarilybeautiful princess named but here the parallelends. The underlying
system lies in the subtle working of its Sambula who was the wife of the heir ap- assumptions however are essentially the
ideologyand in that sense the pativratacon- parent.Unfortunatelythe princecontracted same as both stories deal with the theme of
cept wfs the masterstrokeof Hindu-Aryan leprosyand decided to renouncethe throne suspicionabout the wife if she is awayfrom
genius. It was, in our view,one of the most and liveas a hermitascetic.Everyonelet him the husbandfor any lengthof time. Both are
successful ideologies constructed by any departincludinghis father,and all the wives guardedand protectedby their chastityand
patriarchalsystem, one in which women of the prince, since his open sores were virtue,and by their own internalisednorms
themselves controlled their own sexuality. becomingfoul and rotten.HoweverSambula of true womanhood as lying in devotion to
The actual mechanismsand institutionsof insistedon accompanyinghim whereverhe the husband alone.
control over women's sexuality, and the might go to look after him. So they went By and large most women conformed to
subordinationof women, was thus comple- together to the forest where the man built these internalisednorms, or at least aspired
tely invisibilisedand with it patriarchywas a leaf hut in a pleasant spot. to t4em in theory if not in actual practice.
firmlyestablishedasan ideologysinceit was Dedicating herself to the services of her But in situations wherethe ideological level
'naturalised'. husband Sambula rose early in the morn- of the controlover women was unsuccessful
That the stridharma, or the pativrata- ing to gather fruit and vegetables for his law and customs, as prescribed by the
dharmawas a rhetoricaldeviceto ensurethe food and bathe his wretchedskin with cool brahmanicalsocialcode, wereevokedto keep
socialcontrolof women,especiallychastity, water.One day whileshe was gatheringfood woTnen firmly under the control of the
is now well accepted. As outlined by Manu deep in the forestshe noticeda pleasantpool patriarchalkinship network. The right to
and elaboratedand repeatedby Tryambaka in a cave and bathed herself. As she step- control a womnan's total existence,especial-
in the stridharmapaddhatithe stridharma ped out her radiancelit up the forest and ly regulatingher impulses vested firmly in
was clearly an ideological niechanism for an ogre noticedher.He immediatelywanted the male membersof her family, first in her
sociallycontrollingthe biological aspect of her for himself but Sambularefusedhis ad- natal household and then in her conjugal
women.Women,as biologicalcreatures,are vances.The ogre then threatenedto eat her. household.This is a position statedmost ef-
representatives of a wild or untamednature. Sambulastruggledagainsthim and sinceher fectively by Manu but reiteratedby all the
But through the stridharmathe biological spiritualpowerwas so great(due to the ac- major brahmanicalcodes. Manu's dictum,
woman can be convertedinto woman as a cumulated merit of her good virtuous ac- "day and night women must be kept in
social entity, in whom the biological has tions)the god Sakkanoticedher distressand dependence(and guarded)by the males of
been tamed".In contrast in the Kali age came down to earth to help her. their families"is an explicitstatementof the
especiallythereis an inversionof the system When she returnedafter her misadven- need for stringentcontrol upon women to
in which women lapse into unrestrained turesto her leproushusbandshe wasgreeted safeguardthemand savethem fromtheir'in-
behaviourdisregardingthe stridharmaand with suspicion; the husband would not nate' addiction to sensual enjoyment. He
throwingoff all morals. The wicked and believe her story and reminded her of the rulesfurtherthat if theyarenot guardedthey
essential nature of women then must be wiles of women. The desperate Sambula bring sorrow to two families, the one into
subordinatedand conquered by the virtue cried "Oh my husband, what can I do to which a woman is born and the one into
of the ideal wife. Once the tension between convinceyou of my devotionto you and you which she is given [Manu IX.2-51.
'nature'and 'culture'is resolvedwomencan alone!'Then a solution occurredto her and Special responsibilityis guardingwomen
emergetriumphantas paragonsof virtue.It she decided to perform the ancient ritual is laid upon the husbandwho is represented
is evident from Tryambaka's text that called the 'act of truth' in which a person as most vulnerableto the loss of his progeny
ultimatesocial control is achievedwhen the of great virtueproclaimsthe basis of virtue throughthe infidelity of women. Consider-
subordinated(herewomen)not only accept and if the claim be true;the powerof virtue ing it the highest duty of the husband(and
their condition but consider it a mark of will prove sufficient to work any miracle here he dictates explicitly for all castes)
distinction. requested. So then she proclaimed aloud, Manu enjoins that even 'weak' men must
Muchattentionhas been focusedin recent "MayI be protectedby this truth:that I have strive to guard their wives [Manu IX.61.
yearson the ideologicalcontrolupon women neverheld anyone dearerthan you. By this Baudhayanaalso enjoins that the wives of
throughthe idealisationof chastityand wife- spoken truth, may your disease be cured" men of all castes must be guarded more
ly fidelity as the highest duty of women, Tocompletethe ritualshe pouredwaterover carefully than wealth (Baudhayana
reinforcedthrough custom and ritual, and the diseased skin of her husband and im- 11.2.3.34-35].Occupyinga central place in
through constructions of notions of mediately his sores were washed away. the enforcementof controls upon the wife's
womanhoodwhich epitomise wifely fideli- Cured thus the husband returnedto his behaviour alongwith the husband is the
ty as in the case of Sita, Savitri, Anasuya, kingdom and in due course was installed father-in-law whose. authority keeps the
Arundhati and a host of other similar king while the old king retiredto the forest. daughter-in-law in check. According to
figures in Indian mythology., We shall Uncaring of Sambula'sgreat sacrifice the Medatithi'scommentaryon Manu'sanalysis
thereforenot labour the point. However, newly installed king ignored her and spent of the six causes of the ruin of women are
Manu'sdictumevenhereoutlinesthe impor- more and more time with the youngergirls includedassociatingwith wickedpeopleand
tanceof the ideological mechanism;in his in his entourage.Sambulabore the insult in sleeping at unusual hours; these are
viewno mancan completelyguarda woman silence but her miseriescaused her to grow reprehensibleacts as they erase the fear of
by force [Manu IX.l0] and therefore it is thin and frail.One day the king came to the the father-in-law[Das 1962: 1701.
womenwho of tliir own accord keepguard palaceand noticingher sad state discovered The authority-of the male kinsmen is
overthemselvesthat arewellguarded[Manu her plight. He reprimandedhis son saying backedby the potentialrightto use coercion
X1.13]. Further Manu points out that a "A good wife is hard to find, but you have and physical chastisementof women who
woman who "*controllingher thoughts, a virtuous wife so treat her according to violate the normsestablishedfor them. The
speech,and acts violatesnot herduty toward dharma' The husbandapologised and pro- fear of physicalpunishmentmay appearto
Economic and Political Weekly April 3, 1993 583
be only a deterrent in the normative functionsareassociatedwith kinshipin early house of the husbandand going to a neigh-
literaturebut that it wasactuallyusedis clear India: punishingthose who commit crimes bour is an offence; even kinsmen of the
from the narrative literature. In the against the family,i e, adulterers,and those women are included among the homes of
CulapadumaJalaka the adulterouswife is who commit crimes against property, i e, people women are forbiddento visit unless
described as a harlot by the husband who robbers.Even before the state emerged we they are being ill-treated.Only in case the
first expounds that women deserve to die, haveevidenceof the notion that controlover house of the kinsmanhas been the scene of
then recommendsand executes the cutting women'ssexualityis the concernof the com- death, illness, calamity, or a childbirthis a
off of the adulterouswoman'snose and ears munity of men that constitutes the clan in woman permitted to go there but even in
[Jataka 1.193]. Similarly in the Gahapaii whom political authorityis vested. An inci- such situations the visit must be made with
Jataka, the errantwife of a gahapati when dent recordedin the VinayaPitakadescribes the consent of the husband (3.4.13-15).
caught by the husband is thrashedby him. how a woman,who hadcommittedadultery, The Arthasastra clearly suggests that
He seizes her by the hair, knocks her down flees from her husband who has been husbandswereaided by the coercivepower
and threatensher "If you do this kind of authorised by the clan to kill her seeks of the state in ensuring a firm grip on the
thing again, I'll makeyou rememberit' He shelterin the BuddhistSanghato escapethis 'impulse' control of women, and that
also demandeddamages from the adulterer punishment IVinaya IV: 225-26). through its punitive measures on the free
saying "Damages please for injurydone to After the emergence of the state the movementsof womenopportunityfor viola-
the chattelsunderanotherman'swatchand brahmanicalnormative literatureand the tions of the sexual code were effectively
ward". The narrative concludes with the semi-secular Arthasastra laid down minimised. There is thus very little discus-
statement that following the physical punishments for violations of the sexual sion on adultery itself in the Arthasastra.
chastisementthe wife did not daretransgress code whichthe kitigwasexpectedto enforce. However,sections outlining the duties of a
even in thought [Jataka 11. no 1991. These texts reflectthe more generalanxiety king, or those that concern laws in the
Another story in the Jalaka combines about the husband's need for progeny to brahmanicallegal literaturedwell at length
humiliation with physical punishment to completethe religiousrequirementsof men, upon adultery,as well as upon the violation
cure a woman of her evil ways. Describing and the need to ensure 'legitimate'succes- of the principlesgoverningpermittedunions
a woman's errant behaviour the Kosiya sion to pass on propertybut there is also a betweenmen and women.Violationsin both
Jataka tells us that the bad wife of a good concern about the maintenance of the cases are considered bad but what is con-
brahmanaspenthernightsin gaddingabout, hierarchicalsocial order, based on caste, sidered most mrprehensible is the case of a
and feigningillness duringthe day. She did which must be reproducedwithout diluting high status woman involved with a lower
not do a strokeof work while the husband the purity principle. The burden of caste man. Gautama lays down that a
slaved all day to get her the luxuries that reproducingit lay upon womenand adultery
she demanded.The momentthe brahmana's thus took on an added significance. Manu
back was turned the women flew into the states this explicitly while discussing HOUSING IN
arms of her paramours.The brahmanawas adultery.Accordingto him, "By adulteryis THE THIRD WORLD
advised to assert his control over her, star- caused the mixture of castes among men;
ting with giving her pickled cow dung to hence follows sin, which cuts up even the
Analyses and Solutions
eat and then taking rope or a stick, and roots and causes the destruction of
threateningthe wife with either swallowing Editedby
everything"[Manu Vil: 3531. Leslie Kllmartin & Harjlnder Slngh
the dung or by workingfor her food. If she
refused she was to be given a taste of the The king, who nereacts as an executorof
class power, is howeveronly the ultimate Rs. 400
rope or stick and simultaneo4sly the hus-
band was to drag her by the hair while he agency by which women'ssexuality is con-
trolled.To s4ccessfullyestablishthis control The papers in this collection pro-
pummelled her with his fists. The woman
was thus brought to heel and became as therearea varietyof waysin which women's vide overviews of the topic and
good as formerly she had been wicked 'impulses' are to be -urbed and these are contain contributions which refer
outlinedin the Arthasasira.The Arthasasira
[Jataka 1.284]. to contexts in which the provisions
In all the above-mentioned narratives regulatesthe punishmentenforced by male
whetherphysicalpunishmentis actuallyused kinsmen in inculcating modest behaviour, of housing may be usefully set. It
or not there is an explicit injunction to the which is considered their prerogativebut also proposes solutionstothe prob-
effect that it is advisableto use violence to mustconformto the normslaid down by the
state Thus the text statesthat in inculcating lem of housing provisions.
punish women, particularlywives, to make
them conform to the requirementsof wife- modest behaviourcertain abuses are to be
ly fidelity. avoided. But while verbal restraintis to be
exercised,the use of force itself is permit- HOUSING THE POOR IN
The powerto use violencevestsin the hus-
band and it is recommendedas the means ted. Accordingto Kautilyaone can strikethe THIRD WORLD CITIES
to ensure control over the wife's sexuality, back of a woman three times with either a Choice Behaviour
in particular, and in monitoring her split bamboo cane, or a rope, or else by the
hand.Similartreatmentis prescribedfor the and PubiIc Policy
behaviour more generally. But what if
wife who 'enjoys'herself outside the home by Kamlesh Mlsra
husbandsdo not succeed, even throughthe
use of violenceto bringwomen to heel? For [Arthasastra3.3.7-10].
such situations a third mechanismof con- Wiveswho, though prohibited,indulgein Rs. 200
trol was envisaged in the ancient Indian the sport of drink, or go by day to a show
patriarchalstructure,with the king being by women,or evengo on a pleasuretripwith The book is an importantand timely
vested with the authority to punish errant other women are to pay fines rangingfrom contribution to the resolution of
wives.The king functionedas the thirdlevel three to six 'panas. The 'offence' is con- housing problems in third world
of control over women through whom the sideredmuch more seriouswhen committed
coercivepowerof the patriarchalstate was at night;the fine is then to be doubled.Most cities.
articulatedand used to chastisethose wives serious are those offences that relateto any Publishers: Phones: 5504042,5554042
who flouted the ideological norms for form of interactionwith men other than the CONCEPT
PUBUSHING
COMPANY
women and also subvertedthe control of husband. If womenconverseWithmen in a
BLOCVMoWIGCaaE
A/115-16,COMMERCAL
male kinsmen. 'suspicious'placetheycan be whippedin the
The patriarchalstate of early India view- centre of a village by a 'chandala'instead NEwDEULe-10059
ed adulteryas one of the major 'crimes'in of being mnerelylet off with a whipping StiowRook: 3272187
PHO0NE:
society. In the Buddhistliteratureonly two privately(Arihasastra3. 3.27). Leavingthe 4788/23,AN~S.'
Ro.w,DARYGAN,
NEWDEU*59
woman who has connection with a lower portant substantiation of the overarching R P Kangle, Motilal Banarasidass, Delhi.
caste man becomes an outcaste; if she com- support of the state for patriarchalcontrol Baudhayana Dharma Sutra (tr) (1986), George
mits adultery with a man of the lower caste over women. Patriarchycould thus be esta- Buhler, Sacred Laws of the Aryas, Vol 11,
the king shall cause her to be devoured by blishedfirmlyas an actualityand not merely Motilal Banarasidass, Delhi.
dogs in a public place [XXIII: 141. as an ideology. The archaicstate was clear- Bhagavad Gita (1968), S K Belvalkar (ed),
It i.s noteworthy that according to ly both a class state and a patriarchalstate; Bhandarkar Oriental Research Institute,
Gautartia whereas the lower caste adulterer in the case of India there has been a close Poona.
should be killed the woman is to be publicly connectionbetweencaste,class,and the state Chakravarti, Uma (1985), 'Of Dasas and
which together functionedas the structural
humili4ted and suffer a more ghastly death. Karmakaras: Servile Labour in Ancient
Vasistha on the other hand reverses the onus framework of institutions within which India' in Utsa Patnaik and Manjari
of the guilt somewhat and while the woman gender relations were organised.
Dingwaney (eds), Chains of Servitude:
escapes the death penalty which the low To sum up, a preliminary analysis of Bondage and Slavery in India, Sangam
caste man must face (he is to be thrown in- Brahmanicalpatriarchyin earlyIndiareveals Books, pp 35-75, New Delhi.
*o the fire) the king is enjoined to punish that the structureof social relations which -(1987), The Social Dimensions of Early Bud-
and humiliate her by shaving off her head, shapedgenderwas reproducedby achieving dhism, Oxford UniversityPress, New Delhi.
placing her-naked on a donkey, and parading the complianceof women. The compliance Das, R M (1962), Women in Manu and His
her along the highway. According to Vasistha itself was producedthrougha combination Seven CommentatorsmVaranasi.
following this punishment she is rid of her of consent and coercion as we have tried to
impurity [XXI: 1-2]. outline above. While the elaboraterules of Das, Veena (1976), 'Indian Women: Work,
The case of a maiden violating the caste normativeliteratureand descriptionsin the Power and Status' in B R Nanda (ed),
rules for sanctioned unions between men narrativeliteratureindicates the failure of Indian Womenfrom Purdah to Mfodernity,
and women is considered less reprehensible. brahmanicideologyto producethe realcon- pp 129-45.
In Manu's view the king may overlook the sent of women to brahmanicalpatriarchy Digha Nikaya (1976), E Carpentier (ed), Pali
offence of a 'maiden' who makes advances (therebyrequiringa recourseto coercion)the Text Society, London.
to a man of a high caste (this was obviously values of the caste system were apparently Ganesh, Kamala (1985), 'Women's Seclusion
a permitted lapse) but in the case of a acceptedby both men and womenof the ap- and the Structureof Caste',paper presented
maiden who courts a man of a lower caste per castes. Women's perpetuation of the at the Asian Regional Conference on
the king should force her to remain confin- caste system was achieved partly through Women and the Household, New Delhi.
ed in the house [Vill: 365]. The maiden's their investmentin a structurethat reward- Gautama Dharma Sutra (tr) (1975), George
crime is of less gravity than the wife's, since ed them even as it subordinatedthem at the Buhler, Sacred Laws of the Aryas, Vol 1,
there is no pativraladharma that she has same time. That they too subscribedto the Motilal Banarasidass, Delhi.
violated, but Manu reserves the highest ideology of the caste system-isevident from The Jataka (1957), R B Cowell (ed),
punishment for the wife who though aware an account in the Jalakas of two high caste R Chalmers, etc (tr), Pali Text Society,
of the 'greatness of her relatives' (i e, of their womenwho ranto washtheireyeswhenthey London.
high status) violates the duty that she owes sighted two low caste untouchables[Jataka
IV:No 391].All the anxietydisplayedby the Lerner,Gerda (1986), The Creation of Patriar-
to her lord, i e, her siridharma or her
early texts to monitor the upper caste chy, Oxford University Press, New Delhi.
pativratadharma. In such a situation Manu
like Gautama rules that the king should woman'ssexualitymaintain her purityand Leslie, Julia (1989), The Perfect Wife: The
cause her to be devoured by dogs in a place thus of the caste would become somewhat Orthodox Hindu Wjfe according to the
frequented by many [ViII: 377]. In punishing unnecessaryonce women becamecomplicit StridharmaPaddhatiof llyambakayajavan,
such 'deviant' women the king was up- in the larger structurein which their own Oxford University Press, New Delhi.
holding the existing structure of relations subordinationwas embedded. Mahabharata (1993-39), V S Sukhtankar et al
pertaining to land and the caste order. The (eds), Bhandarkar Oriental Research In-
purity of women ensured the purity of caste Notes stitute, Poona.
and thus of the social order itself. Manu Dharma Sastra (1984), George Buhler
Much of the evidence cited in support of I The attempt made in this paper to trace the (tr), Laws of Manu, Motilal Banarasidass,
the role of the state in monitoring the im- workings of brahmanical patriarchy should Delhi.
pulses of women is in form the normative not be seen as a single chronological develop- Neumayer,A (1983),Prehistorc Rock Paintings
literature and therefore one cannot be cer- ment. The evidence relates to different in Central India.
tain about its working and its effectiveness. regions and different groups of people O'Flaherty, Wendy D (1976), The Origins of
However, if we go by the basic principle of located in specific material cultures. I am Evil in Hindu Mythology, University of
Mimamsa philosophy that something can be therefore not arguing for a monolithic California Press, Berkeley.
prohibited only if its occurrence is possible development of patriarchy given the range -(1985), Tales of Sex and Violence, Motilal
then the role of the state becomes clear. of social formations. Banarasidass, Delhi.
Further a reterence in the narrative lite- Ramayana of Valmiki (1958), S Kuppuswami
2 Notions of the excessive sexuality of women Sastrigal et al (eds), Madras.
rature suggests that kings did regard them-
were not unique to brahmanical literature Rig Veda (1971), R T H Griffth, Varanasi,
selves as responsible for punishing wives
who violated sexual norms. A Jataka story and were widely prevalent in the Buddhist Chaukhamba Sanskrit Series Ofrice.
recounts that when a wife's misbehaviour is texts too, indicating the permeable boun-. Roy, Kum Kum (1987), 'Women in Early India',
brought to the notice of the king he sends daries of the two textual traditions. unpublished type script.
a message back to her stating that the wife Satapatha Brahmana (1964). A Weber, (ed)
must realise "that there are kings in the Chaukhambha Publishers, Varanasi.
land". He tells the messenger to say "she References VasishthaDharma Sutra (1975), George Bu
must dwell with her husband and if she does Sacred Laws of the Aryas, Vol II, Moisaa
not let her have a care; the king will cause Amore, R C and LarryShinn D (1981), Lustful Banarasidass, Delhi.
her to be seized and she shall die" [Jataka, Maidens and Ascetic Kings, Oxford Univer- Vinaya Pitaka (1879-93), H Oldenberg (ed),
I: 214]. Even if the Jataka story is indicative sity Press, New York. London.
only of the widespread social sanction for Apastamba Dharma Sutra (tr) (1975), George Yalman, Nur, 'On the Purity of Women in the
the king's authority rather than as firm Buhler, Sacred Laws of the Aryas, Vol 1, Castes of Ceylon and Malabar: Journal of
evidenceof the king'sactualenforcementof Motilal Banarasidass, Delhi. the Royal Anthropological Institute of
authorityover women'sconduct, it is an im- Arthasastra (1986), edited and translated by
GreatBritainand Ireland,93, pp 25-28.

585
Buddhism and Caste System
Author(s): Y. Krishan
Source: East and West, Vol. 48, No. 1/2 (June 1998), pp. 41-55
Published by: Istituto Italiano per l'Africa e l'Oriente (IsIAO)
Accessed: 03-03-2015 07:14 UTC
Istituto Italiano per l'Africa e l'Oriente (IsIAO) is collaborating with JSTOR to digitize, preserve and extend access to East and
West.
Buddhism and Caste System
by Y. Krishan
I. In Buddhist Texts
It is a common and widespread belief among scholars that the Buddha had taught
that all men were equal, that social superiority based on varna (colour, race) and j?ti
was untenable. The implicationis that theBuddha and Buddhismwere opposed
(birth)
to the Brahmanical hierarchically graded castes which were endogamous, non-commensal
and governed by a discriminatory legal system with unequal rights and obligations.
A priori, in a Buddhist society, there is also no justification for untouchables outside
the caste group.
A critical examination of the Buddhist suttaswhich deal with varna, and j?ti
and the related social features, kula and gotra, indicates that the lay Buddhists
accepted the caste system and never challenged it and the discriminatory laws.
1. Varna, j?ti, kula and gotra as Basis of Castes
(a) Varna as the Basis of Castes
TheMadhura sutta (84), theKannakatthalasutta (90) and theAssal?yana sutta(93)

of theMajjhima-nik?ya, the Cullavagga ix.1.4 of the Vinayapitaka, etc. all recognize
the existence of four castes: Catt?ro vann?, Ksatriya, Br?hmana, Vaisya and S?dra.
In the Kannakatthala sutta the superiority of Ksatriya and Br?hmana castes
is recognized: dve vann? aggam akkh?yanti: the two castes are said to be chief and
therefore deserve respect and service by the other two castes.
(*) This subjecthas been dealt with by the author previously inY. Krishan, 'Buddhism and Caste
System',JIABS, 9, 1986, pp. 71-83. However a furtherand freshexamination of this subjectcompletely
confirms the conclusions arrived at and more removes all doubts or ambiguity
previously importantly
about the attitude of the Buddha towards the institutionof caste in India. The prevalence of caste
systemin the present-dayBuddhist societies of Sri Lanka, Nepal, Ladakh (India) and Burma provides
corroborative evidence that Buddhism did not challenge, nay accepted, the caste system.
[1] 41
(b) Jati as Basis of Caste
In the context the term j?ti is used, itmeans birth, descent. The only
in which
exception sutta
is the V?settha of the Suttanip?ta.
(i) The Vinayapitaka IV.6 enumerates htna j?tis, viz. Cand?la, Vena (basket makers),
Nes?da (hunter, trapper), Pukkusa (scavanger), Rathak?ra (maker of chariots), etc.
(ii) Again theVinayapitaka IV.6 declares that calling an ordainedmonk as having
come from a htna jati? Cand?la, Vena, etc. ? tantamounts to abusing him ? calling
names for which pr?yascitta is essential. The word j?ti in this context refers to a
distinct category of human beings based on birth or descent.
It is significantthat in theP?ccittyaII of theSuttavibhangaof theVinayapitaka,
the mischievous monks enumerate the criteria for testing the dignity of a monk on
the basis of, interalia, (i) j?ti, n?ma (name), gotra (clan) karma (kammam) (workor
avocation), and silpa (craft).
(iii) In the Anguttaranik?ya III.57.2 the Buddha states that human beings are Ksatriya,
Br?hmana, Vaisya, and S?dra, Cand?la and Pukkusa (both outcastes) on the basis of
theirbirth, j?tiyam.
(iv) The expressions j?tiya n?ma gottena {VinayapitakaIV.6) and j?tito, n?mato,
gottato {Dighanik?ya11.18), that is, by descent, by personal name, and by gotraor
as regards one's
family, and the expression j?titthaddho (Suttanip?ta 104), conceited
j?ti (descent), wealth and gotra, clearly indicate that the word j?ti connotes birth only.
(v) In thePabbajj? suttaof theSuttanip?taIII. 1,Ananda describes the ordinationof
Gautama the Buddha. Bimbis?ra tells Gautama in sutta 97: Vann?rohena sampanno,
j?tim?viya khattiyo: fromthe complexion and physiognomyyou (Gautama) appear
to be a Ksatriya by j?ti (descent). The Buddha replies thatby gotra,he is S?ryavamsi
{Adicca) and comes from the S?kya j?ti (clan or tribe).
(vi) The Dighanik?ya 1.120 avers: y?va sattam?pit?mahayug?akkhito anupakkuttho
j?tiv?dena, 'of unblemishedparentage upto the previous seven generations'.
Thus in all these references the word j?ti undoubtedly means birth or descent
as determinant of caste (l).
(vii) In theKhuddakap?thaVI.35 of theKhuddaka-nik?ya (2),persons belonging to
the hina j?ti, 'low castes', viz. Cand?la, Vena, Nes?da, Rathak?ra, and Pukkusa, are
of the seven which are produced for the
totally excluded from possessing any jewels
use of a Cakravarti King of the Ksatriya Caste 'Well born on both sides'.
t1)Majjhimanik?ya 93 (Text ii, 157): na mayam j?n?ma gandhabbo khattiyov? br?hmanov? vesso
v? suddo v?, 'gandharva (the transmigratingpudgala) in the intervalbetween death and rebirth, is
not known to be a Ksatriya, Br?hmana, Vaisya or Sudra'. This only emphasizes that a disembodied
transmigratingpudgala has no j?ti (caste); j?ti is a featureof the embodied human beings only.
(2) Bhikkhu N?n?moli, TheMinor Readings:Khuddakap?tha of theKhuddakanik?ya, London 1972.
42 [2]
(c) Kula as a Factor in Caste Distinctions
Again castes are also distinguished by the appellation kula and kulina. The word
kula, inter alia, means tribe, caste, a noble or eminent family, or race (3), a collection
of cognates and agnates, a high social grade, a 'good family' in the sense of line of
descent (4). The derivate kulina means belonging to a good family,of a high or
eminent descent or good breed. Thus there are kulas of Br?hmana, Ksatriya, Vaisya
or Vanija, ?
S?dra and of outcastes Cand?la Nes?da, Vena, etc.
So the terms often used in Buddhist literature, kula, kulina, and kulinat? are
clearly racial or familial and hence indicative of varna or birth-based j?ti and not of
professional or vocational castes.
(d) Another Social Classification Based on Birth
The Anguttaranik?ya ii.85 (Samyuttanik?ya ii, 84) makes yet another classification
of human beings. It says that puggalas, beings, are of four categories:
(i) tamo tama par?yano, of miserable birth and bound for misery.
(ii) tamo jotipar?yano,of miserable birth but bound forhappiness.
(iii) joti tamapar?yano,of happy, good birth but bound formisery.
(iv) joti jotipar?yano,of happy, good birth and bound forhappiness.
It is significant that tamo puggalas are identified as belonging to ntca kula, low
families of Cand?la, Vena, Nes?da, Rathak?ra and Pukkusa. Thus these four categories
of pudgalas are suh-yonis of the manusya yoni. Hlna (inferior) j?tis are dysgenic and
ucca (superior) j?tis are eugenic. This classification of human beings is again based
on birth or descent; that is j?ti.
(e) Gotra
Gotra is an obscure and controversial concept. Broadly speaking persons of the

same gotra are considered as having either descended from a common ancestor, who
may be a religious teacher and whose disciples are bound by common religious-school
ties or descended in an unbroken male line from a common male progenitor. In either
case gotra is an aspect of j?ti or birth as would be amply clear from the earlier citations
regardingj?ti fromtheBuddhist texts (5).
(3) M. Monier-Williams, A Sanskrit-English Dictionary.
(4) T.W. Rhys Davids & W. Stede, Pali-EnglishDictionary.

(5) Rhys Davids & Stede, op. cit., define gotra as ancestry, lineage, all those descended from a
common ancestor, an agnate.
[3] 43
2. A New Interpretation of the Vasettha sutta
The meaning of the term j?ti in the V?settha sutta of the Suttanip?ta
=
( Majjhimanik?ya 98) requires special considerationkeeping in view the commonly
accepted meaning of this term.
In the V?settha sutta, it is explained that the four castes of Ksatriya, Br?hmana,
Vaisya and S?dra are not distinct species or differentiated by descent or blood but
are merely descriptive labels of occupational groups in society. They all belong to
one common j?ti, the manusya j?ti.
In this sutta, 291-307 the Buddha explains that trees, various insects and ants,
quadrupeds, reptiles, fishes, birds, etc. are j?tis (species) as they are distinguished by
linga, distinctive anatomical features; linga-bheda, distinctive anatomy, leads to j?ti bheda,
distinctive species. The four castes, on the other hand, are distinguished from one another,
not by linga and j?ti but by sarhjn?labels or appellations as such. The four castes represent
not different species of human beings such as different races and tribes, not even genetic
variants of an organism but are mere nomenclatures of socio-economic groups (vargas)
based on their occupations. Thus in the V?settha sutta 307, the Buddha observes:
Yath?et?su j?tisu, linga j?timayam puthu /

Eyam natthi manussesu, Ungarn j?timayam puthu //
Just as these j?tis (plant, bird, animal, etc. species) are distinguishable anatomically
(linga bheda), the human species have no anatomical differences based on their birth.
The distinctions between men, rather groups (varga) of human beings, is based on
their occupational labels or appellations (sarhjn?). So the Buddha explains in suttas
611-619 that the same human beings make a living through cattle rearing (go raksa),
farmingand are called krsaka (farmers),throughcrafts (silpa) and are called silpi
(craftsmen); through trade and commerce and are known as vanija (traders); those
who make living by rendering service to others (para pessena) are known as presaka
(servants); those who make a living through warfare are called yodh?jivi (warriors,
i.e. Ksatriya); those who make a living by performing priestly functions, that is of
a purohtta, are called Br?hmanas. In this sutta, 648-650 the Buddha again declares:
samann? hes? lokasmim, n?magottam pakappitam /

sammucc? samud?gatam, tatha tatha pakappitam //
Gotra an imaginary name or appellation,

is merely a means of identification in
society. Designation so conceived by human beings have become conventional in

society. So the Buddha concludes: na jacc? brahmano hoti, na jacc? hoti abr?hmano,
'a person is a Br?hmana or not-a-Br?hmana not on the basis of birth' (6).
(6) The Madhura suttaof theMajjhimanikaya 84 is another sutta inwhich Kacc?yana declares that
varna (caste) ismerely a designation which has been or declared just a nomenclature.
proclaimed (ghoso),
44 [4]
Again inV?settha suttathebr?hmanasV?sistha and Bh?radv?ja ask theBuddha
whether a person is br?hmana by birth or by kamman?: j?tiya br?hmano hott, ud?hu
bhavati kamman?. The word kamman? means profession of avocation and not karma.
This will be amply clear from the replyof theBuddha.
The Buddha defines the kamman? of various castes twice in this sutta in the
course of his reply: in the V?settha sutta 612-619 and again in 650-652.
In suttas 612-619, the Buddha identifies the various professions or avocations
as sources of livelihood of various castes. Again in the suttas 650-652 he identifies
castes ? In the sutta 650, the Buddha
the by their professions kamman?. says
whether a person is a br?hmana or a non-br?hmana is to be decided on the basis of
his kamman?: kassako kamman? hoti, sippiko hoti kamman? / Vanijo kamman? hoti,
pesiko hoti kamman? //Coro api kamman? hoti, yodh?jtvoapi kamman? //Y?jako
kamman? hoti, r?ja api hoti kamman? (vv. 657-59).
The term kamman? in these verses means 'by profession'
(7) and is synonymous (8)
with jivik? 'livelihood', jtvati'to live by, subsist', upajtvatVto depend on, to live by'.
These synonyms are also used in the earlier suttas 611-616. It is inappropriate to
translate it as 'work' or 'deed' as done by some scholars (9). Though literally correct,
they fail to distinguish between the term kamman? used in suttas 650-652 and in
sutta 654.
In sutta 654, this term is used in the sense of punya and p?pa karma, merit and
demerit, moral karma which creates a transcendental potential to be experienced
in future lives. The term kamman? of sutta 654 generates sancita karma which
will mature {vip?ka) at an unknown time in future. It is in the nature of causal
karma as is evident fromsutta654 itself (10). The termkamma vip?ka used in the
immediately preceding sutta 653 confirms it inasmuch as karma vip?ka has a technical
meaning, 'the ripening of actions', i.e. the good or evil consequences in this life of
human acts performed in previous births (n). The professional actions, except of a
thief, on the other hand, are in the nature of kriy?s which do not generate sancita
karma and hence are non-causal. In the alternative, if the term kamman? is to be
taken to mean karmas in the moral sense, then the suttas 650-652 mean that by
the type of acts a person does now and here, he will in a future birth become a
(7) Rhys Davids & Stede, op. cit.

(8) See also Rhys Davids & Stede, op. cit., for the synonyms.
(9) Various scholars have translated the word kamman? of these verses as under:
'what he (farmer, merchant, craftsman, soldier, priest, king) does': Saddhatissa (transl.), Suttanip?ta,
London 1985.
work': Fausb?ll (transl), Suttanip?ta (SBE, Vol. X), Delhi 1980.
'deed': LB. Horner, The Middle Length Sayings (Majjhimanik?ya),V?settha sutta 98.
(10) Kamman? vattati loko kamman? vattati paj?.
(n) Monier-Williams, op. cit. We have dealt with this aspect previously in Krishan, op. cit.,
pp. 79-81.
[5] 45
Br?hmana, a farmer, a craftsman, a trader, a servant, a thief, a priest, a soldier
or a ruler.
To us it appears that the term kamman? in suttas 650-652 signifies livelihood
activity, profession or avocation. And that is how it has to be since there are no moral
acts or karmas which can be distinguished as appertaining to the four varnas/j?tis.
So the suttas 657-669 are translated as under:
By profession one is a cultivator or farmer

By profession one is a craftsman
By profession one is a servant
By profession one is a thief
By profession one is a soldier
By profession one is a priest (performer of yajnas)
By profession one is a King or ruler.
The word j?ti is derived from the root "Jjanwhich means 'to generate, to beget,
to produce, to cause to be born' (12). The word j?ti therefore means birth, descent.
Etymologically j?ti is an aspect of birth or descent. Hence in the V?settha sutta, the
word j?ti was artificially construed and equated with the designation of a profession
or avocation (13).
The meaning of the termj?ti in theV?settha suttahas to be understood in this
context. It appears that in this sutta, the Buddha was explaining the historical genesis
of castes as propounded in the Purusas?kta of the Rg Veda. It is therefore erroneous
tomaintain on the basis of this sutta, that the Buddha challenged a caste system based
on birth. This also goes against the overwhelming mass of evidence that j?ti clearly
stands for birth, descent by birth and not for occupations.
3. The Doctrine of karma and Caste System
As has been explained previously (14), according to the Buddhist Canon, moral
acts ? ?
good (punya) and evil (p?pa) karmas have no effect on the current caste
(12) Monier-Williams, op. cit.
(13) Itmay be relevant to recall that in the previous article (Krishan,op. cit.) ithas been brought
out that in actual practice, the members of the three upper castes, followed multifarious professions and
not necessarily those which they must necessarily practise under vama vyavasthd or j?tiv?da. In particular
the laymencould followany professionexcept thosewhich cause pollution or impurityorwhich involved
commission of violence and destruction of life. See also Anguttaranik?yaCXCII (iii, 223 ff.)wherein
the Buddha says that in his days the Br?hmanas were doing farming, cattle-breeding, trade, crafts, etc.
(14)Krishan, op. cit., pp. 79-82 dealt with this point at some length. There he had cited the
C?lakammavibhanga suttano. 135 and theAssal?yana suttano. 93 of theMajjhimanik?ya in support
46 [6]
Status of an individual; a person's caste does not undergo a change with reference to
his conduct; only moral acts affect the caste in which he might be reborn and that
as and when such karmas mature.
4. Varna samkara a Vice, and Maintenance of Caste Blood Purity a Virtue
The Buddha condemned miscegenation and upheld the Virtue' of caste blood
purity.
In the Assal?yana sutta the Buddha explains that when a mare ismated with an
ass a hybrid new j?ti ? a mule ? is born. From this passage it appears that the
Buddha did not approve of mixed marriages. The Buddha concludes:
First you (Assal?yana) went about birth, leaving birth, you went about mantras (he
who knows and can recite the Vedas), leaving mantras you arrived at the purity of
four castes which is just I lay down. (15)
Here theBuddha was referring

undoubtedly to purityof (caste)blood and not
moral purity. And purity of blood is the most essential features of j?ti v?da.
In the Anguttaranik?ya iii, 221 f., the Buddha severely criticizes the Br?hmanas
of his days for contracting marriages indiscriminately with women of other castes as
it vitiates the purity of their blood. He observed:
In former times br?hmanas approached only a br?hmani (a Br?hmana lady), never

a non-br?hmani; now they go to the br?hmani and non-br?hmani alike.
of his thesis. To thiswe would add the Puggala-Pannati IV 19. The usual text is: k?yassa bhed?
paramamaran? app?yam suggatim [...] duggatim niryam suggam lokam uppajjeyyam.
We also invite attention to two further references. First is the Vasalasutta of the Suttanip?ta which
narrates that M?tahga, who was a Cand?la and was honoured of the higher
by the members castes,
Br?hmanas, etc. by virtue of his conduct could attain the Brahma world after his death. The Vasala
sutta 139 says: na nam j?ti niv?resibrahmoloka upapattiy?, 'low birth is no impediment to being born
(afterdeath) in the Brahmaloka'. The sutta 141 reiterates:na te j?ti niv?retiduggaca garah?yav?, 'the
j?ti or birth caste in thisworld does not protect him against ill reputation in thisworld and evil birth
in the next'. Second, the Kammakatthala sutta 90 of theMajjhimanik?ya restates that the caste differences
in theexistingworld will disappear in the future,in thenextworld (sampar?yika)ifthepersonsof different
castes possess the five qualities of striving (sila, sam?dhi panna vimutti, and vimuttin?na dassana) in equal
measure.
(15)The text is reproduced forconvenience:pubbe kho tvam,Assal?yana, j?timagam?si,j?timgantv?

mante agam?si, mante gantv? tarn etam tvam cattuvannim suddhim paccagato yam aham pannapemiti.
The Dighanik?ya 1.120 has already been cited wherein the Buddha asserts the superiorityof the
Ksatriyas over the Br?hmanas on the ground of purity of blood of the formerupto seven previous
generations.
[7] 47
The Buddha adds thatdogs onlymate with dogs and notwith any other species
suggesting thereby that even dogs maintain the purity of blood better than the
Br?hmanas.
Again the Anguttara CXCII (iii, 223 ff.) re-emphasizes that a true Br?hmana
should be well born on both mother's and father's sides (ubhato suj?to m?tito ca
pitito), pure in descent as far back as seven generations (y?va sattam? pit?mah?yug?).
It furthergoes to describe five typesof Br?hmanas: (i) Brahma-like; (ii) deva-like;
(iii) those who observe the maryad? (the limits of morality, propriety, rule or custom);
(iv) thosewho violate the rule ofmorality (sambhinnamariy?dam);
and (v) Cand?la
like. The description of the last two types of Br?hmanas is very significant.
(a) sambhinna mary?d? br?hmano, violator of the rules of morality-Br?hmana.
So br?hmanim [...] khattiyam [...] vessam [...] suddam [...] cand?lam [...] nes?dam
[...] venim [...] rathak?ram [...] pukkusim pi gaccati. Such Br?hmana goes to a
Br?hmani, the daughter of a Ksatriya, Vaisya, S?dra, Cand?la, Vena, Rathak?ra,
and Pukkusa.
(b) Br?hmana-Cand?la. Such a Br?hmana is one who makes his living rightfully
(dhammena) or unlawfully (adhammena) from farming, trade, cattle-breeding,
archery, as State official, from a craft or by begging, or who takes a wife according
to law (dhammena) or unlawfully [adhammend), who goes to women of all castes
(Br?hmani, Ksatriy?, etc. as above).
In short the Buddha condemned varna samkara and thus upheld the cardinal rule
of j?ti-v?daof ensuringpurityof blood throughthe rule of endogamy. That iswhy
in the Br?hmanadhammikasutta of the Suttanipdta 315 the Buddha accuses those who
are Br?hmanas by birth but not by karma of violating or repudiatingj?tiv?da:
Khattiy? Brahmabandh? ca, ye canne

gottarakikhat?, j?tiv?dam nirahkatv?
i.e., 'the caste regulations governing the Ksatriya, Br?hmana and other gotra (castes)
were repudiated by them'. We are justified in concluding that this disapproval of
the violation of caste rules means unqualified approval of the caste system by the
Buddha.
Tosum up, there was no essential nexus between caste (varna, j?ti) and various
professions and avocations. The varnaIj?ti of a person was determined by his birth,
his parentage, his descent, irrespective of his profession or the avocation followed by
him. A person's caste in the present birth is fixed and continues to be determined
by the j?ti inwhich he is born and his caste in the next birth by his karmas. The
members of the four varnas, Br?hmanas, Ksatriyas, Vaisyas and S?dras, could lose their
caste identities in present life, that is before death, only on joining the Sarhgha.
Again the birth-based caste structure was hierarchical. Likewise the old 'Hindu'
law was discriminatory inasmuch as punishments for the same offence varied with
the caste of the offenders. The law favoured the dominant castes and was severe
48 [8]
towards the lower castes especially the S?dras and the outcastes. The lay Buddhists
accepted this caste system and the discriminatory laws.
In theKun?lavad?na of theAsok?vad?na 21A kingAsoka is creditedwith the
advice: ?v?hanak?leatha viv?hak?lej?tehpariks? na tudharmak?ledharmakriy?y?hi
guna nirmitt?, i.e., 'caste may be considered when it is a question of invitation (to
a function or dinner) or of marriage but not when it is a question of religion {dharma)
as it concerned with virtues'.
It deserves to be noticed that, according to the Lalitavistara, S?kyamuni as
bodhisattva takes birth in the family of a Ksatriya or a Br?hmana; his birth as a
S?dra or an outcaste was excluded. In the Prajn?p?ramit? s?tra II. 11.2, the Buddha,
speaking of the grave consequences of the 'Deeds conducive to the ruin of the
Dharma' says that such deeds may lead the doer to 'acquire a human body and to
be reborn, inter alia, among blind families or in the families of outcastes or refuse
workers, or among keepers of oxen, hogs or in families which are mean, contemptible
or low castes' (16). In short the low-caste births were retributory yonis, an essential
feature of sams?ra.
II. In Present-Day Buddhist Societies
la. Caste Systemamong theBuddhistsof Sri Lanka
In medieval Sri Lanka (3rd century A.D. to 15th century A.D.) the lay Buddhists
were divided socially into three major groups {vargas):
(a) kulin?: thosewho were deemed to belong to high families,noble men.
(b) hind kula\ commoners or ordinary folk who were considered as belonging to
inferior or low families.
(c) Canddlas: outcastes, now known as Rodiyas.
Occupationally also these constituted three vargas:
(a) utthaka varga consisted of landowners, farmers, Goyigama or Vell?lay?s, and cattle
breeders Gopall?. They also included the warriors and socially were kulind.
(b) hind varga consisted of two sub-groups:
(i) sippika: artisans, craftsmen such as weavers, oil pressers, potters, blacksmiths, etc.
?
(?) those who rendered service on hire pessika, such as barbers, washermen, workmen
or labourers {kammak?ra)
(c) canddla varga: those engaged in unclean, polluting jobs or jobs which caused loss
of life, that is hims?. Such professions or avocations are scavenging, hunting, fishing,
slaughter of animals, working in hides and skins, tapping of toddy, etc.
(16) The Large Sutra on Perfect Wisdom {Prajnaparamit? S?tra), ed. and transl. E. Conze, Indian ed.,
Delhi 1979, p. 289.
[9] 49
Each of the vargas were endogamous and exogamous marriages outside the varga
were taboo.
Professions were hereditary. Hence the social and occupational divisions were
also linkedwith birth.
Thecanddla varga constituted the outcastes. Intercourse with this varga was
strictly avoided by the upper two vargas. They lived in seperate hamlets (17).
It should be noted that even in present day Sri Lanka, essentially the caste system
as set out above prevails. It is significant that 'the regular servitors of the places of
worship, those who sweep the platform, carry the dead leaves, broken branches and
litter generally and keep the place in order are not only slaves but are regarded as
outcastes with whom the rest of the community will have no dealings and whose
society is contaminatory. An outcaste is a parakyum. Anyone marrying a pagoda
slave [...] becomes himself one with all the children he may have had by a previous
wife'. Images of gods are also guarded against pollution by outcastes. Hence the
inner services in temples are performed only by Goyigamas, while outer services are
performed by lay worshippers.
The rituals are performed by monks of Goyigama caste (18).
Comparing the caste system in Sri Lanka and India, Cartman observes: Though
in Ceylon, the caste divisions among the Sinhalese are different from those among
the Hindus, the farmers forming the top class, artificers among next, followed by low
castes such as barbers, potters, washermen and outcastes (Rodiyas), they have the
essential features of the Hindu caste
system, viz. endogamy, hierarchical grading
and hereditary character of castes, restrictions on commensality between members
of different castes and the concept of "pollution" of members of higher castes by
contacts with the lower, and occupational immobility. Even though the castes of
Hinduism are absent (in particular there are no Br?hmanas among the Sinhalese, the
cultivators constitute the highest class called Goyigama, and caste distinctions are
not based on any religious sanctions), the caste system embodies all the essential
characteristics of the caste system among the Hindus' (19).
lb. Caste System in theBuddhistSamgha or CommunityofMonks of Sri Lanka
According to 'Hindu' law, a layman, on renouncing lay life, that is, on becoming
a parivr?jaka, a Buddhist hhiksu or a Jaina yati, severs his connection with lay life and
(17)The above facts are derived fromW. Geiger, Culture of Ceylon inMedieval Times, ed. H.
Bechert, Stuttgart 1986, pp. 23-37.
(18)H.D. Evers, Monks, Priests,and Peasants. A Studyof Buddhism and Social StructureinCentral
Ceylon, Leiden 1972, pp. 41, 46, 62-64.
(19)Cartman, 'Hinduism inCeylon', inE.R. Leach, ed., SomeAspects ofCaste inSouth India,Ceylon
and North-WestPakistan, Cambridge 1960, p. 22.
50 [10]
thereafter has no caste or varna or j?ti. In actual practice, however, themonks continued
to be conscious of their caste ? j?ti and gotra
?
before becoming a monk. Thus in
Sri Lanka itwas a part of the ordination ceremony for joining the samgha, to enquire
about one's j?ti/gotra. After being ordained, the monks 'continued to be highly
?
conscious of their castes [...] they use expressions, their caste appellations Sal?gama
(cinnamon peeler) monk and Kar?ve (fisherman) monk. It an important historical
fact that in the 18th century the chiefmonks in Kandy refused to grant higher
{upasampad?) ordination to those monks who had non-Goyigama social origins, a
refusalwhichwas legitimizedby a royaldecree attributedtoKirti SriRajasimha' (20).
The Samgha informedtheDutch Governor (21) (1765-85) ImanWilliam Flack
that 'at present time in the Sinhalese Rata (Kandyan Kingdom) the admission of a
person of low nagaran (caste) is prohibited thoughit could not findauthorityfor this
practice in the scriptures'. In the result the membership of these elite monastic
fraternities was exclusively reserved for 'the members of the Goyigama caste, that
is the land owning aristocracy'. As Malalgodaconcludes, 'the religious profession
was in practice, if not in theory, a monopoly of the higher strata of society' (22).
'Thus the Kandyan dominated Malvalla and Asguiya nik?yas steadfastly refused to
ordain the new regions) who came into existence with the advent
elites (of the coastal
of the Europeans of non-Goyigama origins. This led to the Sal?gama (cinnamon
peelers) a caste of low land regions and the Kar?ve (fishers) and Durave (toddy
tappers) having to go to Burma to obtain valid ordination' (23).
This lead to the establishmentin the 18th centuryof Siy?mNik?ya (thosewho
had obtained ordination fromSiam) which is themonopoly of thehighest caste, the
cultivators, and the rival Amarpura Nik?ya in the 19th century which is dominated
by the relatively low 'cinnamon peeling' and 'fishing castes' (24). As Geiger (25)
observed, 'It is only in comparatively modern times that the Buddhist samgha split
into caste groups'.
2. Caste Systemamong theBuddhistsofNepal (26)
The Newari Buddha m?rgts, that is, followers of the Buddha, are laymen who are
divided into three castes:
(20)KirtisiriMalalgoda, Buddhism in Sinhalese Society,California U.P., 1976, p. 88.

(21) Ibid., p. 91.
(22) Ibid., p. 43.
(23) Ibid., p. 374.
(24) Ryan, ed., The Sinhalese System in Transition, pp. 39-41.
(25)Geiger, loc. cit. He cites J.Davy, An Account of theInteriorofCeylon, London 1821, p. 219.
(26)The factual data are based on H.A. Oldfield, SketchesfromNepal, repr.,New Delhi 1987;
M. Shepperd Slusser, Nepal Mandala: A CulturalHistory of theKathmandu Valley, Princeton 1982,
[11] 51
(1) the Banhra or Vandya consisting of nine subcastes. Those who perform
priestly functions are Vajr?c?rya (Gubhaju), the highest among the Banhras and
bhiksu or S?kya bhiksu. They have been described as 'BuddhistBr?hmans'.
All these nine subcastes of Banhras eat together, and intermarry. They are
endogamous but exogamous so far as the lower castes are concerned. Only a Br?hmana
of the Sivamdrgi community can be admitted to the rank of Banhra and perform the
duties of Vajr?c?rya. Otherwise Vandhya is a closed caste.
(2) The second caste among the Newar Buddhists corresponding to Ksatriyas,
is that of Bare or S?kya. They are high class artisans ? gold- and silversmiths,
wood and ivory carvers, sculptors, etc. They are called S?kya bhiksus (literally monks)
because in their childhood they had received ordination as bhiksus for four days
whereafter they reverted to lay life. They are grhasthas or family men.
(3) Udas (Uray): Traders and business men are divided into seven subcastes
who eat together and intermarry but not with Buddhists of lower castes. The udas
correspond to the Vaisyas among the Hindu castes.

(4) Jaffus or Jyapus (farmers and occupational castes): They are mostly agri?
culturists. They may be both Buddha m?rgis and Siva m?rgis. They are divided into
thirty subcastes. These again are subdivided into two subgroups: six (cultivators,
potters, etc.) and twenty-four (blacksmiths, barbers, cowherds, carpenters, etc.). Each
of these subgroups marry and dine among themselves. They correspond to the S?dras
of Hindu caste system, and comprise 'clean' and 'unclean' castes.
Below the three upper castes and the 'clean' s?dras among the Buddhists are the
unclean castes who are treated as outcastes. They are divided into eight groups. They
are butchers (Kasai), workers in leather (Sarki), fishermen, sweepers (Pode and Cyame
Halahuhu), Kusle, washermen etc. They are also heterodox, being Buddha mdrgi
and Siva mdrgi.
It is noteworthy that in the Nepalese Buddhist caste system, there is no nexus
between castes and professions/avocations excepting the excluded avocations which
are polluting and degrading or involve violence. Thus a Banhra or a Vajr?c?rya,
besides performing priestly functions in a Buddhist temple or performing the samskdras
(birth, name-giving, marriage, etc. ceremonies) in Buddhist households, can also
undertake avocations, crafts, which have generally been treated as hina, low, degrading
in Buddhist canonical texts and also by other Buddhist communities. Thus a lay
Buddhist Banhra can engage in agriculture, metal work (making gold and silver
ornaments, brass utensils, etc.), manufacture of implements, guns, tailoring, etc. In
short the lay Buddhists of Nepal follow a caste system essentially identical to the
Hindu caste system in Nepal.
Vol. I; Dor Bahadur Bista, People ofNepal, 4 th ed., Kathmandu 1980; Ch. von F?rer-Haimendorf,The
Sherpas ofNepal BuddhistHighlands, London 1964; D. Gellner, Monk, Householder and T?ntricPriest,
Cambridge 1992.
52 [12]
Town planning of Newar towns takes into account the caste hierarchy in which
the Newaris are organized. The houses of the highest castes are concentrated in and
around the centre where the temples of gods, and palaces and durbar of rulers were
located; the houses of lower castes are located away from the centre or core area. The
?
houses of outcastes are outside the periphery of the town outside the town wall.
Ethnic communities which are predominantly Buddhist: Tamangs, Gurungs and
Sherpas are peoples ofMongoloid stock speaking Tibeto-Burman dialects. These tribes
are exogamous.
In some settlements Tamangs are divided into b?r? (twelve) j?ts (castes) and ath?r?
(eighteen) j?ts: intermarriage between these two groups of castes is not favoured. The
offspring of marriages between Tamang men and non-Tamang women are considered
lower in social hierarchy and are not allowed 'to share the common cup' with the
Tamangs. L?m?s usually marry daughters of other l?m?s.
Likewise theGurungs aredivided intoch?r (four)j?tsand sor? (sixteen)j?ts. These
castes are exogamous, the former being higher and the latter lower. Generally the
villages settlements of ch?r j?t Gurungs are exclusive.
The Sherpas followa Lamaistic formof Buddhism. They are divided into two
endogamous groups: (a) Khadeu ('mouth good') and Khamendu ('mouth bad'). The
former are original Sherpas and the latter, later immigrants.
There is a ban on intermarriage between Khadeus and Khamendus. In the case
of mixed marriages between persons of these two groups, the Khadeu loses his superior
status and the children of such marriages are also Khamendu. Persons of superior clans
do not drink from a vessel
touched by the mouth of a Khamendu. Khamendus are
'semi-untouchables'. Likewise Yembas, persons of slave descent or freed slaves, are
treated as inferior like Khamendu.
Both the Khamendus and Yembas are debarred from being ordained as l?m?s.
3. Caste Systemamong theBuddhistsof Ladakh (27)
The Buddhists of Ladakh consist of twomajor groups: (a) Ladakhis or Bhotos

of theTibetan stock and (b) Dolba, consistingof Gara (garba,blacksmiths),Mon
(musicians and dancers) and Beda (musicians, drummers and jugglers).
The Ladakhis or Bhotos are peasants primarily practising agriculture and animal
husbandry. Before the occupation of Tibet by China, they were also engaged in trade
with Central Asia.
(27)R.S. Mann, The Ladakhi, A Study inEthnographyand Change (AnthropologicalSurveyof India),

Calcutta 1986, pp. 10, 11, 13, 16-20; ShridharKaul & H.K. Kaul, Ladakh throughtheAges: Towards
a New Identity,New Delhi 1992, pp. 153-54;Amar SinghChohan, Historical StudyofSocietyand Culture
inDardistan and Ladakh, New Delhi 1983, p. 67.
[13] 53
The Dolbas are considered as low communities (28) as they practise professions
which are considered inferior. M?ns and Bedas are generally landless and work also
as hired labourers.
These social groups are endogamous, there being no intermarriages between them.
In fact inter-caste mixed marriages, known as vama samkara in the Indian caste system,
were strongly disapproved (29), commensality is 'as a rule' absent and interdining is
occasional. The Gara, Mon and Beda on the one hand and Ladakhis on the other
avoid entering into their respective houses and prefer to stand outside their houses
because of status differential. For the same reason the lower castes avoid offering
chang (a beverage) and food (gur-gur) to Ladakhis. When a worship or feast is organized
in a Ladakhi household, the guests are seated in the hierarchical order: first the Ladakhis
followed by Gara, then Mon and lastly Beda. The tables placed before the Garas are
smaller than the tables before the Ladakhis; theM?ns and Bedas are not served food
on tables but in their hands or on leaves or in their own utensils which they carry
to the function. However, the houses of the various communities are not located in
specifically earmarked areas as in Indian villages. Thus there are no untouchables in
Ladakh and there is no segregation of different communities but close contact between
the superior Ladakhi and the inferior Gara, Mon and Beda is considered polluting.
The Buddhist Dards, called Drogpas (30),who originallycame to Ladakh from
the mountains are also another despised community.
of Gilgit, Tn the mouth of a
Tibetan the world "Drogpa" is actually an insult'. In fact the Drogpa, Gara, Mon
and Beda are 'scorned names': these names 'are given to children as protective names
in the naive belief that an evil spiritwould be scared offbecause he would take the
owner of the name
for a genuine "dirty Drogpa" or a "filthy smith"'.
It is worth noticing that craftsmen (31) are treated as inferior castes. This is in
line with the concept of hina j?ti and hina silpa of the Indian caste system in general
(28) These low communities are also called Bern. Gazetteer of Kashmir and Ladakh, repr., Delhi
1992. See also The ImperialGazetteer of India, Vol. XVI, Oxford 1908, pp. 91-92.
? The
(29)F. Grenard, Tibet Countryand Its Inhabitants,
English transl.,Delhi 1974 (repr.),p. 268,
records: T was given the instance of a man of Ladak who, having acquired considerable fortune, succeeded
in allyinghimself by marriage to an aristocratic family;during his life-time,he was able to command
more or less respect by virtue of his money; but, after his death, there was none to attend his funeral;
neither the nobles nor the commoners, whom he had disowned and who now disowned him in their
turn [...]'.
(30) S.H. Ribbach, Culture and Society inLadakh, Delhi 1986, p. 20, n. 11.
(31)The position ismore or less similar in the neighbouringHimalayan Buddhist communitiesof
Lahul and in upper Kinnaur of Himachal Pradesh. are the upper caste and Chwang
In Kinnaur, the Negis
are the lower caste. are usually ?
Damang In Lahul, the Thakurs superior to the artisan castes Gara,
Shipi, Beda who till the soil, forge agricultural implements, weave cloth, play musical instruments.
Krishan Nath, 'The Buddhist Tradition and Contemporary Changes in the Himalayas', in N.H. Samtani
& H.S. Prasad, eds., Amal? Prajn?: Aspects ofBuddhistStudies (Prof.P.V. Bapat Felicitation Volume),
Delhi 1989, pp. 491-500.
54 [14]
and of the Buddhist canonical texts in particular. Secondly the Ladakhis are migrant
Tibetans and Drogpas are an 'Aryan' people who brought civilization to Ladakh:
they introduced agriculture and village settlements with houses and castles. The M?ns
are a people of Indian origin who brought Indian culture to Ladakh by introducing
Buddhism and building monasteries. The Tibetan Ladakhis learnt agricultural practices
from Drogpas and M?ns but either drove them out or enslaved them and reduced
them to an inferior social status (32). The evolution of society in Ladakh appears to
be a re-enactment of the Aryan-Dasyu conflict in Vedic India.
4. Untouchabilityamong theBuddhistsof Burma
The of Buddhist pagodas or temples inMyanmar

attendants (Burma) keep the
approach roads or pathways to the temple, the surroundings and the temple proper
clean. They are, however, looked down upon as 'unclean'. Sir Arthur Phayre (33)
held that this was an anomaly in the Buddhist community. He observed: 'It is a
strange anomaly in the Bhoddhist religion (as it prevails in Burma) that the servitors
of temples are invariably outcastes with whom the rest of the Community will hold
no intercourse'. He goes on to add that even where servitors have become cultivators
'in consequence of their former condition, they are regarded by the people with as
much disgust as they would be from their low caste by the Hindoos'. This provides
yet another example of Buddhist societies accepting birth-based (jdti) caste system.
(32) Ribbach, op. cit., p. 30, n. 11.
(33)Gazetteer of Burma 1880, repr.,Delhi 1983, Vol. I, p. 388.
[15] 55
Ethics, Religion, and Social Protest in the First Millennium B.C. in Northern India
Author(s): Romila Thapar
Source: Daedalus, Vol. 104, No. 2, Wisdom, Revelation, and Doubt: Perspectives on the First
Millennium B.C. (Spring, 1975), pp. 119-132
Published by: The MIT Press on behalf of American Academy of Arts & Sciences
Accessed: 03-03-2015 07:15 UTC
The MIT Press and American Academy of Arts & Sciences are collaborating with JSTOR to digitize, preserve and extend
access to Daedalus.
ROMILA THAPAR
Ethics, Religion, and Social Protest in the

First Millennium B.C. in Northern India
The first millennium b.c. witnessed a spontaneous burst of new

seemingly
ideologies in areas that subsequently became nuclear regions for major civilizations.
The impression is one of a chain of apparently similar developments linking the then
known world. The geographical reach of these civilizations was relatively confined,
and allowed the formation of a network of connections resulting from conquest, trade,
and religious missions. The almost simultaneous and sustained period of speculative
thought throughout this area resulted either from the juxtaposition of a number of
seminal regions and their interconnections or from internal developments within
each society that broke the relative quiescence of the earlier bronze-age cultures.
The sixth to third centuries B.C. in northern India saw the emergence of patterns
of thought that were embryonic to the evolution of what was called in later centuries
the Indian ethos. This paper is an attempt to observe the historical anatomy of this
period and to point up the intellectual processes that gave a legitimacy .to these
patterns. The focus is narrow and concentrates on Buddhism, seen not merely as the
teaching of a single individual but rather as a wider response to a particular doctrine
and as a reaction to the changing milieu
with which it was associated.
The middle of introduces a new ideological perspective,
the first millennium
which, although touched upon marginally in Vedic literature, ismore fully developed
in the teachings of what came to be called "the heterodox sects." To the extent that
Buddhism subsumes this new perspective, it is convenient to juxtapose the polarity of
Vedic thought with that of Buddhism. The primary concern of the new attitude is
with the perception of change, the recognition that the context during this period was
different from any that had existed before. The outcome of this recognition was the
growth of ideologies that were at the same time innovative and germinal to the social
and religious philosophy and ethical thought of subsequent periods. This carried
within it the elements both of pessimism at the passing of the old order and of op
timism in having discovered a way to deal with the changed situation. The "way" as
perceived by the Buddha was arrived at through an innovation in ideology?the no
tion of causation. Causation in turn highlighted other aspects of innovative thinking,
some entirely new, others resulting from the extension of existing ideas.
To understand the perception of change at this time and the need for a new
ideology the authors of these ideas have to be seen in a historical context. The
priorities in their questions and the kinds of avenues which they explored in a search
for answers were not unconnected with the historical milieu inwhich they lived. They
appeared in response to the essentially urban civilization of the Ganges valley. This is
often termed the "second urbanization" of early India, the first having been that of
119
120 ROMILA THAPAR
the third millennium in the Indus valley. The antecedents to this second urbanization
point to a shift in geographical location from the nuclear area of the third millennium.
The Indus civilization had declined by the middle of the second millennium B.C., and
the new culture of the Ganges civilization grew and matured on the other side of the
watershed during the first millennium B.C., seemingly unconnected with the earlier
copper-age civilization. Technologically the new urbanization was based on iron, the
widespread domestication of the horse, the extension of plough agriculture, and a far
more sophisticated market economy than that of the earlier period. Until recently it
was believed that the new civilization grew under the aegis of nomadic pastoralists
speaking an Indo-European language, Sanskrit, who had conquered the existing in
habitants, possibly destroyed the bronze-age cities, and had given rise to the new
civilization in the process of settling down in the Ganges valley?thus moving, as it
were, from the age of the heroes to that of princes and traders. But fresh evidence has
suggested that this discontinuity ismore imagined than real: many aspects of the later
culture bear the impress of the earlier civilization in spite of the considerable
difference in the physicallocation.1
Technological changes were not the only indication of a new historical context, for
these changes coincided with various other developments. Tribal identity gradually
gave way to territorial identity. The territorial units, or janapadas, that emerged were
named after the janasitribes) settled thereon, such as Gandh?ra, Kuril, Pa?c?la, Mat
sya, Cedi, K??i, Ko?ala, Magadha, etc. Lineage, speech, and customary law were the
three criteria of identity and status in the earlier tribal society, with lineage being cen
tral to political control and land ownership. The ksatriya tribes were the land-owning
tribes, who belonged to either the Candravams'i (Lunar) or the Suryavamsi (Solar)
lineages that in later centuries were to become the royal lineages. The location of the
two was distinct, with the Candravam?i lineages centered in the Doab, and extending
southward and westward, and the Suryavamsi centered in the middle Ganges plain.
The separate identity of the Doab and South Bihar is evident at every point.
Cultivated land was initially owned in common by members of the k?atriya
lineage?the khattiyas of Buddhist literature?although much of the actual tilling
appears to have been done by the d?sas (slaves) and bhritakas (hired laborers and ser
vants).2 Lineage rights thus included land ownership, and lineage connections were
carefully recorded. This accounts for the predominantly ksatriya oligarchic political
organization in many janapadas.
The stress on kinship ties was further emphasized by the use of the word j?ti
('assigned by birth'). It occurs first in a late text and is used in the sense of an extended
family.3 In time, the references to jana ? tribe') decreased and those to j?ti increased,
until in Buddhist literature j?ti is used in the sense of caste, implying an endogamous
kinship group, ranked in a list of specialized occupations and service relationships
reflecting an increase in social stratification. The bi-polarity of purity and pollu
tion remained an important characteristic by varna, but this
of the classification
classification was of a more theoretical kind
involving initially four (brahman,
ksatriya, vai?ya, and ??dra) and subsequently five (with the addition of the pa?camma
or 'untouchable') groups in society, and eventually became more closely related to
ritual than to social status. J?ti slowly became the gauge of a more precise assessment
of the socio-economic status of a group, but the criteria of status continued to include
ritual status (varna).
RELIGION, AND SOCIAL PROTEST IN INDIA 121
ETHICS,
During the time of the Buddha (sixth to fifth centuries B.C. ) a major change in the
agrarian structure was the emergence of large estates owned by individual ksatriya
families; the criterion of wealth came to be associated more with land and money and
less with cattle, which had been the measure of riches in earlier Vedic literature. The
transfer of land took place largely within the same social group that had earlier main
tained joint ownership. As an adjunct to this development of a landed class, there is a
noticeable increase in the- categories of wage laborer, hired laborer, and slave. The
slave had the varna status of a ??dra, which was particularly necessary for those who
worked as domestic slaves, the more common category met with in the Indian
sources. A text of the late first millennium B.C. mentions the price of a slave as
being a hundred pieces of money; by comparison, a pair of oxen was twenty-four
pieces.4 Slaves were probably expensive even in earlier centuries and could not
therefore be used too extensively in production.
The intensification of agriculture provided the economic base for the growth of
towns in the Ganges valley. Many of the cities, apart from being important commer
cial centers, were also the capitals of the janapadas such as Kaus?mbi, K??i, ?yodhya
and R?jagriha. These were not the temple-cities of bronze-age civilization, but were
the nuclei of the affluent and the natural habitat of the setthi-gahapatis, the im
mensely wealthy traders and financiers. The flexibility of a market economy was
facilitated by three innovations?the use of a script, the consequent issuing of
promissory notes, letters of credit, and pledges, and the introduction of money in the
form of silver and copper punch-marked coins issued initially, it has been suggested,
by traders' guilds. These, in turn, resulted in the new profession of trading in
money, and the appearance of the banker deriving his wealth from usury. Unlike
the Buddhist texts, the Brahman sources disapprove of usury, although the
censure is restrictedto Brahmans' fraternizing with those who live off usury.5
Apart fromthe archaeological evidence, another indication, albeit indirect, of the
growth of cities is the rapid rise of Jainism when, with the prohibition on
agricultural professions and restriction on ownership of land, trade became the

predominant occupation of the Jainas. The discovery of new routes and the
revival of old routes were further incentives to trade.6
The city produced its own social stratification, where the sresthin ('merchant or
banker') was the most powerful and where the institutional base was that of the ?reni
('guild'). This explains why various religious sects competed for the patronage of the
sresthins.1 Yet in Brahmanical literature the trader is not included among the superior
social groups. The varna-ranking of the vai?ya ('trader') in the third position may have
been irksome to those who had such access to wealth. Furthermore, power in
Brahmanical terms was connected with ownership of land; although not forbidden to
the sresthin, land was by no means his primary source of wealth. Up to a point there
was a distinction between the urban and rural elite?the setthi / sresthin and the khat
tiya/ksatriya?because they derived their income from different sources. But some of
the khattiyas who owned estates were also town-dwellers, and thus formed another
group alongside the traders and merchants.
The guild was emerging as an essential institution of urban life, acting as a center
of both professional and kin cohesion. The recognition of ?reni-dharma (the
customary law of the guilds) as legitimate law by the end of the millennium is another
indication of the powerful status of the ?reni.8 Ultimately it evolved into an agency of
122 ROMILA THAPAR
caste organization where the larger and better established guilds took on j?ti status;
no less important was the role of the guilds in later centuries as patrons of the
heterodox sects.
The lower orders of the guild were the karmak?ras ('artisans') and the antevasikas
('apprentices'), who were nevertheless still superior to the dasabhritaka ('slaves and
hired laborers'). These, together with the cultivators, were all included in the rank of
??dras. In Brahmanical texts, their low rank was maintained by the legal fiction that
they were of mixed caste origin. The gradation among ??dras ran from the sacch?dra
('clean ??dra) to the edge of untouchability. The untouchables constituted the fifth
major group. Their untouchability derived from their being considered polluting
either because of their occupation as scavengers, such as the C?ndalas and Doms and
those who maintained the cremation grounds, or because they belonged to primitive
tribes such as the Nis?da and the Bhilla. Their speech was alien and their manner of
life was strange. Even the Buddhists despised the C?ndalas. The inequities of city life
further aggravated the degradation of these groups, already declared impure on ac
count of ritual pollution.
The rise of political authority as symbolized in systems of government and the
concept of the state were explained in a variety of ways. Vedic literature had con
nected the emergence of kingship with the emergence of government and stressed
that the qualities of leadership in battle and elements of divinity were essential to
kingship.9 By the middle of the first millennium, tribal egalitarianism had sur
rendered to the evolution of a system of government that, whether oligarchic or
monarchical, was explained as concerning itself with the problems of social disharmony,
the need for authority, and the justification for revenue collection. The Buddhist
theory emphasized the perfection of society in the pre-government age, thus imply
ing that government had become an unfortunate necessity,10 through the diffusion of
social disharmony resulting from family discord and private property. Seeking a
solution, people had gathered together and elected a leader?the mah?sammata 'the
Great Elect'?in whom they invested the authority to maintain law and order; in pay
ment for this service the mah?sammata was paid a share of the revenue. Sig
nificantly the Buddhist theory emphasizes contract and seems not to have had any
notion of royal divinity. The Mahabharata expresses a similar idea, but with a
greater emphasis on the notion that societies without governments result in anarchy;
the anarchic society is described as a state of matsy?ny?ya 'the law of the fish,' where
the big fish devour the smaller ones.11 In this theory, the king also contracts to main
tain law and order, but an element of divinity is introduced in his actual
appointment as king.
These theories reflect an increasing sense of alienation where it becomes necessary
to enforce harmony, since the pristine natural harmony of society has disappeared.
They also reflect the acceptance of the idea of authority based on power and not
necessarily on kinship alone. The janapadas were coalescing into territorial states. By
the fifth century B.C. competition for power had already developed among the
stronger of the major janapadas, such as K??i, Ko?ala and Magadha, where even close
kinship ties were ignored to further political gains. Magadha was to emerge as the
most powerful, ultimately becoming the nucleus of the Mauryan empire, which was
built on the conquests of Chandragupta Maurya in the fourth century B.C. and com
prised, during the reign of his grandson A?oka, almost the entire Indian sub-continent
ETHICS, RELIGION, AND SOCIAL PROTEST IN INDIA 123
and eastern Afghanistan. With the growth of political authoritarianism and a com
plicated state machinery, it is not surprising that the justification for the emergence of
government came to be based on the necessity for taxation and the need to maintain
law and order.12
Two co-existing systems of economic redistribution came into being and
sometimes into conflict as well. One, at the level of the state, derived its income from
taxes, tributes, and fines and redistributed it through awards, salaries, grants, and ex
penditures on public works and ceremonies. But the redistribution was not equitable,
and the prestige economy, particularly in the monarchical states, consumed a large
part of the income. The second system, on a lesser scale, was confined to the
merchants and bankers of the cities; among them the ethic of redistribution was such
that substantial sums were retained as capital for further investment.13 They were
doubtless irritated by the prestige economy of the state. That the second system could
function in the cities points to their more diffused political authority; this is also
suggested by the absence of citadels in these cities. To some extent money liberated
the financier from overarching political control.
Caste structure at this time grew out of a variety of interrelationships between
groups. The purity-pollution dichotomy, which above all demarcated the Brahman
from the untouchable and which was absent in the earlier period, is by now well es
tablished. The ?rya-d?sa dichotomy deriving from ethnic, linguistic, and cultural
differences in the Rg Vedic texts was now replaced by the ?rya-??dra dichotomy,
where the ethnic differences are minimal and the main criteria are the use of Sanskrit
and the observance of the varna rules. Non-?ryas are mleccha ('the barbarians' or 'the
impure' ) and are generally ranked as ??dras except in later centuries, when foreign
conquerors such as the Indo-Greeks had to be given the dubious status of "degenerate
ksatriyas." The formation of new castes, theoretically resulting from the intermixing
of the original four, was probably a more open system than has hitherto been
recognized. The evidence from subsequent centuries suggests that newj?tis arose as a
result of incorporating tribes and guilds and, still later, religious sects into
caste society.
The complexity of the new society is clearly reflected in the need for codifying the
laws of the various social groups, which is what is aimed at in the Brahmanical
dharma-s?tras. The purpose of the laws is to differentiate between the various social
groups generally identified as those oijana, jati, and varna. These, however, are made
part of a cohesive view of society. There is an implicit belief that the demarcation of
differences would lead to a resolution of tensions, an attitude that could only have
been feasible in the absence of a situation of conflict. Also implicit in the dharma-s?
tras is the Brahmans' claim to being the arbiters of the law. There was no overt
challenge to this claim since the codification did not aim at a uniformity of laws, but,
on the contrary, to the recognition of their diversity. The Buddhist social code, on the
other hand, stressed broad ethical principles of general application to a variety of
social groups in an attempt to integrate earlier as well as more recently emergent
groups into the new patterns. The integration was easier at the theoretical level. At
the practical level there was a tendency to separate ritual status (varna) from actual
status (j?ti). Social roles were not entirely dependent on the one or the other. The
older traditions and norms were thus placated, and the new entrants into the social
hierarchy were not entirely disappointed. However, the demarcation was in fact by no
124 ROMILA THAPAR
means facileor simplistic. Many of the later subtleties and intricacies of caste
relationships emanate from this early attempt at demarcation.
It was apparent that a condition of permanence was neither feasible nor possible
in the world of reality where all was flux. Even the above brief survey of the historical
scene shows that the condition of constant change could not be ignored.14 It affected
the assumptions of the philosophers of the time and is still reflected in the prevailing
intellectual systems. The consciousness of change is perhaps seen most clearly in the
fundamental problem of human salvation or liberation in which three interrelated
aspects were emphasized?the ethic of the individual in terms of his own moral con
sciousness and his search for release from the bonds of human existence, the verifica
tion of ultimate knowledge so essential to the working out of a means to salvation, and
finally the discovery of a path to salvation. The prime motivation was to find an
answer that would subsume changing material conditions and yet remain viable. The
Buddhist attempt to analyze these problems makes a point of contrasting the attempts
of other groups of thinkers similarly involved.
That these concerns were widespread is apparent from the rise of a variety of
"heterodox sects," among which Buddhism was included. These sects were not
merely a reaction of Vedic religion, as is often suggested, because within the Vedic
Brahman framework there had also been a diversification of views as evidenced by the
texts of the Upanisads and the ?ranyakas. These were the discourses of the renouncers
who had isolated themselves from society and lived in forest retreats. They stood
apart, disenchanted with the world, seeking ultimate truths. Their discourses show a
liberation of the speculative consciousness from the burdens of magical sacrifice and
ritual. However, the universalistic basis of their thinking had some limitations. They
recognized the need for individual salvation. In isolation and through sanny?sa
('asceticism'), the individual could find his moksa ('salvation') which would release his
?tma (individual 'soul') and enable it to unite with the brahma ('all-soul'). Asceticism
was motivated both by a desire to escape from the insecurity of a changing society and
by the conviction that meditation was an effective means of acquiring the knowledge
that furthers self-realization as well as the power (tapas) to become superior even to
the gods. Gradually asceticism came to be regarded as a more powerful force than
sacrifice, thus admitting the ineffectiveness of a community attempt to reach moments
of magic and power. Perhaps more important, asceticism resulted in total freedom, a
break with family ties and social regulations, provided sexual needs could be sub
limated. Hence the correlation between asceticism and asexuality. This freedom in
sured the renouncer a moral status far higher than that of even a sacrificing Brahman.
Some sects, such as the ?jivikas, based their thought on determinism and saw
renunciation as the only and ultimate path to moksa. The Buddhists and the Jainas
had both philosophical and social concerns. Access to knowledge did not lie through
the authoritative voice of the Vedas, for what is not personally verifiable is unaccep
table. Nor is skepticism a path to knowledge; the skeptics for the Buddhists were "eel
Even asceticism was not possible as a path to salvation for everyone. Both
wrigglers."
the Buddha and Mahav?ra, though seeking enlightenment through isolated medita
tion, nevertheless returned to the world of the cities and villages to preach the path of
salvation to the householder who could not become a monk owing to his social
In the case of the Buddha, the emphasis on "the middle way," the path
obligations.
devoid of excesses, emphasizing moderation and a moral life, was indicative of his
concern that the path suggested by him be compatible with the real problems of social
existence. Not surprisingly, the early supporters of Buddhism were not only the
ascetics but also, and in larger numbers, the setthis and members of the ksatriya clans.
At the other extreme were a number of lok?yata sects, particularly the C?rv?kas, who
were based primarily in the towns and who taught a thoroughgoing materialism,
such as the teachings of Aj?ta Ke?akambalin, the end result of which was seen by
others as an idealization of hedonism.
The thread of social protest winding through these heterodox teachings was in
dicative of a perception of change: of existing change, the recognition that further
changes were imminent, and toward change itself. For the Buddhists, change was
symbolized in two strands, which occasionally intertwined, the cosmic and the
historical. The universe is transient and in a state of continuous flux. Buddhist cosmic
time was cyclical, starting with a pristine Utopian society, which had gradually
decayed and was slowly reaching its nadir of sorrow and suffering, the direction in
which contemporary society was moving. Eventually it would rise upwards again and
begin a Utopian phase. Brahmanical sources, also positing cyclical time, attempted a
mathematical measurement of it, albeit of an infinite magnitude, as did the Buddhists,
who indicated infinite eons by spatial descriptions.15 Time was seen as an unending
continuity of which historical time was but a fraction. Within this continuity the in
dividual consciousness also moved unceasingly from one lifetime to the next birth un
til liberated from the chain of rebirth. It is compared to the flame of a lamp used to
light another lamp and so on, ad infinitum. In each case the flame of the lamp is both
the old and the new flame, and so it iswith the perception of change in the continuity
of time.
Change, therefore, cannot be seen as a sudden break. But within historical time
there is a far sharper awareness of the past and the future. Other "enlightened ones"
have trod the same path in the past as the Buddha. Was this allegorical, or was it a
reference to earlier teachers with a similar doctrine? There is also the reference to the
Buddha Maitreya, who will reawaken the world to the Dharma ('the law) many cen
turies after the present Buddha.16 This was to develop in the first centuries after Christ
into an almost messianic movement within Buddhism, no doubt further stimulated by
contact with the messianic message of Christianity and Manichaeanism. The decline
from Utopian beginnings is not accidental. There is a concern with moral decay,
which, although partially inherent (the very state of nature having evolved from
luminosity to dross), is nevertheless caused by changes in the material content of life.
It can be circumvented to some extent by the individual's choice in the manner of
adapting to changing social situations.
Central to the awareness of change is the law of causality, and it is around this that
much of Buddhist doctrine revolves, claiming to derive from rational arguments and
examples. At the individual level, the interconnection between desire, suffering, and
rebirth is explained by causality. The elimination of dukkha ('suffering') lies in the
elimination of tanha ('desire'), and this can be achieved by observing the precepts of
the Dhamma/Dharma ('the Law as taught by the Buddha') and the eight-fold path.
Social change is also explained by causality and becomes a part of the underpinning,
as it were, of the universal applicability of the Dharma, for, once the causal connec
tion is known, change comes under human control.
This led to a new perspective on the significance of the individual. The heterodox
126 ROMILA THAPAR
teaching, and Buddhism in particular, turned the earlier perspective inside out, and
the focus shifted to the individual rather than the social group to which he
belonged.
Up to a point this encouraged a nihilistic trend, as in the case of the ?j?vikas. But
nihilism was not characteristic of all sects. On this question the central core of the
Buddhist Dharma is very clear. Where renunciation or opting out is not feasible, the
individual, whatever his social status, had the choice of becoming a lay disciple and
observing the rules of "the middle way." Furthermore, the moral responsibility of the
individual was seen in the choice of action made by him through his chain of rebirth.
The Brhad?ranyaka Upanisad described rebirth as consisting of samsara, the
transmigration of souls, to which was added the notion of karma ('action' ), the out
come of the activities of one life
affecting the next. The Buddhists modified the notion
of samsara to exclude the soul and to refer to consciousness as the element that con
tinues, and they appropriated the doctrine of karma in its entirety. Thus not only
was the individual responsible for the nature and condition of his present and
future lives, but the doctrine of karma also became a useful means of explaining
the origin of social inequality and the creation of caste society.17 Not only was a man's
social condition a reference point in social justice, but disease, physical pain, and
even death were seen as aspects of social justice, although the moral responsibility for
this condition rested with the individual. Thus the sting of social protest was
numbed by insisting that there was no tangible agency responsible for social in
justice, or even an abstract deity against whom man could complain, but that re
sponsibility belonged with man himself. This in turn tended to curb nonconformity
in behavior for fear of the consequences in the next life.
It is notaltogether fortuitous that Buddhism was popular among the en
trepreneurs and the life-affirming groups in Indian society?the merchants and the ar
tisans. Nor should it be forgotten that at the political level Buddhism registered its in
itial success in the period of the first empire, that of the Mauryas. The life-asserting
aspect of karma is that, if the rules are observed, the next birth can at least bring a
better and more prosperous life, if not freedom from the chain of rebirth. There can
be, therefore, considerable motivation for observing the rules. That the onus was on
the individual is further emphasized by the necessity of being born a man, rather than
any other creature, before salvation can be attained. Moral responsibility was not
developed into a philosophy of radical change, which would have meant challenging
the existing system. The Buddha made a distinction between caste as the frame of the
socio-economic structure, which he accepted, and the notion of the relative purity in
herent in the upper castes, which he rejected.18 The emphasis was on an individual's
choice of an ethic, but the end result of this had its social implications.
Fundamental to Buddhist teaching was the notion of the interplay of acts of merit
(punya) and demerit (p?pa, literally 'evil, wickedness'), and punya becomes central to
ethical thought from this time onwards. The constituents of merit for the layman are
activities motivated by the need to further social good, such as harmonious social
relationships and charity, but, above all, sexual control and non-violence. Harmony
in social relationships referred not only to those between parents and children, but
also between master and slave, and employer and employee in general. This had a
clear correlation with the large estates of the khattiyas and the new urban culture.
Although the Buddha associates the growth of evil in the world to (among other fac
tors) the institutions of the family and private property, both of which, he argues, en
couraged sentiments of possessiveness and consequently aggression, he nevertheless

projects an undisturbed continuity for both institutions. In spite of its evils, the family
did weaken the sense of alienation, and hence there is a stress on respecting
kinship
ties. Charity was seen not only as a means of the suffering of the materially
alleviating
poor, but also as the giving of gifts (d?na) especially to the Sangha (the order of
monks). This had the additional advantage of strengthening the monastic organiza
tion and its relations with the lay community.
Both sexual puritanism and non-violence became controversial issues in the
debate among the various sects. The Buddha was not loathe to accept the devotion of
the more renowned and accomplished courtesans of the towns, such as Ambapalli, but
it took considerable persuasion for him to agree to admit women into the Sangha.
ties were a obstacle to and women were
Family major renunciation, symbolic of these
ties. Yet during this time it was only in the Jaina and Buddhist orders that nuns were
permitted, and the women were drawn largely from urban society and the
royal
households.
Non-violence (ahims?), the central focus of Buddhist and Jaina ethics, was less im
portant in other religious sects. Veiled, ambiguous references can be culled from the
Upanisads, but the exposition of the idea as an ethical value was that of Mah?v?ra and
the Buddha. The Jaina of ahims? appears to be an extreme position in
understanding
volving all created beings and the attempt to preserve them from destruction,
whether deliberate or accidental.19 The Buddhists tend to stress the ethical question of
man's actions in furthering or preventing violence.
Ahims? can be viewed in association with many facets of contemporary life. It has
been seen as an objection to the sacrifice of animals during the yaj?a, the sacrificial
ceremony essential to the Vedic Brahmanical religion. There is repeated mention of
the futility of killing animals as a religious ritual.20
Possibly this coincided with the
rapid transformation of pastoral groups into agriculturalists, which resulted in a
depletion of animal wealth.21 The debate on the inviolability of the cow is referred to
en passant in the Satapatha Br?hmana, but it is again largely due to Buddhist and
Jaina disputation that the prohibition is extended from cattle to violence per se.22
Ahims? can also be explained as a reaction among the new urban groups to the
prestige economy of non-urban societies, who were willfully destroying wealth to no
purpose. Sacrifice, it was argued, is essentially an offering; consequently it lies not in
the destruction of life but in the embodiment of moral values that become the founda
tion for ethical behavior?in
honoring parents, in honoring all the members of the
household from the highest to the lowest, in having patience, meekness, and self
control. The values listed are both conservative and conciliatory. Yet the element of
radicalism in this view is the inclusion of slaves and workmen as deserving of honor.
At another level ahims? would have suited those who were inter-tribal
discouraging
warfare and encouraging the expansion of settled agriculture and trade?activities
from which both the khattiyas and setthis stood to
gain.
Ahims? also included a discouragement of the use of coercion and violence to
justify political authoritarianism?very pertinent to the transformation of the
janapadas into kingdoms laying political claim to large territories. The suspicion of
political authoritarianism may have to do with the fact that the heterodox sects often
had their genesis in the relatively more egalitarian tradition of the oligarchies and
republics, such as that of the S?kyas and Vrjjis. Those most directly affected by war
128 ROMILA THAPAR
would bethe cultivators, whose fields were the prey of marauding armies, and the
traders, who would be unable to transport their goods, or, even worse, whose centers
of production would be destroyed in the devastation of a town?so often the sym
bolic final act of a successful campaign. Possibly ahims? could also undermine the
ritualized wars?the campaigns that were fought subsequent to the a?vamedha sacri
fice, when a king claiming sovereignty over a region would release a sacrificial horse
and would then be duty bound to conquer all the lands over which the horse wan
dered. Implicit in ahims? at the political level is an objection to even the legitimate
use of coercion (danda) by the political authority of the state. The king in his role as
protector should avoid coercion, modeling himself after the ideal universal
monarch, the cakravartin, who is adanda 'not to resort to coercion.'23
having
Ahims? might have had an ameliorative influence in situations of tension, which
were by no means rare. Ultimately there was also the ethical and philosophical level.
Conscious non-violence (not to be confused with cowardice) was expressive of the
highest ethical stand. The credibility of non-violence can only stem from a belief in
man's innate virtue. It has been argued that the Buddha's ahims? represents the
negative philosophy of pacifism. To the extent that the Buddha was not preaching
rebellion, but rather a conciliatory ethic, as a solution to social ills, the negative aspect
of pacifism can be justified. But if ahims? arose from an awareness of varying levels of
comprehension and reaction, then pacifism alone cannot be the complete explanation.
As a method of social protest, the objection even to ritual sacrifice involving the
destruction of life takes on an active and affirmative role, as is evident from the con
tinuing debate on this subject up to recent times. It is perhaps also worth remember
ing that the Brahmanical insistence on vegetarianism dates to the post-Buddhist
period.
The significance of renunciation has its own role in the Buddhist moral position
and relates to the moral and political authority of the renouncer. There has been a
tendency to see renunciation as a purely life-denying process. This it may be if the
renouncer moves away from society and lives in isolation, though, even here, the
negative aspect to the isolation is rarely foremost. But if the renouncer, after a period,
of isolation, resumes a function in society, in spite of his having renounced his ties, his
influence can become both powerful and positive. Moral and political authority are
separated and the former becomes the censor of the latter. This separation can be
crucial to the establishing of an independent intellectual tradition, as was the case in
the lifetime of the Buddha, provided that the independent relationship between the
two is not eroded by the requirements of patronage. If the renouncer is also in sym
pathy with the aspirations of a community and if he comes from a social background
not generally associated with life-denial and renunciation, but rather with political
authority and social status (such as the khattiyas of the time), his moral authority is
almost unlimited. In such situations the renouncer forsakes one life-style to take on
another.
Recruitment to the heterodox sects was not limited to any particular group. Those
who had an organized body of adherents, enlisted monks, and built monasteries en
couraged people of all castes to join the organization and, in theory at least, did not
bar any caste. In the Buddhist Sangha the adoption of a new name by the monk was
symbolic not merely of a new birth in the Sangha but also of a removal from his caste
and status.24 The proximity of all castes within the monasteries ran counter to the
Brahmanical ideal of the segregation of castes in daily living. Commensality among

such monks from the lay community broke the food taboos so essential to the varna
system. Outside the monastery and among lay followers the problem of integrating
social groups remained.
Each j?ti had its own religious observances. Religious differentiations were
preserved through the mechanism of caste, as were the observance of rituals per
taining to local cults, some of which were assimilated into Buddhist practice.25 The
Buddhist shift of emphasis from deities to the more abstract notion of Dharma ('the
law' )was an attempt in part to undermine these religious differentiations. It may be
argued that the absence of a deity in Buddhism inherent in the doctrine strengthened
the idea of a universal religion. In some ways, however, Dharma almost took on the
characteristics of an omniscient presence symbolized in the turning of the wheel of
the law. Dharma was the eternal Law?ultimate, timeless, temporal, transcendent,
immanent. In spite of changing human society the Law remained changeless. It in
tegrated within itself the ethic of the individual, the verification of ultimate
knowledge, and the path to salvation. Change was perceived, recognized, and un
derstood. But it was not the changed situation that was to be subjected to radical
social alteration as much as the law that was to be applicable in all situations. The
law was above the particular and was universally viable. Enlightenment lay in the
discovery of this law and identification with it. But the law was not to be kept to
oneself in isolation. The enlightened ones must return to the cities and the villages
and preach the law. There is the repeated parable of the raft, where he who has dis
covered the law (the raft) must help others to cross the waters on it.26
The arbiters of the Dharma were the Elders of the monastery. The Buddhist
monastery was both a retreat for meditation and an institution for action. The early
monasteries had to be located close to large concentrations of population, because of
the requirement that the monks feed off alms, since they were forbidden to do any
manual labor including cultivating their own food. Begging for alms and preaching
the doctrine brought them into contact with the lay community. The Sangha was thus
a collection of renouncers but not of ascetics. The monks took on a new way of life
based on communal sharing and dedication to poverty, evidenced by the prohibition
against personal possessions and by the name they adopted, bhiksu 'mendicant.' Cen
tral to the organization of the Sangha was the emphasis on the equal status of every
monk, influenced perhaps by the more egalitarian political organization of the
oligarchic janapadas, familiar to the Buddha. But this insistence on equality did not
apply to the world outside. It was almost as if the creation of a radical, egalitarian
society within the monastery exhausted the drive toward such a society in the world
outside, or at least weakened the urgency of radical change?assuming, however, that
this had been intended. Celibacy and the discouraging of manual labor for the monks
point up the bi-polarity between the monk and the householder where the latter
qualifies himself in part by his ties to family and property.27 The monk and the
householder lived in worlds apart.
The Sangha gradually acquired a strong sense of mission. This is evident from the
frequency of the councils determining the true doctrine and the splintering off of
sects within the Sangha after the death of the Buddha, each claiming to represent the
true doctrine. Even more significant was the system of maintaining records and
historical accounts not only of the major events in the history of Buddhism, but also of
130 ROMILA THAPAR
the more important sects within the Sangha that encouraged the polemics of Buddhist
sectarian thought. The community of renouncers was not altogether unaware either of
its political role or of its role in the new ethic that they were promulgating.
The sense of mission was encouraged by the literate monks. The monasteries
developed into centers
of learning. This was again a point of opposition to the Vedic
Brahmanical approach for which literacy was the preserve of the socially-determined
few and which in any case laid greater stress on the oral tradition. As a part of the
appeal to the wider audience, the Buddha preached not in Sanskrit but in ardha-m?
gadhi?a pr?krit of the middle Ganges plain. At the same time as Brahmanical culture
was seeking an ?rya identity and exclusivity, the Buddha was breaking away from it.
The extension of literacy was symbolic of much that the new ideologies stood for,
the insufficiency of faith and ritual and the incorporation of reason and moral action in
a manner that would have wide applicability to large numbers of people of diverse
social origins. The new teachers arose as individuals and not through an institutional
base. But the continuance of the new ideologies required the building of their own in
stitutional base. The perception of change and the need to come to terms with itwere
not seen as synonymous with a radical ideology in favor of a total change. The
Buddhists, for example, were more analytical than earlier thinkers in their views on
man and society, but they did not feel it necessary to suggest a complete reorganiza
tion of the social structure. To that degree, Buddhism in its historical role touched the
chords of social protest but went no further. This was perhaps because the groups for
which it was projecting a new ideology ceased to be the protesters at a certain
historical point and became the heirs. The element of social protest in Buddhism was
therefore limited to providing the intellectual encouragement and justification for the
formation of a new elite. It can be argued that in the historical context of those times
even this was a radical position and it was not necessary to extend causation to its
logical limits. The lok?yatas who insisted on natural causation and opposed the doc
trine of karma were either subsumed into the new system or were left on the fringes
as anarchists.
The historical mission of Buddhism took it far afield. The monasteries, irrespective
of sectarian differences, acted as networks of acculturation and contact within the In
dian subcontinent reaching out into the remotest corners, monks traveling either
in isolation or accompanying the traders. In the first millennium A.D. the sig
nificance of the mission of Buddhism was that it acted as a catalyst in many
parts of Asia. Its major orientation was in Central Asia, China, Japan, and Southeast
Asia. The period when Buddhism took root and prospered in these new areas
coincided with its fading in the country of its origin. Can this be regarded
as a historical demonstration of the Buddhist notion of change and continuity?
the analogy with the flame of one lamp lighting the flame of another before
being extinguished?
References
1
evidence of the post-Harappan period, particularly in Gujerat, Malwa, the Banas
Archaeological
valley, and parts of the watershed and upper Doab, points to some continuities of cultural traits from the
Harappa culture. Small settlements of primitive agriculturalists in the Doab or the western Ganges plain
(the Ochre-Color Pottery culture) were superseded toward the end of the second millennium B.C. by
larger settlements of more advanced agriculturalists gradually taking to iron technology by the earlier
part of the first millennium B.C. (the Painted Gray-Ware culture). Further east, in the middle Ganges plain
and south
Bihar, the impetus for using an iron technology is associated with an apparently different
group of people (the Black- and Red-Ware culture), who appear to have had some links with western
India via the northern part of the central Indian plateau. The Doab was the geographical focus of
the later Vedic literature and was identified (in the main) in Brahmanical literature with the ?rya-varta
or the land of the ?ryas (the pure, respectable people), those who spoke Sanskrit and observed the
caste laws. South Bihar, which included the territory of Magadha, was
to a greater extent the
geographical focus of "the heterodox sects." In the Buddhist and Jaina texts, south Bihar was
the core of the ?rya-varta, since these texts tended to give a more easterly location to the "pure land."
An area of high precipitation, the Ganges plain was at that time covered with forests. It has been argued
that settlement on any appreciable scale would have been virtually impossible before the introduction of
iron technology, the monsoon forest being relatively impervious to the tools of copper technology. That the
introduction of iron coincided with an increase in population is clear even from fairly impressionistic
archaeological data. The iron-age precondition to urbanization is evident from the number of settlements
of iron-using cultures that developed into towns. Increase in population not only assisted in the clearing of
more land for agriculture in the Ganges plain, but could also have acted as a lever toward encouraging a
change to iron technology and more particularly to plough-use agriculture.

2
Pata?jali, Mah?bh?sya on P?nini IV.1.168
3
K?tyayana ?rauta S?tra XV.4.14; XV.2.11
4
Nanda J?taka 1.98; Gamani Canda J?taka 1.207
5
?pastamba Dharma S?tra 1.6.18.22; Baudh?yana Dharma S?tra 1.5.93-94.
6
Maritime trade with west Asia was revived in the first millennium B.C. Close contacts were established
between Iran and northwestern India. Within the sub-continent, overland and maritime routes to the south
(Daksinapatha) were
being explored.
7
N. Wagle, Society at the Time of the Buddha (Bombay, 1966).
8
Manu Dharma??stra VIII.41
9
Rg Veda VIII. 35.17; 86.10-11
10
Digha Nikaya III 93
11
Mah?bh?rata, S?nti Parvan 67.19-24.
12 a tribute or booty and even
The words used for the two basic taxes were bali, originally meaning
tually coming ato mean
tax on land, and bh?ga, meaning 'a share' and applied to the produce of the land,
reflecting the more stable distribution of settled times. An early term for the king was bh?gadugha,
literally, 'he who milks the share.'
13
As suggested by the Buddha in Digha Nik?ya III, p. 188 (P.T.S. ed).
14
For a correlation of material conditions with the rise of Buddhism, R. S. Sharma, "Material Milieu of
the Birth of Buddhism," paper presented to the Twenty-Ninth International Congress of Orientalists, Paris,
1973. The Proceedings of the Congress have not yet been published.
15
Samyutta Nikaya, 15ii. 178-193
16
Digha Nik?ya III 76
17
Majjhima Nikaya 1.289; A?guttara Nik?ya V.288-291. This ismade even more explicit in the preemi
nent text of Brahmanism, the Bh?gvad-G?t? IV. 13, composed in the period after the Buddha.
18
A?guttara Nik?ya III.214; Samyutta Nikaya 1.167; Majjhima Nik?ya II 128-30
19
Ac?ra?ga S?tra 11.1.1 -4.
20 *
A?guttara Nik?ya IV.42-45.
21
D. D. Kosambi, The Culture and Civilisation of Ancient India in Historical Outline (London, [1965]),
p. 105. Cattle provided both labor and fertilizer in agricultural societies, and any depletion was a serious
132 ROMILA THAPAR
loss. That cattle were singled out for protection is clear from the emphasis in some sections of the Buddhist
texts, such as Majjhima Nik?ya, 1.220.
22
The is perhaps
debate best symbolized by passages in the early text, the Satapatha Br?hmar\a
III. 1-2.21 the eating of meat
where (the flesh of the cow is the case in point) is defended by Yaj?avalkya,
who represents an important point of view, and the later prohibition on it, as for example in the text of
Manu dating to the first century B.C.
23Cakkavati Sinhan?dasutta, Digha Nik?ya III, p. 58 (S.B.B. ed).

24
Vinaya Pi\aka 11.239; A?guttara Nik?ya IV.202.
25
Dates for important events such as the vassa (the rainy season when the monks had to return to the
the uposatha (the days for the hearing of the confession of the monks), etc. were calculated on
monastery),
the basis of the lunar calendar, although the solar calendar was also in use at the time.
26
Majjhima Nik?ya 1.134-5; Samyutta Nik?ya IV. 179-181; A?guttara Nik?ya 11.201.
27
S. J. Tambiah, Buddhism and the Spirit Cults in North-East Thailand (Cambridge, 1970), pp. 81 ff.
Revaluation of Tradition in the Ideology of the Radical Adivasi Resistance 30
Article
The Making and Studies in History

31(1) 30–50
Proliferation of Jāti: © 2015 Jawaharlal Nehru University

SAGE Publications
sagepub.in/home.nav
A Historical Inquiry DOI: 10.1177/0257643014558461
http://sih.sagepub.com
Rajan Gurukkal1
Abstract
The article seeks to argue that the making of jāti was homologous with the open-
ing up of deltas for agriculture, involving integration of agro-pastoral descent
groups into hereditary specialists of occupational identity, and formation of
stratified relations of production transcending kin labour. It emphasises that
productive relations in the deltas had preconditions such as hereditary occupa-
tions, asymmetrical social relations, amenability to differential allocation of sta-
tus, and the dominant presence of the Brāhman.a-s with tacitly recognized ritual
supremacy, resource potential, social control, political influence and cultural pre-
eminence for the emergence of jāti hierarchy. A related argument is that the
emergence of hereditary occupation groups and promulgation of sāstraic norms
must have been processes of mutuality and concurrence. It has been understood
the context of the Jāti institution of coercive control and seigniorial jurisdiction
over the labouring body was crucial. As regards proliferation of jāti-s, the argu-
ment is that it had been an ongoing process ever since jāti became the dominant
paradigm of identity construction for occupational groups and service personnel
claiming socio-cultural distinction on the basis of their association with the seig-
niorial power. Rewarded under land-tenure, the personnel in service to the king
and the local chieftains became hereditary for stability of service as well as per-
manence of family landholding. Illustrating the historical experience of the Tamil
South in general and the Kerala region in particular, the argument found feasible
is that the proliferation of jāti-s happened as a land tenure based phenomenon
under the three seigniorial streams represented by the king, the chieftain and the
temple – brahmadēya combine, as realized in terms of the sāstraic norms.
Keywords
kin-labour, descent-groups, productive relations, hereditary occupation, seigniorial
power, jāti
1
Visiting Professor, Center for Contemporary Studies, IISC, Bangalore, India.
Corresponding author:
Rajan Gurukkal, Visiting Professor, Center for Contemporary Studies, IISC, Sir. C.V. Raman Avenue,
Bangalore 560 012, India. E-mail: rgurukkal@gmail.com
Gurukkal 31
It is a bit outlandish to go about inquiring into the origins of institutions of long

continuity, such as jāti in the postmodern context of knowledge production, which
offers no special immunity to historians’ teleological exercise. Although an
attempt to discuss the historical process of the formation of jāti is sure to be a
tenuous exercise yielding no new knowledge, it is fruitful to contextualize the
incidence and proliferation of jāti practice by conceptualizing what the institution
meant and how it functioned in the socio-economic past of southern India. This
article, therefore, seeks to contextualize the incidence of the institution of jāti
against the historical background of the rise of wet-rice agriculture in the deltas
involving integration of agro-pastoral artisans, craftsmen and tillers; the coming
into being of a new system of ownership and control; the formation of a class-
structured society; the institutional manifestation of the jāti as a strong fetter of
productive relations; and the socio-economic process of the proliferation of jāti.
Meaning and Context

The etymology of the Sanskrit word, jāti, gives the meaning, ‘that which one is
born in’ (jāyatē asmin iti). It denotes a trans-familial identity of relative status
commonly shared by many families. Extensive use of the English word caste
adapted from the Portuguese ‘casta’, meaning breed has not only made the mean-
ing of jāti simple and its historical process overlooked, but also internationalized
the misconception that the institution is best exemplified in India even to the
extent of its being unique there.2
Analytically, the notion of the self and the other is perhaps the property, most
archaic and fundamental, to jāti, which comes from pre-jāti descent communi-
ties with the habit of distinguishing each as the pure from the impure other. So
long as every descent group had autonomy to sustain the notion, it was mutu-
ally exclusive, horizontal and fluid. What happened in the historical process of
the constitution of jāti as a system of social organization across descent groups
was imposition of the dominant other’s notion of the pure and impure, upon all,
2
Most studies in caste vouch for this, despite their original contribution to knowledge in the subject
matter. To cite some of the major studies, C. Bougle, Essays on the Caste System (first published in
1908, republished Cambridge: Cambridge University Press, 1971); E.A.H. Blunt, The Caste System
of North India, first edition in 1931 by Oxford University Press, London, new edition by S. Chand
Publishers, Delhi, 1969; G.S. Ghurye, Caste and Race in India (London: Kegan Paul, Trench, Trübner,
1932); J.H. Hutton, Caste in India: Its Nature, Function and Origins (Cambridge: Cambridge
University Press, 1946); F.G. Bailey, Caste and the Economic Frontier: A Village in Highland Orissa
(New York: The Humanities Press, 1958). A. Beteille, Caste, Class and Power: Changing Patterns
of Stratification in a Tanjore Village (California: University of California Press, 1965, rev. edn,
New Delhi: Oxford University Press, 2010); E. Leach, ‘Caste, Class and Slavery: The Taxonomic
Problem’, Caste and Race eds., A. De Reuck and J. Knight (London: J & A. Churchill Publishers,
1967); L. Dumont, Homo Hierarchicus: The Caste System and Its Implications (Chicago: University
of Chicago Press, 1980). J. Wilson, Indian Caste, 1 (Delhi: K.K. Book Distributors, 1985); G. Harold,
The Hindu Caste System. Vol. 1, The Sacralization of a Social Order (Delhi: Chanakya Publishers,
1987); also D. Quigley, The Interpretation of Caste (New Delhi: Oxford University Press, 1993).
32 Studies in History 31(1)
and the fallout, was a structure inevitably vertical, hierarchical and rigid. Thus
structurally, the jāti consisted of two sets of binaries as its constituents—the
pure >< impure and the higher >< lower. Each jāti sustained its historically con-
tingent notion of pure and impure but to be applied solely in relation to the lower
in the hierarchy. It is this zeal of each jāti to brag about its status as far above the
lower though just below the higher, which has been ensuring solidarity across
the unequal in the hierarchy.
It is this significant factor of the pure and impure as well as the high and low,
which persists all along the course of the formation, consolidation and prolif-
eration of the jāti, irrespective of the causal and contextual differences of each.
The formation of the jāti had its causal and contextual links with the spread
of plough agriculture into the delta; while its institutional consolidation was
linked to the productive relations’ systematization into a hierarchical social
structure; and the proliferation linked to the expansion of deltaic agriculture.
The pure >< impure and the high >< low binaries do simply map on to one
another at all the three phases despite the steady development of a series of
complex intersections.
Jāti is rooted in the institution of kulattoḻil or the system of kulam (descent
group)-based division of specialized occupations (toḻil), which gave rise to descent
groups of occupational identity.3 The most crucial factor about jāti was the institu-
tion of labour realization and the entailing form of servitude that it embodied. As
an institution it ensured permanence of labour to contemporary productive rela-
tions by fettering the descent group in specific crafts, which although provided
occupational stability to them, denied the natural right to adopt the livelihood of
their choice. It was reproduction and perpetuation of productive relations with
all allied instituted means of subordination, subjugation, alienation and de facto
control over labour that the jāti ensured as its fundamental service. Hence, the
context of jāti formation is that of the disintegration of descent groups, which
happens in the process of their integration for wetland agriculture of wheat, rice
and sugarcane requiring the technology of plough with its auxiliary arts and crafts
of full-time specialization. The spread of agriculture along the fertile landscape
ecosystems of the river valleys, first in the Gangetic region and subsequently
along the deltas of various other rivers all over the northern, central and eastern
regions was the historical context of the productive integration of agro-pastoral
descent groups in the subcontinent. That the relations of production and reproduc-
tion would define social groups in the historical social process constitutes the
explanatory framework enabling us to contextualize the jāti phenomenon.4
3
For a discussion of the origins of jāti in the light of literary references, see S. Jaiswal, Caste: Origin,
Function, and Dimensions of Change (New Delhi: Manohar Publishers & Distributors, 1998).
4
In fact, historical materialism constitutes the central framework of explanation for the present study.
Marx discussed caste in the context of the institutional as well as ideological aspects of the form
of exploitation of labour in the relations of production. See, German Ideology, original publication
1845–46, Moscow edition, Part I, 63. Also see Capital, vol. I, Moscow Publishers, 1974, 321. For a
creative response to the concept in the context of Indian history, see D.D. Kosambi, Introduction to
the Study of Indian History (Bombay: Popular Prakashan, 1975), 99–101. He related the formation of
Gurukkal 33
At no point in history did the jāti-based social organization encompass all,

leaving no residues. It was an organization with its own spatiality, distinguished
as the veḷḷānvakai (non-Brāhmaṇa) the brahmadēya, the nādu (non-Brāhmaṇa
agrarian locality) and the nagara (the marketing centre) as opposed to the kāṭu
(forest).5 Those accommodated as the menial as well as hard manual labourers
who constituted the main workforce in agriculture were in clusters of jāti exclu-
siveness with their space in the cēris. Purity >< pollution observances in their
extreme form manifested in the institution of untouchability make the spatial
dimension of jāti explicit. Nonetheless, it operated only within the outskirts, but
integral to the space, of productive relations. People, normally settled in the forest
or along the seashore, who had nothing to do with the space of relations of pro-
duction, were outside the jāti system.
Antecedents
Cultivation in the alluvial deltas of the Ganga, which goes back to the mid-first
millennium BCE, was a slow but significant process involving integration of
functionaries and occupation groups who were, by and large, part of agro-pastoral
descent groups.6 This is not to mean that there was already an evolved division of
labour in the agro-pastoral settlements, presupposing the existence of hereditary
specialization of arts and crafts in any institutionalized form. It would not have
been more than a kin-based system of productive relations with a relative special-
ization of arts and crafts with or without the hereditary system but within the
clan ties, precluding hierarchy. Integration of arts and crafts people as part of the
the Sudra farmers to the rise of new forces (iron plough) and relations of production in the Gangetic
plains, which led to the emergence of endogamous caste groups. This has been dealt at length in
K. Roy, ‘Kosambi on Questions of Caste’, Economic and Political Weekly XLIII, no. 30 (July 2008):
78–84. For a theoretical reflection relevant to the context, see I. Habib, ‘Castes in Indian History’,
in Essays in Indian History: Towards a Marxist Interpretation, ed. I. Habib (New Delhi: Tulika
Publication, 1995), 161–79. Also his ‘Note towards a Marxist Perception of Indian History’, The
Marxist XXVI, no. 4 (October–December 2010): 46.
5
See arguments relating to homology between socio-spatial relations in the introduction of D. Harvey,
Explanation in Geography (New York: St. Martin’s, 1969). More insights are available in D. Massey,
Spatial Divisions of Labour: Social Structures and the Geography of Production (London: Macmillan,
1984); D. Gregory and J. Urry, Social Relations and Spatial Structures (London: Macmillan, 1985);
and H. Lefebvre, The Social Production of Space, trans. Donald Nicholson-Smith (Oxford: Wiley
Blackwell, 1992).
6
An exhaustive review of the archaeology and socio-economic processes of the region is given in R.S.
Sharma, Material Culture and Social Formations in Ancient India (New Delhi: Motilal Banarsidas,
2004), 195–99. He argued that around first millennium BCE the Gangetic plain began to be deforested,
thanks to the knowledge of iron technology, and expanded agriculture leading to the formation of class
and state with a steady growth of specialization and subjection of the Sudras. See pp. 236–40. With
little differentiation between the primordial and advanced, there is an attempt to link the caste system
with agriculture in M. Klass, Caste: The Emergence of the South Asian Social System (Philadelphia:
Institute for Human Issues, 1980), 62–65. Klass presumes the transformation of the tribe into caste, but
without any clarity about the historical process thereof.
making of productive relations meant their incorporation into the Sudra varṇa.
This process gave rise to three social groups in competition for dominance,
namely, the ruling aristocracy (Kshtriyas), the priestly group (Brāhmaṇas) and the
traders (Vaisyas), each with its own service personnel of the dāsa-bhrutaka cate-
gory under servitude close to that of slavery.7 Competitions and conflicts, mainly
between the Kshatriyas and Brāhmaṇas, had acquired greater philosophical and
institutional dimensions as exemplified by the constitution of heterodox religious
orders. This long history has been discussed extensively by eminent historians as
part of the early history of socio-economic development with special reference to
the formation of the jāti and State in the region.8 Large-scale expansion of agri-
culture in the deltas using the Sudra labour began only in the third century BCE,
the constitution of which was the result of the dissolution of many agro-pastoral
descent groups into artisans, craftsmen and tillers with occupational identity of
hereditary nature.
In the process of the long-protracted competition and conflicts, the Brāhmaṇas
acquired the highest ritual status that acted as the cultural capital for ideologi-
cal control over political power through the prescriptive knowledge generated
and codified by them into sāstras.9 Kshatriyas accepted the dominance of the
Brāhmaṇas over others and sought to follow sastraic knowledge, which was, in
fact, a power–knowledge combine capable of turning the Brāhmaṇas themselves
into its discursive subjects first and subsequently powerful enough for ordering
of the society as a whole.10 Formation of the jāti system as a hierarchy with the
four varṇas as its rudimentary constituents accommodating artisans and crafts-
men within the fourth varṇa, namely, the Sudra, and keeping the dāsa-bhrutakas
outside it was a discursive consequence of Brahaman domination involving
production of explanatory as well as prescriptive knowledge about the origins and
proliferation of the jāti.
The old time patron–client ties between the Kshatriyas and Brāhmaṇas contin-
ued as transformed into a symbiotic relationship in which the former resorting to
the latter for ritual status and political legitimacy, thanks to their authority over
rites, monopolistic custodianship over textual accounts and the rich heritage of
7
See the process conceptualized in C. Meillassoux, ‘Are There Castes in India?’, Economy and
Society 2, no. 1 (1973): 89–111.
8
To cite the two major studies: R. Thapar, From Lineage to State: Social Formations of the Mid-First
Millenium BC in the Ganga Valley (New Delhi: Oxford University Press, 1991); and R.S. Sharma,
Material Culture and Social Formations in Ancient India (New Delhi: Motilal Banarsidas, 2004).
9
Reference here is to the concept of cultural capital enunciated in P. Bourdieu, ‘The Forms of capital’,
in J. Richardson (ed.), Handbook of Theory and Research for the Sociology of Education (New York:
Greenwood, 1986), 241–58. In fact, it is the combination of the embedded, objectified and institution-
alized forms of cultural capital that the Brāhmaṇas inherited, a rare historical legacy that enabled them
command social capital. Also see his Distinction: A Social Critique of the Judgment of Taste, trans.
Richard Nice (Cambridge: Harvard University Press, 1984).
10
It is the concept of discourse, enunciated by M. Foucault who defined the term as the power–
knowledge combine establishing orders of truth in multiple ways to be accepted as ‘reality’ in a given
society. For details see his The Archaeology of Knowledge and the Discourse on Language, trans.
A.M.S. Smith (New York: Pantheon Books, 1972), 21–39.
Gurukkal 35
traditional wisdom. Having no such ritual strategies, institutional devices and tex-
tual knowledge as source of cultural capital to share political power, Vaisyas or
the traders got subordinated to the other two and remained the third in the race for
domination. Cultural resources, particularly the explanatory and prescriptive wis-
dom being overwhelming in the case of Brāhmaṇas, organization and imposition
of a hierarchy of status with themselves on the top was easy for them. Heterodox
religions had an ideology, but that was against the notion of hierarchy and per-
tained to a parallel politico-cultural domain. With the result Brahamanical values
and passions became hegemonic soon in matters pertaining to every aspect of
human existence. This accounts for the feasibility of subsequent social ordering in
areas of structured productive relations, exactly as construed by the Brāhmaṇas.
The notion of hierarchy in the relations of production though implied in the nature
of entitlements to means of production and the produce, the status implications
thereof were decided according to the prescriptions in the Brahmanical texts. In
other words, Brāhmaṇas decided the status of functions/occupations and ordered
them hierarchically with the repressive institutional means called jāti.11 As wet-
rice agriculture expanded to areas beyond the deltas through the spread of irriga-
tion technology, the same system of productive relations and status hierarchy got
replicated there as well. Ordering of society into a hierarchy according to the jāti
status as superimposed on productive relations was a gradual process over several
centuries across the agrarian tracts.
The Deccan and Beyond

Regions of the Deccan and further south were not varṇa-structured till the
opening up of the deltas, which seems to have begun not earlier than the mid-first
millennium AD, unlike often made out. Archaeological remains on the surface
and out of excavations show a continuous history of human settlements in differ-
ent parts of the region from the turn of first millennium BC, onwards, as adapted
to the multiple landscape ecosystems through an array of unevenly evolving
techno-economic strategies of subsistence and survival.12 It was an ensemble
structured by the dominance of the agro-pastoral culture that had a long continu-
ity, of course with certain changes. Agriculture included both shifting as well as
sedentary, with the latter mostly centred on arid highland crops such as millets and
11
It was the fallout of a historical process rather than a pre-planned scheme. Perhaps the earliest
characterization of caste as a practice ‘born and grew from the concurrence of spontaneous and
collective tendencies’, rather than out of Brāhmaṇa conspiracy, is made in C. Bougle, Essays on the
Caste System, first published in 1908, republished by Cambridge University Press, 1971, 59–61.
12
See, H.D. Sankalia and S.B. Deo, Report on the Excavations at Nasik and Jorwe 1950–1951 (Poona:
Deccan College, 1955). H.D. Sankalia, B. Subbarao and S.B. Deo, Excavations at Maheshwar and
Navdatoli, 1952–53 (Pune: Deccan College, 1958). Dhavalikar, First Farmers of the Deccan (Pune:
Ravish Publishers, 1988). Also V.V.K. Sastry, The Proto and Early Historical Culture of Andhra
Pradesh, AP Archaeological Series 58, Government of Andhra Pradesh, Hyderabad, 1983. S.R.K.
Pisipaty, Andhra Culture: An Obscure Phase in the Early Historical Archaeology of Andhra Pradesh
(Delhi: Agama Kala Prakashan, 2010).
limitedly on fertile lowland crops, such as, rice, wheat and sugarcane. The socio-
economic processes of the region during the immediate post-Mauryan period
(first century BCE) were primarily of the continuation of interactive coexistence
amongst these unevenly evolved forms of subsistence based on the technology of
iron and high-tin bronze. There was specialization in craft production and
exchange, but largely within the clan–kin nexus.13 Most population in the region
must have belonged to the settlements along the black-soil tracts of the Ghats and
upper reaches of the rivers, suitable for agro-pastoral means of subsistence and it
consisted of descent groups and their chiefs of clan–kin ties. Some of them were
inhabitants of small fertile pockets of fields around watersheds, living on wet-rice
agriculture, but the basis of productive relations remained kinship as in the case of
other descent groups. Networking across these settlements of the hill tracts rich in
forest goods, mineral resources, crafts production, dry-land crops such as millet,
pastoral goods and rice, there were trade routes from the north, north-west and
east frequented by long-distance itinerant merchants.
Trade and trade routes had enabled circuit of merchants, monks, mendicants
and others, while their convergence at areas of settlements in turn led to the rise
of monasteries and growth of urban centres. There has always been exaggera-
tion about the agrarian expansion, social stratification and productive surplus in
the settlements and over-generalization about its connection to trade, monastic
establishments, urban development and state formation with differences in the
precedence of one or the other over the rest.14 In fact, these Buddhist monu-
ments were in the upper reaches of the Godavari–Krishna rivers, the landscape
of agro-pastoral settlements.15 Indeed, numismatic material, both local and
non-local in plenty indicates the importance of the area as a township of mer-
chants and monetized exchanges.16 The population in these strategic areas of trade
13
For the standard characterization, see A. Parasher-Sen, Social and Economic History of Early
Deccan: Some Interpretations (New Delhi: Manohar, 1993).
14
For an exaggeration of the material culture, nature and surplus potential of local settlements as well
as for a criticism against generalizations about the connections amongst the various phenomena, see
the discussion in K.D. Morrison, ‘Trade, Urbanism, and Agricultural Expansion: Buddhist Monastic
Institutions and the State in Early Historic Western Deccan’, World Archaeology 27, no. 2 (1995):
203–21. Also P.K. Reddy, ‘God, Trade and Worship: A Glimpse into the Region of Early Andhradesa’,
East and West 48 (1998): 291–311.
15
See J. Burgess, Report on the Buddhist Cave Temples and Their Inscriptions, Archaeological Survey
of India, New Series, 4, New Delhi, 1883. For a comprehensive account of sites, see O.C. Kail,
Buddhist Cave Temples of India (Bombay: D.B. Taraporevala, 1975). Also, I.K. Sarma, Studies in
Early Buddhist Monuments and Brahmi Inscriptions of Andhradesa (Nagpur: Dattsons Publishers,
1988). D. Mitra, Buddhist Monuments (Calcutta: Sahitya Samsad, reprint 1980), 155–57. For Andhra
an exhaustive chronological list of Buddhist and other religious sites, see R. Prasad, ‘Cultural Map of
Andhradesa from Earliest Times to AD 300’, in Comprehensive History and Culture of Andhra
Pradesh, Vol. II: Early Historic Andhra Pradesh, ed. I.K. Sarma (New Delhi: Tulika Books, 2008),
287–308.
16
See P.L. Gupta, ‘The Coinage of the Sātavāhanas: Types and their Regional Distribution’, in
Coinage of the Sātavāhanas and Coins from excavations, ed. A.M. Shastri (Nagpur: Nagpur University
Press, 1972), 41–62; V.V. Mirashi, ‘Wategaon Hoard of Sātavāhana Coins’, JNSI, 1972, pt-ii, 205–12;
M.R. Rao, Sātavāhana Coins in the Andhra Pradesh Govt. Museum, A.P. Govt. Series no. 2,
Gurukkal 37
and markets was largely made up of a few prominent households of big merchants
and chiefly personages such as the Raṭhikas, Bhōjas and Peṭeṇikas, their warriors
and the dāsa-bhrutakas besides some Brahāmaṇas as Vēdic priests and precep-
tors. Specialized merchants, such as, dhānikas (corn dealers), gandhikas (perfume
dealers), mālakaras (florists), suvarṇakaras (gold dealers), odayantrikas (irriga-
tion device dealers) and so on, organized into nigamam too were there as the
cave inscriptions would have us believe. Chiefly households seem to have had
no systematic relations of appropriation with, rather than predatory control over,
the settlements of descent groups. Other prominent households do not seem to
have any intermediaries in land placed below them, presupposing that their land
control was hardly beyond what could be cultivated by their dāsa-bhrutaka work-
force. Being itinerants of brief sojourn at points of exchange, the traders were
not integral to the local society. In short, the system of social relations, though
differentiated was yet to be clearly varṇa-structured. This is a situation evidently
that of the pre-jāti, pre-state, best represented universally by the tribal chiefdoms
of unilineal descent with patterns of power relations varying across tribes, clans
and lineages.17
Sātavāhanas appear to have been a chiefly lineage with Brāhmaṇic preten-
tions in the central Deccan, wielding control over the southern trade route as
its lord (Dakshiṇāpatha-pati) but without any consolidated political authority,
probably till the ascendancy of Gautamiputra Sātakarṇi who could transcend the
gōtra nexus and assert himself as a king.18 There are inscriptional references to
Sātakarṇi and his successors paying attention to the maintenance of the varṇa
system, probably a need in the wake of varṇasamkara and the emergence of
sankirṇajātī. It appears to be a phase witnessing dissolution of the clan/kin base
of productive relations into varṇa base. Nonetheless, there is no indication of
Hyderabad, 1961; D.R. Reddy and P. Suryanarayana, Coins of the Satraps of the Sātavāhana Era
(Hyderabad: Numismatic Society of Hyderabad, 1983), 70–77; also their Coins of the Mēghavāhana
Dynasty of Coastal Andhra (Hyderabad, 1985); S.J. Mangalam, ‘Coins of the Feudatories and
Contemporaries of the Sātavāhanas’, in The Age of the Sātavāhanas, 2 vols., ed. A.M. Shastri
(New Delhi: Aryan Books International, 1999), 360–90. A general appreciation of the situation is
given in H.P. Ray, Monastery and Guild: Commerce Under the Satavahanas (New Delhi: Oxford
University Press, 1987). For a discussion of the nature of urban settlements, see A.P. Sen, ‘Urban
Settlements in the Deccan and Sātavāhana History’, in The Age of the Sātavāhanas, vol. I, ed. A.M.
Shastri (New Delhi: Aryan Books International, 1999), 159–89.
17
See discussion in M.H. Fried, The Evolution of Political Society (New York: Random House, 1967),
236–41. Also, E.R. Service, Origins of the State and Civilization, 14–16. For details of the tribal
polity, see M.D. Sahlins, Tribesmen (New Jersey: Prentice Hall, 1968), 22–25. See variations in
tribal polity discussed in the introduction to J. Middleton and D. Tait (eds.), Tribes without Rulers
(London: Routledge & Kegan Paul Ltd., rpt. 1970), 1–32.
18
See the brief but clinching discussion in B.D. Chattopadhyaya, ‘Transition to the Early Historical
Phase in the Deccan: A Note’, in Archaeology and History: Essays in Memory of Shri A. Ghosh, ed.
B.M. Pande and B.D. Chattopadhyaya (Delhi: Agam Kala Prakashan, 1987), 727–35. This has been
reproduced in his Studying Early India: Archaeology, Texts, and Historical Issues (New Delhi:
Permanent Black, 2003), 39–47. For a view almost the same but arrived at differently to this, see
Y. Subbarayalu, ‘Contacts between the North and the South: An Epigraphical Perspective’, Foundation
Day Lecture, ICHR, New Delhi, 2012.
the expansion of agriculture into the alluvial deltas of the major rivers in the
region. The situation in the Kalinga region was more or less the same, with agro-
pastoral settlements of the arid highlands having precedence over those in the
plains engaged in wet-rice agriculture. It was predominantly tribal despite its long
tradition of cultural sharing with the Gangetic region and subsequent large-scale
marches of people from there with the Mauryan techno-economic culture of agri-
culture and trade after Asoka’s conquest of the region. Socio-economic processes
of the region were the same as those of the upper reaches of the Godavari with
little scope for large-scale transformation of descent groups into class-structured
society and state formation that corresponded to organized agrarian expansion
after three centuries. Integration of the agro-pastoral artisans and craftsmen
became necessary in the Andhra–Kalinga regions only around fifth to sixth cen-
turies CE, when the deltas were being converted into fields of paddy, wheat and
sugarcane, providing the context for the sāstraic ordering of them into the jāti.
The Tamil Region

In the Tamil south it is evident from both archaeology and the heroic poems that
the people were of descent communities, predominantly agro-pastoral, headed by
small and big chieftains, such as, the Kiḻār (Ur chieftain), Vēḷir (hill chieftain) and
Vēntar (regional chieftains, namely, the Pāṇdya, Cēra and Cōḻa) from the turn of
the CE to the third or fourth century.19 The principal social mode of labour realiza-
tion was familial and cooperative with little scope for specialization. However, a
few crafts such as metal-working and pottery were full-time trades of specialists
and hence probably hereditary.20 As the most extensively used metal, iron had a
central place among metals as the base of weapons whose significance in a
society of predatory operations is explicit. Moreover, the practice of burying iron
objects along with the dead had pushed a great deal of iron out of circulation,
presupposing continuous iron working as a full-time occupation of hereditary
specialization. Similarly, production of earthen pots, a characteristically brittle
artefact, was obviously a continuous full-time activity too, for their use was exten-
sive both for the living as well as the dead. Moreover, the fabric, polish, glazing,
slips, paintings, texture and decorative designs of pottery suggest that it was a
full-time technology of specialized expertise. Nonetheless, these crafts people by
and large belonged to the clan–kin ties.
The number of such full-time artisans and craftsmen of hereditary occupa-
tions must have been more in the headquarters of bigger chieftains of the Vēḷir
19
For details, see R. Gurukkal, ‘Forms of Production and Forces of Change in Ancient Tamil Society’,
Studies in History, vol. 2 ns, New Delhi (1989), 159–75.
20
This has been discussed at length in R. Gurukkal, ‘From Clan and Lineage to Hereditary Occupations
and Jāti in South India’, Indian Historical Review, XXII, nos. 1–2 (1993–94): 22–33. It has been
reprinted in his Social Formations in Early South India (New Delhi: Oxford University Press, 2010),
255–71.
Gurukkal 39
and Vēntar levels.21 As the major redistributive pools of resources, the chieftains’
settlements could support more full-time crafts. Another full-time function of
a hereditary nature was that of the warriors, who had existed as a clan. Every
settlement (Ūr) needed full-time warriors since the main mode of political appro-
priation of resources was predatory. In association with the chiefly households,
there were three other full-time hereditary functionaries: the pāṇar (bards),
paṛaiyar (who play a kind of raid drum called paṛa) and tuṭiyar (who play a small
drum called tuṭi). None of these was represented in terms of varṇa in the source
material of their times.
Tolkāpiam mentions Antaṇar (Brāhamaṇas), Aracar (rulers) and Vaṇikar
(traders), but not on a par with the trivarṇikar unlike as often made out by
historians.22 Specialized dealers in arts, crafts and other products, probably as
organized into corporation (nigamam) were present in marketing centres, coastal
towns and ports, but being mostly part of the long distance itinerant merchant
community, they were not integral to the local society. However, it is likely that
the overseas and inland merchants had required servile people (vilainjar) at
the place of sojourn for various menial jobs. Such people at service under con-
ditions of coercion were workers representing a system of relations of labour
transcending kinship, but more or less as the Tamil counterpart of the dāsa-
bhrutaka workforce in the Deccan and northern India. In the process of
predatory operations and redistribution, some kind of differential allocation
of new position, status, roles and prestige within the complex redistributive
relationships was likely in the agro-pastoral settlements. Differential allocation
of positions and roles at the instance of the Vēntar level chiefly authority had
a tendency towards formation of a hierarchy. In fact, functions or occupations,
although not in any elaborate form, were already there as organized around
chieftains in the case of certain functions/occupations and in the case of certain
others, especially traders, artisans and craftsmen, they themselves were organized
into the nikamam. However, the poems do not contain any clues to the existence
of a stratified society.23 They show social differentiation of a simple kind con-
fined to the binary between uyarntōr (the highborn) that comprised Brāhmaṇas
and iḻipiṛappāḷar (the lowborn) that comprised all people. This suggests that
contemporary social division was too fluid even to be varṇa-structured.
Opening Up of the Deltas

A perceptible feature of the opening of the deltas, which began during the sixth to
seventh centuries CE, was the founding of brahmadēya villages through royal
land grants along the fertile tracts of river valleys. We cannot say that these gift
21
Ibid.
22
Tolkāpiam, porul (Chennai: Saratha Pathippagam, 2010), 625–26. For details, see N. Subrahmanian,
Snagam Polity (Bombay: Asia Publishing House, rpt 1966), 258–60.
23
For a detailed discussion, see R. Gurukkal, Social Formations of Early South India (New Delhi:
Oxford University Press, 2010), 224–41.
lands were entirely uncultivated, but it is reasonable to presume that there were a
lot of fallow lands waiting to be brought under the plough. This is an indication of
the absence of productive relations appropriate for maximizing agrarian land use.
There must have existed a specialized workforce in the form of descent groups but
not integrated into a system of relations involving protection from above and
obligation from below. Expansion of agrarian settlements through the creation
of brahmadēyas often involved superimposition of the superior rights of the
Brāhmaṇas over communal holdings and clan families of the locality. It must have
been indeed a coercive process of transformation of clan settlements into farmer
settlements of wet-rice agriculture necessitating stratified relations of production,
which is implicit in the foundation of agrarian villages under royal initiative.
The rise of a new political formation represented by the Simhavarman line of the
Pallavas and the Kdungōn line of the Pāṇdyas, with genealogies celebrated in
the prasastis of copper-plate charters registering the foundation of brahmadēyas
coincided with the process of agrarian expansion along the deltas. Local chief-
tains and prominent households also must have functioned as the source of
coercive power behind the process of integration of agro-pastoral artisans, crafts
people and farmers.24
It appears that Veḷḷāḷas emerged from the farmers of agro-pastoral settlements
and were descent groups with lands communally owned and controlled. Most
brahmadēyas were founded in the neighbourhood of the Veḷḷāḷa settlements,
after the latter were taken into confidence. Agricultural expansion must have
naturally led to the formation of huge agrarian localities called nādus, which
had involved integration of settlements (Ūr) originally bound by kinship and
cultivated by Veḷḷāḷas. It accomplished a uniform structure of productive
relations in agrarian villages irrespective of whether they were brahmadēyas
or vēḷānvakais. It was a hierarchical structure with landholders (brahmadēya-
kiḻavar in the case of brahmadēyas and ūrār/nāṭṭār in the case of vēḷānvakai
settlements) at the apex and leaseholders (kārāḷar) consisting mainly of artisans
and craftsmen in the middle as placed over the primary producers (aṭiyāḷar) at
the bottom. Almost parallel to the leaseholders there were many who held small
strips of land as hereditary holdings (kāṇi). Agricultural produce in given shares
moved up following a system of appropriation under extra-economic coercion
along the structured path from the tillers through categories of different levels of
entitlement in the ascending order. As part of the social mechanisms of ensuring
goods and services to the landholders through the notion of obligation, all artisans
and craftsmen were subjected to immobility. In this hierarchy of productive rela-
tions involving de facto control over human body from above, the institutional
role of the jāti was crucial. In the process of the proliferation of brahmadēyas and
the spread of stratified productive relations, the Veḷḷāḷas seem to have emerged as
a landed jāti of significance due to their kinship with local chieftains on the one
24
It was not altogether without any protest. See certain instances discussed in R. Gurukkal,
‘Non-brahmana Resistance to the Expansion of Brahmadeyas: The Early Pāndya Experience’, Indian
History Congress Proceedings, Delhi, 1984.
Gurukkal 41
side, and the absence of a similar situation for the rise of traders into a powerful
group of hereditary occupation, on the other.
In Kerala the opening up of the deltas was somewhat different. Absence of
royal land grants for the founding of brahmadēyas is perhaps the most striking
among them. However, Brāhmaṇa households had sprung up on their own as
independent production units adjacent to agro-pastoral clan settlements in the
early centuries of CE.25 Formation of Brāhmaṇa settlements along the fertile river
valleys of the region was the result of organized migration of select families from
previous settlements rather than an enterprise under royal initiative as the legend
of Parasurama would have us believe.26 Another remarkable difference was the
region’s waterlogged and marshy landscape ecosystem in the delta, which neces-
sitated extensive mobilization of hard labour for the reclamation of agrarian fields.
Labour mobilization for reclaiming productive lands out of the relatively inhos-
pitable landscape ecosystem, must have been a long-term activity. Agro-pastoral
clan families living along the hilly fringes of the marshy wetlands and engaged
in the cultivation of millet and highland paddy must have been the main source
of workforce. Occupational groups of arts and crafts must have been attached to
the Brāhmaṇa land on a permanent basis under the institution of bonded labour as
enabled by the high ritual status, scholarship and charisma of the Brāhmaṇas. This
was the context of the beginnings of social stratification in the region.
Social Stratification
With the steady expansion of rice agriculture across the wetlands during the sixth
to seventh centuries CE, social relations began to be structured along the line of
productive relations. What began taking shape in the Brāhmaṇa households was
crucial for the beginnings of a hierarchy of status. It was natural for the service
personnel attached to the Brāhmaṇa household as a new unit of production with
stratified relations to be conceived in the form of a status hierarchy in accordance
with the sāstraic prescriptions of social differentiation. In fact, notion of hierarchy
was implicit in the system of productive relations in which the land owning
Brāhmaṇas and the landless tillers constituted the two objectively antagonistic
classes with an intermediary of relative differentiation in economic as well as
social status. Hence, the period, a temporal juncture that witnessed the expansion
of a technology of production and social strategies of labour realization leading
to the proliferation of hereditary occupations, was a turning point in terms of
stratification and hierarchical ordering, which took more than two centuries to
25
See discussions in M.G.S. Narayanan, Perumals of Kerala (Thrissur: Current Books, 2013), 271.
For a detailed study of the Brāhmaṇa settlements see V. Kesavan, Brahamin Settlements in Kerala
(Thrissur: Current Books, rev. edn., 2013), 43–59.
26
The legend forms an integral part of the traditional chronicle Keralōlpatti. For the full text see,
H. Gundert, Keralōlpathi, originally published in 1843, a later edition by Balan Publications,
Trivandrum, 1961.
characterize the social aggregate, for the domination of the social aggregate by
relations in plough agriculture, proliferation of hereditary occupational groups
and their ordering into a hierarchy, was a long process.
The integration of descent groups with the identity of hereditary occupations
into the system of stratified productive relations consequent on Brāhmaṇa-headed
agrarian expansion under royal initiative was the ongoing process inevitably lead-
ing to the constitution of a stratified society. Agrarian expansion in the deltas and
the emergence of stratified productive relations accomplished a hierarchy of land
rights and entitlements. It eventually changed the structure of the ruling author-
ity represented by the Kshatriyas in symbiotic alliance with the Brāhmaṇas, into
the state. Through this alliance of mutual benefits, the Kshatriya got higher status
and ranking at the instance of the sāstraic concept of kingship as ordained by
the Brāhmaṇas, who in turn got land and gold as reward for it. Both shared repres-
sive power, the former through the politico-military source and the latter through
the religio-intellectual source. It helped the Brāhmaṇas to become the cultural
and eternal power, the power of conventions, ethics and morality, sufficient to
take precedence over physical power. The Brāhmaṇa thus became the ideal and
ideally the highest and could decide hierarchy with the self-acquired top. State,
the main repressive institution, delegated its coercive power to the landed along
with proprietary control over villages, which led to their integration as ensembles
of settlers with hereditary occupations with attached entitlements.
Land grants to the Brāhmaṇas were responsible for bringing the deltas of the
major rivers extensively under plough by integrating agro-pastoral settlements,
enhancing surplus and developing a differentiated economy as the foundation
of the state in peninsular India. It was at the instance of the Brahamanas that
the creation and imposition of the hierarchy of varṇa happened there too. In that
sense the Brāhmaṇas were instrumental in organizing the constituents of stratified
society into a hierarchy of status and ranking. Their Vēdic, itihāsic, Purāṇic and
sāstraic ideas and institutions were effective devices of social control and domi-
nation. It is reasonable to presume that the instituted relations of labour realization
by the Brāhmaṇa laid the foundation of a stratified society based on an objective
antagonism between landlords and tillers. In fact, this process involved a series of
transitions: transition from kin labour to non-kin labour in productive relations,
from the primacy of arid crops to the dominance of wetland crops in agriculture,
from itinerant to immobile and obligatory services, from uncertainty to certainty
in occupational rewards, from the fluidity of general functions of choice to the
rigidity of specialized functions of hereditary occupation, from clans to jātis, from
the horizontally structured descent group settlements to the vertically structured
agrarian villages, and from chiefdom to the state.27
27
For details, see R. Gurukkal, Social Formations of Early South India (New Delhi: Oxford University
Press, 2010), 242–54.
Gurukkal 43
The Temple Society

Proliferation of brahmadēyas corresponded to the emergence of temples in
South India, for each of them got organized as temple-centred with the temple
acting as its agrarian headquarters. In due course, the temple itself became a
landed magnate through royal land grants, private land endowments and gifts of
gold that enabled purchase of productive lands. Other forms of resources donated
by devotees also made the temple a fabulously rich institution.28 With a stratified
social order made up of tenants in the temple land, functionaries in its service and
artisans and craftsmen of obligatory responsibilities of their trades, the temple
commanded a large part of the local people as its dependents. All of them were
rewarded in the form of one kind of land right or the other or an entitlement
to land use or a share of the crop. It brought forth a system of service tenure as
attached to most of the services and functions to the temple. A hierarchy was
already implicit in the nature and size of the rewards of the temple servants, but
the order of status implications thereof was decided by the Brāhmaṇas, who were
the custodians of the temple wealth.
The structure and composition of the temple society is best illustrated in the
temple epigraphs of Kerala. What emerges from them is a set of data enabling
us to visualize it as a hierarchical structure of entitlements to varied rights over
land institutionalized by service tenements. At the apex was the king’s seigniorial
jurisdiction (koima), the highest right tacitly recognized by all. Recognized and
exempted from taxes by the king, the Brāhmaṇa land (brahmasvam) and temple
land (dēvasvam) enjoyed autonomous rights of the landlord as granted by the
royal authority. Local rulers (nāṭuvāḻis) and probably other non-Brāhmaṇa land-
lords had proprietary control over their hereditary lands. Below the ownership of
landlords was the leaseholder’s right (kārāṇma). Artisans and craftspeople were
entitled to occupy the land (kuṭiyāyma) by way of reward for their service to the
settlement. At the base were the tillers attached to land, a people of bonded ser-
vitude and immobility. These varied entitlements in the hierarchy superimposed
in the ascending order were tightly bound by ties of dependence and protection
besides relations of exploitation and subjection. People in every stratum were
locked up in a system of hereditary service obligation and land-related reward.
Rewards of the following personnel in the temple service such as the Potuvāl
(the temple secretary), the Vārier (the personnel of supplementary services in
the sanctum), the Bhaṭṭas (Vēdic, itihāsic, purāṇic teachers), the Caṭṭas (the
Brāhmaṇa pupils of Vedic itihasic, puranic studies), the Cākyār (the performers
of kūttu), the Naññyār (female of the Cākyār family), the Uvaccakal (musicians),
28
For details, see Burton Stein, ‘The Economic Function of a Medieval South Indian Temple’, Journal
of Asian Studies 19, no. 2 (1960): 163–76. Also The South Indian Temples: An Analytical Consideration
(New Delhi: Vikas Publications, 1977). The economic position of the temple is well brought out in
N. Karashima, 1896, ‘Land Donations to Hindu Temples in Medieval South India’, Journal of
Asia and Africa Studies, 2, Tokyo (1987): 3–31. Also, D.N. Jha, ‘Temples as Landed Magnates
in South India, c. ad 700–1300’, in Indian Society: Historical Probings, ed. R.S. Sharma and V. Jha
(New Delhi: People’s Publishing House, 1974), 202–16.
the Naṭṭuvar (dancers), the Kāndarppikar (dance teachers), etc., were in the form
of land tenures (tenancy).29 For each of the above services, the specific tenure
called virutti or jīvitam (life expenditure) was instituted under contemporary
land system, which kept at the disposal of the service personnel certain plots
of the temple land on lease. The Potuvāḷ and Vāriyar are the two most impor-
tant functional categories in the non-Brāhmaṇa order, of which the first literally
meant the common man (madhyastha) between the custodians of the temple and
the devotees for all their transactions with the temple, and hence acted as the
secretary of the temple. As a functionary shouldering managerial and executive
responsibility, the Potuvāḷ was relatively close to the Brāhmaṇas and received
virutti land as reward for his service to the temple. Anyone who was a member
in any of the temple committees (vāriyams), could be called a Vāriyar. However,
there is a specific functional group referred to in the temple inscriptions as Vāriyar
by profession and not by virtue of membership in any vāriyam. Members of this
group were also given virutti land as a reward for their services in the temple.
Both the Potuvāḷ and the Vāriyar, two functionally specific groups attached to
the temple, are now the names of two endogamous jātis in Kerala.30 Evidently
the process was that of the turning of the service into hereditary occupation
for retaining its reward in the form of land tenure. Then, as groups of hereditary
professions with economic stability, ritual status as the temple secretary and
ranking due to close interaction with the Brāhmaṇas, they could distance
themselves from others through rules of inter-marriage and inter-dining in order
to constitute themselves into two endogamous jātis.
Similarly, the drummers, dancers and musicians of the temple, paid through
service tenure constituted themselves into separate jātis through the same process
of becoming hereditary professionals first. Drummers of the temple are addressed
in the inscriptions as Koṭṭikaḷ or Uvaccakaḷ, which as such have not survived
as jāti names. Temple dancers, mentioned in the epigraphs as Cākkaimār (male
dancers) and Naññaimār (dancing girls) became a jāti, namely Cākkiyār with the
female members addressed by the old name, Naññyar.31 These jātis are generally
called the temple jātis (ampalavāsis) or the antarāḷa-jātis, the jātis between the
Brāhmaṇas and the non-Brāhmaṇas just below. It was convenient for the temple to
make the services hereditary, for it ensured stability of service. Likewise, it was an
added incentive for the family of the service personnel, for it brought stability of
landed property through service tenure. Thus, service tenure was responsible for
turning the above service personnel families into hereditary occupational groups
and endogamous jātis. A higher service tenement meant a better economic sta-
29
See discussions in M.G.S. Narayanan, Perumals, 272–73. Also R. Gurukkal, ‘The Socio-economic
Milieu of the Kerala Temple: A Functional Analysis, c. AD 800–1200’, Studies in History II, no. 1.,
New Delhi (1980): 159–175.
30
See discussion in R. Gurukkal, ‘Formation of Caste Society in Kerala: Historical Antecedents,’
in Social Formations of Early South India (Oxford University Press, 2010) 312–13.
31
Ibid. Also see ‘Proliferation and Consolidation of the Temple Centred Social Hierarchy in the Cēra
Period’, Journal of Kerala Studies VI, nos. III and IV, University of Kerala, Thiruvananthapuram
(1979): 333–46.
Gurukkal 45
tus and proximity to the managers of the temple-centred village, the Brāhmaṇas,
implying a higher ritual status. Such a system of status differentiation in terms
of economic as well as ritual values was instrumental to the naming of the
hereditary occupational groups with the occupational name as a means of distinc-
tion. Subsequently, the occupational name became the name of an endogamous
jāti integrated to social hierarchy.
The same principle of service stability, material incentives and status holds
good in the case of others such as the merchants (Vāṇiyar), craftsmen/artisan
groups (Kammāḷar) too whose trades became hereditary and the name of the
trade, the jāti-name. Inscriptions refer to groups such as Taccar (carpenters),
Kollar (blacksmiths), Kalavāṇiyar (potters), Vāṇiyar (oil mongers) and Vaṇṇār
(washermen) as the main occupational groups of hereditary identity. In fact, they
were brought and settled along the fringes of the villages with the obligation
to render services to the temple as well as the landlords. Their reward was in
the form of a land tenure (kuṭiyāimai), providing the entitlement to settle down
in a plot of land for their unfailing services and at the pleasure of the landed.
These people were caught up in an inescapable trap of immobility and functional
obligations. At the slum of the bottom, were the servile group (aṭiyāḷar) of vary-
ing names, such as, Īḻavas, Pulayas and Ceṛumas. Removed from the mainstream
as untouchables, they formed the actual tillers, the most exploited group fated
to be in perpetual servitude (aṭiyāima). Attached to agricultural lands, they were
transacted along with land as goods or livestock.32
With the primary producers at the base, the temple signified a gamut of social
relations into which the principles of jāti were introduced for the first time in a
full-fledged form. Formation of a hierarchy was a natural consequence of the
system of social differentiation, based as it was on varying levels of ritual status
positions in the orbits around the Brāhmaṇas. Those enjoying these varied entitle-
ments in the hierarchical order seem to have begun the practice of undertaking
their vocations on a hereditary basis primarily for retaining the land rights as their
family property. As people of hereditary occupations, they began to be addressed
with the name of their occupations. These occupational names subsequently
became jāti names, a process indicating their transformation into endogamous
jātis. Relations with the upper-class categories determined the material status of
the service personnel concerned; these relations depended on the nature and form
of reward that varied from service tenure to kind. Similarly, the nature of reward
might also have mattered in the determination of status. Same occupation or func-
tion thus got differentiated rewards and in such cases of status differentiation
the name was changed that subsequently marked them a different jāti. Kindred
descendants of the upper-class categories formed themselves into closed groups
of jāti and status exclusiveness, zealously guarded through judiciously arranged
marriage alliances and rigorously observed relationships of inter-dining.
A higher service tenement meant a better economic status and proximity to
the Brāhmaṇas, implying a higher ritual status. Such a system of status differ-
32
Ibid.
entiation in terms of economic as well as ritual values was instrumental in the

transformation of hereditary functionaries into an endogamous jāti hierarchy.
The system of service tenure under the king and the local rulers also gave rise
to hereditary offices, generating jātis and sub-jātis with economy and royalty as
determinants of status hierarchy. Its extension into non-Brāhmaṇa villages, and
even to market towns, is attested by records. In short, jāti appeared as an insti-
tutional manifestation in the hierarchically structured agrarian society in which
services were paid for in the form of land rights. Other institutional forma-
tions such as service tenements both in the domains of the temple and the king,
crystallized hereditary occupations into jāti hierarchy.33
Jāti Hierarchy
Productive relations in the deltas had preconditions such as hereditary occupa-
tions, asymmetrical social relations, differential allocation of status and dominant
presence of Brāhmaṇas for the emergence of jāti hierarchy. The dominant position
of the Brāhmaṇas proved to be crucial in the process. Brāhmaṇas’ domination was
based on materially, socio-culturally and historically contingent authority. Tacitly
recognized ritual supremacy, resource potential, social control, political influence
and cultural pre-eminence of the Brāhmaṇas accounted for their dominance.
Status as custodians of higher wisdom about the universe and calendar of seasons,
enabling prediction of natural changes, had added to their charisma. They embod-
ied the collective norms, controlled all cultural channels of communication and
commanded ideological structures of legitimization. These made them a determi-
nant force of political authority enabling to take precedence over the ruling power.
All this explains how the Brāhmaṇas succeeded in being hegemonic to prescribe
socio-economic and politico-cultural normative for ordering the society.
Generating knowledge about Daksiṇāpatha, its peoples and cultures; spread-
ing a new pattern of thinking and transforming the local modes of social existence
constituted the historical context. Brahmanism became the dominant discourse
in peninsular India towards the end of the first millennium CE. Hegemony of
the Brāhmaṇas was a discursively engendered outcome and Jāti hierarchy its
direct fallout, rendered plausible by the sāstraic mode of social representation.
Brāhmaṇas, the primary subjects of the discourse, were directly under the con-
trol of the Vēdic–sāstraic–itihāsic–purāṇic knowledge-power combine, which
decided their perception and appreciation of societal relations. They could not
have conceived the society of functionally specific families attached to their
households except in accordance with the sāstraic prescriptions about the order
33
C. Bougle had shown hereditarily determined occupation, hierarchy and distancing strategy
fundamental to the caste system in his Essay on the Caste System (Cambridge: Cambridge University
Press, 1971), 60–61. For an elaboration of the same perspective see L. Dumont, Homo Hierarchicus:
The Caste System and Its Implications (Chicago: University of Chicago Press, 1980). An expansion
of Bougle’s views is best represented in S.V. Ketkar, History of Caste in India’, Vol. I (Jaipur: Rawat
Publications, 1979).
Gurukkal 47
of occupations and social statuses expressed in terms of the jāti hierarchy. Hence,
it was more of historically and culturally contingent fallout although articulations
of Brāhmaṇas as a hegemonic community were decisive in the process of ordering
the differential relations of production into a hierarchy of jātis.
It is too simplistic to reduce the whole process into a conspiracy of Brāhmaṇas,
for the society owed formation and proliferation of jātis to the material conditions
that gave rise to stratified productive relations. Nonetheless, the discursive role of
the sāstraic organization of status and ranking along categories of economic dif-
ferentiation depended solely on the decision of the Brāhmaṇas. What remains to
be causally linked to the Brāhmaṇas is the imposition of their notion of purity and
pollution upon all with themselves as the purest and hence the infallible point of
reference for determining the relative status of each jāti. Initially, status and eco-
nomy converged but status soon decided the economic privileges as well. In short,
it is not jāti but its status hierarchy, which history owes to the Brāhmaṇas. Even
hierarchy as such, was not their notional construct, for it related to the objective
conditions of differential relations of production, technology of agriculture and
mode of labour realization. It is true that these necessitated institutional or struc-
tural devices of social stratification for stabilizing productive relations through
fetters. However, there is no doubt about the fact that Brāhmaṇas’ ideological
coercion was a key factor of leavening influence.
Proliferation of Jātis
Proliferation of jātis had been an ongoing process ever since jāti became the
dominant paradigm of identity construction for occupational groups and service
personnel claiming socio-cultural distinction. There was a perceptible increase in
the proliferation of jātis during the early decades of second millennium CE. It was
a process at work among the occupational groups and service personnel attached
under land tenure to the three lines of seigniorial jurisdiction, namely, the king,
the chieftains and the Brāhmaṇas. The seigniorial line of the Brahamanas, which
consisted of two service sectors—one of the brahmadēya (brahmasvam) and the
other of the temple (dēvasvam)—managed by the Brāhmanas, seems to have set
the paradigm for the unilineally integrated occupational groups and service per-
sonnel in the other two lines to constitute as well as proliferate jātis. It was natural
because the Brāhmaṇas signified both the source and authority of sastraic norms
according to which the occupational and service people were accommodated
into the jāti hierarchy. However, the unilineal status ordering among them with
degrees of purity based on a set of weird criteria, mimetically borrowed from the
seigniorial domain of the Brāhmaṇas and adapted by independently negotiating
with sastraic norms, was a gradual process.34 The ritual status of the service per-
sonnel within the seigniorial sphere of the Brāhmaṇas was higher for obvious
reasons. Similarly, the political status of the service personnel of the king and
34
C. Meillassoux takes it as an opportunistic strategy. See ‘Are There Castes in India’, 105.
chieftains was higher too. A very significant point to be noted here is that the pro-
liferation of caste was not solely a service-tenure driven social phenomenon,
because there was the dynamic of caste exclusiveness preventing accommodation
of those pursuing occupations other than the ones traditionally given by the caste.
This resistance as an extraneous pressure must have been at work encouraging
hereditary occupational groups to form themselves into endogamous castes.
Rewarded under land tenure, the personnel in service to the king and the
local chieftains became hereditary for stability of service as well as permanence
of family landholding. Exactly as in the case of the temple service, the names of
hereditary offices in the royal and chiefly services became jāti-names of varying
status as determined by their socio-economic and politico-cultural importance as
several instances from the Kerala region demonstrate. Most significant offices
in the royal and chiefly services, such as that of the warrior–chief (Paṭai-nair or
subsequently Kuṛup and Paṇikkar) and of the supervisory accountant (Mēnon).
Nair was a generic term for headship, but the Nair who signified the office of
the warrior–chief was of great importance not only politically, but also economi-
cally, since the Nair signified the largest leaseholder (kārāḷar) group. Most of
the temple lands (dēvasvam), the individual Brāhmaṇa holdings (brahmasvam),
royal lands, and chiefly holdings were leased out to the Nair families. Largest
among intermediaries in land, with the entailing privileges of the nobility, the
Nair matrilineal families provided women for the Namputiri–Brāhmaṇas under
the sambandham system of marriage, which helped them acquire better status,
although, were considered part of the Sudra-varṇa.35 Some of the personnel in
royal service, who eventually became local chieftains, distinguished themselves
from the rest of the Nairs by adopting Kshatriya titles such as ‘vaṛma’. Likewise,
the warrior-heads called Paṇikkar in the royal service acquiring land control
through service tenure distinguished themselves from the Nair and became an
endogamous jāti. In the same way, the office of the supervisory accountant,
Mēnon in the service of the ruling aristocracy as well as Namputiri landlords
who distinguished from the Nair became an endogamous jāti of land control,
socio-political power and ritual status. This process of proliferation of jāti was
continuing even to the late medieval and early modern times under conditions of
service tenure.36
35
Sambandham was a system of keeping concubines by the Brahmins with the Nair women mentioned
in the inscriptions as kaṭṭileṛutal. But it acquired the status of a loose marriage since among the Kerala
Brahmans (namputiris), only the eldest male son was entitled to marry from the same community.
Thus, all the junior members of the family were forced to have sambandham alliance with the Nairs.
See M.G.S. Narayanan, Perumals of Kerala, 270.
36
For a study on a later instance see, F.F. Conlon, ‘The Birth of a Jāti’, in Caste in India, ed. I.B. Dube
(Oxford University Press, 2008), 79–92. There are several instances of service positions under British
colonialism turning into names of endogamous jātis. One instance from Kerala is that of a junior
personnel in the army of Malabar addressed as ‘boy’ (kiṭāvu) became a hereditary occupation
and subsequently a jāti with the name Kiṭāvu (literally child in Malayalam). Proliferation of
jātis continued as a consequence of the colonial practices of classification and grouping as an official
activity under the census administration.
Gurukkal 49
Proliferation of jātis within the Brāhmaṇa varṇa was extensive all over the
subcontinent. An interesting feature in the case of the Kerala region was internal
divisiveness within the Namputiri caste under the Brāhmaṇa varṇa. Namputiris of
Vēdic tradition and land control distinguished themselves āḍhya with all notions
of purity and pollution, from the rest separated as āsya almost to the extent of
constituting them a separate jāti within the jāti.
Arguments
The central argument is that the historical context of the incidence of jāti was that
of the opening up of deltas for agriculture, integration of agro-pastoral descent
groups into hereditary specialists of occupational identity and formation of strati-
fied relations of production transcending kin labour. It appears that the historically
evolved coercive power of the seigniorial control had enabled integration of the
agro-pastoral descent groups into full-time specialized labourers of hereditary
identity as required by the technology of iron plough inevitable for agriculture in
the deltas. Emergence of hereditary occupation groups and promulgation of
sāstraic norms must have been processes of mutuality and concurrence. This was
the context of the Jāti institution of coercive control and seigniorial jurisdiction
over the labouring body for ideologically turning it into the impure and hence
inferior in terms of the sāstraic norms or as ordained by the Bhrāhmaṇas in terms
of purity >< pollution. Organizational constituents of the jāti institution, already
present in the varṇa-structured community, seem to have gradually started acting
as fetters on artisans, crafts folk and tillers whose control was becoming inevita-
ble in the wake of expansion of agriculture into the deltas.
Other arguments relate to antecedents of the jāti hierarchy, discursive
dimension of hegemonic Brāhmaṇism, organization of status hierarchy through
sāstraic prescriptions and service tenure-based proliferation of jātis. A connected
argument is that productive relations in the deltas had preconditions such as
hereditary occupations, asymmetrical social relations, amenability to differen-
tial allocation of status and the dominant presence of the Brāhmaṇas with tacitly
recognized ritual supremacy, resource potential, social control, political influence
and cultural pre-eminence for the emergence of jāti hierarchy. As regards ante-
cedents of jāti formation, the argument is that the service personnel under the
three streams of juridico-political power streams—the king, local chieftains and
the Brāhmaṇas—had laid the foundation of the varṇa-structured society. In the
case of the Deccan and further south, it has been argued that even in the age of
Sātakarṇi and his successors, who took efforts to maintain the varṇa system, the
dissolution of the clan/kin base of productive relations into varṇa base had not
been advancing, for efforts to cultivate the deltas of the major rivers in the region
began only by the mid-first millennium CE. Regarding the situation in the further
south, the argument is that the society was not varṇa-structured till the founda-
tion of brahmadēyas, which coincided with the process of agrarian expansion
along the deltas. A further argument is that the Brāhmaṇas had total hegemony
to prescribe socio-cultural conditions of status and organize its hierarchy with

themselves on the top, and Brahmanism had become the dominant discourse in
peninsular India towards the end of the first millennium CE. What has been argued
in this context is that Jāti hierarchy was the discursively engendered outcome, the
major fallout rendered plausible by the sāstraic mode of social representation and
knowledge production. An argument relating to the process of the formation of
jāti in the wake of the proliferation of brahmadēyas is that the Veḷḷāḷas might have
emerged as a landed jāti due to their close relation with local chieftains.
As regards proliferation of jātis, the argument is that it had been an ongoing
process ever since jāti became the dominant paradigm of identity construction
for occupational groups and service personnel claiming socio-cultural distinc-
tion on the basis of their association with the three streams of seigniorial power.
Rewarded under land tenure, the personnel in service to the king and the local
chieftains became hereditary for stability of service as well as permanence of
family landholding. Illustrating the historical experience of the Tamil south in
general and the Kerala region in particular, the argument found feasible is that
the proliferation of jātis happened as a land tenure-based phenomenon under the
three seigniorial streams represented by the king, the chieftain and the temple—
brahmadēya combine, as realized in terms of the sāstraic norms. This would
not mean that proliferation of caste was solely a service-tenure driven social
phenomenon but there was the dynamic of caste exclusiveness too that pre-
cluded an easy accommodation of new hereditary occupational groups into the
system. It must also have been a compulsion for hereditary occupation groups
to be endogamous castes.

Articles On Caste and Class in Early India 2016

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Articles On Caste and Class in Early India 2016

Uploaded by

Copyright:

Available Formats

On the Translation of the Term Varna

Author(s): Thomas R. Trautmann

Al-Madkhal ild tanmiyatal-a'mdl bi-tahsinan-nyiydt,and also to a lesser extent in

*) This paperwas readat the z6thInternationalCongressof Orientalistsin New Delhi,

ON THE TRANSLATION OF THE TERM VARNA

I) J. H. Hutton, Castein India (3rd ed.), Oxford, I961, p. 64.

or to be byjdti a Chandila or Nisdda, etc., as belonging to the despised part of the

I) A. A. MacDonell, "The Early History of Caste" in The American Historical Review,

meanings in medieval Europe and goes on to say, "That which, in medievalthought,

AN INSCRIBED WEIGHT FROM MATHURA

I) Both specimens are in the Curzon Museum, Mathurd. I wish to acknowledge my

A GLANCE AT THE WORD JĀTI IN THE VEDIC LITTERATURE

The wordjãti formedthe rootjan 'to be born' or 'cause to be born'

We intendtohave a look now atthejãti as itoccursintheVedic literature.

when developed; first there is development, then birth.' 4 Šatapatha

Kãtyãyana Srautasütra(XIV. 2. 32) uses jãti meaning'species' while

Baudhäyana Dharmasütra(1. 1. 10) uses thewordjãti as thepopularly

A surveyof the numberof use of the two words varna (= brightness,

VARNA No. of ufce JATI No. of use

Rgveda 23 JATI nil

It is no denyingthe fact thatthe social conditionsas reflectedin the

8 Tambiah, S. J.'FromVarnatoCastethrough MixedUnions'inJackGoody(ed.)Char-

Social Stratification in Ancient India: Some Reflections

WiththeIndus scriptstillundeciphered,1 in spiteofnearlysixtyyears

power in the Harappan cities being of a republican,oligarchic

That Malik's view of Historyas a disciplineis singularlynarrow-

pursuits,theiragriculture beingsubsistence-oriented. Usingox-yoked

copperor bronze.The taksanor tastribuiltchariotsforwar,65prepared

in thetext,oftenin thesenseofopenand fluidfunctional categories,

iron(syamaor krisnaayas)artifacts on a modestscale about900 BC or

SatapathaBrahmana82 pointto his belongingto theclass of labourers.

One majordevelopment in theLaterVedicperiodwas thebeginning

observedthe ritualsof penitenceand redemption. Theywere to lose

superioritywas challenged;divinesanctionbehindit was questioned;

NOTES AND REFERENCES

1. Despite numerous attemptsand more than fortyclaims to success, the

fullof challenging problems;see 'InterpretingtheIndusScript',in B.B. Lal and

in D. Sen and A.K. Ghosh (eds.), Studies in Prehistory, Firma K.L.

Introduction to theStudyofIndianHistory (hereafter Introduction), Popular

familiarelementof all ancientpolitiesincludingthe Harappan (The Indus

38. S.C. Malik,Indian Civilization:The Formative Period-A Studyof Archaeology

HarappaCultureand categorically statesthattherewas no consciousness ofcaste

54. No othergrainis specified.

83. Kosambi,Introduction, p. 91;Culture, p. 24.

thecountry.Fora detailedexposition ofmyviewsin thisregard,see myarticles

Towards a Historical Sociology of Stratification

? 1992 by The University of Chicago. All rights reserved.

Others have insisted on definitional clarity. Among those who have

Constituting a canon (a finite set of "texts" that are regarded as foun-

or to vac, "speech" or "the word" also in the sense of "truth."Vac is the

10 For the three Vedas as the three forms of

or connections that govern Vedic philosophy,14 the two tripartitestruc-

associated with the three social classes: earth = Brahmins, atmosphere

In the beginning,Prajapatiwas the only one here.He desired,"MayI be, may

The gods saidto Prajapati,"If thereshouldbe a calamityin oursacrificedue to

And in yet another version of the establishment of these same basic

[Prajfpati]heatedup thesethreeworlds.Fromthis worldhe emittedAgni, from

20 Manu 1.23 also

Thenhe developedthe essence of fieryluster(tejas) for this threefoldwisdom,

Combining the components of these two texts to those established

This worldis light (bharga),the atmosphericworldis might(mahas),the sky

And now for this secret teaching (upanisad)concerningall speech.... The

Prajapatigenerated this [world by saying] "bhuh," the atmosphere [by saying]