Académique Documents
Professionnel Documents
Culture Documents
Social History
Review
http://ier.sagepub.com
Published by:
http://www.sagepublications.com
Additional services and information for Indian Economic & Social History Review can
be found at:
Email Alerts: http://ier.sagepub.com/cgi/alerts
Subscriptions: http://ier.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Right
and Left
as a
Root
Paradigm
As an explict division of South Indian society,~ the dual classification of many of its constituent groups into Right Hand (Valangai)
and Left Hand (Idangai) castes, has a history which commences no
later than the eleventh century and ceases to exist, except in vestigial
forms, after the nineteenth century. Its precise historical origin has
not been agreed upon and is in any case secondary to the argument
of this paper. The argument I will try to make is that, as a classification, it has been invoked in the context of several major conflicts or loci of systemic strain during this period. The history of
its appearances is, consequently, a record of its applications, invocations and expressions in a series of local cases of conflict, anomaly
or
competition.
The fact that lists of the constituent groups of the two sides have
varied significantly from each other has generally been a source of
bewilderment to scholars searching for a consistent explanation.
With afew notable exceptions, previous scholarly attempts to
explain the dual classification have been far too partial. They have
suffered from the presumption, and this is true even of the most
sophisticated and comprehensive explanations, that it is some single,
*For their suggestions and criticisms on an earlier draft of this paper, I am grateful
Inden, Nelson, Ramanujan and Tambiah, all of the University of Ch cago. as
well as to N. Dirks and P. Pessar. also of the University of Chicago.
1. By South India is meant that portion of peninsular India covered by the Dravidian language group, specifically by Tamil, Tulegu and Kannada. Kerala, with Malayalam, as its dominant language, does not yield any evidence of the dual classification.
This might be linked to its relatively constant and rigorous integration, under the
ruling Nambudiri-Nayar alliance, which has prevented the sort of vertical cleavages,
whether based in mobility movements or otherwise, that have characterized the other
regions of South India.
to Profs
217
consistent, and substantive property which underlies the diverse appearances of the dual classificatian. Thus we have a host of explanations based on the particular theorists preference for, or proximity
2.
218
The above list, which attempts to be synthetic and includes groups
from the Tamil, Telugu and Canarese countri-s, nevertheless begins
to appear idiosyncratic when compared with other lists, from
other spatial and temporal contexts. Although this list, like most
others, excludes Brahmans from the dual classification, other lists
do include Brahmans, although some of them have been treated,
on prima facie grounds, as erroneous.
While this list suggests that
and
Pallis, Chakkiliyans
(implicitly) Vaniyans have been internally
divided according to the dual classification, other lists suggest that,
in other contexts, Komatis, Kai-Kolans, Saliyans and Gollas have
been similarly partitioned. Likewise, the variation in the number
of castes that are held to belong to either side is also enormous5.
There is also some contextual variation in the side to which a particular caste is said to belong,. Some lists include immigrant Northern
groups like Gujaratis while others do not. Unlike most other lists, ,
this one completely excludes the higher agricultural castes, whereas
others often include at least Tamil and Canarese dominant agricultural castes. This is all meant to suggest that, inspite of certain
broad consistencies, the contents of the two sides of the dual classification vary significantly over space and time. This variation,
improperly understood, has generated a host of explanations which
variously reduce the basis of the division to a conflict between arti-
6.
Oppert.
couver,
219
and merchants, artisans and agriculturists, local merchants
and regional merchants&dquo;. mobile, urban elements and fixed, landed
elements and so on. It is not only with respect to the contents of the
classification that the data are heterogeneous. They are no less so in
the matter of the issues over which these two groups have come into
conflict. Though, at a high level of generality, these disputes tend
to be around the respective monopolies of the two sides over certain
practices, emblems and spaces, here again contextual variation is the
dominant motif.
To understand how the dual classification performs its cultural
function, it is necessary first to locate the metaphor on which it is
based in the broader context of the Indian cultural system. The
Tamil terms Valangai and Idangai (as well as their Telugu and Canarese equivalents) refer to the left and right hands or arms of the
human body. But the terms are best understood to refer to the
right and left sides of the body. Let us consider now the meaning
of the metaphor of the body for society, in the Indian cultural
sans
system.
In Manu Dharma Sastra, the classical moral codebook of later
vedic society (200 B. C. to 200 A. D.), we are presented with the
image of the body of Purusa, the original man, as a symbol of
society. Purusa was divided by the gods into four varnas. The
Brabmana, born from the highest part of Purusa, his mouth, was the
highest varna : he was to teach the Vedas, perform sacrifices for the
Ksatriya and Vaisya, and accept gifts from them in exchange.
The Ksatriya, born from the arms of Purusa, possessed royal
power and was to fight enemies, give gifts and food to the Brahmana.s, and protect the Vaisya; in exchange, he received a share in
the leavings of the sacrifice from the Brahmana and wealth from
the Vaisya. The Vaisya, born from the thighs of Purusa, possessed
productive power : he was to produce wealth, and was to give a
share of it, as taxes, to the Ksatriya in exchange for protection.
Rice, op. cit., p. 223.
9. This tends to be the most frequent explanation : see, for example, Manual of the
Administration of the Madras Presidency, op. cit., Vol. I, p, 69; also, Oppert, op. cit.,
8.
p. 58.
10.
Srinivasa
Aiyangar,
op.
cit., p. 100.
220
While these three varnas were &dquo;twice-born&dquo; men divinized through
a second, ritual birth, the fourth varna. the
Sudra, was &dquo;once-born&dquo;,
from the lowliest part of Purusa, his feet, and was capable only of
performing service; i.e. labour, for the three &dquo;twice-born&dquo; varnas.
This four-fold, horizontal division of the social body, rooted in the
ranking and exchange activities of the Vedic sacrifice, provides the
basic symbolic template for ordering the plethora of jatis of which
actual Indian communities are composed. However parochial the
classification and description of jatis within local and regional ranking systems, they are, in Vedic theory, products of the miscegenation of the four original varnas and responses to the requirements
of particular historical contexts.
Two major peculiarities characterize the realization of the Varna
scheme in South Indian caste systems. In most ,of South India,
because of the high regard given to female bodily substance, both
parents contribute to the generic definition of their oflspring : as a
result, endogamous castes tend to divide into numerous smaller
circles, who often mark their distinctions with visible attributes.
Thus the relatively small spread of each caste permits any given caste
to be ranked relatively consistently by the other castes in only a few
contiguous localitiesll. The other maor peculiarity, whose historical reasons are outside the scope of this paper, is that in South
India there are no clear-cut occupants for the second two categories
in the varna scheme, the Ksatriya and the Vaisya, although there
are perfectly distinct Brahman and Sudra jatis.
These two regional
the
have
a
to
do
with
conflicts
that lie at the
deal
peculiarities
great
core of the classification of South Indian society into right and left
hand castes.
The notion of right and left hand castes bifurcates this horizontally
segmented social body into two sides. Just as the horizontal layers
of the social body in Vedic symbolic representation refer to an ideal
state of social integration, characterized by exchange and solidarity,
so also the notion of a vertically divided social body is the symbolic
The entire preceding discussion of the metaphor of the social body in the Vedic
period and its model of social differentiation, is taken from McKim Marriott and
Ronald B. Inden, "Caste Systems", Encyclopaedia Britanaica, 1974, (forthcoming,)
11.
pp. 6-8.
221
the South Indian cultural system orders
the conflicts of real social situations, whether these be cognitive,
economic, political or social. Just as the segmented body of Purusa
is the basis for the cultural ordering of social groups, so the metaphor of the bifurcated social body is the specific response of the
South Indian cultural system to the anomalies, oppositions and
conflicts generated by the imperfect actualization of this cultural
model in the social and historical reality of South Indian society. As
in other cultural systems, the left hand in South India has connotations of impurity whereas the right hand has powerful and positive
normative associations. In India this polarity is reinforced by the
non-duality of the moral and natural orders and their free capacity
to exchange attributes, meanings and properties. This asymmetry
adds considerable affective and emotional power to the symbol.
But the metaphor of the vertically divided social body is no
ordinary metaphor. In its capacity to confer meaning on, and inspire
action in, a variety of historical situations, it is what has been
called a root paradigm!!. : these are consciously recognized cultural
models which emerge during the life-crises of individuals or groups,
and have reference to the social relationships of those involved, as
well as to the cultural, ideological or cognitive patterns which incline
them to alliance or diviseness. Root paradigms, in Turners definition, are neither precision tools of thought nor are they stereotyped
222
be human or congnitive, is the formal function of the notion of right
and left hand castes, which I shall henceforth refer to as the root
paradigm.
To understand its formal structure, let us consider the strict definition of the notion of a &dquo;paradigm&dquo;, in linguistic theory 13 :1
can
By contrast is meant that quality of certain linguistic units whereby the substitution of one for the other in a certain context alters
that context, and the context of a linguistic unit is specifiable in
terms of its syntagmatic relations&dquo;. The dual classification is in
this strict sense a paradigm, whose constituent units are themselves
a series of contrasted pairs (viz. Komati and Beri Chetti; Mala and
Madiga; Kammala and Vellala etc : see following section III).
Each of these contrasting pairs activates the root paradigm in a
different context (i, e. at a different level of the caste hierarchy as
well as in varying historical and social circumstances), causing a
wide variety of sytagmatic expressions (i.e. particular episodes of
conflict) in each context. This formal fact underlies both the variety
of historical expressions of the root paradigm as well as the variety
of indigenous stories concerning the origin and meaning of the root
paradigm.
There is thus no single substantive meaning for the dual classification. There is a formal structure which is contrastive, contextual
and paradigmatic, and a formal function which is integration, that
constitute its essence. It is thus a single paradigm, but in each of
John Lyons, Introduction to Theoretical Linguistics, Cambridge, 1968, p. 73. I
to Prof. A. K. Ramanujan, University of Chicago, for suggesting to me
that this set of linguistic concepts would clarify my argument.
14. Ibid. p. 67 and p. 74.
13.
am
grateful
223
its contextual applications, it has a different meaning. To understand how this root paradigm has functioned in South Indian history,
I turn now to mapping some of its contextual and syntagmatic
-
expressions.
II.
South Indian
History
B. A.
p. 68.
16.
224
paradigm
consistently
225
.
lar in the Tamil country, and Kamsalis in the Telugu country. This
group was probably based on a pre-Aryan artisan guild, which was
most likely low-ranked until the great temple-building epoch starting
in the tenth and eleventh centuries, when the demand for their services suddenly increased21. Their subsequent affectation of Brahmanical status, and the negative reactions that these struggle have
provoked, usually from the dominant Tamil agricultural caste, the
Vellala, remain the dominant motif of the history of the root paradigm. This particular struggle, and the position of the artisans on
the left hand side, often in a leadership capacity, have been sufficiently consistent across space and enduring over time, that many
analysts have been tempted to make it the basic explanation of the
dual classification22. In fact, it is only one of its more ubiquitous
manifestations, and probably the occasion of its legislation. It must
21. C. S. Srinivasachari, "The Origin of the Right and Left-Hand Caste Divisions",
Journal of the Andhra Historical Research Society, Vol. IV, Pts. 1 and 2, (July and
October 1929), p. 80; for discussion of efforts on the part of these artisan jatis, to
improve their status. particularly in the context of medieval temple-building, see Srinivasa Aiyangar, op. cit.,
pp. 108-109; also,
especially pp. 18-20, 26-28, 35.
K.
Sundaram,
op.
22. Of the many analysts who have offered this explanation, the most recent,
elaborate and comprehensive is Brenda Beck, in "The Right-Left Division of South
Indian Society", Journal of Asian Studies, Vol. XXIX, No. 4, August 1970, pp. 779-798
and in Peasant Society in Konku, op. cit., Becks argument, which makes the contrast
between artisans and Vellalas in the region of the South called K onku Nadu, the basic
explanation of the dual classification, is a penetrating analysis of one regional variant
of the dual paradigm, but is far too partial to work as a general explanation. The
problems that this analysis encounters are instructive. The fact that Beck found
Komati Chettiyars classified on the left, whereas others classifiy it on the right
(Peasant Society in Konku, p. 74), is not simply an ethnographic anomaly but an
indication of the regional and contextual variation which lies at the heart of the root
paradigm. Similarly, the fact that the low est castes were the most ambivalent in their
choice between the two models (Brahman-ritual and Kavuntar-instrumental) that
underlie, in her argument, the right-left division, and yet were the ones who were
most vociferous in the right-left disputes, suggests that these two models, while conceivably represented in divergent lifestyles, should not be overstressed. Hindu social
thought, which is not characterized by such dualities as nature and morality, sacred
and profane, this-wordly and other-worldly etc., does not contrast ritual and instrumental behaviour. Rather, both Brahmins and Kshatriyas, in the fulfillment of their
226
,
also be remembered that the artisans struggles for higher status are
a much more general upward mobility amongst craft and
trading groups in medieval South India, some of which is connected
to the inclusion of other groups, such as weavers, in the lists of the
root paradigm. Incidentally, medieval mobility movements, especially of artisan groups associated with temple-building, cast some
light on why the dual classification never took root in Kerala, the
Malayalam-speaking part of South India : Kerala did not undergo
the massive temple-building phase experienced by the rest of South
India, and thus did not have to respond to a sudden and powerful
disturbance to the social order.
Finally, there is some evidence that the struggle for supremacy
between Brahmans and Jainas, which affected particularly the Pallava
and Kadamba countries in the ninth and tenth centuries, eventually
expressed itself in terms of the root paradigm23. In a Mysore inscription of 1368 A. D. we are told that when the Brahmans and
Jains fought over certain symbolic privileges, the then king of
Mysore, Vira Bukka Raya, effected a compromise between the two
sides, whereby the Jainas were classified with the right&dquo;.
It is thus not hard to see that, even from the ninth to the thirteenth
centuries, during the Pallava-Chola period, the remarkably common
interpretation that the division of society into right and left hand
castes represented the struggle between landed, stable agricultural
groups on the one hand and mobile, urban artisan and merchant
part of
..
227
groups on the other, is really a rather strained synthesis of what
were, in fact, widely varying local and contextual variants of a single
cultural paradigm.
During the Vijayanagara period, the integrity of local peasant
institutions for taxation and ordcr was destroyed, and replaced by
considerably more direct and far-reaching control of agrarian resources by Telugu Nayakas, with peasants bearing the heavy burden
of being on the bottom of a vertical system of tributes2~. In this
context, there is considerable evidence to suggest that the root patadigm provided the basis for the formation of stable, corporate
groups, particularly of the lower castes, with a distinctly political
and self-protective character26. They apparently often combined
together to oppose excessive taxation, whether by local Brahman
and Vellala landholders or by ofhcials sent by political chiefs. The
interesting thing about these contextual expressions of the root
paradigm is that they provide an empirical demonstration of the
fact that the two sides of the social body are contrasted but also,
ultimately, two sides of a single, unfied social body.
Secondly, there was, during the Vijayanagara period, an unprecedented degree of urbanization, which had three sources. In the
first place, both the requirements of defense and the wish to emulate
the Imperial Court dictated that the Telugu nayakas of the period
create fortified headquarters. Also, the autarchic economic policies
of this warrior elite, their wish to maximize control over resources
in their particular territories, led them to attract artisans from other
places in order to achieve self sufhciency in the stable artisan products, particularly in cloth. Finally, these warriors supported and
subsidized the enormous growth in temple centers during the Vijayanagara period. These temples, qua pilgrimage centers, also often
acquired a market function, and as a means for the acquisition of
prestige, invited endowments both from these warrior elites, as well
As a consequence, temples
as from other upwardly mobile groups.
became themselves involved in redistributive and developmental
tasks in the agrarian economy, of both a technological and a finan,
Burton Stein, "Integration of the Agrarian System ...", op. cit., pp. 191-192.
T. V. Mahalingam, Administration and Social Life Under the Vijayanagara
Empire, Madras, 1940, p. 221 and pp. 92-95.
25.
26.
228
cial sort. This urbanization appears to have placed a strain on the
dual classification leading to a series of disorders in Vijayanagara
cities : these disorders are the first in a continuous urban phase in
the manifestations of the root paradigm2.
The most broad and interesting of the spatial variations in the
appearances of the root paradigm is the one we have already
observed, namely a tendency for the root paradigm, at least in its
more violent manifestations, to appear more often in urban rather
than in rural contexts. In addition to this broad shift, however,
there are more disaggregated spatial variations. One way of approaching this spatial variation is by considering the spatial contexts in
which the root paradigm has been activated, by contrasts or variations at different levels of the caste hierarchy.
At the upper levels of the caste hierarchy, sectarian conflicts of
various sorts seem to have activated the root paradigm. One example
would be the conflicts between Brahmans and Jainas in those urban
centers in which the Jainas had the strongest following. The conflict
between Vaishnava and Saiva Brahmans also apparently expressed
itself through the metaphor of the vertically divided social body28 :
this could have occured in these religious centres in the Canarese
country, where both Vaishnavism and Lingayat Shaivism had their
most radical expression in the medieval period. According to a
recent ethnographic study in the Telugu country, the connection of
the Shaivite/Vaishnavite opposition to the root paradigm, although
attenuated and modified by changes over time, is still recognizable.
In certain districts of this area, Vaishnavite Brahmans are recorded
as having openly associated themselves with the right hand group
(although Brahmans are not supposed to be involved in these schisms)
and three groups of the left hand side, the Jangam (Lingayat priests),
Devanga (weavers) and Kamsali (artisans) are extreme Saivites.
Perhaps the most important conflict at this highest level of the caste
hierarchy, however, is the one that I have already described, which
Stein, "Integration of the Agrarian System...", op. cit., pp. 195-196.
Saletere, op. cit.,
pp. 66-67; see also story No. 4 in Section III.
N.S. Reddy, Transition in Caste Structure in Andhra Desh with particular
Reference to Depressed Castes. Ph. D. Thesis, University of Lucknow, 1952, pp.
166-167.
27.
28.
29.
229
involves artisan claims
to Brahman status. In the form of the conflict between the artisans and the dominant agricultural castes, the
dispute seems to have been most common in the Tamilcountry, and
to some extent in the Canarese country. The most important Telugu
agricultural castes, such as the Reddis and the Kammas, seem to
have avoided this classification. It ts interesting, as well, to note
that artisan claims to Brahman status have not always been challenged by agricultural castes, but have occasionally been challenged
and protested by Brahmans themselves3. It could be said, in general,
that the root paradigm has been activated at this level of the caste
hierarchy, as a response to a variety of threats, both internal and
external, to the homogeneity, exclusiveness and primacy of Brahmanic status in indigenous systems of rank.
The formal absence of clear occupants for the next two slots, i.e.
that of the Ksatriya and the Vaisya, in the varna system as it was
established in South India, sets the stage for the root paradigm to
express an extremely heterogenous set of contrasts and conflicts
among agricultural, commercial and craft groups. In this sense, the
extent to which Vellalas, the dominant Tamil agricultural caste,
resist artisan claims to Brahman status, is, in part, a measure of
their claim to Ksatriya or Sat-Sudra status3l.
When we consider the alignment of various merchant groups
according to the dual classification, the disaggregated picture is very
complex. We have already referred to the original appearance of the
Banajigas and Nagartas at the head of military organizations in the
Chola ,period. In the modern period, in Mysore, these two groups
express their conflict through the dual classification. The Banajigas
are a Canarese trading group, who along with the Linga (i.e. Saivite)
Banajigas lead the right hand group, while the Nagartas and the
Panchalas lead the left hand group. In Mysore, this split is associated with two regional peculiarities. In the first place, the total
number of groups that are alleged to be divided into right and left
hand groups are eighteen, and are called panas (professions),
30. G. Oppert,
op. cit., pp. 111-112; see also James F. Kearns, "The Right and the
Left-hand Castes", Indian Antiquary, Vol. 5, December 1876, pp. 353-354.
31. Mahalingam,
op. cit., pp. 239-240.
230
although lists in practice contain many other castes : there was also
apparently competition between the leading groups of the two sides
over which panas belonged to their respective sides.
Secondly, it is
in the Mysore region that the two sides are termed Desa (outsider
It was this specific local
or foreigner) and Nadu or Pete (natives).
variant which led one observer to hypothesize that the division was
based on a struggle between &dquo;followers of the old-established handicrafts and innovators who brought in the exchange of commodities with other parts, supported by producers and ministers to
luxury&dquo;32. One variant of the tradition concerning the origin of the
root paradigm suggests that the Balijas, one of the two great Teluguspeaking pan-South Indian trading castes, and the Nagartas, urbanbased Telugu traders mostly based in Canarese areas, also expressed
their conflicts, possibly in Canarese trading centers, in terms of the
dual classification33. Also mentioned on many of the lists, and
almost always on the left-hand side with the artisans, is a merchant
group called the Beri Chettis : this used to be the name for the entire
Nagarta community, but in the modern period has come to denote
a Vaishnavite subsection of the Nagartas, almost invariably engaged
in urban retailing activities. Finally, among the most important of
the commercial oppositions to coalesce around the root paradigm,
which also is important in Madras city in the seventeenth and eighteenth centuries, is that between the second great Telugu trading
caste, called Komatis, and the Beri Chetti caste. Ever since their
eclipsing of the great medieval merchant communities under the
patronage of the Telugu warriors of the Vijayanagara period, the
Komatis have been vociferous claimants to Vaisya status34. Most
often, these claims are made in a mutually exclusive fashion with
those of the Beri Chettis, who made equally vigourous claims to
Vaisya status35. In general, it could be said that conflicts and oppo-
231
country.
.
Finally,
sorts
36.
Reddy,
op.
cit.,
pp.
167-180.
232
I
is that which occurs between several homologus pairs of &dquo;untouchables&dquo; : Pariahs and Chakkiliyans in Tamil-speaking areas; Malas and
Madigas in Telugu areas; and Holeyas and Madigas in the Canarese
country37. Both the longevity and the relative consistency of this
type of dispute across regions, are no doubt connected to the relative
immobility and static identity of these groups in the last eight centuries in South India. Existing explanations of the passionate
involvement of precisely these groups in conflicts of right and left,
which relate this fact to their unenviable structural position, seem
to be correct.
The unity and essence of the dual classification lie in its formal
and functional aspects, namely to express a series of conflicts (viz.
Komatis vs. Beri Chettis), anomalies (artisan claims to Brahmanic
identity would be an example), and schisms (such as the one between
Vaishnava and Saiva Brahmans) in terms of a cultural paradigm
whose core is the notion of a single social body, albeit one vertically divided into two sides. By being conceived in terms of this
particular contrastive paradigm, a variety of regionally defined and
potentially dysfunctional stresses in South Indian history are rendered both culturally meaningful, and, in principle, adjudicable. These
stresses presented challenges to healthy social integration because
they probably were rooted in such processes as migration, corporate
fission, economic competition and social mobility, which are not
confined to medieval and modern South India. But the cultural
context in which they occured i. e. the peculiar realization of the
varna scheme in South India, and the particular historical development of South Indian society, exemplified, at a macro-level, by the
fundamental cultural and institutional role of the South Indian
temple, do indeed distinguish South Indian society from the eleventh
to the nineteenth century, from other places and times. It was this
particular configuration of widely observed processes and pressures
with a specific cultural and institutional history, that provided the
conditions for the creation and currency of the cultural paradigm
37.
op.
G. Oppert,
cit., passim.
op.
cit.,
pp.
65-66; Thurston,
op.
233
of right and left hand castes, and its restriction to the Tamil, Telugu,
and Canarese areas of peninsular South India.
This section has been dominated by a discussion of some of the
spatial and temporal parameters according to which the root paraof its contextual expressions. But there is no
understand, at the cultural level, how a single paradigm generates many syntagmatic expressions, than by considering a
sample of the stories that have been recorded at various places and
times regarding the origin and nature of the distinction between
right and left hand castes. This sample, as well as a brief analysis,
are contained in the following section.
digm acquired
some
better way to
III.
Indigenous Explanations
of the Root
Paradigm
ment.
regional fragments38 :
In the time of the Soren Raja Parimalan, Veda Vyasan endeavoured to induce the king to allow his family to perform the
38.
James F.
234
sacred officers for the royal family; but the Raja declined, saying,
The Panchalar (Visva Brahmans) perform them very very well,
and he desired Vyasan to take his leave. The Raja died shortly
afterwards, and his brother succeeded him, whereupon Vyasan
made another attempt to have his family appointed, but the new
king repelled him rudely. Vyasan then went to the illegitimate
son of the late Raja, and by false stories stirred him up against
the Raja, and the Panchalar, and obtained from him a promise
that he should be made priest for the royal family on condition
of his deposing the Raja, and raising him to the throne.
Accordingly,
illegitimate son
the
was
-.
235
-,
sion
and
to carry
p.
109.
236
sects enter into
4. Another variant appears in a manuscript called Idangai Valankai Kaifiyut, described in Taylors Catalogues Raisonees, III, p. 7
and appears to have been produced in the context of disputes between Vaishnava and Saiva Brahmans:41
This relates to the great dispute between the Vaishnava Brahmans with their followers who have the epithet of right-hand,
and Saiva-Brahmans, with their followers, termed left-hand.
The dispute is stated to have arisen from the usage of a garuda
banner, or flag bearing the eagle or kite of Vishnu, as a device.
The right of bearing this banner, and the question of which of
the two classes it belonged, created so hot a dispute, that the
matter was referred in arbitration to Vicrama-Chola-DevaPerumal, in Cali-Yuga 4894, Paritabi cyclic year. That prince
caused the old copper-plate records at Conjeevaram to the disinterred and examined, and legal authorities ti be consulted.
As a consequence the claim of Saivas to the Garuda banner
was admitted; but another result was, the more accurate dis41.
Saletore,
op.
cit., p. 66.
237
comes
from
from G.
India43:
comes
or
Oppert,
on
The five classes of artisans-the carpenters, goldsmiths, blacksmiths, braziers and masons, well-known in South India as
Panchalar or Kammalar-regard themselves. as the descendants
of the divine artificer Visvakarma, call themselves Visva Brahmans. They assume the title of Acharya, wear the holy thread,
42.
43.
Srinivasa
238
right
themselves, especially
at
pp. 117-118.
generated
one
44.
239
best of the division, and all the other castes joined in, and
took the side either of the Kammalan or the Balija.
(69 in all).
9. Conflicting and mutually exclusive claims for Vaisya status
between the Beri Chettis and the Komatis provide the context for
the following two variants, which inspite of the fact that they are
etc.
.-
240
not
explicitly
to me,
on
seem
waterpandal,
46.
&dquo;
241
.
are
called the
Pillaipuntha Komati,
or
10.
allocation in Conjeevaram to the time of Karikala Chola, whom they date in the 12th century. In the days
of the Nabobs of the Carnatic they seem to have been frequently brought by their self assertion into conflict with the authorities, and a story is current, which appears to bear the impress
of truth, of their persecution by one Salva Naik (said to have
been a Brahman), the result of which was that large bodies of
them were forced to flee from Conjeevaram to Tanjore, Madura
and Tinnevelly, where their representatives are still to be found.
They are violent partisans in the disputes between the right and
They
left hand
castes.
The tradition is, that Karikala Chola divided the people into
these two parties, assigned 98 tribes to each, and appropriating
to their use distinct flags and musical instruments for festivals
and funerals. This distinction, established at first for political
reasons, or to prevent disturbances, has for the last several centuries been the constant source of contention, and has even
given rise to serious tumults. ,
11.
the following variant of the tradition, said to be current amongst the lowest untouchable groups in
Mysore, the Holeya and the Madigag8 :
At a remote period, Jambava Rishi, a sage, was one day questioned by Isvara (Siva) why the former was habitually late at
the Divine court. The Rishi replied that he had personally to
47.
C. S. Crole, Manual of
48.
242
attend to the wants of his children every day, which consequently made his attendance late : whereupon Isvara, pitying the
children, gave the rishi a cow (Kamadhenu), which instantaneously supplied their every want. Once upon a time, another
rishi, named Sankya, visited Jambavas heritage, where he was
hospitably entertained by his son Yugamuni. While taking his
meals, the cream that had been served was so savoury that the
guest tried to induce Jambavas son Yugamuni, to kill the cow
and eat her flesh; and, inspite of the latters refusal, Sankya
killed the animal, and prevailed upon the others to partake of
the meat. On his return from Isvaras court, Jambava found
the inmates of his hermitage eating the sacred cows beef; and
took both Sankya and Yugamuni to Isvaras court for judgement. Instead of entering, the two offendors remained outside,
Sankyarishi stood on the right side and Yugamuni on the left
of the doorway. Isvara seems to have cursed them to become
Chandalas or outcastes. Hence Sankyas descendants are, from
his having stood on the right side, designated right-hand caste
or Holayas; whilst those sprang from Yugamuni and his wife
Matangi are called left-hand caste or Madigas.
243
for
244
two groups, followed by the dual division of society itself, is
the theme of the fundamental unity and complementarity of these
two halves of the social hody. This theme finds various kinds of
expression in these various traditions. Spatial symbolism is the
most direct way in which this complementarity is expressed : the
right and left sides of the King or Goddess, and the right and left
sides of the entrance of the Divine Court are examples. In some of
the stories, involving artisan claims to Brahman status, the unity of
the warring halves is expressed in a normative way : thus, the primacy of the Brahman status in a single social system of rank, is
never contested, although there are disputes concerning who the
&dquo;real&dquo; Brahmans are. Similarly, the several references to a copperplate or set of inscriptions, also constitute a set of legal and moral
symbols of the ratified and complementary (and presumably consensually accepted) rights and privileges of the two halves. Even
more interesting as expressions of the basic unity of the two halves,
are the several cases where competing claims to dependancy are
expressed in kinship terms : examples would include the conflicting
claims of Kammalas and Vellalas about being each others jatipillaigal 50 (sons of the caste); the Beri Chetti characterization of the
Komatis as Pillaipuntha Komatis (Komatis who became sons); the
common paternity, in one variant, of the Kammala and Balija sons.
The two most concrete expressions of the unity of the opposed
halves are the symbol of the literal dismemberment of the kings
body by the Balija and Kammala sons and the sharing of the cows
ween
245
this problem than through the king, who is charged with the proper
maintenance of the social order. Thus, the fact that, in many of
the variants, the king is weak, indecisive, killed, illegitimate, unjust,
dismembered or remiss in his duties, is the clearest indication that
the genesis of these various conflicts, and the subsequent creation of
two distinct social halves, is a product of weak integration.
Having established some of the basic themes that seem to run
through these stories, let me now consider some of the ways in which
they differ from one another. The most important principle by
which these traditions vary from one another is part of my general
argument : it is the principle of context, whereby each particular
contrast expresses the root paradigm in its own way. The traditions
express this variation both thematically and symbolically. The highlevel oppositions, particularly those which involve artisan claims to
Brahman status, since they affect the &dquo;head&dquo; of the social body, are
replete with themes of authority, legitimacy and rule, both priestly
and kingly. The middle-level stories, particularly the ones that
involve commercial groups, do not raise these themes explicitly, but
assume them, and are explicitly preoccupied with competition for
ritual objects, like the waterpandal or technical instruments like the
weighing pan. Also, their natural and contextual preoccupation is
with such commercial themes as fines, debts and loans. At the very
bottom, and consistent with the general position of untouchables in
the ranking system, the predominant theme is a shared crime, the
eating of beef, and its just punishment : these are pervasive elements
in the belief-systems of untouchable groups across India : Thus,
even in terms of the level of the caste hierarchy at which the primary
contrast occurs, the root paradigm of the bifurcated social body
acquires its cultural content on a disaggregated and contextual
basis.
-
IV.
The Case of
Early
Madras
246
of which became economic centers, attracting artisans and commercial groups from surrounding areas. These cities, like pre-modern
Indian cities in general, owe their growth in population to what one
analyst has pithily described as &dquo;protection, prestige and proximity.&dquo;bt Their attraction, in other words, was provided by the
protection they provided, especially when they were fortified, to the
native population against the insecure civil conditions of much of
the countryside; by the opportunities they offered ambitious groups
and communities, particularly in the middle, commercial levels of
the caste hierarchy to view with each other in the emulation of royal
and courtly lifestyles; and by the proximity that most of them represented to opportunities for the pursuit of economic gain, both by
productive and entrepreneurial activities as well as by commerce.
In these three respects, the early port settlements of the period from
the sixteenth to the eighteenth centuries served as &dquo;similar points of
attraction&dquo;. That the European, and particularly the English, port
settlements did not alter this basic urban pattern is not only established by empirical comparison, but makes considcrable deductive
sense, given three basic factors. Firstly, the colonials rapidly involved themselves in indigenous notions of diplomacy and landcontrol. Secondly, they had no especial wish to alter the internal
social or political attangements of the native inhabitants of these
settlements. And thirdly, it was not until the nineteenth century
that they profoundly altered the technological, commercial or social
basis of the indigenous economy.
This broad morphological continuity between the urban centers
of the Vijayanagara and early colonial periods provides the background for the fact that conflicts around the root paradigm of right
and left hand castes seem to be especially violent in these settings52.
51. John Brush, "The Morphology of Indian Cities", in Roy Turner (ed.) Indias
Urban Future, Berkeley, 1962, pp. 64-66.
52. Throughout the colonial period, a host of conflicts around a wide range of
ritual and spatial issues are recorded in cities all over South India : in the neighbourhoods of Porto Novo and Cuddalore in the nineteenth century; at Kanchipuram in the
18th century; at Tegnapatam, Nagapatnam andPulicat at various times in the early
period of British rule; at Dummagudam in the Godaveri District in the second half of
the 19th century; at Sriringam soon after the British won it from Tipu Sultan in the
second half of the 18th century; and in Madras city in the middle of the 17th century
247
53.
pp.
248
on
249
in terms of direct extortion, but even sometimes, as in the
of the painters, by being themselves saddled with some of this
downward flow of &dquo;debts&dquo;.
This was thus a sort of dual economy. In theory it consisted of
a flow of capital, generally money rather than goods, from the East
India Company to its representatives in Madras, through the native
intermediaries, down to the textile-producing castes, resulting
in the return of textiles to England. In practice, it consisted of a
complex network of extra-legal transactions centered on the private
trade and profit both of the English factors as well as the native
intermediaries, which was made possible by the misappropriation
of the Companys capital and by a variety of pressures on the
native population. The paucity of the salaries of the factors, the
rudimentary development of the notion of conflict of interest, and
the loose control of the parent company over the activities of its
agents in the East, all supported and sustained the basic contradiction of this economy, which was between the official pursuit of
profit by the East India Company, as a joint economic entity, and
the private pursuit of profit by its agents in the East.
This peculiar and stressfull economic situation had, as its most
serious social consequence, the several riots of 1652-4, whose basis
was the organization of the native population into two factions.
The reason for their genesis was that the native intermediaries
needed followings in order to compete with each other, to display
and trade their respective strengths in their dealings with the
English, and to command the allegiance of both the textile-producing castes as well as of such commercial castes as the Beri Chettis.
The political and economic standing of these intermediaries eventually turned on their capacity to control the labour of the producer
castes, and the commercial outlets of the shopkeepers.
These factions worked in a number of ways. The faction leaders,
the Brahman brothers Venkata and Kanappa on the one hand, and
the Balija merchants Seshadri and Koneri Chetti (with the support
of Rudriga and Timanna) on the other, constantly competed for
the control of various groups of painters and weavers, as well as
of other castes like the Beri Chettis, the Komatis, and the Pallis,
since the allegiances of these groups did not seem to be cleancut or
not
only
case
250
continuous.
persuade the
to
251
in which the root paradigm of right and left hand castes
provided strong affective impetus to action is best seen in its
expression in conflicts over spatial questions. Because Madras had
not yet completed its process of growth and spatial and residential
definition, and because of the migration of relatively widespread
groups to early Madras, space provided the perfect medium in which
to give concrete symbolic expression to factional conflict conceived
in terms of the root paradigm. This was so in three respects. In
early Madras, as in indigenous Indian cities in general, caste groups
tended to live in relatively segregated areas, and thus it was not
difficult for factions to express their conflict through spatial monopolies and definitions. A documentary award of 1652, allotting
particular streets to the right and left hand sides, both for residential and ritual/festive purposes, seems to have been supported and
defended by leaders of both sides, occasionally altered or manipulated, and even suppressed and redisplayed, depending on the particular juncture in the factional conflict. Existing spatial definitions
were extended to include new potential participants into the factional dispute, probably on the basis of prior regional expressions
of the root paradigm: &dquo;a difference occuring between the painters
and a Palli, at Greenhills order the Brahmans procured a written
agreement that the pallewaru should go with their weddings into
any streets, only reserving the Committee (Komati) Streests for the
The
sense
a
Belgewars honour&dquo;59.
Secondly, particular spatial divisions, once established for residential or ritual purposes, were a perfect medium for confrontation
and provocatlon in the course of factional conflict. When, at one
stage of the conflict, the Brahman brothers were opposed to the
Beri Chettis, at the wedding of their accountant, they &dquo;tried to
bring a palankeene into the Berewars Streete&dquo;60, and the resistance
of the Chettis caused some of them to be imprisoned. When
Seshadri Chetti succeeded in getting the mooree weavers on his
side but failed to win over the caingaloone weavers, he &dquo;made a
59.
60.
Foster,
op.
cit., p. 59.
252
Madras at this time, and the root paradigm of right and left hand
castes, it is not surpprising that the English considered spatial ordinances to be the best method of dealing with these disturbances.
Neither is it surprising, however, that these measures rarely succeeded.
But it is important to note that the English were, in part, incapable of controlling the conflict, because they were themselves internally divided on a factional basis, along lines which were both
isomorphic with, and connected to, those which divided the native
population. President Baker appears to have been a supporter of
the Brahman brothers partly out of weakness, and partly out of
Ibid. p. 258.
253
Government6 2.
Let me now restate my argument concerning the relationship of
the root paradigm of right and left hand castes to the economic
and social context of Madras in 1652-4. The social and economic
system of Madras in this period was under severe strain, manifested
in factional developments. To the extent that this factionalization
was built around the oppositions between castes from diverse local
systems that had conceived of their local conflicts in terms of the
root paradigm (viz. Balijas and Beri Chettis; Komatis and Beri
Chettis; Pallis and Painters etc.), the cultural schema of right and
left hand castes seems to have been activated. It worked in two
ways. On the cognitive level, it served to define, in a neat dual
opposition, a highly complex (and presumably cognitively strenuous) network of competitors and alliances. On the affective level,
it was given expression in highly specific (though unstable) spatial
privileges and monopolies. As symbols, these spatial definitions
served to express the self-definition, solidarity and prestige of each
side. The affect inspired by this aspect of the root paradigm was
at the existential core of the violence which characterized the situation. Thus we have a seeming paradox : in responding to a political and economic situation that harbours dangers of disintegration,
the
Although it is beyond the scope of this paper to provide a detailed examination of the evidence from other urban situations in
the colonial period, these cases do seem to fit my general argument.
In general, a large number of these conflicts appear to have commercial conflict between indigenous commercial groups at their
core : this is true of Madras, Pulicat, Tegnapatnam and Nagapatnam in the eighteenth century and Chinna Ganjam in the nineteenth cent~y.g3 The Madras cases from the eighteenth century
62.
Love,
op.
254
merit notice, with respect to a few striking points.&dquo;
These later cases provide further evidence that, in urban situations, the root paradigm becomes an overall cultural schema which
is applied to a host of situational and traditional oppositions, which
might have had only limited, pre-urban meaning for some groups.
In a Madras conflict between &dquo;right and left handed&dquo; castes in
1707, the weavers and oilmen claimed to be assigned to the two
sides by the English. The British seem also to have to some extent
realized the locality-bound nature of the contrastive components
of the root paradigm, for, in one case, they attempted to import
washermen from Vizagapatnam to replace the ones at Madras who
seemed susceptible to factionalization along the lines of the root
paradigm. Within the cultural scheme of right and left hand castes,
moreover, we get additional evidence of the manipulations of the
indigenous intermediary groups, their attempts to recruit lower
groups into their factions, and the strikes and appearance of unionization that this sometimes gave.
The Madras cases from the eighteenth century also give us some
insight into the nature of the symbols and privileges over which
conflict around the root paradigm seems to arise. These vary, but
seem to fall into three major categories.
Firstly, as in the case of
early Madras, conflicts between the two groups arise over their
respective rights to space, particularly for weddings and funeral processions. Secondly conflicts arise over their respective rights to
privileged symbols and practices, such as the right to wear red
ribbons and build houses with flat roofs, to use the &dquo;tom-tom,
spoon and bell&dquo;, to display flags of various colours and with various emblems.
Thirdly, conflicts arise over their respective rights
to the performance of particular ritual activities in particular
temples. This variation in the symbolic focus of conflicts appears
to be linked to a number of factors, such as the particular composi-
do, however,
63.
A
113-115, seems to present a situation similar to that of early Madras, particularly with
regard to the role of native intermediaries of the English administration.
64. Love, op. cit., Vol. II, p. 29; my discussion of the disputes in 18th century
Madras relies on Love for the facts.
255
tion and
pre-urban affiliations of the groups involved in any partiepisode; the relative stability and identity of particular urban
contexts, which varies with the expansion of cities into previously
rural areas, with new construction, particularly of temples, and of
demolition caused by battle; the relative endurance or ephemerality of particular alignments of groups classified under the root
paradigm; the relative extension of the root paradigm beyond the
cular
core
or
individuals
to encompass
other
groups.
Finally,
256
their distinctive
housebuilding
etc.
op.
cit.,
p.
65,
p.
73,
p.
87; Srinivasachari,
p. 80,
257
V Conclusion
258
259
there