SIMON CHARSLEY

University of Glasgow

Interpreting Untouchability
The Performance of Caste in Andhra Pradesh,
South India

Abstract
The performance of a living caste purana is outlined for the light it throws on its
“Untouchable” owners’ placing in the world of their own caste identity as Madigas. This
is shown as taking the form of a confrontation between the Madiga and the Brahman,
the former seeking to undermine the claimed superiority of the latter. Going beyond this,
it calls on cosmogonic traditions emphasizing the female Shakti and making secondary
and junior the great male gods of contemporary Hinduism. Madigas’ practical impor-
tance to others in relation to leather and to performing, and the relationships resulting,
are emphasized. An embraceable caste identity at odds with sheerly negative conceptions
of “the Untouchable” is constructed. The problematic aspects of Madiga identity are not
ignored; exclusion, Untouchability and poverty are accounted for, but contextualized
within powerfully positive elements. It is suggested that this should not be seen imme-
diately as an answer to outsiders’ questions but, in a Geertzian spirit, as a chance to
analyse a complex and changing story that people tell themselves about themselves.
Keywords: India—Untouchable—purana—caste myth—performance

Asian Folklore Studies, Volume 63, 2004: 267–290

 N THE Indian state of Andhra Pradesh members of a small caste, Chindu

I by name, act out a confrontation between Jambava, the ³rst ancestor of

Madigas, and “Brahman.”* Madigas are one of the two major, formerly
“Untouchable” caste clusters or blocs of linked endogamous groups of this
region. Their main traditional occupations have been focused on leather.
Chindus are not themselves leather workers; they are performers with the
right and duty to perform for village communities of Madigas as part of their
hereditary occupation.
The confrontation of the “Untouchable” and the Brahman, at ³rst sight
the key feature of the performance, displays the centrality and opposition of
these two sections of Indian society which DUMONT (1980) so memorably rep-
resented as at the root of the caste system and which DELIÈGE (1989, 1993) has
found repeatedly juxtaposed in his important investigation of Untouchable
caste myths. Here, however, the phenomenon offered for examination is
signi³cantly different to the caste myths which have been used as a route into
understanding the Untouchable “mode of thought” in addressing the predica-
ment of exclusion, discrimination and exploitation (DELIÈGE 1993: 534,
546–47; See also CHARSLEY 1998: 61–62). A living caste purana here provides
a complex of richly contextualized and related stories and contentions, pre-
sented by those for whom they are directly signi³cant and to informed and for
the most part personally involved audiences. It contrasts, that is to say, with the
evidence hitherto available, episodes at best extracted from such larger narra-
tives, often elicited in response to outsider’s puzzles, or even perhaps casually
encountered in different contexts and periods. Already familiar episodes often
appear, but far more of the essential contextual basis for their interpretation is
delivered with them. The task of interpretation is not simpli³ed; on the con-
trary, the illusion of simplicity that may accompany homogenized notions of
“the Untouchable” as the epitome of deprivation is dispelled. This paper has
a limited aim therefore. It seeks to summarize the contextualizing whole and
draw out some implications for the understanding of Untouchability in gen-
eral from this particular group’s worked-out response to their own situation.
[ 268 ]
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 269

The ways the performance moves beyond—and behind—its apparent starting

point in the confrontation of the Brahman and the Untouchable become clear.
A longer standing agenda of research into the nature and effects of
“Untouchable” as a label lies behind the discussion here (CHARSLEY 1996,
1998; CHARSLEY and KARANTH 1998). I have argued elsewhere that, though
it was grounded in ancient ideas and practices of pollution and exclusion
which are appropriately termed Untouchability, as a label for a major section
of the Indian population as a whole it was constructed only in the twentieth
century. It came then—it is important to note that this was from perspectives
external to the groups concerned—to override in of³cial and then academic
discourses the diverse individual caste identities of those it embraced. Of
these the major kinds were those based on traditions of agricultural labor
and service to farming families, of leather trades, and of “sweeping” or san-
itary work. The scope of the label, and its force, were largely the product of
campaigning, mobilization, and legislating in the twentieth century, estab-
lishing key elements of the legal, political, and social order of postcolonial
India. Hitherto entirely separate caste clusters and individual castes were
consolidated—in aspiration if not in reality—into a single section of society.
The classi³cation has always been motivated by liberal and humanitarian
intentions and has been of immense value for the lives of many. On it was
built the vast apparatus of positive discrimination or af³rmative action
attached to the of³cial designation “Scheduled Caste.” It has derived its
power largely, however, from a characterization that is two-edged. Those so
labelled have been identi³ed as victims of an iniquitous social order, exclu-
sively victims with absolutely nothing of which they could be proud. A prime
signi³cance of this caste purana, and a basic reason for displaying it here, is
in providing a forceful reminder of a previous and here persisting order in
which the sheer negativity of the Untouchable category had yet to take hold,
in which, in varying degree, the castes so reduced were able to distinguish
their identities with pride. What emerges is a worldview which does not seek
to explain away the discrimination under which they labor but contextual-
izes it and focuses on presenting an embraceable identity.1

MADIGAS, CHINDUS, AND THE JAMBA PURANAM

Today’s association of leather-working with chappals, the characteristic
Indian form of leather footwear, often obscures the former importance of
leather items in irrigation, farming, cattle husbandry, transport, and many
manufacturing processes. In a world where technological substitutes had yet
to be developed, leather goods were of immensely greater signi³cance. The
tanning of animal skin to produce leather transforms a messy, smelly, and
rapidly decaying substance into one that can be experienced as sweet and
270 SIMON CHARSLEY

clean and of great utility. It is one of humanity’s major historic technological

transformations, along with the smelting of metal. In India, complex values
related to cattle, the main source of the skin, generated a problem but did not
prevent the development of a major industry carried on in almost every vil-
lage and town across the land. Pro³tability of leather has at times balanced
ideas of religious pollution associated most strongly with its basic production
process and made leather-related castes one of the major sections of the pop-
ulation, the most numerous and best known being the Chamars of North
India (BRIGGS 1920, COHN 1990, KHARE 1985). For the South, the great fam-
ily of Telugu leather-related castes has been analyzed in a small but pioneer-
ing volume by T. R. Singh as one of a number of “constellations” of castes
(SINGH 1969, 31). Madigas, as represented here by Jambava, a ³rst ancestor,
are a large central caste in the constellation and they have other small “satel-
lite castes,” sometimes regarded as sub-castes, attached to them. Apart from
Madigas in villages collecting cattle carcasses, tanning the skins to make
leather, and then producing items of everyday local use from it, other castes
of the constellation specialized in tanning, in high quality manufacturing of
particular articles—with the repair of footwear so familiar in the recent past
as a generally available fallback—and in the trading of skins, leather, and
³nished products. The importance of leather and the secure livelihoods to be
obtained from it meant that even leather workers based in villages might sup-
port satellite castes operating over wider areas, amongst them the Chindus
and other performing castes to be noted.2
The caste myth (kula pur„«a½) here is focused on Jambava, or often
Jambavamuni, the ancestral Madiga, and is generally termed the Jamba
Purana. Such myths and their hereditary tellers are well known by and for
many castes in Andhra Pradesh (SUBBACHARY 2000, 2001; compare SHAH
and SHROFF 1959; DAS 1968). Madigas are distinctive here only in having
several different groups owning and telling their own versions of the caste
myth in their own ways. Apart from the Chindus, on whose version this
paper focuses, there are Nulakachandaiahs, who are caste gurus, keepers of
legal records and genealogies, as well as narrators. There are also Dakkalis,
Baindlas, Asadis, and Masthis, each with their distinctive—but changing—
specialisms. Versions from all these have been recorded but only a Chindu
text, to be discussed here, has been published.3 The Chindus are performers
of yakshaganas (yak¤ag„na½), an ancient genre of musical theatre with sto-
ries drawn from the major epic traditions, Mahabharata and Ramayana, and
the puranas. These have taken oral and, from ancient times, written forms.
Puranas were de³ned as early as the ³fth century CE as dealing with ³ve
themes: genesis, destruction and regeneration, genealogy, cosmic cycles, and
dynastic histories (COBURN 1984, 21). The list remains usefully indicative of
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 271

the content of this Jamba Purana. Yakshagana is one of the ways in which
the puranic tradition lives on, if now more strongly in some regions of India
than in others.4 In the area in question, Chindus are one of two small castes
with performance of this kind as their specialism; there are also numerous
amateur troupes of enthusiastic yakshagana performers, Madigas prominent
amongst them.
This wealth of explicit, complexly developed, and expertly delivered
discourse on caste origins and identity contrasts, as noted above, with the
impression hitherto given by the literature on Untouchable caste myths of
individual stories, often fragments weakly remembered and casually told.
Here they reveal themselves as substantial ideological and artistic resources,
surprising if their “Untouchable” owners and tellers are thought of merely
as devalued objects of deprivation and discrimination.
The Chindu treatment of the Jamba Purana in yakshagana style is gen-
erally known amongst performers and audiences as Gosangi Vesham
(go¤„ªgi v†sa½). Vesham signi³es a made-up and costumed role, here in the
³rst place and most obviously Jambava, the Madiga ³rst ancestor. Its per-
formance involves dialogue, song, and dance with accompanying musicians.
Amongst the components of the Jambava role that will be noted is, howev-
er, the power to pacify the enraged goddess through his dancing. This iden-
ti³able and, as it turns out, separable role within a role is Gosangi. In the
narrative of the performance it is passed on from Jambava to his Chindu son
and hence to the Chindus as a whole. They, as performers by caste, are pro-
viding the player for the Jambava role, as well as for the other roles required,
Brahman, and whatever supporting characters may appear in the perform-
ance known as a whole by this title, Gosangi Vesham. The name Chindu,
which here represents the caste, means also the dance steps which are a key
part of the Gosangi performance. From this association the caste itself may
also be called Gosangis.
Though its style is closely linked to yakshagana, Gosangi Vesham is dis-
tinguished from it in several important ways. Chindus are not summoned to
perform but follow their own programme of visits to a series of villages in
which they have rights (mir„si) of support and performance. They stay for
several days, at ³rst performing overnight as is usual for most village per-
formances. On the last day or possibly two days of their visit they perform
instead in the daytime, ³rst Gosangi Vesham and then Yellamma Vesham.
The central vesham or role in the latter is of the living idol of a goddess with
whom Madigas have a special but not exclusive relationship. Yellamma her-
self, a popular goddess to whom animal sacri³ce is directed, is vividly
described in the Gosangi Vesham but does not appear there. When she does
appear subsequently, she is played by a female Chindu performer and
272 SIMON CHARSLEY

becomes the object of worship (pðj„) in a procession from house to house

around the Madiga area and beyond.5 The two performances together
become a forceful assertion, directed both to Madigas themselves and to
other villagers more generally, of the special signi³cance of Madigas.
Through Yellamma, their importance for the religious interests of many
other villagers is displayed; through the Gosangi Vesham, they assert
Madigas’ cosmic signi³cance, practical importance, learning, and piety in
the face of everything for which the Brahman tradition, in their perception
of it, stands.

THE GOSANGI VESHAM

The content of the living purana has now to be outlined, drawing on two
main sources, both from Nalgonda District of Telangana, the north-eastern
region of Andhra Pradesh. One is a village performance observed, recorded
and videotaped in March 2003, the other a published Telugu text
(VENKATESWARLU 1997).6 In addition, an as-yet unpublished paper and dis-
cussion with its authors have brought in a wider range of experience (REDDY
and HARISCHANDRA 2002).7
Preparations for the village performance took place at the house of the
Madiga leader of a yakshagana troupe in a village of about four thousand
people, one hundred and forty kilometers from the state capital Hyderabad.8
The Jambava himself was the main focus of attention, his costume and
make-up heavily laden with explicit symbolic reference (REDDY and
HARISCHANDRA 2002). Two subsidiary performers, playing members of high
and low status trading castes, Komati and Balija, and the other main pro-
tagonist, Brahman, joined him. Four dappu drummers arrived,9 led by one
of the two elders of the Madiga community. Their ³rst task was to deliver
Komati and Balija to the performance space at a street corner in the neigh-
boring section of the village belonging to the Goud or Toddy Tapper caste.10
Once arrived and already attracting an audience, the two played out a short
comedy of interactions between themselves and a farmer character in every-
day dress. Such supporting caste veshams—performances of stereotypical
caste roles—are usual, but the particular choice was unexpected and unfa-
miliar to most: Washerman and Barber had been expected. Though such
opening episodes are essentially about gathering an audience in preparation
for the arrival of Jambava, they establish the inter-caste context and theme
from the beginning.
Meanwhile the ³nal stages of Jambava’s own preparation and the ritu-
alization of the event were completed at the gate of the house yard. Turmeric
and pink patches representing the smallpox with which Yellamma is associ-
ated were applied to his legs, water poured around his feet, a coconut broken
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 273

and vermillion (kumkum) applied to foreheads of the other participants, sig-

nalling their participation in worship. Decked in leaves of the neem tree
(Azadirachta indica), preceded by the drummers, and with a canopy carried
along behind him, Jambava then advanced out of the colony brandishing his
cutlass and a heavy bamboo stick and dancing ³ercely (raudra). Brahman,
with book, spectacles, an elegant black stick, and his umbrella open against
the sun, walked calmly along in the same procession.11
Themes and Format
The performers distinguished three puranas as making up their Gosangi
Vesham. The ³rst is the Adipurana („dipur„«a), dealing with origins, the
second, the Shaktipurana, dealing with the feminine principle of force or
energy (šakti), the Goddess and village goddesses, and the third, the
Basavapurana,12 dealing with the signi³cance of leather. Performance is con-
tinuous however. The three sections start in the order suggested but inter-
weave as they are developed. Confrontation as a narrative theme is directly
relevant mainly for the last: in the two preceding, Brahmans as such are
signi³cantly absent from the stories told. The Trimurthis—Brahma,
Vishnu, and Shiva—have their place in the Adipurana but it is, as will be
seen, a subordinated one. The form throughout is, however, confrontation
moderated by cooperation: Brahman facilitates Jambava’s explanations from
the beginning, respectfully asking him to explain himself and prompting the
next stage of the narrative. Jambava intersperses his dialogue with songs,
mostly with responses (vanta) from the musicians, and the cindu footwork.
The songs are more or less relevant, more or less readily understood, some-
times subsequently explained. Brahman keeps mostly to dialogue and sup-
porting action. In the observed performance but not in the published text, he
was the more μuent performer, at times helping Jambava out with his own
points and explanations. At the same time he was intermittently provocative.
Jambava responded physically, once or twice chasing Brahman escaping
from the performance site altogether. His response to the most offensive and
outrageous of Jambava’s assertions was to cry exaggeratedly, to the amuse-
ment of the audience. They also tricked each other on occasion, with acted
humor relevant to their role. Brahman offered Jambava his ³ne scarf if he
could catch it; but then threw it bundled into a ball into the air, in the oppo-
site direction. Jambava got Brahman to bend before his seated self, as
described below. He interacted with a young boy in a sketch about wanting
to buy sweets as another family had done; this had been for their grandfa-
ther’s death anniversary. “Grandfather, when will you die?” responded the
boy. He enquired of his fellow Chindus about their performing of lady roles,
and parodied the way it should be done. The performance ended, in typical
FIGURE 1. Jambava at a house in the Madiga colony applies base color for face
using a mirror. Brahman to his left prepares his sacred thread. Harmonium.
Ankle bells and ornaments on the ground.

FIGURE 2. Brahman making up, Komati

behind him, Jambava to his right. Jambava
is almost fully dressed including saris and
neem leaves.

FIGURE 3. Jambava in the course of the per-

formance, with full cowrie shell ornamenta-
tion, garland, bell, bamboo stick and cutlass,
and smallpox (Yellamma-related) ornamenta-
tion of legs.
FIGURE 4. Brahman performing: umbrella,
spectacles, shawl, book, fine black stick.

FIGURE 5. Chindu musicians and singers.

FIGURE 6. Brahman, Jambava,

and Madiga drummers return
down the street after the per-
formance, back to the Madiga
colony.
276 SIMON CHARSLEY

yakshagana style, with ma«ga}am, auspicious praise for the gods, and for the
sponsors of the event performed by Brahman. The performers then returned
as they had come, in procession to the Madiga settlement.
Adipurana
At the beginning of the text13 Jambava claims that he was sent to earth one
night by Brahma, Vishnu, and Ishwara (Shiva). Almost immediately, in a
song he proclaims himself as “born six months before the earth itself.” Then
there is a lotus in the waters and its μowers are producing fruits. Jambava
cuts one open and Adishakti emerges, crowing and dancing like a peacock.
She delivers three eggs. What follows expands and ³lls this out with a sec-
ond and apparently different genesis narrative of the emergence from noth-
ingness of the sound Om with lights μashing colors. Names are acquired
and these names are the original god, Adidevudu („did†vudu). From it/him
the waters formed and the lotus in the waters. And in the lotus, again six
months before the birth of the earth, Jambava was born.
In the main development the precedence of Jambava is therefore boldly
established. He is present at the origin of everything. Though he is going on
to proclaim himself the grandfather (t„ta) of Brahma, Shiva, and Vishnu—
the Trimurthis—he has from the beginning a separate identity distinguish-
ing him from the gods.
Though not carried systematically all through, a claim that Madigas rec-
ognize eighteen ages (yugam), as compared with Brahmans’ four, provides the
frame for ordering this and the following narratives. Their names are recited in
the ³rst example of the naming and listing which occurs throughout. It is one
of the favored forms of learning—or performance skills—displayed repeatedly
by Jambava. It was in the ³rst age that the original god was born. Subsequently
he is most often called Parabrahmasvarupam, which may be glossed as “the
essential self-created spirit,” and he has super-anthropomorphic form: ³ve
faces, ten eyes, ten hands, and all the vedas in him. A lotus and other plants in
the water were created in the second age. Jambava himself emerged from the
water looking like a hairy animal and carrying out austerities (tapas), and this
marked the end of the fourth age. In the ³fth it was the creation of Adishakti
who also, like Parabrahmasvarupam, had ³ve faces and ten hands but also an
extra eye (palanetram) in her forehead, which was capable of burning up what-
ever it blazed on. She had ³ve names, stars on her tongue and mantras in her
mouth, and a jewel granting all wishes (cint„ma«i) in her navel. Again she is
in the lotus but otherwise the story is now different and Jambava and Adishakti
have not met.
In the form of a bee, however, she searched the lotus looking for some-
one else and found Jambava sleeping. Growing up, she tried to persuade
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 277

him to have sex with her but he refused: he was her elder brother, born in
the preceding age. Instead she should undertake austerities for the original
god. By such means she would summon him to ful³l her desire. This was
accomplished and, in the form of peacocks, she and he played in the waters.
He impregnated her with three drops of his tears and she laid three eggs.
From the eggs Brahma, Vishnu, and Shiva were born.14 Providing for them
in the watery world required an island and the building of a nest on it, and
Adishesha, the original serpent, to guard them. From the shells remaining,
the rest of the form, furniture, and living beings of the world were created.
Their resonant names and vast numbers are impressively listed, arriving
³nally at Kamadhenu, the cow that provides unlimited milk—or perhaps
anything that can be wished for; Kalpavriksham, the tree that is similarly
bountiful with fruits; and Palasamudram, the ocean of milk the churning of
which was to yield immortality for the gods. They were to provide suste-
nance for the newly born Trimurthis. At this point the purana of origination
can be regarded as completed, with the Shaktipurana already grounded
within it.
Shaktipurana
Jambava is still the muni, holy sage and adviser of the gods, in this next stage,
but the narrative now goes beyond this to link him with his Madiga and
other descendants and provide charters (MALINOWSKI 1926) for their rela-
tionships, privileges, and pains. The three young gods have grown up and
Adishakti now wants sex from them. Though she claims to be only the
mother of the eggs, hence their grandmother, they regard her as their own
mother. They therefore refuse her and she seeks to kill them. They turn to
Jambava for protection. His advice is to agree to her request but to ask her
to bathe ³rst. She will then remove her dangerous third eye before entering
the water—the gods should seize it and use it to burn its owner herself up.
They do as instructed and she is reduced to ash. From the ash
Parabrahmasvarupam, who now re-enters the story to provide wives for the
young gods, creates a new image of Adishakti. He divides it into ³ve parts.
From the head, chest, and navel he creates Saraswati, Lakshmi, and Parvati,
as wives for Brahma, Vishnu, and Ishwara (Shiva) respectively. From the
fourth part, two wives are created for Jambava, and from the last part Kali,
not as a wife but to help the gods (d†vata) ³ght the anti-gods (asura) and
demons (r„k¤asa). She would do so in a succession of forms and would be
worshipped and receive animal sacri³ce under so many names: forty-three
are listed in this text, most of them the names of village goddesses. The pas-
sage locates the wives of Jambava amongst the goddesses but without being
goddesses themselves. Their clearest signi³cance is that they express the
278 SIMON CHARSLEY

relationship between the Madigas and their Chindu performers: the ³rst
wife is the ancestress of the former, the second wife of the latter.
Having created wives for the Trimurthis, the task then is to get them mar-
ried. This necessitates jewellery, requiring the creation of precious metals for
it. Vishwabrahma and other craftsmen are required to make it, and Jambava
is to provide the leather bellows needed for melting the metal. Leather was
however lacking: this is its ³rst appearance in the narrative and it is noted that
he did not have cows. Having given his word to provide the bellows nonethe-
less, a son, Yugamunindrudu, was born from the right side of his own stom-
ach, to be killed by his own father so that his skin would provide the necessary
material. Once killed and his skin—indeed his bones too—used for necessary
implements, he still came back to life and cursed his father: “You will forget
your austerities and will become a Chandala”—the puranic outcaste. Nobody
would touch him. Eventually the curse was limited to ³ve thousand years, but
Jambava nevertheless cursed him in return: “You will be called Dakkali. You
will beg for food in the houses of my descendants.” The episode is completed
with the successful making of the jewellery and the creation of a Kummari
(potter caste) to make the pots needed for the marriages.
This is a passage of immense signi³cance for Madigas. It establishes the
link with skin or leather, only it is not the skin of cattle but almost Jambava’s
own. It exempli³es the dependence of others on the leather he provides, which
is to be the major theme of the last section of the performance. And it ³xes
another inter-caste relationship, that between Madigas and Dakkalis, another
of their satellites. At its center is the rooting of untouchability, not in working
with leather or even in anything to do with the cow, but in a human tragedy.
Such generous pledging of help as Jambava displays and its potentially tragic
consequences are a well-known spring of action and emotion in the epics and
the rendition of stories from them in yakshagana: Lord Krishna, for instance,
commits himself to avenge the spat-upon Gayodu in the popular
Gayopakyanam. In yakshagana, however, ultimate tragedy is avoided through
the powers of gods: here however, it is the outcome in mutual cursing amongst
humans which is envisaged as having the direst of consequences for them and
their everyday world. Curses are, in the universe of Hindu thought, neither in
themselves wicked nor the historic fantasy that they are in contemporary
Western thought. It is not a mistake, nor wickedness, nor is it portrayed as
entrapment by the gods: it is a tragic consequence of a good man trying in
extraordinary circumstances to do what is both necessary and right.
Basavapurana
It is the following episode that sets up a dangerous relationship between
Jambava and the cow. The story now brings him into relation with the key
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 279

source of the leather, dead cattle, indeed with the skinning of the dead
Kamadhenu. It is framed in a way which avoids any implication that this in
itself is other than a valued skill. From this Madiga perspective, the impend-
ing problem lies elsewhere, and blame will be ³rmly attached to a non-
Madiga. Parvati and her husband were moving in the forest. Parvati injured
herself and blood from the wound became a tree. At its foot a boy, Chennaiah,
was born.15 The implication of his birth is that he is the child of Parvati, and
quasi kinship with Jambava is established too. The gods (d†vata) told
Chennaiah to address Jambava, “born six months before the birth of the
earth,” as grandfather (t„ta). Chennaiah is ³rst taken back with Parvati to
Kailasam, Shiva’s abode, to be the keeper there for Kamadhenu. He, howev-
er, forms a desire to taste her meat, at which she dies of sorrow in front of
Lord Shiva. The gods are not up to the task of removing and cleaning the
body. Chennaiah is therefore sent to fetch Jambava from his own place where
he is sitting, high up and engaged in the worship of lingams, the symbol of
Shiva. Chennaiah means to call him: “Grandfather, come down!” but instead
mispronounces the intention, saying instead “Grandfather Madiga come!”
The portents are bad: the disturbance has caused him to lose his
lingams16 and to be called “Madiga.” He descends, interpreting events as the
working of the curse of his Dakkali son. He opens his treasures and hands
on to his Chindu son, Jihmamuni, a bell which Adishakti had given him. It
is a conspicuous ornament worn by the Chindu playing Jambava in Gosangi
Vesham. His Chindu son is to come twice a year to the houses of the
Madigas to receive his dues. Jambava and Chennaiah return to Kailasam
where he and the gods cut up the carcass. Chennaiah is to cook it for eating,
but in the course of the cooking a piece falls from the pot. He picks it up,
blows on it to clean it and pops it back. This is the offence: he has polluted
it with his breath and the gods reject it: “It is for Jambava and his grandson.”
By implication, the gods are ceasing to eat polluted beef; it is to be for those
cursed with untouchability only, but the restriction is linked to the polluting
of food, not to the sacredness of the living cow or danger from the dead one.
Pollution as a phenomenon is not denied, but any understanding of the cow
in its death as intrinsically dangerous is turned aside. Disaster hangs over the
whole sequence of events and is expressed in a tirade of blame aimed at the
unfortunate Chennaiah. Jambava proclaims him a Mala, the progenitor of
the other main “Untouchable” caste of the region.
The impending end of the last age before the present age of ultimate
degeneration, Kaliyugam, is signalled. The gods will withdraw, leaving him,
says Jambava, helpless in the bad place to which they had summoned him.
He begs from them the boon of the provision for his Madigas of paddy and
other foods. He lists thirteen kinds of residue from harvesting and threshing,
280 SIMON CHARSLEY

as well as other items that people are to give them annually. Thus the ³nal
element in the relationship of untouchability for the present age is set up. In
it Madigas will continue to deal with the successors to Kamadhenu as they
die;17 they will continue to have the beef to eat but will be regarded as pol-
luted for doing so; and they will be provided with paddy and other means to
life by those growing it. They will have a livelihood but its cost in pollution
and status loss is not disguised.
Performance and the Goddess
Next we are taken into the world of war between the gods and the anti-gods
and demons. Shankara (Shiva) is killing them but with the resulting problem
that from every drop shed of their blood another opponent springs up.
Adishakti reappears to provide a solution: she prevents the blood falling to
earth by catching it on her tongue and drinking it. However, this blood-
drinking enrages her and she turns on the gods themselves. Jambava is
recruited to pacify her. To carry out this great service he is presented with
thirty-two badges of distinction or “honors” (birudamu) by the gods. Vishnu
creates for him the distinctive Madiga drum (dappu); Shiva presents a tiger
skin; Virabhadra the ankle bells (gajjelu) that all performers who dance wear;
Shanmuka peacock feathers; and so on. These are more or less closely con-
nected with performing and were displayed in the performance observed.18
Jambava, supported by his sons and his wives, paci³es the fury of Adishakti
by his Gosangi Vesham and returns to his home, Jambalagiri.
Only now does Goddess Yellamma enter the narrative for the ³rst time,
in her Brahmanical form as Renuka, wife of the Brahman sage Jamadagni.
She is destined to be killed at her husband’s behest by her son Parashurama.
The focus is now on Chindus rather than Madigas. Renuka/Yellamma is
represented as μeeing from her son to Jambalagiri and seeking Jambava’s
protection. Parashurama, however, comes to the town, manages to ³nd his
mother and kills her. She becomes a demon and chases the people to kill
them. Jambava therefore needs to protect his people from her but is unable
to do it himself. He calls ³rst on his guru, Rudramahamuni. He could not
help, but his mention emphasizes yet again Jambava’s standing and piety.
Likewise his ³rst wife’s Madiga sons and his Dakkali son declared them-
selves unable to help. The Dakkali parried the request: his father having
made him Untouchable, how could he possibly have that strength? Finally
Chindu Jihmamuni, the son of his second wife, was asked. He accepted the
task but on condition that all his father’s thirty-two honors received from the
gods should be passed on to him. This was agreed and, so equipped,
Jihmamuni and his wife Sridevasalani danced before Renuka and ³nally she
was appeased. Yellamma—as she is now named in the text—blessed the
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 281

couple and gave Jihmamuni seven gifts of her own ornaments. His wife was
to perform Yellamma Vesham and they were to take it round people’s hous-
es every year. In return, they would receive offerings from the people.
The text appears then to revert to the perhaps older story which pre-
ceded this Renuka episode. Chindu Jihmamuni is again being given per-
formance instructions but now by Parabrahmaswarupa and the other gods.
Yellamma is not referred to again until the very end of the performance.
Taking up the earlier story, the Chindus are to perform for Jambava’s people
and to take round with them the honors Jambava had handed on to them.
There should be no marriage between Madigas and Chindus. The Dakkali
son receives instructions too: he is to beg in the houses of Madigas and to
live to the east of them. Five Jetti castes, attached to Madigas, Malas, Toddy
Tappers, Weavers, and Washermen—apparently descended from those to
whom the falling drops of blood mentioned above gave rise—are also called
for instruction. They are to make the people happy with their performances,
except perhaps for the Madiga Jettis who are to be guards.19
Learning, Caste Schemes, and the Challenge to Brahmanism
The Shakti purana theme has been completed and the stories of originations
(Adipurana) are about to be with one short section grounding links with pas-
toral Gollas (VENKATESWARLU 1997, 20).20 Jambava’s claims to superior
knowledge of caste and genealogy come to the fore. Only the Basavapurana
theme remains to be ³nally developed. Cattle and leather are emphasized as
the source of Madigas’ importance for others, simultaneously as the source
of problems for themselves. The latter is identi³ed as essentially
Brahmanical and is challenged. This is the theme that comes to dominate
the last part of the performance.
The ³rst major demonstration of superior knowledge here is the listing
of castes in three separate schemes. These start with strong quasi-Brahmanic
reference but move into the immediately experienced caste universe of the
region, before taking off ³nally, after immense displays of virtuoso listing,
into an imaginative assertion of Madiga superiority. The beginning is a
scheme which Jambava represents as set out by Vyasa, the classic “arranger”
of Brahmanical lore for the present Kaliyugam. The three “Twice-born” var-
nas of the well-known Brahmanical scheme of four, or ³ve, lead the listing,
but it then continues with eighteen more. They are named with Brahmanic
terms ³rst, each then equated with a contemporary and locally familiar
name: for example Lingadharaka (those wearing lingams) are Jangams,
Gorakshaka (cow protectors) are Gollas, Nartaka (dancers) are Bhogams,
Matanga are Madigas, and Chandala are Malas. The orthodox fourth varna
term, “Shudra,” does not appear; Madigas and Malas are not placed at the
282 SIMON CHARSLEY

end of the list; and there is no division for “Untouchables.” The second list-
ing gives striking expression to the caste-constellation concept (SINGH
1969). Seventeen major divisions are shown as made up of one hundred and
forty-two sub-groups and, in one case, a sub-sub-group. Brahmans are still
put ³rst, with ten kinds distinguished by their specialized activities: per-
forming rituals, working as village heads and accountants, narrating vedas,
and so on. Some of the names at each level correspond with the ³rst scheme,
but the two lists are very partially consistent. The third scheme leaves
Brahmanic categorizations entirely behind, setting out minor or satellite
castes which thirty-three major castes have attached to them. Brahmans are
now mentioned in third place, with Vipravinodulu, a known performing
group, attached to them. All but two are shown with one satellite only.
Gollas are shown with three, and when Jambava arrives at his own Madigas
he sets off on a fanciful stream of dependants and the dependants of depen-
dants. It starts with the by now familiar Dakkali and continues for ten levels
beneath them. As a μight of listing fancy and a μourish of greatness in the
published text, it is magni³cent; in the observed performance, however,
none of this was even attempted.
The theme of caste knowledge and the relationships of Madigas and
other castes does not end here. Two further strands are woven together, one
concerning puranic relationships of origins and genealogy—another
Brahmanical sphere of knowledge to be contested—the other representing
the importance of Madigas’ leather products for people of all castes.
Relationships with Gollas lead the ³rst strand: the idea that Lord Krishna
was a Golla rather than—the usual Brahmanic version—only brought up by
pastoralists (O’FLAHERTY 1975, 204–13) leads into genealogical issues. The
Madiga version has Jambava giving his daughter, Jambavathi, to Krishna,
making him the Madigas’ son-in-law. The son of Jambavathi and Krishna,
Sambudu, then married Lakshana, daughter of Dhuryodhana, the Kaurava
king in the Mahabharata.21 The extended and contentious debate refers to
inter-caste marriages, mixed descent, and anomalous births. Jambava uses it
to ridicule Brahman ideas of proper descent and any radical separation
between Brahmans and Madigas.

Don’t say that you are superior and Madigas are inferiors. Oh fool! Do
you know who was superior and who was inferior in the past? Don’t
talk without knowing the past! Oh Brahman, you are the son of a don-
key. Brahmans are the slaves of other castes…. To whom was Vyasa
born? Was he born to a Madiga or to a king?… Tell me who are bas-
tards, Brahmans or Madigas? (VENKATESWARLU 1997, 28).
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 283

Brahman bounces back with protests and counter-assertions. The text

runs on in sharp banter and mutual abuse. Sexual morals and practices—
whether it is Brahmans or Madigas who were born to prostitutes (lanja);
who practise abortion to hide sexual irregularities; who refuse remarriage to
child widows—these and more are brought into vigorous contention.
Madigas’ identi³cation with leather is the last strand running through
the ³nal section of the performance. The dependence on them of other
castes as leather users is argued, with song and acting, culminating in a ³nal
mocking of Brahmanical notions of leather as impure. Even Reddy land-
lords, Jambava asserts, though they will not touch Madigas, use the footwear
made by them. Others depend on their leather for their traditional callings:
Gouds use a leather sling made by Madigas to enable them to climb their
toddy trees; Bhogams as dancers use it as skins for their drums; Washermen
(s„kali) as blinkers for their donkeys. Even Brahmans, when they draw water
for their daily bath, must come into contact with leather. But not nowadays,
responds Brahman: today we press the button on an electric motor. Ah,
Jambava answers, but the belt on the motor is prepared by Madigas.
“Everything in the world can only work with the help of Madigas”
(VENKATESWARLU 1997, 34). Mingled with this are direct challenges to
Untouchability. Brahman: “Hey, you son of a dog, don’t touch me!”
Jambava: “Hey, I touched you—you are impure!” He rebukes Brahman for
calling him “Untouchable” and educates him in the work of Sri
Virabrahman, a popular seventeenth-century saint who denounced ideas of
pollution and Untouchability.22
The published text ends with the acting out of a scene between Jambava
as a contemporary village Madiga and Brahman as a would-be customer for
his chappal repair services. Brahman asked him to mend his chappal;
Jambava denied that Madigas do such work any more. Brahman tried to
remind him that he wanted him to ³x the propitious time for his son’s mar-
riage.23 Jambava denied it: they are knowledgeable too and can ³x it for
themselves. Brahman appealed to his good nature and kicked the chappal to
be mended over to him. In the performance this was a symbolic chappal, a
stick with a thread tied round it: his own potentially polluting leather chap-
pals remained ³rmly on his feet. Though the represented insult was supposed
to anger Jambava, it also cued a word-play song proclaiming the central
importance of footwear for life and for Goddess Yellamma herself. He then
distracted Brahman with talk of Madiga marriage customs and his son’s
wedding before asking him to try the repaired chappal. This necessitated his
bending, only to be teased for bending before a Madiga. “Yes, yes! That is
how you pay the respect that everyone offers to Madigas. Bend down again
and repair your own footwear” (VENKATESWARLU 1997, 35). This was too
284 SIMON CHARSLEY

much: Brahman announced that he was going to Kashi, otherwise Varanasi,

the holy city of Hindu pilgrimage and death on the Ganges (PARRY 1994).
He enacted his going, ³nally overtly complicit in the undermining of
Brahman ideas, values, and even dignity. He comes to a river; his chappals
will be soaked; the water is rising; he will carry them in his hands; the water
has reached to his shoulders, to his head. So ³nally Brahman is carrying his
leather footwear on his head. He demonstrates where the real priority lies,
not in issues of purity and pollution but in keeping one’s chappals dry. In the
performance, though the action was over with this ³nal discom³ture of
Brahman, it was he who led the ³nal singing of praise to the gods and gave
a concluding speech. In the past they used to perform Jamba Purana for the
Madigas, he said. In return they would receive donations. But now these
people from Hyderabad had come to watch their performance. God had sent
these people from the city to help them. All Gods are the same: Adam, Eve,
Allah, Rama, Krishna, all are the same. People from different places worship
the God with different names. Finally, everyone set off back whence they
had come, to the Madiga settlement for the working out of the “help” which
was to be provided.

CONCLUSION
This paper has outlined the performance of a living caste purana and the
light it throws on its “Untouchable” owners placing of themselves in the
world. It lives in the sense of being still performed, if probably less frequently
than in the past, and in a dynamic adjusting of its legacy of narrative, song,
and issues to current circumstances. The account here has shown how the
performance is framed as a confrontation between the “Untouchable” and
the Brahman, seeking to undermine the perceived claims to superiority of
the latter in learning, in descent, and in purity. It challenges notions of puri-
ty and pollution in terms of which the owners of the purana as traditional
leatherworkers, associating these ideas particularly with Brahmans, know
themselves to be devalued. Its claim, however, is deeper and more radical.
What does not appear is any assertion of primeval kinship with Brahmans,
so common a theme in previous discussion of “Untouchable myths,” nor
with the gods. It is precedence rather than coevality which is claimed, call-
ing on cosmogonic traditions emphasizing the female Shakti and making
secondary and junior the great male gods of contemporary Hinduism. This,
it should be noted, is not claimed on behalf of “Untouchables” as a whole
but for the particular caste. The other major and rival “Untouchables” caste
or constellation of the region, the Malas, are provided in the purana with an
altogether different ancestry in the person of a quasi-son of Goddess Parvati,
with none of the primeval quality asserted for Madigas.
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 285

If precedence is the fundamental theme here, and the challenging of

Brahman hegemony a contemporary relevance with strong local roots, the
practical importance of Madigas for others is the second major element. It
has two parts to it, leather and performance. Dependence of others on
Madigas as the source of leather goods is established cosmologically in the
provision, as so great a cost to themselves, of bellows for furnaces, making
metal-working possible, and in relationships with Visvakarmas in recogni-
tion of this. The relationship with pastoral Gollas, at least in the region in
question, is the major practical though less immediately telling example
claimed. Relationships with Toddy Tappers and a mass of other secondary
claims mentioned are becoming ever more tenuous with the decline in prac-
tical dependence on leather in recent times. This appears to be the point of
greatest vulnerability in the puranic tradition here: as leather becomes
increasingly residual in the historical experience even of Madigas them-
selves, its signi³cance as a mythological concern is likely to be fading. A
third theme is then performance signi³ed in the Gosangi role not merely as
entertainment, though there is certainly an element of this, but in relation to
the goddess, her pacifying, and worship. The purana, as has been seen, rep-
resents the power of Gosangi performance as having been passed on from
the ancestor of the Madiga cluster as a whole to one branch, the Chindus,
but as regards performing more widely, Madigas have remained active in
their distinctive style of drumming and, since the mid-twentieth century,
often in yakshagana performance. More recently they have also been active in
contemporary performing arts as promoted by the cultural wings of parties of
the left. Their identity as performing people therefore remains signi³cant
and, like the purana itself, alive to current change.
It is these aspects together which make the embraceable caste identity at
odds with sheerly negative conceptions of “the Untouchable,” asserting
instead a multi-stranded tradition of which people can be proud. The prob-
lematic aspects of Madiga identity are not ignored, but there is no reason to see
them as, for Madigas, what the purana is primarily about. Exclusion,
Untouchability, and poverty are accounted for, but contextualized within pow-
erfully positive elements. The ability to move and skin dead cattle is given a
forcefully positive evaluation. The eating of beef is not represented as a mis-
take. Its con³nement to certain castes is the consequence of a change of heart
by the gods which at once secured a source of nourishment but also set apart
those bene³ting from it. A mistake had been made. It was not, however, to eat
beef but to pollute it for others by blowing on a piece thrown back into the pot.
Even this somewhat tangential mistake had been made not by a Madiga but
by the progenitor of their rivals, the Malas.
It was in any case not this, nor the gods, which caused Madigas downfall.
286 SIMON CHARSLEY

Jambava is caught up in an epic tragedy. It is the tragedy of his Dakkali son,

born to be killed by his father to provide the leather needed for the bellows,
cursing him in return, and the curse working out in exclusion through the
intervention of Parvati’s Mala “son.” Analogies with the fate of the Pandavas
at the root of the Mahabharata epic and of Rama in Ramayana, both well
known here through yakshagana and other performance genres, come to
mind. Most particularly it resonates with the fate of the ever-faithful King
Harischandra, condemned to serve an “Untouchable” master in the burial
ground and the subject of one of the most popular and moving dramas of all
the many performed today in Andhra Pradesh.24
For Madigas and Dakkalis the cursing was, it will be recalled, mutual,
resulting in the Untouchability of Madigas for nearly all others, and the
Untouchability of their Dakkali “sons” even for themselves. What seems
therefore a neat package of pollution, infringement, and exclusion from a
Brahmanical perspective is presented in the purana in an altogether differ-
ent light. It can be said to be deconstructed as long as this is not understood
as conceding a priority to the orthodox upper caste perspective which it
allows us to see beyond. To see it as an answer to outsiders’ questions, such
as how Indian “Untouchables” could have been so generally passive in the
face of the discrimination and exploitation imposed upon them (MOORE
1978; FREEMAN 1986), or whether they accept the values in terms of which
they are themselves devalued (MOFFATT 1979; DELIÈGE 1989, 1993; MOSSE
1994), is to miss the opportunity here offered to understand a perspective
which is different but, as so often in anthropology, not in quite the ways
expected. It is, in a Geertzian spirit, a chance to analyse a complex and
changing story that people tell themselves about themselves. A start to this
major task has been attempted here.

NOTES

*The research reported in this paper was carried out as part of an Economic & Social
Research Council project on “Popular Performance and Social Change in Indian Society”
(R000239063). Their ³nancial support and practical assistance are gratefully acknowledged,
as also the support of the University of Hyderabad at which it was based, and in particular its
Department of Sociology.
1. ILAIAH’s Why I am not a Hindu (1996) drew widespread attention to the possibility of
non-Brahman perspectives attributing positive value to the achievements and traditions of
castes normally regarded as “low.” At the same time, the Madiga Dandora movement was
proclaiming the caste’s identity as never before in Andhra Pradesh and mobilizing impres-
sively in pursuit of a caste-wise revision of government provision for Scheduled Castes
(BALAGOPAL 2000). These events were signi³cant in shaping the research on which the pres-
ent study is based.
2. This is a different and non-Brahmanical conception of the phenomenon on which
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 287

MOFFATT (1979) was subsequently to focus his debate-inducing notion of “replication.” See
also DELIÈGE 1992.
3. Professor J. Tirumala Rao of Telugu University has a major Dakkali text already print-
ed but not so far circulated or published.
4. Yakshagana μourished in court circles in the late-sixteenth and seventeenth centuries
and subsequently developed rather differently in the Telugu, Tamil, and Karnataka regions
(SATYANARAYANA 1983, 151–52, 439–41; RAO and SHULMAN 1991; BAPAT 1998).
5. Others are frequently possessed in her presence, but she herself is not. It is therefore
more appropriate to regard her as performing the goddess role than as embodying her in any
more radical sense.
6. This was derived from a performance set up for recording in 1994. The performers
were selected members of local troupes and the version offered appears to have been thought
out in terms of the perceived audience and purpose. The original translations from the
Telugu text used here were made by I. Narasaiah. He co-authored the preliminary reading
on which the present paper draws freely.
7. Published references to various versions of the set of stories which ³nd their place here
were once quite numerous: see for example OPPERT 1893, 464–74; RAUSCHENBUSCH-
CLOUGH 1899, 13–16; THURSTON and RANGACHARI 1909, 315–16, citing Mysore Census
1891; SIRAJ UL HASSAN 1920, 410–11; WHITEHEAD 1921, 127–34.
8. Name and exact location are withheld from general publication.
9. Dappu, a μat, single-skinned drum played with two unequal sticks, is the instrument
primarily identi³ed with Madigas. It is important both in their own cultural activity and in
the provision of services for others.
10. Those historically associated with tapping toddy palms and selling the fermented sap.
Here they have not been regarded as Untouchable and today are members of a large, inμuen-
tial, and, in part, wealthy caste.
11. It has sometimes been thought that the performance would be a private affair for
Madigas, carried on in their own colony. While it is likely that this has happened on occasion,
it is not what the present research found. It is more possible to imagine performance as an
open and licensed “ritual of rebellion” (GLUCKMAN 1954) in contexts of normal repression.
12. Somanatha (1160–1240 CE) wrote a Basavapurana, one of the early Telugu classics,
focused on the life of the Basaveshwara, the founder of the Virasaiva movement. The name
Basava is also that of a bull sacred to Lord Shiva. In its early stages the movement challenged
concepts of pollution and caste discrimination, and an historical link is possible despite the
absence of reference to Basava in the Jamba Purana as recorded here (BROWN 1852, 606;
ISHWARAN 1992, 26–30; SATYANARAYANA 1999, 340).
13. The summary that follows is based mainly on the published text of a three-hour
performance (VENKATESWARLU 1997; CHARSLEY and NARASAIAH 2004).
14. MCCORMACK (1959, 123–24) reports a Virasaiva Devipurana from Karnataka which
includes a similar narrative. Eggs, with varied contents and uses for their shells, are a major
cosmogonic theme.
15. The observed performance offered a different mechanism, involving the dung of
Kamadhenu and Parvati’s shadow, but the effect was the same, a quasi son for Parvati.
16. The worship of personal lingams, the symbol of Shiva, points to a link with
Virasaivism. In a later re-capping of this incident, one particular lingam protecting his life is
distinguished. This also fell down, but he appointed another son, Sangaiah, the progenitor of
the Nulakachandaiah caste gurus, to protect it (VENKATESWARLU 1997, 29).
17. Madigas do not scavenge other dead creatures, an altogether different activity.
288 SIMON CHARSLEY

18. Make-up, costuming, and accoutrements are analyzed in detail by REDDY and
HARISCHANDRA (2002).
19. These would be examples of satellite castes: Gouda Jettis are indeed today the tellers of
the Toddy Tappers’ caste myth in the region, but other Jettis are not currently known.
20. Relationships with Gollas—often keen to be known nowadays as Yadavas—are found
in a variety of contexts. For the village performance, a major and essential item of costume,
the gajjela lagu, shorts with bells attached, were borrowed from Yadavas locally.
21. This is the subject of Lakshana Parinayam, a yakshagana which is a favorite of
Madigas.
22. Sri Virabrahman, popularly known as Brahmamgaru, and his doctrines live on as a
center for pilgrimage and one of the most popular subjects for yakshaganas and other dra-
matic genres in Andhra Pradesh.
23. That Madigas have the service of Brahmans in this respect is taken for granted.
24. See BRIGGS (1953, 44–49) and PARRY (1994, 25–26).

REFERENCES CITED

BALAGOPAL, K.
2000 A tangled web. Subdivision of SC reservation in Andhra Pradesh. Economic &
Political Weekly, March 25: 1075–81.
BAPAT, Gururao V.
1998 Yak¤agaana: Performance and Meaning—A Semiotic Study. Udupi, Karnataka:
Regional Resources Centre for Folk Performing Arts.
BRIGGS, G.W.
1920 The Chamars. Calcutta: Association Press.
1953 The Doms and their Near Relations. Mysore: Wesley Press.
BROWN, Charles Philip
1852 A Dictionary. Telugu and English. Madras: S. P. C. K. Press. (Numerous reprints
including Hyderabad, Potti Sreeramulu Telugu University, 1999).
CHARSLEY, Simon R.
1996 “Untouchable”: What is in a name? Journal of the Royal Anthropological Institute
(n. s.) 2: 1–23.
1998 Caste, cultural resources and social mobility. In Challenging Untouchability: Dalit
Initiative and Experience from Karnataka, ed. Charsley, Simon R., and G. K.
Karanth, 44–71. New Delhi: Sage.
CHARSLEY, Simon R., and G. K. KARANTH
1998 Dalits and state action: The “SCs.” In Challenging Untouchability: Dalit Initiative
and Experience from Karnataka, ed. Charsley, Simon R., and G. K. Karanth, 19–43.
New Delhi: Sage.
CHARSLEY, Simon R., and I. NARASAIAH
2004 Jamba Puranam: A preliminary reading of a published “caste myth.” http://www
.gla.ac.uk/departments/sociology/units/anthrop/dalit/jamba–puranam.ht,
accessed 16 October 2004.
COBURN, Thomas B.
1984 D†vi-m„h„tmya: The Crystallization of the Goddess Tradition. Delhi: Motilal
Banarsidas.
COHN, Bernard S.
1990 An Anthropologist Among the Historians and Other Essays. Delhi: Oxford University
Press.
 UNTOUCHABILITY AND PERFORMANCE OF CASTE 289

DAS, Veena
1968 A sociological approach to the caste puranas: A case study. Sociological Bulletin 1:
141–64
DELIÈGE, Robert
1989 Les mythes d’origine chez les Paraiyar. L’Homme 109: 107–116.
1992 Replication and consensus: Untouchability, caste and ideology in India. Man (n.s.)
27: 155–73.
1993 The myths of origin of the Indian Untouchables. Man (n.s.) 28: 533–49
DUMONT, Louis
1980 Homo Hierarchicus: The Caste System and its Implications. 2nd ed. Trans. Sainsbury,
M., L. Dumont, and B. Gulati. Chicago: University of Chicago Press.
FREEMAN, James M.
1986 The consciousness of freedom among India’s untouchables. In Social and
Economic Development in India, ed. Basu, D. K., and R. Sisson. 153–71. New
Delhi: Sage.
GLUCKMAN, Max
1954 Rituals of Rebellion in South-East Africa. Manchester: Manchester University Press.
ILAIAH, Kancha
1996 Why I am not a Hindu. Calcutta: Samya (Bhatkal & Sen).
ISHWARAN, K.
1992 Speaking of Basava. Lingayat Religion and Culture in South Asia. Boulder: Westview
Press.
KHARE, R.S.
1985 The Untouchable as Himself: Identity and Pragmatism among the Lucknow Chamars.
Cambridge: Cambridge University Press.
MCCORMACK, William
1959 The forms of communication in V‡rašaiva religion. In Traditional India: Structure
and Change, ed. Singer, Milton, 119–29. Austin: University of Texas Press.
MALINOWSKI, Bronislaw
1926 Myth in Primitive Psychology. London: Kegan Paul.
MOFFATT, Michael
1979 An Untouchable Community in South India: Structure and Consensus. Princeton:
Princeton University Press.
MOORE, Barrington
1978 Injustice: The Social Bases of Obedience and Revolt. New York: M. E. Sharpe.
MOSSE, David
1994 Idioms of subordination and styles of protest among Christian and Hindu Harijan
castes in Tamil Nadu. Contributions to Indian Sociology (n. s.) 28: 67–106.
O’FLAHERTY, Wendy Doniger
1975 Hindu Myths: A Sourcebook Translated from the Sanskrit. Harmondsworth: Penguin
Books.
OPPERT, Gustav
1893 The Original Inhabitants of Bharatavar¤a or India. Westminster: Constable.
PARRY, Jonathan P.
1994 Death in Banaras. Cambridge: Cambridge University Press.
RAO, Velcheru Narayana, and David SHULMAN
1991 The powers of parody in N„yaka-period Tanjavur. In Gender, Genre, and Power in
South Asian Expressive Traditions, ed. Appadurai, Arjun, et al., 428–64.
Philadelphia: University of Pennsylvania Press.
290 SIMON CHARSLEY

RAUSCHENBUSCH-CLOUGH, E.
1899 While Sewing Sandals: Or Tales of a Telugu Pariah Tribe. London: Hodder &
Stoughton.
REDDY, Y. A. Sudhakar, and R. R. HARISCHANDRA
2002 Story, performance and metaphors: Explorations in the art form of Cindu
Madigas. Unpublished paper for National Seminar, “Performers and their Arts:
Changing Lives, Changing Forms,” Department of Sociology, University of
Hyderabad 14–16 March 2002.
SATYANARAYANA, K.
1983 A Study of the History and Culture of the Andhras. Vol. 2. New Delhi: People’s
Publishing House.
1999 A Study of the History and Culture of the Andhras. Vol. 1. Hyderabad: Visalaandra
Publishing House.
SHAH, A. M., and R. G. SHROFF
1959 The Vah‡vanc„ B„ro¦s of Gujarat: A caste of genealogists and mythographers. In
Traditional India: Structure and Change, ed. Singer, Milton, 40–70. Austin:
University of Texas Press.
SINGH, T. R.
1969 The Madiga: A Study of Social Structure and Change. Lucknow: Ethnography and
Folk Culture Society.
SIRAJ UL HASSAN, Syed
1920 The Castes and Tribes of H.E.H. the Nizam’s Dominions. Vol. 1, Madigas. 409–20.
Bombay: The Times Press.
SUBBACHARY, Pulikonda
2000 A Study of Caste Myths and Dependent Caste System of Telugus (telugulõ
kulapðr„«„lu „¤rita vyavastha). Hyderabad: Prajashakti Book House.
2001 The tradition of parallel puranas in Andhra Pradesh. In Dravidian Folk and Tribal
Lore, ed. Reddy, B. B. Ramakrishna, 210–17. Kuppam: Dravidian University.
THURSTON, Edgar, and K. RANGACHARI
1909 Castes and Tribes of Southern India. 9 vols. Madras: Government Press.
VENKATESWARLU, Bittu
1997 J„mba pur„nam (M„diga kula s„mskritika chitram). Hyderabad: Janapada Kala
Parishodhana Vedika.
WHITEHEAD, Henry
1921 The Village Gods of South India. 2nd ed. Calcutta: Association Press.