University o f London

The London School of Economics and Political Science

Department o f Anthropology

Witchcraft, Religious Transformation,

and Hindu Nationalism in Rural Central India

Amit A. Desai

Thesis submitted for the degree o f Doctor o f Philosophy

2007
 UMI Number: U615660

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

Dissertation Publishing

UMI U615660
Published by ProQuest LLC 2014. Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
 Abstract

This thesis is an anthropological exploration of the connections between

witchcraft, religious transformation, and Hindu nationalism in a village in an Adivasi (or
‘tribal’) area of eastern Maharashtra, India. It argues that the appeal of Hindu nationalism
in India today cannot be understood without reference to processes of religious and social
transformation that are also taking place at the local level.

The thesis demonstrates how changing village composition in terms of caste, together
with an increased State presence and particular view of modernity, have led to difficulties
in satisfactorily curing attacks of witchcraft and magic. Consequently, many people in the
village and wider area have begun to look for lasting solutions to these problems in new
ways. A significant number have joined a Hindu religious sect, the Mahanubhav Panth,
seen as particularly efficacious in matters of healing. Membership of this sect however
alters the values and practices of adherents which not only causes conflict with non-sect
neighbours and kin but also resonates powerfully with the messages promoted by Hindu
nationalist agents in the area.

The thesis engages key areas of anthropological concern: the relationship between
individual action and social structure; kinship and sociality; State activity; and religious
conversion.
 Acknowledgements

Thanks are due to many people in many places. First and foremost are those owed
to the people of Markakasa who found a stranger in their village, welcomed him, and
transformed him into a friend. Instances of their warmth, curiosity, and generosity are
abiding memories of my time there. I thank in particular Brijlal, Champabai, Jankibai and
Bar Singh, Hemu, Mohan, and Mahananda who gave me a home and made it easy for me
to become part of their family. I thank my Mahaprasad, Ganeshyam for his friendship,
and the rest of the Koreti household for their kindness. There are many others in
Markakasa whose companionship and affection requires acknowledgment but I hope they
will forgive me for not mentioning them individually. I would also like to thank Virendra
Anjankar for introducing me to Markakasa, and Shravan-y/ Taram for hosting me in
Deori.

Fieldwork is an arduous business and every anthropologist needs fairy godparents. While
Brijlal and his family fulfilled that function in Markakasa, my uncle and aunt, Avinash
and Nandini Sahasrabudhe, and my cousin Ashwin supported me in Nagpur. Without
their encouragement, enthusiasm, assistance, and love, it would have been nigh
impossible to complete my research. Their home provided a relief from the intensity of
the experience and allowed me to keep a sense of perspective. On several occasions when
doubts about my work set in they were on hand to encourage or distract. Also in Nagpur,
I would like to thank the Arkatkar family, Harshwardhan Nimkhedkar, and Ramdas,
Rohit and the other members of the Dixit clan.

The research presented here was made possible by the financial support of several bodies.
I was fortunate to receive a postgraduate studentship from the Economic and Social
Research Council (no. PTA-030-2002-00731). I would also like to thank the Royal
Anthropological Institute of Great Britain and Ireland for awarding me an Emslie
Homiman studentship and for giving me an unexpected but very welcome grant from the
Sutasoma Trust. I thank the Central Research Fund of the University of London for
giving me a travel award and lending me a camera. The Kale Memorial Trust also gave
generous assistance for which I am grateful.

I could not have hoped for better supervisors than Jonathan Parry and Veronique Benei.
They have inspired and challenged me in equal measure and I thank them for tolerating
my laziness and occasional obstinacy. I was fortunate that they both visited me in the
field and offered encouragement, ideas and new perspectives. I also thank Laura Bear,
Michael Scott, and Norbert Peabody for leading the writing-up seminar and for their
comments on versions of some of the chapters. In addition to my supervisors, I also
benefited from discussions with Chris Fuller, Peggy Froerer, and Laura Bear in the early
stages of planning the research.

My peers and friends at the LSE have been of tremendous support and a constant source
of intellectual excitement. I thank Michelle Obeid, Girish Daswani, Casey High, Jason
Sumich, Florent Giehmann, Eve Zucker, Clafinda Still, Evan Killick, and Magnus Course
for generously reading and commenting on parts of my work. I am grateful to Maxim
Bolt for expertly and speedily proof-reading the thesis for me at a busy time for him
personally.

My conversion to anthropology would have been impossible without the financial and
emotional support of my family. I thank my parents, Ashok and Yashaswini, and my
sister Aarti for their constant love and encouragement while I was writing the thesis and
for respecting my desire to do it.
 Table o f Contents

Abstract page ii
Acknowledgements iii
List of Figures viii
List of Plates ix
Map of India xi
Locator map of India showing Maharashtra and Gondia district xii
Aerial view of Markakasa xiii

Chapter One: Introduction 1

Orientations 5
Religious experience and nationalist identification 5
Poles apart? The idea of the Adivasi and Indian modernity 8
Discussing conversion 14
Experience and embodiment as method 17
The experience of fieldwork 19
Outline of chapters 24

Chapter Two: The Setting 30

Space, caste and religious affiliation 35
Gaita Gond 38
Mahar 39
Teli 41
Commensality and Untouchability 42
Livelihoods 47
The land 47
The forest 50
State-funded casual work 53

v
Chapter Three: Ritual Friendship 56
Ritual friendship in Markakasa 59
Three friends 65
The content of ritual friendship 69
The problem with brothers 74
Conclusion 83

Chapter Four: Anti-Witchcraft and its Impediments: Migration,

the ‘Mixed Village9 and the Indian State 86
Witches, sorcerers and ghosts 90
Resolving disputes, migration, and land reform 95
Angadev: the collective response to village dis-ease 102
The demise of panbuddi: sex and marriage, detecting
witches, and ‘looking bad’ 112
Police, Naxalites, and the State 119
Conclusion 134

Chapter Five: Sect, Self and Sociality 138

The Mahanubhav sect 140
Suffering and the self 141
Dealing with illness 143
Salvation in Sukdi 148
Joining the sect and returning home 152
Satsang and the articulation of difference 158
Old kin, new kin? 164
Pilgrimage and the idea of Maharashtra 174
A Christian coda 178
Conclusion 183

vi
Chapter Six: From Protection to ‘Improvement’: The Commitment to
Vegetarianism and Visions of the Future 188
Food and the dangers of commensality 193
The power of vegetarian practice: protection not prestige 196
Vegetarians and meat-eaters: Gonds, sociality and commensality 203
The case of the missing horse: animal sacrifice and‘superstition’ 207
Food, drink and the ‘backward Adivasi’: the neighbouring
village of Kedanar 212
Hindu nationalism and a coincidence of values 217
The holy man’s sermon and Hindu modernity 222
Conclusion 227

Chapter Seven: Conclusion 229

Ritual friendship and the sect 232
State and nation 234
A question of conversion 237

Bibliography 242

vii
 List o f Figures

Figure 1: Table showing village composition by caste page 35

Figure 2: Landholding (in acres) by household and caste 50
Figure 3: The relationships between Naresh, Mukesh and Buddharu 66
Figure 4: The mahaprasad relationship between Manu and Devaram 70
Figure 5: The relationships between the Thakur’s family and Brijlal’s 146

viii
 List o f Plates

Plate 1: A view of Markakasa page 28

Plate 2: A village street 28
Plate 3: A large Markakasa house 29
Plate 4: A village street 29
Plate 5: Village children playing kabbadi* 34
Plate 6: A Gaita Gond sian* 34
Plate 7: Harvesting rice 54
Plate 8: Harvesting rice 54
Plate 9: Taking a break from harvesting 55
Plate 10: Laying out bundles of tendu leaves 55
Plate 11: My Mahaprasad 64
Plate 12: Women gathered for a chikat 85
Plate 13: Markakasa boy at chikat 85
Plate 14: A powerful baiga 93
Plate 15: The Angadev with attendants 105
Plate 16: The Angadev at ‘play’ 105
Plate 17: A baiga becomes possessed while the Angadev ‘plays’ 106
Plate 18: A Chhattisgarhi Teli bridegroom 117
Plate 19: Markakasa people building a road 136
Plate 20: Taking a break from the road-building work 136
Plate 21: Returning home after the day’s work 137
Plate 22: The Mahanubhav temple, Sukdi 150
Plate 23: Midday worship at the Mahanubhav temple,Sukdi 150
Plate 24: A Markakasa man and his son* 167
Plate 25: Building a tomb during karsad 185
Plate 26: The affines play ‘mud’ on the tombof a kinsman 185
Plate 27: Women carrying pots during karsad 186
Plate 28: Young men and women dancing and drumming during karsad 186
Plate 29: A baiga sacrificing a chick on a tomb during karsad 187

ix
Plate 30: Markakasa deity* 208
Plate 31: Markakasa deities* 208
Plate 32: Offering worship during Sanjori* 209
Plate 33: The Thakurdev, principal tutelary deity of the village 209
Plate 34: Village boy 228

* Photographs courtesy Ajay T. G.

x
 #Kondo?

— International b o unda ry ............ E xpress*

— State or unton territory boundary -------------- Road

National capital — —— » Railroad

State or unton territory capital

ThacPy of Chandigarh <s Urn capital at Ciwndtgarti Unton TanHory

and * a * o ttm c a p * a J of Punfab a r t Haryana Start*
Shiquanhe

M W t CDrtbmtf Onuc AqKMe SP 12NI33H

JlehiaJHw

NEPAL
ItJnagar
Lucjwuw

KsHM

HWia iT '
IJabaipt*

Portandaf"

afOwtt:* e*

MAHARJ
DAMAN AND DIU baneswai

^AJJIWRA , ,
P R A D E SH i/'''

Mangalora
Bangalori
HKtalWTY
>tRRY

K a/aratt Andaman
keralm ANDAMAN AND
r I '.Madura* . . . .
LAKSHADWEEP Cochin I L - NICOBAR ISLANDS

mCOBAH
ISLANDS

Laccadive

INDONESIA'

B ase 8 0 2 8 1 2 Al (C0 0 2 1 3 ) 1 0 01

Map 1: India

xi
 Gondia
District

Map 2: Locator map for Maharashtra State showing Gondia District.

xii
Aerial view of Markakasa

xiii
 Chapter One
Introduction

Brijlal was a happy man and not only because the rains had been good. He was
happy because a period of great sadness in his life appeared to have passed. As he tells it,
for about seven years beginning around 1993, there had been nothing but pain. Two sons
had passed away in infancy and then he himself had been at death’s door, struck down by
a mysterious illness, sapping him of strength leaving him unable to hold his own baby
boy. He tried many remedies, visiting doctors, healers and sorcerers, but to no avail.
Finally he was persuaded to go to a temple one hundred kilometres away where he was
assured he would find relief. He stayed there for several weeks along with his wife and
spent his days in prayer and devotion. Sorcery was determined as the cause of his illness
and on Brijlal’s behalf God began to fight the affliction. The magic loosened its hold and
his health improved; he and his wife decided to join the Mahanubhav Panth, the religious
sect that had established the temple. They became disciples of a guru there and returned
to their village as changed people. Their lives, so recently characterised by intense pain,
sadness and danger, were now transformed into ones of happiness, contentment and
safety. When I got to know him, there was no sign of his past woe: he had fully recovered
from his illness and he now had two healthy young children. He could farm his fields,
drive a bullock cart and ride a bicycle again. His life had changed in other ways too: as a
result of membership of the sect he had made a commitment to abstain from meat and
alcohol and worship One God (Bhagwan) to the exclusion of all others.

Brijlal was my landlord during my stay in Markakasa, the village in central India where I
carried out fieldwork. I slept and wrote in a room of the house and took my meals with
his family and in time came to be considered as one of them, “a person of the house”.
And Brijlal was one of my closest friends and a key informant. One of the few
‘intellectuals’ of the village, Brijlal loved puzzling over the issues of the day both local
and global, and would often come into my room after lunch while I was writing up my

1
notes and quiz me about the headlines and stories in my rather sporadically delivered
English-language newspaper. He is a Gond, which is a tribal or Adivasi (lit. original
inhabitants) community, classified under the Indian Constitution as a Scheduled Tribe
(ST), and is the largest tribal group in India spread over several states in the central
region of the country. He is not a particularly wealthy man, owning only three-and-a-half
acres of farmland, but he is influential, and though only thirty-four years old is
considered a sian (elder) and an important member of the panch (the village’s dispute
resolution body)1.

Political activity is an important part of Brijlal’s life. He is a volunteer worker

(karyakarta) for the Bharatiya Janata Party (BJP) and is regarded as the BJP’s man in
Markakasa. The BJP is an integral part of the Sangh Parivar2, an association of
organisations that promote a Hindu nationalist agenda in nominally secular India. Until
the general elections in 2004, the BJP had been the largest party in the ruling coalition
known as the National Democratic Alliance; though it is the party of government in
several States, it now sits on the opposition benches in the federal parliament. The other
prominent organisation of the Sangh Parivar, and the ideological engine of the entire
movement, is the Rashtriya Swayamsevak Sangh (RSS) which established the BJP as its
party political wing in the 1980s, and of which Brijlal became a member at about the
same time that he fell ill. Another member of the ‘family’ which has a particular presence
in my fieldwork area is the Vidarbha Vanavasi Kalyan Ashram (VVKA), an organisation
which works in areas in eastern Maharashtra (Vidarbha), where Adivasis predominate.
Once again, Brijlal is involved in its local activities, attending meetings and acting as a
guide for visiting dignitaries. The VVKA, which exists in different incarnations in many
parts of India, is principally concerned with what it calls ‘uplift’ or social development
and dharma jagran3. To this end it runs a boys’ boarding house near a high school in a
large village 11km away (where Brijlal intends to send his son when he is old enough). It
also organises local sports tournaments and distributes free medicine to villagers, and

1 This is a separate body from the gram panchayat. The latter is the elected village council.
2 Literally ‘the family o f associations’. It has no formal structure and includes workers, students and traders
groups as well as religious and anti-globalisation associations. Their common denominator is their
engagement whether partial or full in Hindu nationalist ideology.
3 ‘Awakening o f Moral Order’.

2
encourages its members to hold regular meetings (satsang) where religious matters are
discussed and correct Hindu practice taught. The VVKA’s overall objective is to keep
‘vulnerable’ groups such as Adivasis within the Hindu fold and safe from the attentions
of Christian missionaries and other ‘anti-national elements’.

This simple and partial account of the life of a man I got to know well raises one of the
principal questions of this thesis: what, if anything, does the narrative of illness and
healing have to do with Brijlal’s involvement in the Hindu nationalist movement? At first
sight the connection between the two may seem contingent: surely the two parts of
Brijlal’s life have very little to do with one another; and even if there were some
connection, how would the analysis of it be of wider anthropological import? It is my
concern here to look at the connections between the personal and the political; between
the local world of the people I lived with, and the political and ideological battle of ideas
and interest being fought in the larger arena of the nation.

More specifically, this thesis is an exploration of how and why Hindu religious sects have
become so popular in Adivasi areas of eastern Maharashtra and a consideration of the
implications of this religious transformation for both the self and local society. Contrary
to approaches which suggest that this is part of an inevitable process of a “Hindu mode of
tribal absorption” (Bose 1941), or of Sanskritisation (Srinivas 1966), it proposes that
membership of religious sects help solve ‘being-in-the world’ problems, which have
arisen because of rapid changes in local society. The consequences of sect membership,
however, are not limited to those who join. The practice of devotion restructures sociality
with kin and neighbours, and also reveals the dangers of such relationships. Adherence to
the sect also provides a commitment to the values of vegetarianism, teetotalism and
regular worship, which are understood by new sect members as key to improving society
more widely. An approach which explores the experience of religious transformation not
only helps us better understand the complex motivations for changing religiosity, but also
forces us to rethink key questions of traditional concern to Indian sociology such as status
and prestige. It also puts one in a better position to analyse the ways in which

3
commitment to certain practices can be mobilised for political projects by Hindu
nationalist activists.

At the level of people’s lived experience, there appears to be a crisis in the control of
witchcraft and magic, a crisis caused by the unavailability of satisfactory solutions for the
resolution of illness caused by malignant mystical attack. This crisis has its roots in the
. increased presence of the Indian State but also in the inability to migrate in order to
escape misfortune. This had led a growing number of people in the area to seek relief, as
Brijlal did, from a Hindu religious sect called the Mahanubhav Panth, which is regarded
as particularly efficacious in dealing with attacks of witches and sorcerers. Membership
of this sect protects people from further attacks of witchcraft and this in turn produces a
commitment to practices such as vegetarianism, teetotalism and time set aside for daily
worship. The sect and these practices not only have the effect of transforming sociality
with others and creating a new sense of self but also become part of a broader discourse
of the constitution of modernity. Full participation in modernity for those who come from
marginal areas is bound up with a distancing from the wild (jangli) ways of the past and
of others in the present, which serve to stigmatise Adivasi people and prevent total
national identification. These attitudes resonate with practices which are locally regarded
by Hindu nationalists as representative of ‘correct Hindu behaviour’ that serves to
strengthen the Hindu community by marking people as unambiguously Hindu. I
emphasise then the apparent coincidence of values that enables people like Brijlal to see
in Hindu nationalist ideology some resonance with the experience of their own lives, and
which makes the nationalist message all the more appealing. I describe it as coincidental
because the Hindu sects like the Mahanubhav that have become so popular have no
discemable connection to the Hindu nationalists; nor do Hindu nationalist activists seem
aware of the growth in sectarian membership and the implications that has on their level
of support.

In the next section I suggest some ways in which the ethnographic material presented in
this study can be framed. It is not intended as a summary of the chapters which comprise
the thesis but rather a discussion of some of the themes which cut across chapters. A brief

4
exposition of the experience of doing fieldwork follows that, and a chapter outline
appears at the end of the Introduction.

Orientations

Religious experience and nationalist identification

The idea for this study was bom out of a frustration that Hindu nationalism in India was
being discussed in the academy largely as a political force that swept all before it and that
those marginal groups in society the nationalist movement targeted - in particular
Adivasis and Untouchables - were represented as being somehow duped through social
or economic enticements into believing in the righteousness of such a movement (e.g.
Hansen 1999: 103-6). Some of the analysis that followed the massacres of Muslims in the
state of Gujarat in 2002 centred on the fact that it seemed to be those such as Adivasis
and Untouchables who were doing much of the killing in rural areas: the conclusions
were invariably drawn that years of Hindu nationalist activity in villages in the area and
anti-Muslim propaganda had radicalised elements of the population (see Yagnik and
Sheth 2002; Balagopal 2002; Lobo 2002). I do not want to suggest that such
propagandising in the state which has been called Hindutva’s laboratory is of no
importance at all. But such accounts which place action and agency firmly in the Hindu
nationalist activists’ court effectively prevent us from asking other salient questions: why
might Hindu cultural and religious nationalism as ideology and practice appear attractive
and meaningful, not only for people who took part in the violence or are members of
nationalist organisations, but also for those numerous others who have no clear party
political affiliation?

These accounts of Hindu nationalist activity mentioned briefly above do not explore in
any great depth the local contexts of social, economic, and religious change that are
taking place in areas where Adivasis and other marginal communities live. This thesis
offers an ethnographic exploration of the environment in which Hindu nationalism
operates and attempts to gather support. In so doing, it takes the phenomenon of Hindu

5
nationalism as its background and its end point, rather than its beginning. In particular, I
focus on people’s changing religious lives and look at how people who come to adopt
certain practices might find the message of Hindu nationalism appealing. In common
with others such as Van der Veer (1994) and Kapferer (1997), I suggest that the logic and
symbolism of nationalism operates at a level beyond the immediately political and
resonates with the most fundamental existential crisis human beings can face: that of
falling ill and being close to death, and then being cured. At the heart of this is ritual. As
Van der Veer writes: “What seems to happen in religious nationalism is that ideological
movements give a new interpretation to the cosmological understandings communicated
in religious ritual. The nation is presented as an extension of the self and nationalism as
part of religion, dealing with shame and honour, the illness and death of the person...”
(1994: 84).

I do not suggest that an instrumentalist approach is of no value. On the contrary it is

essential, as the study by Froerer (2002, 2006) demonstrates. She examines the ways in
which political Hindu nationalism is introduced to a mixed Hindu/Christian Adivasi
community in Chhattisgarh. She focuses on how RSS activists use local disputes to
promote a nationalist agenda by transforming ethnic identity in the image of wider
concerns of the threatening ‘Other’. But such an approach is not sufficient to understand
the complexities of the issue. The question one needs to ask, in addition to how Hindu
nationalist agents use local disputes to spread their ideology, is why the targets of their
ideology may be attracted to such ideas in the first place. Searching for and examining
the answers to this question are essential if we are to understand the long-term
sentimental attraction for people of Hindu nationalism. To neglect to do so reduces
religion wholly to politics, and the ‘Hindu’ element of ‘Hindu nationalism’ becomes a
hollow shell manipulated by those wishing to retain or capture power. Hindu nationalism,
in that interpretation, is the pursuit of political dominance conducted by other means.
Anthropologists have to be wary of this reductionism: the religious is just as important as
the political and they implicate each other profoundly. The Hindu-ness of the nationalism
has to be explained and accounted for in contemporary ethnography. Such unwillingness
to engage in the implications of religious transformation leads to accounts of the

6
relationship between marginal communities and Hindu nationalism which privilege the
ostensibly political and economic over the religious. Baviskar, for instance, asserts that
the “multiple spiritual and cultural meanings embodied in the practices of being bhagat [a
member of a Hindu devotional sect] now matter less [my emphasis] than its potential for
recruiting Adivasis to political Hinduism” (2005: 5108). But it is precisely in unpacking
those multiple spiritual and cultural meanings, in examining why and how Adivasis and
others undergo such radical religious transformation, that one can begin to understand the
attraction of Hindu nationalism, and how relations of power are really configured. If it is
true that religious or ethnic identities are being mobilised by elites (Brass 1974, 1979),
one at least needs to examine how those religious identities not only come to be formed
but have such power. As Van der Veer suggests, Hindu devotional movements have long
played a role in the formation of religious identity, and though the movements themselves
“remained bound to certain regions or social groups so that Hindu identity continued to
be fragmented” (1994: 50), a common denominator of something that is recognisably
Hindu can be discerned in the discourse and practice of devotionalism (ibid.). Thus it
seems that an examination of these devotional sects is especially pertinent.

It may sound like a truism, but Hindu devotional movements are clearly not the same
thing as Hindu reform/ revival movements. The importance of Hindu reform movements
in the creation of a Hindu identity unified in practice and ideology has tended to be
stressed, both in the anthropological literature and in the historiography (see Bhatt 2001).
An emphasis on these sorts of politico-cultural movements, such as the Arya Samaj
which emerged during the course of the nineteenth century, tends to obscure the
importance of already existing religious formations in Indian society such as the
devotional sects. Some of the latter, such as the Mahanubhav, were established in
mediaeval times. Indeed, the focus on these more recent religious formations naturally
privileges the political in any analysis of the emergence of Hindu nationalism because of
their undeniably political aims. It is these organisations which are involved in mass
programmes of religious change: the Arya Samaj organises shuddhi (purification)
campaigns of ‘reconversion’ of Adivasis from Christianity to Hinduism. Similarly, the
RSS runs a ghar vapsi (homecoming) campaign with much the same end. Away from

7
these spectacular campaigns with overtly political messages and concerns, devotional
sects continue to attract adherents for multiple reasons and it is some of these reasons that
I look at in detail.

There are resonances at a cosmological level between the conceptions of moral order and
disorder of local society, which are primarily associated with the disruptive effects of
witchcraft and sorcery, and the symbolism of Hindu nationalism more generally.
Witchcraft and sorcery can affect individuals, as the example of Brijlal demonstrates, but
people also attribute a sense of dis-ease of the wider community of their village to the
actions of witches and sorcerers. This sort of witchcraft is largely attributed to people
within the village, and is thus an internal problem. But the threat is also exteriorised by
generalised discourses of difference which emphasise the influx of ‘different kinds of
people’ whom one cannot really trust, and which not only change the composition of the
village but also make it harder to combat witchcraft itself. This sort of logic has
considerable parallels with the logic of Hindu nationalism, with the obsession with an
interior ‘Other’ that needs either to be incorporated or made exterior and expelled
(Jaffrelot 1993: 522). There are similar processes at work: just as individuals and villages
are attacked from within, and are healed and thereby transformed, so the nation as
conceived by Hindu nationalist ideology is under attack and needs to be healed and
transformed (Hansen 1999: 10-13). I am not suggesting, however, that the resonances at
this cosmological level are what cause people to act in the way that they do. I am simply
proposing here that, for instance, the symbolism and ritual of healing is part of a larger
cosmological system that goes beyond the individual or the local and is found at the (not
necessarily more interesting) level of the nation.

Poles apart? The idea o f the 'Adivasi ’ and Indian modernity

The question of Adivasi-ness runs through much of this thesis. While it is true that in the
particular village in which fieldwork was conducted, Adivasi people formed a minority,
the area as a whole was nevertheless described as an ‘Adivasi area’ by locals and
administrators alike. Though the condition of Adivasi-ness was located largely in those

8
who defined themselves as Adivasi or members of Scheduled Tribes, there were
important restrictions and expansions of this idea. Restrictively, the use of the term
‘Adivasi’ in this area is meant to refer to members of the Gaita Gond caste. As a
consequence of their distinctive life-cycle rituals, their language and dress, and their self-
perception that they are the ‘pure Adivasi’ (which in turn is founded largely on their own
perceived distinctiveness), they have come to be identified with an irreducible Adivasi
character. By way of contrast, in its most expansive definition, all people resident in an
Adivasi area have something of the Adivasi about them. Thus the widely held view that
people of any caste in this area were perceived as having less formal education than in
other parts of the region, or that people were poorer or more superstitious, was attributed
by officials and villagers alike to their predicament of living in an Adivasi area. In all
cases, whether restrictive or expansive, the concept of Adivasi is unquestionably linked to
that which is backward (magaslele) or wild (jangli), as opposed to the ways of other
Indians in other areas who are more civilised or advanced. Chapters Four and Six
highlight aspects of the association of Adivasi with ‘backwardness’.

Chapter Four discusses the impediments that have developed in the past fifty years or so,
which make the identification and public accusation of witches and those who employ
sorcerers much more difficult in Markakasa. Two ritual forms, Angadev and panbuddi,
were used in the past to identify witches and other sources of mystical attack. These
forms are no longer available to Markakasa people largely because of what they connote
as examples of Adivasi ‘backwardness’. Angadev, a witch-detecting device, is seen as
evidence of Gond superstition and lack of modernity by both non-Adivasi villagers and
government officials. In a mixed caste village such as Markakasa then, the Angadev is
unavailable as a remedy because of a lack of agreement as to its efficacy. Similarly, the
ritual of panbuddi in which a bride is tested for witchcraft, performed before a Gond
wedding, has largely been abandoned because it expresses an ideology of sexuality which
runs counter to that which is regarded as modem or civilised by many in Markakasa. In
both these cases, the relationship between the idea of the Adivasi and the notion of the
primitive is clear. A consequence of this particular articulation of the Adivasi leads to the
unavailability of certain measures of witch-detection and conflict resolution.

9
The discussion of vegetarianism and teetotalism in Chapter Six also demonstrates this
association of Adivasi with backwardness. Though Markakasa sect members adopt these
food practices in order to fulfil the functional ends of protection from witchcraft,
vegetarianism and teetotalism come to express key ways in which Markakasa people
distinguish those who are backward from those who have improved (sudharlele). People
who are sudharlele are regarded as more able to participate in modem India as
represented through politics and education. The largely Gond neighbouring village of
Kedanar is often discussed as more backward than Markakasa, not only because the
former lacks the mixture of castes of the latter and is more recognisably Adivasi in
composition, but also because of their excessive consumption of meat and alcohol.

There has been considerable discussion throughout the last century and into the present
one about the category of ‘tribal’ and ‘Adivasi’. The famous debate between G. S.
Ghurye and Verrier El win (Ghurye 1943; El win 1943; Guha 1999) about the place in
Indian society o f people classified as tribal was primarily concerned with questions of
policy. Were they, as Ghurye proposed, to be encouraged to assimilate or, as Elwin
campaigned for, were they in need of special governmental protection to ‘preserve their
way of life’? Underlying this debate was a fundamental disagreement as to who those
called Adivasi were. Ghurye saw them as low status and marginalised Hindus and
stressed continuities between ‘tribal culture’ and mainstream Hindu culture, whereas
Elwin, for much of his career, seemed to consider the two in opposition to each other.
Others such as Bailey (1960, 1961), Dumont (1962) and Beteille (1987) discussed the
sociological usefulness of the term and whether one could oppose it to the category of
caste. There have also been several studies which explore the colonial construction of
tribe particularly in contrast to that of caste (e.g. Skaria 1997b; Guha 1998). Rather than
consider these issues, which have already been extensively explored, I want to focus in
this section on the notion of the ‘Adivasi’ through the context of two political movements
that were active in the area where I did fieldwork: chauvinistic Hindu nationalism and the
Maoist-inspired Naxalite movement. Both these political projects offer re-interpretations
of the notion of Adivasi or tribal which diverge from the conflation in symbolic terms of

10
Adivasi with that which is primitive and radically other in Indian society. While the
former seeks to discourage the use of the word altogether, the latter appears to offer a
symbolic re-interpretation of it.

As others have noted (e.g. Froerer 2002; Sundar 2006a), Hindu nationalist organisations
such as the RSS contest the word ‘Adivasi’. That it means ‘original inhabitant’ offers a
challenge to Hindu nationalist ideology, which stresses that all the people resident within
India today are autochthonous to it and have always been Hindu.4 VVKA activists around
Markakasa told me that ‘Adivasi’ was a mischievous invention by India’s erstwhile
colonial rulers, which had the purpose of dividing the population. In their interpretation,
the concept of Adivasi needs to be seen less as a specifically ethnic term (as suggested by
the restrictive understanding above) and more as a condition of geographical being (as
suggested by the expansive understanding). In this respect, contemporary Hindu
nationalist thinking recalls Ghurye’s position. For the RSS and VVKA, ‘vanvasi’ (forest-
dweller) is a more appropriate label since it does not differentiate on the basis of caste. It
is merely a descriptive geographical term that can be opposed to the categories of
saharvasi (city-dweller) and gaonvasi (village-dweller). This neologism was met with
hostility from some in Markakasa. On one occasion, a couple of VVKA activists went
from door to door handing out leaflets to villagers explaining the function and activities
of the VVKA. A Gond man, politically active in the Congress Party,5 publicly challenged
them on the use of the word ‘vanvasi’. Returning the leaflet he had been handed, he
argued that the replacement of ‘Adivasi’ was part of a plot by the BJP to rewrite the
Constitution and revoke rights to government benefits and reserved public sector jobs that
Adivasi people were entitled to. Other villagers gathered around and started to hand back
the leaflets too. By erasing the concept of ‘Adivasi’ as a special category of people, the
Hindu nationalists were accused of denying that Adivasi people were in need of more
assistance than other people; in a sense therefore, and paradoxically so, the activists

4 Which has the implication o f course that Muslims, Christians and others are ‘really’ Hindus waiting to
come back to the fold.
5 Campaign speeches in the run up to the 2004 General Election by the Congress Party, the Nationalist
Congress Party (NCP) and the Bahujan Samaj Party (BSP) all mentioned this apparent BJP promotion of
vanvasi over Adivasi and the threat it represented. The BJP speeches I heard never mentioned ‘vanvasi’ but
always spoke in terms o f Adivasi. Vanvasi was used far more by VVKA activists.

11
appeared to deny that Adivasi people as a particular group or caste were more ‘backward’
than any other in the area.

At the other end of the political spectrum, the Naxalite insurgency in central and eastern
India has gained ground in largely Adivasi areas. Maoist-inspired, its ultimate aim is the
overthrow of the Indian state. While it is presented as a class rebellion of the peasantry
against exploitative landlords, moneylenders and government officials, it has also taken
on the characteristics of caste conflict in many parts of the country, particularly in Bihar
and Jharkhand. Though I discuss the Naxalite insurgency in more detail in Chapter Four
and argue that it has drawn the State ever closer in both imagination and fact to
Markakasa people with implications for how witchcraft is dealt with, I want to focus here
on the idea that the Naxalites in and around the Dandakaranya region (the Gond areas of
eastern Maharashtra and southern Chhattisgarh) are interpreted by local people as an
ethnic movement of Adivasis (principally Gonds) against non-Adivasi outsiders. This
identification of the Naxalites and Adivasi ways of being in turn suggests a possible
symbolic re-working of the ‘Adivasi as primitive’ into the ‘Adivasi as modem’.

The Naxalites have been present in and around Markakasa since the early 1990s and visit
the village periodically, though there have been no general meetings held for a number of
years. I had the opportunity to meet local Naxalites but declined to do so because I would
have had tremendous problems with the police had they found out. Thus the analysis
which follows is based on Markakasa people’s impressions of the Naxalites they
encountered.

Naxalites were seen as Adivasi people and in turn Adivasi people were regarded as their
most fervent supporters. In Markakasa, small squads would regularly visit a number of
Gond houses and ask for food or shelter, and enquire about goings-on in the village.
When the names of captured or killed Naxalites appeared in newspapers, there would be
considerable commentary that the insurgents always seemed to have Gond surnames.
Their close identification with Gond-ness was often highlighted through language.
Markakasa people said that the Naxalites spoke ‘pure Gondi’ and would not readily

12
switch to other languages. They asked Adivasi people to speak Gondi to their children in
order to protect their sanshruti (culture) from others. When they visited villages, all the
songs they sung at the beginning and end of meetings were in Gondi.6 While their Gond-
ness was stressed, the Naxalites also seemed comfortable in the milieu of modernity.
They were credited with having special telephones which they carried around in the
forest and which they used to communicate with one another. On their feet they wore
special boots which helped them run silently through the forest to avoid detection by the
police. According to Markakasa people, their leaders were well-educated men who spoke
English and had gone abroad. This interface between modernity and the Adivasi was well
expressed in an Indian film that was fairly popular in the village while I was there.

Laal Salaam (Red Salute) tells the story of a Gond man who hails from a small village in
an Adivasi area7. He becomes one of the first fully qualified medical doctors in the state
and is doing rather well for himself. One day he hears of an atrocity perpetrated by police
and forest officials on people in his home village and when he returns to investigate
becomes the subject of violence himself. Outraged and angry he joins the Naxalites and
begins to wreak revenge on those who abused him and his fellow villagers. What is
interesting is that this man had been the epitome of modem India for Markakasa people:
being a Gond he had nevertheless become a doctor, a position he had given up to enlist
with the Naxalites. Though characterised by Gond-ness, the Naxalite movement in the
film is presented to village viewers as profoundly modem.

The relationship between modernity, Naxalites and the idea of the Adivasi also becomes
apparent in the treatment of gender relations. Naxalite women were often described by
Markakasa villagers as taking leading roles in meetings, and they were often the
perpetrators of violence (against men). A Markakasa man recounted to me how he was
walking down one of the village lanes in the middle of the day and out of nowhere a
bicycle came hurtling towards him and almost knocked him over. He shouted insults at
the woman who was riding the cycle. The woman just looked at him and rode off. He

6 Nandini Sundar (2006b) on Salwa Judum and Bastar - promotion o f Gondi by Naxalites.
7 The filming was actually carried out in neighbouring Gadchiroli district.

13
recognised her as a Naxalite that he had seen several weeks earlier as he was walking
through the forest, and for the next few days he was terrified that she would come back
with the others and beat him up. When I asked people about why women became
Naxalites, one of the most common answers I received was that the women who joined
were Gonds, and Gond women were strong and unafraid of men. This perceived strength
and independence are paradoxically a source of shame, which has led to the abandonment
of the ritual of panbuddi, (discussed further in Chapter Four). Panbuddi expresses the
idea that women (and men) do have choice in matters of sex and love, an attitude which
is regarded by Markakasa Gonds as ‘looking bad’ in front of other castes who do not
share that view. The ritual is seen as a component of Gond-ness that enables other castes
to regard them as backward or primitive or lacking proper culture. The fact that Gond
women are seen to participate in the Naxalite movement then perhaps reconfigures this
attitude. Instead of the association of Gond-ness or Adivasi-ness with ‘backward’, the
modernity of the Naxalite movement lends an aura of modernity to a view of gender
relations and sexuality otherwise seen as paradigmatically primitive, shameful and
‘other’.

Discussing conversion

A focus on religious experience to understand both social change and changes in

sociality, leads one naturally to an examination of religious conversion in India and
elsewhere. It is striking that conversion within Hinduism (from non-sectarian to sectarian,
or from one sect to another) has received far less attention8 than conversion from
Hinduism to Christianity (e.g. Eaton 1997; Viswanathan 1998; Clarke 2003), or to
Buddhism (e.g. Burra 1996; Fitzgerald 1997). Yet as the discussion in Chapters Five and
Six demonstrate, becoming a different kind of Hindu has profound implications for self
and society too.

The anthropological literature on conversion is vast and it is not possible to review it all
here. Rather, I want to discuss the ways in which the material I present on religious

8 Though see Sharma (1977), Dube & Chatteijee Dube (2003), Hardiman (2003), and Sarkar (1999).

14
transformation may be usefully examined in the context of some current anthropological
thinking on conversion.9 There appear to be three sets of issues.

Firstly, as others have noted (e.g. Hefner 1993: 20-25; Robbins 2004: 84-87), theories of
conversion tend to fall into one of two broad camps: the so-called ‘utilitarian’ approach
and the ‘intellectualist’ approach. The former tends to emphasise the social benefit,
whether material or moral, of converting to a new religion (e.g. Laitlin 1986; Peel 1977).
The latter instead tends to see conversion as an attempt at meaning-making on the part of
persons or communities faced with changed social and cultural circumstances (e.g.
Horton 1971).

Secondly, there is a debate in the literature between those scholars who wish to highlight
the discontinuities produced by conversion - and thus the radical nature of the religious
transformation for self and society - (e.g. Meyer 1998; Robbins 2003; Engelke 2004;
Keller 2005) and those who rather emphasise the essential continuities in society and
culture despite the fact of changed religious affiliation (e.g. Sundkler 1961). As Robbins
points out, this emphasis on continuity naturally lends itself to the utilitarian approach:
since religious affiliation is changed for reasons of gain, people who convert “...rarely
come to understand that world religion as anything more than a means to locally defined
worldly ends” (2004: 85). Robbins notes that anthropologists can claim therefore that so-
called ‘traditional religion’ survives despite the appearance of a world religion such as
Christianity (ibid.).

Thirdly, there is a divergence among scholars between those who direct their analysis to
the act of conversion itself, and those that focus on why people stay committed to a
particular religion once they have converted (e.g. Keller 2005). This latter approach
recognises that what makes people join a particular religion may be quite different from
what makes them stay affiliated. In his study of the conversion to Christianity of the

9 Anthropologists studying conversion in India, whether from Hinduism to Christianity or from ‘little
tradition’ Hinduism to ‘great tradition’ sectarian Hinduism would stand to benefit from paying attention to
these wider debates in the anthropological corpus. The recent collection by Robinson and Clarke (2003) for
instance makes little reference to this wider literature.

15
Urapmin of Papua New Guinea, Robbins (2004) attempts a synthesis of these two periods
in the life of converts. He suggests a two-stage model of conversion, one in which
utilitarian concerns give way over time to intellectualist ones as people begin to
understand the religion that they convert to (ibid: 87).

These debates have a number of resonances with the sort of religious transformation I
describe in the thesis. As I show in Chapters Five and Six, joining the Mahanubhav sect
produces profound discontinuities in social life: becoming an adherent of the sect is not
only the consequence of social change but is truly transformative, changing society (and
notions of society) in return. The kinds of practices demanded of sect members are
markedly different from the observances of non-sectarian village Hindus and are
tremendously disruptive of kinship and other relations. New conceptions of sociality and
personhood, which are supported by the practices and doctrines of the sect, come to be
important. And yet, new adherents do not talk of their ‘conversion’ (dharma suikarna) as
a ‘complete break with the past’, as the Ghanaian Pentecostal Christians in Birgit
Meyer’s (1998) study do. Rather, while accepting that they have become different people,
Mahanubhav devotees Still see themselves as full members of their caste group and work
hard to be recognised as such. Similarly, while kinship relations are transformed in some
areas (such as in the worship of household gods and sharing a kitchen), they remain
unchanged in others (such as the pooling of household labour).

When we turn to look at motives for conversion we can discern elements of both the
‘utilitarian’ and ‘intellectualist’ approaches. All the Mahanubhav adherents I spoke to had
joined the sect for a definite and unambiguous purpose: to help cure an intractable illness
caused by witchcraft or magic. But also evident was a clear sense of an ‘intellectualist’
motivation: that the incurability of the illness led to a loss of ability to make meaning and
gain personal satisfaction from the range of actions (from medical doctors to sorcerers)
one would ordinarily have pursued. In my analysis in Chapter Four, this loss of meaning-
making is in part a consequence of state policies, both in the colonial past and the post­
colonial present, such as increased police (and thus state) activity because of the decade-
old Maoist insurgency, and effective implementation of legislation designed to protect the

16
integrity of Adivasi landholdings. As I discussed above, these state policies are motivated
in part by the consideration that Adivasi people and the areas in which they live are more
‘backward’ (and thus in need of more attention) than the rest of India.

Finally, like Robbins and others, I found that the reasons why Markakasa people stayed
committed to the Mahanubhav sect needed to be examined in addition to the
circumstances relating to the initial conversion. This is what I emphasise in Chapter Six
in my discussion of the practice of vegetarianism as a consequence of membership of the
sect. Whereas Keller (2005) sees a significant difference and discontinuity in motivation
between the point of conversion and afterwards for Malagasy Seventh-Day Adventists, I
show that the desire to be protected from further attacks of magic and witchcraft is what
produces a commitment to the sect in the long term, too. But this is not all that is going
on. Practices and attitudes promoted by membership of the sect become part of a local
imagination of an ‘improved* and ‘modem’ society, resonating too with the message of
various Hindu nationalist actors in the area. Thus, the realisation that membership of the
sect leads one to become a ‘good Hindu’ makes one’s commitment to its practices and
doctrines all the stronger.

How can we begin to examine ethnographically the questions of motivation for religious
change, and the reasons why people remain committed to new religious practices and
doctrines? One way would be to use as a tool a focus on the bodily experience of those
who join the sect.

Experience and embodiment as method

The body is the key way in which Markakasa people experience and talk about mystical
attack, healing and protection. First, the malicious attentions of others, whether they are
sorcerers or witches, are generally felt bodily, ranging from the merely irritating to the
catastrophic: headaches, exhaustion, blackouts, nightmares, temporary paralysis,
madness, death. Then, a cure is sought at a Mahanubhav sect temple, where the body of
the sufferer is subject to new disciplines of prayer, time, diet and space. On joining the

17
sect and returning to one’s village, these new bodily practices of diet and worship learnt
at the temple have to be observed in order that the devotee can continue to be protected. It
is appropriate then to approach the question of religiosity and its social consequences in
terms of anthropological thinking on embodiment, and the way in which the classic
mind/body dualism is collapsed. This method provides an alternative perspective on the
reasons why religious change in the form of membership of a religious sect is attractive.

There has been a growth in anthropological thinking about the place of embodiment and
experience in the analysis of culture and society (see Connerton 1989; Csordas 1999,
1994, 1999; Strathem 1996). Though we might see this new focus as part of a Western
theoretical preoccupation brought about by the increasing instability, of the category of
the body in Euro-America, the role of the body in social life has a long history in the
anthropological corpus as Hertz’s (1960) analysis of the right hand and Mauss’s (1934
[1979]) famous essay on the techniques of the body amply demonstrate. Recent
approaches have, however, emphasised the notion of the ‘mindful body’, a body which
plays an important part in perception and thought. Contrary to certain Foucauldian
analyses, they recognise that the body is not only the outcome of social processes but,
through experience and emotion and a process of embodiment, is the site of meaningful
social production itself. These theories find a natural ally in phenomenology and, taking
their cue from Merleau-Ponty (1989), attempt to explain the life worlds of social actors
through a focus on experience. In this way, social actors emerge as agents of change.
Religious transformation, for example, is not only individually meaningful but also
socially transformative. By looking at the bodily and experiential imperatives that lie
behind certain practices, a reinterpretation or refinement of existing work on the
relationship between religion and politics can be proposed.

In his analysis of the Devi movement in southern Gujarat in the early twentieth century,
Hardiman (1987) argues that the Devi was essentially a way in which Adivasis could
compete in the struggle over power. It did this by selecting those practices (such as
vegetarianism and teetotalism) which were dominant as representative of a powerful
strand of pious Gujarati being, and which enabled Adivasis to contest local Parsi

18
oppression (ibid: 164). Similarly he suggests that Adivasis converted to Christianity in
north-eastern India because of the desire to emulate and appropriate the power of the
Christian colonial authority there. His analysis, however, fails to consider why such
practices might be meaningful at an existential and experiential level removed from the
question of local politics. One could suggest a functionalist refinement of Hardiman’s
position: people are not necessarily drawn to practices in themselves but invest them with
meaning because of the reasons they come to practice them. Thus, Mahanubhav sect
members in Markakasa see the practices of vegetarianism, teetotalism and daily worship
as meaningful because they are part of a package of measures that protects them from
harm. The experience of pain and suffering, and the transformation to happiness and
safety one effects through membership of the sect, produces a commitment to these
practices and the desire never to suffer in the same way again. The importance of such an
approach is that we can observe the reinterpretation of dominant values by people who
join the sect for functional ends.

This thesis is written in the traditional anthropological fashion of breaking down barriers
between concepts other disciplines tend to treat discretely, and aims to show how
domains of social life such as ‘religion’ and ‘politics’ are intimately connected. The
orientations I have outlined above set the scene for the ethnography which follows in the
main body of the thesis, and I will return to them in the Conclusion.

The experience of fieldwork

I conducted fieldwork in Markakasa for approximately eighteen months between

November 2002 and June 2004, and again for a month in April 2005.1 had been to India
several times during my childhood but my experiences had largely been limited to urban
India. I decided to conduct fieldwork in Maharashtra, the part of India from where my
family hails, largely because it was a state that has a substantial Adivasi population,

19
particularly in its eastern districts, and also because I was already fairly familiar with the
Marathi language.10

My main criteria in determining a fieldsite were the presence of VVKA activists and a
mixed Adivasi/non-Adivasi population. On my arrival in India in October 2002,1 lived in
Nagpur, the main city in eastern Maharashtra and stayed with my maternal uncle (mama)
and aunt (mami). After talking to friends of my Mama who knew the area well, and
consulting the Census for population information, I decided that a village in Tahsil Deori
of District Gondia would be a good place to do fieldwork. I learnt that not only did Hindu
nationalist activists in the form of the VVKA have a presence in the area but the mix of
people was also unusual: Chhattisgarhi migrants had settled the land alongside Marathi
and Gond people.

It was through contact with the Vidarbha Vanavasi Kalyan Ashram (VVKA) that I was
able to settle on a fieldsite. I presented myself at their office in Nagpur and asked if I
could meet one of their workers in the southern tahsils of Gondia district. He asked me
several questions, including the name of my Mama, and after getting permission from the
director of the Ashram wrote me a letter of introduction to a schoolteacher and VVKA
worker called Mr. Anjankar.

I went to see Anjankar at his house in a small lane where several other teachers lived. As
often happens on these occasions, he was not at home and I waited in his living room
while his wife found a village boy to go and fetch him from school. He did not take long
to arrive and I explained why I had come to see him. I mentioned the names of villages
which from the Census and their location seemed suitable, and he offered to take me on a
tour of them. He said that he was friends with a man in Markakasa (one of the villages on
my list) who could probably rent me a room in his house. We jumped onto his motorbike
and set off. After about twenty minutes of winding roads through the forest we came to
Markakasa and parked the bike in front of a large house. Anjankar explained to Brijlal

10 Though this turned out to be less useful than I thought because o f the range o f different languages spoken
in Markakasa.

20
who I was and why I was there and he agreed to take me in. He had had many people stay
with him in the past he told me, mostly teachers posted to the village school: his was a
good house, and I would have no problems living there.

I used the VVKA to find a fieldsite in order to allay their fears about my intentions: I
thought that since I had arrived in the village through their introductions I was less likely
to face problems from local activists. My reasoning was not entirely sound, as I realised
when I was interrogated, around a month after my arrival by visiting higher-ups to the
area, about my work and reasons for choosing to do it in this area of the country. The
Sangh regarded this part of the district as ‘sensitive’, since Christian missionaries had
worked hard to convert parts of the population but had largely been unsuccessful and the
Sangh was anxious to keep it that way. They were concerned that my research would in
some way be used to help the Christians discover why they were meeting with such little
success. The fact that I am a Maharashtrian Brahmin11 partly allayed their fears, and I
was bluntly told that had I been a European the Sangh would not have permitted me to
continue fieldwork in the area. Nevertheless, the men I met were deeply suspicious of my
supervisors, who in their eyes were in all probability Christians who had sent me to India
as their unwitting pawn.

As is the case with most fieldworkers, some villagers were rather suspicious of my
presence too. The timing and manner in which I arrived in the village may have not been
entirely propitious. After I had made my initial visit to Markakasa with Anjankar, I told
my landlord Brijlal that I would return in a couple of weeks with my belongings.
Meanwhile, a friend who was travelling in India came to Nagpur and I suggested she
accompany me to the village and spend a night there. We arrived at Markakasa, dumped
our luggage in my room and then set off on a walk around the village. We decided to go
around the perimeter, walked through a patch of forest and emerged on the other side of
the village. The lanes were deserted. I later learnt that some of the villagers had seen us

II Many members o f my extended family in India are active or were at some time involved in the RSS and
its affiliates. The closeness o f my fam ily’s association with the Hindu nationalist movement was brought
home to me by my discovery o f my maternal grandfather’s photograph album where I found old pictures o f
V. D. (Veer) Savarkar, the principal ideologue o f Hindutva, attending family gatherings and birthday
parties. I did not explain all this to the Sangh people I met.

21
walk out of the forest, remarked that my friend was wearing green combat trousers and
concluded that we were Maoist fighters and rushed into their homes: the insurgents had
called for a three-day bandh in the area, which meant that villagers could not work in
their fields, and we were assumed to be to checking that the order was being observed.

Some of the older men, on learning that I was from England, became concerned that I
was working for the British, attempting to resurrect the Raj using their village as a base.
My identity never seemed clear to them and was a source of much confusion. I am British
of Indian origin, having never lived in India before and from a Maharashtrian Brahmin
family. I assumed that when I replied to questions of my caste status I would cease to be
viewed as a complete foreigner. But it was clear that people could not conceive of a
world without caste and assumed that all people wherever they lived had it. Paradoxically
then the fact of being a Brahmin did not naturally mark me as an Indian or a Hindu.
Bizarrely, for me, I was often asked if I was a Christian. My woeful language abilities
must have made me appear even stranger. Suspicions quickly abated after my Mami’s
first visit to Markakasa a couple of months into my fieldwork. She was an instant hit,
and, coming from a farming family herself had plenty of appropriate and interesting
questions to ask the villagers. From then on, I was seen as less strange, as having a
connection to India that the villagers understood better. Mami visited several times after
that and when my cousin got married in Nagpur, Brijlal and my ritual friend Ganeshyam
(see Chapter Three) were invited to the wedding. Ganeshyam in turn telephoned Mama
and Mami to invite them to his wedding in Markakasa in May 2006. On the part of
Markakasa people, my Mama and Mami represent a link to me even though I have now
left the village. They know they can contact them and have a message passed onto me.

I found the combination of novelty and familiarity disorienting at first. I did not ask
questions about the meaning of puja (worship) or why people ate only with their right
hands: these were questions I had asked and had had answered throughout my own life.
Asking questions such as these would have been as ridiculous as asking what feet were
for. But I quickly realised that my lack of questioning on issues that I thought I already

22
knew the answer to would prejudice my data more than necessary. So I swallowed my
pride and started acting dumb.

I quickly settled into the rhythms of village life. As I mentioned above, I rented a room in
Brijlal’s house and he provided me with meals and his wife washed my clothes. His
household consisted of his wife and their two young sons; sharing the same courtyard and
entrance was the house of his brother-in-law and elder sister and their three children. It
was winter when I arrived, and for most people the main agricultural season was already
over and the grain bins were full of newly harvested rice. December was a perfect time to
begin fieldwork: the weather was temperate; people were mostly idle or occupied in light
tasks and so more willing to talk; and the cool mornings and evenings meant that there
was a reluctance to move away from the fire around which one sat and chatted. Brijlal’s
was a lively house and there were always people dropping by to exchange gossip and
news. It is placed somewhat outside the main settlement of the village and beside a path
which leads to the school, the village council {gram panchayat) building and the fields
beyond. To the front of the house is a large open space (maidari), which leads down to a
small lake circled by mango trees where people bathed and washed their clothes. People
liked coming to his house because from the veranda one could watch the world go by and
enjoy the cool breeze in summer, away from the stifling heat and still air of the main
village (<basti). I was fortunate that Brijlal is the sort of man who likes collecting people;
he took me under his wing and I quickly got to know a wide range of villagers.

I restricted most of my research to Markakasa and did not conduct any significant amount
of fieldwork in any other place. In many senses however, the fieldwork was multi-sited
since I often accompanied villagers on trips to other places. Much of the fieldwork was
conducted relatively informally. I spent the first six months trying to identify what
seemed to be important issues for people themselves and building relationships with
Markakasa’s inhabitants. My first task was to make a plan of the village and I recruited a
village youth who was recovering from a bout of malaria to take me on a walking tour,
making sure I included every house. I showed this to my landlord who over several
evenings patiently gave me the name of every head of household in the village and their

23
caste. In those first few months I also collected the bulk of the thirty genealogies of
village families. They were invaluable, not only in determining how people in the village
were related, but also as tools for eliciting narratives of in-migration to Markakasa, life-
histories and marriage patterns. Sometime further into my fieldwork, I consulted the land
register held by the village accountant (talathi; patwari) and recorded ownership of land
by household. In my final two months I conducted an extensive household survey
recording education, land ownership (including how and when the land was acquired),
rice production, income, debt, migration, marriage age, location of wedding, rate of
brideprice, incidence of divorce and ritual friendship. Apart from these fairly structured
forms of research, the majority of my time was spent visiting people in their homes,
accompanying them to the fields and forest and talking to people I found at gathering
places in the village such as the tea and paan stalls and the shop.

My research was conducted in a mixture of languages. As I explain in greater detail in the

following chapter, four languages were spoken in Markakasa on a daily basis:
Chhattisgarhi, Gondi, Marathi and Hindi. Of these, I had reasonable fluency in Marathi
when I first arrived in the village. Though the village is located in Maharashtra state,
Marathi is certainly not the most common language spoken there, the lingua franca being
Chhattisgarhi, a dialect of Hindi. At first all my conversations were carried out in
Marathi, and though my proficiency in that language increased greatly I was restricted in
the range of people I could speak to: most people understood Marathi but they were
unable to conduct detailed conversations in it, feeling more comfortable in Chhattisgarhi.
I picked up the latter fairly quickly and began to talk a strange mixture of Marathi and
Chhattisgarhi. I only ever had a basic knowledge of Gondi, a Dravidian language spoken
mainly by Gonds, and certainly could not follow detailed conversations.12

Outline of chapters

Chapter Two introduces the village of Markakasa where I carried out fieldwork and
provides a brief sociology of caste, livelihoods, space and religious affiliation.

12 Gondi was sometimes used in conversations specifically to hide what was being said from me.

24
Chapter Three begins a discussion of relations between villagers by looking in detail at
how and why people form ritual friendships with one another and introduces the
importance of witchcraft. It argues that ritual friendship is a way of extending sociality
beyond one’s immediate household in a manner that is safe and free from the risk of
magical attack or witchcraft. In this respect it is contrasted with relations between agnatic
kin and brothers in particular, relationships which are seen as particularly prone to
malicious spiritual attack. What is emphasised in the comparison is the essentially
disinterested nature of ritual friendship contrasting with the interested relationship of
brotherhood. The existence of cross-caste ritual friendship also reminds us that locality-
based forms of belonging are important in qualifying the difference implied by caste, by
the recognition of similarity between ritual friends.

Witchcraft and sorcery are important phenomena among the people of Markakasa.
Chapter Four looks at witchcraft in more detail, but in contrast to much of the work on
witchcraft in India, which focus largely on witchcraft accusations, I examine here the
impediments that have developed over the past fifty years to dealing publicly with this
sort of mystical attack, both for individuals and for the village as a whole. I suggest these
impediments have developed in this part of Adivasi central India for the following
reasons. Firstly, migration to another village in an attempt to escape witchcraft or
unexplained misfortune has become more difficult, partly as a result of changes in
legislation, which have reduced the supply of alienable land in Adivasi areas. Thus,
whereas in earlier times people could have ‘solved’ the problem of malicious attack by
migrating, nowadays, short of moving permanently to the city, which is generally seen as
undesirable, people who suffer such attacks cannot seriously consider escaping.
Secondly, earlier waves of migration to the area have changed the villagers’ perception of
the nature of their community: from being a village of one type of people, it has become a
‘mixed village’ (panchrangi gaori) with different castes,' languages, and ways of doing
things. This has meant that there is a lack of consensus as to the value or effectiveness of
certain village-wide anti-witchcraft measures, such as the clan-god Angadev, that could
have been employed in the past. Similarly, on an individual level, due to differences

25
concerning ideas about sex and marriage, testing potential wives and daughters-in-law for
witchcraft has become less desirable as Adivasis try to avoid “looking bad” in front of
other castes. Thirdly, the increased engagement of the state in the area, and in particular
of the police, as a result of the Maoist insurgency, has made dealing with suspected
witches more problematic. The police bring with them an alternative vision of Indian
modernity, one which has no space for what they regard as anachronistic and
superstitious practices such as witch-testing and ghost-finding. This has come to be
expressed in their opposition to the use of the Angadev, which to the police (and other
local administrators) is symptomatic of the failure of Adivasis and the Adivasi
‘community’ to develop or improve, despite the wide range of government policies in
place to ameliorate their condition.

The cumulative effect of these impediments has led to the increasing popularity of a
particular Hindu religious sect, the Mahanubhav Panth, which has gained a reputation
locally for healing people affected by malicious spiritual attack and for providing
permanent protection. In Chapter Five I examine how people are healed through ritual at
the sect’s temple and the resulting transformation from suffering to happiness they
undergo. I also demonstrate how membership of the sect affects one’s relations with kin,
neighbours and caste-fellows, particularly how there is an attempt to limit sociality (with
both divinity and man), seen as the cause of much misfortune in the past. In addition, the
increased incidence of religious meetings (satsang) in the area indicates that a clearer
sense of sect belonging is being articulated. Pilgrimage back to the temple where one was
cured is an important part of devotion but it also creates a sacred geography proper to
Maharashtra and an engagement with its language, through which identification with the
nation in religious terms is made. Widening the discussion somewhat I also look at how,
within the past year, a couple of households have converted to Christianity for the same
reasons as those that join the Mahanubhav sect: to cure intractable illness caused by
witchcraft.

In Chapter Six I look at the consequences of sect membership in terms of changed dietary
practice - namely vegetarianism and teetotalism - and how the power of the

26
transformation from suffering to happiness makes these practices meaningful for sect
members. I examine how it is a concern for protection from further harm which drives
adherence to these practices rather than the conventional explanations of social mobility
or Sanskritisation. But it is also clear that these practices become part of a more general
way of thinking about ‘wild’ (jangli) and ‘improved’ (sudharlela) which finds resonance
not only with local VVKA activities to discourage meat-eating and drinking, but also
with the sermons of sadhus or ‘world-renouncers’ associated with the Hindu nationalists
who visit the area from time to time. I suggest that the message of these politico-religious
Hindu nationalisms is appealing to those who have embraced the practices advocated for
different reasons.

27
Plate 1: A view o f Markakasa. Note the blue Gond tombs.

Plate 2: A village street. The road was widened and rebuilt by the Border Roads
Organisation (BRO).

28
Plate 3: A large Markakasa house, home to three households. The upper floors
are used to store grain.

Plate 4: A village street.

29
 Chapter Two

The Setting

As one drives east on the road to Calcutta from Nagpur, central India’s largest
city and winter capital of the state of Maharashtra, the busy plains watered by the
Wainganga River give way after the small town of Sakoli to more hilly terrain, and the
air cools slightly as the forest cover increases. Signs appear on the roadside asking
drivers to be careful as the highway rises and curves round and falls. The tall hills of the
Navegaon National Park loom in the distance. Spied from the road, blue tombs topped
with small statues of horses begin to appear in the fields and groves and one knows one
has entered Gond country. After two sets of hills, one turns off the national highway at
the small town of Deori and heads forty kilometres south to the village of Chichgarh
nestled on a plateau surrounded by dense forest. Following the road which skirts the
village, one passes a Ram temple on a hill before plunging back into the forest. One
emerges into a landscape of rice fields, the forest having receded to the north, and
Markakasa announces itself with a set of three blue tombs by the side of a lake
surrounded by mango trees.

Markakasa is a small settlement in central India at the eastern fringes of the state of
Maharashtra, in the borderlands with the new state of Chhattisgarh. Located in Gondia
District, the official administrative village (gram) is actually composed of two hamlets
(tola), Markakasa and Herpar, though in local imagination they are regarded as two
separate villages (gaon). The former is much larger, and this is where I lived and did
fieldwork. Markakasa is surrounded on one side by forest and on the other by paddy
fields, and has a population of 626 evenly split between the sexes and distributed among
122 households. Approximately 120 of its inhabitants are below the age of sixteen. Gram
(administrative villages) in this area tend to be small, having populations of no more than
eight hundred, and many are composed of several hamlets (tola), often regarded as
villages (gaon) in their own right. Distances between one village (gaon) and the next are

30
not great, three or four kilometres at the most. The area is overwhelmingly rural; the
nearest town is Deori, the site of the sub-district administration, which has a population
of 10,000 and is forty kilometres away.

In colonial times, Markakasa was part of the Chichgarh zamindari (estate) in Bhandara
District13 in the Central Provinces. The District Gazetteer of 1908 describes the area thus:

“This estate is situated in the south east comer of the Sakoli tahsil
bordering on the Nandgaon State and the Chanda District. Its area is 240
sq. miles, a large proportion of which consists of unculturable [sic] hill
country covered by forest” (Russell 1908: 196-7).

The village of Chichgarh itself had a population of about 500 in 1908 and the area
surrounding it was one of least attractive in the district due to the density of the forest.
Revenue was comparatively low. As the Gazetteer continues,

“the worst country is that around Kaoliwara, Nishani and Mahsuli and
again between Rajoli and Chichgarh; this consists almost entirely of hill
and forest” (ibid: 225).

The zamindari was held by a family of the Halbi14 caste who had received their sansad
(grant) from the Gond kings of Chanda. The family still lives in the area, in a village
outside Chichgarh. Heavily indebted after having mortgaged much of his land, the scion
has political aspirations and has campaigned for several years to be allowed to contest the
Maharashtra Assembly seat for the Congress Party. In colonial times, the administration
of the zamindari was regulated by the Wajib-ul-arz or Record of Rights which stipulated
that there was to be a manager (diwan) for the estate. This Diwan was hereditary and the
post occupied by a Marathi Kunbi family. The descendants of the Diwan live and hold
land in the neighbouring village of Kedanar. The Zamindar, who was referred to as the

13 Bhandara District was split into two in 1999 - the easternmost portion became the new Gondia District.
14 A caste categorized as a Scheduled Tribe. There are very few members o f the Halbi caste in the area.

31
Raja by Markakasa people, was permitted to take free labour (begar) of “four days work
in the year from each plough with ploughman and bullocks for his cultivation in the same
village and of one days work of a man from each house for thatching his houses” (Russell
1908: 159).

The village was first settled by a man called Burri, who left his natal village in what is
now Chhattisgarh, and Burri’s descendants became the Thakurs (hereditary headmen) of
Markakasa. The current Thakur, Drukh Singh, was bom in 1936, and Burri is his great-
great-grandfather (FFFF). Markakasa was most likely founded by Burri between 1860
and 1880. It seems that for many years the village remained small, consisting of no more
than two or three households at the edge of dense forest. Over the past sixty years
however, the village has grown rather dramatically, with many migrants arriving and
buying up farmland. Migration has slowed of late, partly because of the enactment of a
law in 1974 restricting the alienation of Adivasi-owned land.

Markakasa has two parallel modes of governance: the gram panchayat and the sian
panch. The gram panchayat is the lowest tier of government and is a body consisting of
seven members, each elected by one of the wards of the gram. It is headed by a sarpanch
who as primus inter pares is invested with more authority and powers than the individual
panchayat members. While I was living in Markakasa, the post of sarpanch was reserved
for a Scheduled Tribe woman and occupied by a Gond widow from one of the more
powerful families in the village. In the past the post of sarpanch has been reserved for
Scheduled Castes (SCs, former Untouchables), Other Backward Castes (OBCs) or left
open. Some of the wards of the gram are also reserved (by caste and gender). These are
reviewed every four years by the district election authorities, and the reservations rotated
between the various categories of people. At the time of fieldwork, the gram panchayat
consisted of five men (one Gaita Gond, one Mahar, one Dessau Gond, two Telis) and two
women (one Gaita Gond, one Mahar). The gram panchayat is responsible for
administering local government works projects, allocating communal land, collecting
taxes and charges and recording births, deaths and marriages.

32
The sian panch (‘council of elders’)15 is an informally composed body of village elders,
though not all village ‘elders’ are actually old. Some men in their thirties and forties are
regarded as sian. And some elderly men are not ‘village sian' and do not attend these
meetings. The sian panch organises village festivals and rituals, but its main function is
resolving disputes of all kinds between villagers. It is a multi-caste body (both ‘clean’
and Untouchable) but its members have authority as individual men and not as
representatives of their caste. Even if there is a dispute between members of a single caste
(e.g. a husband and wife, or two brothers), a multi-caste panch will meet to interview the
parties and facilitate a solution. As such, the panch is not dominated by any one caste,
and two of the most respected sian are members of the Untouchable Mahar caste. The
panch is acephalous: all members are equal and decisions are taken by consensus. The
Thakur’s presence (or that of one of his sons or brothers) is generally desirable,
especially when arranging village rituals, but his word does not carry any more weight
than that of the next man. The panch generally consists of men, but the village sarpanch,
who was a woman at the time of fieldwork, was often asked to attend its meetings too.

In many respects these two bodies of village governance operate independently of each
other. Though the sarpanch and certain gram panchayat members also attend sian panch
meetings, they do so in their individual capacity as sian and they are not regarded as any
more important than other sian. There seems to be a fairly strict separation of powers and
domains. While both the gram panchayat and the sian panch employ villagers as
assistants (chaprasi), their roles are clearly demarcated too. Thus, while the gram
panchayat chaprasi draws a government salary and records the death of villagers, it is the
gaonti chaprasi, paid in grain by his fellow villagers and directed by the sian panch, who
is the town-crier informing the village of a funeral.

15 Linguistically, compare sian with the Latin senex (old man; elder) from which we derive Senate, senatus
or ‘Council o f Elders’.

33
Plate 5: Village children playing kabbadi in preparation for an inter-school
tournament.

Plate 6: A Gaita Gond sian.

34
Space, caste, and religious affiliation

Markakasa is a mixed village, a panchrangi gaon (‘five-coloured village’) in local

parlance, and consists of twelve castes16. Most of the villages in the area do contain such
a mixture of castes but in many, Gaita Gonds tend to be in an absolute majority.
Markakasa is interesting in that respect because the three largest castes in the village
(Gaita Gond, Teli, Mahar) have roughly the same strength in numbers (see figure 1).
Though the Gond Rawat cannot marry Gaita Gond, in Markakasa they are co-opted at
life-cycle events as members of that caste. The two Untouchable castes are the Mahar and
Mochi.

Figure 1

Caste (Traditional Households Language Government

Occupation)17 Category
Gaita Gond 29 Gondi Scheduled Tribe
(farmers)
Mahar 29 Marathi Scheduled Caste
Teli 25 Chhattisgarhi Other Backward
(oil-pressers) Caste
Dessau Gond 11 Chhattisgarhi Scheduled Tribe
(farmers)
Rawat 9 Chhattisgarhi Other Backward
(cow herders) Caste
Sidar . 6 Chhattisgarhi Scheduled Tribe
(farmers)
Mochi 4 Chhattisgarhi Scheduled Caste
(cobblers)

16 The sole member o f the Rajput caste is not usually counted in this reckoning: she is seen as a temporary
resident.
17 According to Markakasa people.

35
Kevat 3 Chhattisgarhi Nomadic Tribe
(fishermen)

Gond-Rawat 2 Gondi Scheduled Tribe

(cow herders)
Lohar 2 Chhattisgarhi Other Backward
(blacksmith) Caste

Nai (barber) 1 Chhattisgarhi Other Backward

Caste
Rajput18 1 Marathi Forward

Kallar (distiller) 1 Marathi Other Backward

Caste

There is currently no clear spatial division of the village by caste. All castes, Untouchable
and clean, live mixed up and side-by-side and there are no parts of the village which are
associated with any particular caste. This is unusual further to the east in plains
Chhattisgarh but was a common feature of villages around Markakasa. It has not always
been the case however, and though the village seems seamless to an untrained eye there
are at least five different neighbourhoods (para). There is a part of the village which is
still referred to as mochi para (the cobbler’s quarter) and though it now has a mixed
population of clean and Untouchable, there was a time about sixty years ago when the
shoemakers and tanners would have lived separately from the rest of the village. The
oldest part of the village is the area around the Thakur’s house, thakur para, which only
has two castes: overwhelmingly Gaita Gond along with the descendants of the kotwal
(village watchman), the first Mahar to settle in the village. Even here Mahar houses are
found among a cluster of Gond ones. As the village grows, however, residence along
caste lines may be becoming more salient. Members of the Mahar caste for instance have

18 This is the village kindergarten teacher (anganwadi masteriri), who has lived in Markakasa for eleven
years.

36
built new houses in one particular area of the village, which is now called 'SC para19’.
But even this is in no way apart from the rest of village, and nor does it mean that other
Mahars do not continue to live alongside ‘clean’ castes. Locality and proximity are
important even within a village of six hundred people, and members of each para
(regardless of caste) have strong senses of belonging and prejudices about the other
paras20.

There is no clear hierarchy of castes in Markakasa and people of all castes only make a
clean-Untouchable distinction. The dominant caste both in terms of numbers and land-
holding is the Gaita subdivision of the Gonds, though it is not necessarily seen as the
most pure or prestigious: Brahmins, for instance, used to refuse to perform rituals for
them, while at the same time servicing the Teli caste21. Among the Untouchable castes,
Mahars rank higher than Mochi. Among the clean castes, there is very little interest in
ranking one caste higher or lower than another. Castes which straddle the boundary
between clean and Untouchable are the service castes of Blacksmith (lohar) and Barber
(nai).

There are no parts of the village public space which are off-limits to the Untouchable
castes. All use the village teashops and wells, though this did not used to be the case.
Even as recently as the late 1980s, Mahars and Mochi were handed water from the well
and were not permitted to touch it. But people come and go from one another’s houses
without regard to caste and close friendships develop across the clean-Untouchable
divide (see Chapter Three).

The ethnography presented in the chapters which follow is not of ‘a people’ (Gond
people; Adivasi people) but of a village in an Adivasi area and of the people who live in
it and identify themselves as ‘Markakasa people.’ Nevertheless, Markakasa is a village of
many castes and caste is an important source of identity. What follows is a brief

19 ‘SC’ referring to Scheduled Castes, the administrative term for Untouchable castes.
20 Froerer (2006b) makes the same point in her discussion o f friendship among children in central India.
21 This in any case has little effect in practice since Gaita Gonds never express the intention to employ
Brahmins.

37
presentation of the three main castes in the village: Gaita Gond, Mahar, and Teli, and an
exposition of changing religious affiliation.

Gaita G onS2

As mentioned above, the Gaita Gond caste is the dominant caste in Markakasa in terms
both of numbers and land-holding, a pattern which is repeated in the majority of villages
in the area. From discussions with various Gaita Gonds, it seems that the areas in which
they live in greatest concentration are southern Gondia District, northern Gadchiroli
District, and Rajnandgaon and Kanker Districts in Chhattisgarh. Beyond these areas, I
was told, one would encounter other types of Gond (Raj Gond, Maria Gond) with whom
marriage was not desirable. Gaita Gonds claim Gondi as their ‘mother-tongue’
([matrubasha) and distinguish themselves from the other sub-caste of Gonds in the village
- the Dessau Gond - by asserting that the latter’s mother tongue is Chhattisgarhi. Though
some of the Dessau denied that this is what made them different, they, unlike the Gaita,
spoke Chhattisgarhi at home and their life-cycle rituals were more akin to those of other
Chhattisgarhi castes than those of the Gaita. In addition, whereas the Dessau Gonds
divide into four exogamous phratries, Gaita Gonds have only two - saha dev (six-god)
and sath dev (seven-god). Their kinship world is thus fairly simple and is divided
between bhau (brothers) and sorye (affines). Gaita Gond kinship terminology is almost
identical to that collected by Simeran Gell for the Muria Gonds of Bastar (see Gell 1992:
43). No doubt influenced by the politics of reservations (positive discrimination), many
Gaita Gonds like to claim that they are the ‘real’ or ‘pure’ Adivasis as opposed to the
other Scheduled Tribe castes in the area.

22 Gond sub-castes have been the subject o f various anthropological studies throughout the last century: e.g.
A. Gell 1980,1982, 1986, 1997; S. Gell 1992; Furer-Haimendorf 1948, 1979; Elwin 1943,1947; Grigson
1938.1 am not aware o f any work done among the Gaita subcaste.

38
Mahar

The Mahars are the largest Untouchable caste in the village, and the most widespread in
the region. Indeed they are found right across the state of Maharashtra and constitute
over a third of the state’s Untouchable population. Karve notes a Marathi saying, “jithe
gaon, tithe maharwada”, translated as “wherever there’s a village, there are Mahars”
(cited in Zelliot 1996: 87). Equivalent to the English proverb, “there’s a black sheep in
every flock”, it expresses their ubiquity and extremely low social status within the local
hierarchy.

In common with several other north Indian Untouchable castes, the Mahars’ traditional
role was that of village servant (kotwal) who acted as watchman, messenger and, most
importantly in the context of his untouchability, the removal and disposal of dead cattle.
In return for these services the Mahar kotwal received an annual grant of grain and a
small amount of land. In Markakasa, the first Mahar family to settle in the village became
kotwals and they are still referred to as the “kotwal household” (kotwal ghar). They are
one of the wealthiest families in the village, with fourteen acres of good quality land.
After 1960, the kotwal ceased to be paid in grain and became a government employee
drawing a salary. In 1985, the functions were taken over by the gram chaprasi, and the
kotwal became the village land registrar’s assistant.

The Mahars of Markakasa identify closely with India’s most celebrated Untouchable
leader, Bhimrao Ambedkar. Ambedkar was a Mahar, educated in India and overseas, who
from the 1930s exhorted his caste-fellows to protest against their status. In contrast to
Gandhi he argued that no reform of Hindu religion could eliminate Untouchability and in
October 1956 in Nagpur he converted to Buddhism. Mahars in and around Markakasa
followed suit a few years later. The conversion was talked about in terms of a changed
religiosity as well as a more general change in a mode of life. For example, the traditional

23 While the same title is used in Chhattisgarh, it denotes a different office. In Maharashtra, the kotwal was
and remains an assistant, nowadays to the village land registrar/accountant (talathi, patwari). In
Chhattisgarh, the kotwal is an official with important local powers, akin to the position o f gram sevak in
Maharashtra.

39
Mahar occupation of goat-herding was abandoned following the conversion because
Ambedkar asked Mahars not to pursue demeaning occupations. All Mahar homes in
Markakasa contain images of Ambedkar and the Buddha and it is to these that worship is
offered. All but three of the Mahar households in the village describe themselves as of the
Buddhist religion (bouddha dharma): two households have converted to Christianity and
one to the Mahanubhav sect.

The Mahar conversion to Buddhism has not changed their position as Untouchable in the
village or local hierarchy, although, as mentioned above, their access to water and other
facilities has changed. Nevertheless the conversion and association with Ambedkarite
politics and social action has transformed the Mahars perception of the caste system and
they rail against the unequal treatment they sometimes receive. While I was living in
Markakasa, several village meetings were held at the instigation of the Mahars to discuss
what they perceived as slights by the clean castes. The discussions focused on
commensality, which I discuss a little later in the Chapter. Though in many respects the
discrimination is blatant, there is not in Markakasa that extreme division between clean
and Untouchable one finds in other parts of rural India.

Despite the poverty of most Mahars in the village, they are aware of the progress made
by some of their caste-fellows, especially in the field of education. It has been estimated
that, while they comprise only 35% of the Scheduled Caste population of Maharashtra,
Mahars hold up to 60% of public sector jobs reserved for Untouchables (Deliege 1999:
197). This perceived success is commented on by both Mahars and people of other castes
in the village. The first and only teacher produced by the village is a Mahar man (from an
exceptionally poor household) and, though he does not now live in Markakasa, his name
is mentioned frequently when discussing the relative success of the caste in taking up
opportunities offered by the system of reservations. Similarly, the first girl in the village
to pass her Standard XII24 examinations (which happened in 2003) is a Mahar. Outsiders
too provide examples of Mahars ‘getting on’: two of the four teachers at Markakasa
Primary School are Mahars from other parts of Vidarbha.

24 Roughly equivalent to English A-levels.

40
Teli

The Telis in Markakasa are all Chhattisgarhi-speaking and most trace their origin to
various villages in what is now Chhattisgarh. Though traditionally oil-pressers, they have
all abandoned that occupation and now see themselves primarily as farmers. The first Teli
family to move to the village came from the neighbouring hamlet of Herpar. They left
following a series of thefts from their house which they attributed to a ghost (bhut) and so
decided to set up home alongside the handful of Gond families in Markakasa. Sometime
in the 1940s, a trio of Teli brothers arrived fleeing scarcity and poverty in the
Chhattisgarh plains and, as they had no land, began making roof tiles which they sold in
the larger and more prosperous villages of Chichgarh and Kakodi.25 Their respective sons
are now prominent Telis in the village.

The Teli caste is closely associated with the Kabir sect and the membership of the latter
in the village is exclusively Teli. The three brothers and their families were the first to
join the Kabir sect some sixty years ago, after they had moved to Markakasa. The reasons
for this conversion were not clear and were rather contradictory: some talked about their
forefathers’ desire to become closer to God through giving up sin; others mentioned
illness and the power of prayer to heal. One can only speculate. One reason might have
been to differentiate themselves from their new Gond neighbours, whom they might have
regarded as jangli (wild) and ‘backward’. Certainly there was a considerable amount of
tension.

People of all castes would tell me of a time about fifty years ago when the Telis and
Mahars came into conflict with the Gonds over the performance of village rituals. This
must have been at about the same time that both the Telis were beginning their
involvement with the Kabir sect, and the Mahars had converted to Buddhism. The Telis

25 Roof tiles began to be commonly used by Markakasa people themselves sometime in the mid 1980s. Up
until then most houses were covered with grass. The scarcity o f good grass in the forests around the village
and the prestige associated with tiled roofs caused the shift, according to Markakasa people.

41
and Mahars protested the use of animal sacrifice at collective rituals such as Sanjori ,
arguing that it was unnecessary and ‘backward’, and that the Gonds ought to ‘improve’
(sudharna).27 The Gonds refused to accept their demands, and for a number of years only
the Gonds performed collective village ritual. After several village meetings, involving
people from other villages too, the villagers agreed to stop animal sacrifice in order to
enable the participation of all. As we shall see in Chapter Six, there has been a return to
animal sacrifice recently, perhaps as a result of the recognition by non-sectarian, non-
Buddhist villagers that the search for unity of purpose, expressed through collective
ritual, is no longer attainable given the diversity of religious affiliation.

Markakasa is a village of multiple sectarian formations. The Kabir sect was the first
sustained appearance of a different form of religiosity but was restricted in its appeal to
the Teli population. The Mahars too formed a ‘sect’ of their own by becoming Buddhists,
most no longer participating in certain village rituals. More recently, as this thesis will
explore, significant numbers of people of all castes in Markakasa have joined the
Mahanubhav sect, and they too have refused to participate in village ritual. In numerical
terms, these three ‘sects’ taken together constitute nearly half the households in the
village. That half the village is reluctant to take part in activities which are seen by the
other half to promote the well-being of the village as a whole is matter of concern for this
latter group. These problems of unity of purpose and attitude will be discussed in greater
detail in Chapters Four, Five and Six.

Commensality and Untouchability

The brief exposition of some the main castes in the village naturally raises questions of
commensality: who eats with whom, when and where? While one would expect to see a
clean/ Untouchable caste distinction concerning this form of sociality, things are not
really all that simple. Public and private events have different rules regarding eating

26 A ritual conducted before the threshing o f the paddy can begin. There are heavy penalties for those who
thresh before the ritual is performed.
27 O f course, this is how the past was presented to me in the present. I have no way o f knowing exactly
what happened.

42
together. It seems that an event classed as ‘public’ (sarvajanik) adheres to a strict equality
between clean and Untouchables castes: no distinction is made and all sit together in an
unbroken line (pangat). At an event which is ‘individual’ (vekthigat) and private,
however, no rules of any kind apply; indeed it seems it is up to the man hosting the meal
whether Untouchables are sat apart from clean castes or with them. Tensions have grown
in Markakasa precisely because of this division of domains. There is disagreement
between the Mahars and the other castes, in particular, over whether a village wedding
can be considered a public or a private event.28

Let me provide some examples of sarvajanik and vekthigat situations to help illustrate the
distinction. Sarvajanik (public) events include collective village rituals centred on the
worship of the tutelary deity Thakurdev. At some rituals a meal of rice is cooked in the
sacred grove and the attendant men of all castes, including Untouchable, eat the rice side-
by-side. Similarly, the members of a multi-caste (again both clean and Untouchable
castes) committee, which was organising the annual bullock-cart races, came to my
landlord’s house where they were served food and, again, all members sat together mixed
up in an unbroken line.

Vekthigat (personal) events include gatherings at individual households. For instance, the
wife of a Gond man died and he held a number of feasts in the days following the funeral.
Most feasts were attended by large numbers of kin and caste-fellows. For the last feast
however he invited only those villagers to whom he felt close. Among these was an old
friend of his, a Mahar man, who sat amid the other men and ate with them. No distinction
was made. As I examine in Chapter Three, affection somewhat mitigates caste and
Untouchability. Likewise, a young Mahar man, Munya, who had worked in my
landlord’s fields to bring in the harvest alongside men of clean caste, was invited to the
house to eat a special feast called barona. He too was seated in the middle of the group.
Another time, Munya had come to the house and was chatting with us all and was asked
to stay for dinner. As it was an ordinary dinner we all took our meal in the kitchen,

28 In common with the general Indian ethnography, women seem to be less catholic in their dining partners
than men, and so their definitions o f what is a public and what is a private event are in some respects
different to those o f men.

43
including him. Since it is the place where food is prepared, the kitchen is generally seen
as susceptible to pollution. Yet my landlord and his wife had no qualms about Munya’s
presence in this most sensitive of rooms. But nor did they regard him as ‘clean’: after
Munya left, my landlady Champa did not wash his dish or his cup and Brijlal’s unmarried
niece was asked to do it instead. Whether Munya knew this would happen was difficult to
determine and I certainly did not ask him for fear of offending him. What was clear was
that Brijlal and Champa hid from Munya that he was being treated differently, and this in
itself is relevant. It demonstrates that the practice of Untouchability is no longer really
legitimate, even in this ‘individual’ sphere. If one invites an Untouchable to eat with one,
one cannot ask him to sit apart.

The series of disputes between the Untouchables (principally Mahars) and the ‘clean’
castes concerned the issue of village weddings: were they sarvajanik or vekthigat events?

The tensions flared up at a Gond wedding in April 2003. It was a double wedding:
brother and sister were getting married at the same time in order to save money. As a
result, there were more guests than usual, with both sets of new affines (in addition to
their guests) present.29 The Mahars, who would have normally eaten at the wedding feast,
were told that there was not enough food for everyone and that they would have to be
content with a small amount of uncooked rice (chawal). The Mahars were furious. They
held a meeting and promptly returned the rice, something which had never happened
before. The Mahars I spoke to ridiculed the notion that there was not enough food to go
around and said that the only reason they were not welcome to eat was because the Gonds
regarded them as achhut (untouchable). Keeping the rice would have compounded the
insult. Relations between the two sides worsened, with the Mahars saying that they would
not now attend any Gond weddings, nor accept rice from them. After talking to some
Gonds, it became clear that the Mahars were right: the reason given of scarcity was just a
cover. The Gonds were uneasy about the prospect of the Mahars attending the wedding
feast because instead of sitting and eating separately, they feared the Mahars would

29 Gaita Gond weddings take place virilocally or uxorilocally: both are equally acceptable and there is no
particular prestige associated with uxorilocal marriage. Among other castes in Markakasa the rule is clear:
the groom travels to the bride’s village to be married in her house.

44
“come and sit in the middle with everyone else”. “How would that look to our guests
from other villages?” the Thakur asked me. He told me that their behaviour had changed:
in the past they knew their place. But he also repeated the initial assertion that there were
now too many mouths and the custom of feeding the whole village was becoming
financially impossible. When he was a boy (he is in his late sixties), there were only four
or five Mahar households, so feeding them was no problem at all. Now there are thirty,
he commented. The Thakur and others also mentioned a perceived lack of reciprocity as
another factor which affected the way they saw the issue. They complained that, though
Mahars eat at Gond weddings, Gonds cannot eat at Mahar ones. The Thakur gave the
example of the recent wedding of Vijay, a Mahar man. Vijay’s family had not made
separate dining arrangements for Gonds and others, so people found it impossible to
attend. From the Mahar point of view, Vijay and his family were unwilling to
accommodate an attitude they find deeply offensive.

The tensions rumbled on throughout the summer wedding season. Mahars were generally
not invited to Gond or Teli weddings, and matters deteriorated further when the Mahars
threatened to withhold contributions to the agricultural ritual of bali conducted before the
new rice of the year can be eaten. A large village meeting was held at the end of
September to discuss the problems between the Mahars and the ‘clean’ castes and to
attempt to resolve them. A senior Mahar man, Asaram, began the proceedings. He
explained to the other castes that the Mahars had noticed a change in attitude towards
them over the past year: they were no longer invited to communal feasts and generally
felt excluded by the Gond and Teli communities. “When people claim that there is not
enough food to go around why does that only apply to the Mahars?” he asked them. “If a
Teli household has enough food to feed the Gond caste, why does it not have enough
food to feed the Mahars?” He complained openly about the ‘special treatment’ meted out
to the Mahar community. They felt excluded, he said, and their only course of action was
to withhold their contribution to village ritual. By expressing themselves thus, the Mahars
were clearly stating their opinion that it was the right of every villager to attend a village
wedding: it was a sarvajanik event and they ought not to be excluded from it. And the

45
implication, which everyone understood, was that if they attended a public event such as
a wedding they would not tolerate being seated apart from the others.

The response of the ‘clean’ castes to this speech was mixed. Some people reiterated the
view that the Mahars were not being reciprocal themselves: they were demanding to eat
at Gond and Teli weddings but were not asking Gonds and Telis to eat at theirs. Others
were angry that the Mahars could even consider withholding their contributions to village
events in such a brazen fashion and suggested that all exchange (len-den) with the
Mahars ought to cease. One sian, a Teli man, suggested the following compromise: that
in cases where the household was poor it should only invite members of its own caste to
the wedding feast. If the household had means, then they ought to feed everyone
irrespective of caste. This was agreed upon by most present, Mahar, Gond and Teli alike,
as the most sensible way of resolving the matter. Curiously though, the issue which
seemed to lie at the heart of the problem - the fact of Mahar Untouchability - was not
dealt with head on, either by the Mahars or by the others.

A couple of the younger Mahars at the meeting sensed this too and raised the issue
directly. The essential problem, they said, was that the other castes thought in terms of
nich-uppar (low and high), and regarded them as achhut and asprushya (untouchable)
and so discriminated against them on that basis. This was not really taken up and
discussed by either the Mahars or the others, but speaking to one of the young Mahars
later on, he nonetheless felt a sense of victory in having expressed himself in this manner.

The situation was not really resolved however, and the compromise agreed to quickly
broke down during the wedding season the following year. The Gonds and Telis began to
argue that weddings were not sarvajanik (public) at all but vekthigat (personal), and one
could invite whomever one wanted depending on the affection one felt for them. It
became apparent that very few Mahars were receiving invitations.

46
Despite this breakdown in relations at the level of the caste, individuals still continued to
behave with one another as they had always done, particularly with their neighbours,
friends and, importantly, ritual friends.

So far I have only described situations where people of clean caste do or do not sit
alongside Untouchables while eating. Do people of clean caste, however, eat food
prepared by Untouchable hands? Here again the rules do not seem as rigid as one might
expect. At the annual village fair (mandat) for instance, when several households set up
stalls selling tea and snacks, Markakasa Mahar stall-holders had both ‘clean’ and
Untouchable clientele. Markakasa Gonds and others did not have to eat and drink at a
Mahar stall; other options existed. That they did so demonstrated their lack of concern
about pollution at this public event.

Commensality and Untouchability remain important areas of conflict in Markakasa. As

will be discussed in Chapter Three however, the institution of ritual friendship serves to
deny caste in many ways by permitting ritual friendships which cross the
clean/Untouchable divide to eat with one another without comment or penalty. As
commensality is a key element of sociality, the other transformations it is undergoing will
also be discussed in the chapters which follow.

Livelihoods

The land

The major livelihood here is rice agriculture and, since the land is not irrigated, the crop
is dependent on the annual rains. Consequently the village only produces one crop a year
and the agricultural season is the half-year from June to December. The soil is not
particularly fertile: it is unlike both the famous black soil of Nagpur used to grow cotton
and the red soil found to the east in the Chhattisgarh plains. Most of the farmland was
once forest or scrub, and this is given as a reason by the villagers for its relative lack of
productivity. No-one in the village owns a tractor, and rare is the household that hires one

47
from a neighbouring village for agricultural work. Ploughing is done by oxen and the
crop transported by cart. Sowing begins in late June following the first monsoon showers
moving up from the south-west. There are two methods for sowing used, depending on
the quality of the land, its position (high, low, sloping, even) and its ability to hold water.
The first, called perni, is a method whereby the seed is scattered randomly over a
ploughed field; the second is rohona where the seed is sown intensively in a very small
patch of land, allowed to grow green shoots and then transplanted in a regular fashion
over the field. The latter is initially more labour intensive, requires a good supply of
water and is ultimately more productive than perni, especially for the better quality grains
such as HMT. A field that has been sown using the perni method, however, requires
weeding (nindai) more frequently and also needs an additional process called biashi (a
flattening of the shoots) in order to ensure good growth of the rice.

By mid October, the lower quality grains have ripened and harvesting begins in earnest,
usually before the festival of Diwali so that some rice can be sold to finance the purchase
of new clothes to celebrate in style. After harvesting, the rice needs to be threshed and
winnowed before it is packed into sacks or stored in earthen bins ready for husking ,or
selling. Threshing cannot begin until the village collectively performs the Sanjori ritual.
The threshing and winnowing is also done without the aid of tractors or machines. An ox­
cart is driven round and round over the stalks in order to dislodge the grain. Then the
grains are gathered up in wicker trays (supa) and dropped in front of hand-driven fans to
separate out the chaff.

There are two main buyers in the area: the Co-operative society [English word used] run
by the state-financed Tribal Development Council (Adivasi Vikas Mahamandal), and
local shopkeepers. Though no private shop can buy rice while the society is open, this
rule is not observed by farmers, shopkeepers or local officials. The shopkeepers offer
lower rates than the society, but payment is received almost immediately. This is not the
case when one sells to the society. The society lends money to farmers for the purchase of
crop or fertiliser and these loans are taken into account when paying for the rice grain.

48
Thus farmers may be forced to pay off their debts and not receive much cash in return.
Shops are therefore always attractive as ready sources of cash.

Cash is most definitely in short supply. Very few people have savings of any
consequence, and though some people have bank accounts their use is restricted. Most
farmers store their wealth as grain and sell throughout the year as and when cash is
needed.

Those who have basic irrigation such as a well or proximity to a lake can grow a second
crop after the rice has been harvested. Paddy is not sown again (the water required far
exceeds what is available). Instead farmers grow chickpeas, chillies, or vegetables (or a
combination of all three), which they then sell at market or to other villagers.

There are large disparities in the size and quality of land-holdings, though there are only a
handful of households who are technically landless. Others have such small holdings that
they often talk as if they had no land at all. Predictably, the largest single landowner in
the village is the Thakur. He has approximately twenty acres of excellent land and his
two brothers each have a slightly smaller share. See Figure 2 below for size of
landholding per household by caste.30

30 Does not include all the households in the village.

49
Figure 2: Landholding (acres) per household by caste in Markakasa

No 1-2 2-4 4-6 6+ Total

land31
Gaita Gond 0 3 5 7 10 25

Mahar 3 13 8 2 1 27

Teli 6 7 2 5 5 25

Dessau 0 3 2 1 0 6
Gond

Rawat 1 0 3 3 0 7

Gond 0 0 2 0 0 2
Rawat

Kevat 1 0 1 1 0 3

Sidar 1 0 2 • 1 1 5

Rajput 1 0 0 0 0 1

Kallar 1 0 0 0 0 1

Nai 0 0 0 0 1 1

Lohar 1 1 0 0 0 2

Mochi 0 1 1 0 2 4

Total 15 28 26 20 20 109

The forest

People whose ancestors had settled in Markakasa said they came to the area because of
the forest: there would always be plentiful firewood, and timber to build houses, and the

31 Also includes those who have less than one acre.

50
sale of bamboo was lucrative. The forest was cut down to make farmland, and one could
cut as much as one wanted. In times of famine, one gathered roots, tubers, mushrooms,
bamboo shoots, and mahua32 flowers from the forest, or went hunting for meat.
Nowadays there is very little to hunt and Markakasa people only go on expeditions when
invited to by people in other villages with more game. The mushrooms and bamboo
shoots that grow following the onset of the monsoon are much prized, however, and men
and women alike will grab a basket in the morning and head to the forest in order to pick
them.

The forest also provides a great deal of freedom from the rhythms of village life. When
the agricultural season is over, groups of women often make day-long trips to the forest
to collect grass to make brooms and reeds which are woven into mats. They also go in
search of huge quantities of certain types of leaves which are made into plates. These
come in handy for wedding feasts which take place in the summer. The women are often
full of stories on their return home, having spent a lot of time gossiping and exchanging
information about neighbours and kinsmen.

A source of gossip to be sure but the forest is also a good source of money, especially in
the summer months between March and June when there is very little other work. A
major earner at this time is the collection of tendu leaves, which are used to make Indian
cigarettes (bidi). Households generally earn Rs. 1000-3000 per season. The money comes
at a critical time in the agricultural year, and is used to mend the dykes of rice-fields and
buy better quality grain and fertiliser in preparation for sowing. For those without land,
the tendu leaf money can be used to rent some for the coming agricultural season. The
Forest Department sells the tender of an area of forest to a contractor who then sets the
rate paid per bundle to villagers. The work is prepared for several weeks in advance:
women collect a particular grass which is almost bark-like in appearance and will be used
to bind bundles of leaves; similarly, once word reaches the village that a tender has been
awarded, everyone sets out to the forest, axe in hand, in order to cut down the old leaves
and encourage new growth. Once the tendu season proper begins, most households spend

32 Bassia latifolia. A flower also used to make alcohol.

51
all day on this activity. People set out very early, often at three or four o’clock in the
morning. This not only gives them time to reach the deepest parts of the forest where they
will encounter greater quantities of good leaves and less competition, but it also means
that they will return by late morning before the intense heat of a summer’s day can take
hold. After lunch, people begin bundling their leaves and then take their bundles to be
counted and registered by the munshi, a villager paid by the contractor to be their
representative.

Tendu collection has been disrupted in recent years with the intensification of the
Naxalite insurgency. The latter demand a cut from the contractors in order to allow the
villagers to collect the leaves without molestation. The amounts demanded in recent years
have been rising. The tendu season in the summer of 2004 was cancelled because the
contractor refused to pay the protection money. The Naxalite involvement in the tendu
business is long-standing and had been uncontroversial: villagers recount how the rebels
in the past forced the contractors to increase the rate paid to villagers. Now however, with
the withdrawal of the contractors, and the reluctance of the state to take over the contract
itself, the Naxalites are blamed by many Markakasa people for denying them an income.

Households are also busy in April collecting mahua flowers, which bloom and fall at this
time. The flower is used to make alcohol and is sold to local traders and distillers. It is
collected legally from one’s own trees, and illegally from State-owned forest. Men and
women head out to their trees in the morning, spend an hour or two collecting the flowers
which have fallen to the ground and return home where the flowers are spread out in the
courtyard or garden in order to be dried by the sun. Cattle find mahua flowers an
irresistible snack and people have to be vigilant if they are not to see their money
disappear into the belly of a cow. Income from mahua sales can be high, with most
households making between Rs. 500 and Rs. 1500 over the course of two or three weeks.

52
State-funded casual work

Work undertaken as part of the Maharashtra State Employment Guarantee Scheme (EGS,
or Rozgar Hami Yojna) is an important source of money too: labourers are paid a
standard rate of Rs. 47, some of which is paid in grain. The work includes building tanks
(small reservoirs), roads and other development schemes, primarily designed to provide
income during the lean months between March and June when grain bins are emptying
and, apart from the few weeks of the mahua and tendu seasons, there is little economic
activity. Work teams are often mixed caste affairs too and people generally choose to
work alongside their friends and immediate neighbours.

Having sketched rather briefly these various facets of Markakasa life, let us now take a
closer look at the social transformations underway and their implications.

53
Plate 7: Harvesting rice.

Plate 8: Harvesting rice.

54
Plate 9: Taking a break from harvesting.

Plate 10: Laying out bundles of tendu leaves (used to make Indian cigarettes,
bidi> to be counted by the munshi (secondfrom left).

55
 Chapter Three
Ritual Friendship

Towards the end of my fieldwork, on a hot summer’s day in May, I was sitting
chatting to Shamrao and his wife, enjoying the cool breeze on their verandah, when he
suggested that since Ganeshyam, the village shopkeeper, and I got on so well, we ought
to become ritual friends (mahaprasad). We clearly had prem (love; affection) for one
another, he said, and that was the most important thing in these mahaprasad
relationships. We were both at the same stage in our lives, young (jawan) and unmarried
(kunwara), so it would be a good match (Jodi) and make the other villagers happy. It
would also mean, he added, that I would have to return to Markakasa at some point. The
pull would be all the more powerful if I had a mahaprasad to visit and I would of course
have to attend his wedding. The best type of mahaprasad relationship was when the two
friends lived far apart, and how much further could Ganeshyam get than having a
mahaprasad in London? After all, what was the point, Shamrao continued, of having a
mahaprasad in the same village: the importance of visiting each other was lost and one
just ended up meeting at the teashop? It was not something I had given much thought to
before, but it seemed that people were discussing the prospect of my departure, debating
whether or not I would keep in contact with the village and were trying to find some way
of making me establish a permanent and meaningful link.

That evening after dinner, my landlord and I ventured out to Birbal’s paan shack to sit
and listen to the gossip. Ganeshyam was there, as was Shamrao, and on seeing me he
launched into an appeal to both of us to ‘form mahaprasad" (mahaprasad badna). He
repeated much of what he had said to me earlier that day and went on to add that not only
our prem, but the prem between our respective families, would increase. I was a regular
visitor to Ganeshyam’s house and knew his mother and brothers well, and he had
attended my cousin’s (MBS) wedding in Nagpur. Now, if he ever went to Nagpur again

56
he, as my mahaprasad, could always call on my mama (MB) and so the prem would
grow, Shamrao explained. The others sitting there thought it a great idea but the two of us
remained strangely non-committal. I, for one, felt uncomfortable agreeing to something
as seemingly important as this without discussing it properly with my mahaprasad-io-bQ
to gauge his feelings. Having investigated this form of ritual friendship over several
months, I was well aware that not all people approved of this particular social institution
and indeed not everyone who was ‘proposed’ to by a friend necessarily accepted.

A few days later, I was sitting out in front of Ganeshyam’s shop (one of the village
haunts) when my landlord came looking for me and asked whether we had decided to
become mahaprasad or not. Some of the village elders, including Ganeshyam’s mother,
had been discussing it and were keen on the idea. If we agreed, we could perform the
ceremony on the morning of the day I was scheduled to leave, he suggested. We had a
quick chat and Ganeshyam said that, although he was not really interested in “all this
mahaprasad stuff33”, the other villagers seemed to think it was a good idea and that we
ought to go ahead with it. And so, about a week later, the day I left Markakasa, witnessed
by almost the entire village, Ganeshyam and I became mahaprasad.

This chapter seeks to explore the structural role of the social institution of ritual
friendship, and examines both how people talk about such a relationship and how they
practice it. I hope to raise questions about how ritual friendship is opposed to ideologies
both of caste and of brotherhood. Regarding the latter in particular, I suggest that the
experience of disputes between brothers and the expectation that they will fall out with
one another, coupled with a heightened fear of the power of witchcraft and sorcery,
especially used within agnatic kin groups, leads people to see ritual friendship as a form
of association which is safe from dispute. As such, in ideological terms it is constructed
as founded purely on sentiment, unencumbered by material concerns and thus free from
the sorts of entanglements that relationships with agnatic kin tend to suffer from. In a
society characterised by a fear of malignant spiritual attack, one creates a relationship of
ritual friendship that is ‘like brothers’ but is ideally and ideologically disinterested.

33 Maza lakshaya goshti 'in var jaasta nai ye.

57
Indeed, whereas the relationship between brothers is based on reciprocity, the
relationship between ritual friends, one could argue, is ideologically based on that of the
gift given without expectation of a return. Sentiment, the notion of prem, brings this into
sharper focus.

The question of sentiment as the basis of a relationship is crucial since it opens up a

wider exploration of personhood, individualism and exchange. Carrier (1999) makes a
comparison between two broad categories of self, the western individual and the
Melanesian dividual, the latter being one where the self is constituted by the relationships
one has with others. Carrier claims that in societies where the latter type of self is in
evidence, ‘it is difficult to conceive of the spontaneous affection that emerges from
internal sources that characterise the western notion of friendship’ (ibid: 28). Instead, in
Melanesia, for instance ‘those who are close to each other are likely to speak about the
relationship...in terms of a common situation or structure of relationships that
encompasses the people involved’ (ibid: 31). The idea that a particular society produces
only one particular type of self is unappealing. It is far better to imagine several
ideologies of the self existing in any particular society and even different ideologies held
by people for different purposes. As I demonstrate below, an analysis of how people talk
about ritual friendship reveals that, in Markakasa, it involves two types of self (and thus
two ideologies of the self). In one, there is spontaneous affection; in the other, sentiment
is discussed as emerging from an existing set of relationships.

Taking the discussion back to ritual friends and brothers, one can also see that, whereas
the former is always seen in terms of sentiment and affection (and, as a corollary,
involving exchange which is not predicated upon calculation of a return), the latter is
characterised precisely by the give-and-take of daily life: sentiment is present but finds its
basis in existing kinship relations. Where sentiment is bound up in other spheres of social
life, as it is in the relationship between brothers, there is the risk of dispute over land or
other resources and jealousy has plenty of opportunity to rear its ugly head. And where
jealousy walks, the fear of witchcraft and poisoning surely follows. The ideology of ritual
friendship, on the other hand, is characterised by a lack of dispute or argument between

58
the friends. Through the emphasis on affection, ritual friendship in Markakasa becomes
in a sense abstracted from a social life grounded in materiality - especially if, as Shamrao
said, the best ritual friends are those who live far from one another.

Sentiment is seen as the basis of ritual friendship, founded in the affection two people
have for one another. It is not, however, restricted to these two people. Contrary to Pitt-
Rivers’ (1973) assertion that sentiment among friends is confined and therefore socially
unimportant, we can see here that the ties of affection that bind people together are also
conceptualised as bringing different families closer together. One could also argue that
that the sum of these individual ties of sentiment produces wider bonds between all
people in a community such as a village beyond their own personal relations of kinship
and friendship. For India in particular, others have looked at locality-based senses of
belonging (e.g. Lambert 1996; Froerer 2006b) and have argued that the traditional
emphasis on caste in India has relegated the importance of links between people who
share a common residence (e.g. a village or a neighbourhood). Taking this line of
reasoning further, I would argue that the existence of locality-based relationships enables
people sharing the same physical space (a village) to think of one another as essentially
the same, something that the ideology of caste works hard to negate. Sentiment is valued
here between individuals who are friends, but one could argue that the consequences of
sentiment have a wider effect on the imagination of a village community bound together
by the ties of locality (as opposed to caste or kinship-based relationships). The
importance of such an approach is that it recognises that an ideology which ignores caste,
as ritual friendship does, can exist alongside a contrary ideology which affirms caste.
Ritual friendship, the sentiment it involves and the ties of affection it builds beyond
individuals express the recognition of a fundamental affinity of people (see Rezende
1999) and the assertion of people as the same, an idea that caste denies.

Ritual friendship in Markakasa

Ritual friendship is rather common throughout the Chhattisgarhi cultural zone (Babb
1975; Jay 1973; Prakasam 1992), and though Markakasa is technically in Maharashtra it

59
is only twelve kilometres from the border. Due to migration the majority of people are, in
linguistic and cultural terms, recognisably Chhattisgarhi. These then are the borderlands:
just ten kilometres away to the west, deeper into Maharashtra, one is hard pressed to find
people who have formed these specific types of friendship. The generic Chhattisgarhi
term for ritual friendship is phul-phulwari, while the word mentioned above,
mahaprasad, is one of several types of phul-phulwari ?A The name given to the
relationship depends in most cases on the substance exchanged at the ceremony creating
the ritual friendship. Thus, in mine and Ganeshyam’s case, we exchanged prasad (ritual
gift of the deity) from the Jagannath Temple in Puri in Orissa, which is known as
mahaprasad. This then comes to signify the name of the relationship and also the title by
which one addresses the other. Taking one’s ritual friend’s name is not permitted, and a
small monetary fine or a coconut is imposed for transgressions. Other common
substances exchanged include ganga ja l (holy water from the River Ganges) and tulsi ja l
(water sprinkled into the mouth using leaves of Indian basil, a holy plant). Both men and
women can form these types of phul-phulwari, but the friendships must be between
members of the same sex. Sakhi is a type of ritual friendship that occurs between girls
only. There is also sanghi, which is used by men and is the title one gives to one’s
mama’s son (MBS). A sanghi is another form of ritual friend but whereas the
relationships created by mahaprasad, tulsi ja l and ganga ja l could be thought of as
‘parallel or agnatic friends’, a sanghi is technically the son of a man one calls mama
(MB) and so is a ‘cross’ category of friend. These last two, sakhi and sanghi, do not
require the exchange of any special substance. Of these five types of phul-phulwari,
mahaprasad is the most common, at least in Markakasa.

People become ritual friends for different reasons and in different circumstances. For
some, the friendship was already of long duration before the ceremony and they wanted
to formalise and publicise their prem (love; affection); for others, often those people who
had struck up an acquaintance based on working together in another village or town, the

34 Jay (1973) refers to the word mitan which he translates as ‘friend’ and which is used in the same way
Markakasa people use phul-phulwari; Parry (personal communication) too reports the use o f this word in
and around Bhilai. To my knowledge, Markakasa villagers never used mitan to refer to these specific ritual
friendships, and indeed the title mahaprasad was sometimes used, incorrectly, as a generic term.

60
ritual friendship marked the beginning of a more profound relationship. People who had
performed the ceremony in their youth or childhood often mentioned that they became
ritual friends because they used to walk to school together, play together or share food
from one another’s lunch boxes. For others it was because they had met working in the
same gang on a government work project building a road or reservoir. One young Mahar
man became tulsi ja l with a man from a neighbouring village because they both liked
racing bullock chariots and always ended up competing against each other. A Teli woman
told me that she became mahaprasad with a woman who was visiting relatives in
Markakasa because everyone told them they looked so similar and would make a good
pair.

For a number of people the prem came after the ceremony, since the friendships were
arranged by their fathers or grandfathers in order to keep the families close and to
strengthen the relationship (samanda vadna). In this situation the ritual friendship
between individuals is a symbol of a wider union. Pardhu, a middle-aged Gond man with
two mahaprasad relationships, told me that his first one was made when he was a child
of about six or seven years old with a Rawat (cow-herder) boy from a neighbouring
village. It was arranged by their parents to ensure that the relationship between the two
families would not break {samanda tutnar nai). He met his second mahaprasad while
they were both servants {naukar) at the village headman’s house and lived and worked
there together. They became friends and decided to become mahaprasad. Interestingly,
Pardhu’s younger son has become ritual friends with the son of his father’s mahaprasad.
Though the sentiment for the two ritual friends in cases of ‘arranged friendships’35 seems
to be absent, what is in fact being affirmed and continued is the original sentiment that
caused the old ritual friendship to be made.

That the children of ritual friends can themselves become ritual friends suggests a tension
in the elaboration of the relationship as one of kinship36. Brothers or other close relatives

35 This is my category, not one that Markakasa people used.

36 In his study o f a Satnamis in rural Chhattisgarh, Prakasam finds that the children o f mitan regard each
other as brothers and sisters and cannot become mitan to one another (1992: 202-3). This is clearly not the
case for the people among whom I carried out fieldwork.

61
cannot become mahaprasad to one another, and since mahaprasad is conceptualised as
being ‘like brotherhood’, it would follow that their respective children are also to be
regarded as agnatic kin, and barred from forming ritual friendships with one another (cf.
Prakasam’s data in footnote 4). My data demonstrate that existing ritual friends do
arrange the friendships of their children and in doing so they implicitly deny that they are
‘real kin’ or ‘real brothers’. It suggests that in reality what is being created in ritual
friendship is a relationship that looks very much like agnatic kinship or brotherhood but
is in fact something rather different.

Whereas caste is relevant in much thinking in other spheres in village life, people are
emphatic that it is not a consideration when choosing a ritual friend. Even to suggest it in
connection with phul-phulwari seemed distasteful to those I spoke to. Those who in other
conversations and contexts were the most disparaging of the village Untouchables
(Mahars and Chamars) were clear that phul-phulwari was blind to caste. Most ritual
friendships are formed out of caste, unlike marriage of course which is (almost) always,
and ideally, contracted within caste. Jay (1973) claimed that ritual friendships are not
made between people, of the same caste living in the same village, and are thus clearly
about bringing different castes together. I found no such rule, either stated or observed, in
Markakasa. Whether this is due to change in rules over time or a question of regional
variation is difficult to discern, though Markakasa people certainly did not regard ritual
friendship as something changeable. Of the forty-five relationships within the village,
thirty-six were contracted with members of a different caste and nine were contracted
within caste. Of the seventy relationships that Markakasa people contracted with people
from other villages, thirty-eight were with different castes and seventeen within caste37.
Ritual friendships also crossed the clean-Untouchable divide: of the thirty-six
relationships contracted within the village but outside one’s own caste, thirteen were
between Untouchables and non-Untouchables.

37 In two cases, the Markakasa villagers did not know the caste o f their ritual friend. For the remaining
thirteen relationships outside the village I have no data as to the caste o f the ritual friend.

62
Similarly there does not seem to be a concern to make or avoid making ritual friends with
people of the same class: wealthier villagers made ritual friendships with poorer ones;
poor with poor and wealthy with wealthy. All in all, the impression is of a rather chaotic
‘system’. It was important that the two friends be at a similar stage in their lives: though I
was closer friends with my landlord than with Ganeshyam, it was never suggested that I
become ritual friends with the former, in large part because he was married and had
children and I was not.

The ceremony is not complicated and takes about five minutes to complete. A ritual
specialist is not always required. Most people made do with a member of the family or a
friend; some people called the village baiga or the barber (nai). People often chose to
perform the ceremony at the time of the annual village fair (mandai) or at weddings, in
part because it is at these times that the largest number of witnesses will be gathered in
one place, and partly because the feasting that goes on in any case at these times can be
hijacked to celebrate the friendship ceremony too. Also, as mentioned above, the ideal,
according to some people, is that the ritual friend live in another village, and it is at
weddings and mandai that guests {saga) visit the village and old acquaintances from far
away are given the opportunity to become something more intimate. The two parties sit
on wooden blocks (pidi) facing each other and two mounds of clay or mud representing
the goddess Gauri (Parvati) and her son, the god Ganesh, are placed on a plate. These
then are the divine witnesses to the ceremony. Each friend brings with him a plate which
contains a coconut, some money (five or ten rupees, but the same amount in each), a
sprig of bound grass, red vermilion powder and a heap of husked rice grains {chawaT).
The two friends anoint the gods with vermilion powder (gulal) and then one another on
the forehead (tika). The sprig of grass is placed behind the other’s left ear and then the
plates are exchanged between the parties an odd number of times (five, seven, nine), so
that they each end up with the other’s plate. The friends feed each other the ritual
substance (be it mahaprasad, ganga ja l or tulsi ja l) and then embrace {beht). They then
feed each other paan. The coconuts are broken up and distributed among the assembly as
prasad (the ritual gift).

63
Plate 11: My Mahaprasad (with monkey).

64
Becoming a ritual friend is not necessarily an individual act. As mentioned above, it can
involve a wider class of persons than the two friends. And the relationships that are
created are not restricted to these two people. One’s ritual friend’s brothers and sisters
become one’s own siblings; their parents are referred to, both when speaking to them and
about them, as phul babu (flower father) and phul dai (flower mother). If the ritual
friends become married then their respective wives also automatically become
mahaprasad or tulsi ja l to one another without a separate ceremony being required, and
the same is often though not always true for husbands of ritual friends. Ritual friendship
seems to borrow many of the attributes of ‘real’ kinship. The following example, of three
friends tied by various bonds of ritual friendship, demonstrates both the
interconnectedness of ritual friendship, kinship and other forms of belonging, and the
differences in terms of caste and class that can exist among them.

Three friends

Naresh, Mukesh, Buddharu and I were sitting together in the middle of Naresh’s fields
where he had taken us to eat the chickpeas (channa) his family had planted there.
Wonderfully sweet when eaten raw and chewy if roasted in hay, inviting people to one’s
fields and giving them bunches of channa was a great way to consolidate old friendships
and make new ones. Going to the fields to eat channa was often employed as an excuse
for conducting private conversations and secret love affairs. These three young
Markakasa men spend a lot of time together and are good friends, yet they have rather
different backgrounds. The ways in which they are ‘related’ is summarised in Figure 3.

Naresh is a 23 year old Gond from a middle income Markakasa family. They have five
acres of farmland which, in a good year, produce sixty-five sacks of unhusked rice. The
land is also partly irrigated by a well. This enables them to sow a second crop of
chickpeas or chillies once the rice has been harvested in November. They also have a

65
Figure 3: the relationships between Naresh, Buddharu and Mukesh

Gond K evat Rawat

tulsi ja l

mama (M B)

mahaprasad sanghi

Daya Naresh Buddharu Mukesh

L
mahaprasad

K ey
Figure 3
Ritual
Village relationship
friendship
(gaon Ise riste)

‘R eal’ kinship Gond Caste

66
very large garden which is used to grow all manner of vegetables and pulses. The
importance of all this is that the family do not need to sell a lot of grain in order to buy
food. Rather, they can sell their surplus vegetables, chickpeas and chillis for money.
Naresh has studied up to the tenth standard (equivalent of GCSEs) and, though he failed
the first time, he recently retook the exams after a four year interval and passed. He
dreams of leaving the village permanently and getting a job as a policeman like his mama
(MB), who is with the Central Reserve Police Force. As a result of this highly valued and
unusual connection, Naresh has frequently travelled to Nagpur and further afield to the
city of Pune in the west in order to work.

Naresh has been friends with Buddharu for most of their lives, ever since they were both
children. The families were close and their fathers are tulsi jal. At their fathers’
suggestion, Naresh and Buddharu performed the mahaprasad ceremony when they were
eleven years old. What was the importance of becoming ritual friends when you were
already good friends, I asked. Permanence, they both replied: friendship (<dosti) can break
but mahaprasad is for life. It’s also a question of trust or faith (vishwas), said Naresh: the
level of trust you have with a mahaprasad is of a different quality.

Buddharu is of the fisherman caste (Kevat-Nishad), though this is not how his household
principally earn their living. His father is a lamsena, a man who has come to live and
work in his father-in-law’s household and, though the latter is now dead, his property has
not been disposed of among his heirs and Buddharu’s father remains landless. They are
among the poorest families in the village, renting two acres of rather undesirable land
from a Markakasa Gond which produces thirteen sacks of rice in a good year. The family
eats ten sacks worth and so relies on a cash income to buy vegetables38. Buddharu’s
father had been fortunate enough to be employed by the local Tribal Development
Council rice procurement centre, and both he and Buddharu have migrated in the past to

38 One sack o f unhusked rice (dhan not chawat) weighs approximately 70kg; the average price at which
rice was bought in 2003-04 was Rs 500 per quintel (100kg).

67
the industrial centres of Nagpur and Chandrapur where they have worked on building
sites and unloading coal trucks .

Buddharu’s sanghi is Mukesh, from one of the wealthier families in the village. He is a
Chhattisgarhi Rawat (cow-herding caste) and a B.A. student at a college seventy
kilometres away. Mukesh’s father has a government job (sarkari naukari) as a
caretaker/dogsbody (chaprasi) for the District’s Tribal Boarding Schools (Adivasi
Ashram Shala). The family have recently constructed a new (mud) house at the edge of
village, complete with a posh brick wall and flower garden. Mukesh and Buddharu
became sanghi to one another because Mukesh already regarded Buddharu’s father as his
mother’s brother {mama) in village relationship terms {gaon tse rishte). In the course of
our conversation I asked them if they could have become mahaprasad instead of sanghi.
Yes, they could have done, they both replied. Naresh, however, disputed it since
Buddharu and Mukesh already have a relationship {rishte) through him and so it would
not have been possible. Naresh and Mukesh have a village kinship relationship {gaon tse
rishte): it is an agnatic ‘uncle-nephew’ relationship, Mukesh calls Naresh, kaka (FyB)
and in turn Naresh calls him bhatija (BS). This in turn has its roots in a ritual friendship
between Naresh’s FeBS and Mukesh’s FyB. Thus, Mukesh and Buddharu could not have
become mahaprasad because they are not in the same structural relationship to one
another. Wherever there appear to be inconsistencies, such as a confusion of generations
as in this case, the parties invariably opt for the sanghi relationship, which seems to
suggest it is a more flexible category of ritual friendship. So all three friends are
connected and connect themselves to one another in ways which involve a mix of ritual
friendship, village kinship and real kinship. This example is important because it reveals
the catholicity in people’s formal friendship relations. The three friends are of different
castes and of different income levels (one might say class). Mukesh’s family earn their
primary income from formal sector government employment. They can be considered
members of the aristocracy of labour, though admittedly of the shabby variety. Buddharu

39 Seasonal migration occurs in Markakasa but is not common. According to the data from my household
survey, about ten people (both men and women) migrate every year in search o f work for approximately
three months. They are joined by others who are not regular migrants but who have had a bad harvest or are
seeking to escape the village for a while.

68
at the other end is from an exceptionally poor family even by the rather low standards of
the village. And yet their relationship is not based on an economic model of patron-client
relations. Sentiment does indeed seem to be the sole reason for creating these formal
relationship ties. It also demonstrates that ritual friendship in practice is a mix of
relationships, all involving ideas of sentiment but also spoken of as being constituted
differently: sometimes the friendship is the result of spontaneous affection; at other times
and instances it is the result of existing ties of friendship and wider association embedded
in other social relations, themselves based on sentiment.

The content of ritual friendship

What then are the obligations that ritual friends have to one another? These are only
loosely articulated. What is emphasised is not what ritual friends should and could do for
one another, but simply that they have love (prem) for one another. People generally say
that if ritual friends live in different villages they ought to visit one another, which is in a
sense a result of the prem which exists between the two. As mentioned above, life-cycle
events such as weddings and death rituals are often occasions when those ritual friends
who live a distance from one another may visit. The act of visiting friends and relatives in
other villages is an important part of both men and women’s lives. Through this act it is
possible to create an imaginary landscape of relatedness, in which one comes to
conceptualise other villages as linked with one through ties of sentiment. But this has
implications between people living in the same village too: not only is the tie between the
Markakasa person and his friends or relatives in the other village, but also with people in
Markakasa who have friends or are related to residents of that other village. Thus,
contrary to Pitt-Rivers’ assertion, ritual friendship is capable of creating sentiment that
extends beyond the two people who have directly formed the relationship. My landlord
Brijlal’s family’s ties of friendship with a family in another village provide a good
example (see Fig. 4 below).

69
About eighty years ago, a man o f the Sidar40 caste from a village called Patratola, fifteen

kilom etres away to the east in what is now the state o f Chhattisgarh, bought som e land in

Figure 4: the mahaprasad relationship between Manu and Devaram

Sidar G ond

LI

A
S a m la l
A
Devaram
dp I ,H 1
Manu Didi Brijlal

# # #

mahaprasad
F ig u re 4

In Markakasa Ritual friendship

In Patratola

40 A Scheduled Tribe under the Indian constitution and therefore ‘Adivasi’ but not Gond.

70
Markakasa and became friends with Mayaram, my landlord Brijlal’s father. As he had
land in Patratola, too, he would only occasionally visit his holdings in Markakasa and
when he did would bring his son, Devaram, with him. Mayaram’s eldest son, Manu,
worked with Devaram and they helped each other out with their respective fields. In time
they became good friends and decided to become mahaprasad to one another. Their
fathers were happy too since it cemented the relationship between the two families.
During Brijlal and his sister Didi’s childhood there was much coming and going between
Markakasa and Patratola. They would often spend several days visiting each other’s
households and they considered Devaram their older brother; his father was their phul-
baba (lit. flower-dad), the term used for a mahaprasad1s 41 father. The links were
strengthened by Devaram’s sister marrying and settling on her father’s land in Markakasa
as a lamsenin42. Their grandchildren today are very close to Brijlal and Didi’s family:
they often work together both in work gangs on government projects and as agricultural
labour, and are among the most common visitors to the house. In addition, Devaram’s
FBS, Samlal, who also inherited land in Markakasa, set up a house there and even now
shuttles between the village and Patratola. His family is close to Brijlal’s, too. Samlal’s
wife, for instance, cooks and cleans for Manu, Brijlal’s brother, when his adopted son and
daughter-in-law are away, and nursed his late wife before she passed away. Thus, though
Manu’s FyB lives in Markakasa, and they have cordial relations, it is to Samlal and his
wife (his mahaprasads FBS and a member of a different caste) whom he turns to in
times of need. When Devaram’s wife came to Markakasa to attend the wedding of one of
the ‘grandchildren’ referred to above (her HZSD), she contrived to spend as many nights
as possible with Didi’s family rather than sleep in the houses of her affines and caste-
fellows. When I asked her why, she replied that there were tensions with both of the other
households relating to old disputes in Patratola and so she felt more comfortable staying
with Didi. And this reveals another facet of the nature of ritual friendship that I
highlighted at the beginning of the chapter: it is set up in contrast to the kinship relations

41 Though it was Manu who was Devaram’s mahaprasad, the terminology used extended to his siblings.
42 ‘Lamsenin’: the wife o f a lamsena, a man who lives uxorilocally and farms the land o f his father-in-law.

71
one might have with one’s real brothers (or as in this woman’s case her husband’s
brothers).43

Along with visiting, also central is the idea that ritual friends should dine together
whenever they can; it was said that one should always invite a mahaprasad to dine at
one’s house. Festivals such as Diwali were important occasions when dining between
ritual friends living in the same village would take place. In cases where their ritual
friends are members of an Untouchable caste, Markakasa villagers pursue different
strategies. People often said that young, unmarried men and women could eat at the
house of an Untouchable without censure or pollution regardless if they were ritual
friends or not. When young people marry, however, they ought to stop. But this was not
an uncontested discourse, as we saw in the discussion of commensality and
Untouchability in Chapter Two. When it comes to ritual friends, all those who had
conducted ritual friendships across the Untouchable/ ‘clean’ divide said they had no
qualms about eating at untouchable households and they did not think about such things
when it came to ritual friends. And of course, Untouchables friends were invited to eat at
clean caste households. Caste councils, though active in other areas of life such as
marriage and divorce, did not police the ‘breaking’ of these dining rules.

There is also the feeling that ritual friends should help one another in times of need. The
following story illustrates the way in which many people thought about the ideal
friendship engendered by this type of relationship.

On more than one occasion (and indeed after my own mahaprasad ceremony was
performed), I was told the story of Lord Krishna and Sudama. Krishna and Sudama were

43 Bloch’s (1973) discussion o f affinal and agnatic kinship may suggest a complementary interpretation. He
argues, following Fortes, that affinal kinship relationships need to be constantly activated and ‘used’ in
order to be maintained: thus they have force in the short-term. For the long-term however, one knows that it
is agnatic kinship relationships which endure and have the greatest moral force. This may be the case here,
with ‘affinal kinship’ replaced with ‘ritual friendship’ in the analysis. There are a number of differences,
however, not least that ritual friendship acts to make up for the failings o f brotherhood and in the process is
transformed into a different type o f social relationship altogether. It is because one knows that the long­
term experience o f brotherhood may be ultimately disappointing that the ritual friendship based on
sentiment has more moral force.

72
childhood friends (though not ‘ritual friends’) who had played together and studied
together. When Krishna defeated his evil uncle and took back his throne at Mathura he
had told Sudama that if ever he wanted anything he need only ask. Sudama was a poor
Brahmin and his ever-growing number of children only increased his poverty. His wife
was something of a shrew by all accounts and pestered him to go and see his old friend
Krishna, now a wealthy king, to ask for assistance. Sudama wanted to take him a gift,
since it had been so long that they hadn’t seen each other. All they had in the house
however was poha (flattened or beaten rice), hardly a suitable offering to a king, but
Sudama grabbed a bagful and set off for Mathura. On his arrival, seeing the splendour of
Krishna’s court, Sudama grew ashamed of his meagre gift and decided not to give it to
his old friend. Krishna greeted Sudama affectionately and washed the feet of his guest.
But he had noticed the bag of poha that Sudama was unsuccessfully trying to hide and
snatched it from him. He opened the bag and began eating the poha with such joy that his
courtiers began whispering about this mysterious gift. “It’s the tastiest food I’ve ever
had”, announced Krishna. Sudama was happy but felt uneasy about asking Krishna for
help. He left without mentioning the reason for his visit. When he got home, it was
unrecognisable; where his hut had once stood there was now a beautiful palace. Krishna
had known what his friend had wanted and provided it. This, I was told, was how ritual
friends ought to behave with one another.

Not everyone thinks phul-phulwari is a wonderful institution, however. There are those,
admittedly few in numbers, who dislike it because of its caste blindness; in particular the
types of associations it creates between clean castes and Untouchables, and the emphasis
placed on commensal dining in these sorts of relationships. In this the opponents seem to
recognise the social value of these sorts of friendships, that the sentiment they create and
express has implications beyond the individuals concerned and their households. There
may also be opposition to the institution of ritual friendship because it is not seen as
entirely respectable. Jonathan Parry (2001) reports from Bhilai that among (Untouchable)
Satnamis ritual friends had rights of sexual access to each others wives. Though this was
never clearly articulated in Markakasa, a visiting man from Chhattisgarh told me that
when making ritual friendships the criteria people used was to consider his friend’s

73
“wealth and wife” (dhan ao daiki). For Bhilai, Parry’s data suggests that people see ritual
friendship as an old-fashioned and un-modem social institution. In Markakasa, however,
this idea seems to have little resonance: young men and women who in other respects are
concerned with the trappings of urban modernity, as they see it (e.g. clothes, cosmetics,
Bollywood film and music), continue to form ritual friendships with one another and
most regard these relationships as serious and not to be entered into lightly.44

In creating connections between people of different castes, whether they live in the same
village or not, people express the idea that sentiment has a value that transcends that of
caste. Put in other terms, the idea of difference that is promoted by an ideology of caste is
countered by the idea of similarity involved in the ideology of ritual friendship.

The problem with brothers

“The name of brother is truly a fair one and full of love... But sharing our
property or dividing it up, with the wealth of one becoming the poverty of
the other, can wondrously melt and weaken the solder binding brothers
together. Brothers have to progress and advance by driving along the same
path in the same convoy: they needs must frequently bump and jostle each
other. Moreover, why should there be found between them that congruity
and affinity which engenders true and perfect friendship?”
- Michel de Montaigne in On Friendship

When I asked people in the village about the nature of the relationship between ritual
friends, they would often reply that forming mahaprasad creates a bond like that of
brothers, like the closest of brothers, and that it is even closer than brothers because there
is no self-interest involved or reason to argue; there is sirf prem (only love, affection).
The proof that the relationship is only ‘like brothers’ is that the children of ritual friends
can themselves become ritual friends, an act that would be prohibited if in fact what was

44 The mahaprasad relationship is classed as ‘non-joking’: it is not done to make fun o f your mahaprasad
or to joke around with him. A certain decorum and respect applies.

74
conceptualised was the making of brotherhood. Thus it is misleading at the level of
analysis to lump together ritual friendship and kinship (especially the agnatic forms of
friendship and agnatic kinship), despite the apparent similarities in the content of the
relationships between the two types of relatedness. I suggest that we can uncover the true
meaning of ritual friendship if we instead oppose the categories of ritual friendship and
brotherhood. Let us take a closer look at the relations between brothers.

Brothers are regarded as the closest of all kin relations but the relationship is also
recognised as the most difficult to maintain successfully, especially after the brothers
marry. Before marriage, brothers are seen as working happily together for the good of the
common household; after marriage, however, with the arrival of wives, brothers become
selfish and quarrelsome. This is, of course, as it is seen from the male point of view, but
women as wives share these views too: that the relationship between brothers is fraught
with difficulty. The ideal household is where all brothers, along with their wives, live
together and work together without division or jealousy. But this is seen as too lofty an
ideal to realise and as a consequence people often said that having more than one brother
was a recipe for trouble.45 In fact, most households in the village have at one time or
another been seriously divided because of disputes between brothers. Two of the castes in
the village, the Mahars and the Desau Gonds, have each split into two factions as the
result of fraternal quarrels. Brothers (and their wives) are seen as intensely jealous of
their siblings. The most common cause of disputes is over land and inheritance. At least
until their father dies, brothers and their families farm the land together and hold the grain
produced in common, though if married they often keep separate stashes of petty cash
earned through selling labour. Maintaining this arrangement can, however, be difficult
and in several cases a division of property Qiissa) has been made in the father’s lifetime,
so that brothers farm separately and hand over a portion of their grain to their parents.
Moreover, accounts are certainly kept of how much labour one brother’s household
contributes to the other. The disappointing experience of brotherhood for many is in
sharp contrast to the idealised conception of ritual friendship.

45 Thus, though strictly speaking the ideology o f brotherhood involves harmony, the experience o f failed
fraternal relationships leads to the creation o f an almost-ideology o f brotherhood as problematic.

75
I returned to the village in April 2005 for just under a month’s worth of fieldwork, and a
series of incidents involving my mahaprasad, Ganeshyam, highlighted the
interested/disinterested dichotomy between brothers and ritual friends. Ganeshyam, as I
said before, is the village shopkeeper. In a tiny space less than two metres by two metres,
he stocks a rather wide range of household goods; he also buys and sells unhusked rice,
mahua flowers (used to make alcohol), tamarind, and other farm and forest produce. The
shop turns a handsome profit, and many people praise Ganeshyam not only for his fair
rates and prices but also his commercial skill. He is the youngest of four brothers and a
recent incident affected relations between them.

April is the time when the mahua flower, which is used to make alcohol, blooms and falls
and is collected by villagers to sell. The sale of mahua flowers to private shops, however,
is illegal; all mahua must be sold to the State’s Tribal Development Council (Adivasi
Vikas Mahamandal), which guarantees a .‘fair rate*. For Ganeshyam, the buying and
selling of mahua can be highly profitable but it is also dangerous. Forest officials
regularly conduct raids, and the transportation of mahua between merchants which often
takes place at night is risky and subject to interception. Tractors transporting mahua had
been caught on two separate occasions in the area that month and had been confiscated
by the Forest Department along with the flowers. According to Ganeshyam all the traders
were jittery. But he could not sit on his mahua any longer; he was short of money and
needed to sell it on. So one day before dawn he sent three bullock cart loads off to his
buyer. Unfortunately, fifteen forest officials were waiting for them and impounded the
carts and the produce. Ganeshyam rushed to the Forest Office as soon as he heard and
managed, with the help of a local leader from a neighbouring village, to persuade the
officials to release the carts and the mahua in exchange for a large sum of money. On his
return his brothers, Durga and Dayal Singh, berated him for his lack of judgement in
selling the mahua. The income that he would have made was eaten up by the bribe he had
handed over. His sister-in-law, Durga’s wife, also began to criticise him suggesting that
the shop had lost enough money, and that they ought to shut it and Ganeshyam should do
something else. Durga and Dayal Singh agreed. This was a surprising statement since

76
everyone knew how well the shop did and that it was as a result of the income from the
shop that the household had grown so wealthy. Ganeshyam said as much to his brothers
and sister-in-law. Angry and upset, he went to open his shop for business. Later that day,
he came to my house and told me what had happened. “I can’t understand why they’re
behaving like this”, he said. “I had to sell the mahua. And that’s what business, running a
shop, is like: you have to pay people off all the time. My brothers don’t understand that
because they don’t run the shop. How can they say I should close it down? They don’t
like seeing me doing so well, that’s what the problem is. They can see that I’m getting to
know lots of important people, traders, and they can’t bear it. They can’t bear the fact that
it’s because of me that the household runs at all and if any of them want money it’s me
they come to.” That evening he stayed and ate with me and came again the following
evening too. He said that he just did not feel like eating at home with his brothers. The
atmosphere was bad and his mother was crying all the time. Though he has other friends
in the village, it was striking that he chose to come to my house, the house of his
mahaprasad while he was fighting with his brothers. Compared to the intensely
interested relationship he had with his brothers, our relationship was disinterested in that
neither of us could constrain each other’s action, nor were we involved in each other’s
household affairs. Significantly, whereas brothers are expected to fight and fall out, it is
said that ritual friends never do.

This discussion of disputes and arguments among brothers (or agnatic kin more
generally) is important because such conflicts run the risk of involving magic or
witchcraft if not resolved satisfactorily. Malignant spiritual attack (discussed in more
detail in Chapter Four) is of two broad kinds: that perpetrated by witches (tonhi or
saude), who may or may not need a reason to attack whom they do, and that initiated by
laymen with the help of a sorcerer/witchdoctor (baiga or pujari). This latter kind of
spiritual attack is, according to my informants, most commonly committed by close kin,
close agnatic kin in particular, of which one’s real brother is the ideal type. While a
brother is often accused of attack, it is also common to suspect his wife, who also has an
interest in harming her brother-in-law and his family. Consider the following example.

77
My landlord Brijlal had a longstanding dispute with his elder brother Manu that has only
recently been resolved. The relationship between the two brothers has never been good
by all accounts. Its most longstanding expression is in a dispute over inheritance and
property. Manu, Brijlal’s brother, has no children of his own. Instead he adopted a boy
some fifteen years ago when the child was eight. Adoption is common in cases of
childlessness but the adopted child is usually someone within the family, a brother’s son
or daughter for instance. The reasoning is simple: ‘an “outsider” (bahar ka admi) should
not eat one’s land’. In this case, Latchu, the adopted boy, though a relative (Manu’s
FZDS), is not of their lineage. Brijlal felt great bitterness over this: Manu ought to have
adopted one of his own children so that the land would remain in the family; Manu’s four
acres would instead go to this ‘outsider’. In the midst of all these claims and accusations
of betrayal, Manu’s wife, according to Brijlal, went to see a baiga (sorcerer) to ask that
he cast a spell (mantra) to kill him. Brijlal duly fell ill, with recurring headaches and
unexplained weight loss, and was at death’s door for the good part of a year.

One could argue that interested relationships such as those involving brothers are
susceptible to the use of witchcraft and magic, especially as weapons in the course of a
dispute. Ritual friendships are ideally disinterested relationships, and as a result could be
interpreted as being free from the types of risks associated with interested ones. The
terms interested and disinterested suggest that ideas and theories about the gift and
exchange may come in useful in understanding ritual friendship.

In a recent paper discussing ritual friendship in Orissa, eastern India, Uwe Skoda hints at
a link between the ideology of ritual friendships and the ideology of the pure gift but does
not develop the argument any further (2004: 176). In another paper on ritual friendship
in Orissa, Pfeffer (2001), cited in Skoda, makes the intriguing observation that “love,
nothing but pure love is expected from and given to the [ritual friend]. They will never
demand a buffalo but their alter egos will surely provide for them.” Though not expressed
in quite those terms, the ideology seems to be very similar to what I have described
above. Firstly, that love, only love (sirf prem), is the most important factor in a ritual
friendship, and secondly,' that gifts are given without any expectation of return. This is in

78
sharp contrast with the type of exchange that takes place between brothers: though
brothers and their families may appear to give to one another with no expectation of
return in the short-term, accounts are certainly kept for the long-term. Over the course of
one’s life, generalised reciprocity eventually becomes balanced. Sahlins (1974) argues
that generalised reciprocity as open-ended responsibility is the sort of arrangement that
exists between close kin. The experience of agnatic kinship in Markakasa seems to
suggest that over the course of time it has to become balanced, or at least have the
semblance of balance, for any sort of relationship to endure. A permanent imbalance will
lead to conflict.

Take the example of my landlord Brijlal again. His elder brother Manu and he have now
reconciled and they now live and work together. The adopted son Latchu and his wife
were asked to go and live in her natal village and renounce their claim on inheriting his
land. Manu is much older than Brijlal. He is in his late sixties and can only perform the
least demanding of tasks, such as collecting tamarind or mahua. Brijlal supports his
brother now; he pays for his medicine and the repairs to his cycle, and sweeps and cleans
his elder brother’s living quarters, but this is part of the deal. When the childless Manu
dies, Brijlal will inherit his land. The acrimonious dispute of more than twenty years
standing appears to have been forgotten.

No such accounts are kept with ritual friends: people are expected to give with no
expectation of return. When I returned in April, Ganeshyam would insist that I eat with
him every evening; after the fifth dinner in as many days I remarked that I felt awkward
that I could not reciprocate. (As I ate my meals with my landlord and lady, I found it
difficult to invite people to dinner.) “But we’re mahaprasad,” he replied. “That sort of
thing doesn’t matter, you should eat with me everyday”. The motivating force is
sentiment and affection; there is no ‘looking to the future’.

Seen in terms of reciprocity and the gift, the story about Krishna and Sudama, recounted
earlier in the chapter and told to me almost every time I asked someone specifically about
ritual friendships, can be considered in a new light. What is appealing about the story is

79
not, as I had initially supposed, that it is a general statement about how friends should
behave with one another. Rather, the essential point of the story might be that Sudama
cannot ask his friend for anything but Krishna knows what he wants and gives it to him
without Sudama’s knowledge.46 As Parry states: “the genuine gift is never solicited and
the gift should be made in secret” (1986: 461). I am not suggesting that this holds true in
reality for the types of exchanges that go on between ritual friends, merely that the
ideology appears in contrast to the types of interested exchange that go on between
brothers. The gift between ritual friends looks a lot like Maussian ‘total prestations’, gifts
which do not need to be repaid. As Graeber puts it,

“unlike competitive gift exchange, ‘total prestations’ created permanent

relationships between individuals and groups, relations that were
permanent precisely because there was no way to cancel them out by a
repayment...No accounts need to be kept because the relation is not
treated as if it will ever end [my italics]” (2001: 218).

The ritual friendship captures an “image of eternity” (ibid). Unlike the expectation of
disputes with brothers (raised almost to the level of ideology), it is said that ritual friends
never fall out. This can be seen in the exchanges which take place in the ritual which
makes people into friends. As mentioned above, each party has a plate containing some
money and rice and this plate is passed back and forth between the parties five, seven, or
nine times. The significance of odd numbers is not only that at the end of the process,
each party holds the other’s plate, but that as a result of this long back-and-forth, the
parties become confused as to whose plate they are holding at any point during the
exchange. This has two implications: firstly, it suggests that this type of exchange is
represented in ritual as continuing indefinitely and as an ‘image of eternity’ (what else
explains the prolonged series of exchanges?); secondly, the fact of exchange becomes

46 That Krishna is a god is fairly irrelevant in the telling o f this story, and the contexts in which I was told it
make this clear. It is not Krishna’s divinity that is emphasised but his quality as a friend, and he realises
what Sudama wants because he has prem for him. That the story featured a god is not at all surprising:
throughout much o f India stories o f gods are told in exactly this fashion to signify exemplary behaviour to
be aspired to by ordinary mortals.

80
irrelevant because of the confusion caused by the passing back and forth - the knowledge
that one has given to the other seems to be enough.

I should say here that I am not suggesting that the relationship between brothers is
characterised by a complete lack of an ‘image of eternity’. I am simply proposing that
type of reciprocity that goes on between brothers (or other agnatic kin) is subject to
change from generalised to balanced (or in Graeber’s terms, from ‘open’ to ‘closed’
[2001: 220]47), whereas the ideology of exchange which characterises relations between
ritual friends remains permanently ‘open’.

This brings us neatly back to a discussion of equality and hierarchy, and the opposition of
the ideology of ritual friendship to the ideology of caste. Once again, taking my cue from
Graeber, gifts do not need to be repaid between ritual friends because the relationship is
not identified with inequality between the actors: “Gifts have to be repaid when
communistic relations are so identified with inequality [as in the case of brothers of the
same household] that not doing so would place the recipient in the position of an inferior”
(ibid: 221). The relations between ritual friends are not characterised by inequality but the
relations between brothers (especially those who share a household) certainly are. By not
repaying the gifts received from a brother, one places oneself in a position of inferiority
in relation to that brother. Take the example of Brijlal and Manu once again. Brijlal
provides for Manu and receives nothing in return for the moment. Manu (though older) is
put in a position of inferiority as long as no return is made. This situation will change
once Manu is dead and makes his return by giving his land to Brijlal. That ritual friends
are regarded as equal is supported by the fact that, since sentiment is the basis of their
relationship, they are not compelled to make a return: there is no question of an inferior
or superior friend.

And yet, are ritual friendships in reality, as they are experienced and not as they are
idealised, ultimately about self-interest and is that why people make them? The comment

47 Where ‘open reciprocity’ means that which “keeps no accounts because it implies a relation o f permanent
commitment” (Graeber 2001: 220); ‘closed reciprocity’ occurs when “a balancing o f accounts closes the
relationship off, or at least maintains the constant possibility o f doing so” (ibid).

81
made by a Chhattisgarhi visitor to the village, referred to above, that one looks at ‘wife
and wealth’ in making ritual friends, would certainly seem to support such a view. And
the view of some, expressed by Shamrao right at the beginning of the chapter, that ritual
friends ought to be from different villages, ideally far away, seems to indicate that
friendships are made to expand social networks beyond the confines of one’s village and
‘to get to know more people’ - people one would not otherwise encounter. Interestingly,
in answer to my questions about why people married less either within the village or to
cross-cousins than in the past, I was often told that to do so meant one did not get to
know new people, and knowing new people was seen as valuable. By bringing the fear of
witchcraft back into the discussion, however, one could alternatively suggest that the
greater the distance between ritual friends, the lesser the chance the relationship could
become interested and thus susceptible to the types of attack that occur among kin and
neighbours. One could look at benefit in two ways: material and moral. Daya, a forty-five
year old Teli man, has a mahaprasad relationship with a bullock seller who visits the
village once, sometimes twice, a year. When I asked them and others why they had
become ritual friends, they gave the stock answer: prem (love or affection). But it is also
clear that for the bullock seller, having a ritual friend in Markakasa means that on a cold
winter’s night he is fed and housed during his stay while the other members of his party
shiver outside under the mango trees. He also has the moral benefit of being treated not
as a complete outsider, as his fellow bullock-sellers are, but explicitly as a saga
(guest/relative)48 of the village because of his relationship with Daya. It is impossible to
ascertain, however, whether this moral benefit would translate into material benefit, in
the form of improved sales of bullocks for instance, but certainly in the eyes of Daya’s
fellow villagers there was a sense that they would get a fairer deal from a man connected
with them in this way. Nevertheless, the cornerstone of the relationship is still idealised
as being one of love and affection (prem), despite the actual experience of the ritual
friendship once it is formed, which may involve material or moral benefit. The practice
and reality of ritual friendship can be as disappointing as the relations with brothers:
friendships made in youth can fall away in later life and, though I never heard of ritual

48 This is the same sort o f relationship I had with the village, as a saga, though my status I suspect was both
because I had a mahaprasad in the village and because I had spent an extended period o f time there. Saga
is an extraordinarily flexible word, meaning ‘guest’, ‘caste-fellow’ or ‘relative’ depending on the context.

82
friends arguing or fighting, simple avoidance can be the expression of a disagreement or
of a growing-apart. After all, ending a marriage or building a separate house from one’s
brother signals the end or at least suspension of those respective relationships. But how
does one publicly end a friendship? In contrast, however, to the relations with brothers,
and wives for that matter, there is a strong presumption that ritual friends do not argue
and that it is a lifelong association. Though undoubtedly idealised, the point worthy of
note is that it is my informants themselves who do the idealising, and not the
anthropologist who could quite naturally be accused of viewing their social world
through rose-tinted spectacles.

Conclusion

Friendship should indeed be seen as a process (Loizos & Papataxiarchis 1991), as a form
of belonging that changes over the course of a lifetime: at different points in one’s life
certain relationships are privileged over others, and parents, peers, siblings, spouses,
children, as well as friends fade in and out. Such process has important social
significance. But as I have shown here, looking at friendship as structure, and as
functional, can also provide valuable insights into personhood and sociality.

By looking at the function of this particular form of friendship in the context of social life
in Markakasa, it should be clear that ‘ritual friendship’ is not at all the same thing as
‘fictive kinship’. Despite the fact that ritual friendship is modelled on kinship to a certain
degree (e.g. the use of kin terms), what is constructed in ideological terms is a type of
association which is ultimately contrasted with kinship, not assimilated to it. The fear of
witchcraft or magic, the expectation of disputes between brothers and the accompanying
disappointment one might feel about the fraternal relationship contribute in large part, I
suggest, to the construction of an ideology of ritual friendship that is typically
disinterested and based purely on affection. Where there is no ‘interest’, at least in
ideological terms, one can create a safe relationship not subject to spiritual attack or
spectacular dispute. Ritual friendship also demonstrates that sentiment between people of
different castes acts as a counter to the ideology of caste in certain spheres of social life,

83
and also that this sentiment extends beyond the two individual parties to the friendship. In
addition to affirming the social body of the village by connecting people across caste, kin
and class lines and thereby emphasising a fundamental affinity as members of a common
humanity (see Rezende 1999) localised in a particular space, ritual friendship creates a
landscape of relations with people outside one’s locale, which has the effect of
reinforcing ties with related people within one’s village. This identification does,
however, have its limits, and though sentiment and affinity are shown to be possible
between people who are in other respects regarded as very different, this difference
emerges as significant when people consider the problems of living in a ‘mixed village’,
elaborated further in later chapters.

This chapter has also begun a discussion which will continue throughout the rest of the
thesis about the social importance of malicious spiritual attack, either in the form of
witchcraft or magic. Here I have shown how it enters into an analysis of the intimate
relations between members of a given community and how it informs understandings at
an abstract level about the nature of relationships such as friends and brothers. In the
chapter which follows, I examine witchcraft and magic more closely and discuss the
impediments which have developed to combating spiritual attack, both at the level of the
person and the village.

84
Plate 12: Women gatheredfor a chikat. A Chikat is an occasion where a gift
(usually rice) is made to a household where a wedding is taking place. The gift is
generally made by friends and kin and is used to alleviate the cost offeeding
guests at the wedding feast. This gift is in addition to that which is offered when
the bride and groom are formally greeted by friends and kin during a ceremony
known as tikan.

L ..

Plate 13: A Markakasa boy at the chikat.

85
 Chapter Four
Anti-Witchcraft and its Impediments: Migration, the ‘Mixed
Village’, and the Indian State

I was only four months into fieldwork and already thoroughly sick of the food.
The rains that year had not been good and consequently very little had grown in my
landlord’s garden. In normal years, he assured me, there would be runner beans,
tomatoes, bittergourd, and various pulses. Nor was anything of interest available in the
markets held in villages around Markakasa. I had been eating nothing but aubergine curry
noon and night for five days. On the sixth day, as I sat down to lunch and my landlady
served me a steaming mound of rice and then the dreaded aubergine, my appetite failed
me. I managed a couple of mouthfuls and then asked if I could leave the rest. I hated
wasting food but my landlady Champabai reassured me: “Don’t worry Desai, we can give
your leftovers to the cattle”. Evening came round and when I was called to eat I replied
that I wasn’t hungry. My landlord Brijlal made his way into the kitchen from the
courtyard where I was sitting and Champabai said to him, “I think there’s something
wrong with Desai; he didn’t eat lunch and now he says he’s not hungry. I think he’s got
deet.” “Surely not”, replied Brijlal, “he hasn’t eaten at anyone else’s house recently, has
he?” “Well”, piped up Brijlal’s seventeen year old niece Nanda, “he’s been going to the
Bakchoriya house a lot to watch cricket on their television. He must have had something
there”. Deet, as I learned through the conversation that ensued, was a type of witchcraft
that was performed while the subject of the attack was eating. The witch would watch the
victim eat and then would lick her lips and swallow her own saliva. This act would cause
the victim (almost always a young person) to lose his appetite; if the deet held then the
victim would eventually die. “The old Bakchoriya woman, Mukesh’s grandmother, is

86
bad”, Nanda explained. “Yes, it’s true”, confirmed Champa, “but don’t say anything to
anyone, it’ll cause badnami (infamy; a stain on one’s name49).”

This was my first encounter with witchcraft. Until then, I had been aware neither of its
existence nor the problems that it could cause. As the months passed, I began to
investigate witchcraft and magic in a more systematic way. I did not find evidence of a
‘witchcraft panic’50, as much of the academic writing describes for southern Africa (e.g.
Comaroff and Comaroff 1999). Rather, the general feeling among Markakasa people was
that witches in particular were bolder now than in the past, because they knew that the
villagers could not hold them to account. But this did not translate into a dramatic
increase in the incidence of attack compared to the past. The problem was simply that
witches or those who employed sorcery could neither be identified nor dealt with
publicly. I also observed that the membership of a particular Hindu religious sect, the
Mahanubhav sect, had grown considerably in the past fifteen years both in Markakasa
and the surrounding area more generally. The connection between the two processes was
that people joined the sect principally in order to protect themselves from malicious
spiritual attack. The question naturally posed itself: what had changed about the way in
which people dealt with witchcraft and why had it changed? Had witchcraft and magic
become more difficult to combat?

As I discussed in the previous chapter, fear of attack from witchcraft and sorcery,
resulting from the closeness and interestedness of the agnatic kinship relationship,
contributes to some extent to the form and ideology of ritual friendship in Markakasa,
marked as it is by distance, permanence and affection. In that analysis, the phenomenon
of witchcraft was examined (and used as a tool) for what it can tell us about ‘traditional’
fields of anthropological concern: kinship, reciprocity, the gift and sociality (see also
Kapferer 2003: 24). This chapter continues in a similar vein and looks at witchcraft and
sorcery somewhat obliquely.

49 Though it w asn’t entirely clear whose name would be besmirched - that o f the Bakchoriya household or
that o f my landlord’s.
50 Such ideas are in any case hard to quantify and may not have much analytical value (see also Kapferer
2003).

87
I begin with a general discussion of witchcraft and sorcery in the village. The principal
task of this chapter, however, is to examine the impediments that have developed,
preventing people from dealing effectively with these sorts of harm. These impediments
have become more pronounced over the past twenty years, as a result both of increased
state interest and engagement in the region and of the changing composition of the village
(and other villages in the area like it) in terms of caste and ethnicity. As we will see in the
chapters which follow, the fact that witchcraft can no longer be dealt with in satisfactory
ways means that an increasing number of people in the area are seeking relief through a
particular religious sect, membership of which creates a transformation in sociality. It is
also evident that the values and practices promoted by the sect have affinities with the
ideology and practices of local Hindu nationalist activists (see Chapter Six).

While the academic literature on witchcraft and magic abounds with studies from Africa
(e.g. Geschiere 1997, 1999; Auslander 1993; Moore and Sanders 2001) where the work is
engaged with anthropological debates of modernity, economic transformation and.
religious change51, work on witchcraft in South Asia has focussed largely on the issue of
accusations and violence against suspected witches, particularly in the context of gender
(e.g. Carstairs 1983; Kapur 1983; Bailey 1997; Kelkar and Nathan 1991; Nathan et al.
1998; Mishra 2003; MacDonald 2004)52. There has, however, been little analysis of the
changing ways in which people deal with attacks on an everyday level as problems of
existence, and the reasons for and implications of these changes (though see Skaria 1997a
for a historical account). This chapter is therefore not a general discussion of witchcraft
as witchcraft in the manner of Evans-Pritchard (1976 [1937]) and does not seek to raise
questions of rationality and causality. Nor is it an examination of the pattern and logic of
witch accusations (see Sundar 2001; MacDonald 2004).53 It seeks to document the
problems people face in the context of social, economic and legal changes when

51 See also Meyer and Pels (2003) for discussions o f magic and modernity in a variety o f settings.
52 MacDonald’s (2004) study is in part an examination o f varying perceptions o f Indian modernity,
especially among different echelons o f the police force.
53 Part o f the problem in Markakasa, discussed in more detail below, is that witchcraft accusations by
individuals very rarely become officially public (even at the level o f the village) because to do so in the
context o f these new conditions o f state surveillance and village heterogeneity would mean a report being
filed against the accuser by the Police Patil at the police station, and open conflict between village factions.

88
attempting to deal with mystical attack attributed to the actions of others.54 This chapter
can be read as an exploration of these social changes through the lens of witchcraft and
sorcery.

There appear to be three broad reasons why impediments to dealing with witchcraft have
developed in this part of Adivasi central India. Firstly, migration to another village in an
attempt to escape witchcraft or unexplained misfortune has become more difficult, partly
as a result of changes in legislation which have reduced the supply of alienable land in
Adivasi areas. Thus, whereas in earlier times people could have ‘solved’ the problem of
malicious attack by migrating, nowadays, short of moving permanently to the city which
is generally seen as undesirable, people who suffer such attacks cannot seriously consider
escaping.

Secondly, earlier waves of migration to the area have changed the villagers’ perception of
the nature of their community: from being a village of one type of people, Gaita Gonds, it
has become a ‘mixed village’ (panchrangi gaori) with different castes, languages and
ways of doing things. This has meant that there is a lack of consensus as to the value or
effectiveness of certain village-wide anti-witchcraft/sorcery measures, such as the clan-
god Angadev, that could have been employed in the past to detect witches and ghosts.
Similarly, at the level of the household, differences between Gonds and others regarding
the control of female sexuality and marriage have meant that a ritual for testing potential
wives and daughters-in-law for witchcraft, which also expresses a divergent attitude to
sexuality, has become less desirable as Gonds try to avoid ‘looking bad’ in front of other
castes.

Thirdly, the increased engagement of the state in the area, and in particular of the police,
as a result of the Naxalite insurgency, has made dealing with suspected witches more
problematic. Road-building projects and new bus routes have brought Markakasa people
ever closer to local administration; in turn, local administration has better access to

54 Following Kapferer (1997,2003), I hesitate at this stage to see witchcraft as resistance, or as a folk
explanation for misfortune, or indeed as a metaphor for wider social processes o f economic change or
discontent with the consequences o f millennial capitalism (cf. Sontag 1991, Comaroff and Comaroff 1999).

89
Markakasa. Police patrols on the look out for insurgents make this arm of the state more
involved in local life than ever before. While local police may share some of the beliefs
of villagers in witchcraft and sorcery, they go to great lengths to distinguish themselves
from the ‘backward’ Adivasis who in their blindness support the rebels. Consequently,
they attempt to regulate any evidence of ‘superstition’ in areas under their protection, as
part of a wider plan of bringing development and progress to the benighted Adivasis in
order to wean them off the Naxalites. The suppression of anti-witchcraft activities is part
of a larger political and social project of modernity in India that has a long genealogy
stretching back to the colonial period. In contemporary India, it finds expression in the
activities both of state actors such as the police, and also of non-governmental
organisations such as the Maharashtra Committee for the Eradication of Superstitious
Practices who have recently promoted legislation aimed at suppressing ‘superstitious
practices’.

Before turning to an examination of these three broad impediments, let us briefly

consider local ideas of witches and sorcerers.

Witches, sorcerers, and ghosts55

The general category of jadu-tonha (witchcraft) is wide and includes acts performed by
witches proper, magic and spells cast by baiga (sorcerers and ritual specialists) and
laymen, and extends to the employment of ghosts/spirits (bhut). Witches can be both men
(tonha; saude) and women (tohni; saudin), though most people agree that women are

55 Douglas’ (1970, 1973) proposition that sorcery and witchcraft are distinctive categories and arise in
different social conditions is instructive. Sorcery, in her definition, is conscious, intentional, performed by
specialists trained in the art o f magic and ambivalent, in that it can be used for both protection and harm. It
occurs in societies with complex hierarchies and widely shared codes controlling kin activity and relations
with strangers. Witchcraft on the other hand, is always malevolent and is present in societies characterised
by looser social arrangements, and where people are more mobile and less stratified. In India, therefore,
one would expect to see more sorcery than witchcraft. While noting the importance o f Douglas’ typology,
Kapferer (2003) prefers to see her distinction between witchcraft and sorcery as Weberian ideal types (ibid:
11). “Ethnographically, their character as typified here is often intimately related...The forms o f sorcery
and witchcraft are likely to occur in different social dynamics and social spaces within the one broad social
context” (ibid: 12). While aware o f the distinction between the two, witchcraft and sorcery are used fairly
interchangeably in the main body o f the chapter, largely because Markakasa people seldom themselves
distinguished the different types o f magic.

90
more likely than men to be witches. People can be bom with the qualities of the witch
and it can be hereditary, but in most cases a witch has been trained and taught spells
(;mantra) by another witch56. Witches are most often encountered at night57. It is at night
that a witch really becomes recognisably a witch in the sense of taking on her
characteristics: completely naked, she floats above the ground, and her saliva drips from
her mouth, shining like silver. Most of the accounts of witch-attack involve examples of
vampirism: the most common method is that the witch will climb onto the roof of one’s
house and position herself above her sleeping victim. She ‘binds’ the victim using a spell
so that he cannot move or speak, and then lowers down a pipe into the victim’s navel
through which she sucks his blood. This continues over several nights or weeks until the
victim eventually dies. Throughout this experience, the victim is generally unaware of
these night visits but gets progressively weaker and sicker. Sometimes, however, the
victim does remember what has gone on during the night and experiences it as a dream. If
the dreams reoccur several times, the victim becomes aware of the attack and can begin
to seek a remedy.58 The tonha or tohni is a basic-type of witch. The saude or saudin is a
witch with a greater mastery of spells, and is therefore more powerful and dangerous.
Once the victim has been killed, either by the process described above or simply by the
casting of a spell, the saude exhumes the body after burial and brings it back to life, a feat
that requires a fairly advanced knowledge of magic. In an act of cannibalism, he/she
proceeds to eat the flesh and drink the blood of the victim.

What motivates witches (tonhi and saude) is complicated, and no clear picture emerges:
they act interestedly and disinterestedly, the former often in the course of a dispute,
though they are just as likely to attack ‘without reason or motive’ {bind karan; bind
matlab). There is also the question of intention that is separate from that of motive. Deet,
the particular spell allegedly cast on the unfortunate anthropologist, is always done
intentionally by witches, whether or not they are acting interestedly. But ordinary people

56 See also Das (1988: 229) for a brief description o f the training o f witches in Chhattisgarh.
57 The time just before dawn is especially dangerous for women; the latter are usually the first to wake in
their households, and as they make their way to the fields or forest for their morning defecation there is a
risk of being accosted by a witch seeking an apprentice.
58 Another sign that one is being attacked by a witch is the visit to one’s house o f an owl. Owls are seen as
particularly powerful creatures, with knowledge o f special mantra which they teach to witches and baiga;
the former often employ them as familiars.

91
who are not witches and have no knowledge of mantra can also cause deet
unintentionally by the act of watching someone eat, desiring the food and licking their
lips: children eating in front of one another at school were particularly susceptible to
this.59

People who are not witches and have no knowledge of spells do have recourse to
sorcerers known as baiga or pujari, who attack others in exchange for payment. Baiga
have two principal functions: firstly, they are the main officiates at village rituals and at
(Gond) life-cycle events60; secondly, they are wielders of powerful spells (mantra), and
are employed as sorcerers, diviners and healers. Not all baiga are competent to fulfil both
functions, however: the main baiga of Markakasa, for instance, was simply a ritual
officiate with no knowledge of powerful spells, and was referred to rather dismissively as
a ‘coconut-breaking baiga\ Those who are sorcerers are sought out to divine causes of
illness and suffering, check for possible witchcraft attacks, prescribe cures and to remove
misfortune pertaining to the planet Saturn (shani). Baiga can work for a client who
wishes ill on others. They are asked to cause illness, discomfort, death or material
hardship for one’s enemies. Baiga can of course also work on their own behalf, and
powerful ones are feared in their communities and among their acquaintances for the
potential damage they can do.

Witchcraft is an ‘open secret’ (see Auslander 1993) in Markakasa. Everyone knows it

exists and most (including children) know the identity of established tohni and saude in
the village. Nevertheless, it is not generally a topic that is discussed in public except
when talking about witchcraft elsewhere. If the subject is raised in public, people express
a marked discomfort, refusing to utter the words ‘tonhV or ij adu-tohna\ and instead use
hand gestures to signal what they mean. The number and identity of witches (tohni or
saude) in the village differed depending on who one talked to, but there were at least five

59 Adults rarely cause deet (except if they are witches), largely because, unlike children, they are expected
to control their desires. Moreover, they do not often find themselves in situations where they are eating
food that is different to that o f their co-diners.
60 However, at certain life-cycle events such as the Gond funerary ritual known as karsad, magic in the
form of mantra (spells) is used to determine the cause o f death o f the deceased. The performance o f this
requires a baiga (and often several) with knowledge o f magic. See also the discussion o f the pre-wedding
ritual o fpanbuddi, below.

92
Plate 14: A powerful baiga from a neighbouring village.

93
on whom most were agreed: an elderly Gond widow, two Teli women, a Teli man and a
Chhattisgarhi Rawat woman (the last being the witch who attacked me). Beyond these
were a number of ‘unknown’ witches, again a mixture of men and women, about whom
there was no sense of an ‘open secret’. Individuals had particular grievances against
certain people, often kin or close neighbours, whom they suspected either of having
learnt witchcraft or of regularly using a baiga to work on their behalf.

Ghosts (bhut) form a separate category of powerful magical beings which can cause great
harm. Though they are most commonly encountered near water crossings or in the forest,
they do venture into settlements where, if invited, they will take up residence. Ghosts, I
was told, are always looking for a home. In exchange they offer their services: they can,
for instance, increase the wealth of the host household or torment an enemy. They appear
at night, while one is asleep, often in the guise of a traveller looking for shelter.
Typically, they promise to become a servant and, if invited in, they do their master’s
bidding for a while. Ultimately however, the ghost turns on the master and destroys him.
Ghosts are mainly employed for stealing other people’s wealth in the form of grain and
their use is often cited as explanations of sudden prosperity. The scion of one of the
earliest Teli families to migrate to the village was said to have employed bhut in this way:
they arrived landless and made roof-tiles for a living, and yet they managed to purchase
several acres of land and a (since-sold) scooter. The thinking was that their ghost had
stolen grain from other households in the village. Ghosts can also torment one’s enemies
by throwing stones at their homes or by appearing in dreams. Whereas ghosts that are
employed by individuals can be troublesome for others individually, ghosts that are not
employed by anyone yet live close to or in a settlement can be the source of harm for the
village as a community. As we will see in the sections that follow, ghosts are just as
much objects of village-cleansing measures as witches.

94
Resolving disputes, migration, and land reform

Markakasa is not a particularly old settlement, having been founded sometime in the late
nineteenth century61, and has grown largely through migration from other places within a
fifty kilometre radius of the village, principally in the last sixty years or so. In the course
of fieldwork I collected migration histories of almost all households in the village. For
the vast majority of people, names of villages and the circumstances of their departure
were well remembered and indeed contact with ‘original’ villages had in some cases been
maintained, often over periods of seventy or eighty years. There was no sense that people
were involved in actively forgetting where their ancestors had come from.

One reason for leaving one’s home village was to escape circumstances and an
environment which seemed to one to promote misfortune, whether articulated in terms of
malicious attack or a more general feeling of unease. Examples include the death of a
series of children, the inability to recover from a prolonged illness, the failure of harvests
and the death of cattle. Another reason was the desire to escape a dispute. As we saw in
the last chapter, relationships with kin here, as in many other parts of the world, are often
fraught with difficulty. And conflict between kin - agnatic in particular - is common,
especially as a result of disputes involving access to land and other resources. Though it
has often been analysed solely in terms of economic motivation, migration, either to other
villages or to urban centres, is often a way in which people can, through a policy of
avoidance and separation, attempt to resolve a dispute. Indeed, the desire to ‘resolve’ a
dispute was a common reason given by many people as to why their father or grandfather
(or in some cases themselves) packed up and left. Disputes and witchcraft or magic are
intimately connected: the former leads in most cases to an attack of the latter.

Take the example of Bihari, a man of the Teli caste who lives in the same part of the
village as I did. About fifty years old, his three sons are all married and Bihari has
already divided up his property (land, cattle, plough) among his children, who each give

61 This is an estimate based on genealogies o f the oldest families to settle the village. O f the two villages
(gaon) - Markakasa and Herpar - that make up the administrative village unit (gram), the latter is the older
(and smaller), though I have no information as to when it was first settled.

95
him a share of their harvest. He has in the past couple of years become one of the
village’s entrepreneurs by opening a largely unprofitable paan stall, which he takes on
tour during the bull-chariot racing season. The family are originally from a village called
Mohgaon, some twenty kilometres away to the north-west, and moved to Markakasa in
1979. Bihari’s father had died many years earlier and the family was plagued by
problems. Bihari’s FFyB (dokra-baba) was claiming Bihari’s father’s land, saying that
the division of property had not been correctly performed. This was an old dispute,
according to Bihari, one that went back to his grandfather’s time. It was a time of great
pain (dukh): children were falling ill, and the crop failed in years when those of others
succeeded. This confirmed what they had suspected: that Bihari’s FFyBW (dokri-dai)
was a saudin and was directing her attacks against them in order to seize their land.
Unable to live there any longer, the family - Bihari, his wife, mother, sister and FyB’s
household - sold their land and left Mohgaon. Having bought farmland in the village of
Kedanar, they moved to neighbouring Markakasa, Bihari’s mother’s natal village, and
initially stayed with her brother. They then acquired land in Markakasa from the Thakur
to build a house of their own. Indicative of the type of obstacles many people face when
settling on land in the area, the forest department claimed that it was the original
landholder of the plot bought by Bihari and that the Thakur had no right to sell it. He paid
a fine and the matter appears to be resolved62. Misfortune, however, seemed to follow
Bihari: having moved away from a source of spiritual attack in Mohgaon, Bihari suffered
magical attacks in Markakasa itself, this time from a non-kinsman but again as a result of
a dispute over land.

The attempt to escape witchcraft attacks was not always articulated as related to a
dispute. Sometimes there was a feeling of unease and incurable illness, together with the
awareness that one was being attacked. Consider Pardhu, a middle-aged Gond man,
whose story in some ways parallels that of Bihari. Hailing from Bonde, a village to the
south in an area regarded by Markakasa people as wilder territory (jangalpatti), Pardhu’s

62 The overwhelming majority o f houses in Markakasa are built on forest land and are therefore illegal.
Most o f these householders have at some point been summoned to the sub-district administrative office to
answer a charge and pay a fine. The payment o f the fine de facto regularises the illegal occupation, and it is
understood that the forest department will not ask that the house be demolished.

96
father died when he was a boy. His mother took the children to her natal village,
Markakasa, and stayed in the house of her brother. In order to ease the burden on his
mama (MB), Pardhu became a naukar (long-term servant) for another village Gond,
Buddu Koreti, grazing his buffaloes. After two years of service, Buddu asked Pardhu if
he would marry his daughter and become a lamsena, a married man who lives
uxorilocally. He agreed, and for a while they lived in his father-in-law’s household.
Tensions appeared to have developed and Pardhu did not like the way he was treated by
Buddu. He left Markakasa with his wife and mother and returned to his natal village.
Within a couple of years of his return, he fell seriously ill: spots and boils developed on
his legs and feet and he felt very weak. The sickness came and went and he found it
difficult to farm his land. His mother told him that his father had died of a similar illness
and that they should leave Bonde permanently and return to Markakasa. There he
recovered, but for several years they had no land to farm as their own. It was his wife’s
duty, he said, to ask her father for a share of his estate but she never did so. Nevertheless,
Buddu did give them some land to build a house, and Pardhu’s wife eventually inherited
farmland from her mother’s family in a neighbouring village.

Of course, there were other reasons why people left their natal villages and came to
Markakasa. Famine resulting from crop failure at home was repeatedly mentioned, and
this concurs with the historical record of several famines in the late nineteenth and early-
to mid-twentieth centuries in both the Chhattisgarh plains and Chanda and Bhandara
forests (Russell 1908: 142-44). The area around Markakasa was seen as an attractive
place to come for those from the plains because, although the soil was inferior, there was
an abundance of forest and unsettled land. The forest itself could provide an income
through the sale of bamboo and other produce, and where necessary could be cut down
and ploughed.

Overall, there was an impression among Markakasa’s inhabitants that people were more
willing to move their homes from place to place in the past than now.63 Of course, these

63 See also Nandini Sundar (1997: 44-5), who similarly suggests that the rate o f migration in the
neighbouring region o f Bastar in the past was high and the population mobile.

97
days people migrate to Nagpur or Mumbai to work. I was told that this rarely happened in
the past and even now very few end up staying there permanently. Of those who had left
the village permanently, there were only two men (and their families) who, as informal
sector workers, had done so for reasons unconnected with taking up a government job
(which would require relocation). For many young men who had worked as seasonal
labourers, life in the city was seen as exciting, with the distractions of girls, ‘hotels’
(restaurants) and the cinema; but it was ultimately unliveable. Money was plentiful in the
city but one’s quality of life suffered. One had to buy firewood - spindly sticks and
branches one would never consider using in the village. The city was seen as a dirty
place, full of illness and mosquitoes, where there was nowhere to seek relief from the sun
and heat; a place where bathing was difficult and where one had no choice but to defecate
in full view of women.

Non-seasonal mobility was bound up with availability of land: in the past, one could
arrive in the village, cut down some of the forest and, if one farmed the land for two
years, one was permitted, according to custom, to assert ownership over that land. In any
case I was told, people in the past (especially Gonds) disliked staying in any one place for
too long and farming the same land over again. If they suffered misfortune or attack,
received a bad omen or felt uneasy they would leave the village and move to another one.

Moving permanently to another village to escape spiritual attack or a sense of unease has
become progressively more difficult in areas where Adivasis live, largely because the
supply of farmland has, from the 1960s (in Maharashtra at least), been restricted by state
legislation. Under these laws, an Adivasi may not alienate his land to a non-Adivasi
without the written consent of the Collector, the highest-ranking administrator in a
district. These restrictions were put in place by the Maharashtra Land Revenue Code of
1966 in accordance with certain provisions of the Indian Constitution. It quickly became
apparent, however, that in many districts Collectors were granting permission as a matter
of routine. The State then enacted two measures to tighten up the regime of non­
alienation: the Maharashtra Land Revenue Code and Tenancy Laws (Amendment) Act
1974 and the Maharashtra Restoration of Land to Scheduled Tribes Act 1974. The former

98
imposed stricter restrictions on future alienation while the latter acted retrospectively to
invalidate any land transfers made from Adivasi to non-Adivasi in the period from 1966
to 1974 and return the land to the original Adivasi landholder. These laws were enacted
in order to protect Adivasi landowners from exploitation by moneylenders and traders,
and by non-Adivasis migrating to Adivasi areas and settling the land.64 Over the past 100
years or so, Adivasi land tenure has decreased considerably. In some parts of central
India, Gonds, who just a century ago were the principal landowners have now been so
thoroughly displaced that many work as labourers on land they once used to own. Bates
(1985: 574), for instance, reports that the number of Gond villages (those where the
landlord was a Gond) fell from 205 to 141 in Narsinghpur district between 1866 and
1907 and from 294 to 129 in Raipur district between 1869 and 1912. Indebtedness to
moneylenders and shopkeepers was one of the major causes of this process of
appropriation: land would be demanded in order to settle debts or bills. Though abuses
still occur, they have largely diminished in most areas, due in part to the growth of the
Naxalite movement (see below)65. In Markakasa, although the Gaita Gonds comprise less
than a quarter of the total village population, they hold nearly sixty percent of the
available land: their combined landholding is 203 acres while that of the Telis and
Mahars (non-Adivasi castes which have comparable population in Markakasa), is 57 and
45 acres respectively. Whereas pre-1974 (1966) Gond-held land was alienable and was
sold to incomers, since 1974 this land has ceased to be available. This is the case, of
course, for many of the villages in the area. To be sure, though the supply of land
available to purchase is far more restricted nowadays than in the past, it is still possible to
lease land for less than five consecutive years according to the 1974 law. Farmland
available for lease in this way is, however, often of inferior quality, being situated on

64 Many o f the Adivasis I knew, while appreciating the need to protect their landholdings, did resent the
fact that they were effectively barred from selling land, which they could no longer farm effectively
through lack o f labour, and were unable to convert into money which could be used to buy agricultural
capital such as tractors or pumps. Non-Adivasis would often complain that while Adivasi people had a lot
o f land, they were weak (kamjor) and unable or unwilling to farm it as effectively as others.
65 The abuses now generally centre on the benefits Adivasi people receive under government policies to
assist Scheduled Tribes, though the perception is that these too have diminished as a result o f the Naxalite
presence. I was told o f several cases, for example, when a local trader ‘persuaded’ an Adivasi to apply for a
subsidised tractor (the cost o f which the Adivasi could not afford) so that the trader could use it to transport
his goods. The increased availability o f expensive medicines in recent years has also caused some local
people to go into debt in an attempt to cure protracted illness. Unable to pay the pharmacist, they agree to
lease their land for a certain number o f years.

99
uneven and sloping ground and thus less productive, and any crop sown here is more
prone to failure. The Jhetmal family, a group of four Teli households and relative
newcomers to the village, were often discussed as having succeeded in cultivating leased
land productively and profitably. Their history demonstrates to some extent the
intersections of migration, witchcraft and land law reform.

The Jhetmal family are Chhattisgarhi Telis and live in the newest part of the village,
located at its northern end close to the Gond graveyard where the forest has recently been
cut into to make way for more houses. The four brothers, together with their father and
respective households, came to Markakasa in 1987 from the neighbouring village of
Somutola where they had been resident since 1968. Before that they had lived in
Alewada, a village thirty kilmetres away to the north-west. After arriving in Somutola,
the Jhetmals had bought land from a Gond in 1973. Unfortunately for them, the
enactment of the Scheduled Tribe land legislation in 1974 meant that the sale was
retrospectively annulled and their eighteen acres of land reverted to the original owner.
They left Somutola because land was available to rent in Kedanar, a village neighbouring
Markakasa, and also because Somutola had become an uncomfortable place to live, as
one of the brothers, Atmaram Jhetmal, explained:

“Somutola is a Gond village. There used to be some households of other

castes there but they had left over the years and we were the only Teli
family left. We weren’t treated well by the other villagers because we
were the only non-Adivasi there. We didn’t feel right. We managed to rent
some land in Kedanar [after our land in Somutola was taken away] but we
didn’t want to stay there because that too is a Gond village and none of
our caste-fellows live there. Markakasa has Telis so we moved here.”

When I asked why they had moved from Alewada in the first place, the old man Jhetmal
told me that the family had been suffering ‘pain’ (dukh) there. When I pushed him on the
kind of dukh, he said that his family had been the targets of magic (jadu-tohna)
performed by other villagers. When I repeated this to Didi, my landlord’s sister, she said

100
that that was rubbish: the reason why the family had left their village was that Old Man
Jhetmal is ‘bad’ (kharab), a practitioner of magic (jadu khor) and the Alewada villagers
had driven them out. Of course, it is impossible to know which version of the migration
story is true. What is interesting is that in both accounts witchcraft and a dispute feature
as the reasons for moving.

From 1987 to the present day, the Jhetmals have rented the land of others, mainly Gonds
in Markakasa and Kedanar, and are regarded as being very successful. They are admired
by many other Markakasa people for working hard on land that Gonds could not make
productive, and the amount of money they have in their bank accounts is the subject of
much speculation. In the year 2003-04, the four brothers produced between ten and
thirteen sacks of rice per acre farmed. And, over the course of several years, they have
between them acquired three teams of bulls and three carts, which represent major
investments and signifiers of wealth for Markakasa farmers. Nevertheless, the Jhetmals
buck the pattern: renters of land are poorer than those who own. Certainly no other
landless villager who farms, or farmer with small landholdings of less than an acre, owns
bulls or carts. Whereas those whose land-holdings in the village stretch back two or more
generations produce between ten and fifteen sacks of dhan (unhusked rice) per acre of
land farmed, those who have acquired land more recently from non-Adivasis, or who
farm rented land, produce between seven and nine sacks per acre.66

Those who migrate to Adivasi areas (in Maharashtra) post-1974 face a chronic shortage
of land. The best farmland is often owned by Adivasis who, watched by the village
talathi (government land registrar), are unable to sell. The process of being allocated land
by the state is lengthy and expensive, and the land is largely of poor quality. Migrating in
order to escape the attentions of witches and attacks of magic in the course of a dispute
has become far more difficult, because the security of a living is no longer guaranteed.
Those who live exclusively by leased land need to work much harder, and the prospect of
buying land has significantly diminished. Though the Jhetmal sons are keen on

66 One sack o f unhusked rice (dhan) weighs in at about 70kg. The 2003-04 rate for the ordinary grade rice
which the overwhelming majority o f farmers sow was between Rs. 480 and Rs. 520 per quintel (100kg).
The figures quoted above are based on a household survey I undertook between March and May 2004.

101
eventually acquiring land of their own, their elderly father, perhaps affected by the
experience of having land taken away from him, is more equivocal: “If you’ve got gur
f\ 7
[molasses], it’ll attract flies; no gur, no flies”.

Migration did occur in the past, however, and on a large scale. The result in Markakasa is
the creation of a panchrangi gaon, a ‘mixed village’ where people of twelve castes and
four languages live side by side. The founding story of the settlement is clear: Burri
Netam left his home village of Barritola to the south-east (now in the state of
Chhattisgarh) and cut down some of the jungle after obtaining permission from the local
malguzar. His was the only house at first, but he was soon joined by more Gaita Gond
families and then also by a Mahar. The arrival of waves of Chhattisgarhi (and Marathi)
immigrants has radically altered the village and the way it does things. The ghotul-kuria,
a dormitory for youth, ceased to function at some point in the 1930s, a result of the influx
of other types o f people, I was told (discussed further below). Similarly, the lingua franca
of the village has now become Chhattisgarhi, whereas in the recent past it was Gondi and
even the non-Gond castes spoke it. On our return from a Gond meeting in an all-Gond
village, where Gondi was the only language spoken throughout the proceedings, my
landlord lamented (making a comparison): “Our village has become so dhamdi
[derogatory term for Chhattisgarhi people], even the Gonds are forgetting how to speak
their language. It’s not pure anymore, it’s mixed with Chhattisgarhi.” The presence of
other castes with other world views meant that certain measures that were in the past used
to combat witchcraft collectively at the village level are no longer available. One
important example is the Angadev, a deity that can detect witches but cannot be deployed
in a mixed village such as Markakasa, ostensibly because of differences in opinion over
its efficacy.

Angadev: the collective response to village dis-ease

The principal weapon the village as a collective has at its disposal in combating witches
and ghosts (bhut) is to request the services of an Angadev (or in Gondi, Angapen), a

67 In other words: having something will ensure it is coveted by others; having nothing solves the problem.

102
powerful Gond deity that is adept at detecting troublemakers and unquiet spirits. The
Angadev is called to uncover the causes of suffering felt by the village as a whole, such as
in cases of illness or death of a large number of young people. But the Angadev will only
visit the village if the village has, as a whole, invited it to come; it will not attend if there
is the slightest hint of dissension or unwillingness. In a ‘mixed village’ such as
Markakasa this unity of purpose has proved hard to come by and, despite repeated
attempts by certain groups within the village to organise a visit by the Angadev, other
groups have been equally persistent in their opposition to its presence. The unity
demanded by the Angadev is significant. A village which invites the god acknowledges
that it is somewhat disunited, and that there are within the body of the village people such
as witches and others who are causing harm. In a sense, therefore, the arrival of the
Angadev is evidence that disunity has already been overcome, and its presence is a trace
of prior processes of negotiation which re-established village cohesion. Unity of purpose
also has a more practical aspect: inviting the Angadev to one’s village is extremely
expensive and all villagers need to be agreed in order to contribute money. The villagers
have to bear the cost of transport for the deity and its attendants, animals (goats, pigs,
chickens) for sacrifice and feasts, and alcohol. Depending on how long the Angadev
stays, the whole exercise can cost between Rs. 20,000 and Rs. 50,000, an enormous sum
of money.68

While I was conducting fieldwork in Markakasa, an Angadev was invited to the

neighbouring village of Gujarbarga, located 3km away along a dirt road. Though I had
heard descriptions of it from my informants, I was thus fortunate enough to be able to see
it for myself. The Angadev resembles a wooden bier69, comprising three long, thick
wooden poles crossed by two smaller pieces that connect the others across its width. The
middle length of wood is said to be imbued with power (shakti) and is carved at the front

68 This converts to approximately £250 and £650 respectively. The daily wage for a (male) farm labourer is
Rs. 25 (30p).
69 The Angadev's similarity to a funerary bier is noted by Elwin (1943), who suggests that the actual tool o f
divination may in the past have been a corpse. That the word anga means or suggests ‘body’ in most north
Indian languages confirms this for him (anga [body] + Gondi. pe«/Chhattisgarhi. dev [god]). In Markakasa
the use o f a corpse for divination is known as heyr and is a distinct ritual. During heyr, the corpse o f the
person struck down by witchcraft is borne by four bearers who are guided to point out the wrong-doer.

103
with the head of an animal - in the case known to me, of a horse. The latter is specifically
referred to as the pen or dev (god). The Angadev that I saw at Gujarbarga was decorated
with a large number of peacock feathers at the places where the length and the width
poles crossed; the god itself (the head) was painted bright orange, and the poles were
black. This particular Angadev had come from its shrine 40km away to the south,
accompanied by several baiga. Elwin (1943) writes that the Angadev is a Gond clan god:
the Gonds are divided into a number of clans, the clan name being what other Indians
would call the ‘family name’ (or surname). Each clan has an ultimate home territory
(bhumi), a place of origin, and this is where the Angadev resides70. Gell (1980) does not
dwell on the clan aspect of the god and refers to the Angadev as ‘log-gods’. In
Markakasa, the Angadev is regarded primarily as a tohni (witch)- and bhut (ghost/spirit)-
detecting device, and classed by Adivasi and non-Adivasi alike as belonging to Gond
sanskruti (culture).71 The Angadev is carried by four bearers, who describe being driven
by the dev to places where witches live or ghosts are to be found. The landscape is no
obstacle for the deity: Angadev have been known to take their bearers through lakes, into
wells, far out to threshing floors and up onto the roofs of houses in search of bhut and
tonhi.

I had conflicting reports about why the people of Gujarbarga had invited the Angadev.
Some in Gujarbarga denied that there was anything currently wrong in the village and
asserted that the Angadev was coming in order that it stayed that way. Thirty years ago, I
was told, a lot of cows in the village stopped giving milk and started dying, and many
people were sick. The Angadev visited, rooted out the tonhi and bhut that were causing
the problems and everything has been fine ever since. At that time, the elders of the
village promised to ask the Angadev to return again in the future. The present-day elders
had decided, therefore, that now was the time. Others, however, assured me that the

70 The idea o f bhumi was not particularly well articulated among the Gonds in Markakasa; it was regarded
as specialist knowledge that only baiga would know. A couple o f households have Gond calendars, which
contain a list o f all Gond surnames (or clans) and their bhumi, but for the Gonds o f Markakasa this seemed
to be little more than a curiosity. For them it was more important to know die ‘home’ village o f their
particular lineage or family, for this is where their household gods (lonchtun) reside.
71 See Sundar (2001: 437), who describes how the state facilitated witch-hunting by renting out the
Angadev housed in the Maharaja o f Bastar’s palace at Jagdalpur. She, too, states that the Angadev is a clan-
god o f the Muria and Maria Gonds.

104
Angadev had been called to detect existing trouble-makers, and that the solution provided
thirty years ago needed to be sought once more.

Plate 15: The Angadev with attendants

Plate 16: The Angadev at ‘p la y ’ (khel,), driving its hearers

105
106
There are three principal stages of the Angadev's visit to a village. The first is the binding
of the village’s boundaries (Chh. sima bandhna), whereby the Angadev travels the length
of the borders, stopping several times along the way while the attendant baigas sacrifice
animals to it. The binding is said to be done in order to prevent residents leaving the
village while the Angadev is in residence. Informants described it as a powerful spell that
would kill or harm anyone trying to leave. All the inhabitants of a village visited by an
Angadev must be in residence for the duration of its visit; people working as labourers in
nearby towns are contacted and instructed to return home. This residence rule is
extremely important and if the Angadev is not convinced that all inhabitants will be
present, it is said that it will refuse to come at all.

The second, and central, stage is the detection of witches and ghosts (Chh. tonhi-bhut
khojna). The Angadev goes from house to house searching for wrong-doers - witches and
ghosts. When it detects a witch72, it stands in front of him and knocks him with one of the
poles that make up its bier. Interestingly, there was lack of agreement as to the
significance of this action: some people said that this was merely the act of identifying
the witch; others believed that the blow was actual punishment and that the witch would
eventually die as a result.

The third and final stage is ‘play’ (Chh. khel; G. karsana) and is the entertainment aspect
of the visit: the Angadev puts on a show. A crowd of spectators gather in the village
square, and as the drums beat the Angadev rushes around swaying and dancing,
sometimes bumping into members of the audience. Here, however, there is no risk of
being accused of witchcraft: the Angadev is merely ‘playing’. At the same time, induced
by the drumming, the attendant baigas and certain other people, both villagers and
visitors, begin to get possessed by the gods (Chh. dev jhupna), who come to ‘play’
alongside the Angadev. The scene is incredibly chaotic. Men possessed roll around on the

72 Technically a witch can be male (tonha) or female (tonhi), but the former term is less common in daily
usage than the latter, which is often used to refer to male witches too.

107
ground, jump into the crowd or flay themselves with barbed chains and all the while the
Angadev is running and swaying.73

Of these three stages of an Angadev visit, I was only able to witness the final one. But
this was not through lack of trying. Though I had confirmed the dates of the visit on a
number of occasions, and the Gujarbarga people I spoke to led me to believe that I would
be able to witness the detection stage as well as the ‘play’ stage, they had in fact misled
me and other ‘outsiders’ (people from other villages) and given us all the wrong dates. I
arrived in Gujarbarga at the appointed time accompanied by several friends from
Markakasa. To our dismay we were told that the first two stages had been carried out
during the night when only the Gujarbarga villagers were present and that all the ‘work’
(kam) had been done; all that was left now was the element of spectacle. We visited
several households (those with whom my companions had some kinship or friendship tie)
and asked questions about what had occurred the night before, but received no
satisfactory information. We watched and returned to our village by early evening, and
the talk in the teashop was dominated by how no-one who had been to Gujarbarga had
understood what had gone on. To the extent that we can learn a great deal about a given
social situation by ‘reading’ that which is lied about or left unsaid, that the Gujarbarga
villagers had done this was extremely interesting. Why would they lie? Naturally, I was a
little put out when I discovered that I would not be seeing any witch-finding and I asked
an acquaintance, a man called Girdari, who was the Gujarbarga sarpanch (elected
headman), to explain why I (and others) had been misled. The reason, he said, was that
the activities of the Angadev in Gujarbarga were a secret matter, and secrecy needed to be
maintained in relation to two groups: people from other villages and the police. As
Girdari explained, the concern with keeping things from people from other villages was
bound up with potential stains on honour and reputation:

“It was held at night because we didn’t want people from other villages
here. If word gets out that such-and-such person’s house had a ghost in it

73 A. Gell (1980) has an interesting discussion o f the Angadev's characteristic movement in the context o f
an argument about the importance o f vertigo and dizziness in Hindu-Gond ritual practices.

108
 or that so-and-so’s wife is a witch then the reputation of that household
suffers in relation to outsiders. They might have problems marrying their
sons and daughters. So it’s best to keep these things within the village;
outsiders ought not to know. We don’t want the village to get a bad
name.”

Thus it would seem that though the invitation of an Angadev to one’s village is a sign to
the outside world that something is not quite right within and needs to be dealt with, the
honour of the village and of the people in it requires to be maintained nevertheless. The
Angadev's insistence on village unity is striking. Not only does it bind the village along
its boundaries, thus identifying it physically in terms of geography, but it also prevents
any villager from being absent, which marks the village socially. In a very real way then,
the Angadev, through ritual, brings into being or confirms the existence of the body of the
village, a body which is described as having fallen sick.

Girdari’s desire to keep the Angadev's visit secret from the police is far more complex,
and is indicative of the type of impediments that exist to combating witchcraft and magic
at the village level. The problem was that Gujarbarga had been refused the required
police authorisation to invite the deity. In the particular case of the Angadev, villagers
seek assurances from the police that those who accuse others of witchcraft will not be
subject to criminal investigation. The villagers in return undertake not to use violence in
dealing with a suspected witch. Essentially, the police are asked to adopt a ‘hands-off
attitude and the villagers are given control over witchcraft accusations and their
consequences.74 Just as police permission was refused for Gujarbarga, so it was denied
about six months later for another all-Gond village, Piparkhar, about 10km from
Markakasa. Young people there had been falling ill and dying at an unusually high rate,
perhaps eight in as many months, and no medical explanation had proved satisfactory to
Piparkhar’s inhabitants. While negotiations were taking place with the guardians of an
Angadev in Bastar, the police refused the villagers’ request. On learning of this, the

74 See Macdonald (2004: appendix) for an example o f just such an agreement between police and villagers
in plains Chhattisgarh.

109
Angadev guardians declared that they would not attend a village which had not received
official permission. Police approval seemed to be increasingly difficult to come by. In the
past, the Angadev would come and go and asking for permission was just a formality; in
most cases the village would not bother to inform the authorities at all. The police were
stationed far from the village and visited infrequently. Why were the police no longer
granting permission for Angadev visits? I will return to possible answers to this question
below. Let us for now consider why, unlike in Gujarbarga which is an all-Gond village,
the use of the Angadev to combat village dis-ease in mixed villages such as Markakasa is
no longer (at least publicly) considered.

As we saw above, the Angadev requires that a village be united in its desire to invite the
deity. One reason for this requirement might be so that the Angadev is not used in a (at
least overtly) political way against a particular household, family or faction within the
village75. This kind of unity, according to people in Markakasa, was difficult to come by
in a mixed village (panchrangi gaon) and so the dis-ease that afflicted it was difficult to
combat. Many people I spoke to confirmed that “things were not right” in Markakasa and
that there was no solution. Bhanu, the owner of the Markakasa tea-stall and a
Chhattisgarhi Rawat76, gave an example of the village’s malaise.

“About fifty years ago when I was a child,” Bhanu narrated, “Markakasa had plenty of
cows and plenty of milk; indeed it was famous throughout the area for its milk - ask
anyone: people used to come here from villages all around for it. Then suddenly over a
period of a couple of years all the cows began to get ill and died.”
“Surely”, I said, “they just contracted some illness (bimari) that was going around.”
“No, this was not a disease. How can you explain the fact that cows were kicking away
their calves, not letting them suckle and leaving them to die? That’s no disease. Do you
see any milk in Markakasa today? We’ve got the whole forest to graze our cattle and still
they can barely produce half a glass. Surely this is a sign that something’s wrong here.”
“What’s wrong?”

75 It is acknowledged, however, that the bearers o f certain Angadev can be bribed into identifying certain
people as witches or troublemakers.
76 TTiere are two types o f Rawat (traditionally cattle-herders) in Markakasa - Chhattisgarhi and Gondi.

110
“Some people here are bad (kharab). That’s why the village has been ruined (gaon
bigadgehein).”

There were a number of village men sitting at the tea-stall as this conversation took place,
listening and agreeing with what Bhanu said. A Gond man, Buddharam, with whom I had
had several conversations about the Angadev previously, joined in the discussion:

“Remember what I told you about the Angadev. Well, about three years
ago, some of us wanted it to come here but there was no agreement. Some
people, the Telis, said that the Angadev was a waste of money and a load
of superstition (<andha-shradda), and that it couldn’t find tonhi or bhut.
But that’s all rubbish. They’re just worried that the Angadev will find out
what they’ve been doing. We’ve been trying to invite the Angadev for
years. But because the village is not united, we cannot bring the Angadev
here.”
\

Though it was certainly not the case that all Telis regarded the Angadev as evidence of
‘superstition’, when I discussed the Angadev visit to Gujarbarga with several Teli men,
the prevailing attitude appeared to be one of condescension. The Angadev was
essentially a hoax, they said, and the Gujarbarga Gonds, in spending such a vast amount
of money in inviting it, were prisoners of their own ‘blind faith’. But the Teli opposition
to the Angadev also stemmed from anxieties about the risk of being singled out by the
deity as practitioners of witchcraft.

In much of the area around Markakasa and in other parts of Chhattisgarh, the Teli caste
has a particular reputation for employing ghosts (bhut) and for its women to be witches
(tohni or saude). Villages in Chhattisgarh with majority Teli populations were regarded
by other castes as risky places in which to marry daughters, who might either become the
target of a witch or, more likely, would be tempted into learning the ways of magic (jadu)
themselves by becoming someone’s apprentice. To my surprise, the idea that Telis (and
in particular Teli women) have a greater propensity to be witches was shared by some of

111
my Teli informants, too. The presence of a large Teli community in Markakasa, then, is
regarded as problematic. Not only is witchcraft more likely to be performed and ghosts to
be employed by such people, but their presence makes the use of anti-witchcraft
measures such as Angadev more difficult, if not impossible. This difficulty, which is the
result of living in a mixed village with other castes, is replicated at the household level
where, among Gonds at least, the use of a pre-wedding ritual to test a new wife or
daughter-in-law for witchcraft has been abandoned because of what it connotes about
Gond sexuality.

The demise o f panbuddi: sex and marriage, detecting witches and ‘looking bad’

There are impediments which exist at the level of collective village action against
witchcraft and dis-ease, the most striking being the difficulty experienced in employing
the Angadev. But impediments exist at the individual level too, and are intimately bound
up with ideas of ‘improvement’ and of ‘looking bad’ in front of other castes and other
types of people, ideas which emerge out of the experience of living in a ‘mixed village’.

One of the more individual ways in which to test for witchcraft is a ritual called
panbuddi. Like the Angadev, this is a ritual identified as Adivasi or Gond; unlike
Angadev, however, it is only Gonds who perform it. Panbuddi is carried out the day
before a man and woman are married (for shadi). It has two purposes. The first is to
impress upon the couple the importance of their exclusive commitment to one another.
Among Gonds, in sharp contrast to many of the other castes whom they live alongside,
virginity is not generally regarded as a burden (cf. Parry 2001: 795-6), and neither men
nor women are assumed to be sexually inexperienced when they marry. By undergoing
panbuddi, the couple formally renounce any previous ties of affection they may have had.
One man explained it thus:

“The baiga [Gond ritual specialist] asks them both to come out to the
maidan [open space] and there they walk around him seven or nine times'
while he says a prayer and sprinkles water over them. He makes them pure

112
 [shuddh] from anything they may have done in the past in their villages
before their marriage. They have to forget the past [their previous loves]
now that they are married.”

Panbuddi also serves a second purpose: that of detecting witches. During the course of
the ritual, the bride-to-be is tested by the baiga to determine whether or not she has learnt
witchcraft, or whether she is associated with any spirits/ghosts {bhut) that may work for
her and that she would bring to her new household. The test is performed on behalf of the
groom’s family and the baiga is employed by them. Concerns about possible witches are
rather prominent in choosing a new daughter-in-law, and extensive enquiries are made
through contacts in her natal village to detect any rumours about her or her family,
especially if the natal village is one with a reputation for witches and ghosts. Despite
these enquiries, many families also ask that a panbuddi be performed before the actual
wedding ceremony takes place. Panbuddi has, however, ceased to be practiced in
Markakasa and in several surrounding villages. The reasons given to me by Gond
villagers for its abandonment is that it is considered as ‘looking bad’ (Chh. accha ni dike)
in both its aspects: that it is a public yet highly individual test for witchcraft, and that it
affirms an ideology of sex and marriage that runs counter to that held by other groups.
Let us look a little more closely at this latter point and how the expression of a different
ideology might appear shameful or less prestigious.

Some of the most frequent questions I was asked by villagers about life back in Britain
were about marriage: how, when, why and whom do we marry. Playing the good
anthropologist and trying desperately not to generalise, I would reply that some married
at nineteen or twenty, others not until their thirties or later. I added that recently more
people were choosing to cohabit before marriage and would even have children before
any wedding might take place. Men and women married for love (prem), I said. Rather
than be shocked by this or find it incomprehensible, as I had experienced in other parts of
India, villagers of all castes would comment half-jokingly that British people seemed a

113
 77
lot like Gonds in this regard . Gonds, too, were allowed (i.e. they would not suffer any
punishment by their caste council) to live together without getting married and would
70
also often have children out of wedlock. This Gond ideology of sexuality and marriage,
in relation once again to specifically Gond activities (in this case, of dancing), found
expression at a meeting I attended held at a neighbouring village.

In most all-Gond villages in the area, adolescent boys and girls from the age of twelve to
nineteen (or older in some cases) regularly meet to practise dancing, singing and
drumming79. Mahua daru, the local brew, is often consumed on these occasions by both
girls and boys. These dances are then performed at Gond weddings and funerals80. In the
past, dancing and singing would take place in the area outside the ghotul, a dormitory
principally for boys. Though the ghotul had ceased to exist as an actual building in all the
Gond villages I visited during the course of fieldwork, this aspect of adolescent
association seemed to have survived in the form of these dancing groups. In most of the
mixed villages such as Markakasa, the Gond youth no longer danced or sang at all, even

771 consistently encountered the view that the English language and Gondi had some sort o f natural
affinity, perhaps in part because o f the difference in rhythm and tone between these two on the one hand
and the relatively similar sounding group o f Marathi/Hindi/Chhattisgarhi on the other.
78And love (prem) appeared to have value as a basis for marriage, though it seemed to have different
meanings depending on the context. On one level, prem was regarded as something rather risqu6, not
respectable and also somehow disrespectful to the families concerned on both sides. But this was a rather
narrow, sensationalist view o f a ‘love marriage’. As a result o f this connotation, people would often talk
about prem in terms o f the slightly colder, clinical phrase man-pasand (lit. ‘agreeable to the mind’), a
flexible usage that covered all sorts o f motives, from what we would understand as romantic love to merely
seeing the prospective spouse and liking the look o f him or her. Talking to two Gond young men, Hemu
and Latchu, about love and marriage one day, Hemu said that Latchu and his wife Mukta had had a ‘love
marriage’. Latchu’s response was immediate. “No, it wasn’t like that at all; w e’d been ‘talking to one
another’ for a while but I went to her house and spoke to her parents (i.e. asked their consent for the
marriage).” In effect, Latchu and Mukta did have prem before they married but Latchu did not want me to
think that he had eloped with his wife or acted disrespectfully.

More o f the Gond elders than those o f other castes had married their wives as a result of paithu. Paithu
occurs when a woman enters the house o f her prospective husband and demands that he marry her. It is
then his responsibility to send word to her parents o f their intention to marry, and also pay a special bride-
price known as sukh (ordinary brideprice is karchi). Often, this ‘entering’ will have been pre-arranged
between the lovers, who have in effect accomplished an elopement.

79 These include the ‘rello’ and ‘hulki’ dances described by Elwin (1957) and Simeran Gell (1992). These
meetings for dancing and singing were collectively referred to as ‘play’ (G. karsana).
80 Indeed, this is how I discovered the continued existence o f the ghotul, an institution I assumed had
completely died out. Attending a Gond wedding in Markakasa, I watched a group o f young men and
women from a neighbouring village dance, and I was struck by how the dancing boys and girls sat together
all mixed up to eat their meal at the wedding feast in a way that Markakasa young people certainly did not.

114
 o 1
though a ghotul building may have existed in the past (as indeed it did in Markakasa) .
The responses to questioning about why Gonds in these mixed villages no longer danced
was fairly uniform: “that in front of other castes, boys and girls dancing together didn’t
look good (accha ni dike) and they laughed at us”.

Mahaka is an all-Gond village located about 6km from Markakasa. A meeting of local
Gonds from forty villages across three districts (Gondia, Gadchiroli, Rajnandgaon) had
been called there to discuss a relationship between a Mahaka boy and girl.

The couple had started ‘talking to one another’ while dancing and struck up a
relationship. This went on for approximately three years and, towards the end of that
period, the girl became pregnant and gave birth to a girl. They began living together in an
abandoned house in the village and maintained cordial relations with their respective
parents. The rest of the village was not particularly concerned that this couple were living
together or had had a child before being married; the meeting had been called because the
villagers believed the couple were committing incest. Both the boy and girl were from the
sath-dev (seven-god) exogamous division; as such they were to regard one another as
brother and sister (bhauband). In past village meetings, the couple had been repeatedly
called upon to renounce one another but they had refused and been thrown out of caste.
Now the council of village elders (panch) convened this general meeting to force a
solution, armed with the knowledge that the boy had changed his mind, was willing to
break the relationship and return to the Gond caste fold. This general meeting took no
issue with the pre-marital relationship or with the birth of the child. Their sole concern
was the breaking of the union of an incestuous pair, and it was for this that they were
both eventually fined. The assembled Gonds did not offer commentary on the morality of
the relationship except to criticise the transgression of the incest rule. The connection
with a ‘typically Adivasi’ activity such as dancing is significant. In the close associations
and love affairs (though not necessarily eventually marriage) that it promotes, it

81 When I asked one o f the village elders about the ghotul in Markakasa, he said it still existed in the form
o f a Gond Youth committee set up in the 1980s, which includes all unmarried but mature Gond boys and
girls. Its principal function is to organise the serving o f food and refreshments during Gond weddings.
Interestingly, the Gonds are the only caste in Markakasa in which both young women and men serve guests
at a wedding; the other castes use men only o f any age.

115
expresses the relatively unproblematic nature of pre-marital sex for Gonds , a notion that
would of course ‘look bad’ in front of other castes holding contrary views in a mixed
village.

A brief exposition of child marriage and bride-service also suggests a divergence in

attitudes to sexuality. During discussions of marriage with villagers (of all castes), it
would always be impressed on me that the Chhattisgarhi castes (and especially the Telis)
married their children very young, in the past at the age of eight or nine and nowadays as
young as fourteen (though more usually at eighteen or nineteen).83 Gonds did not do this,
I was told: they waited until children were accha jawan (fully mature), and had always
done so. As Parry (2001: 795-97) has observed, the performance of child-marriage in
Chhattisgarh in the past expressed the idea that the parties to a marriage ought to be
virgins, though he also makes clear that, as a result of the time lapse between marriage
and actual cohabitation, one did not necessarily need to lose one’s virginity to one’s
spouse. Though it was now seldom performed, Markakasa people reported that, in the
absence of performing a pre-pubescent marriage, Chhattisgarhi castes married the girl to
a pot or an arrow (see also Dube 1953; Gell 1992: 245-247; Parry 2001: 796). While
some interpret this as a ‘holding operation’ which attempts to control the girl’s
burgeoning sexuality (Gell 1992: 246; Yalman 1963), Parry (2001) suggests that the
primary marriage (shadi) in Chhattisgarh in fact liberates female sexuality. He reports
that after shadi is performed, girls are free to have sexual relations, unburdened by the
need to be chaste. In contrast to Gond attitudes, sexuality is only released after marriage.

That the Gond notion of sexuality differed from those of the Chhattisgarhi castes is also
suggested by the fact that Gond men who performed bride-service before becoming

82 Though see S. Gell (1992) for a discussion o f the tensions that the existence o f the ghotul sustains
between love, closeness and marriage.
83 Though I conducted a survey on marriage age, I suspect the findings were not all that reliable: many
older Telis, even those who, I learnt from other sources, had had child-marriages would, on seeing my
notebook and pen, insist that they had got married at eighteen or twenty. I felt that, in these cases, more
informal methods o f questioning on this topic resulted in more accurate information. Though child-
marriages were no longer performed, o f the marriages that had taken place more recently, Chhattisgarhis
(both girls and boys) married at a younger age than Gonds, and there were no Chhattisgarhi men over the
age o f 20 still single, and no women over the age o f 18, in stark contrast to Gond men and women.

116
Plate 18: A Chhattisgarhi Teli bridegroom being anointed with turmeric
the day before his wedding.

117
lamsena (uxorilocal husbands) were permitted to have sex with their wives-to-be without
the performance of any prior sacrament.84 In the case of one lamsena couple resident in
Markakasa, the ‘wife’ was pregnant and the marriage had yet to be celebrated. One
cannot assume, either, that this sexual access is granted in the full expectation that the
lamsena will marry the woman intended: it is not unheard of for lamsena either to leave
of their own accord or be kicked out by the erstwhile prospective father-in-law during the
period of bride-service. Significantly, I was told by several Teli men that, though men
could become lamsena in their caste, it very rarely happened, and they would not perform
bride-service. Out of the seven current lamsena in the village, none were Teli (or indeed
of any other Chhattisgarhi caste), and while Gond and Mahar genealogies were replete
with lamsena, the Teli ones I collected contained none.

That Gonds see no reason for child-marriage or for suppressing sexual access during
bride-service suggests that the notion that the parties ought to be virgins has little
importance for them. Faced, however, with a ‘mixed village’, which contains a majority
of Chhattisgarhi people for whom the virginity of their offspring is of concern, the public
affirmation of a counter attitude through the performance of panbuddi leaves Gonds
‘looking bad’. Unfortunately, since panbuddi is also a test for witchcraft it also leaves
Gonds without a convenient and useful method of testing for potential witches.

The ‘mixed village’ then, or at the very least the idea of it, makes the control of witches
and magic more difficult, both at the level of the village and for the individual household.
For the village, the lack of unity of opinion about the Angadev militates against its
invitation; for households, especially Gond ones, certain measures to detect witches
among in-marrying women appear less attractive because of what the ritual connotes
about (female) sexuality.

84 This sexual access before marriage was not entirely free o f problems, however. Sometimes the father-in-
law would govern rather strictly the frequency o f sexual relations, leading to conflict with the lamsena.
This is what happened in the case o f the lamsena and pregnant lamsenin I refer to above: a dispute which
arose partly over sexual access led to the lamsena threatening murder and suicide.

118
From the idea of the mixed village, let us turn to the third reason why witchcraft
detection and control in this part of India may have become more difficult: the increased
presence of the state and of the police in response to the Naxalite insurgency. Recall the
comment of the Gujarbarga sarpanch, Girdari, who said that the visit of the Angadev was
to be kept secret from the police since permission had not been granted. How has the
relationship between Markakasa villagers and the state changed, and what impact has that
had on the control of witchcraft?

Police, Naxalites, and the State

Over the past fifteen years, the eye of the state has become increasingly keener in this
relatively neglected and under-governed part of India. This is largely a result of the
presence and growth of a violent Maoist-inspired revolutionary movement, often referred
to as the Naxalite movement. The police (as representatives of the state) have literally
come physically closer to villages such as Markakasa and Gujarbarga than ever before,
and bring with them an alternative world-view which is unwilling to tolerate practices
(such as the hunting of witches and bhut) that reek of superstition or ‘blind-faith’ (andha-
shraddha), and are antithetical to their idea of what constitutes ‘modem India’. In her
recent study of witchcraft accusations and violence in Chhattisgarh, Macdonald (2004)
examines the differences that exist between the higher and lower ranks of the Indian
police force regarding belief in these types of being: the former subscribe to a ‘modernist’
vision which takes a dim view of belief in witches and magic, regarding it as a throwback
and part and parcel of ‘traditional India’; the latter, on the other hand, are more willing to
countenance suspicions of magical activity and share in the belief as to their existence
(131). Though her argument is entirely convincing and applies equally well to police in
the area of my fieldwork, it is premised on the fact that, in much of plains Chhattisgarh
(and I suspect in rural India more generally), the lower-ranks of the police are drawn
from similar backgrounds to the citizenry they serve. In the Adivasi districts of eastern
Maharashtra however, and particularly in the Adivasi dominated areas of southern
Gondia and Gadchiroli districts, there is a sharp difference between the lower echelons of
state administration (police, teachers, forest officials) and the local population.

119
Recruitment of the latter into the former, though increasing, is at much lower levels than
in other non-Adivasi districts, and thus policemen and others are overwhelmingly
‘outsiders’, people who regard themselves, and are regarded by locals, as coming from ‘a
different area’ with ‘a different atmosphere’ (vatavaran).85 Despite Macdonald’s
evidence that objectively witchcraft is not simply an Adivasi or jangli issue and that it is
just as prevalent in much of non-Adivasi plains Chhattisgarh, the fact of the matter is that
members of the local bureaucracy and police in Adivasi areas do, as a matter o f
subjective understanding, believe that it is more of a concern in these ‘backward’ areas
where they work. This attitude predominates regardless of lower-level policemen’s own
beliefs in witchcraft and magic. Confronted with people paradigmatically regarded as
backward and superstitious, the policeman’s response is to deny any similarity.

The Naxalite movement began in the late 1960s in eastern India as a student-led agrarian
rebellion agitating for land reform. It exploded into large-scale violence between 1970
and 1971 in Bihar, West Bengal, southern Orissa, northern Andhra Pradesh and Punjab,
which receded after the death of the movement’s charismatic leader Charu Mazumdar in
1972 (Singh 1995: 133). The People’s War Group (PWG) was founded in 1980 in
Andhra Pradesh and over the next ten years emerged as the most formidable Naxalite
formation in the country (ibid: 106). The founder of the PWG, Kondapalli
Seetharamaiah, was arrested in 1984 but after orchestrating a dramatic escape set about
consolidating and expanding the movement. It spread its area of operation into eastern
Maharashtra and southern Chhattisgarh (erstwhile Madhya Pradesh), both predominantly
Adivasi areas. Its ultimate aim appeared to be the elimination of the activities of the
Indian state in areas it controlled. The PWG programme consisted of the redistribution of
land; enforcing the payment of minimum wages to farm labour; imposing taxes and
penalties; holding people’s courts; destroying government property; kidnapping
government functionaries; attacking policemen (ibid: 108). By the early 1990s, Naxalite

851 was unable to obtain official statistics on the percentages o f policemen who were ‘locals’ as opposed to
‘outsiders’. Based on my own interviews with junior policemen, however, I would suggest that no more
than five percent o f the personnel in either the police station at Chichgarh or at the armed outpost nearer
Markakasa were from Gondia or Gadchiroli Districts. Most were from other parts o f Vidarbha or from
western Maharashtra. Conversely, o f the three Adivasi men from Markakasa who were policemen, two
were stationed in non-Adivasi areas in the northern part o f the District.

120
violence peaked in Maharashtra with 30,22, and 37 deaths (police and civilians) recorded
in 1991, 1992 and 1993 respectively, and by then the PWG had gained access to
landmines as well as guns (ibid: 133-4). The turn of the century has seen a renewed
growth in Naxalite activity in Maharashtra, spurred most recently by the influx of Maoist
or
fighters from Andhra Pradesh following the ceasefire there . The figures for Maharashtra
are as follows:

2004 2005 2006 (to June)

Incidents (Naxalite) 84 94 56
Civilians killed 9 29 24
Policemen killed 6 24 1

Source: Parliament Q & A Home Ministry, 2nd August 2006. See

http://l64.100.24.219/annex/208/A U952. htm

In 2004, the PWG merged with another important Naxalite faction, the Maoist
Communist Centre (MCC) which operates principally in Bihar and Jharkhand, to form
the Communist Party of India (Maoist). Their objective has been the creation of a
Compact Revolutionary Zone stretching from the Nepalese border in the north through
the spine of central India to Tamil Nadu in the south.87

While the Naxalites declare that their struggle is against the perceived oppression of the
Indian state and its agents and their ultimate aim is the liberation of the Adivasi peasantry
from a neo-colonial yoke, the growth in Naxalite activity has succeeded, at least in this
part of India, in bringing the state closer to its citizens. While this is certainly not the
case in the ‘Liberated Zones’, such as parts of southern Chhattisgarh, where the Naxalite
presence and dominance appears so overwhelming that the state (as schools, police,

86 A ceasefire which has since broken down.

87 Since 2005, the Naxalite presence in southern Chhattisgarh has been met with a violent response by a
vigilante movement known as Salwa Judum, a people’s militia sponsored by the state government. The
intense fighting has led to the deaths o f many people, and an estimated 40,000 people (mostly Adivasi) are
internally displaced, now living in refugee camps. See Sundar 2006b.

121
health care, roads and transport) has effectively withdrawn (though see footnote 88), in
areas where there is more of an engagement between the two sides, there appears to have
been an intensification of state activity. Shantha Sinha (1989) observes the same process
in rural Andhra Pradesh where, as a result of greater and closer police and administrative
contact, “people for the first time realised that their standard of living depended not, as
they had so far believed, on the local landlord or money lender but on a much larger and
infinitely more powerful entity - the State” (317). An example of this in Markakasa was
the construction of a new network of roads in the area by the Border Roads Organisation.

One of the largest public works projects undertaken by the State in this area in recent
years was the building, repairing and tarring of the road network. The work was carried
out by a central government agency, the Border Roads Organisation (BRO), a military
organisation headed by a Lieutenant-Colonel, and organised into units which complete
the commissions. Renowned throughout India for its work on building roads in the most
inaccessible and dangerous parts of Kashmir and the insurgency-hit states of the North-
East, and along India’s tense borders with Pakistan, China and Bangladesh, the BRO was
contracted by the government of Maharashtra to build a network of all-weather roads in
the Naxalite-affected districts of Gondia and Gadchiroli. A network of tarred roads would
allow better access for police and security forces in areas where rebels were known to
operate, and also minimise the risk of landmines which could be hidden more effectively
on dirt or stone roads. Moreover, the BRO as a military organisation had soldiers at its
disposal to guard the construction work from attack by rebels opposed to the project. The
work was carried out in phases and is still going on in both districts and across the state
border in Chhattisgarh. While the BRO supplied the machinery, materials and engineers,
labour was recruited from the areas in which the roads were built. Many Markakasa
villagers, both women and men, were employed as labourers on the road-building work
during the early phases of construction at various periods between 1995 and 1998.

The BRO left a lasting impression on those in Markakasa who worked for it, and the
phrase “BRO ke zemana mein...” (in the time of the BRO...) was the reference point used
by many to define that period of four years, even when the topic of conversation was not

122
related to the road construction work. The BRO paid its labourers generously and it was
seen as a time in which most people were awash with cash. The monthly wage was Rs.
2200, more than double the amount that could be earned on ‘ordinary’ government works
projects, and three times the wage of a farm labourer. When payday came, everyone
would receive crisp five hundred rupee notes; for some it was the first time they had seen
a denomination that high. Many workers bought consumer goods such as cycles and
radios, or a new pair of bulls. It was during the time of the BRO, I was told, that people
began to get a taste for eating snacks and drinking tea in roadside cafes, and indeed it was
at this time that Markakasa’s first paan and tea stall opened. Everyone smoked
sophisticated and expensive cigarettes (‘Bristol’), not the cheap, rustic and more popular
Indian bidi. Since labourers often worked far from home on other sections of the project,
the work provided plenty of opportunities to conduct love affairs too: one young (and
pregnant) Markakasa woman recently eloped with a man of a different caste she had met
at that time. The BRO did not benefit everyone however: those who requiredlabour for
their fields found it difficult to engage anyone at the wages they offered. This ishow one
man, Ganaram, remembered it:

“I was sitting in the teashop when Asaram came up to me and asked if I

wanted to come and work on his fields for a few days. I had a ‘Bristol’
hanging out of my mouth and I said that I didn’t think I could work for
less than sixty rupees a day! [he laughs]. I didn’t need to work as a
banihar (daily wage labourer); I had plenty of money then.”

It was not only the generous wages that people remembered about the BRO but the style
and efficiency of the operation. The project represented a different ‘state’ to the one
which they had had experience of. The machinery the BRO brought with them was
impressive, quite unlike the equipment used by the local Public Works Department
(PWD). The engineers and officers, who, I was told, all spoke English with one another,
did not tolerate complaints from farmers whose fields bordered the road, and when they
realised that supplies were taking too long in coming from elsewhere, they established a
cement factory close to their operations. The roads built by the BRO are constantly

123
praised and compared favourably with those built by the district PWD: the latter’s roads
begin to disintegrate after a couple of monsoons, because of the poor materials used by
the contractors (tekedar) who lined their own pockets. That sort of corruption did not go
on with the BRO, I was told, and as a consequence their roads are of better quality. In
short the BRO represented not only a different kind of State, but also a very generous and
powerful one, and the project has continued to have great symbolic value. As
representative of the modem, good roads are seen by Markakasa people as key indicators
of the extent of development of a particular area. Thus, as the neighbouring state of
Chhattisgarh was seen in many respects as representing the ‘traditional’, it was also seen
as more backward because of the poor condition of its roads. Indeed, Chhattisgarh’s
roads were a standing joke and descriptions of journeys there would often include the
phrase, “well, you know what their roads are like”. The importance of the BRO project
was largely symbolic in demonstrating the power of the state in a hitherto neglected part
of the country. As I examine later in this section, the engagement of the police has been
far more striking and closely involved with the issue of witchcraft.

The Naxalites88 are a definite presence in Markakasa and the immediate area and have
been active here since the early 1990s. Small squads visit the village regularly, going
principally to those households where they know they will be fed and watered, and also
come to the village shop (owned by my mahaprasad, Ganeshyam) after hours to buy
soap and biscuits. A general village-wide meeting has not, to my knowledge, been called
since 2001, approximately a year before I arrived for fieldwork. On that occasion, the
visiting squad woke all the villagers and asked them to gather in front of the Thakur’s
(hereditary headman) house. There they dragged a village man, Hira, and beat him in
front of the rest of villagers. The Naxalites had been told that Hira, who was renting a
room in his house to the local forest guard, was passing on information to the latter about
the other villagers’ illegal timber-hunting forays into the jungle. Substantial bribes were
being demanded by the guard, and one or more of the villagers felt aggrieved enough to
call in the guerrillas. The Naxalites had let it be known that Hira should ask the forest

88 The villagers used a variety o f terms to refer to the Naxalites: naxalwadi (‘naxal-ist’); jangalwalle (‘those
o f the forest’); lal salaam watte (‘red salute people’); or simply by saying o-man (‘that lot’). They were
also referred to non-verbally by making the hand gesture for a gun.

124
guard to leave his house. Hira, apparently not believing the information he was given, did
nothing. A couple of weeks later, the Naxalites showed they meant business.

Every village in the area has a story like this to tell, and in many the Naxalites have also
killed local people. Both the sarpanch (elected head of the village council) and the police
patil of the nearby village of Piparkhar - referred to above - were killed by the Maoists
about ten years ago. In January 2003, a couple of months into fieldwork, two policemen
were blown up by a landmine on their way to investigate a Naxalite arson attack on a
timber depot fifteen kilometres from Markakasa; the following month, a new, as yet
unoccupied government building located about twenty kilometres to the south was
attacked and burnt to the ground. In May 2005, a landmine killed seven policemen in a
jeep outside a village ten kilometres away. The sarpanch of Gujarbarga, Girdari - whom
I quoted above - was abducted several years ago and held in the forest for a night. These
are just some of the incidents which have taken place locally involving kin, friends and
acquaintances of Markakasa people. Information about Markakasa people’s direct
participation in the Naxalite movement was murky at best, but it seemed that only one
person in the village, a woman, had herself been a Naxalite and left the movement a
number of years ago; another had been a regular cook for a squad which visited her natal
village.

The idea of unity and the mixed village is also a concern whenever the Naxalites are
discussed. According to my Markakasa informants, the Naxalites only rather reluctantly
engage with panchrangi gaon (mixed villages); they prefer villages where Gonds
(Adivasis) are the only caste, or at least form a dominant majority. The reasoning is
rather straightforward: mixed villages are likely to be more divided and quarrelsome, and
Naxalite security and secrecy is compromised where one faction uses the threat of
informing the police about the others’ links to the guerrillas. The neighbouring village of
Kedanar is well-known among locals for being ‘open’ to the Naxalites: with the village
bordered on several sides by dense forest, they come and go as they please and are more
visible there than in Markakasa. Kedanar is also seen as a ‘united’ village in ways that
Markakasa is not, principally because the Gond caste are in the overwhelming majority,

125
comprising approximately eighty percent of the households. The Gond-Gowari89 caste
forms a substantial minority in Kedanar. Though marriage between the latter and Gonds
is not permitted, the two caste groups nevertheless often work together at each other’s
life-cycle rituals, especially in villages where there are only a handful of Gowaris (as in
Markakasa). Though there are twenty Gowari households in Kedanar, they have for a
number of years worked together with the Gonds. Recently however, the Kedanar Gowari
community split into two factions (‘parties’), with one remaining within the larger Gond
fold and the other asserting its desire to operate separately. It appears that a disgruntled
Gowari approached the Naxalites complaining about this state of affairs. Two or three
guerrillas then visited the village informally and in plain clothes, and let it be known to
the two parties that they would rather everyone made up. The leader of the ‘separatist’
party refused to resolve the issue and, when an ‘official’ visit was paid a few days later,
the leader was roughed up and the factions merged. “The Naxalites talk about unity (ekta)
when they visit us. They say that all the people in the village should stay united and not
let anything divide us.”

In response to increased Naxalite activity in the area, in 2003 the police established an
Armed Outpost (AOP) in Ganutola, a village four kilometres from Markakasa. AOPs are
heavily fortified camps, containing between 25 and 50 policemen led by a Police Sub-
Inspector (PSI). AOPs have none of the principal functions of police stations; their sole
purpose is to provide a base closer to the forests and villages where Naxalites operate,
and from where patrols can be carried out more easily. In an average week, the police
patrol for about three days, in a squad of approximately fifteen men, during which time
they spend the nights out of base either in the forest or in a village. While I lived in
Markakasa, a band of heavily armed policemen in fatigues, looking more like soldiers,
would often arrive in the late hours of the evening, use the house of a villager as a base to
eat their meal and then retire either to an abandoned threshing floor or the roof of the
school to spend the night, heading into the forest before sunrise. The patrol would then
reappear in the village a day or two later on their way back to the AOP.

89 Gond-Gowaris are classed by the Indian Government as Scheduled Tribe (ST) and are traditionally
cattle-herders. In this area they speak a language akin to Gondi. Their surnames are typically Gondi -
Netam, Madavi etc.

126
Not only are the police more visible to villagers than ever before, they are also more
accessible and less threatening. The policy appears to be one of engagement with the
local populace in order to diminish the attraction of the Naxalite movement. People
themselves say that, even fifteen years ago, everyone would either run into their homes or
flee to the forest on hearing of the arrival of police in the village. The police were
regarded as brutal and being hauled to the station would inevitably involve violence.90
Nowadays, I was told, even if the daroga (chief constable) visited the village, people
would not bother to get up and give him a seat. This was bluff of course, and when high-
ranking police officers did visit Markakasa they were treated with a certain amount of
deference. But the element of fear that people described for earlier times clearly did not
exist to the same extent. This was not only the result of their increased visibility but also
of a policy of positive engagement. In many ways, local police attempted to position
themselves as defenders and advocates of the local Adivasi population vis-a-vis other
local state actors such as the tahsil (sub-district) office and the Adivasi Development
Board.91

An example of such ‘engagement’ was the policy of helping villagers in their dealings
with the local administration over the renewal of ration cards. Ration cards are required
in order to purchase cheaply priced essentials such as rice, wheat, and kerosene from
Government shops. The renewal of these cards had to be done at the sub-district
headquarters, some forty kilometres away, and travelling to this town on government
work tended to be troublesome and expensive. Typically, the applicant would be asked to
submit several copies of various documents and told to return. This would happen
repeatedly to the same applicant; bribes were also demanded to process the paperwork.
These difficulties came to the attention of the policemen stationed at the AOP at Ganutola
as a result of their conversations with villagers in the course of patrolling. The ranking
officer, the PSI, decided that they, the police, would collect all the ration card

90 In 1990, the local police beat a village man so severely that he died o f his injuries.
91 Many scholars have noted that the notion o f the ‘state’ as a unitary, abstract entity is problematic.
Different state actors often have different projects and consequently ‘the state’ is experienced in multiple
ways by the citizenry. For more on the Indian state and society see Fuller and Benei (2001), Gupta (1995).

127
applications and submit them en masse to the tahsil office on behalf of the applicants.
The PSI at the time was a young man in his late twenties called Sachin. Hailing from
more prosperous western Maharashtra, Sachin had seen service in Naxalite-affected areas
in the east of the state for three years. His marriage had only recently been agreed and so
had his request for a transfer back west, delayed however because of the prospect of a
general election92. A talkative and inquisitive man, his visits to my house during the
course of a patrol always made me slightly uncomfortable: the conversation would
naturally turn to the topic of the Naxalites, and he would gently probe me as to whether I
had heard any news lately or had an encounter with them. Though I always replied in the
negative to both sorts of inquiry, truthfully in the case of the latter, dishonestly in the
former, I would then be anxiously questioned by Markakasa villagers at the teastall about
what I was asked and whether I had said anything about the jangalwalle. Sachin
explained why the police had decided to help the locals over the ration cards:

“That’s what’s good about being at the AOP in Ganutola: you’re in much
closer contact with the local people - we see them everyday at the teashop
or paantella - and then we meet people in other villages when we go on
patrol. We heard complaints about how difficult it was to get ration cards
renewed. I asked my superiors in Deori [the tahsil town] if we could do
something to help: it would help the people but it would also help us [to
encourage people to turn against the Naxalites]. We could show them that
the police are friends of the people and not their enemies. It’s true that the
police in this area have behaved badly in the past so we have to change
what people think of us.”

In their own minds then, they, the police, had become the protectors of the local populace
against the petty exploitation perpetrated by local officials. As I mentioned above,
villagers were conscious too of a change in attitude and viewed the police with less

92 Since the Naxalites had threatened to disrupt the elections, and had left flyers in the village urging people
not to vote, the level o f security was high. On election day itself, the officials walked several kilometres to
the polling centre (the Markakasa school) from a secret location, guarded by a group o f ten heavily armed
policemen. When the ballot closed, a convoy o f armed policemen arrived in several lorries to transport the
electronic voting machines to the sub-district headquarters.

128
trepidation than in the past. On the part of the police, this level of involvement, which
extended beyond the maintenance of order, was indicative of a wider process of
engagement that saw the Adivasi locals as ‘backward’ and in need of a guiding hand
along the path of development and modernity.

One such area is in the combating of what are regarded as superstitious practices,
including disapproval of the use of the Angadev. As we saw earlier in the chapter, the
police have been reluctant to grant permission to the visit of Angadev to villages in the
area. Whereas in the past such visits would routinely be approved, nowadays the police
are less willing to tolerate the practice. A Markakasa villager told me the following story
of an Angadev visit to a neighbouring village which took place some ten years ago.

A group of policemen on patrol happened to pass through the village at a time when the
Angadev was in attendance.93 They asked what the curious looking structure was and on
being told it was a god which had the ability to detect witches, the policemen began to
laugh. They challenged the villagers, saying that witches were not real and that the
Angadev could not detect them; they were examples of backward superstition and the
Angadev was guided by its bearers, not the other way round. How could the villagers not
see this? No, countered the villagers, the Angadev has considerable power (shakti): for
instance, they suggested, it would not permit just any person to pick it up. The police
took up the challenge but were unable to lift the deity. What really happened on that
occasion is difficult to determine but the truth of the matter is irrelevant. The man who
told me this story was trying to emphasise the difference in attitude between the local
populace and the police who patrolled in their areas. What is interesting is that the police
were seen as presenting the local populace as bound by superstition and they took on the
role of challenging such erroneous belief.

The police are just one part of a larger class of local state officials and administrators who
see the area in which they have been posted, and the people whom they govern, as

93 This is echoed by Sundar (2001: 441), who describes how the fact that a witch had been murdered only
came to the attention o f the police in Bastar when they were combing the area in an anti-Naxalite operation.

129
fundamentally un-modem and undeveloped in comparison to western Maharashtra or
other parts of Vidarbha, where the vast majority of personnel are recruited from. It was
constantly impressed upon me that the people here were uneducated, simple and easily
taken advantage of, unlike people from other parts of the state or country. This brings me
back to my earlier point: though witchcraft is, as Macdonald (2004) suggests, widespread
throughout rural (and perhaps urban) India, it is nevertheless associated with
‘backwardness’ in the minds of officials trained in the ideology of modernisation. The
combination of ‘backward practices’ (witchcraft) performed by paradigmatically
‘backward people’ (Adivasis) provides a potent composite image of backwardness. The
Naxalite insurgency has made certain state actors more urgently and intimately involved
in what they see as the problem of ‘backwardness’: Adivasi people are seen as ‘simple’
and ‘trusting’ - characteristics of their ‘backwardness’ - which, the police suspect, leads
to Adivasi collaboration with the rebels. Official intolerance of a belief in witchcraft and
of the corresponding measures needed to tackle it, and their much closer involvement in
people’s lives, means that villagers and villages that are afflicted by witchcraft and a
sense of unease find it increasingly difficult to use remedies which would have been
effective in the past.

Macdonald’s research on witch accusations and police authority in the Central Provinces
(where Markakasa was located94) in colonial times suggests that, in contrast to other parts
of India, police presence there was minimal and largely ineffective (2004: 114), and most
policing was left to village law agents who colluded in keeping witchcraft accusations at
the level of the village (ibid: 108-14). Even as recently as sixty years ago, at the time of
Independence, the police station with responsibility for Markakasa was in the small town
of Sakoli, eighty kilometres away to the north-west along bullock cart trails through
dense forest. As recounted by elderly villagers, police officers (and others such as forest
guards) visited only sporadically, staying at the abandoned ghotul building which came to
be known as the sepoy hangla (the soldiers house). The 1908, Bhandara District
Gazetteer mentions in passing that the southern part of the district (where Markakasa was

94 Markakasa was located in Bhandara District in the Central Provinces. In 1999, Bhandara was split into
two and the eastern portion, containing most o f the forest, and Adivasi population, became the new district
o f Gondia.

130
located) was ‘Very jungly and remarkably free from crime” (Russell 1908: 171). In 1906,
the proportion of police engaged in the detection and prevention of crime in Bhandara
District was one policeman for every 13 sq. miles and 2139 persons (ibid). This
compared to the all-Central Provinces figures of 9 sq. miles and 1061 persons
respectively (ibid). By 1968, the figures for the District were one policeman to 10.13 sq
km and 1370 persons.95 Thus even in the late 1960s this district was less well-covered
than the average district in the Central Provinces at the turn of the twentieth century. One
can conclude that the forested areas of Bhandara District were even less intensively
policed than the already rather sparsely administered general Central Provinces, as
MacDonald demonstrates (2004: 108-14).

If we examine MacDonald’s evidence of a limited police presence throughout much of

the rural Central Provinces in the light of Skaria’s analysis of witchcraft in the Dangs and
Mewar in colonial western India, we can make some speculative observations about the
present-day situation. Skaria (1997a) demonstrates how the districts of Mewar and the
Dangs differed: whereas the police and administrative presence in the former was in
many respects similar to that of the rural Central Provinces, the latter was intensely
policed and governed from an early stage. In the Dangs, individual witch killings became
more pronounced as village-wide witch detection was suppressed (ibid: 137). In Mewar,
by contrast, detection conducted by bhagats (diviners) and others continued, and witches
could be dealt with effectively at the village level (ibid). This resonates with
MacDonald’s contention that in the Central Provinces more generally, news of anti-witch
activity very seldom reached the ears of the administration. It may also account, rather
speculatively, for that curious comment in the 1908 Bhandara District Gazeteer that the
jangli southern part was crime-free, perhaps indicating that indeed in this part of the
Central Provinces, ‘crimes’ were not being reported but were dealt with at the village
level.

In the erstwhile Central Provinces, the process which took place in the Dangs in the early
part of the twentieth century has only happened more recently with the expansion of

95 From District Gazetteer (1979) Bhandara District, Maharashtra State (Bombay), p. 580.

131
police presence gradually since Independence, and especially in the past fifteen years
with the explosion of the Naxalite movement. Not only are the police more present and
accessible, but public policy and discussion in India oppose witchcraft and sorcery
practices to the idea of modernity: detecting and dealing with witches, and employing
sorcerers is seen as hampering India’s emergence as a developed modem nation.

Public policy on the specific question of witchcraft has had an effect on how police
approach the issue of witchcraft and witch-detection, and how ordinary people come to
be aware that in many instances their views on witchcraft are at odds with those of other
parts of Indian society. My repeated questions about the fate of witches once they were
identified were met with verbal equivalents of shoulders shrugged. A typical response
which I heard from several people: “What can we do? Nothing. If we beat them or even
have a meeting about them, they can go to the police station and file a complaint against
us. It didn’t used to be like that. The siart (elders) of the village would tie the person to a
tree and beat her to stop her doing her badmashi (wickedness).”

One recent example is the approval of a Bill for the Eradication of Black Magic and Evil
Customs by the Maharashtra Legislature which explicitly bans what are called
superstitious practices96. The Bill still needs additional approval in order to be formally
enacted and the purview of the Bill seems rather wide. It remains to be seen how the
courts will choose to interpret it or the police enforce it. The Bill aims to punish those
who promote superstitious rituals and contribute to the physical, mental and financial
exploitation of people in society. Its main targets are people like baiga in their capacity as
diviners and sorcerers. Also targeted are the institutions like the Mahanubhav temples
which, in the eyes of the sponsors of the Bill, prevent people from seeking proper
psychiatric assistance when ill and encourage a belief in witches, ghosts and demons97.
The moving force behind the Bill is the increasingly active and high-profile Maharashtra
Committee for the Eradication of Superstitious Practices (Maharashtra Andhashraddha

96 Hindustan Times, 19th December 2005: “One man’s crusade led to Anti-superstition Bill”.
97 Ibid, Interview with Dr. Dabholkar, founder and chair o f MANS. Indeed, people at a Mahanubhav temple
town told me that MANS have visited them and tried to persuade people who had come to be healed to
consult a mental health expert instead.

132
Nirmulan Samiti, or MANS). The Committee is a rationalist, humanist organisation
which campaigns across Maharashtra to demonstrate the falsity of what it sees as
superstitious practice and unnecessary religious ritual.

Perhaps this increased awareness of the hostility of the state has led to the progressive
marginalisation of baiga as identifiers of witches and sorcerers, and thus to an important
impediment to the proper resolution of mystical attack. Apart from the common
complaints about their greed, the most striking disadvantage from the afflicted person’s
point of view is that baiga are not willing to disclose the identity of the attacker for fear
of causing disputes. It is only in this sphere of suspected magical attack that they show
such reticence. In other circumstances, such as that of trying to divine the whereabouts
and identity of a thief, the baiga has no such qualms about describing the wrong-doer in
detail. This reticence has been noted for contemporary Chhattisgarh by MacDonald
(2004), who ascribes this change in role to “the police, media and administration seeking
to locate responsibility for witch-killing” (ibid: 186). She sees it as a relatively recent
development in rural plains Chhattisgarh, following a witch-killing case in 1995 that was
seized upon by the media and campaigning organisations. The police, she reports, have
publicly changed their stance and have been far more vigilant in tracking down those
(including witch-identifying baiga) who can be held responsible for violence against
witches (ibid: 140). She writes,

“Witchcraft accusations have transformed from being merely cognizant cases

identified through the Indian Penal Code to a political object for discourses that
challenge ideas of development, progress and modernity” (ibid: 152).

In a climate such as this, it is understandable that baiga are wary of identifying witches
and opening themselves to possible arrest by the police should the victim of magical
attack seek a violent remedy. Unfortunately, their reluctance to name witches or
instigators of sorcery makes them less satisfying for people who visit them in search of
answers about causation. And this applies to the use of the Angadev too. Though police
were refusing permission to invite the Angadev, the example of Gujarbarga demonstrates

133
that a village, if united, will ask the deity to visit regardless. The problem appears to be
that villagers cannot take satisfactory measures against any witches identified by the
Angadev: they run the risk that the accused will file a complaint with the police.

Conclusion

An examination of the impediments that exist to controlling or combating witchcraft,

sorcery and magic reveals several important attitudes and social processes. Not only does
it demonstrate that the categories of witchcraft, political economy, the State and sexuality
implicate each other in the lives of Markakasa villagers, but more specifically it offers a
methodologically satisfying explanation of why membership of a religious sect which
promises to protect one from mystical attack has grown so rapidly in the last fifteen
years.

Running through the discussion in this chapter has been the wider opposition of
‘wildness’ and ‘civilisation’ that characterises various projects of Indian modernity, and
the place that Markakasa people see themselves (and are seen by others) as occupying
along that continuum. The articulation of the idea of the ‘Adivasi’ and of the ‘Adivasi
area’, through the interpretative device of witchcraft, is highly relevant.

Recent historiography on witchcraft practices and their suppression in colonial India may
provide further insights and points of comparison to the processes I have outlined in the
chapter above. As mentioned above, Skaria (1997a) reports that the late nineteenth
century in western India (the Dangs) witnessed a decline in ‘traditional’ forms of witch-
detection such as ‘swingings’ and the use of bhagats (diviners) due to pressure from
colonial authorities. In addition, the colonial administration was particularly vigilant in its
pursuit of those suspected of killing witches. As a result, victims of witchcraft had no
remedy: they could neither detect witches nor retaliate (1997a: 140-41). One of the
consequences, Skaria suggests, was the use of religious reform movements in the early
twentieth century, such as the Devi movement (see Hardiman 1987), as substitutes to
control witches and magic. He reports that, as the movement reached the village, the Devi

134
would ask the women to assemble and they would be tested for witchery (ibid.). The
control of witchcraft then was intimately linked to the popularity of these religious
reform movements: since other forms of addressing the issue had become unavailable,
people saw in the Devi movement a way in which to deal with this pressing problem.

In her study of witchcraft in Chhattisgarh, MacDonald (2004) argues that violence in

cases of witch-accusations paradoxically produces rupture even though it aims tobring
resolution and order. She describes the process thus:

“By employing the solution of either physical or indirect violence, the

perpetrator(s) seeks to transform the social environment in several ways:
by immediately halting the magical strike of the witch; by seeking
resolution and order to the disruption produced by the witch; by exacting
retribution through murder/torture or forced relocation of the witch; and
finally, by forestalling any future ‘magical attack’ by the witch” (2004:
28).

Witchcraft and sorcery in Markakasa are increasingly difficult to deal with using existing
modes of defence. In the absence both of effective means of witch-identification and of
violent retributive remedies such as beating, killing or expulsion, people subjected to
mystical attack have sought out another solution. This solution - joining a religious sect -
enacts exactly the same process MacDonald outlines. By seeking relief at the sect’s
temple town and then by joining the sect, Markakasa people attempt to stop the current
magical attack, claim retribution against the witch through the actions of God, and ensure
that any future magical attack against them will be ineffective. Sect members in search of
protection transform both their social environment and themselves. In the next chapter, I
trace the rise in popularity of this particular Hindu religious sect, the Mahanubhav Panth,
which has been drawing adherents from Markakasa and the surrounding villages, and I
examine the ways in which membership of this sect produces rupture by altering people’s
perception of society, sociality and self.

135
Plate 19: Mar/cakasa people digging earth to build a road.

Plate 20: Taking a break from the road-building work.

136
Plate 21: Returning home from working on the road.

137
 Chapter Five
Sect, Self and Sociality

The previous chapter discussed the constraints that have developed in dealing
with witchcraft and magic over the course of the past fifty years, at the level of the
person, the household and the village in this area of central India. This chapter goes on to
consider how people have responded to those constraints by seeking out an alternative
means of resolving the problems of malignant mystical attack. This alternative consists of
joining a Hindu religious sect, the Mahanubhav Panth, which protects adherents from
these sorts of attacks. Membership of the sect fulfils the functional purpose of protection
and is also transformative in several ways. As I discussed in Chapters Three and Four,
witchcraft and magic are often associated with closeness between persons. There is a
heightened fear of attack in situations of intense sociality such as at wedding feasts, or as
a consequence of disputes which mar the fraternal bond. Becoming an adherent of this
particular sect leads to the regulation of closeness and the restriction of sociality in
certain spheres of life. Not only is sociality limited, but also the transition from suffering
to happiness effected by the sect leads to the partial articulation of a different kind of
person. This new type of person no longer depends on others to solve these particular
problems of existence: the maintenance of his relationship with God is all that is
necessary. In order to preserve this relationship of devotion, adherents are required to be
unswerving devotees (atal bhakt) and refrain from offering worship to any other deity but
God (Parmeshwar or Bhagwari). Moreover, they must abstain from eating meat and
drinking alcohol.

There is a vast anthropological literature on the Hindu religious and devotional orders
(e.g. Burghart 1996; Van der Veer 1988; McKean 1996), and a certain amount on the
Mahanubhav sect in particular (Raeside 1976; Feldhaus, 1984, 2003; Feldhaus and
Talpule 1992; Skultans 1987). Little of it, however, has focused on the context of those

138
sects in local society, away from the temple towns and ashrams. Indeed, Raeside states
quite categorically, “I have no information on how far professing to be a lay
Mahanubhava affects one’s life at all” (1976: 596). I therefore look at the ways in which
membership of a religious sect (and of the Mahanubhav panth in particular) changes
one’s relationship with society and creates a new sense of self.

Understanding the experience98 of joining a religious sect is important in contemporary

India, since so many devotional movements offer perspectives and commentaries on
society. Through gurus who give sermons (pravachan) and provide models of exemplary
behaviour, religious sects serve to fashion ‘good Hindu’ society as much as they create
‘good Hindu’ selves. Devotional sects in India have also been linked more explicitly with
the project of Hindu nationalism (McKean 1996), and the association is especially
important when marginal communities such as Adivasis and other low status groups are
involved, since the latter are targeted by organisations such as the Vidarbha Vanvasi
Kalyan Ashram (VVKA) using much the same rhetoric. This Hindu nationalist
organisation, while professing to be apolitical and only interested in ‘social work’, is
intensely involved in confirming the Hindu nature of the Adivasi people among whom it
works. It does this locally by encouraging Sanskritic daily worship, abstention from meat
and alcohol and the promotion of local holy men (sadhu-sant), who themselves advocate
such practices. VVKA activists also make periodic visits to villages to monitor possible
Christian proselytisation. Through a policy of ‘uplift’ of Adivasi people, the VVKA and
other nationalist organisations set up an opposition between those who are civilised and
those who are wild.

This chapter is largely an examination of how sect membership changes people from the
point of view of the person-in-society. In the chapter which follows this, I expand on how
the functional requirement for protection is transformed in people’s narratives into a more
generalised discussion of the nature of the ‘improved village’, which feeds into a wider
debate about wildness and civilisation. ‘Good Hindu’ selves in this analysis can create

98 By this I mean a phenomenology which takes sociality as its focus o f analysis, as Kapferer (1997) has
proposed in his discussion o f Sinhalese healing rituals.

139
‘good Hindu’ society, an attitude which resonates with the nationalist projects of
organisations such as the VVKA. What I want to emphasise is that people’s values and
attitudes as to what constitutes the ‘good Hindu’ and ‘good Hindu society’ are not
‘introduced’ by Hindu nationalists, but arise as a consequence of searching for a
permanent solution to bodily attack, socially understood.

The M ahanubhav Sect

The Mahanubhav sect developed in the thirteenth century at around the same time as
other religious movements in western and central India which emphasised bhakti, or
devotion to God. In common with many of these others, its key doctrines are monotheism
(or more accurately, monolatry), rejection of caste and the refusal to acknowledge the
authority of Brahmins. Though the Mahanubhav scriptures emphasised key differences
from other forms of contemporary worship and belief, and its adherents were often
subject to local Brahminical persecution, it is now considered just one of many Hindu
sects. In geographical and cultural terms, the Mahanubhav is primarily a Maharashtrian
phenomenon with adherents concentrated in northern and eastern Maharashtra (Raeside
1976: 585). According to Enthoven (1920: 427-33), Mahanubhav devotees numbered
22,000 at the turn of the twentieth century; present-day figures are hard to come by
because Mahanubhav sect members are no longer listed separately in the census, coming
as they do under the category Hindu. Secrecy had been a hallmark of the sect right up
until the 1920s, with Mahanubhav mahants (senior sect leaders) refusing to discuss
doctrine with outsiders (Raeside 1976: 586). This attitude changed and sect texts became
available to non-sect scholars, who discovered that these writings represented the earliest
and most complete examples of Marathi prose before the seventeenth century (ibid).
Nowadays, in those parts of Maharashtra where it has a presence, the sect and its temples
are associated with the healing of illness caused by witchcraft (see Skultans 1987).

Of the 122 households in Markakasa, thirteen are adherents of the Mahanubhav sect",
eight are Kabir and three are Ramanandi (see Chapter Two). The Mahanubhav devotees

99 Number o f households by caste: 5 Teli, 3 Gond, 2 Rawat, 2 Mochi and 1 Mahar.

140
in the village and elsewhere in the area are an eclectic group comprising members of all
castes, both clean and Untouchable, and people of all wealth levels. Not only is the
Mahanubhav sect the largest in the village, but it also appears to be fastest growing: just
ten years ago, there was not a single adherent in the village. The rate of growth in other
similar-sized villages in the area has been comparable, and each can count five to ten
(though in some cases more) Mahanubhav households. Interestingly, there are no
Mahanubhav sect members in the all-Gond neighbouring village of Kedanar, a fact on
which I will elaborate in the next chapter.

I focus principally on the Mahanubhav for two reasons. Firstly, as mentioned above, it is
not only the largest Hindu sect in the village but is also the sect that has shown the fastest
rate of growth. That it should become so popular in a relatively short space of time
demands a more detailed investigation, into how the rise in membership of this particular
sect affects those who join, in their relations with other sect members and with non-sect
kin and neighbours. Secondly, the diversity of its membership, in terms of caste and
wealth, makes it an interesting sect in which to examine the processes through which
existing sociality is restricted and new forms of association expanded. This would
perhaps have been more difficult to observe in a sect such as the Kabir panth, which has
an almost exclusively Teli membership in the area.100

Suffering and the Self

Closer analytical examination of the causes of suffering for those who join the sect
suggests that the actions of other people - witches, sorcerers and ancestors - are to blame.
Conversely, they become happy and protected from suffering by entering into a direct
relationship with God through devotion. How can one analyse this change in terms of the
place of human action in society? Graeber’s discussion of magic marks out a potentially

100 The Kabir sect’s membership, at least in this area, does not seem to have risen as dramatically (if at all)
as that of the Mahanubhav sect. Though this is entirely speculative and would require more research, the
Teli association with the Kabir panth may be connected to the issue o f witchcraft too. Since members o f the
Teli caste have a particular reputation for being witches in many parts o f Chhattisgarh, the act o f joining the
Kabir sect may in the past have been an attempt to distance themselves from possible accusations. See also
my brief discussion in Chapter Two.

141
fruitful path. He begins with Marx’s view of the human condition: he states that, for
Marx,

“human beings are creatures. The social world we live in is something

we have made and are continually remaking but we never see it fully
that way and therefore can never take control of the process. [Religion]
is the prototype for all forms of alienation since it involves projecting
our creature capacities outward onto creatures of pure imagination and
then falling down before them asking for favours” (Graeber 2001: 239).

For Graeber, magic is just the opposite of this: it is an un-fetishised view of the workings
of society “in that it recognises that the power to transform the world ultimately goes
back to human intentions” (ibid: 245).101

Following the logic of this analytical framework, what is it that Mahanubhav sect
members are doing by joining the sect and living by its rules? By refusing to
acknowledge that the actions of witches, sorcerers and others have any power over their
lives they are in a sense rejecting an unfetishised view of society and instead placing faith
in God and ‘religion’ - the ‘prototype alienation’ - to direct the course of their lives. Not
only does membership of the sect appear to minimise the role (or at least impact) of other
human action because of the fetish of God, it also enhances the role of the self. It is the
individual self who becomes responsible for his well-being: as long as he keeps to the
rules of the sect regarding diet and worship, his protection from the actions of others will
be maintained. Safety and certainty in the course of one’s life is entirely in the hands of
im
the devotee and is no longer dependent on the actions of others . This concentration of
power in the hands of the self entails a restriction of sociality. Moreover, this restriction,
as I show below, is promoted by the rules and practices of the sect.

101 In his discussion o f Sinhalese sorcery and healing, Bruce Kapferer (1997: 298) makes a similar point:
“The phenomenon o f sorcery and witchcraft are mirrors to humanity. Above all, they address the
problematics o f human existence in worlds that human beings make and unmake.”
102 Though Mahanubhav sect members venerate their gurus, the latter are more like moral exemplars than
intercessors with God on behalf o f the disciple.

142
Dumont’s (1980) discussion of religious sects in India is also instructive and indicates
that the type of self that is constituted by adherence to the Mahanubhav Panth is not
necessarily an ‘individual’ in the sense understood in Euro-America. For him, the
existence and pervasiveness of caste belonging in Hindu India means that the notion of
this individual, the ‘individual-in-the-world’, cannot emerge in society. Rather, it is
constructed ‘outside the world’ in the figure of the world-renouncer who rejects society
and caste (ibid: 273-75). According to Dumont, the ‘man-in-the-world’ who follows the
teaching o f the world renouncer, becomes a member of his sect and establishes a personal
connection with God does not transform into an individual in the Western sense, because
his new religious affiliation does not really attempt to compete with his caste identity and
make him despise the values of the latter (ibid: 188). Nevertheless, Dumont states that to
conceive of a personal Lord, as Mahanubhav adherents certainly do, there must also be a
believer who sees himself as an individual (ibid: 282). Thus there appears to be a tension
between the articulation of the self as independent of the actions of others, which is
supported by the membership of the sect, and the ultimate denial of individuality as
understood in the western sense because of the continuing relevance of caste as a source
of personal identity.

People come into contact with the Mahanubhav sect as a result of an incurable illness, but
the sect is seen as a last resort. Before people make the decision to go to the sect’s temple
town, they often seek out a wide range of other remedies.

Dealing with illness

When people fall ill, they have several options. Most people act immediately: very few
people sit around waiting for the illness to get worse. In the first instance, whether the
illness is a persistent headache, a fever or a stomach complaint, people generally seek the
advice of a ‘doctor’. A doctor is simply someone who can examine one and prescribe
treatment, either in the form of tablets or an injection; he or she is not necessarily fully
qualified (i.e. in possession of a M.B.B.S. degree). Markakasa was fortunate in having a
‘doctor’ who had studied under a ‘proper’ doctor for two years and knew how to deal

143
with minor ailments and administer injections. If the sickness appears serious to this type
of doctor, he suggests the patient visit another doctor, a ‘proper doctor’ in a hospital or
clinic located some distance from the village. Unless it is very serious, people are
reluctant to take this step largely because of the cost of visiting such a practitioner. At the
same time, especially in cases where pain exists but there is no external manifestation on
the body, the afflicted person visits one of the several healers in the village who are able
to phuk mama (Chh. ‘blow’). The healer, who is in possession of a mantra that has
proved efficacious, blows ash over the area of pain three times while reciting the mantra
under his breath. People also consult a baiga, a ritual specialist and almost always a Gond
man, who has the ability to divine causes of illness, remove inauspiciousness caused by
Saturn (Chh. shani) or combat witchcraft (Chh. jadu-tohna). If the sick person is a Gond,
the baiga might suggest that offerings be made to household gods (G. lochturi) or to the
ancestors (G. hanal). Payment is often made in alcohol (Chh. mahua darn)103, though in
many cases, especially in those where witchcraft or magic is suspected and needs to be
countered, chickens or even goats and pigs are needed for sacrifice. If, after trying all of
these, the person’s health still does not improve then often someone, usually a sect
member, will suggest they go to the Mahanubhav sect temple town of Sukdi, about a
hundred kilometres away to the north. What is important to note here is that a journey to
Sukdi is seen as the very last resort, after all else has failed.

Here I pick up the story - which I briefly outlined at the very start of the thesis - of my
landlord Brijlal, who lost two infant sons and nearly died himself. Through this partial
life history, I hope to convey some of the experience of his search for a permanent
solution to the problem of malicious mystical attack.

Brijlal is a Gond man in his mid-thirties. His family is one of the earliest to have settled
in the village some ninety years ago and is rather closely related to the family of the
hereditary headman (thakur). His FM was given from the Thakur’s household, and his FZ

103 This is alcohol home-brewed from the mahua flower, which is collected in April. Mahua dam is
frequently used as an offering in rituals involving Gond families and their household gods, and in collective
village rituals. It is a ritual substance, therefore, in a way that government-brewed alcohol (desi dam ) and
beer are not.

144
was married to her MBS, the current Thakur’s father. In other words, women have been
exchanged between the families twice (see figure 5 below). Both patrilateral and
matrilateral cross-cousin marriage is practiced among the Gonds, and is the preferred
marriage. This sort of union, however, is not without its perils. If there is a frequent
exchange of women between two households then a condition called vellarhari104 may
arise, an affliction that only Gonds appear to suffer from. Each Gond lineage has
household deities (G. lochturi), which protect that family and which require to be
worshipped regularly and at important events throughout the year. Vellarhari occurs as a
result of jealousy and neglect: the household gods of the family with whom one has
exchanged women begin to demand attention too. One now has two sets of gods to
appease. If attention is not given, the gods claim retribution by ‘eating’ (killing) male
children “so that the family cannot grow” (M. parivar vadu nai shakat). The trouble with
vellarhari is that one does not know one has a problem until children or young men start
dying or one finds one cannot conceive.

When he was twenty years old Brijlal married Champabai, my landlady, who was then
eighteen. A couple of years later, their first child, a son, was bom. The baby was healthy
and crawling, and things were good for several months. Then, on a windless night, the
large mango tree in front of the house keeled over; by the following morning their son
was dead. They grieved awhile but picked up their lives and a second son was bom a few
years later. This time, ten days after giving birth, Champabai suddenly stopped producing
breast milk. They tried feeding him other milk but the child weakened and, within a
month, he too had died.

104 I did not obtain a satisfactory etymology o f the word vellarhari from my Gond informants, though I
suspect it is the composite o f the Gondi words ‘vellah’ which has the connotation o f witch, sorcerer or
demon, and ‘arhari’ which is the Gondi kin term for FZD (and thus a marriageable cross-cousin for male
ego).

145
Figure 5: the relationships between the Thakur’s family and Brijlal’s

The Thakur’s family Figure 5

The Connections between Brijlal’s
and the Thakur's
family

Mayaram Konde

The current Manu The Didi Brijlal Champa Rajim

Thakur Bhairi
(headman)

Latchu

■ Deceased
adopted by
Manu

N o w they were convinced som ething was wrong: surely mango trees do not just fall over

and m ilk dry up? Unfortunately the death o f male children was nothing n ew in Brijlal’s

fam ily. H is then estranged elder brother, Manu, had been married five tim es and those

five unions had produced only one child, a son w ho suddenly fell ill at the age o f twenty

and died. Sim ilarly, B r ijla fs FyB, Konde, had lost tw o sons and only one now survives.

Brijlal consulted a local baiga (diviner, sorcerer), w ho declared that the cause o f his

children’s death had indeed been vellarhari and suggested h ow best to resolve the

problem. It seem ed that both Brijlal and the current Thakur had to perform a ritual

together in order to appease the latter’s household deities. But the Thakur w as unw illing

146
to allow someone else105 to worship his gods, thinking that his own family would suffer
as a result.

Though not magic or witchcraft in the strictest sense (though see footnote 104),
vellarhari, and the suffering it caused for Brijlal and his family, was the result of human
actions in the past (the marriage of a pair of cross-cousins) over which he had little
control. In addition, the solution suggested to him required the ritual assistance of other
people, and the Thakur’s refusal to help left him without an effective remedy to the
condition. Both the affliction and its solution required engagement with others; one could
not deal with it alone. He eventually managed to deal with vellarhari by turning to
another ritual specialist, a guru of the Ramanandi devotional sect who employed a
different method to that of baiga: instead of placating the angry household gods by
offering them meat and alcohol, the guru used powerful mantra to bind them and limit
their power over Brijlal and his family. The nature of divinity as revealed through
vellarhari and Brijlal’s experience o f it is also significant: household gods are jealous,
but their jealousy comes into being as a result of human sociality, in the act of two cross­
cousins marrying one another. Though Brijlal resolved the problem of vellarhari, his tale
of suffering did not, unfortunately, end there.

In 1998, Brijlal became very ill. He began to lose weight and suffered from terrible
headaches. He consulted doctors, who were at a loss to explain the cause and disagreed
among themselves as to the diagnosis. At the same time he went to see baigas, on one
occasion travelling a hundred kilometres away to seek a remedy from a particularly
famous one, and though he himself does not eat meat or drink alcohol, he had to offer
chickens and goats and bottles of mahua daru (alcohol). All the baigas said the same
thing: ‘someone is doing something to you’ (i.e. witchcraft \jadu\) but were unable or
unwilling to tell him who. This is one of the main problems people talked about when
discussing baiga: that they will not reveal the name of the tormentor because to do so
would lead to disputes (see Chapter Four). All Brijlal’s efforts appeared to be in vain: his

105 There are two exogamous divisions among the Gonds in Markakasa: seven-god {sath deo) people and
six-god (che deo) people. People o f the same division are regarded as brothers (bhauband), people o f the
other as affines (sorye). Brijlal is a seven-god Gond and the Thakur a six-god one.

147
health was still deteriorating. A member of the Mahanubhav sect, a rather wealthy old
Mahar man from a neighbouring hamlet, suggested that Brijlal go to the sect’s temple
town, Sukdi, to worship God (Parmeshwar or Bhagwan) and seek relief. Brijlal was
reluctant: he had heard strange stories of this particular sect, of its exclusivity and its
strictness, and was not all that keen to get involved. But he was also rather desperate; the
villagers seemed to be mourning his passing already, telling him his time was up, that
though he walked he was dead. The Mahar man suggested a test: Brijlal should light an
oil-lamp and pray in front of a picture of Lord Krishna and if he felt any better he ought
to think seriously about going to Sukdi. Brijlal did feel stronger and told his wife to
prepare for their trip.

Salvation in Sukdi

Sukdi is a rather large village in the north of the district about one hundred kilometres
from Markakasa. By bus it takes a good five hours, with several changes along the way.
Nor is the fare cheap, costing seventy rupees, the equivalent of two days wages106. The
temples of the Mahanubhav sect are located at places where the fourth (Govindprabhu) or
fifth (Chakradharswami) incarnations (avatar) of the Lord, and the founders of the sect,
stayed during their medieval wanderings in the thirteenth and fourteenth centuries. These
towns are ranked according to how long the avatar spent there: thus the more important
temples are those where they stayed a month or more. Sukdi is a rather minor temple
town in the large scheme of things: Chakradharswami spent only one night here. There
are five places of worship in Sukdi, each corresponding to a place where Chakradhar
stopped or gave a sermon. There are no idols here however; the objects of veneration,
known as ote (M. sing, ota) are raised stone and marble structures representing the dais
from which Chakradar preached, and there are five such ote at Sukdi. It is at the principal
ota in Sukdi, in the main temple, that the act of worship (puja) is performed, with hymns
sung and the Deity greeted with light (aarti). Worship takes place three times a day: once
at 4am, then again at noon and finally at 6pm. It is during these acts of aarti that women
receive byan. Byan (lit. statement) is the name given for the type of trance women enter

106 Daily wages ranged from Rs. 25 to Rs. 47 depending on gender, work, and employer.

148
when the aarti is performed.107 It takes the form of a conversation between God and the
baiga or witch who performed the spell that caused the illness or madness in the afflicted
person. Byan is the means by which God reveals the name and actions of the tormentor,
and thus represents a radical departure from the stance of baiga who refuse to name
perpetrators of magic for fear of causing disputes. Receiving byan is exhausting but an
essential part of the healing process. Through the body of the woman, God attempts to
dispel the magic and, during byan, women are thrown around the temple grounds,
crashing into walls or repeatedly smacking their heads on the marble dais which form the
focus of worship; sometimes they attempt to scale the temple buildings and hurl
themselves from the roof. What occurs during these intensely physical experiences is not
remembered by the woman herself, who claims that she feels no pain resulting from
byan, but is witnessed by those around her, including the man or woman who is the
subject of the witch’s attack. If the witch’s magic is especially potent, the woman will be
instructed by God to take herself to a lake at an appointed time where she will submerge
herself until the witch is defeated. Since there is a risk of drowning, attendants from the
temple accompany her and ensure that after the fifth submerging the byan-TQCQiver
surfaces. This is the final stage in the long process of healing the afflicted person, and is
not only physically exhausting for the woman receiving the byan but also somewhat
terrifying for the people watching. These intense experiences of healing involve the total
body, and also become defining motifs in people’s biographies as they explain the
transformation from suffering to happiness they underwent at Sukdi.

My Mahanubhav informants in Markakasa tended to have spent between one and two
months at Sukdi on their first visit, though of course people do go there for longer.108
Supplies of rice are taken from home, as are cooking vessels, and visitors find a spot
within the temple precincts and cook and sleep there.109 Going to Sukdi is seen as very
expensive; if both husband and wife go (as is almost always the case if the afflicted
person is married), the loss of income can be considerable. In addition, vegetables all
need to be bought from market rather than picked from one’s own garden. The decision

107 Skultans (1987) describes how in taking on the byan, women transform themselves into their patients,
transferring in a sense the illness to themselves.
108 When I visited Sukdi, I met one couple who had lived in the temple for six months.
109 In early 2004 when I visited, a row o f rooms that could house ten or so families had just been built.

149
Plate 22: The Mahanubhav temple at Sukdi

Plate 23: Midday worship at the Mahanubhav temple, Sukdi

150
to go to Sukdi is, as I emphasised above, not taken lightly; it really is seen as the last
resort. Moreover, visitors are aware of the transformative nature of the sect: that, if they
are cured and decide to join the sect formally in order to protect themselves from future
attacks, their lives when they return to their villages will change.110 Indeed, the initial
motivation for going to Sukdi - to deal with a malicious mystical attack - is largely
glossed over in initial discussions with devotees. What is emphasised instead as
motivation in ‘post-conversion’ narratives is the desire to improve and cleanse one’s life
and one’s soul and to move from a period of pain (dukh) to one of happiness (sukh). But
let us pick up the thread of Brijlal’s story.

Brijlal, his wife Champa, and their infant son stayed in Sukdi for five weeks. They slept
and cooked on a verandah behind the main temple, taking all their supplies from home.
The structure of their day was provided by the thrice-daily aarti and the rest of their time
was spent in prayer reciting the names of the five avatars of the Lord. By following the
example of other people there, and after receiving instruction from a guru of the sect,
they began to learn how to pray and worship the Mahanubhav way. Yet several weeks
passed and still Champa did not experience byan. Eventually, a friend of Brijlal’s from a
neighbouring village who was also at Sukdi at the time suggested that his daughter take
the byan for Champabai, a potentially dangerous proposal since the magical attack could
easily have switched to her. Brijlal and Champa thus came to know the name of their
attacker: it was the Bhairi (Chh. ‘the Deaf One’), Brijlal’s estranged elder brother’s
wife111.

The relationship between the two brothers, Brijlal and Manu, had not been good for well
over twenty years. Though they are now reconciled, its most longstanding expression was
in a dispute over inheritance and property that lasted ten years (discussed more fully in
Chapter Three). In the midst of all the claims, counterclaims and accusations of betrayal,

1,0 No-one was aware o f any person (and I certainly did not meet anyone) who had gone to Sukdi and not
been cured. Nor did I ever hear o f anyone being cured and then not joining the sect.
111 Though everyone referred to this woman as his ‘wife’, in fact they had not performed any sort of
marriage ceremony, not even a lesser marriage (Chh. churripahnana, lit ‘putting on bangles’) used for
widows or divorcees. She was still married to her first husband.

151
Manu’s wife, according to the byan, went to see a baiga to ask that he cast a spell
(imantra) to kill Brijlal. Bhagwan (God), however, punishes evil-doers, those who harm
his devotees, and when the Bhairi died rather suddenly last year, there were whispers in
Brijlal’s household about the byan having sealed the old woman’s fate.

Joining the sect and returning home

Membership of the Mahanubhav sect leads to a transformation in the lives of its

members. Not only is there a transition from suffering to happiness, but notions of
sociality change dramatically too, linked to a heightened fear of others to be sure, but also
because the sect requires the devotee to take responsibility for his own misfortune.
Having looked at the reasons for suffering and found in them the idea that the actions of
other humans can be dangerous and problematic for people, I go on to show how the
practices and requirements of the sect lead devotees to restrict sociality with others, in an
effort to protect themselves better from renewed malignant spiritual attack. Through this
process, I argue, the adherent also becomes a different kind of self in society.

During their five weeks at Sukdi, Brijlal felt his strength returning: his headaches ceased
and he began to put on weight. He and his wife decided to join the sect and became
disciples (shishya) of one of its gurus, a man who lived in Sukdi, and who was already
guru to several of Brijlal’s acquaintances back home. The process of becoming the
disciple of a guru is known as kan phukna (lit. blowing on the ears); what is ‘blown’ is a
mantra or spell that the adherent repeats at times of crisis and which he or she must not
reveal to anyone outside the sect. Joining the sect means following its rules regarding
diet, prayer and conduct within society, the observation of which protects one from
further attacks of magic or witches. The types of behaviour these strictures demand,
however, are radically different (and are seen as such) from the Hinduism practised by
the rest of the village. Devotees must refrain from eating meat or drinking alcohol: this
has proved problematic in that it has disrupted understandings of commensality, and is
discussed in greater detail in the next chapter. No images of gods or goddesses are
permitted in the devotee’s home, which is in sharp contrast to other houses, plastered as

152
they are with printed images of deities. Like the God of Moses, this Bhagwan of the
Mahanubhav is a jealous god and devotees must not offer worship to any other deity.
Twice daily puja (worship) is required before the family take their meals. Though the
object of worship is Bhagwan, who is in theory formless, devotees are also given a small
black stone, called vishesh, which is said to represent God and, according to some, a
piece of an ota. It is this vishesh that is placed on the altar and is the focus of devotion
during aarti. The vishesh is powerful and has the ability to protect one from witchcraft,
and is also treated with great tenderness (the act of devotion) by being kissed and stroked
repeatedly before the aarti commences. This twice-daily puja is a practice no non-sect
person performs and indeed many people say they are unaware of how to perform
worship. The contrast was especially striking during the festival of Diwali on the night
when the goddess of prosperity, Laxmi, is traditionally worshipped. While my landlord
Brijlal’s household spent nearly fifteen minutes on an elaborate form of worship112, his
sister’s household, who do not belong to any sect, waved an incense stick in front of a
newly purchased poster of Laxmi and that was that. Of course this too is a form of
worship, but Brijlal’s sister was acutely conscious that she did not know which phrases to
utter or how to sing hymns.

The consequences for disobedience of the rules of the sect are dire: the vishesh disappears
(gaiap), protection from magic is withdrawn and the erstwhile devotee will go mad. The
disappearance of the vishesh is a sign to the devotee that the strictures of the sect are not
being adhered to. I was told the story of one woman in a neighbouring village who started
eating meat once more. Her vishesh disappeared immediately. She stopped and it was
found.

The division between sect and non-sect villagers is made even more apparent when the
former follow the injunction to worship Bhagwan alone. For members of the Gond
community this means that household gods (lonchturi) are removed from the house, as are
the spirits of the ancestors (hanal) (see further below); for Mahars who join the sect and

112 They were not, o f course, praying to Laxmi but to Bhagwan; nevertheless, the purpose of the worship
was identical.

153
who were Ambedkarite Buddhists, the worship of Lord Buddha and Dr. Ambedkar
ceases.

Jadhuram, who lives in the neighbouring hamlet of Herpar, is the wealthy old Mahar man
who suggested that Brijlal go to Sukdi to seek relief. His house, though non-descript from
the outside, is rather impressive once inside, with large carved pillars and a stone (as
opposed to mud) floor. Jadhuram also owns a small, rather decrepit rice-milling machine
that does not seem to be making as much money as it used to. He has been a member of
the sect for more than fifteen years, one of the first adherents in the area. He joined it
along with his wife, who had suffered from a number of illnesses and sought relief at the
temple. Her illness was the latest in a long line o f misfortunes that befell the family: their
cows got sick and died; a new rubber belt for the mill snapped after only four days of use,
as did a second belt that was bought to replace it; and one of their sons fell seriously ill
too. Convinced that they were targets of witchcraft, the family went to Sukdi for a couple
of months and came back cured and members of the sect. As with all devotees, Jadhuram
was required to give up the gods that he had worshipped before; in this case, he removed
pictures of Ambedkar from his house and stopped the worship of Buddha. His caste-
fellows were not at all pleased, and demanded he reinstate the images and stop the
worship of his new God; they accused him of trying to give up his religion (dharma).
Jadhuram refused, and the local Mahar caste-council slapped a fine on him and he split
from the rest of the caste-community, though he was not formally excommunicated by
them.

But Jadhuram also insisted that he had not given up his status as a Buddhist: he pointed to
the fact that he still had a picture of Lord Buddha in the house, though he admitted he did
not worship Him as he had before. This seemed contradictory to me since he, along with
other adherents, used the phrase dharma suikarla (‘accept the faith’) to describe the act of
accepting the Mahanubhav teaching and joining the sect. Could one be a Buddhist and a
Mahanubhav at the same time? I realised however that Jaduram was explaining that his
identity as a Mahar had not changed in becoming a member of the Mahanubhav sect.
Since being a Mahar is so closely identified locally with being a Buddhist, keeping the

154
un-worshipped picture of the Buddha in his house meant he wanted the world to
recognise his continued identification with the Mahar caste. The sect changes many
forms of sociality, and the caste of fellow adherents is said to be irrelevant in the
community of the sect. Mahanubhav sect members will all sit in one unbroken line
(pangat) to eat, Untouchable and clean caste alike. Thus commensality rules are relaxed,
but only on public occasions of Mahanubhav devotion (e.g. at satsang, see below). Caste
identity is still extremely important to sect members: a marriage between members of
different castes (whether they are both Mahanubhav devotees or not) is still a
transgressive union. As Dumont writes, “a sect cannot survive on Indian soil if it denies
caste” (1980: 269). It demonstrates the limits on the transformation the self undergoes by
joining the sect: completely denying caste is not an option. Like ritual friendship, the sect
provides an alternative to the ideology of caste in certain spheres of life, but cannot
eliminate its relevance.

When I asked non-sect villagers what they thought about the Hindu sects and their
membership, the answers I received were broadly of two types that seemed to conflict
with one another and were often offered on different occasions by the same people. One
type of answer was a broadly critical one, that the sects (and in particular the
Mahanubhav) were ruining the village; another common response was that people get
sick and need to find cures, and going to Sukdi seemed to be very effective. One man,
Ponwar Singh, a senior Gond who had been the village’s sarpanch (elected headman) in
the past, was intensely critical of the sects and their members because he saw them as
destructive of the established fabric of village life, in that they refused to take part in
village rituals or acted in ways that were different to their caste-fellows. “How is this
village supposed to work, it’s become full of little, little sects”113, he remarked as a
meeting convened to ponder its response to the theft of one of the village tutelary
deities.114 A more personal criticism this time, directed at my landlord: “Brijlal doesn’t
follow our Gondi customs (rithi-rivaj) anymore. It started when he joined that panth
(sect). It was a scandal what he did with his father’s funeral pyre: he just let it bum and

113 Hamara gaon mein nane nane panth bowth ho ge hein, gaon chalana kaise?
114 This event is examined more fully in the next chapter.

155
didn’t collect the ashes - that’s what they’re told to do by their gurus.”115 But the tone
was markedly different on another occasion as we were cycling together to another
village: “People go to places like Sukdi because they are very sick, because they’re going
to die. So they go and then they come back and they’re cured. People need to go to places
like that; how can I say that it is a bad thing? You didn’t see Brijlal when he was sick, he
was so thin. Now look at him, you would never know that he had been through something
like that.” What can we learn from these seemingly contradictory statements? Perhaps we
can say that, when it comes to ways of thinking about the village or the caste-group, the
sects are discussed disparagingly and regarded as dangerous. But at a more individual and
perhaps more personal level, when one thinks in terms of sentiment and affection, people
who in other situations are opposed to the Hindu sects for their perceived exclusiveness
recognise that people who are sick need to be healed and need to search for solutions, and
that a temple such as that at Sukdi has proved highly effective.

While at Sukdi, the byan, as we have seen, is an individualised measure: it describes the
circumstances of spiritual attack against one particular person, usually a relative of the
woman who enters the trance. The nature and function of the byan, however, changes
once the women and their afflicted kin return home to their villages. There, byan is
received by one woman on behalf of all the members of the sect in any given space (a
village for instance). Thus she acts as a warning device, alerting her fellow sect-members
to the possibility of spiritual danger. Non-sect members are not privy, however, to the
information revealed in the state of trance and so Mahanubhav devotees regard
themselves as more able than non-sect members to fight attacks of magic or witchcraft. It
is generally the case that one woman at any given time will be receiving byan; when her
byan ends another woman inevitably takes it up and begins to trance during the aarti
performed in her household. In Markakasa, the reception of the byan seemed to go back
and forth between two particular women, a Teli woman and a Rawat woman. Needless to
say, receiving the byan is regarded as a mixed blessing; through the byan, one becomes

1,5 This statement was puzzling in the extreme, firstly because Brijlal’s father had died several years before
Brijlal joined the sect; secondly, because Brijlal’s mother, who had died after Brijlal joined, was buried
according to Gond ritual; and thirdly, Mahanubhav panthis do not cremate their dead but rather bury them.

156
aware of danger, but the experience is intensely exhausting for the woman concerned and
she is considerably weakened during it.

From what I have sketched above, I hope to have given the impression that the reason
people go to Sukdi and then subsequently join the sect (which is seen as a difficult thing
to do) is to find a lasting and effective solution not only to the wider, more abstract
problem of malicious mystical attack, but, more concretely, to the sickness they suffer
from at that particular moment in their lives. But I ought to make clear that sect members
do not always explain their reasons for joining the sect in these terms. I spoke to many
Mahanubhav sect members over the course of my fieldwork: those that I knew well,
especially those from Markakasa, told me rather plainly and straightforwardly that they
had gone to the temple in order to cure intractable illness or madness caused by
witchcraft. Others, most often those from other villages who were relative strangers,
would be rather vaguer. A typical explanation was: “My life was full of pain (dukh) and
sin {paap). I wanted to change my life and become ‘improved’ (sudharlele). Before I
joined the sect I ate meat and drank alcohol and told lies. Now I am a better person and
I’m happy. My soul is pure (shuddh); I’m pure on the inside.” Later on, either by asking
the person concerned directly or asking another more familiar sect member I would learn
that in fact such-and-such-a-man, for instance, had been vomiting blood for several
months and had therefore decided to go to Sukdi to get it sorted out. That the
explanations differed so strikingly were not, I believe, simply the result of my informants
feeling uncomfortable talking about what might be perceived as sensitive matters (though
of course this played an important role). Rather, the types of narrative offered by sect
members demonstrate a desire to emphasise the break with the past they have
accomplished in going to Sukdi and joining the sect. The statement of the unnamed
informant above may be interpreted as a search for a more prestigious way of life in line
with theories such as Sanskritisation (see Chapter Six), but one must keep foremost in
one’s mind that the primary (maybe only) reason why people beat a path to the Sukdi
temple door is to seek a cure from a life-threatening, painful and intensely experienced
sickness.

157
It is here that we can bring in the discursive motif of the body. Concepts such as sin,
misfortune and mystical attack are discussed and controlled in relation to the bodily
processes of eating and drinking inappropriate and dangerous substances (meat and
alcohol). And it is through the body that both the pain of suffering is experienced, and
healing through byan is achieved.

We have seen that the return home to the village from the temple involves a shift in
understanding: whereas people enter Sukdi as one type of person, they leave it as another,
with an additional sense-of belonging to the fellowship of their sect. Let us now go on to
examine how that sense is more precisely articulated.

Satsang and the articulation of difference

The guru had been ailing for some time now. On his last visit116 to the village he had not
been able to engage in any conversation and had seemed very weak. He whispered the
words of worship and let his wife do most of the talking, nodding or grunting to signal
assent. Several months later, as they learnt of his death, Brijlal and twelve other shishya
(disciples) from Markakasa and surrounding villages hired a car at considerable expense
to take them the 100km to Sukdi in order to attend his funeral.

The death of the guru had a profound impact on his disciples in Markakasa and in other
villages in the area. For the members of the Panth, their guru was a remarkable man,
worthy of great veneration and imbued with the power to keep them on the path of
righteousness. There was a fear that, with his passing away, his disciples would become
lax in their practices and forget what he had taught them. An effort was made, therefore,
to arrange religious meetings, satsang, as regularly as they could, something that had not
been organised while the guru had been alive. Indeed, the inspiration behind the satsang
appears to have been the guru himself. On the same day he died, he appeared in a dream

116 This act o f visiting was significant. Indeed, other commentators have remarked on the importance
placed in the Mahanubhav sect on constant movement from place to place in emulation, one supposes, o f
its founder, the wandering Chakradar Swami (Raeside 1976; Skultans 1987).

158
to a man called Phul Singh, the son of Bhagat, one of the earliest disciples in the area,
and instructed him to gather all the shishya together and perform a satsang.

The purpose of meeting in satsang is to discuss and debate matters of dharma (religion),
and doing so is fairly common throughout North India especially among members of
devotional sects. The satsang is typically composed of lay men and women, though
exceptionally it will be directed by a religious figure such as a guru. In such a setting,
questions are asked about the importance of practices and rituals along with more
complicated issues about the nature of life and o f death, and those members of the lay
community that are regarded as being more learned than others in such matters are called
upon to give their views. Essentially, however, it is a forum in which all the faithful, both
men and women, are encouraged to talk and share their thoughts.

I attended two such satsang, one in October 2003 and the other two weeks later. The first
was held in a village called Ganutola located four kilometres from Markakasa and the
second in my landlord’s house in Markakasa. What follows is a detailed exposition and
analysis of what went on. I examine the second satsang first.

The second meeting I attended was the fifth to be held following the death of the guru.
After dinner, at about eight o’clock, people began arriving at the house; men came first,
from Markakasa and other villages, and the women followed some time later after having
washed the dinner plates and tidied the kitchens in their homes. As the devotees entered
they greeted everyone else present individually, both men and women, with their
characteristic salutation, the dhandavat (see below for full description). The front door to
the house was closed, sealing off those gathered inside from the rest of the village. Front
doors to houses are very rarely closed in this way, especially if people are awake inside.
However, doors are often closed when a baiga is visiting the house and performing a
ritual to expiate misfortune, illness or witchcraft, which is done to prevent any force from
outside the house interfering with the ‘work’ (kam) going on inside. In a sense, therefore,
this small act was a way of keeping what went on inside secret and unpolluted and not

159
subject to possible spiritual interference; it was a ‘sealing o ff of the community of the
faithful.

The discussion moved on to a topic which appeared to have been bothering a number of
the devotees for a long time, a question which was asked as ‘should we contribute money
(barar) to village festivals such as Ganesh?’ and which one could paraphrase as: ‘how are
we to live in society with people who are not members of our sect?’

Phul Singh, the man who had had the dream, answered first, and it was a categorical
“no”. He said that it went against the principles of their dharma (religion) to worship any
other God but Bhagwan, and that they were atal bhakt (unswerving devotees) who did
not stray from devotion. “By making a contribution to the village’s Ganesh festival you
are in fact worshipping Ganesh {ganeshtsa puja karat ahe). You may think that you
haven’t participated in puja because you don’t eat the prasad or do the aarti, but with the
contribution you have made other people will, and so you will end up worshipping him,”
Phul Singh explained. For him, worshipping Ganesh was tantamount to betraying the
promise to be an atal bhakt. He proceeded to ridicule the festival: “what do people do
after they’ve finished worshipping him for ten days?” he asked rhetorically. “They throw
him [his image] in the lake! Is this how God (bhagwan) should be treated? Is this really
the Lord (parmeshwar)?” Another man quoted a saying by Tukrodi Maharaj, a sant (holy
man) and an immensely popular religious and moral figure in Maharashtra (though not a
Mahanubhav panthi)\ “If it is made of wood it can bum; if it is made of mud it can be
dissolved; if it is made of gold it can be stolen; if it is made of stone it can be broken.” By
contributing this he was affirming the Mahanubhav injunction against the worship of
images and agreeing that the worship of the idol of Ganesh was not something that a
devotee of Bhagwan should participate in. Phul Singh then went on to elaborate what he
saw as the problem with Hindu dharma, its multiplicity of gods:

“Four men - a Hindu, a Sikh, a Christian and a Muslim - wanted to cross

a river. They saw a boat hitched to the bank and got in and made their
way across. Suddenly the water became rough and the boat began to

160
 sink. Each man started to pray. The Muslim prayed to Allah and was
saved, the Christian prayed to Christ and was saved and the Sikh prayed
to the “Sikh God” [sic] and was saved. The Hindu was sure of being
saved too: after all he had far more gods than the other three and they
had survived. ‘Which one should I choose to save me?’ he wondered.
T il choose Ganesh because he is the god that is prayed to first.’ So he
prayed, ‘Save me O Ganesh!’ ‘Why should I save you?’ the god
answered. ‘Every year you worship me for ten days and then you throw
me in the lake and forget me. I will not save you.’ The Hindu then
prayed to other gods but each one refused because they were upset at not
being asked first. The Hindu gave up hope and drowned. The other three
men were saved because, unlike the Hindu117, they were all atal bhakt
[devotees of one God].”

But people had questions. Several adherents reiterated the need for all the Mahanubhav
adherents in this area to act as one, especially regarding contributions. “If we act
differently then the other villagers don’t understand what it is we are. It matters what they
[non-sect villagers] think about us. If one man in a particular village says that he will pay
the barar and another man in the same village says that he will not, then other [non-sect]
people will wonder to themselves as to the nature of our panth.” There is a feeling among
adherents that their panth in particular is demonised by other villagers as being strange
and difficult, in a way other sects such as the Kabir or Ramanandi are not. “They [non­
sect villagers] are always prodding us (humala towsthat) to try and discover what our
dharma [religion] is,” one of the others said. They seemed to be extremely conscious
about the inconsistencies in their behaviour and how these inconsistencies appear to other
people. Unity was being sought, and the purpose of the satsang was to achieve it in
relation to the rest of society by ensuring that all the adherents in the area were reading
from the same page.

117 Despite this formal assertion o f being distinct from Hindus, Mahanubhav adherents regularly described
themselves as Hindu and as belonging to the wider Hindu tradition rather than outside it. Thus my landlord
Brijlal would say “My dharma [‘religion’] is Hindu, my panth [sect] is Mahanubhav and my ja t [caste] is
Gond. As a point o f comparison, even Buddhist Mahars in some contexts described themselves as Hindus.

161
The example of a neighbouring hamlet, Gujarbarga, was given. Out of around forty
households, six are members of the Mahanubhav sect. And of these six, five are fairly
recent ‘converts’ and the remaining household has been a member of the sect for much
longer. When they were asked to contribute barar to the Ganesh Festival, the latter paid
up but the former refused. The confusion arose because of the different attitudes of the
two groups: the senior sect member was accustomed to paying because up until recently
the members of his household were the sole adherents of the sect in the village; the newer
members on the other hand, confident in the solidarity of their group position, felt able to
refuse to pay. This confusion was troubling to the gathered sect members because it left
them open to charges of inconsistency to those ‘other people’ (bakilog) unwilling or
unable to understand their dharma.

It is significant that the initial question about the payment of contributions was asked
specifically in relation to the Ganesh festival {ganesh utsav). Ganesh (also called
Ganapati and Vinayak) is the pot-bellied, elephant-headed god and an immensely popular
deity in Maharashtra. His festival is held in the Hindu month of Bhadrapad which
generally falls sometime in August-September. Originally a household festival observed
almost exclusively by Brahmin families from the Marathi cultural heartlands in and
around the cities of Pune and Mumbai, it was promoted and transformed by the
nationalist leader Lokamanya Tilak towards the end of nineteenth century into a public
celebration in order to circumvent colonial laws which outlawed the assembly of large
numbers o f people (Cashman 1975). Regarded as one of the early successes in mobilising
the masses during the nascent movement for self-rule, the Ganesh festival remains to this
day an intensely political festival. Though initially a uniquely Maharashtrian urban
phenomenon it has since spread to rural areas in the state and to other parts of India (see
e.g. Fuller 2001). The Ganesh festival is particularly well-suited as an expression either
of village solidarity or of village conflict because there is no single permanent site in
Markakasa, such as a temple, for his worship; as in many places in rural India, temporary
shrines are constructed, which are then installed with the idol {murthi) and where the puja
is performed for the duration of the festival. Thus, no single shrine can claim to be more

162
important than any other. In other words, there is no main shrine from which the others
legitimate their authority. The Ganesh festival has in recent years become a site of
struggle between the clean castes and the Mahars. The latter have refused to pay barar
for the celebration of the festival and have instead privately supported the construction of
a rival shrine in the village. Aware of this history, for the Mahanubhav panthis to suggest
withholding contributions to the festival, though admittedly for different reasons, is to
challenge the idea of the festival as somehow representative of the entire village, and
acknowledge it as a legitimate site at which to demonstrate their lack of sociality in the
common worship o f ‘false gods’.118

This exclusivity in relation to non-sect members, and the doings of other dharma,
unfortunately extendetj to the anthropologist too. At one point in the evening, the
discussion outlined above stopped, and people began talking about what they wanted to
do next. It seemed they wanted to discuss the mantra, or formulae, given to them by their
guru. My landlord looked over to me and asked me what the time was. I replied that it
was midnight. “You must be very tired,” he said, “don’t you want to go to bed?” “No,
I’m fine,” I said, rather stupidly. A few minutes later, he asked me the same question.
This time I got the hint and took my leave. My landlord’s sister and her eldest son, who
had also been listening up until that point and are not members of the sect either, also got
up and went to bed. The mantra are of course secret and are not to be revealed to the non­
initiated. The following morning, my landlord filled me in on what I had missed. After
praying to themselves quietly for some time, where they repeated the names of the five
avatars of the sect, they held what seemed to be a question-and-answer session on the
practices, rituals and beliefs of the sect in order to establish that everyone present
remembered and performed what their guru had taught them. Brijlal told me that some of
the women, especially the older ones, had demonstrated a considerable lack of this
important knowledge and that they had been set tasks to improve it by the next satsang
when they would be tested. The satsang was thus not only a forum for discussion but the

118 It was striking too that they sought to follow the Markakasa Mahars’ lead on this, especially as non-sect
villagers disparagingly called the sect “the Mahar sect”, principally, I was told, because the first few
adherents in the area were Mahars, and also because the usual caste rules regarding commensal dining
restrictions with Untouchables were relaxed with members o f one’s own sect.

163
site at which, paradoxically, a ‘closing of the canon’ was taking place: the principles and
practices that bound the devotees together were being emphasised and knowledge of
them confirmed in order to mark them out from non-sect others.

Old kin, new kin?

Now, someone at this satsang at my landlord’s house raised another issue: “what about
relations with our mother and father, and our sas-sasur (spouse’s parents)?” The context
of this question was clear to most of those present, especially those from Markakasa.

During the nine days of the festival of Navratri, which had just been celebrated, the
adherents of the sects are required to abstain from eating food cooked in another
household, refrain from eating with people not of their sect and avoid spending the night
in a house other than their own. In the week gone by, the village had been abuzz about
the dispute emanating from Gangaram’s house. Gangaram is an elderly Teli man who
lives with his wife and three sons. All the sons are married and the younger two, Dinu
and Lomu, are members of the Mahanubhav sect, along with their wives. Though Lomu
is often referred to as haf-dhimak (half-witted) and has a violent temper, Dinu is well-
liked throughout the village, and is seen as a good-natured, hard-working and generous
young man. The other villagers were surprised to learn, therefore, that both Dinu and
Lomu and their respective wives were refusing to feed their parents. When asked why,
they justified their action by saying that the rules of the sect forbade them from doing so
for the duration of Navratri. This answer seemed to confirm in the minds of many non­
sect villagers whom I spoke to that the Mahanubhav sect was radical and dangerous and
that it overturned even basic notions such as the obligation to feed one’s parents. This
very example was raised in the satsang that night; interestingly neither brother had turned
up, but now Dinu was brought from his house and asked to explain himself. He did so,
and several others present agreed with him and said they had not fed anyone but their
immediate household either; that was how they had understood the teaching. Again, the
consciousness of what outsiders would make of this was immediately apparent: “People

164
will ask, ‘Is this what your panth is about? These sukdi-walleh119 aren’t allowed to feed
their parents!’ That’s what people will say.” Again, the devotees were asked to be
flexible and that, even if their parents were not members of the sect, they could not deny
them food; by neglecting to serve (seva) their parents, they were neglecting to serve God.
Yet men and women needed to be wary of their non-sect parents too. “My sas-sasur (in­
laws) complain that their daughter has changed since she joined the sect and try and
convince her that what she is doing is wrong and that she should give it up”, one man
said to a general murmur of agreement and acknowledgement. Likewise, women were
asked to stay on guard against attempts by their in-laws to change the mind of their
husbands. Both husband and wife were exhorted to help each other and protect one
another from such challenges to their dharma (faith). The emphasis here seemed to be on
the importance of the marital bond and the ‘sealing o ff of husband and wife from the rest
of society. This ‘sealing’ finds expression in the sect’s ritual practice too. When
adherents perform one of the most important acts of worship at a sect temple, known as
ota, it is the husband and wife that perform it together with others standing well away. It
is only the couple who hold the plate (thali) containing the aarti (flame) used to worship
at the ota. When I accompanied Brijlal, his family and several other villagers to Sukdi
and witnessed them perform this worship, Brijlal gave specific instructions to the rest of
us that at no point during the puja were he or his wife to be touched. This is in marked
contrast to the way in which worship is offered during village rituals, such as at Ganesh
festival where as many people as possible hold the plate or, even more strikingly, ritually
touch someone who is holding the plate. One could argue that it is the marital unit
through which the sect is in a sense experienced. Married people who visit Sukdi in
search of relief are often accompanied by their spouse and, as we have seen, spend more
than a month together surrounded by strangers and other sick people. Both men and
women devotees whom I spoke to often dwelt upon this aspect: the time spent in one
another’s company and of cooking and eating together in such a small space. As Skultans
(1987) has shown, the receiving of the byan, which is done exclusively by women and in
most cases by wives, is seen by them as a sort of sacrifice on their part performed for the

119 ‘sukdi-walleh’ mean those that are devotees o f the sect associated with the town o f Sukdi i.e. adherents
o f the Mahanubhav panth.

165
well-being of their families. Children do accompany their parents to Sukdi especially if
they are young, but they do not become members of the sect until they are older (jawan),
and then only if they choose to. The sect also seems to disapprove of marital infidelity; if
a husband sleeps with another woman for instance, this too will be revealed in the byan
that either the wife herself or the village trancer receives. Thus, as a member of the
Mahanubhav sect, one could no longer conduct an affair in secret: Bhagwan would reveal
all. In all this then, the marital bond is emphasised, sometimes explicitly, sometimes
implicitly, perhaps to the detriment of other sorts of relationships. As the adherents of the
sect seem to seal themselves off from the wider village community, so husband and wife
seal themselves off in certain instances from ‘outsiders’. This ‘sealing’, however, also
goes one stage further, to the level of the person afflicted by the attack. In a rite
conducted while people are at Sukdi, the guru instructs the sick person to perform special
worship at the ote after having observed a rule: for three days the temple visitor must not
let anyone touch his sleeping mat (chatay). The sleeping mat in Sukdi indicates one’s
presence and one’s location in the temple complex. By sealing the mat in this way, even
from one’s own family, one is in a sense indicating one’s separateness and one’s mistrust
of other people: the mat is sealed for much the same reason that the door is closed on the
satsang, in order to stop interference with a ritual process.

Though it is the family (and more importantly the husband and wife unit) that prays
together morning and evening and in so doing asks for protection from God, in some
senses the worship is highly individual. It is not the household (or the husband-wife unit)
that receives the vishesh but sect members as they are themselves, as individuals. Thus,
Brijlal has a vishesh of his own, as does his wife Champa. When a devotee is away from
home for the night visiting relatives or friends in a distant village, for instance, he carries
his vishesh with him and worships it there, and his spouse worships her vishesh
separately at home. In striking contrast to other accounts of devotionalism (or indeed
general Hindu religiosity), which has tended to view women as prime movers in
household religiosity, among the Mahanubhav members (and other sects) in the village,
men and women are charged equally with religious responsibility, with men on the whole
taking a more active role (both within the household and publicly) than women. The

166
woman o f the household does not pray on behalf o f her fam ily; both husband and w ife

perform the act o f worship together morning and evening.

Plate 24: A Markakasa man and his son on a trip to a local beauty spot.

The practices o f the sect, w hich devotees are required to adhere to i f they are to maintain

continued protection from witchcraft and magic, em phasise the individual s e lf as a site o f

power. The binding o f the sleeping mat and the exhortation to sleep in o n e’s ow n bed

suggests the importance o f the individual s e lf in society: on the one hand the isolated

being is less likely to suffer, but on the other hand, the individual with his own vishesh is

responsible for his ow n level o f protection from future suffering. Dum ont seem s to agree:

“ in order to con ceive o f a personal Lord there must also be a believer w ho sees h im self as

an individual” (1980: 282).

Disputes as a consequence o f sect membership could be more serious than the exam ple o f

the two brothers given above. Although Dinu, Lomu, their elder brother and their father,

Gangaram, eventually split from each other and established separate households, relations

between them remain good and there is much com ing and going betw een the various

hom es. One reason for this is that ‘conversion’ is not as dramatic as in other cases; here,

167
Gangaram has been a long-time adherent of the Kabir Panth, also a devotional sect.
Another family in the village, however, had a much more significant falling out, precisely
over this issue of sect membership.

The Koreti are Gonds and there are three households in the village, all agnatic kin. The
late Budda Koreti was the first member of his family to come to Markakasa from what is
now the new state of Chhattisgarh, approximately sixty years ago when he was just a boy
of about ten or eleven. His village was suffering from famine and drought so he came and
worked as a house servant (naukar) in the village headman’s (thakur) household. His
brother and sisters followed him and, when they could, they built a house in the village
and acquired some farm land. Budda developed into a charismatic man and a powerful
pujari (or baiga: a sorcerer), and someone whom the gods had chosen to possess. The
Koretis had done well in Markakasa: from nothing they had managed to become one of
the wealthiest and powerful families in the village. This was not only the result of hard
work and good fortune, they said, but also because they had not abandoned the dev-
dharmi (gods and religion) of the Gonds120. As a consequence, worshipping the Gond
deities and household gods (lonchtun)121 seemed very important to them and, of all the
Gond households in the villages, the Koreti employed baiga the most readily to intercede
with the gods in order to solve problems.

Now, Budda’s brother had two sons, both of whom are married and live in the village.
About five years ago, the wife of Ramdas, the elder of Budda’s nephews (bhatija), went
mad. She would talk nonsense to herself and attack people without warning and only let
people visit her if they brought gifts of food. Ramdas told me he had tried everything to
find a cure: he consulted doctors in hospitals and asked baigas to perform magic but
nothing seemed to work. The longer the madness lasted, the more witchcraft was
suspected. He went to see a baiga in a village about thirty kilometres away and sacrificed

120 In most contexts this was not distinguished from that o f Hindu dharma or Hindu dev; these were seen as
the same as the Gond gods and dharma. The difference as people saw it was in the form and structures of
worship. The only times the term ‘Gond dharma ' was referred to in an exclusive sense was when it was
contrasted with the kind o f dharma practiced by members o f a religious sect, most importantly, the
Mahanubhav sect.
121 Each Gond family (extended) has two shrines, often located outside the house in a dwelling called
kuratandi: one for the household gods (lonchtun) and another for the hanal, or the spirits o f ancestors.

168
a Rs. 1000 goat, and for a time the madness ceased. Then it resumed and, as in the case of
Brijlal discussed above, someone suggested they go to Sukdi, where if they prayed they
would receive protection. Ramdas, his wife Tarabai and their daughter stayed there for
five weeks where they prayed all the time, repeating the names of the five avatars and
performing puja three times a day. The madness eventually left her and their guru
instructed them on how to pray at home. They joined the sect by having their ‘ears blown
upon’ (kan phukna) and returned to the village. Though Ramdas recognised the power of
the place, he did not seem to regard it with much affection: “[Sukdi] is full of mad people
wandering around. People who are happy (sukhi) don’t go there, only those with pain
(<dukh),” he told me.

Unlike Brijlal’s rather exceptional experience, Ramdas and Tarabai’s ‘acceptance’ into
the sect (dharma suikarna) led to immediate conflict with their kin. Like many new
adherents, they began to cook separately, and of course refused to worship both the
lonchtun (household gods) and the hanal (ancestors). By abandoning his gods, Ramdas
was charged by his uncle, Budda, of giving up his Gond-ness and wanting to be separate
from the rest of the Gond community. Ramdas has since built a separate house at the
other end of the village and does not now visit his uncle’s or younger brother’s house, nor
do they visit him, though his relationship with his wife’s parents (also his FZ and FZH)
who also live in the village does not seem as strained, partly because the refusal to
worship household deities does not directly concern them. Yet Ramdas is emphatic that at
no point did he quarrel with his samaj, the Gond community (his caste-fellows); the
dispute in this instance is a matter of the ‘household’ {ghar). His assertion to me that he
was a fully engaged member of the Gond community, a man who attended Gond caste
meetings and helped at Gond weddings, demonstrated his desire that I acknowledge that
his ‘Gond-ness’ had not altered despite the change in ritual practice. Like Jaduram,
Ramdas had no desire to question his caste identity; despite changes in certain spheres of
sociality and in the notion of the self, caste remains the hardcore. A different self which
restricts sociality and is empowered by the ability to resolve existential problems of
illness and pain has certainly emerged, but some ties continue to have meaning. Ramdas
is no individual-in-the-world.

169
Disputes with non-sect members, be they kin or neighbours, are treated philosophically -
as almost part of the process of devotion and of being an atal bhakt. “After all, a country
cannot function with only one [political] party; it needs two to make it work”, Brijlal said
to me the following day as we were discussing the satsang of the evening before. Aside
from demonstrating rather wonderfully the workings of parliamentary democracy, his
statement expressed that conflict and opposition were to be expected and only made
devotees firmer in their faith.

While for many people, the act of joining the sect may have led to many difficulties with
kin, it also results in the adherent becoming aware of another association, that of their
sect. Interestingly, the language used is that of kinship: all the disciples of one guru are
called guru-bhai (guru-brothers) and this is a term that is extended to and used by women
too. The guru is never in fact referred to as such and is instead called babaji or ‘father’. In
both of the satsang I attended, the devotees elaborated the common sense of belonging
further as they talked about themselves as ‘the children of one mother’ (ek mai he
pilla)m and emphasised the need for unity among the corps of devotees even in the face
of public, though more commonly private, hostility. Indeed this seemed to be the main
purpose of the very first satsang held in the neighbouring village of Ganutola thirteen
days after the death of the guru.

Though a hamlet of less than fifty households, Ganutola’s location on a road between two
larger villages, the presence of a government boarding school for Adivasis and an Armed
Police Outpost mean that it has attracted several traders to set up shop and is the site of
the nearest weekly market to Markakasa. The village has a large number of Mahanubhav
panthi devotees and is also the home of a Gond man, Bhagat, who was one of the first
people in the area to go to Sukdi and join the sect about ten years ago. It was Bhagat’s
daughter who took the byan for Brijlal and Champa when they went to Sukdi in order to
seek relief. This first meeting following the death of the guru was held in his house,

122 Being the ‘children o f one mother’ is regarded as the closest kin relationship one can have, closer than
for example being the children o f one father: a father can have many wives all o f whom are mothers but
only one o f whom is regarded as the ‘real’ mother by her children.

170
chosen because of his position in the community and because his house was large enough
to accommodate the forty people who attended. The meeting had several purposes: to
perform the death rites of their guru held thirteen days after death when the soul of the
deceased is dispatched; to perform a satsang, in accordance with the guru’s wishes
revealed in the dream; and to enable all the disciples of the guru and members of the sect
to get to know one another (parichay). As I show below, one of the major concerns of the
devotees gathered together that day was a fear of misrecognition.

Before we turn to an analysis of this first satsang, let us take a look at greeting and the
process of recognition. Men in this region generally greet one another by saying “Ram
Ram”, though most of course dispense with this formality with people (such as fellow-
villagers) whom they see often. Women, when greeting, do not usually say “Ram Ram”
to one another or to men. Castes and religious sects alike have their different greetings;
but one can only know which greeting to use if one knows who that greeted person is.
The process of greeting is important since it is one way in which other people, spectators
to the greeting, can observe the signs of community. Thus, generally, only a member of
the Mahar caste would greet another Mahar (about whose caste identity the greeter was
certain) with “Jai Bhim”, though of course not all Mahars greet each other in this way all
the time.123 Likewise people of the Teli caste usually greet one another by taking the
name “Saheb”, associated with the Kabir devotional sect of which many Chhattisgarhi
Telis are adherents. That people greeted one another in these multifarious ways was
commented on extensively by villagers and was an important way in which to establish
the caste or sect identities of people one had seen or had just met. Another mode of
recognition is the application of tika (vermillion powder) to the forehead. In general the
presence of the red powder in the space between the eyebrows indicates a member of a
religious sect, whether it be Mahanubhav, Ramanandi, or Kabir. Indeed this is one of the

123 Often, greetings such as “Jai Bhim" and “Jai Sew a" (the greeting used by the Gonds) would be used
jokingly between non-Mahar or non-Gond people. Sometimes, the joking aspect was clear; at other times,
especially when the greeted party was o f that community, the use o f the greeting was I felt more ambiguous
- mocking or respectful depending on the context. Though I did not observe such a pattern in my fieldsite,
Burra (1996) notes that Mahars in western Maharashtra tended to use the “Jai Bhim" greeting in public in
front o f mixed company but reverted to “Ram Ram" in private.

171
ways in which non-sect people refer to sect adherents, by imitating the act of applying the
tika.

Though their tika look rather generic, adherents of the Mahanubhav sect do however have
a particular greeting, “dhandavat”, which has the effect of marking them as members of
the sect to observers. All devotees, both men and women, greet one another by saying
“dhandavat” accompanied by a folded hand gesture and a bow124. In form, it is
significantly different to the action which accompanies the “Ram Ram” greeting in which
the greeter touches the hand of the greeted and brings his own hands back to his chest,
thus physically establishing a connection. In the dhandavat no physical connection is
made or sought and the entire act looks far more ritualised than the simple “Ram Ram”.
The saying of it is not rushed; if two Mahanubhav panthis (of either sex) pass one another
in the street, for example, the parties will stop, face one another and perform the
dhandavat even if no other conversation then ensues. When meeting a large group of
fellow devotees, such as at a satsang, care is taken to greet each devotee in turn with a
separate dhandavat, and again the space and time of the dhandavat becomes highly
ritualised with conversation ceasing until the newcomer has finished greeting everyone.
The dhandavat is an acknowledgment of belonging to the same moral community and
acts as an important marker of its presence to those not members of the sect. But the
dhandavat also signals something more: that it is said by each person individually to each
other person individually, even if one is in a group, suggests that, for that ritualised
moment in time, the group falls away and the individual members of the sect relate to one
another as individual beings and not as members of a group. Conversely, the “Ram Ram”
greeting in a group setting is usually uttered once by the newcomer and is meant to be a
salutation to all. In the style of utterance of the dhandavat, the person seals himself off
for a brief moment and through this practice expresses his sociality from a markedly
individualistic perspective. But sect members also aim to establish the grounds by which
a new sense of belonging can be recognised, one which involves being able to recognise
others who are like them. So, let us now go to the first satsang, held shortly after the guru
passed away.

124 The speech and the gestures are collectively known as dhandavat.

172
After the devotees had gathered at Bhagat’s house, and after they had observed a few
moments silence in memory of their guru, all those who were present were called upon to
introduce themselves, to give their parichay. Each man (there were no women present at
this initial satsang) gave his name, his village and the name of his guru (which, apart
from a couple of exceptions, was the same for everyone). While the guru had lived, they
had not needed to know who other members of their panth were, someone said: the guru
had visited everyone and they had been connected through him. Now that he was dead, it
was important to recognise just who one’s gurubhai were. Just before the satsang had
formally begun, for example, I noticed that my landlord Brijlal had expressed surprise on
seeing an acquaintance from a neighbouring village; he had had no idea that this man was
a member of the sect. That Mahanubhav panthis would not recognise one another was
seen as a problem: “Imagine you meet a gurubhai but you don’t know that he is a
member of your sect and you greet him with 'Ram Ram’ instead of ‘Dhandavat’. That
would be a grave mistake (g halaty This misrecognition had the effect therefore of
denying the presence of a community and again, in their opinion, led to the impression of
an inconsistency in practice from the outsider‘s point o f view. Likewise, it seemed to be
desirable to create a map in space of where one’s gurubhai lived, again to be able to
imagine the extent of the community of sect members. It was important, one man
declared at the satsang, to know where their sect-fellows lived so when talking about
such-and-such village, one could say, “yes our bhai (brothers) live in that village”.
Membership of the sect is not spread by active proselytisation on the part of the gurus or
by its lay membership; despite the fact that the former spend a considerable part of the
year travelling from village to village, they meet only members of their own sect and do
not hold public sermons (pravachan) as gurus of other Hindu sects do. Indeed, as we
have seen, the sect is extremely conscious to keep non-initiated people at arm’s length.
Going to Sukdi and becoming a member of the sect is seen as an extremely difficult thing
to do. Not only is it expensive, but once entered into it cannot be cast off easily without
the risk of madness. Thus, as discussed above, it is seen as very much the last resort of
people who have tried everything else. And the defining feature of the decision to go is
the suggestion by a member of the sect that perhaps the time has come to give the

173
temples at Sukdi a try. Information about the sect’s power is spread by word of mouth.
Knowing the social geography of the sect (who is an adherent, and where) is thus an
essential part of increasing the membership of the sect, though this is not explicitly stated.

The restriction of sociality and the creation and recognition of a new sociality have
considerable parallels with the discussion of ritual friendship in Chapter Three. There I
argued that the ideology of ritual friendship asserts that love and affection are the source
of the association and that this enables it to be contrasted with other types of
relationships. Like gurubhai, it too is a relationship that does not distinguish on the basis
of caste or class, and has force because it is a way of expanding sociality safely beyond
troublesome (agnatic) kin. Though I do not suggest that members of the Mahanubhav sect
create more ritual friendships than other types of people, the fact that a model already
exists for recognising people as similar to one another in spite of caste and class may
make the notion of gurubhai all the more meaningful.

I have argued that membership of the sect, and the kinds of practices and doctrines it
requires of its adherents, enable the individual person take control of himself over matters
of illness and well-being. Just as the sect member removes himself from certain forms of
sociality, he engages in new forms with his fellow adherents, though it must be stressed
that this latter is not really necessary for a devotee to maintain either his status as a
Mahanubhav, or his level of protection from harm. In becoming associated with the
Mahanubhav sect however, and especially through the act of pilgrimage, it is possible to
discern another form of belonging, namely a more pronounced sense of Maharashtrian
identification.

Pilgrimage and the idea of Maharashtra

For adherents of the sect, journeys to the temple towns where Chakradarswami preached
and stayed are seen as desirable. Indeed, if an adherent suspects that he has been the
victim of attack and successfully protected through devotion to God, he makes a promise
to Bhagwan to visit a temple and give offerings (Chh. naryal chudhana\ lit. “to raise up a

174
coconut”). Most Mahanubhav adherents in Markakasa visit Sukdi every year, though
more men than women seem to accomplish this. In addition to these ‘individual’
journeys, there are also more ‘organised’ pilgrimages which occur in the Hindu month of
Chaitra (March-April) and which take in several of the larger Mahanubhav temples at
Mahur and Riddhpur in other parts of Maharashtra. A visit to a Mahanubhav temple is
said to be particularly efficacious during the latter half of Chaitra since it is at this time of
the year that the power to fight witches and ghosts is especially strong. During this time, I
was told, Sukdi transforms from a sleepy backwater into a thronging town packed with
pilgrims and the afflicted.

1
Markakasa people see themselves as living in a borderland , between the cultural
worlds of Maharashtra (‘Marathwada’)126 and Chhattisgarh. Neither entirely
Maharashtrian nor entirely Chhattisgarhi, most feel out of place on their visits to other
areas. This sense of belonging nowhere was often expressed in discussions of language.
People would say that, though they might know several languages (Chhattisgarhi,
Marathi, Gondi, Hindi), each was impure (ashuddh) and a ‘mishture’ (mixture). Travel
west to Nagpur and beyond and one would hear ‘pure Marathi’; go east to Raipur to hear
‘pure Chhattisgarhi’; and go south, deeper into the forest areas (jangalpatti) for ‘pure
Gondi’, I was told.

This border identity also finds expression in discussions over agricultural techniques:
whereas in ‘Marathwada’ the dominant method is transplanting rice shoots (rohona), in

125 Interestingly, Chakradhar exhorts his followers to go to these very borderlands. The Mahanubhav texts
remind adherents to emulate the founder and go and live at “the end o f the land” (deshacha shevat). As
Feldhaus writes, “according to them, ‘the end o f the land’ is an area in which people speaking Marathi and
another language intermingle. The periphery is understood linguistically, as a borderland between two
language regions” (2003: 188). Not only should followers travel to borderlands, but the places they go to
should be off the beaten track: “ .. .followers are to keep to ‘miserable little villages’, they are to stay ‘on
hillsides and o ff the road’ and they should sleep under a tree or in an abandoned temple outside a
village...The most frequently repeated o f Chakradhar’s commands in this connection is the command to
stay ‘at the foot o f a tree at the end o f the land’” (Feldhaus 2003: 186).
126 The use o f ‘Maharashtra’ and ‘Marathwada’ in different contexts is significant. ‘Maharashtra’ was used
to identify the political state entity and as such the villagers recognised that they were in Maharashtra.
‘Marathwada’ on the other hand was a description signifying cultural content such that ‘Marathwada’
meant places where people spoke Marathi, built their houses in a Marathi way, and followed Marathi
customs. Thus, though Markakasa was in Maharashtra, it was not ‘Marathwada’. Also, Marathwada as used
by Markakasa people refers not to the specific region o f Marathwada within the state o f Maharashtra but to
all o f Maharashtra where Marathi is the dominant language.

175
Chhattisgarh the main mode is sowing seed by hand randomly over a ploughed field
(perni). Needless to say, in Markakasa and the surrounding area they use both methods.
What is important here is not whether in reality farmers in ‘Marathwada’ or Chhattisgarh
use the methods ascribed to them by Markakasa people, but that their explanation of why
there are two types of technique commonly used by them is this condition of living in a
border zone.

One consequence of joining the Mahanubhav sect is a growing awareness of a sacred

geography proper to the territory of Maharashtra. The imagination of sacred geography
among non-sect villagers is limited to pan-Indian centres such as Kashi (Benares) and
Puri (in Orissa) on the one hand, and local centres such as the temple of the Goddess
Bombleshwari at Dongargarh in Chhattisgarh on the other. As discussed at the start of
this chapter, the Mahanubhav sect is a distinctly Maharashtrian sect. Feldhaus writes,

“ ...the Mahanubhavs are...extremely important for understanding the

beginnings of regional consciousness in Maharashtra. For the
Mahanubhavs were among the first to choose Marathi as a language of
literary expression, and their religious literature was the first to ascribe
religious meaning to a whole region that they called ‘Maharashtra’”
(2003: 185).

Indeed, Feldhaus’s analysis of key Mahanubhav texts leads her to conclude that they
regard Maharashtra as better than other lands: where the natures of the latter are rajasik
(characterised by earthly passions and desires) or tamasik (characterised by ignorance
and baseness), Maharashtra is sattvik (pure and good) and so more conducive to healing
bodily afflictions (ibid: 193). She also highlights the importance the texts give to location
and place and suggests that Mahanubhav pilgrimage is centrally concerned with
remembering the deeds of the founders of the sect. The detailed listing in the Sthanpothi
(The Book of Places) of villages and towns visited by Chakradhar and the other founders,
and careful noting of the places where they stood, sat, stayed, ate, slept and defecated,
suggest to Feldhaus “that rather than being a pilgrim’s guide, the Sthanpothi permits the

176
Mahanubhav devotee the possibility of thinking about the places without actually going
to them” (ibid: 196-8).

Several scholars have noted the importance of region as a way of mediating the
relationship between the local and the nation in India (e.g. Benei 2001; Bayly 1998). The
identification with Maharashtra and Marathi-ness through membership of the sect may
enable this for Markakasa people. Certainly, many Mahanubhav sect members told me
how their proficiency in the Marathi language improves after having stayed at Sukdi.
Everyone there speaks Marathi and that is the language they use to talk with their guru.
For those who are literate, the main texts of the sect such as the Lilacharitra and the
Sthanpothi are in Marathi too. There is awareness in the village of the importance of
Marathi. The medium of instruction in the Markakasa school is Marathi, a language
which only a small proportion of children are familiar with. Teachers, parents and
students alike complained that most people failed their Marathi exams and performed
worse at school than people in other areas, largely because Marathi was not the language
of the home for the majority of pupils127. Even the Mahars who speak Marathi as their
first language complained that it was not pure and mixed with Chhattisgarhi.

Those sect members in Markakasa who had time and money visited the places of
pilgrimage dotted around Maharashtra. The temples at Mahur and Riddhpur, west of
Nagpur, were frequently mentioned both in conversation with me and amongst
themselves, and these became more important than the pan-Indian or more localised
pilgrimages. Membership of the sect draws people unfamiliar with Marathi ways into
closer contact. What effect this has on other spheres of life, such as marriage patterns (do
more people marry Marathi people as a result?), remains to be seen.

127 A study carried out by a local teacher examining primary and secondary education in a group of twenty-
four villages including Markakasa concluded that the lower pass rates and lower level o f formal schooling
in those villages is closely linked to the language o f instruction. Most students speak Marathi (if at all) as a
language after Gondi and Chhattisgarhi, and he proposed that the schools in this borderlands area ought to
devote more time to improving the level o f Marathi in the early years o f schooling (personal
communication, I did not read the dissertation myself).

177
Up until now I have discussed these changes in relation only to the Mahanubhav sect, and
located the transformation firmly within Hindu religiosity. There seems however to be a
local logic for this sort of transformation, which leads to certain people becoming
Christians for the same reasons as those who become ‘better Hindus’. Interestingly then,
viewed in light of political and cultural Hindu nationalism (which aims to make
Christians the ‘other’), both ‘good Hindus’ and ‘othered’ Christians are products locally
of the same social processes, which have led to problems in dealing effectively,
definitively and cheaply with attacks of witchcraft and magic.

A Christian Coda

On my return to Markakasa for a month-long visit in April 2005 (after an absence of one
year), I discovered, entirely by chance a couple of weeks into my stay, that two
households in the village had converted to Christianity. While waiting for his mother to
turn up, I had been helping a boy revise for his up-coming sociology exam. He had been
attempting to memorise key facts about the religions represented in India and then, when
we got to Christianity, mentioned that a few months after I had left the village the year
before, Christians had visited the area and some people had joined them. He seemed
unsure about who exactly had converted but I learnt from further enquiries that day that
the newly Christian families were those of Baburao and Soma, both of the Mahar caste.

Baburao is a Mahar man in his sixties, with two sons: one lives in Mumbai and works in
a restaurant near Dadar station; the other, Durga, lives now in Markakasa and had spent
ten years working as a cook in the State Transport bus station canteen in Nagpur, a job
which helped him buy an acre of land. Though Baburao himself is poor with only a three-
quarter acre of farmland, and a modest house in the new ‘SC para’ of the village, he is the
MBS of one of the wealthiest men in the village. It was the first time I had visited his
house since my return, and fortunately I found Soma there too.

Baburao narrated the circumstances which led to him and his family becoming
Christians. Durga’s seven year old son had been ill for some time with a chest disease

178
and, though they had spent a lot of money on medicines, the boy remained sick. Baburao
too suffered increasingly from shortness of breath and likewise could not be cured. So
they began to suspect that the cause of both their illnesses was something that medicine
alone could not cure. At around the same time, he said, a group of Christians visited the
village and said that if they sought relief they should come to the ‘mandir’ (church) at
Korchi some thirty kilometres away to the south, where Durga’s son would be healed.
They made the trip to Korchi and were taught how to pray. There they also learnt prayers
(prarthana) and hymns {gana), and were given bibles in Marathi and Hindi. Reading the
words in the Bible protected them from harm; Durga’s wife told me that whenever she
felt frightened or had the sense of something was going to happen to her (M. “mala
shanka all, atta kayatri mala honar ahe” i.e. that she was being attacked by witchcraft)
she would recite the Lord’s Prayer.

Soma went to the church in Korchi for much the same reason. His house had been
plagued for a number of years by a bhut (ghost) sent by an enemy (dushman)m . No
matter how good the harvest had been, they never had enough food: the bhut was eating
their rice. Since he began praying to Jesus Christ the situation had reversed, and now they
could feed ten people a day and not run out of grain. This bhut had also stopped the
family sleeping: it would throw things around the house and enter their dreams. Though
he had employed several baiga to fight it, the bhut had always returned. Now it had gone
for good.

The short narratives presented above bear a striking resemblance to the way in which
people who joined the Mahanubhav sect described their experience and how they
explained the importance of their new religious attachment. For both sorts of religious
change, incurable illness caused by witchcraft or magic motivated the decision to seek
sources of healing other than doctors or baiga; and the desire for continuing protection
produced an attitude of commitment on the part of the newly converted. Indeed, Soma
drew the comparison during our conversation: “it is like what Brijlal and those others
Sukdi-walleh did - they were near death and they went to the temple and felt better - it’s

128 An enemy he declined to name.

179
exactly the same thing.” In order to continue being protected from harm, Baburao slept
with the Bible under his pillow. They recognised that the ritual was different too: instead
of lighting incense or offering coconuts, Christians only needed to get down on their
knees with a lighted candle in order to pray. Like the Mahanubhav, Soma’s wife told me,
there are no idols (murthi) in the ‘temple’.

And there is another similarity. Just as Jaduram and Ramdas, discussed above, were keen
to emphasise their ongoing caste commitments of Mahar and Gond respectively, despite
their adherence to the Mahanubhav sect and in the face of challenges from their non­
sectarian caste fellows, so Soma and Baburao denied that they had given up their
‘Boddha dharma’ (Buddhism) despite having become ‘Isai’ (Christian). Though I could
not investigate these new Christians’ changing notions of self and sociality in any depth, I
suspect that this conversion to Christianity accords rather well with Dumont’s discussion
of the continuing importance of caste for members of Hindu sects.

What is also interesting about these recent Christian conversions is that they raise
questions about novelty and secrecy. A consideration of these concepts as seen by
Markakasa Christians and Mahanubhav sect members might help illuminate, in terms
other than those discussed so far, why religious change of this sort might be attractive.

Why did these two Mahar households not take the well-trodden path to the Sukdi temple
door in search of relief from these attacks? When I asked them, they both said that the
Christian temple was closer and less expensive. Unlike at Sukdi, which often required a
stay of five or six weeks, they only needed to stay at the Church for a few days. While
these are clearly important considerations, I wonder if there is not something else going
on too. Perhaps the attractiveness of the Christian path to healing, to transform oneself
from a suffering to a happy person, was located in the novelty of the experience. The
novelty in itself helps restrict the bonds of sociality with one’s kin and caste-fellows,
because one now worships in a different way and has created a different set of
attachments unassociated with the past. As I talked to both Soma and Baburao’s families,
I sensed their excitement in doing something new: their Bibles (both in Hindi and

180
Marathi) were immediately produced, and Baburao’s SW began to recite the Lord’s
Prayer without being prompted before going on to explain in detail what the church
looked like and what they did there. The importance of novelty has been noted by
Kapferer (2003: 20), who suggests, following Douglas (1970, 1973), that for Sri Lanka
and elsewhere in the world, much ritual practice (including sorcery practice) derives its
force from the fact that it is regarded as originating from ‘the outside’: its power inheres
in its newness and separateness from people’s existing social worlds. Conversely,
“sorcery that is well-tried loses its strength”, he claims (ibid).

Connected to novelty is the issue of secrecy, which is a significant element both of going
to Sukdi and visiting a baiga. Markakasa people were insistent that, if one wanted to
consult a baiga, one should not tell anyone that one was going; and indeed this is how
they behaved. On several occasions, I witnessed people who I later learned had visited a
baiga, pointedly ignoring questions asking where they were off to. I once accompanied a
villager on a trip to another village to procure seven chicks of certain colours, required by
the baiga he had engaged to help cure him of his constant headaches and dizziness -
brought on, he suspected, by witchcraft. As we were cycling back home, he asked me not
to tell anyone what we had in our bags: the fact that a baiga would be visiting his house
that evening was to be kept a secret. This not only echoes the discussion of the ‘sealing
o ff of the house discussed earlier in the chapter, but also the way in which Mahanubhav
sect members talked about their visits to the temple. As we prepared to leave on our trip
to Sukdi, my landlady Champa began to complain rather loudly that too many people had
got to know about our journey. Her comments were aimed at me: I had told too many
people and had invited them along.129 At first I thought it was because she was impatient
to leave and fed up with having to wait for these other people to turn up, but in fact she
had a far more serious concern. Her guru had told them that if they wanted to visit Sukdi
they should do so without telling anyone and should leave as early in the morning as they

129 I was coming to the end o f my fieldwork and I had certain debts o f kindness I wanted to repay. Since I
was paying to hire the car to Sukdi, I invited three village friends along (the junior Thakur at whose house I
often ate, a Teli man and my Mahaprasad). My landlady was also upset at the presence o f Latchu and
Mukta, the adopted son and daughter in law o f her HeB, Manu, who were also sect members but whom
Champa suspected o f magic. In fact, I had not asked them to come; they had heard I was going to Sukdi
and invited themselves along.

181
could, before anyone could find out or ask questions. “If someone finds out you are
going, they might try to do something to you”, she explained semi-euphemistically.
Secrecy characterised both the consultation of baiga and going to the temple at Sukdi.

The similarities between the two raise a couple of points. Firstly, in analytical terms, one
can see the Mahanubhav sect (and perhaps the Christian Church) as types of sorcery used
to counter other forms of witchcraft and sorcery. That the Mahanubhav might be seen as
such is also suggested by the fact that many non-Mahanubhav Markakasa people saw the
sect’s temple as a dangerous place for non-sect people to visit. Consider the example of
our trip to Sukdi once more: Champabai complained that too many people were getting to
know that we were going and that this would hamper the effectiveness of the visit. I had
invited my Mahaprasad, Ganeshyam, along too, and though he had initially said yes, as
the time approached to leave he came to the house and said he couldn’t come: his brother
had errands to run and could not watch the shop in his place. This was clearly an excuse
(his brother never had anything urgent to do) and we asked Ganeshyam what the real
reason was: did his family not want him to come? He admitted that his mother opposed
him visiting the temples at Sukdi: “she says that if go to Sukdi I will fall ill or worse, and
that the power (shakti) there is too strong and harms people who are not panth-walleh
(sect members).”

The discussion raises a second point: the secrecy and novelty of the Mahanubhav type of
sorcery was being compromised by its popularity. So perhaps the reasons why the
Mahanubhav sect may not have been attractive to Soma and Baburao were that it was
already coming to be seen as an old and non-exclusive form of sorcery, that there was no
secrecy left in its practice and that a large number of people in the village had already
been to Sukdi and already knew protective mantra associated with the sect. The Christian
church and modes of worship on the other hand were new and uncommon in Markakasa
and the surrounding area. As such it was a new form of ‘sorcery’, all the more powerful
because its practices were strange and unknown. While baiga are repositories of secret
knowledge that can only be learnt after long periods of training, the secrets (and thus
power) of the Mahanubhav sect are available to all simply by joining the sect, becoming

182
the follower of a guru and keeping to the sect’s strictures regarding worship, devotion
and practices. The very fact that membership of the sect (and also perhaps of the
Christian church) means that the self has responsibility for how magical attacks are dealt
with also has the corollary that, as a form of sorcery, it is in fact more widely and
immediately accessible (provided one keeps the faith) than that which is exercised by
baiga, who manage to keep their mantra secret from all classes of people. As the
Mahanubhav sect becomes more popular, therefore, the power that it wields in terms of
secrecy and novelty diminishes and opens the field for other forms of sorcery.

Conclusion

Adherence to the Mahanubhav sect has the effect of reconfiguring sociality, especially
with kin and neighbours. The experience of troublesome and dangerous kin relationships
makes a religious sect which emphasises exclusivity and the restriction of sociality
particularly attractive.

Through this discussion of the changing notions of the self and sociality, one can discern
several tensions. Firstly, though kin are regarded as problematic (which is why sociality
is restricted with them in terms of dining, cooking and worship), the closest of kin - one’s
wife, daughter or mother - are essential elements of the process of healing because they
take the byan which leads to one being cured. Secondly, though the self takes
responsibility for his own level of protection as a consequence of membership of the sect,
a state of affairs seen as highly desirable given the problems which exist in asking for
assistance from others, the sect member is still a man-in-the-world and so cannot restrict
his sociality completely: his caste has enduring importance in the constitution of society.
Thirdly, though the Mahanubhav sect is appealing because it is effective against
witchcraft and sorcery, the power it exerts through its ritual practices can be seen as a
type of sorcery too. As such, the more popular it becomes the less powerful or effective it
may appear.

183
The perspective in this chapter has largely been from the point of view of the sect
member and the partial creation of a new self in society. In the chapter which follows, I
show how a change in food and drink practices as a consequence of sect membership not
only fulfils a functional requirement of protection from harm, but also transforms into a
way of thinking about a person who is ‘more improved’ or developed than others. This
then feeds into a more widespread local discussion influenced by Hindu nationalist
activists and ascetics about the constitution of the virtuous society.

184
Plate 25: Building a tomb during the Gond death ritual o/karsad

Plate 26: The sorye (affmes) o f

the dead playing chikal (mud)
on a tomb during karsad.

185
Plate 27: Karsad. Women affinal kin o f the dead carrying pots which will contain
a frog in water, representing the jiv (life) o f the dead.

Plate 28: A group o f young men and women dancing during karsad.

186
Plate 29: A baiga sacrifices a chick on a Gond tomb during karsad.

187
 Chapter Six
From Protection to ‘Improvement’: The Commitment to

Vegetarianism and V isions o f the Future

A cadem ic writing on the practice o f vegetarianism in India has proceeded on the

assumption that those on the low er rungs o f society - A divasi people and lo w castes -

give up eating meat out o f a desire to change their social position. In its baldest form, as

stated by M. N . Srinivas, the suggestion is that, in em ulating high caste Brahminical

practices, those who are low attempt to improve their position in a purity/pollution-

govem ed caste hierarchy in a process known as Sanskritisation (1 9 6 6 )130. Srinivas also

highlighted the prestige associated with a vegetarian diet, and many scholars suggest that

marginalised people (either as individuals, households or groups) adopt vegetarianism in

order to increase prestige more generally, regardless o f any clear sense o f caste hierarchy

(see e.g. B ailey 1960, 1961; Furer-Haimendorf 1982; Singh 1985; M endelsohn and

V icziany 1998; D eliege 1999). W hile these studies often assum e that m otivation for

changing dietary practice lies in the increase in status and the accum ulation o f prestige, I

aim to shift the perspective som ew hat and exam ine what vegetarianism m eans for the

person as a com ponent o f a wider transformation, part o f w hich w as discussed in Chapter

Five. One o f the consequences o f becom ing a member o f the Mahanubhav sect, in

Markakasa and in the area more generally, is the adoption o f a vegetarian, teetotal diet.

For Markakasa people, the adoption o f a vegetarian diet as a consequence o f m embership

o f the sect so lv es ‘being-in-the-w orld’ problems: by join in g it and living by its rules they

130 It was Srinivas’ (1952) study o f the Coorgs o f south India that suggested the content o f a theory o f
Sanskritisation. Srinivas found that the Coorgs were changing their dietary habits and aspects o f their
religious life. But he did not see this as necessarily a stab at increased status since the Coorgs were then the
dominant caste locally. The way in which Sanskritisation was formulated in that ethnography was in fact
more nuanced than the way it was subsequently developed and taken up both by others (see above) and
Srinivas him self (e.g. 1966).

188
attempt to protect themselves better from attacks of witchcraft and magic. In fact, this
desire for protection is what gives vegetarian practice its power.

Insisting that the adoption of vegetarianism is evidence of a search for social prestige
would be a misreading of the significance of vegetarianism, resulting from a
disengagement from the view of people who adopt the practice. The social scientist often
refuses to acknowledge how and why motivation for action (such as changing dietary
habits) may have its roots in religious transformation because, according to the terms of a
particular version of modernity, this motivation appears either pre-modem or objectively
non-transformative.131 Rather, explanations are sought which are more easily regarded as
true engagements with modernity on the part of marginalised people, explanations such
as the impact of anti-colonial nationalism, Hindu nationalist activity or, in the case of
Sanskritisation, the desire to escape marginal status by acquiring social prestige. The
experience of religion is un-coupled from both society and politics.

Religious movements among lower caste groups, which promoted not only changes to
worship but also vegetarianism and teetotalism, have been analysed by elite nationalist
Indian historiography as expressions of rebellion and resistance to colonial rule. David
Hardman’s history of the Devi religious movement among Adivasi people in southern
Gujarat in the early twentieth century seeks primarily to re-consider this movement from
the bottom-up and not necessarily as expressive of anti-colonial sentiment. He also
disputes the Sanskritisation model, and sees the adoption of vegetarianism and alcohol
abstinence as an appropriation and democratisation of Brahminical notions of purity and
a way in which Adivasi and other low status groups sought to deprive these notions of
their power (1987: 6). My position is that vegetarian diet has a power that is embraced by
people not necessarily because of the Brahminical status of that value but because of its
appeal to experience. But let me be clear: I do not claim that the theory of Sanskritisation

131 See e.g. Wakankar (2005) for a discussion o f the implications of such an approach in examining the
historical figure o f the religious reformer Kabir. In that paper, Wakankar suggests that the causes and
consequences o f a low caste Hindu, Kabir, becoming a Muslim have been erased in official nationalist
historiography in favour o f an emphasis on the ‘typically Indian’ syncretism o f the sect Kabir founded.
Thus the transformative act for the actor (Kabir) is not examined and a religiosity founded in low caste
experience left unexplored.

189
is without value - the evidence presented here cannot possibly do that - only that the
method used to reach its conclusions be made more rigorous.

A focus on embodiment, on the way in which bodies, through experience, become

socially and culturally creative, can reveal the power of a practice such as vegetarianism
in this context. As we have seen, the experience of pain and suffering and of being healed
at Sukdi has importance for Mahanubhav sect members in terms of the body. The body is
the way in which suffering is experienced, whether it be madness, headaches, weight loss
or sleeplessness. Healing at Sukdi also occurs through the body, seen in the violence of
byan and the bodily disciplines imposed while one is in residence there. It is the
remembrance of these experiences and the desire never to suffer again that make the
practice of vegetarianism so powerful for sect adherents.

By suggesting that low status people act largely in reference to elite groups and elite
modes of being, one not only denies the ability of those people to make sense of their
lived worlds but one also pursues a methodology which leads to a fundamental
misreading of a given social situation. Thus one would not be in a position to interrogate
adequately why people give up meat, nor the power or consequences of that act. What I
propose is, rather than assume that the adoption of vegetarianism is ‘about’
Sanskritisation or nationalist projects, one should be sensitive to the ways in which its
practice may in fact be immediately irrelevant to the practitioner in terms of status or
power.

The issue is, however, complicated, not least because notions of purity and sin and the
link between these and a vegetarian life are important both for Hindu religious sects and
in Indian society more generally. In this ethnographic context however, these important
notions are merely elaborations rather than full explanations. Since the period of
suffering, as they describe it, categorically does not result from being sinful or impure but
from a sustained assault by magic or witchcraft, the purity and virtue of the vegetarian
diet cannot but be encompassed by (and thus understood in relation to) the greater
motivation for and power of membership of the sect and the resulting change in diet.

190
While the practice of vegetarianism and teetotalism is powerful for the individual sect
member because of their experience of pain and suffering, these practices do have a
wider social significance. Not only do they need to be examined against the background
of changed notions of sociality of the kind discussed in the previous chapter, but what
people eat and drink also becomes part of and contributes to the articulation of what
constitutes sudharna (‘improvement’), a local category which describes social change
with the view to making ‘a better village’. Villages in which sectarian Hindus (of any
caste) live are seen as more ‘improved’ than those which contain none. This idea of
improvement resonates strongly with local Hindu nationalist discourses on ‘uplift’
espoused by the Vidarbha Vanvasi Kalyan Ashram (VVKA) but also by holy men,
ascetics who promote the notion that the ‘good Hindu’ who gives up eating meat and
drinking alcohol can create ‘good Hindu society’.

But vegetarian and teetotal practices are not uncontested and have led to division and
tension in Markakasa, particularly among Gonds. What appears at first sight to be a
divide between vegetarian and meat-eating villagers is less about the nature of the food
being consumed than about a radical transformation in the nature of sociality itself
brought about by membership of the religious sect. Ideas about what constitutes Adivasi
culture and its relation to eating meat and drinking alcohol are used by some meat-eating
Adivasis in order to demonstrate that a vegetarian, teetotal Adivasi is a contradiction in
terms. But this, I argue, is just a way of talking through the issue; the real problem is not
the dilemma of Adivasi-ness and its association with certain food practices, but a change
in the way people relate to one another as a result of membership of the sect. Once again
therefore, the methodological focus on the transformative act for persons in their
particular local social setting reveals not only the power of vegetarianism as a way of
being but also that its practice is indicative of other more general processes of change.

The conception of ‘the Adivasi’ is not, however, irrelevant. It serves to frame the
discussion of sudharna and what makes a ‘developed village’. Being located in a

191
 1
borderlands area, which is neither ‘plains’ or ‘jungle’ , means that Markakasa notions
of sudharna, linked to vegetarianism and teetotalism, can also be seen as an expression of
distancing from the idea of jangli (wild), and thus participate in a vision of Indian
modernity through national identification.

Late in the nineteenth century, religious reformist groups such as the Arya Samaj became
increasingly identified with the growing nationalist movement which sought to raise the
consciousness of Indians as Indians in order to make demands on the colonial state. An
Indian identity was being forged which drew on the ‘rediscovery’ of the ways of the
ancient Hindus. Practices seen as superstitious and deviant from ‘pure Hinduism’, such as
blood sacrifice, were targeted by nationalists, and consequently the view that good
Hindus were vegetarian came to claim greater attention. Srinivas, for instance, reports
that in Coorg district in south India in the 1940s, “the local leaders of the Indian National
Congress have been everywhere more or less successfully opposing the propitiation of
village deities with blood-offerings” (1965: 182, cited in Fuller 1992: 99). The link
between what is eaten by the deity and what is eaten by man has been commented on, by
Fuller among others: “There is a saying in the Ramayana that ‘whatever food a man eats
the same is the food to be offered to his deities’ (Kane 1968-77, 2: 733). The reformist
corollary must be that if all deities are vegetarian, so too are all Hindus” (Fuller 1992:
102). A specific view of what it meant to be a modem Indian and a-modem Hindu was
being promoted, one which saw meat-eating as a barbaric custom and a result of the
degradation o f the ‘Indian spirit’. Alcohol, which led to drunkenness and idleness, came
under attack by the Congress nationalists too. Verrier Elwin, for instance, writes at length
of the Congresswallahs’ attempts to wean the Adivasis of central India off daru, as part
of their project of bringing the tribals into twentieth century independent India (see Guha
1999). Reforming eating and drinking practices was therefore seen by some in the
nationalist movement, especially those on the neo-traditionalist right, as important to the
task of building a modem society.

132 These are not necessarily accurate topographical descriptions. Markakasa was in a forested area
regarded as jangli by many from the ‘plains’ (in Chhattisgarh and Maharashtra).

192
So let us first take a closer look at what Markakasa people eat, how it has changed and
the social and symbolic significance given to different sorts of food.

Food and the dangers of commensality

Rice is the principal crop in this part of central India and people eat what they grow: most
households operate subsistence agriculture. Though there are significant wealth
disparities between villagers, there is little difference in the type of food ordinarily eaten.
Meals are taken twice a day and consist of freshly cooked boiled rice and one vegetable
dish or lentils. Sometimes this is accompanied by homemade lime or mango pickle. Food
is cooked on mud stoves (zula) on the floor, which are fed by firewood. In the summer,
when cooking indoors becomes unbearable, most households build stoves in the open air
either in their courtyards or back gardens. Though women are the main cooks in the
home, men do on occasion participate in the preparation of meals. They will often take on
slicing, peeling or chopping tasks, while it is the wife who grinds the chillies to make the
masala paste. Men are expected to know how to cook and will cook for their family,
especially but not exclusively when their wives are menstruating and there are no girls to
cook for the household. Diners (both men and women) sit on the floor and are served by a
female member of the family, who eats after the others have eaten. If the rains are good
then gardens can supply some of the villagers’ vegetable needs: runner beans, pumpkins,
bittergourd, tomatoes, aubergine; some of the wealthier households with larger plots of
land or access to water can grow spinach, fenugreek, cauliflowers, cabbage, chillies, and
various types of lentil. Otherwise, vegetables are available in weekly markets, held in
several villages around Markakasa and attended by most every week or fortnight.

Apart from the increased incidence of vegetarianism, there appears to have been little
1
change in people’s eating habits, at least over the past twenty years. Certainly the types
of choices that now exist for city-dwellers are not available to Markakasa villagers: to
this day most people have neither tasted ice cream nor drunk Coca-Cola. There are a

133 Given the prevalence o f regular years o f famine in the past (e.g. in the late 1950s as reported by
Markakasa people), what has changed over a longer time span is the experience o f serious hunger.

193
couple of small shacks in a village four kilometres down the road which serve snacks
such as fried bhajia (dumplings), but for a meal one has to travel forty kilometres to the
tahsil (sub-district) headquarters.134 It would be fair to say then that, for Markakasa
people, food is essentially that which is eaten in one’s own home, cooked by a member of
one’s family.

The majority of people in Markakasa, whether men or women, adults or children, eat
meat, the most common being eggs and chicken; goat and pork (both wild and domestic)
are also consumed. Though the consumption of carrion beef is not publicly admitted to,
some Chamars (those involved in the flaying of dead cattle) as well as several Mahar
households (involved in the transportation of cattle to slaughterhouses) certainly do eat
this otherwise locally abhorred meat. On occasion more unusual meats such as owl,
snake, tortoise, rabbit, or mouse will be eaten. From informal inquiries, meat (including
eggs) is generally consumed once a week and it is not meat that makes the meal but rice.

Those who are vegetarian are for the most part members of one of the three Hindu
religious sects (Mahanubhav, Kabir, Ramanandi) represented in the village. There are
also a couple of individuals who do not belong to any Hindu sect and have given up meat
in order to become disciples of a local sadhu (ascetic).

As Maurice Bloch (1999) has pointed out, food is often the best conductor both of
sociality and of anti-sociality. The meal shared creates closeness but closeness carries
with it vulnerability and the risk of the perversion of that closeness through poisoning or
witchcraft (ibid: 145). Indeed, “the better a food is as a conductor that creates bodily
closeness, the better it is as a medium of poison” (ibid). Just as with the Zafimaniry of
Madagascar whom Bloch describes, a fear of witchcraft (jadu-tohna) and poisoning often
enters into discussions about food and eating in Markakasa. It is generally thought that
one is most vulnerable to such attacks when eating. Precautions are taken to avoid eating

134 Even eating in a ‘hotel’ (restaurant) in this small town is thought of as incredibly expensive for the vast
majority o f Markakasa people. Such meals can cost Rs 20 per person, which is just under the daily wage for
a farm labourer. Foods common in much o f (urban) India such as dosa and paneer are unheard o f by most
Markakasa people.

194
in such a way that one may be seen by people from outside the household: either a door is
partially closed, a curtain hung, or diners move out of the line of sight of visitors. In my
landlord’s household for instance, where I took my meals, my landlady would whisper to
me to move further into the kitchen when visitors to the house hovered outside in the
courtyard. Likewise I was warned by my landlord’s eight year old son not to drink my
morning tea too far from the house in full view of passers-by for fear of being bewitched.
In Markakasa, paan (betel nut leaf package containing betel nut, tobacco, and lime) was
most often cited as the carrier of poison or other magical concoction, and for that reason I
was often reminded not to accept paan from strangers but from people I knew or trusted.
Buying one another paan is a highly symbolic act and a mark of closeness or intention to
create closeness; the power of its potential as a conduit for dangerous substances lies, as
Bloch suggests, in this strong symbolic value.

Meat has a symbolic value in Markakasa135. It is food which is regarded as the most
social and the clearest indication of sociality. Whenever guests from other places visit
meat-eating households, at least one of the meals they are served during their stay
contains meat. Even among Markakasa people, friends invite one another to eat on those
days when their household has meat to cook. Sharing a vegetarian meal has social value
too of course, but the value of meat seemed to rank higher. Meat-eating guests who came
to Markakasa and stayed with vegetarian households were said not to enjoy themselves as
much, and would stay fewer days on account of the lack of meat. They would spend their
evenings at other people’s houses, newly struck-up acquaintances, where meat would be
served, or would disappear to a neighbouring village where a relative lived. Following
the logic of Bloch’s argument then, one could argue that, since meat in Markakasa has a
strong association with the expression of sociality, it also carries with it the risk of
danger.

Bloch also proposes that it is commensality with outsiders which is regarded as

dangerous among the Zafimaniiy: the closeness created by eating together forces people
who are otherwise strangers (or even enemies) into one another’s sociality. In Markakasa,

135 Bloch (1999) also notes that meat in many societies performs an ultra-social function.

195
similar sentiments were also expressed, as the example of paan above demonstrates.
Weddings, too, were seen as prime sites for poisoning and witchcraft, since one would sit
in a dining line (pangat) with people one might not necessarily know or trust. But as we
saw in Chapter Three, and contrary to Bloch’s ethnography, one also has reason to fear
and mistrust those who are closest - brothers and other agnatic kin - since these are the
types of people who are most involved in one’s life and with whom one is most likely to
enter into dispute.

The power of vegetarian practice: protection not prestige

Whilst the actions of the body in the process of suffering and healing are striking for
those who go to Sukdi (see Chapter Five), no less striking are the bodily disciplines
imposed on them while they are in residence at the temple. Former meat-eaters and
alcohol-drinkers have to change their diets and have to learn how to pray at the temple
and participate in the worship there. Membership of the sect and observance of its values
and practices protects people from the attention of witches and other malicious spiritual
agents; thus the practices adopted while in residence at the temple are brought back to the
village where they are believed by sect members to be part of the package that protects
them from harm. In short, the practice of vegetarianism is seen as essential to living a life
free from pain and hardship. Here then, we find ourselves far from the position where
vegetarianism is seen as a marker of social prestige; rather it is a practice that is a
fundamental component of the life lived safely, and one which has power because it is
practiced by people who have experienced the worst of times. The intense bodily
experiences involved in both suffering and healing create dominant motifs in the
biographies of people who have become vegetarian; the power of the latter is found in the
remembrance of that set of experiences, and the desire never to suffer in the same way
again (see also Connerton 1989; Csordas 1990, 1994, 1999). As discussed in the previous
chapter, the sign of their devotion is the vishesh, the stone said to be from an ota (dais),
which each adherent possesses and to which they offer worship. The vishesh is also a
mark of protection; its disappearance suggests that the devotee has somehow strayed
from the path of the sect and protection from witchcraft and sorcery is withdrawn. To eat

196
meat is to err and one risks returning to that time of pain and suffering before one joined
the sect.

Since they are required to give up eating meat and alcohol on joining the sect, new
Mahanubhav adherents come to see these two forbidden substances differently to before.
Not only is abstention from meat and alcohol seen as contributing to a virtuous life
characterised by the principle of non-violence (ahimsa), an idea that would have had little
currency for them as meat-eaters before joining the sect, but these foods are seen as
dangerous. My landlord, Brijlal commented that since joining the sect even to smell
alcohol makes his head spin, a sensation he never used to feel before. This comment was
made after he had told me that if Mahanubhav panthis begin to eat meat and drink
alcohol again then they will eventually go mad. Once again then, what is emphasised is
the bodily sensation of contact with the forbidden substance and the effect that has on the
status of one’s vulnerability to attacks of magic and witchcraft. Not only is the body
instrumental in one’s experience of forbidden substance through the remembrance of the
pain and process of healing, but it engages the entirety of one’s senses and reminds us of
the nature of the sensual power through which food is experienced (see Jackson 1989;
Stoller 1989, 1997).

As mentioned at the beginning of this chapter, there is also a complementary notion of

purity: membership of the sect and observance of its practices has effected a change from
the impure and sinful to the pure and virtuous. One sect member from another village
offered this typical explanation: “My life was full of pain (dukh) and sin (paap). I wanted
to change my life and become a better person. Before I joined the sect I ate meat and
drank alcohol and told lies. Now I am a better person and I’m happy. My soul is pure
(shuddh); I’m pure on the inside.” Does this desire to remain pure motivate the practice
of vegetarianism? And in expressing this desire, do not vegetarian Mahanubhav devotees
claim some sort of prestige that has a Brahminical referent? Does this undermine the
proposition that it is the power of experience which motivates a vegetarian diet? The
connection, after all, appears clear: purity is associated with a vegetarian diet throughout

197
much of India and, as scholars have observed, this combination of purity and
1
vegetarianism reaches its apogee in the figure of the Brahmin.

But where does vegetarianism as a value come from for Mahanubhav sect members?
Does it have a Brahminical referent? We can draw on Dumont’s (1970) famous
discussion of the relationship between the Renouncer and Brahminical thought in Indian
religions to illuminate this question. Dumont suggested that innovations in Indian society
and Hindu thought were due to the individual-outside-the-world or the World Renouncer.
The Renouncer formulated new ideas, which were then canonised in the sect that he
created. As the sect and the renouncer posed a challenge to Brahminical power and status,
since they were often non-Brahminical in character, Brahminical thought then
incorporated values and ideas propagated by the sect as Brahminical values which then
became markers of prestige in society.

Let us take the example of vegetarianism. The Renouncer declares that to be virtuous and
lead a good life one has to be vegetarian. This practice is then advocated by the sect he
(or his disciples) establishes. Brahminical thought takes up vegetarianism and makes it a
marker of prestige in society. Dumont sees this is a historical process, as a moment or
series of moments that, were we so inclined, we could verify historically. Thus in joining
the sect now, in contemporary time, what devotees are actually doing by practicing
vegetarianism (regardless of what the sect says, regardless if the sect is explicitly anti-
Brahminical, regardless of their motives in joining the sect) is reinforcing Brahminical
values, since sometime in the past the idea of the renouncer has been incorporated into
Brahminical ideology.

However, we could see Dumont’s theory of the role of world renunciation as dynamic
and applicable in the present-day, and not simply as a historical process completed
sometime in the past. Surely, if Dumont’s proposition is correct, Brahminical ideology is

136 Parry (1991: 269) notes the considerable difficulties in attempting to study values such as pollution,
purity, sin and misfortune in Hindu India: one should not assume that different words necessarily indicate
different concepts. In this case however, the motivations o f the people in seeking a remedy to illness do
make it possible to distinguish between sin/virtue and purity/impurity.

198
not fixed for all time but needs to be maintained by constantly adopting values and ideas
from present-day World Renouncers. If present-day renouncers were to stop saying that
vegetarianism is a virtuous practice, how would Brahminical ideology react? One could
conclude that the act of joining a sect such as the Mahanubhav in the present day is to
reinforce the values of the Renouncer and not of the Brahmin. Brahminical values cannot
be completely hegemonic in the present: the relationship between the Renouncer and the
Brahmin is dynamic, and the ideology of the latter has to be constantly maintained.

One could take it so far as to say that by joining the sect of the World Renouncer,
Mahanubhav devotees become (perhaps only for an instant) symbolic World-Renouncers
themselves and give vegetarianism a value which is then transmitted to Brahminical
thought anew. The source of the power that the Mahanubhav adherent gives to the value
of a practice such as vegetarianism is the Renouncer, not the Brahmin. To argue
otherwise would be to deny the power of the World-Renouncer in modem India; his
power cannot be reduced to that of Brahminical ideology.

But what about purity? Here we can take up Hardiman’s argument outlined at the start of
the chapter, that religious reform movements such as the Devi were an attempt to
democratise Brahminical notions of purity: by making everyone pure (not just Brahmins),
purity lost its social power. This seems to bear some affinity to what we know about the
history of the Mahanubhav sect. Like many other religious sects which began in medieval
India, the Mahanubhav positioned itself against Brahmin ritual hegemony (Raeside 1976;
Skultans 1987). Unusually, the Mahanubhav was the subject of intense Brahminical
persecution up until the late nineteenth century when it was recognised as a Hindu sect
and, perhaps as a result, the sect writings are vehemently anti-Brahmin (Raeside 1976).
Similarly, the nuns and monks of the sect whom I met both on their visits to Markakasa
and on my visit to Sukdi would often define one of the chief characteristics of their sect
as being against both Brahminical authority and privilege and the figure of the Brahmin
more generally. At this point in the discussion, Hardiman’s contention of an attempt to
deprive purity of its specific Brahminical power, by making it a value available to all,
appears entirely applicable.

199
However, one can distinguish purity as a value for the sect and as a value for its
adherents: the two are not necessarily the same. The practice of vegetarianism and
teetotalism from the view of the sect may indeed be an expression of purity in relation to
Brahminical power. But purity as a principle from the sect’s point of view is not the same
thing as purity from the lay devotee’s perspective. There are two possible explanations
why the desire to be pure might be attractive to people in Markakasa. There is, firstly, the
explanation which derives from the well-established argument running throughout the
anthropological literature, that purity as a value in India is important in terms of social
prestige - whether in terms of caste hierarchy or status more generally: in being members
of a Hindu (devotional) sect and thus being vegetarian, people attempt to emulate the
purity of Brahmins, a caste and status position denied them by the fact of their birth. A
second possible explanation, advanced here, is that people maintain purity in order to
protect themselves better from attacks of witchcraft. The motivation explained in terms of
prestige may indeed be valid but it cannot exist without the corresponding motivation to
have continuous protection; if the latter is untrue then both the decision to go to Sukdi to
be cured and the consequent bodily experiences lose meaning. This latter explanation can
however exist without the former, while the former is incomprehensible without the
latter. Thus the values of purity and holiness, which the sect doubtlessly promotes and is
associated with, are encompassed by the requirements of protection.

That purity is a value defined by and encompassed by the needs of protection against
malicious spiritual attack is demonstrated by an episode involving my landlord Brijlal. As
well as being a Mahanubhav sect adherent, Brijlal is active in the local Bharatiya Janata
Party (BJP), one of two major national political parties of India. Though he does not hold
any official post within the local party organisation, he is nevertheless regarded as the
BJP’s man in Markakasa. In this part of India, as in many other areas of the country, it is
customary for the major political parties (most notably the Congress and the BJP) to
distribute alcohol to voters during elections. The purpose in Markakasa at least was not to
‘buy votes’ as it is often represented in the media; the alcohol is distributed after the polls
have closed. But a link is established and generally, though not necessarily, BJP

200
supporters drank the BJP alcohol and Congress supporters drank Congress-supplied
alcohol. As the main BJP worker (karyakarta) in the village, Brijlal took delivery of the
alcohol when the party jeep visited Markakasa. He refused to receive it into his house,
however, and instead used a room in the adjoining house of his brother-in-law (ZH),
whose household are not members of the sect, to store the liquor. After the polls closed
and the election officials had left, Brijlal organised the consumption of this alcohol in
front of his house. Though people wanted to drink in his courtyard, regarded as ‘inside
the house’, Brijlal made it clear to those assembled that no alcohol was to be consumed
there; in addition none of his cups were used and his sister supplied some, rather
unwillingly it has to be said, from her own kitchen. Nevertheless, it was definitely Brijlal
as representative of the BJP who served the alcohol and organised its distribution. Clearly
he was concerned with purity but I doubt whether it was a matter of prestige. Prestige,
after all, is a value in relation to others in society and yet Brijlal was not concerned about
appearing to lose prestige by publicly serving alcohol in front of his house. It was his
purity and that of his household that he was anxious about, and feared that by receiving
the alcohol, a forbidden substance, into his home he would be breaking a key
Mahanubhav tenet. Would the vishesh have disappeared if he had stored and served it
within the walls of his home? It would be reasonable to conclude that he had to maintain
purity in order to maintain protection.

But is not purity intimately connected with protection from misfortune in other parts of
India too, even Brahminical India? Parry (1991) shows that Benarasi Brahmins regularly
attribute illness to an impure diet; pollution, he claims, is a potent source of ill-being in
many parts of India (ibid: 271). The position looks similar to that of the Markakasa
Mahanubhav adherents but there is a qualitative difference. Before they join the sect,
Markakasa people do not link impurity of diet or sin to the cause of their affliction. The
illness is caused by witchcraft or magic in a situation completely outside their control,
and regardless of their own behaviour. It is only in ‘post-conversion’ narratives that the
question of sin and impurity becomes relevant. But sect members, if pushed as to the
causes of their illness, quickly abandon the stance of impurity as cause and explain it in
terms of mystical magical attack by other human beings. Pollution does not cause illness

201
in the same way that it does for Benarasi Brahmins: for the latter, a person in an impure
state seems to attract misfortune without the intervention of another actor - impurity and
illness are inextricably linked; for Markakasa Mahanubhav people, on the other hand, the
impurity attached to meat or alcohol which they come into contact with is seen as
breaking their promise to God to be good devotees, and it is this broken promise which
withdraws protection.

And there is a difference in the importance placed on prestige too. For the Benarasi
Brahmins that Parry describes, prestige is just as important as protection. Indeed, funeral
priests who expose themselves to more impurity than others are not only regarded as
more susceptible to illness or death but are also lower in status than other Brahmins in the
city (Parry 1994). This does not apply to Markakasa Mahanubhav devotees, for whom the
prestige associated with a vegetarian diet is entirely irrelevant in their search for a lasting
cure to their illness.

The practice of a vegetarian diet in Markakasa is the consequence of powerful needs

rooted in painful bodily experience. The concerns with purity, which might have
suggested an ultimate link with high-caste ways of being, are shown to be subordinate to
the superior demands of protection from malicious spiritual attack. And though
vegetarianism is associated with Brahminical ideology, a dynamic reading of Dumont on
world renunciation demonstrates that sect members reference the values of the
Renouncer, not of the Brahmin.

What then are the consequences of this change in diet for relations between vegetarians
and meat-eaters? Indeed, is this latter distinction between the consumers of different
types of food one that Markakasa people see as ultimately meaningful in structuring
relations among themselves? Bringing notions of embodiment back into the analysis, one
may be able to observe how “the body is not only an outcome of social processes but [is]
a transformer of social processes itself’ (Lyon and Barbalet 1994: 49).

202
Vegetarians and meat-eaters: Gonds, sociality and commensality

Gond-ness was often unsurprisingly expressed in the idiom of difference: their language,
their life-cycle rituals, their gods and their enjoyment of meat and alcohol. So, roasted
fish curry was an essential part of a naming ceremony (satti), as were copious amounts of
daru (alcohol) for both women and men. At the other end of life, during the Gond death
ceremonies known as karsad often performed several years after the actual death, when
tombs for the dead would be built and the spirits of the ancestors given a room in the
house (kuratandi), huge expense is incurred with the slaughter of animals both for
sacrificial purposes and for the needs of the guests. Indeed, this is the main reason not
only that the karsad takes place several years after a death, but that the karsad is
performed for several deceased agnatic kin at one time. Daughters of the family who
have since married and become members of other households return at this time, and
make offerings of juvenile chickens to their dead ancestors which are then sacrificed by
the officiating baigas (ritual specialists) and cooked. Likewise, alcohol flows freely: an
important role is played by the deceased’s affines (sorye) in the preparation of the tombs,
and in the transportation of a pot said to hold the deceased’s life force or soul (/7v), and
every now and again at various points during the ritual the sorye all sit down and refuse
to take any further part until they are given alcohol to drink. Eating meat is so bound up
with one’s identity as a Gond that it was said that even Gonds who were vegetarian in
their lifetimes would ask for meat on their death-beds. Talking one day with a Gond man,
Dayal, he narrated an apocryphal story of Brijlal’s father, a vegetarian, who, as he was
dying, called Dayal over and asked him to get hold of a small chicken for him, cook it
and feed it to him. “Why,” I asked Dayal, “would a vegetarian do this?” “He was a
Gond,” he replied, “and Gonds need meat.”

The Gondness of eating meat and drinking alcohol appears to be so intuitive that even
children are aware of it. I take my meals with my vegetarian landlord’s family. The
adjoining household, separated only by a courtyard, is that of my landlord’s sister and
brother-in-law who are not members of the sect. While my landlord’s two year old son,
Anil, and I were eating lunch one afternoon, his uncle (the aforementioned brother-in-

203
law), pottering around in the courtyard outside, asked us what we were eating. We
answered and then Anil, prompted by his mother who was serving us, in turn asked his
uncle what he had eaten that day. “Fish curry”, came back the teasing reply. “You lot are
Gonds” said Anil. His mother and I both laughed. “What are we then?” she asked her son
rhetorically.

And despite this, the association of Gondness with meat and alcohol made by some
people cannot really be sustained; in conceptual terms a vegetarian Gond would cease to
exist as a Gond, something which clearly does not take place since vegetarian Gonds are
regarded as full members of their caste group. What is at the root of these anxieties over
Gond-ness, expressed by meat-eating Gonds, is not dietary difference but conflicting
notions of sociality resulting from membership of the sect. Consider karsad, outlined
above. Mahanubhav sect adherents would not require karsad for themselves but are quite
willing to organise one for their non-sectarian parents. One man told me that the structure
of the ritual stayed precisely the same: the affmes (sorye) would periodically refuse to
take part in the procession and instead of alcohol were given lemon sherbet to drink to
persuade them to continue. Likewise the sacrifice of chickens and goats was replaced by
lemons and coconuts and vegetarian food was served at the various feasts. Yet the karsad
as a ritual was valid and successful: tombs were built and the ancestors (hanal) housed.
That the karsad, a distinctive local marker of Gondness, could still be performed suggests
that a vegetarian Gond is not really a contradiction in terms at all.

As we have seen, the practice of a vegetarian diet and abstention from alcohol in
Markakasa are part of a package of measures which Mahanubhav devotees embrace in
order to better protect themselves from spiritual attack now and in the future. This
transformation distinguishes them from their kin and neighbours who are not members of
the sect and do eat meat and drink alcohol. Vegetarianism needs to be seen in the context
of a range of other changes in local notions of sociality. To neglect to do so would be to
hold up the practice of vegetarianism in and of itself contra meat-eating as the key
opposing tropes in society, and the contrasting dietary practices themselves as the
distinguishing point of analytical departure. Approaches to embodiment which base

204
themselves solely in experience and emotion (e.g. Obeyesekere 1981) are susceptible to
the criticism that society and sociality (or how one relates to others) drop out of the
analysis. I want to show instead how individual experience and practice (such as that of
vegetarianism) can be socially as well as individually transformative.

There is a marked change in the practice and attitude to commensal dining. As discussed
above, there is a general fear of magical attack while eating and one is most vulnerable at
someone else’s house, eating someone else’s food. Sect members often expressed their
reluctance to eat at the homes of others who were not Mahanubhav panthis. A common
pattern among new devotees is the culinary division of the household: sect members
break away from brothers or parents and cook separately (though they may continue to
cooperate in agricultural work). This exclusivity with regard to food and commensal
dining has its ultimate expression in the ‘sealing off* which takes place during Navratri
(see Chapter Five).

Consider the following example. My landlady Champa, a Mahanubhav Gond, returned

home from a baby’s naming ceremony (satti) which she had attended in a neighbouring
village along with several other Markakasa Gond women, and was not at all happy. “I tell
you, Desai, it was awful”, she complained, “our Gonds are really filthy; I couldn’t eat a
thing.” “Didn’t they have anything vegetarian?” I asked. “Yes, they’d made potato and
aubergine curry”, she replied, “but all I could taste was fish, they were probably using the
same spoons and pans.” Champa felt her purity had been compromised: even though
vegetarian food was served, doubts entered her mind regarding contamination. She had
felt very uncomfortable and left early. The following day I went to visit another Gond
woman, Ganeshyam’s mother, Saitrabai, who had also attended the naming ceremony.
She was not at all pleased at Champa’s behaviour the day before. What irked her was
Champa’s desire to be separate from the rest of the guests and her refusal to eat the
vegetarian food she was served. “This is our sanskruti (culture)”, she said, “and it’s
because of people like her [members of a sect] that we’re losing it.” Saitrabai’s notion of
sociality consisted of people eating together, happily without too much fuss. In her view,
Champa had unproblematically been provided vegetarian food and so ought to have been

205
satisfied. Instead Champa’s reluctance to partake in commensality was proof to Saitrabai
that membership of the sect had fundamentally changed her and that it was this, and not
the practice of vegetarianism, that was of concern in terms o f ‘culture’.

What is troubling then for non-sect members is not the difference in dietary habits but
that, in several spheres of social life, sociality and commensality seem to have been
transformed. The real dilemma of Gond-ness for Gonds is not vegetarianism, which is
capable of accommodation, but that Mahanubhav sect members no longer worship
household gods or ancestors, nor do they share a common view of sociality.

That vegetarians refuse or are reluctant to eat with others is not in emulation of
Brahminical practice but is part of a wider transformation in sociality as a result of
membership of the sect. Their exclusivity, though it may be expressed in the language of
purity and pollution, does not have a Brahminical referent; instead, in their interactions
with others,, sect members are keen to avoid coming into contact with forbidden
substances because of the risk they pose to their level of protection against witchery and
magic. Similarly, as a result of the remembrance of the painful embodied experience
which led to them joining the sect in the first place, sect members have a heightened
sense of danger and are cautious about commensality in general, fearful of a relapse into
suffering. An interrogation of the causes for the adoption of vegetarianism and its
continuing power, as discussed in the previous section, leads to a better interpretation of
people’s sociality and the place of dietary practice in that.

The dietary practices that Mahanubhav (and other sect) members observe emerge from
the need for protection from future harm and represent an aspect of their performance of
devotion to God. As noted at the beginning of the chapter, vegetarianism and teetotalism
have been historically constructed as representing a true Indianness or Hindu-ness, and
have been associated with civilisation in opposition to the barbarism of meat-eating. Thus
these practices need to be seen as part of a wider set of concerns, which attempt not only
to define who good Hindus are but also to distinguish those who are ‘wild’ (jangli) from
those who are ‘improved’ (sudharlele). The commitment to vegetarianism produced by

206
membership of the sect and rooted in the experience of illness also provides a link to this
wider set of oppositions that come to be represented by different types of diet. The
following example demonstrates how the vegetarianism of the sect members is
transformed into a commentary on ‘superstitious’ and ‘backward’ practices conducted by
non-sect villagers.

The case of the missing horse: animal sacrifice and ‘superstition’

On joining the sect, new Mahanubhav adherents return to their villages from Sukdi as
changed people, and it is in their practices that they appear changed to those on the
outside. In order to affirm their status as atal bhakt (unswerving devotees) to God
(Bhagwan) alone, they take down pictures of Hindu gods and goddesses which
previously adorned their homes. For Gonds who join the sect, the household gods
(lonchtun) and ancestors (hanal) must also be removed from their land; members of the
Mahar caste who, in this area at least, are all Buddhists, remove pictures of the Buddha
and cease to worship him; and sect members (of any caste) no longer worship village
deities. Sociality with divinity is radically transformed. An event that took place during
my stay in the village crystallised the divisions between sect members and non-sect
members. It centred on the type of food one sacrificed to the village gods.

On the outskirts of the settlement, as the houses give way to the fields and forest, are the
shrines of the Thakurdev, the principal tutelary deity of the village, and the gods and
goddesses of the surrounding forests and hills that have been called to Markakasa and
established there (sthapan) to protect it. These latter are represented in the form of clay
horses and are called dev. Thakurdev dwells in the saja tree and is associated

207
Plate 30: A tutelary deity o f the village, around which collective village worship
centres.

Plate 31: Deities.

208
Plate 32: Offering worship to the tutelary deities o f Markakasa during the
agricultural ritual of Sanjori.

Plate 33: The Thakurdev, the

principal tutelary deity o f the
village. It lives in the saja tree.

209
with the headman’s family; the current headman’s ancestor founded the village and the
shrine. Also in this grove is Bhimalpen: he too lives in a tree and is the guardian of the
village boundaries, being the deity that keeps thieves and ghosts away. It is here among
this set of shrines that the village’s communal religious life is centred.

Fifty years ago, animal sacrifice for these gods was common. One of the ways in which
the village as a collective could respond to crop failure or illness determined to have been
caused by angry gods was the sacrifice of a buffalo to the god Bhaisasur. Bhaisasur is
located in an area slightly beyond the grove which houses Thakurdev and the others.
There, the villagers would dig a pit and lower a live young buffalo into it. The pit would
then be covered with a large slab of stone and the buffalo would be left to die. The
Bhaisasur is still powerful today. The land where it resides belongs to Jagdish, a Mahar
and one of the wealthiest men in the village. His wife told me that they do not farm that
part of their land at all. Once a servant from another village had mistakenly ploughed the
land there and had dug up a chain which he attached to one of the bulls. That night,
Jagdish’s wife saw a buffalo running through her dream and knew that something was
wrong. When she woke in the morning, the bull with the chain around its neck had died.

But the god is no longer propitiated. Following a series of disputes with the incoming
Telis about fifty years ago, some of whom were then beginning their involvement with
the Kabir sect (see Chapter Two), blood sacrifice (including buffalo sacrifice) was
abandoned.

There is a close association between the. food that is eaten by men and the food offered to
the gods during worship (puja). Babb (1975), for instance, notes that the principles of
purity and hierarchy that are involved in food transactions, especially in relation to the
notion of polluted leavings (jutha), are present in the offering of food to the deity. Just as
to eat someone’s jutha is to recognise one’s inferiority to them, he argues, the receiving
back of prasad which is shared among the congregants expresses man’s inferiority to the
deity (ibid: 53-60). And while blood sacrifice (bali) is different in form to ordinary
worship (puja), scholars have recognised that they are related categories and that the

210
eating of the meat after the sacrifice is akin to consuming prasad (Fuller 1992: 85-6). Just
as the food that people were eating in Markakasa was changing, so the food eaten by their
gods had to change too. But were the gods willing to accept that change?

A couple of months after my arrival in Markakasa, one of the horses went missing. Now,
this unsettled many villagers; that a god could be stolen and the perpetrator not suffer any
immediate punishment seemed to them a sign that something was terribly wrong, that
perhaps the village gods were angry. A meeting was held and it was decided that a white
goat would have to be given in sacrifice to the village gods; the villagers would ask that
the gods bring back the stolen horse to the shrine and punish the thief by making him ill
or killing him. Several sect members expressed their dissatisfaction with this return to
animal sacrifice, but did not pursue the matter publicly. Goats are rather expensive so a
collection was taken up to raise the seven hundred rupees needed: some sect members
avoided paying; others told me they had reluctantly handed over the money. Many
adherents of the sect regarded the slaughter of a goat in this situation as an example of
superstition or ‘blind faith’ (andha shraddha) and would delight in telling me that the
sacrifice had been in vain: though the horse was later found, no-one appeared to have
been punished. That people should eat meat or drink alcohol was seen by sect members
as undesirable in human beings, an indication that the individual or the community was
not ‘developed’ or ‘improved’; that people should offer blood sacrifice in order to feed
the gods was superstitious nonsense, another indicator of people’s lack of ‘development’
or ‘improvement’. What is important here is that blood sacrifice was not decried using an
idiom of purity and pollution; instead, the vegetarian sect members I spoke to felt that
sacrifice such as this was decidedly unmodem: “Educated people from the cities don’t do
this sort of thing; they pray to God and have done with it”, a Mahanubhav Teli man said.

This attitude did not sit well with those, especially non-sect Gonds, who had trusted in
the efficacy of the ritual. That the horse could go missing in the first place was a sign,
according to many people, that the vegetarians had had their way too long in denying the
gods animal sacrifice all these years. That the sacrifice was only partially successful was
another warning sign. Too many concessions had been made to both the Telis, who had

211
initially introduced sect Hinduism to the village with their advocacy of the Kabir Panth,
and more recently to the Mahanubhav sect and others, with their unwillingness to take
part in any village ritual. While the rituals used to involve animal sacrifice and daru, now
all that was offered was a coconut. The gods were not being satisfied and further
problems could be expected. As one slightly drunk, non-sect Gond man put it to me:

“If you invite someone to eat with you and put a pot of water in front of
them to wash their hands and then tell them there is no dinner how will
they feel? That’s how our gods Qiamara devta) feel when we give them
coconuts; they’re waiting for meat and alcohol.”

Through the accusations of ‘superstition’ and ‘backwardness’ the sect members levelled
at the rest of the village for pursuing the goat sacrifice, they were expressing their
identification with a modem way of being, akin to that lived by people in the cities who
were thought to have more effective ways of dealing with problems such as this. But
alongside the calls of superstition is the fact that devotees, through membership of the
sect, have in any case transformed an aspect of their sociality with these local village
gods. They can call these practices superstitious and ineffective largely because they no
longer need to rely on them for protection from harm.

Food, drink and the ‘backward Adivasi’: the neighbouring village of Kedanar

From my earliest days in the village, people would say to me that it was fortunate that I
had come to live in Markakasa, and not in the neighbouring village of Kedanar. It was
dirty, people said, and their inhabitants were not improved (sudarlele). The people there
were always drunk and did not know how to welcome outsiders. At first I dismissed this
as petty village chauvinism, but when it was repeated to me throughout my fieldwork and
by ‘outsiders’ (forest guards, teachers) too, I began to look seriously at why they would
think that Kedanar, and indeed several other villages in the area, were more backward
(magaslele) than Markakasa. What were the criteria they were using?

212
Kedanar is a predominantly Gond village one km away along the road. The settlement is
larger than Markakasa’s and its forest is denser and more extensive, covering an area of
over twenty square kilometres. Unlike in Markakasa, the Gonds here are by far the
majority community and there is also a large population of Gond Rawat (a cattle herding
service caste, that speaks a different dialect of Gondi). There are very few Chhattisgarhi
castes in Kedanar, though there are several Maharashtrian Mahar households. As a result
Chhattisgarhi is hardly heard and the village lingua franca is, unlike in Markakasa,
Gondi.

People would often say that Kedanar was less ‘improved’ or developed than Markakasa,
because the people there drank too much alcohol (daru). Markakasa is one of the few
villages in the area to have banned the brewing and sale of alcohol, though not the
consumption of it (see more below).137 So, if people want a drink, they often go to
Kedanar where homebrew is never in short supply; I was told repeatedly that Kedanar is
a place where “you can obtain daru in any house” (sab ghar me daru milenge).
Interestingly, the phrase “going to Kedanar”, was used euphemistically in certain
contexts to indicate that the speaker was going for alcohol. Indeed, expecting that these
sorts of statements were just rhetorical, I was struck by their relative truthfulness. On my
first visit to Kedanar, I was offered a drink by three different households, something that
rarely happened to me in Markakasa except on festival days or during a wedding. In
Markakasa and Kedanar alike, drinking daru was not necessarily seen as a bad in itself.
In fact, the local brew, made from the mahua flower, was seen by many, including
teetotallers, as an effective medicine against diverse ailments; “daru me abud bitamin
have” (lit. there are many vitamins in daru). And ‘alcohol’ was not an undifferentiated
category.

As I sat round the fire after dinner one evening, I casually mentioned to my host family
that my uncle (mama) liked a glass or two of whiskey in the evenings, immediately
apprehensive after the words left my mouth that my host family would think badly of a
Brahmin who drank. Instead they seemed rather confused as to what whiskey was. My

137 The ban is not a gram panchayat resolution, but one made by the sian partch.

213
landlord asked me if it was ‘inglish’. ‘Inglish’ [from ‘English’], I learnt, referred to
alcohol that was seen as more modem in that it was drunk by paisewalleh log (people
with money) from the city, such as my uncle. ‘Inglish’ was a different type of alcohol,
and indeed the word ‘daru ’ would not be used to refer to it. Inglish such as beer (Indian
brand lager) and whiskey (or more commonly, ‘whiksey’) was something that was drunk
with ‘taist’ [taste], my informants said; it was a drink to be savoured, instead of knocked
back, which is why city people (like me) always sipped mahua daru as if it were tea.
Though sect members would not touch any alcohol, beer and whiskey, they did not seem
to have the same opprobrium attached to them. Though surprising given his caste, the fact
that my uncle drank whiskey was not necessarily contentious, perhaps because it had no
local salience, perhaps because it was seen as ‘modem’ and therefore acceptable. The fact
that my uncle ate meat was commented on however, and that my landlord did find
curious. Daru, then, in most situations, referred to a limited category of alcohol which
included the local brew made from the mahua flower, mahua daru, and also the more
expensive government distilled alcohol known as desi daru. There was some confusion
over the status of newer drinks such as Coca-Cola138: on several occasions I was asked
whether it was daru129. That the villagers, especially young men, associated me with
modernity, and that this modernity was linked to beer, is illustrated by a curious incident.
A group of young men of all castes, who were on their way to a wedding in another
village, tumbled into my room one afternoon as I was enjoying a nap and produced a
750ml bottle of beer. They had clubbed together to buy this bottle (about eighty rupees,
or two days labour) and needed a place to drink it. They all knew how opposed my
landlord was to alcohol (see above), so it seemed a strange choice deliberately to come to
his house: if he found out he would not be pleased, and in any case there were several
houses nearby which were eminently more suitable. They wanted to drink the beer with
me and show me their sophistication. If it had been mahua daru I doubt they would have
gone out of their way to come to such a risky drinking den.

138 A six rupee (about eight pence) 150ml bottle o f Coke was still regarded by many in the village as a
strange luxury that only teachers and other government servants indulged in.
139 Daru, used here to mean alcohol in the widest sense.

214
The problem with alcohol, I was told, was that some people, especially in Kedanar, drank
to excess; they did not drink with ‘taisf as city people or foreigners did. One only had to
look at Kedanar to see the effects of their drunkenness: “They have so much land in
Kedanar, most people there are bigger landowners than in Markakasa, but they’re just not
interested in working hard; they’re always drunk. That’s why so many people from
Markakasa rent farmland from Kedanar,” said Ramu, a non-sect Gond who himself had a
bit of a reputation as a keen drinker.

People in Markakasa seem to adopt an evolutionist frame of thinking when talking about
Kedanar. Though only one kilometre away, it was, in their minds, further back down the
road of improvement or change, a snapshot perhaps of what Markakasa used to be like.
“What used to happen in Markakasa still goes on in Kedanar”, was a common sentiment.
My Mahaprasad (ritual friend) Ganeshyam visited Kedanar frequently, usually once
every two or three days. His family has a home there, since his late father’s second junior
wife still lives there (Ganeshyam’s mother is the first junior wife, and maintains her
household in Markakasa), and they can count a number of kin in Kedanar: the sarpanch
(elected head of the village council) there, for instance, is his FZD; and Kedanar is the
natal village of his senior mother (bade dai). His mother is the Markakasa sarpanch who
criticised my landlady in the incident referred to earlier in the chapter. Ganeshyam
himself does not drink daru and, though he eats eggs, does not see himself as a meat-
eater. And naturally, since his is a meat-eating household, he is not usually fussy about
where he eats. I was asking him about labour employment in his small shop one evening
as we were cycling back from a visit to Kedanar, and he said he preferred to get people in
from Kedanar to work for him. They were simpal log [from English ‘simple’], he said,
who weren’t interested in working for money: all they needed in payment was a few
glasses of daru and a chicken or goat meal. Labourers from Markakasa on the other hand,
he said, were money-minded and would bargain ferociously over their daily wages. There
was of course no way of confirming this; what is interesting about these statements is not
their veracity but the fact that they are made at all and expressed in terms such as these,
with food and drink playing a part in the construction of the image of the ‘simpal
Adivasi’. In his eyes, the Gonds from Kedanar were to be admired for their simplicity.

215
But Ganeshyam could also be disgusted at times with people from Kedanar, and their
lack of ‘development’ or ‘improvement’ was to be deplored. On another occasion,
accompanied by him, I dropped in on a meeting of the Kedanar sian panch, who were
discussing the arrangements that needed to be made for that year’s village fair (mandai).
Several of the men were noticeably drunk and daru was being served. As guests, we too
were offered a drink: I accepted; Ganeshyam, who is teetotal, refused and was served tea
instead. I drank up, and chatted a while and then left with Ganeshyam, who looked rather
cross. “I feel sick; they served my tea in a glass which had had alcohol in it and now I
can’t get the taste out of my mouth”, he complained as he chewed rather ferociously on a
couple of 50 paise sweets. “Really, these Gonds just won’t improve. A thing like that
would never happen in Markakasa these days; people know better now.”

Ganeshyam’s concern over the daru was expressed in terms different to those used by
Mahanubhav sect members. For the latter, having a drink served in an unclean glass
would not only have been unpleasant, but also would have compromised their purity and
thus their protection. For them, the sensation of drinking water from an alcohol-stained
glass would have recalled both their experience of painful illness and their promise to
God not to come into contact with such forbidden substances. Like Ganeshyam, however,
they would have discussed such an act as evidence of a lack of improvement.

The lack of other castes in Kedanar, and the absence of members of devotional sects,
were often given by Markakasa people as a reason for the former’s ‘backwardness’. The
Gonds there dominated the village and had been left to themselves. Thus they could do as
they pleased and did not have any time for the haatiyas (a derogatory term used for
Chhattisgarhi people). Their excessive alcohol consumption was also used to explain the
physical state of the village; Kedanar was seen as a dirty place where the lanes got
muddy, the houses were poorly maintained and pigs ran amok with their litters. It was felt
that the Gonds in Markakasa had improved precisely because of the presence of other
castes and sects. But the lack of other castes and sects in Kedanar meant that the people
there, and the village as a whole, were seen as more united than Markakasa. Decisions
could be taken quickly about matters such as organising the village fair or, more

216
importantly, dealing with attacks of witchcraft, magic or the displeasure of the gods. In
local discourse, the mixture of castes, sects and people in Markakasa that made
‘improvement’ possible also made resolving disputes, and acting effectively as a
community, more difficult.

The development of these categories of ‘improvement’ (sudharna), linked as they are to

the practice of vegetarianism and teetotalism, find resonance with the activities of a group
of local Hindu nationalist activists who belong to the Vidarbha Vanvasi Kalyan Ashram
(VVKA) an organisation engaged in projects of ‘uplift’ and ‘cultural protection’. In the
sections which follow, I want to demonstrate how the reasons for being vegetarian
discussed above (protection from harm) become tangled up with ideas promoted locally
about what it means to live in a perfect society, a society which has a distinctly Hindu
character.

Hindu nationalism and a coincidence of values

Brijlal and others like him have experienced the problems of dealing with witchcraft and
magic and were healed at the Mahanubhav temple. As a consequence, sect members in
Markakasa have come to regard certain practices as important to them in order to
maintain their protection against further mystical attack. But as I describe above, their
practices, especially their dietary practices, become part of a way of describing social
change, ‘improvement’ (sudharna), which also enables Markakasa people to distinguish
themselves from other people, typically other Adivasis, who are regarded as jangli (wild).
As I mentioned in the vignette which began the thesis, Brijlal is also involved in several
Hindu nationalist organisations which operate in the area. Here I want to discuss some of
the activities of the latter, and also look at Brijlal’s involvement in a little more detail.

Brijlal’s father was a member of the BJP at a time when the Gond community was seen
as the preserve of the Congress Party (as it largely still is). It is fair to say that Brijlal, like
few others in Markakasa, grew up in an atmosphere where party politics was talked
about: he developed an interest early on and when his father passed away in 1993, he

217
took up the old man’s responsibilities. His work for the Party is rather limited and he only
really gets busy around election time. Though I questioned him often about this, his
commitment to the BJP does not seem to be founded on a political ideological
commitment to making India a Hindu nation. For him, this seemed an almost ridiculous
proposition: as he saw it India is already Hindu. His association with the BJP was
inherited from his father and was not something he really questioned, and he did not
consider becoming affiliated to the Congress Party for the same reason. He was recruited
to the RSS in 1997 and to the VVKA a couple of years later by Anjankar, the
schoolteacher who introduced me to the village. He became a devotee of the Mahanubhav
sect in 1998.

The RSS has had seemingly little presence in Markakasa; immediately after Brijlal
joined, he and two other village swayamsevaks (volunteers) started weekly meetings
(shaka) where nationalist songs were sung and the RSS flag was hoisted. But they
stopped holding shaka three months later for lack of interest and it has never since
restarted. RSS higher-ups do visit Markakasa sporadically but tend to visit Brijlal only;
they do not have a wider audience. Nor has the RSS entered the village’s collective
consciousness for all I could gather: some people are aware of its existence but most are
not, and the shakas seem to have made no impact at all. The VVKA is certainly more
visible in the area and organises sport for youths, offers free medical supplies and runs a
boys’ hostel (chatravasa) in the nearby village of Chichgarh; though it is not a constant
presence, most Markakasa people are aware of it. Importantly, it is not seen by most
people as linked in any way to the BJP, and both members of the latter and of the
Congress Party in the area have joined it. This is in part due to the success of the
VVKA’s strategy of persuading people that it is not a political organisation. One local
leader on his visit to the village reiterated to me that “we don’t do politics. Our only
concern is with making vanvasis conscious. Our work is about Hindus and Hindutva.”140
The other side of this is that the BJP is not seen as particularly concerned with issues of
Hindutva or Hinduness. It is regarded locally as a political party similar to the Congress,

140 Ami rajkaran nai karat. Amtsa kam phakta vanvasi jagruti badal ahe. Amtsa kam Hindu ani Hindutva
badal ahe.

218
and a number of local leaders have switched allegiance back and forth between the two;
one prominent local BJP organiser is in fact a Muslim. In short, at least in this part of
Maharashtra, the notion that the BJP is an ideological party with a clear agenda like the
Bahujan Samaj Party (BSP), a low-caste/Untouchable party, seems to have very little
currency.

One aspect which impressed Brijlal about these organisations was their ability to help a
man without connections like him; in short, it provides sifaras (introductions)141. On
several occasions, for instance, VVKA workers in Chichgarh have helped intercede with
the police on behalf of people accused of collaborating with the Naxalites. One such case
involved Brijlal’s salu (WFBDH), Gota, who lives in the neighbouring village of
Gujarbarga142.

Gujarbarga is an all-Gond village and is regarded locally as ‘open’ to the Naxalites: the
latter come and go very frequently, largely without fear of police interference. As the
General Election of 2004 approached, the police seemed to become more vigilant: their
patrols increased and their questioning of villagers became more intense. Gota was at
home one morning, making lunch for his wife who had gone to collect mahua flowers.
As it was summer he was cooking rice on a stove in the back garden. He heard his name
called and went through to the front of the house. There he found several policemen, who
asked if he would come with them to the police station in Chichgarh to answer some
questions about the Naxalites. They told him that they had information that he had fed the
insurgents on several occasions. Gota panicked. Claiming that he had to go out the back
to see to the rice, he slipped away and ran barefoot the three kilometres through forest to
his salu Brijlal’s house in Markakasa. He hid out there for a couple of days and then
Brijlal took him to Chichgarh to see Anjankar, who in turn accompanied Gota to the
police station. Anjankar assured the police that Gota was a good and honest man. The
police held him overnight but let him go the following morning. Whether Anjankar’s
intervention affected the outcome of this case is difficult to determine. Both Gota and

141 The correct Hindi word is sifarish but I write it here as my informants spoke it, probably in
Chhattisgarhi.
142 This is the village to which the Angadev was invited, described in Chapter Four.

219
Brijlal nevertheless attributed the leniency of the police to Anjankar’s assistance. And
Brijlal could call on Anjankar because they were both RSS/VVKA men. There was the
sense, therefore, that being a member of these organisations brought one into contact with
people of influence who could be useful in times of need.

The social benefit of associating with various Hindu nationalist organisations is clear: for
people such as Brijlal, who has no connections to ‘big men’ (bade admi), they provide
access to a wider world, the chance to meet and regard as friends people such as doctors
and teachers. While this aspect of local Hindu nationalism is important, it has been my
contention that we need to look at the connections between the particular practices and
values it promotes and the religious experience of people such as Brijlal, in order to
understand why the movement may have long-term attraction which endures despite the
vagaries of electoral fortune.

What seemed to make the greatest impression on Brijlal was the discipline of the RSS
and VVKA, something he frequently mentioned in our discussions about these
organisations. He was struck by the manner in which meetings were conducted. Everyone
would remove their chappals (sandals/flip-flops) before entering a meeting room and,
instead of kicking them off roughly and leaving them untidily, members would place
them neatly in a row. RSS/VVKA events always ran on time even if ‘important people’
had not yet arrived. An important part of this discipline, as seen by Brijlal, was the
exhortation at many of these meetings that people, especially Adivasi people, ought to set
aside time for daily puja (worship) and that they should give up meat and alcohol.

That Adivasi people ate meat and drank alcohol in immoderate quantities and neglected
to perform daily worship, as he Brijlal, a Mahanubhav sect adherent, did, indicated for
him a lack of discipline that prevented Adivasi people from getting on in the world: these
practices kept them ‘backward’ (magaslele). This is where local notions of progress and
development discussed earlier in the chapter converge with the practices and values
advocated by these Hindu Right organisations. The VVKA is also involved in
‘development’ (or ‘improvement’) work of a sort: it attempts to make Adivasi people

220
more recognisably Hindu in order to both reduce what they see as a gulf in practice
between Hindus, and make Adivasi people more resistant to the attractions of Christian
conversion by suggesting that Christianity makes people un-Indian. Making Adivasi
people more Hindu is important because without it the long-term goal of Hindu
nationalism of a united Hindu nation cannot be realised (see Hansen 1996; Froerer 2005).
The project is profoundly patronising. And yet, as I have emphasised throughout, the
commitment on the part of people like Brijlal to these practices has a different source,
one which is rooted in the problems of everyday existence: relations with kin, friends,
and neighbours; illness; suffering; healing.

Let me elaborate further with the example of a man called Edme, a Gond who lives in a
village outside Chichgarh. Edme is a devotee of Ram and is a member of the Ramanandi
religious sect, and his story in some way mirrors that of Brijlal. Several years ago Edme
fell desperately ill and, as with so many others, could not find a satisfactory cure.
Suspecting witchcraft, he went to a local sadhu-sant (holy man ascetic) who prescribed a
series of measures: worship of God several times a day, and abstention from meat and
alcohol. The sadhu explained that Edme’s illness was ultimately the result of his sinful
life (a position markedly different to those who join the Mahanubhav sect) and that in
order to be safe in future he needed to follow the path of virtue. Edme’s health improved
and he became a disciple of the sadhu.

Edme has very little land, and he and his wife make their living by weaving baskets and
mats from reeds they collect from the forest. He sells to people all over the area, and had
got to know Anjankar and other VVKA activists in Chichgarh. He was asked if he would
like to be the Superintendent of the VVKA boys’ hostel (chatravasa). They told him that
they wanted someone who was aware of the importance of daily worship of Bhagwan in
inculcating Hindutva. They also emphasised that they wanted a vegetarian (shakahari)
person. Unlike at the Government Adivasi hostels, no meat would be served at
mealtimes, since they regarded the eating the flesh of animals as himsa (violence) and
against the idea of Hindu-ness (Hindutva). These were the reasons why they appointed
him, as he presented them to me. The values that the VVKA promoted in the operation of

221
the chatravasa found resonance with the practices Edme had adopted for different
reasons. Edme was receptive largely because of his personal history of suffering, sin, and
illness.

The following section demonstrates how the projects of reform pursued by the VVKA
intersect with local ideas of ‘improvement’ and what makes a ‘good society’, ideas
influenced by the growing numbers of people joining religious sects in Markakasa.

The holy man’s sermon and Hindu modernity

About a month into my stay in Markakasa, a sadhu-sant (holy man ascetic) called
Balakdas came and gave a series of sermons (pravarchan) at the neighbouring village of
Yedmagondi. Only twenty-four years old, Balakdas was from the Pateshwar Ashram in
Jamdipat, in District Durg, Chhattisgarh. A devotee of the god Ram, he was also closely
allied, as it became clear from his sermon, to the Vishwa Hindu Parishad (VHP), a large
organisation of holy men and devotional orders that sees itself as the custodian and
arbiter of Hindu culture and religion in India and throughout the world. His visit had been
organised by the Shri Ram Seva Satsang Samiti143, established by the Yedmagondi
sarpanch* s husband, Girdari, a devotee of the Ramanandi sect. Girdari was also a VVKA
member, and his son attended the organisation’s boys’ hostel in Chichgarh. He ran the
Samiti with the help of other VVKA members in a neighbouring hamlet and with
financial support from the VVKA itself. The Samiti’s main purpose was defined to me as
‘Dharma Jagran’ (Awakening of Faith) among Adivasi people, and organising visits of
holy men-ascetics to the area achieved that end. In addition, the Samiti held religious
meetings (satsang) in six villages in the area where it was explained to people that they
ought to preserve their Hindu sanskruti (culture). This was essentially the message of
Balakdas’ sermon that day, which offers an interesting meditation on his vision of the
ideal society. This is important because of the way in which it resonates with the
discourse of ‘improvement’ and modernity articulated by certain Markakasa people.

143 The Committee for the Service o f Lord Ram

222
About ten people from Markakasa made the six kilometre journey to Yedmagondi that
day, several of them members of the Mahanubhav and other sects, but also a group of
young men who were curious to see this sant and for whom it was an excuse to take a trip
out of the village. Yedmagondi is also a ‘mixed village’ (panchrangi gaori), like
Markakasa, and has a large number of people from all castes who are members of various
devotional sects. Like Markakasa, the largest sectarian group is the Mahanubhav, and
there are smaller but still substantial numbers of Kabir and Ramanandi. Also in
attendance were the boys from the Chichgarh VVKA chatravasa (hostel), their
superintendent Edme and several other local VVKA activists. The connection with the
VVKA is significant. A couple of weeks later, Balakdas was the star turn at a ten-day
religious mela (fair) at the town of Chhuriya about forty kilometres to the east in
Chhattisgarh, where he gave an afternoon sermon every day. The head of the VVKA
came to Chichgarh and from there took a number of local activists, including my landlord
Brijlal and the teacher Anjankar, to Churiya to meet Balakdas. They asked him if he
could go to Melghat, another Adivasi area in western Vidarbha, and give a pravarchan
there as he had done in Yedmagondi. Balakdas’s involvement with VVKA projects and
the coincidence of his ideology and theirs is clear. So what did Balakdas say to the
villagers that day in Yedmagondi?

Balakdas arrived in a small white Maruti van preceded by a line of devotees ringing bells.
He made his way to the dais that had been prepared for him and began to chant a prayer
which he ended by shouting “Hindu dharma ki Jai! [Hail the Hindu Religion!]”, “Bharat
Mata ki Jai! [Hail Mother India!]”, “Gau Mata ki Jai! [Hail Mother Cow!]”, which were
repeated by some of the 250 people assembled. These introductory shouts were an
appropriate beginning, since his sermon was essentially a discussion of these three areas.
He started to speak on the subject of cow protection. The latter was absolutely essential,
he said, if Hindus were to create Ayodhya here; if we did not act to prevent cow slaughter
we would instead live in Lanka. As the Kingdom of Ram, Ayodhya is a common
metaphor for the ideal society just as Lanka, the kingdom of his foe, Ravana, is the
ultimate land of sin and wickedness. He told the following story:

223
 “A tiger came upon a cow grazing in the forest. As the tiger got ready to
pounce, the cowherd said to the tiger, ‘Take me instead’. The tiger could
not believe that a man would substitute his life for a cow and so left
them both alone.”

The sermon moved on to Hindu sanskruti (culture) and Hindu dharma (religion) and he
asked those present to be wary of Christians visiting the area, who encouraged people to
give up the Bhagavad Gita and Ram and take up the Bible and Christ instead. They
would offer food, clothing and money, he said, to persuade people to give up their ‘Hindu
dharma’. Adivasis should not succumb: “If anyone like that comes to your village tell
them they are not welcome. Kick them out!”

Balakdas now broached the difference as he saw it between the virtue of the village
(gaon) and the evil of the city (sahar). In cities there was great violence, and bombs
exploded, and people lived unrighteous lives. The people of the villages were beginning
to behave in a similar fashion, Balakdas warned, and so they and their way of life
required protection. Protection, he seemed to be suggesting, came in the form of hostility
to Christians, no truck with cow slaughter and an opposition to the English language and
its associated foreign culture.

“Why do young people say ‘Happy New Year’ on 1st January? It is not
our New Year yet people in the city say it to one another and hold
parties. Why do they do that? That is the New Year according to the
calendar that the English gave us. But our calendar is different and our
New Year is different. The people of the cities already live by the
English calendar; the people of the village should live by the Hindu
calendar. I was looking around the houses in this village. How many of
them had ‘Sitaram Sitaram’ written on them? None. How many had
‘Welcome’ written on them? Many. Why? What is this ‘Welcome’? It is
an English word. Why do people say ‘Hi’ and ‘Hello’ when they meet?

224
 They should not. These are not our Hindu sanskruti. We should say ‘Jai
Hind!’ instead.”

He then began a discussion of sin and declared that eating the flesh of animals was not
part of Hindu sanskruti either. Combining the attack on ‘Angrezi culture’ and meat-eating
practices, he rounded on the culture of people in the cities “who have begun to celebrate
birthdays by cooking ‘cake’, a type of sweet from abroad (videsh) made with eggs. They
light candles on the top and then blow them out. How wrong! How wrong of them to
celebrate by cooking eggs! And, how wrong of them to create darkness by extinguishing
the light on this joyous occasion!144 How is Ayodhya to be created if people behave like
this?” He ended his sermon with a series of shouts from him: “Ramchandra ki Jai! [Hail
Ram!]”, “Bharat Mata ki Jai”, which were taken up by the congregation.

Balakdas’s sermon was important because of the way in which it highlighted the link
between the daily practices of people (diet, greeting etc.) and the project of creating an
ideal society in India, one governed according to the principles of Ram Rajya. It is also
indicative of the connections made between the type of diet one has and the type of
society one creates. For Balakdas, those who ate meat and drank alcohol were unable to
participate in the (re-)establishment of Ram Rajya. Meat-eating was characterised as a
foreign trait and not part of Hindu culture; as such it implied that those who continued to
eat meat could not be regarded as proper Hindus. Sect members who have embraced the
practices of vegetarianism and teetotalism for reasons of protection from harm see
resonances in the discourses of such holy men with the experience of their own lives. I
asked Om Prakash, who became a member of the Ramanandi sect when his wife was ill,
why he was a member of the BJP and not the Congress: “the Congress people always
serve a lot of meat at their functions and I don’t like it: eating meat and drinking alcohol
is all they do. The BJP don’t do that - no meat and no alcohol, the atmosphere of their
meetings is better.” Thus the commitment to the BJP was not expressed abstractly, as
support for a party which promoted a Hindu agenda. Rather, the idiom of meat-eating and

144 The talk o f birthday cakes totally mystified many Markakasa people. Apart from a few o f the boys who
had noticed them in films, they asked me later to explain what a cake was.

225
alcohol was used to distinguish the two parties, and why one was supported over the
other. Clearly such sentiments were not articulated by every sect member in Markakasa,
and not every sect member necessarily supported the BJP come election time. Rather,
what I want to emphasise here is that the appeal of Balakdas’ message, of the articulation
of what it means to be a ‘good Hindu’ and what constitutes ‘good Hindu society’,
resonates with people who adopt certain practices not because they want to become good
Hindus or create good Hindu society but because they are in search of a permanent
solution to the problem of suffering and pain. Nevertheless, once they recognise that their
practices are constitutive of good Hindu-ness, they may become receptive to the message
of Hindu nationalist actors. Their identity as a particular type of Hindu may then be
mobilised for political ends, even though they have become more identifiably Hindu
(through sectarian membership) for reasons that are far from political.

The implications of such sermons (and, by association, the activities of the VVKA) in
terms of transformations in sociality is well-understood by many non-sect, meat-eating,
(especially Gond) Markakasa people. Ramu Koreti, the estranged brother of Ramdas (see
Chapter Five), criticised the sadhu-sant who give sermons, and the Adivasis who listened
to them:

“Why do these sadhu-sant people come here to Adivasi areas? Why do

they tell us we should give up eating meat and drinking alcohol? This is
our Adivasi culture [sanskruti]. We are Gonds. This is what we do, we eat
meat and drink daru. We should not be listening to sadhu-sant - they are
not for us. They’re ok for Chhattisgarhi people but they are not for us
Gonds.”

As I argued earlier in the chapter, his concerns thus expressed have less to do with the
practice of a particular diet than with what vegetarianism and teetotalism signify in a
transformed sociality that membership of the sect has caused. He speaks from his own
experience of being estranged by a brother who joined the sect and would no longer
worship their household gods nor eat at his house.

226
Whereas the sects, and the Mahanubhav sect in particular, promotes an ideal of Hindu
behaviour which resonates with the Hindu nationalist message, the practice of this ideal
comes into conflict with the non-sect Hindu (and particularly Adivasi) population of the
village. There is perhaps something of a paradox here: whereas the purpose of Hindutva
is to unify the Hindu population nationally, in practice and belief (and Hinduise or
exclude the non-Hindu population), membership of religious sects, which outwardly
espouse many of the ideals of Hindutva, can lead to conflicts at the local level between
different categories of Hindu.

Conclusion

This chapter has attempted to show the ways in which vegetarianism and teetotalism have
meaning for the people who practice them. It argues that the prime consideration is that
of protection rather than a search for prestige. Examining changes in dietary practice need
to be seen in the larger context of religious transformation, which restructures and
reconfigures relations between people. Nevertheless, vegetarianism and teetotalism do
have a value beyond these immediate concerns, and are part of a dialogue about the
constitution of the ideal society and the participation in the latter of people regarded to be
outside the mainstream. Prestige may not have a role to play at the start of this process
but, through a particular articulation of the relationship between ‘wild’ and ‘improved’,
we can see that it comes to have importance. The message of Hindu nationalist
organisations and holy men also resonates with sect members’ own experience, and that
too is involved in an interpretation of what constitutes the modem and what does not. At
the root, however, is still the original motivation for being vegetarian, and focussing on
this, methodologically and analytically, enables us to see the processes of translation that
go on when certain practices are entangled in larger visions.

227
Plate 34: A village boy playing on the verandah o f his house.

228
 Chapter Seven

Conclusion

Though Brijlal is healed and happy, rupture and division seem to characterise
Markakasa society. The ever-present problem of mystical attack at the village-level
remains unsolved and is indeed exacerbated, as protective gods now lie un-worshipped in
their groves, and witches and those who employ sorcerers go undetected and unpunished.
By seeking a personal solution to the problem of attack, Mahanubhav sect members
effectively prevent others in the village from dealing successfully with witches and other
malfeasants because of their opposition to witch-detecting devices such as the Angadev.
Sectarian membership is thus never a matter which only affects the person who becomes
an adherent. Just as the self is transformed, so is society, as a concept understood both
subjectively as experienced and objectively as studied. This change in society cannot be
thought of merely as a consequence of changing religiosity and modes of worship; for
many sect members, the fact of changed sociality and of the restructuring of kinship and
village relations actually makes the sect more appealing. The commitment produced
towards practices such as daily prayer, vegetarianism and teetotalism also resonates
strongly with wider currents in Indian society concerned with the fashioning and fixing of
Hindu identity. The causes and consequences of religious transformation thus raise
important issues, and only by taking religious experience seriously can one fully
understand these concerns and the social transformations they generate.

Before I examine some of the questions raised by the ethnography, and suggest new
questions too, I would like to look at how the analysis develops from one chapter to the
next.

In Chapter Three, I examined the place of ritual friendship in Markakasa society and
suggested that it fulfils a functional need. Not only does ritual friendship enable people to
imagine and create a space where the otherwise overwhelming dominance of caste
ideology does not really operate, but it also provides for a type of relationship that looks

229
very much like brotherhood but is free from the dangers associated with close agnatic
kinship. Rather than examine the important topics of caste and kinship on their own
terms, the intention was to interpret them through the lens of another form of association
- ritual friendship - that is important to Markakasa people. It also illustrated the fear
people have of being bewitched or poisoned, both by those close to them (agnatic kin)
and by those others such as neighbours and ‘ordinary’ friends with whom they are
involved in exchange on a daily basis. Ritual friendship provides a way of expanding
sociality out into the wider world without having to fear dispute or attack. The discussion
also highlighted the play between structure (as represented by caste and kinship) and
sentiment (as represented by the love \prem\ between two ritual friends). This classic
tension in anthropological thought - between structure and agency - is examined further
in the rest of the thesis. Looking at ritual friendship as a functional form of social
association lays the foundation for the discussions of witchcraft and sociality, in Chapters
Four and Five respectively.

Chapter Four developed the theme of witchcraft introduced in Chapter Three. Instead of
looking at witchcraft qua witchcraft, I explored some of the impediments to dealing with
witches and malignant mystical attack that have developed over the last sixty years.
Clearly the main social change that has occurred is the closer presence of the Indian state
in a variety of forms: the effective implementation of land legislation which makes
migration to the area more difficult; increased police presence due to the Naxalite
insurgency; and the promotion of a particular type of modernity (both from inside and
outside the village) which casts Adivasi people in the image of modernity’s ‘other’. The
discussion of changed village composition and state activity is important because it
provides sociological reasons for the kinds of religious transformation people undergo
when they become devotees of a Hindu sect. Read in conjunction with the chapter which
follows it, it demonstrates once again the tension between ‘society’ on the one hand, and
the ways in which people attempt to control their lives and make meaning in changed
conditions on the other.

230
Chapter Five looked in detail at this attempt and at the consequences of religious
transformation. It took us back to the earlier exposition of the dangers inherent in kin
relationships (especially agnatic kin relations), and showed how people use the rituals
and practices of the sect to take control of certain spheres of life which in the past were
problematically reliant on maintaining sociality with others. The analysis was grounded
in the idea that the experience of social life, as well as the pain and suffering it can cause,
is important to understand religious transformation that results from an attempt to solve
problems of existence. This is significant because it explains why certain practices and
ideas promoted by the sect have such power for those who join, and how they might be
mobilised for political ends. But, while the ethnography for the most part details
instances of rupture for those engaging in religious change, it also suggests some
important continuities. Caste, for instance, is a constant, though the relations between
members of different castes who are also all sect members is modified somewhat. A
continuity of a different nature is the structure of the phenomenon of sorcery: one could
interpret the Mahanubhav sect and the Christian Church as types of counter-sorcery
employed in the absence or ineffectiveness of other measures. This too has political
implications, especially in a context where Christians have been made into an ‘Other’.

Finally, Chapter Six considered the practices of vegetarianism and teetotalism, which are
adopted by sectarian villagers and promoted as ideals of Hinduness by local Hindu
nationalist actors. It suggested that these practices have meaning for sect members as
ways in which they can keep their promise to God and thus be protected from further
harm. The chapter raised questions about conventional interpretations of the adoption of
this dietary practice in India, interpretations which emphasise motives such as prestige or
social advancement. Unable to deny the obvious connection between Brahminical
ideology, vegetarianism and purity, the analysis nevertheless showed how vegetarianism
could have a parallel existence in India outside Brahminical thought. This is important
because it explains how the commitment to a practice comes out of the lived experience
of intimate social life, and not because it is imposed or advocated by more powerful
forces in society. It also discussed the ways in which dietary practices are talked about as
markers of ‘improvement’, which does bring with it ideas of prestige and social

231
distinction. Crucially, local Hindu nationalist actors (activists, holy men) see
vegetarianism as a marker of Hinduness and it is articulated as a component of a virtuous
and united Hindu society. This enables sect members in particular to see resonances
between their personal experience and the political and cultural projects promoted by the
Hindu Right. The end of the main part of the thesis thus suggests a beginning from where
we can examine Hindu nationalist projects of ‘uplift’ and integration, and the role played
by people such as Adivasis in them. Once again, what is emphasised is the creativity of
people to develop meaning for practices which become entangled in larger structures of
power.

Having looked at the chapters separately, I want now to pose new questions and suggest
further avenues of enquiry.

Ritual friendship and the Sect

As it was presented in Chapter Three, ritual friendship may have appeared timeless and
unchanging. But, seen in the light of what was discussed in the chapters that follow it,
perhaps we can suggest connections between the religious transformations taking place
and the function and ideology of ritual friendship. More precisely, is ritual friendship a
palliative for the same sorts of tensions and fears that make membership of the sect so
attractive?

We can observe the following. Just as membership of the sect results from and also
promotes a view of certain forms of sociality as problematic and perhaps even dangerous,
so the ideology of ritual friendship functions in opposition to the types of tensions that
exist between close agnatic kin. Members of the sect have come to fear the actions of
others and therefore take the step of finding a radical and extreme solution to the problem
of mystical attack. Neighbours and kin are problematic, especially in a ‘mixed village’
formed in living memory by migration, where the identity and histories of the people one
lives with are never entirely certain. Those who contract ritual friendships, though not

232
articulating them as such, are actually forming associations which are relatively safe,
outside the daily give and take of family and village life.

Also, as ritual friendship provides a different sort of sociality to that found with brothers
or other agnatic kin, so sect devotees develop a new sense of belonging to the community
of their sect, and their relations with kin and others are transformed. However, sect
membership and ritual friendship operate differently in one important respect: whereas
membership of the sect results in the re-structuring of relations with other villagers and a
withdrawal from key areas of village life, the effect of ritual friendship is actually to
make the imagination of the village all the more meaningful by linking peopleand
families together based on ties of residence.145

All of this begs the question: do sectarian villagers form ritual friendships of a different
sort to those formed by non-sect villagers? We saw in Chapter Three that the institution
of ritual friendship in Markakasa was described as founded exclusively on sentiment. As
such it offered an alternative to the experience of brotherhood. With the rise in popularity
of the religious sects and, consequently, vegetarianism, differences in eating preferences
are now also being taken account of when people make ritual friendships. This seems to
fit in with the ideology of sameness that is talked about, but conflicts with the
overarching principle of prem (love; affection) as the deciding factor in these sorts of
relationships. It is love and affection, after all, which gives such ritual friendships their
power.

I was invited to witness a friendship ceremony between Markakasa’s junior headman’s

daughter and a girl from a neighbouring village. The respective girls were good friends,
as were their fathers; both fathers, Om Prakash (the headman) and Sukhram, were also
gurubhai, disciples of the same guru of the Ramanandi sect, and both families strictly
vegetarian and teetotal. Usually friendship ceremonies are performed at a time of some
other celebration such as a wedding or village fair: in this case Sukhram had invited a

145 This happens even if the ritual friends live in different villages. As I show in Chapter Three, it is often
the case that one’s ritual friend from another village is actually a relative o f one’s fellow villager.

233
sant (holy man) of the Ramanandi sect to come to his village and give a two-day
pravarchan (sermon), and so it was decided that the friendship ceremony would be held
at the end of the last day in front of the sant and those who had come to listen to him.
Like Balakdas’s sermon, discussed in Chapter Six, he too emphasised the need for people
to give up meat and alcohol: they were sinful substances and made people sinful. As such
they were obstacles for the establishment of righteousness {dharma). After the ceremony
I was chatting to the girls and asked them why they had become mahaprasad to one
another. In the course of our conversation they mentioned that there was supposed to
have been a third girl with them that day and all three of them were to have become
mahaprasad. The sant, however, had talked to them all beforehand about the importance
of ritual friendship and had discovered that this third girl was not from a vegetarian
household. He suggested that they rethink becoming mahaprasad with one another. He
told them that dining with one another would become problematic and disputes between
the three friends might arise because of the differences in lifestyle. The problem of diet is
in fact one of sociality, which membership of the sect radically transforms. Ritual
friendship, idealised as being founded on sentiment alone, is qualified in this instance
because of the greater demands of a new type, of sociality, one which can only be enjoyed
by sectarian people. As non-sectarian Markakasa people recognise, fashioning ‘good
Hindus’ who lead virtuous and safe lives involves a great deal of rupture.

Fashioning ‘good Hindus’ out of ‘backward Adivasis’ and others is a project of a

particular Hindu modernity, one that is as old as the anti-colonial nationalist movement
itself. The image of the ‘Adivasi as Other Incorporated’ is important not only to this
Hindu vision of the future, but also to projects of the Indian State.

State and nation

The inability to deal with magical attack in the same ways as in the past is the result in
large part of the policies of the Indian state. One could rather crudely argue, as the
Maharashtra Andhashraddha Nirmulan Samiti has in fact done146, that the failure of the

146 Frontline, August 2006

234
Indian state to provide accessible, effective and inexpensive healthcare in ‘backward’
areas such as eastern Vidarbha has meant that many people find medicine deeply
unsatisfying as a solution to illness. The proposition is that if healthcare (and psychiatric
care in particular) were better then the need to go to places such as Sukdi to be (‘falsely’)
healed would be obviated. Such a proposition is undoubtedly simplistic, and the treatment
of illness is generally characterised by a pluralism which exists even in societies which
have well-funded public health systems. And Markakasa people, despite the lack of
affordable medicine, do generally seek out doctors in the first instance and continue
treatment alongside the consultation of baiga or other healers. The importance of the role
of the Indian State rather lies in the particular projects of modernity it pursues through its
actors. In the state of Maharashtra, at least, such a vision of modernity seems to offer no
space for a belief in the efficacy of sorcery and the existence of witches.147 State actors
such as policemen and teachers, as well as forest guards and others, consider the region in
which they work to be ‘backward’ (magaslele) and ‘wild’ (jangli). Though they may
believe in witchcraft and magic themselves, public policy and their status, as outsiders
concerned with maintaining the distinction between themselves and the ‘Adivasi Other’,
effectively preclude their tolerance of superstition. The lack of State recognition of the
validity o f retribution against the witch, a position which has continued virtually
unchanged from the colonial period, means that other forms of protection have been
sought. The Adivasi is a figure of alterity who is seen as a threat to the progress of
modernity. At the same time, the image of the ‘Other’ is necessary to provide a point of
reference to demonstrate how far ‘the rest of us’ have come.

While the State in India (or at least Maharashtra) pursues projects of modernity, how can
we develop the link between religion and nation? That the religious transformation is
rooted in painful experience and inability to deal with malignant mystical attack is
significant for a sense of nationalist identification. As we saw in Chapter Six, the

147 Cf. other parts o f India such as Jharkhand where the Chief Minister publicly professed a belief in the
efficacy o f ojha or sorcerers. The case o f Maharashtra (and India more generally) contrasts too with parts
o f Africa (e.g. Cameroon) where witches and sorcerers can be convicted o f wrongdoing (Fisiy and
Geschiere 1996; Geschiere 1997:169-97). In this vision o f modernity there is certainly space for a
consideration o f wrongdoing o f this sort. See also Siegel’s discussion o f witchcraft and the state in Africa
(2006: 12-26).

235
practices sect members follow and have a commitment to find resonance with the
projects of various Hindu nationalist actors in the local area. But might not the
resonances go beyond what I describe? If followed to its logical conclusion, one can
begin to see how the violence enacted on the (previously unknowable) witch by God
through the devotion of the sect adherent can be compared to the violence inherent in
Hindu nationalist ideology and practice, where the ‘Other’ is made unknowable and is
therefore threatening. The growth of political Hindu nationalism has in part been
signalled by spectacular and violent events. These include the destruction of the Babri
Mosque at Ayodhya in 1992, the riots in Gujarat in 2002 and the general increase in
violence against Christians and missionaries. Similarly, the numerous rath yatrasm
undertaken by senior members of the BJP were both political stunts and symbolic
movements o f ‘warriors’ ready to defend India from internal and external enemies. While
the analysis of this violence is important in understanding how political Hindutva
operates (see Brass 2003; Nandy et al. 1995; Tambiah 1996), my concern here is to
suggest ways in which one can begin to relate it to the experience of religious
transformation and in doing so find possible reasons for the role of violence in the
cosmology of a putative Hindu State.

How can one connect the personal experience of overcoming pain through ritual and the
nationalist logic of violence against the ‘other’? One way would be to examine the matter
symbolically. As Kapferer (1997) suggests in his discussion of Sinhalese Buddhist
nationalism, Bloch’s (1992) theory of ritual and rebounding violence provides us with a
powerful analytical framework to do this. His principal claim is that, rather than
dissipating violence and disorder, ritual as a transformative act actually serves to
intensify it: the transformation is one from being a victim of attack to being an
unrestrained perpetrator of violence. This element of rebounding conquest has two
dimensions: the first is reproductive in nature, where vitality previously lost is regained
and increased through sacrifice, initiation or healing ritual; the second aspect is
aggression, where violence expands outwards (ibid: 61). In state societies such as India,

148 These are journeys traditionally undertaken by Hindu kings or gods on chariots, which have the religio-
political function o f establishing his kingship over a given territory.

236
the two aspects are separated. The former is the outcome of local ritual by the ruled; the
latter is harnessed by rulers from the ritual of state (ibid: 63-4). What is particularly
interesting in relation to the discussion of Hindu nationalism is that the different types of
ritual action set up a communicative link: “...the evocative power of the rituals of the
rulers for the ruled largely depends on the connection these have with the symbolism by
which ordinary citizens express their reproduction through rebounding conquest. What
happens is that similar symbolic sequences are used by the two groups with different
emphases which express different political idioms.” (ibid: 61). The thesis has examined
the first half of the process Bloch refers to: it has looked at the way in which ritual as
healing through the sect can be transformative in the way one relates to society; the way
in which the magic of the witch or sorcerer is conquered through membership of the sect;
and the way in which the power of the new sect member is expanded in an attempt to
refashion society itself. From being a victim of witchcraft, the sect adherent becomes a
wielder of powerful sorcery himself, which is intensely socially disruptive. The violence
of the ritual act, which provides retribution against the malfeasant through the actions of
God, does not lead to resolution or to the restoration of the status quo ante, but creates
more rupture. Though it falls outside the parameters of my study here, following the logic
of Bloch’s theory, connections can be made to the wider cosmological sphere of Hindu
nationalism whereby it seeks to ‘heal’ the nation and thus amplifies violence.

Violence, both symbolic and real, seems to characterise much Hindu nationalist thought
and action. Similarly, the ‘spectre of conversion’ (Dube and Chatterjee Dube 2003)
seems to haunt Hindu nationalism, and has become entangled in political and popular
discourse with the notion of Indian nationality itself (Hardiman 2003: 279). Building on
my discussion of anthropological theories of conversion in the Introduction, I look here
more specifically at conversion in the context of Indian society.

A question of conversion

At the beginning of the thesis I noted the importance of looking at religious experience,
as expressed as problems with society and sociality, in understanding the appeal of

237
various projects of Indian nationalism (both Hindu and otherwise). Political Hindu
nationalism both pre- and post-Independence has been concerned about the activities of
Christian missionaries, and has seen the religious transformations people undergo as
threatening both to the stability of local society and to the unity and integrity of the
Indian nation, conceived of as singularly Hindu. But what kind of religious
transformation is that of Hindu sectarian membership, and is it not just as radical a social
act as that of conversion to Christianity or Islam? Should we not consider this new
sectarian membership as one of conversion? Is this how we should frame Brijlal’s
religious transformation? Is it useful to think of it as a ‘conversion’, even though in broad
terms he considered himself a Hindu before he joined the Mahanubhav sect and stills
considers himself a Hindu now? By seeing Hindu sectarian membership as ‘conversion’,
and understanding the reasons for it, one can better see the connections between it and the
appearance of Christianity in the village. As I show in Chapter Five, there are
considerable parallels between the two. Such an approach undermines the idea that
‘foreign’ religions are somehow more socially transformative and disruptive than
‘indigenous’ ones. Looking at Hindu sectarian membership as religious conversion also
highlights the plurality of Hindu religious experience, an observation which implicitly
criticises a project promoting a singular view of Hinduism.149

In order to examine this question of Hindu sectarian membership as religious conversion,

I consider here David Hardiman’s (2003) historical examination of a religious reform
movement in early twentieth century western India. Looking at Hardiman’s study will
help illuminate the material I have presented, both in its areas of common concern and in
the ways in which my approach departs from his150.

149 See also Dube and Banerjee Dube (2003: 249).

150 Hardiman’s study also highlights what I lack: a properly historicised account o f social and religious
change. Such an account would enable us to examine whether the transformations discussed here are
recapitulations o f processes that have occurred in the past or truly innovative responses to changed social
circumstances. That I have drawn on historical analyses by Hardiman (1987, 2003) and Skaria (1997a,
1997b) cannot remedy this shortcoming. Nevertheless, the use o f their material provides some
understanding o f the historical context (of other places to be sure) o f the ethnography presented here.

238
Hardiman traces the rise of a Hindu reform movement begun by an Adivasi Bhil man
called Govind, and the ways in which it became interwoven into nationalist agitation
against colonial rule. Govind exhorted the Bhils to adopt vegetarianism and teetotalism,
to practice devotion to Sanskritic Hindu deities and to give up Bhil beliefs in witches and
ghosts as causes of illness or misfortune (2003: 264). In his analysis, Hardiman claims
that the movement was, for Govind, a way in which to assert Bhil power in collaboration
with the colonial authorities against powerful Rajput princes and local Brahmins. Indian
nationalists on the other hand attempted to portray the movement as anti-colonial. They
regarded Bhils, and especially Govind’s Bhils, as Hindus, and saw an opportunity to use
Govind’s movement to spread a ‘civilised’ Hinduism among the ‘janglV (wild) Bhils
(2003: 275). He also suggests that some of these nationalists were anti-Muslim and may
have wanted to bolster local Hindu identity through this process (ibid.).

Hardiman asks us to what extent we can define Govind’s movement as one of

‘conversion.’ He points out that, “It is clear... that systems of belief and practice that were
carried on within India frequently competed with each other to attract followers” (2003:
277). In addition, “those who followed broadly Brahminical rites often tried to popularise
such practices amongst groups that were considered to be ignorant of them, such as the
Bhils” (ibid.). In the Indian context, this sort of social action (‘assimilation’, ‘uplift’),
including what later became known as Sanskritisation, represents an attempt to dispense
with the concept of ‘conversion’; the latter came to mean a change concerning
proselytising world religions only (Robinson and Clarke 2003: 10). Hardiman is right to
expand the category of conversion since it forces us to look at the local logic for religious
change regardless of whether those who undergo transformation see it as a ‘conversion’.
This local logic manifests itself in the attractiveness of different ‘sects’ at different times.

Like the comparisons I draw between adherence to the Mahanubhav Panth and
Christianity, Hardiman goes on to compare Govind’s movement and the conversion to
Christianity that was taking place at the same time. He writes,

239
 “[Govind’s movement] required a rejection of existing Bhil cultural and
religious practices, and this often caused great social tension within
villages and families. Such tensions were also commonly found when
Bhils became Christians, so that there were parallels here to what we
may describe as a process of ‘conversion’” (278).

This discussion raises several related concerns, which demonstrate the difference
between Hardiman’s approach and my own. Firstly, Hardiman imputes the same
intentions to Govind’s followers (who became known as bhagats) as to Govind himself.
When the movement is described as asserting Bhil power against the oppression of
princes and Brahmins, why should we conclude that bhagats buy into the movement for
the same reasons?

Hardiman states that Govind exhorted Bhils to give up belief in witches, ghosts and
demons. He argues that this was the result of the criminalisation by the British colonial
regime of Bhil ways of dealing with witches, and so Govind sought to make the whole
issue redundant (ibid: 264). But Hardiman then goes on to say that people came to see
Govind to be healed from illness, illness which he somewhat contradictorily states was
largely caused by witchcraft and evil spirits (ibid: 264-65). His interpretation is not
entirely convincing and I wonder if he has it slightly wrong. Is it not more likely that
Govind’s movement offered a cure for this sort of attack? Rather than renounce belief in
these entities, it might be the case that Govind convinced people that, if they joined him,
these sorts of entities and attack would no longer affect them.151 Giving up fear and
giving up belief are two entirely different things. If my interpretation of Hardiman’s
account is correct, there are considerable parallels with the changes I have discussed in
the thesis. Once again, at the root of people’s adherence to a sect is the search for a
permanent solution to bodily attack. It is more than likely that the bhagats found this
religious movement attractive because it promised solutions to the problems of dealing
with witchcraft and magic, problems which had become difficult to tackle because of

151 The question o f what is meant by ‘b e lie f is a thorny one and Govind might simply have been using
‘belief in the Old Testament sense o f ‘don’t put your trust in witches or sorcerers’ rather than ‘don’t
believe in witches and sorcerers’. See Pouillon (1982).

240
colonial government policy. This eliding of intentions also characterises other studies of
religious movements in India (see e.g. Jones 1989). What I have shown in Chapters Four,
Five and Six is that there is no easy assimilation of the institution of the sect on the one
hand and the power of practices and membership for its adherents on the other. The
motivations of adherents cannot be assumed, they must be explored.

In light of this general point (which is really a methodological one), we can turn the
quotation above from Hardiman on its head. Hardiman sees the rupture with one’s kin
and village as a consequence of their adherence to the movement.152 But could it not also
be true that the types of problems that exist between people and their kin and neighbours
actually contribute to the attraction of the religious movement in the first place? Why
could rupture not also be a caused Social rupture causes one to join the sect and rupture
results.

One of the major themes of my study is that a focus on religious experience in the context
of local society leads to a better understanding of motivation; of how people are actively
involved in restructuring their lives in conditions of social change brought about by the
fact of living in a nation-state; and how the consequences of such actions play out, both
in local society and as part of wider political projects of the State and the nation. By
looking at religious experience in the context of sociality, we can begin to understand its
true social power. The lives of people like Brijlal are important for what they can tell us
about how people attempt to make meaning from a society in flux and how that attempt
ends up changing society further.

152 O f course he is not alone. In the same collection on religious conversion in India, Dube and Chatterjee
Dube describe the same sorts o f kin and village tensions which follow membership o f a religious sect in
Orissa; they too see such tensions as a consequence o f changed religiosity (2003: 237-38). In their
Introduction to the volume, Robinson and Clarke write, “ ...conversion is perceived to be rarely a
transformation in the realm o f ideas and beliefs alone. Other areas of social life may change. New faiths
and creeds throughout Indian history have negotiated and sometimes clashed with prevailing patterns o f
kinship, marriage and inheritance. Food conventions, sartorial codes and occupational arrangements have
manifested mutation” (2003: 18). But what if these ‘mutations’ were the causes for changed religiosity as
well as consequence?

241
 Bibliography

Auslander, Marc, 1993. “‘Open the Wombs!”: the symbolic politics of modem Ngoni
witch-finding,’ in Jean and John Comaroff (eds) Modernity and its malcontents:
ritual and power in postcolonial Africa. Chicago; London: University of Chicago
Press, pp. 167-92.

Babb, Lawrence, 1975. The divine hierarchy: popular Hinduism in central India. New
York; London: Columbia University Press.

Bailey, F. G., 1960. Tribe caste and nation: A study ofpolitical activity and political
change in highland Orissa. Manchester: Manchester University Press.

Bailey, F. G., 1961. “‘Tribe’ and ‘caste’ in India”, Contributions to Indian Sociology 5:
7-19.

Bailey, F. G., 1997. A witch-hunt in an Indian village. Delhi: Oxford University Press.

Balagopal, K., 2002. ‘Reflections on “Gujarat Pradesh” of “Hindu Rashtra’” , Economic

and Political Weekly 37(22): 2117.

Bates, Crispin, 1985. ‘Regional dependence and rural development in central India: the
pivotal role of migrant labour’, Modern Asian Studies 19 (3): 573-592.

Baviskar, Amita, 2005. ‘Adivasi encounters with Hindu nationalism in Madhya Pradesh’,
Economic and Political Weekly 40 (48): 5105-5113.

Bayly, C. A., 1998. Origins o f nationality in south Asia: patriotism and ethical
government in the making o f modern India. Delhi: Oxford University Press.

242
Benei, Veronique and Christopher J. Fuller (eds), 2001. The everyday state and society in
India. London: Hurst.

Benei, Veronique, 2001. ‘Teaching nationalism in Maharashtra schools’, in Veronique

Benei and Christopher J. Fuller (eds) The everyday state and society in modern
India. London: Hurst, pp. 194-221.

Beteille, Andre, 1987. ‘The concept of tribe with special reference to India,’ European
Journal o f Sociology, 27 (2): 297-318.

Bhatt, Chetan, 2001. Hindu nationalism: origins, ideologies and modern myths. Oxford:
Berg.

Bloch, Maurice E., 1973. ‘The long and the short term: the economic and political
significance of the morality of kinship,’ in Jack Goody (ed.) The character o f
kinship. Cambridge: Cambridge University Press.

Bloch, Maurice E., 1992. Prey into hunter: the politics o f religious experience.
Cambridge: Cambridge University Press.

Bloch. Maurice E., 1999. ‘Commensality and poisoning’, Social Research 66 (1): 133-
149.

Bose, N. K., 1941. ‘The Hindu method of tribal absorption’, Science and Culture 8: 188-
94.

Brass, Paul, 1974. Language, religion and politics in north India. Cambridge: Cambridge
University Press.

Burghart, Richard, 1996. The conditions o f listening: essays on religion, history and
politics in South Asia. Oxford: Oxford University Press.

243
Burra, Nita, 1996. ‘Buddhism, Conversion and Identity: a case study of village Mahars’,
in M. N. Srinivas (ed) Caste: its twentieth century avatar. New Delhi: Viking Press,
pp. 152-73.

Carrier, James, 1999. ‘People who can be friends: selves and social relationships’ in
Sandra Bell and Simon Coleman (eds) The Anthropology o f Friendship. Oxford:
Berg.

Carstairs, G. M. 1983. Death o f a witch: a village in north India 1950-1981. London:

Hutchinson.

Clarke, Sathianathan, 2003. ‘Conversion to Christianity in Tamil Nadu’, in Rowena

Robinson and Sathianathan Clarke (eds) Religious conversion in India: modes,
motivations and meaning. Delhi: Oxford University Press, pp 323-50.

Comaroff, Jean and John Comaroff, 1999. ‘Occult economies and the violence of
abstraction: notes from the South African postcolony,’ American Ethnologist 26(3):
279-301.

Connerton, Paul, 1989, How Societies Remember. Cambridge: Cambridge University

Press.

Csordas, Thomas J., 1990. “Embodiment as a Paradigm for Anthropology”, Ethos 18: 5-
47.

Csordas, Thomas J. (ed). 1994. Embodiment and Experience: The Existential Ground o f
Culture and Self. Cambridge: Cambridge University Press.

Csordas, Thomas J., 1999. “The Body’s Career in Anthropology”, in H. L. Moore (ed.)
Anthropological Theory Today. Cambridge: Polity Press, pp. 172-205.

244
Das, N. K., 1988. ‘Tribal priesthood and shamanism in Chhattisgarh, Journal o f the
Indian Anthropological Society 23: 223-231.

Deliege, Robert, 1999. The Untouchables o f India. Oxford: Berg

Douglas, Mary (ed), 1970. Witchcraft confessions and accusations. ASA Monographs 9.
Tavistock Publications.

Douglas, Mary, 1996. Natural symbols: explorations in cosmology. London: Routledge.

Dube, S. C., 1953. ‘Token pre-puberty marriage in Middle India’, Man article 25, pp. 18-
19.

Dube, Saurabh and Ishita Baneijee Dube, 2003. ‘Spectres of conversion: transformations
of caste and sect in India,’ in Rowena Robinson and Sathianathan Clarke (eds.)
Religious Conversion in India. Delhi: Oxford University Press, pp. 223-254.

Dumont, Louis, 1962. “‘Tribe’ and ‘Caste’ in India”, Contributions to Indian Sociology,
6 : 120- 122 .

Dumont, Louis, 1980. Homo Hierarchicus: the caste system and its implications.
Chicago and London: University of Chicago Press.

Eaton, Richard M., 1997. ‘Comparative history as world history: religious conversion in
modem India’, Journal o f World History 8 (2): 243-71.

Elwin, Verrier, 1943. ‘Folklore of the Bastar clan gods’, Man vol. XLIII 83-102: 97-104.

Elwin, Verrier, 1943. The Aboriginals. Bombay: Oxford University Press.

245
Elwin, Verrier, 1943. Maria murder and suicide. Bombay: Oxford University Press.

Elwin, Verrier, 1947. The Muria and their ghotul. Bombay: Oxford University Press.

Elwin, Verrier, 1957. Leaves from the jungle: life in a Gond village. Delhi: Oxford
University Press.

Engelke, Matthew, 2004. ‘Discontinuity and the discourse of conversion’, Journal o f

Religion in Africa 34(1/2): 82-109.

Enthoven, R. E., 1920-22. The tribes and castes o f Bombay.

Evans-Pritchard, E. E. 1976 [1937]. Witchcraft, magic and oracles among the Azande.
Oxford: Clarendon Press.

Feldhaus, Anne, 1984. The deeds o f God in Rddhipur. New York: Oxford University
Press.

Feldhaus, Anne and Shankar Gopal Talpule, 1992. In the absence o f God: the early years
o f an Indian sect. Honolulu: University of Hawaii Press.

Feldhaus, Anne, 2003. Connected Places: Region, Pilgrimage and Geographical,

Imagination in India. New York; Basingstoke: Palgrave Macmillan.

Fisiy, Cyprian and Peter Geschiere 1996. ‘Witchcraft, violence and identity: different
trajectories in postcolonial Cameroon,’ in Richard Werbner and Terence Ranger
(eds) Postcolonial Identities in Africa. London: Zed Books.

Fitzgerald, Tim, 1997. ‘Ambedkar Buddhism in Maharashtra’, Contributions to Indian

Sociology (n.s.) 31 (2): 225-51.

246
Froerer, Peggy, 2002. From local tension to ethnic conflict: the emergence o f Hindu
nationalism in a Hindu/Christian ‘tribal *community in Chhattisgarh, central India.
PhD thesis, University of London.

Froerer, Peggy, 2005. ‘Challenging traditional authority: the role of the state, the divine,
and the RSS,’ Contributions to Indian sociology (n.s.) 39 (1): 39-73.

Froerer, Peggy, 2006a. ‘Emphasising “Others”: the emergence of Hindu nationalism in a

central Indian tribal community.’ Journal o f the Royal Anthropological Institute
(n.s.) 12: 39-59.

Froerer, Peggy, 2006b. ‘Close friends: the importance of proximity in children’s peer
n j

relations in Chhattisgarh, central India’. Unpublished. Paper presented 22 June

2006 Anthropology of Friendship workshop, London School of Economics and
Political Science.

Fuller, Christopher J. 1992. The camphor flame: popular Hinduism and society in India.
Princeton: Princeton University Press.

Furer-Haimendorf, C. von, 1948. The Raj Gonds o f Adilabad. London: Macmillan.

Furer-Haimendorf, C. von, 1979. The Gonds o f Andhra Pradesh: tradition and change in
an Indian tribe. London: Allen and Unwin.

Furer-Haimendorf, C. von, 1982. Tribes o f India: Struggle fo r Survival Berkeley:

University of California Press.

Gell, Alfred, 1980. ‘The gods at play: vertigo and possession in Muria religion’, Man
(n.s.) 15 (2): 219-248.

247
Gell, Alfred, 1982. ‘The market wheel: symbolic aspects of an Indian tribal market’, Man
(n.s.) 17 (3): 470-491.

Gell, Alfred, 1986. ‘Newcomers to the world of goods: consumption among the Muria
Gonds,’ in Arjun Appadurai (ed.) The social life of things: commodities in cultural
perspective. Cambridge: Cambridge University Press, pp. 110-138.

Gell, Alfred, 1997. ‘Exalting the king and obstructing the state: a political interpretation
of royal ritual in Bastar District, central India’, Journal o f the Royal
Anthropological Institute 3 (3): 433-450.

Gell, Simeran Man Singh, 1992. The ghotul in Muria society. Reading, U.K.: Harwood
Academic Publishers.

Geschiere, Peter, 1997. The modernity o f witchcraft: politics and the occult in
postcolonial Africa. Charlottesville; London: University Press of Virginia.

Geschiere, Peter, 1999. ‘Globalisation and the power of indeterminate meaning:

witchcraft and spirit cults in Africa and east Asia,’ in Birgit Meyer and Peter
Geschiere (eds) Globalisation and Identity. Oxford: Blackwell, pp. 211-37.

Ghurye, G. S., 1943. The aboriginals - so-called - and their future. Pune: Gokhale
Institute of Politics and Economics.

Graeber, David, 2001. Toward an anthropological theory o f value: the false coin o f our
dreams. New York: Palgrave.

Grigson, W. V., 1938. The Maria Gonds o f Bastar. London: Oxford University Press.

Guha, Ramachandran, 1999. Savaging the civilised: Verrier Elwin, his tribals and India.
Delhi: Oxford University Press.

248
Guha, Sumit, 1998. ‘Lower strata, older races, and aboriginal peoples: racial
anthropology and mythical history past and present,’ Journal o f Asian Studies 57
(2): 423-441.

Gupta, Akhil, 1995. ‘Blurred boundaries: the discourse of corruption, the culture of
politics and the imagined State,’ American Ethnologist 22 (2): 375-402.

Hansen, Thomas Blom, 1996. ‘The vemacularisation of Hindutva: the BJP and Shiv Sena
in rural Maharashtra,’ Contributions to Indian sociology (n.s.) 30 (2): 177-214.

Hansen, Thomas Blom, 1999. The saffron wave: democracy and Hindu nationalism in
modern India. Princeton: Princeton University Press.

Hardiman, David, 1987. The coming o f Devi: Adivasi assertion in western India. New
Delhi: Oxford University Press.

Hardiman, David, 2003. “Assertion, conversion and Indian nationalism: Govind’s

movement among the Bhils”, in Rowena Robinson and Sathianathan Clarke (eds.)
Religious Conversion in India. Delhi: Oxford University Press, pp. 255-84.

Hefner, Robert, 1993. ‘Introduction: world building and the rationality of conversion’ in
Robert Hefner (ed.) Conversion to Christianity: historical and anthropological
perspectives on a great transformation. Berkeley: University of California Press,
pp. 3-44.

Hertz, Robert, 1960. ‘The pre-eminence of the right hand’, in Death and the right-hand,
trans. Rodney Needham and Claudia Needham. London: Cohen and West.

Horton, Robin, 1971. ‘African conversion’, Africa 41: 85-108.

249
Jackson, Michael, 1989. Paths toward a clearing: radical empiricism and ethnographic
inquiry. Bloomington: Indiana University Press.

Jaffrelot, Christophe, 1993. The Hindu nationalist movement and Indian politics: 1925 to
the 1990s. London: Hurst & Co.

Jay, Edward J., 1973. ‘Bridging the gap between castes: ceremonial friendship in
Chhattisgarh’, Contributions to Indian Sociology 7:144-158.

Kapferer, Bruce, 1997. The feast o f the sorcerer: practices o f consciousness and power.
Chicago: University of Chicago Press.

Kapferer, Bruce, 2003. ‘Outside all reason: magic, sorcery and epistemology in
anthropology’, in Bruce Kapferer (ed) Beyond Rationalism: rethinking magic,
witchcraft and sorcery. New York: Bergahn Books, pp. 8-24.

Kapur, Sohaila, 1983. Witchcraft in western India. London: Sangam.

Kelkar, Govind and Dev Nathan, 1991. Gender and Tribe: Women, Land and Forests in
Jharkhand. New Delhi: Kali for women.

Keller, Eva, 2005. The road to clarity: Seventh-Day Adventism in Madagascar. New
York; Basingstoke: Palgrave Macmillan.

Kipp, Rita, 1995. ‘Conversion by affiliation: the history of the Karo Batak Protestant
Church’, American Ethnologist 22 (4): 868-882.

Laitlin, David D., 1986. Hegemony and culture: politics and religious change among the
Yoruba. Chicago: Chicago University Press.

250
Lambert, Helen, 1996. ‘Caste, Gender and Locality in Rural Rajasthan’, in C. J. Fuller
(ed). Caste Today. Delhi: Oxford University Press, pp. 93-123.

Lobo, Lancy, 2002. ‘Adivasis, Hindutva and post-Godhra riots in Gujarat,’ Economic and
Political Weekly 37 (48): 4844-49.

Loizos, Peter and Evthymios Papataxiarchis (eds), 1991. Contested Identities: Gender
and Kinship in Modern Greece. Princeton, N. J.: Princeton University Press.

Lyon, M. L. and J. M. Barbalet, 1994, ‘Society’s body: emotion and the “somatisation”
of social theory’, in Thomas J. Csordas (ed.) Embodiment and experience: the
existential ground o f culture and self. Cambridge: Cambridge University Press.

Macdonald, Helen, 2004. Resolution and rupture: the paradox o f violence in witch
accusation in Chhattisgarh, India. Unpublished PhD thesis. University of London.

Mauss, Marcel, 1979. ‘Techniques of the body’, in M. Mauss, Sociology and Psychology.
London: Routledge and Kegan Paul.

McKean, Lise, 1996. Divine enterprise: gurus and the Hindu nationalist movement.
Chicago: University of Chicago Press.

Mendelsohn, Oliver and Marika Viciziany, 1998. The Untouchables: subordination,

poverty and the state in modern India. Cambridge: Cambridge University Press.

Merleau-Ponty, Maurice, 1989. Phenomenology o f perception. London: Routledge.

Meyer, Brigit, 1998. “‘Making a complete break with the past”: memory and postcolonial
modernity in Ghanaian discourse,’ Journal o f Religion in Africa 28 (3): 316-349.

251
Meyer, Brigit and Peter Pels, 2003. Magic and modernity: interfaces o f revelation and
concealment. Stanford, CA: Stanford University Press.

Mishra, Archana, 2003. Casting the evil eye: witch trials in tribal India. New Delhi:
Namita Gokhale Editions; Roli Books.

Moore, Henrietta and Todd Sanders, eds, 2001. Magical interpretations, material
realities: modernity, witchcraft and the occult in postcolonial Africa. London:
Routledge.

Nathan, Dev and Govind Kelkar, Yu Xiaogang, 1998. ‘Women as witches and keepers of
demons: cross-cultural analysis of struggles to change gender relations,’ Economic
and Political Weekly 33 (44), Review of Women Studies 58-70.

Obeyesekere, Gananath, 1981. Medusa’s Hair: an essay on personal symbols and

religious experience. Chicago; London: University of Chicago Press.

Parry, Jonathan, 1986. ‘The Gift, the “Indian” Gift, and the “Indian Gift’” , Man (n.s.) 21
(3): 453-473.

Parry, Jonathan, 1991. ‘The Hindu lexicographer? A note on auspiciousness and purity’,
Contributions to Indian Sociology 25: 267-285.

Parry, Jonathan, 1994. Death in Benares. Cambridge: Cambridge University Press.

Parry, Jonathan, 2001. ‘Ankalu’s errant wife: sex, marriage and industry in contemporary
Chhattisgarh’, Modern Asian Studies 35(4): 783-820.

Peel, J. D. Y., 1977. ‘Conversion and tradition in two African societies: Ijebu and
Buganda, ’ Past and Present 77: 108-141.

252
Pffeffer, G., 2001. ‘A ritual of revival among the Gadaba of Koraput’ in H. Hulke and B.
Schnepel (eds) Jaganath revisited: studying society, religion and the state in
Orissa. New Delhi: Manohar.

Pitt-Rivers, Julian, 1973. ‘The Kith and the Kin’ in Jack Goody (ed) The character o f
kinship. Cambridge: Cambridge University Press, pp. 89-105.

Pouillon, Jean, 1982. ‘Remarks on the verb “to believe” ’ in Michel Izard and Pierre
Smith (eds.) Between belief and transgression: structuralist essays in religion,
history, and myth. Chicago, London: University of Chicago Press, pp. 1-8.

Prakasam,G., 1994. The Satnamis o f Chhattisgarh. unpublished thesis, University of

Oxford.

Raeside, I. M. P., 1976. ‘The Mahanubhavas’, Bulletin o f the School o f Oriental and
African Studies 39 (3): 585-600.

Rezende, Claudia B., 1999. ‘Building affinity through friendship’, in Sandra Bell and
Simon Coleman (eds) The Anthropology o f Friendship. Oxford: Berg.

Robbins, Joel, 2003. ‘On the paradoxes of global Pentecostalism and the perils of
continuity thinking’, Religion 33: 221-231.

Robbins, Joel, 2004. Becoming sinners: Christianity and moral torment in a Papua New
Guinea society. Berkeley: University of California Press.

Robinson, Rowena and Sathianathan Clarke, 2003. ‘The many meanings of religious
conversion on the Indian subcontinent,’ in Rowena Robinson and Sathianathan
Clarke (eds.) Religious conversion in India. Delhi: Oxford University Press, pp. 1-
21 .

253
Russell, R.V. (ed), 1908. Central Provinces District Gazetteers: Bhandara District Vol. A
(Descriptive). Nagpur: Central Provinces Press.

Sahlins, Marshall, 1974. Stone Age economics. London: Tavistock.

Sarkar, Sumit, 1999. ‘Hindutva and the question of conversions’ in K. N. Pannikar (ed)
The concerned Indian’s guide to communalism. Delhi: Viking Press.

Sharma, Arvind, 1977. ‘The place of conversion within Hinduism’, Contributions to

Indian Sociology 11 (2): 345-354.

Siegel, James, 2006. Naming the witch. Stanford: Stanford University Press.

Singh, K. S. 1985. Tribal society in India: an anthropo-historical perspective. New

Delhi: Manohar.

Singh, Prakash, 1995. The Naxalite movement in India. New Delhi: Rupa.

Sinha, Shantha, 1989. Maoists in Andhra Pradesh. New Delhi: Gian Publishing House.

Skaria, Ajay, 1997a. ‘Women, witchcraft and gratuitous violence in colonial western
India’, Past and Present, 155:109-141.

Skaria, Ajay, 1997b. ‘Shades of wildness: tribe, caste, and gender in western India’,
Journal o f Asian Studies, 56 (3): 726-745.

Skoda, Uwe, 2004. ‘Ritual friendship in a converging tribal and caste society’, Journal o f
Social Science 8 (2): 167-177.

254
Skultans, Vieda, 1987. ‘The management of mental illness among Maharashtrian
families: A case study of a Mahanubhav healing temple’, Man (N.S.) 22 (4): 661 -
679.

Sontag, Susan, 1979. Illness as metaphor. London: Allen Lane.

Srinivas, M. N., 1952. Religion and society among the Coorgs o f south India. Oxford:
Clarendon Press.

Srinivas, M. N., 1966. Social change in modern India. Cambridge: Cambridge University
Press.

Stoller, Paul, 1989. The taste o f ethnographic things: the senses in anthropology.
Philadephia: University of Pennsylvania Press.

Stoller, Paul, 1997. Sensuous scholarship. Philadelphia: University of Pennsylvania

Press.

Strathem, Andrew, 1996. Body thoughts. Ann Arbor: University of Michigan Press.

Sundar, Nandini 1997. Subalterns and sovereigns: an anthropological history o f Bastar

1854-1996. Delhi: Oxford University Press.

Sundar, Nandini, 2001. ‘Divining evil: the state and witchcraft in Bastar’, Gender
Technology and Development 5 (3): 425-448.

Sundar, Nandini, 2006a. ‘Adivasi vs vanvasi: the politics of conversion in central India,
in Satish Saberwal and Mushirul Hasan (eds), Assertive religious identities. Delhi:
Manohar, 357-90.

255
Sundar, Nandini, 2006b. ‘Bastar, Maoism and Salwa Judum’, Economic and Political
Weekly 4\ (29): 3187-3192.

Sundkler, B. G. M., 1961. Bantu prophets in South Africa. London: Oxford University
Press.

Van der Veer, Peter, 1988. Gods on earth : the mafiagement o f religious experience and
identity in a North Indian pilgrimage centre. London: Athlone.

Van der Veer, Peter, 1994. Religious nationalism: Hindus and Muslims in India.
Berkeley: University of California Press.

Viswanathan, Gauri, 1998. Outside the fold: conversion, modernity and belief. Princeton:
Princeton University Press.

Wakankar, Milind, 2005. ‘The anomaly of Kabir: caste and canonicity in Indian
modernity’ in Shail Mayaram, M. S. S. Pandian, Ajay Skaria (eds.) Muslims, Dalits
and the fabrications o f history (Subaltern Studies XII). New Delhi: Permanent
Black.

Yalman, Nur, 1963. ‘On the purity of women in the castes of Ceylon and Malabar’,
Journal o f the Royal Anthropological Institute, 93: 25-58.

Yagnik Achyut and Suchitra Sheth, 2002. ‘Whither Gujarat? Violence and after’,
Economic and Political Weekly 37 (11): 1009-1 111.

Zelliot, Eleanor 1996. ‘Learning the use of political means: the Mahars of Maharashtra’
in From Untouchable to Dalit: Essays on the Ambedkar movement. New Delhi:
Manoharpp. 86-125.

256
257