ArticlePDF Available

Wild Sacredness and the Poiesis of Transactional Networks Relational Divinity and Spirit Possession in the "Būta" Ritual of South India

Authors:

Abstract and Figures

The ritual practices of the low castes have often been considered through concepts such as Sanskritization as well as consensus and replication, but have also been interpreted as resistance against the dominance of the high castes. The tendency common to these analyses is their interpretation of the low castes’ ritual practices in terms of caste hierarchy and power relations. Focusing on the relational aspect of divinity and the importance of wild sacredness in ritual contexts, this study will provide an alternative perspective from which to view the complementary opposites in the rituals of the low castes. These are not merely a reflection of unequal caste relations, but are the basis of the relationships among all the various actors—including human beings, wild animals, and spirits—personified as būtas that constitute a fluid network in a social, ecological, and cosmological sphere.
Content may be subject to copyright.
Wild Sacredness and the Poiesis
of Transactional Networks
Relational Divinity and Spirit Possession
in the Būta Ritual of South India
The ritual practices of the low castes have often been considered through con-
cepts such as Sanskritization as well as consensus and replication, but have also
been interpreted as resistance against the dominance of the high castes. The
tendency common to these analyses is their interpretation of the low castes’
ritual practices in terms of caste hierarchy and power relations. Focusing on
the relational aspect of divinity and the importance of wild sacredness in ritual
contexts, this study will provide an alternative perspective from which to view
the complementary opposites in the rituals of the low castes. These are not
merely a reflection of unequal caste relations, but are the basis of the relation-
ships among all the various actors—including human beings, wild animals,
and spirits—personified as būtas that constitute a fluid network in a social,
ecological, and cosmological sphere.
: spirit possession—transactional networks—śaktibūta ritual—
South India
M I
Kyoto University
Asian Ethnology Volume , Number , –
© Nanzan Institute for Religion and Culture
Būta  deities and spirits worshipped widely in South Kanara, the coastal areas
of Karnataka.1 They are generally considered deities, such as apotheosized
local heroes or heroines or the spirits of wild animals dwelling in forests. The būtas
are closely related to and also embody the wild, dangerous, and fertile aspects of
divine power.
The būta ritual is mainly constituted of spirit possession, oracles, and the inter-
actions between the devotees and the būtas incarnated in the impersonators
belonging to the Nalike, Parava, and Pambada castes (all designated scheduled
castes). Priest-mediums called pātri or māni of the Billava caste and mukkāldi of
the Bant caste conduct the ritual.2 Among all the devotees at the būta shrine, the
patrons of the shrine play the most important role. Most of them are the landlords
of local manors called guttus, who belong to the Bant caste.3
Since the colonial period, European writers have dierentiated būta worship
from the Sanskritic rituals for Hindu deities and have considered the former as
inferior to the latter. For example, T (, –) described būta wor-
ship as “devil dances” and a būta as a “demon” served by and incarnated in “a
Pombada or a Nalke, a man of the lowest class.
Thurston’s description makes explicit the general view of the colonial administra-
tors and missionaries, a view based not only on Christianity, but also on the Brahman-
ical doctrine of native Brahman informants. Būta worship in South Kanara was thus
regarded as one of the “archaic” Dravidian cults discriminated against by Brahman
priests who conducted Sanskritic rituals (N and F, ix–xx).
Apart from the colonial interpretation of popular rituals, ritual practices of the
low castes have often been understood in academia by contrast with the dominant,
Sanskritic rituals. S () described the customs and rituals of the lower
castes using the notion of Sanskritization, while M () interpreted them
through the concepts of consensus and replication. On the other hand, the ritual
practices of the low castes have often been interpreted as resistance against the ide-
ology and values of the high castes.4
A question common to these studies is this: how should we understand the
ritual practices of the non-Brahman, low-caste people? In order to answer this
question, most analyses focus on the hierarchical relationship and power relations
between the high and low castes. In such analyses the whole meaning and func-
 | Asian Ethnology Volume , Number 
: būta     | 
tion of a ritual is often reduced to a symbolic representation of the unequal caste
hierarchy in the social-political sphere. However, as we will see in this article, these
rituals concern not only caste hierarchy, but also the creation of a network through
which particular actors, substances, and power are related to each other in more
complex ways.
In this article, I attempt to reconsider the hierarchy-centered analyses of the
low castes’ ritual practices, such as theories of Sanskritization, consensus and rep-
lication, and resistance, through a close investigation of būta worship. Extending
A and B’s () notion of a “transactional network” and
reconsidering D’s () notion of relational divinity as well as his insight
into the importance of complementary opposites in Hindu rituals, this study pro-
vides an alternative view to the hierarchy-centered analysis; namely, a perspective
to analyze complementary opposites in ritual contexts not merely as the reflec-
tion of unequal caste hierarchy, but as the core of the relationships among various
actors, including not only human beings but also wild animals and spirits personi-
fied as būtas.
I   
Ritual practices of the low castes have often been considered through
concepts such as Sanskritization, consensus, and replication. S () first
described the concept of Sanskritization as the adaptation of Brahmanical rituals,
beliefs, and ways of life through which low castes seek to improve their position
in the caste hierarchy. For this Sanskritization, the coupled notions of purity and
impurity are important as they systematize and maintain the structural distance
between dierent castes (S , ).
Related to this Sanskritization are M’s () concepts of consensus and
replication. Moatt argues that the Dalit live in consensus with the wider Indian
culture by replicating among themselves almost every relationship from which
they have been excluded. Though Moatt’s consensus and replication theory has
provoked criticism from historical and epistemological points of view (for example,
S , –), his ethnography warrants reevaluation from a new analytical
perspective. Drawing on his ethnographic data, he gives the example of the annual
festival of the territorial goddess Mariyamman in a Paraiyan or “Harijan” colony in
Endavur village, Chingleput district, Tamil Nadu.
Moatt explains that as a Dalit group, the Harijans are excluded from the ūr
(main caste hamlet) and thus from the higher-caste cult of the ūr goddess Mari-
yamman. In response, the Harijans replicated for themselves an identical cult and
a separate image of Mariyamman inside the colony. The colony Mariyamman even
has the same powers as the ūr Mariyamman: to guard the boundaries of her terri-
tory, to protect those inside, and to bring rain (M , –).
The bodily form of the goddess has a Brahman’s head and a Dalit’s body. In her
worship in the colony, as in the ūr, she is alternately present in her low and high
forms; that is, as dangerous, impure, and bloodthirsty, or benign, pure, and tranquil.
 | Asian Ethnology / 
In the colony festival, the goddess is dealt with first in her low, fierce, and impure
form, mediated by the pūjāris (temple servants), and then later in her higher forms,
mediated by the higher-ranking Valluvar purohit (temple priest). She is trans-
formed from a low, angry, and “hot” being into a higher, beneficent, and “cool
being who provides rain and protection to the colony (M , –).
Moatt argues that the Mariyamman festival shows the forms of social-structural
replication by the Endavur Dalit. By
replicating the ūr territorial cult from which
they are excluded, “the Harijans have recreated for themselves the single specific
divine being from who the higher castes have cut them o. This deity’s powers are
thus available to them” (M , ).5
In this way, ritual practices of the low castes have been interpreted as indications
of their internalization of caste hierarchical values that distinguish highness and
lowness, purity and impurity, and superordination and subordination among the
higher castes. On the other hand, religious practices by the low castes have also
often been interpreted as resistance to the ideology of the dominant castes.
U (), for example, illustrates how the ritual practices con-
ducted by the Dalit Parayas and Pulayas of Nagarajanadu, Alleppey district,
Kerala, work as resistance against the dominance of the high castes. The social
structure of Nagarajanadu is expressed spatially with the houses of the higher
castes such as the Nayars in the higher central part of the village and the houses
of the Pulayas and Parayas closer to the punja (vast low-lying water-logged land),
which is seen by the high castes as dangerous and polluted. The ancestor spirits
of Dalits are also believed to be “sitting” in their lineage kāvus (groves) on the
margins of the village.
Though Dalits are socially subordinate to the Nayars, there is a complementary
relationship between the high caste in the center and the Dalit on the periph-
ery, in that the Nayars remove their “sins” and malevolence, and “gift” them to
the Dalits. On the other hand, paddy (oryza sativa) produced in the punja by
the Dalits is brought to the landlords’ houses in the center. Through this process
of mutual gift giving, of distance pollution, and paddy production, the space of
Nagarajanadu is reconstructed into one of a pure, structured center with polluted
and less-structured margins (U , –).
As a result of land reforms in the s and s, the Dalits lost their ancestral
land, and many high castes came to own this marginal land. The Dalits, however,
attempted to retain their control by performing ancestor worship in their lineage
kāvus. By invoking ancestral spirits in the kāvus, they invoked the dominant sym-
bolic system’s caste-specific spaces over the disembedded post-land-reform space.
U (, –) argues that the performance of ancestor worship
thus works as resistance against the encroachment of higher castes into marginal
lands by recreating the symbolic boundaries that separate the pure center from the
impure margins.
: būta     | 
B     :
  
Uchiyamada interpreted Dalit ancestor worship as resistance to the
higher castes, while Moatt analyzed their Mariyamman worship as the replica-
tion of high-caste ritual practice. Although they reached contrastive conclusions,
similar structures can be abstracted from their ethnographic descriptions: these
rituals crystallize and recreate the interdependent, complementary relations of
opposites—center and periphery, inside and outside, highness and lowness, purity
and impurity, protection and danger, and structure and chaos.
In his reexamination of the Mariyamman festivals described by M ()
and B (), F () calls our attention to the relation of opposition
between the high and low forms of the goddess. He points out that rather than
being a cluster of contrasting attributes, the two contrastive forms of the god-
dess embody D’s (, –) principle of divinity as a relation between
ranked complementary opposites. Based on his own research at the Minaksi Tem-
ple in Madurai city, Fuller insists that while divinity within the pantheon of village
deities is always relational, for the Sanskritic deities relational divinity is largely
displaced by substantial, non-relational divinity.
The Sanskritic deities … symbolise a social order in which the keystone of
Dumont’s theory of caste, complementary hierarchical relationships, has van-
ished. These relationships instead organise, in counterpoint, the village deities,
who symbolise the caste system as it exists, albeit one-dimensionally. Further,
they also symbolise resistance to a Brahmanical pretence that the low castes have
no function in the world, by according pre-eminence to the hierarchical rela-
tionships which necessarily link the high and the low. (F , )
Here, Fuller seems to reach a similar conclusion to Uchiyamada: rituals work as
the resistance of the low castes against the high castes by recreating an interdepen-
dent caste relationship. Fuller’s insight into relational divinity is most significant
for the analysis of the low castes’ ritual practices. His conclusion that the village
deity symbolizes the resistance of the low castes to Brahmanical doctrine, however,
is insucient for two reasons. First, the complementary opposites are not only
symbolized statically in the figures of the village deities but also dynamically flow
through space and interact with each other. This can be seen, for example, in the
high and low forms of the incarnation of Mariyamman in Endavur or in the pure
paddy and destructive “gift” in Nagarajanadu. The “sacred geography” (U-
, ) is recreated through the physical interaction and circulation
in space of substances and actors with contrastive characters. Second, when Fuller
interprets Mariyamman worship as the “resistance” of the Dalit, he assumes that
the relationship between complementary opposites in rituals merely symbolizes
the caste relationship in the social-political sphere. Here, the whole meaning and
function of the ritual is reduced to a symbolic representation of unequal caste hier-
archy and power relations.
 | Asian Ethnology / 
It may be true that all rituals concern the hierarchy and power relations of the
people involved. Nevertheless, the rituals described by Moatt and Uchiyamada
concern not only caste hierarchy but also the creation of order in the world and
the cosmos by linking, while separating and circulating particular substances in
space (see D ; R  ). For instance, the Dalits in the Endavur Colony
must conduct the Mariyamman festival themselves in order to access the goddess’s
power to reproduce the boundaries, to repel epidemics, and to bring rainfall. In a
similar way, the Dalit in Nagarajanadu have to perform ancestor worship to recre-
ate the sacred geography consisting of the appropriate circulation of agricultural
products, sins, malevolence, and other substances.
In this sense, it is not sucient to interpret the rituals merely in terms of caste
hierarchy. Rather, they should be considered in terms of the cosmology, ecology,
and ontology of the people who perform them. It may be necessary to consider the
meaning of complementary opposites not merely from the viewpoint of unequal
caste hierarchy in mundane human society, but as the relation of mutually dieren-
tiated, contrastive substances and actors that constitute a fluid network in a social,
ecological, and cosmological space.6
Based on the above theoretical framework, I will examine būta worship in a vil-
lage called Perar in a suburb of Mangalore, Karnataka. I will first investigate the
historical background of būta worship in the village, then examine the process of
the legitimation of Brahmanical caste hierarchy through the village būta shrine in
the colonial period. Third, I will analyze būta worship in Perar from a relational
perspective, focusing on the role of the būtas’ impersonators as well as on the
interrelation between the highest-ranked būta and the other būtas in the annual
festival. Through these investigations, I will show how Brahmanical caste hierarchy
interrelates with, or is absorbed into, the network of substances and actors created
by būta worship.
T   
I conducted fieldwork in the two adjoining villages of Mudu Perar (East
Perar) and Padu Perar (West Perar) in Mangalore Taluk, Karnataka.7 In Perar,
thick forests and shrubby hills surround lowlands intersected by a river. Land in
Perar is classified into several categories according to its soil and humidity. Rice
and areca nuts are produced mainly in the wet lowlands, while several kinds of veg-
etables are produced in the dry highlands. Local manor houses and other landed
farmers’ houses are scattered throughout the extensive paddy fields, while most
wage laborers, who were once the domestic laborers of powerful guttu houses, live
in the highlands.
In addition to the paddy and other cultivated fields, the forests and hills called
guḍḍe are an important resource for the villagers’ lives. People often go into the
guḍḍe to hunt game or gather useful plants. Since most guḍḍe land is under the
control of local manor houses, a villager
who hunts game there shares part of
his bag with the owner. The guḍḍe is believed to be the dwelling not only of
: būta     | 
wild
animals but also of būtas and other spirits. In Perar, several nāga (cobra)
shrines called nāgabana are located inside the groves, and a shrine to Pilichamundi
(pilicāmuṇḍi [a tiger būta]) is located on top of a hill near the village būta shrine.
Because it is believed that various būtas of wild animals as well as other dangerous
spirits are wandering about in the guḍḍe, it is regarded by most villagers as a fertile
but hazardous place.
T   būta   
The territories of Perar, including living premises, cultivated fields,
guḍḍe, and wastelands, are deeply related to the būtas’ power, as illustrated in local
legends. The dana (oral epic) of Perar narrates the legend of Nadu, a tragic
hero who travelled across the country as a human being, and then, after his death,
was revived in a place called baṇṭakaba in Perar as a very powerful būta, Bala-
vandi (balavāṇḍi), the main deity of the village shrine. Balavandi and related būtas
such as Arasu, Pilichamundi, and Brammabermeru (brammabermer) are believed
to be the ultimate owners of Perar land. Thus they have the power to protect the
land as well as to authorize the guttus’ rights to their territory. The dana, out-
lined below, shows the “origin” of būta worship and the highest-ranked guttus’
control over village land.8
Balavandi was originally one of the followers of the supreme būta, Brammaber-
meru, in the spirit world. However, cursed by his master, Balavandi was sent to
this world as a human baby and named Nadu by his foster parents. After grow-
ing up, Nadu travelled across the country with the king of Mangara (present-
day Mangalore). On their way home, Nadu and the king met a Brahman who
suddenly stopped them, taking the reins of Nadu’s horse. The Brahman was
actually Brammabermeru, who had disguised himself as a human being. Nadu
and Brammabermeru glared at each other for a second, and then a great bat-
tle between them took place. As a result of this battle, Brammabermeru forced
Nadu to disappear and become a būta.
After being revived as a būta in Perar, Balavandi invited another būta called
Arasu, the “king” of the būtas, to enshrine himself there. Arasu agreed to come
to Perar on one condition: that Balavandi also invite Brammabermeru. Bala-
vandi accepted his request. The būta of a wild tiger called Pilichamundi joined
them after fighting with Balavandi to acquire the right to enter the village.
Balavandi and Arasu then disguised themselves as human beings and visited
several major houses in Perar. They visited a manor house called the Munda-
bettu guttu, owned by a Jain lady called Koratai Balardi, the leader of the village.
The deities requested that she build a būta shrine in exchange for the saving
grace of the būtas. She accepted their request and built a shrine in cooperation
with the heads of the other fifteen guttus in the village. The shrine consisted of
a a [tall būta shrine] for Arasu, a cāvai [būta shrine with an open hall in
front] for Balavandi, and a guṇḍa [sanctuary] for Brammabermeru. Balavandi
 | Asian Ethnology / 
hierarchically ordered the guttu houses, and the Mundabettu guttu was given
the highest position as well as control of the vast territory of Perar.
To fulfill Arasu’s request, Balavandi went to a place called Kaje to get a
brahmaliga,9 the incarnation of Brammabermeru. After taking the brahmaliga
to the village, Balavandi told all the heads of the guttus, “You people should
perform a daily pūjā for Brammabermeru, hold a dīpa pattuni [light] for us, and
conduct a nēma [festival] every year. If you do all these things, we great daivas
[royal būtas] will protect both you and your land.
Henceforth, the heads of the guttus and other villagers have conducted the
nēma in Perar.
The oral epic summarized above shows the ambiguous relationship between
Balavandi and Brammabermeru. Balavandi, once a follower of Brammabermeru,
was banished to the human world by his master’s curse and then killed by him to
become a spirit. Despite this antagonistic relationship with Brammabermeru, Bala-
vandi took the brahmaliga to Perar to install it in the village shrine.
The dana recounted above also shows the ambiguous character of Bramma-
bermeru himself. In the epic, he is the sovereign of all the būtas, while he appears
in this world in the form of a Brahman. Also, the brahmaliga is promised special
status in the village shrine and oered a daily pūjā, something usually only per-
formed by a Brahman priest for paurāik deities in Hindu temples. These aspects
suggest both the local and Sanskritic qualities of Brammabermeru.
In the village būta shrine today, we can observe the Sanskritic characteristic of
Brammabermeru, as described in the oral epic. For example, while a Bant priest
called a mukkāldi conducts rituals for three būtas (Balavandi, Arasu, and Pilicha-
mundi), a Brahman priest called an asraṇṇa performs rituals only for Bramma-
bermeru in the guṇḍa, which more closely resembles a temple than a būta shrine
in structure. While chanting a mantra, the asraṇṇa cleans the inside of the shrine,
bathes the brahmaliga, decorates it with garlands, and waves lamps in front of it.
When worshippers come, he gives them sandalwood paste or a little consecrated
water as prasāda (consecrated food oering). On the other hand, the cāvai for
Balavandi is usually closed and no daily pūjā is performed beyond an oering of light.
būta ,    ,
   
Būta worship in Perar is based on a sophisticated system called kaṭṭụ
(custom or law). The most privileged families in relation to būta worship are a
Brahman family called Pejattaya and the sixteen guttu families. These families are
hierarchically ranked from the Mundabettu guttu to the sixteenth guttu, Pereer.
Except for one Gowda family and three Billava families, all the other guttu families
are Bants. Each guttu family has various roles and duties to organize the rituals at
the village shrine. Among them, the first and the second guttu (the Mundabettu
and Branabettu) are most responsible for the patronage and management of būta
worship at the village level. The primary patron of the village shrine, the Munda-
: būta     | 
bettu guttu head, called the gaipattunār, has command over all the other guttu
members and ritual workers.
For fulfilling their duties in būta worship, the Pejattaya and sixteen guttus are
rewarded with various rights, privileges, and honors at the village shrine as well
as in village society. During the nēma, all the heads of the sixteen guttus stand in
front of a būta incarnated in a Pambada (būta impersonator), waiting their turn
to be called by family name by the deity. At the end of the ritual, they receive the
būta’s blessings according to their rank. These performances in the nēma show
and confirm their social status.
The ritual roles of these sixteen guttus are complemented by another set of sixteen
families called the ulaguttu (sub-guttu). Under these guttu and ulaguttu families,
dozens of people called cākiridakuru (servants/people in service) execute various
services for būta worship at the village level. These people are from particular fami-
lies of several service castes, for example, Madivara (washermen), Jogi (musicians),
Bandari (barbers), and Pambada (būta impersonators). Among them, one Pambada
family plays an especially important role in Perar būta worship. Its male members are
trained as the dancers and mediums of the daivas, or great būta
s.
Traditionally, each cākiridakuru family was granted a portion of tax-free land
called umbai from the Mundabettu guttu. Some of them settled there and those
 . The gaipattunār and cākiridakuru walking to the
village būta shrine,  August . All photos by Miho Ishii.
 | Asian Ethnology / 
lands were named after the owners’ families, such as pabadelɛi (Pambada’s
hilltop) or jōgilɛ bail (Jogi’s plain). Also, these cākiridakuru families enjoyed the
rights to a share of the paddy produced on particular plots of land called bākimāru.
These plots were the property of the village būta shrine and were managed mainly
by the head of the Mundabettu guttu. All ritual expenses and shrine worker
rewards used to be paid in the form of paddy produced on this land. Apart from
the cākiridakuru families, other families of various castes such as Billava, Achari
(carpenters), and Gowda (cultivators and cattle breeders) also enjoyed rights to
shares of the prasāda distributed during the nēma, in reward for their services at,
or oerings to, the būta shrine.
The system of būta worship in Perar described above can be interpreted as a
“system of entitlements” (T ) that existed in pre-colonial western and
southern India in various forms. In the pre-colonial system of entitlements, Tanabe
argues, members of a local community were granted various rights to shares of
local products and royal and/or community honors and privileges in exchange
for performing dierent duties and functions for the reproduction of the state and
community. In Perar, būta worship has been the core of the system of entitlements
and has thus formed the basis of social-economic relations in the village through
the redistribution of land, local products, honors, and prasāda. It can also be noted
that būta worship in Perar constitutes a “transactional network” (A and
B ) to which all the actors involved in this redistributive process
are related. On the redistributive processes and transactional networks in South
Indian temples, A and B (, ) write as follows:
At the normative level, the deity … commands resources (i.e., services and
goods) such as those which are necessary and appropriate for the support and
materialization of the ritual process described above. But these resources are not
merely authoritatively commanded and received by the deity. On receipt, they
are redistributed in the form of shares (paku) to the royal courtiers, the donor
(yajamāna), and worshippers at large. The authority to command and redistrib-
ute resources places the deity at the center of a transactional nexus in which the
deity is expected to be generous. Ritual which constitutes worship provides the
schematic and elementary unit in which to observe the transactional network
where first the deity and subsequently the donor are the object of gifting activity.
In Perar, the system of entitlements is constituted in, or embodied by, the
mutual gifting activity
between the būtas, as the ultimate owner of the land, and
people in rituals, creating a transactional network among them. The term “net-
work” is appropriate here because as S points out, the concept of a
network provides a way to “link or enumerate disparate entities without making
assumptions about level or hierarchy” (, ). Also, the concept allows us to
focus not only on rules or regularity but also on the dynamism of the flow of cer-
tain substances through the transactional process, which performatively creates
the nexus linking disparate entities (M ; S ; ;
see also G , –; S , –). In other words, it is not always
: būta     | 
the case that the hierarchical system regulates the flow of substances; the flow of
substances itself incessantly actualizes or (re)creates the relationship among the
actors participating in the transaction. As we will see, in the case of būta worship,
the actors who enter into the transactional network through the ritual process
are not only the people and the deity as a sovereign, special person (A
and B , ; A , –), but also various wild
animals such as tigers, wild boar, bualoes, and cobras that are worshipped as
būtas.
būta     
While emphasizing the importance of the system of entitlements in pre-
colonial West and South India, T also points out that this system broke
down with the advent of colonial administration and the new kind of monistic
caste hierarchy that emerged under colonialism, where the ritualistic Brahmanical
caste hierarchy matched the socioeconomic hierarchy (, –). In the case
of Perar, būta worship, which had formed the basis of the system of entitlements in
the village, was unsettled and partly transformed under the colonial situation. To
illustrate this, I will focus on the legal discourse’s legitimation of the Brahmanical
caste hierarchy in the village būta shrine.
The incident analyzed here is a legal dispute between an asraṇṇa named
L. Udupa and the heads of the Mundabettu and Branabettu guttus over the
trusteeship of the
village būta shrine. Before we analyze the details of this dispute,
it is necessary to outline the policy of the Madras government at that time on
temple administration.
The Madras Endowments and Escheats Regulation,  (Regulation  of
) was the first legislation on religious institutions in Madras and was super-
seded by the Religious Endowment Act of  (Act  of ). In , the
Hindu Religious Endowments Board (henceforth  Board) was formed and
the Hindu Religious Endowment Act (Madras Act II of , henceforth 
Act) was passed by the Madras Legislative Council. Through the foundation of
the  Board and the  Act, the state enhanced its administrative power over
local temples and gradually undermined the autonomy and traditional authority of
local temples (P , –). Under these circumstances, the trusteeship of
local religious institutions became a common subject of competition and dispute
(A , –; D ).
As we have seen, the village būta shrine in Perar consists of the a and the
cāvai for Arasu and Balavandi respectively, and the special sanctuary or guṇḍa
for Brammabermeru. The daily worship and rituals for Brammabermeru are con-
ducted only by an asraṇṇa, a Brahman priest who is hired by the Mundabettu
guttu for this purpose. However, in , L. Udupa, who was then the asraṇṇa
of the village būta shrine in Perar, filed a suit in the district court against the then
head of the Mundabettu guttu “for delivery of the properties appertaining to the
Padu Perar institution” (Original Suit [O. S.] No.  of , ).10
 | Asian Ethnology / 
This case was contested before the subordinate judge of South Kanara in ,
with L. Udupa as the plainti and the six main members of both the Mundabettu
and Branabettu guttus as the defendants. There were three main points of conten-
tion in the suit: first, which religious institution in Perar could be identified as the
one over which the asraṇṇa insisted on his trusteeship; second, who had received
the allowance (tasdik or tastik) granted by the government to the religious insti-
tution in question; and third, whether the institution in question was a “hereditary
temple” that could be “excepted” from the provisions of the  Act.11
According to the court record, the Padu Perar institution in question was reg-
istered in Mangalore Taluk in  as “Perar Shastavu Brahma Bhoota” (O. S. No.
 of , ). An asraṇṇa named A. Shibaraya was the then trustee (moktesser
or moktīsare) of the institution and he had an allowance of nine rupees from the
government. Since then, the person who became the asraṇṇa received the insti-
tution’s allowance. In addition, L. Udupa had been appointed by the Mangalore
Circle Temple Committee (later replaced by the South Kanara District Temple
Committee), which had been constituted under the  Act, as a trustee of an
institution in Padu Perar described as “Shastavu Brahma Balavandi Pilichamundi
Daivastanam.” Based on these facts, L. Udupa insisted on his trusteeship of the vil-
lage būta shrine, which he identified as “Shastavu Brahma Balavandi Pilichamundi
Daivastanam” of Padu Perar.
On the other hand, the guttu members insisted that the Padu Perar institu-
tion was known not as “Shastavu Brahma Balavandi Pilichamundi Daivastanam,
but as “Kinni Majlu Ishta Devata Balavandi Pilichamundi Daivastanam,” whose
management was traditionally vested in the hereditary rights of the Mundabettu
and Branabettu guttus. They also insisted that the institution was a “hereditary
temple” exempt from the provisions of the  Act; therefore, neither the Manga-
lore Circle Temple Committee nor the South Kanara District Temple Committee
had jurisdiction over it.
After a number of hearings and detailed investigations into exhibits from both
the plainti and defendants, the court finally settled the suit in September .
Among the total of fifty-seven exhibits, the one the judge regarded as the most
decisive was a report submitted to the Temple Committee in January  by
A. S. Pai and P. V. Rao, both advocates of the court and members of the Temple
Committee. According to the report, they inspected the Padu Perar institution in
question in the presence of the plainti (L. Udupa) and first defendant (J. Naik)
and held an enquiry. B. G. Avargal, the then Subordinate Judge of South Kanara,
stated the following about this report in the court record:
They observed in their report that the form and appearance of the building in
which the idol of Brahma was kept led them to conclude that the building was a
temple and not a Daivastanam [būta shrine] but that the “mada” of Ishta Devate
[another name of Arasu] and the “chavadi” of Balavandi within the same enclo-
sure had the appearance of Daivastanams…. They have further observed as a
result of their inspection and enquiry that the institution for which the Plain-
: būta     | 
ti was appointed trustee by the Mangalore Circle Temple Committee and for
which tasdik was paid by Government is the one in Padu Perar, that the principal
deity therein is Brahma and that the 3 daivas, viz., Ishta Devata, Balavandi and
Pilichamundi, are subsidiary deities. (O. S. No.  of , – [emphasis mine,
to illustrate the hierarchy of the deities presumed by the judges])
Based on the observations of two influential advocates, the judge concluded
that the institution for which L. Udupa was appointed as a trustee was identified
as the Padu Perar institution (that is, the village būta shrine), in which Bramma-
bermeru was the presiding deity and other būtas were attendant deities. The judge
also pointed out that the Padu Perar institution was not an “excepted temple,
since it had at least one oce of trustee that was not hereditary (O. S. No.  of
, –). The judge thus approved the Temple Committee’s right to exer-
cise jurisdiction over the institution and therefore decided that its appointment of
L. Udupa as a trustee was valid.
In this judgment, Brammabermeru was identified as a quasi-Hindu god
enshrined in a “temple” and was regarded as superior to the other būtas. In the
same manner, the judgment guaranteed the status of the asraṇṇa as the priest
of the main deity and a trustee of the institution. It can be noted that this judg-
ment was not only influenced by ideas of Brahmanical caste hierarchy held by the
judge and two advocates, but it also consequently legitimized the Brahmanical
caste hierarchy in the village būta shrine.
As we have seen, Brammabermeru has ambivalent characteristics both in the
dana narratives and in worship at the village būta shrine, for he is treated as a
local as well as a Sanskritic deity. In conjunction with legal discourse, which pre-
sumed the Brahmanical caste ideology, the Sanskritic aspect of Brammabermeru
became evidence of his supremacy and the existence of Brahmanical caste hierarchy
in the village būta shrine. As a result, at least at the level of legal discourse, the vil-
lage būta shrine in Perar became recognized as an institution where a quasi-Hindu
god is the supreme deity and a Brahman priest occupies a crucial status. Does this
mean, however, that the village būta shrine in Perar has been Sanskritized through
the penetration of legal discourse and Brahmanical caste ideology? In the next sec-
tion, I will consider this question by focusing on popular narratives and practices.
F    
Counter to the judgment of the subordinate judge who approved the
supremacy of Brammabermeru, the then heads of the Mundabettu and Branabettu
guttus contended that Brammabermeru was only a minor deity and that Balavandi
was the chief deity there (O. S. No.  of , ). As shown in their statement,
the devotees have through today generally regarded Balavandi as the main deity
in Perar, although they acknowledge Brammabermeru’s special position in the
būta shrine. In the same way, people continue to regard the Mundabettu guttu as
 | Asian Ethnology / 
principal in būta worship in Perar, though they recognize Brahmans such as the
Pejattaya and asraṇṇa as necessary for worship.
Contrary to the presupposition of the judicature in colonial South Kanara, nei-
ther the relationship between Balavandi and Brammabermeru nor the position of
the Brahmans in the shrine can be fully explained in terms of the Brahmanical
caste hierarchy based on a dichotomy between superior and inferior or purity and
impurity.12 In order to understand this from another viewpoint, we first need to
consider the coupled notions of purity and pollution (mai-mayilig
ԑ
or sudda-
asudda) in the context of būta worship. Dayananda Pambada, who is the imper-
sonator of Arasu and Pilichamundi, explained that the impersonator should obey
particular rules to invoke the būta in the ritual:
The būta performance is called nēma. This word originated from niyama [rules
and regulations], in Sanskrit. Only when the performer obeys the rules, will it be
successful. So, we should be in the condition of niyama niṣṭhɛ [devoting oneself
to the rule]. We should follow several ritual practices. For instance, I’m a strict
vegetarian and I never drink alcohol. If you obey these rules, daiva will defi-
nitely come to you. (Dayananda Pambada,  May )
Satish Pambada, the impersonator of Balavandi in Perar, explained the rules he
follows in a similar way:
We, as būta performers, should not eat food oered at a funeral. I’m strictly fol-
lowing this rule. Apart from that, we should not eat food prepared by a woman
who is menstruating. Also, we should not eat food in the houses of Achari,
Catholics, or Muslims. Though it is not easy for us to obey all the niyama today,
we try to be in a state of purity [sudda] as much as we can.
(Satish Pambada,  June )
As shown in these narratives, for both, the notions of niyama and sudda are
crucial to being a proper impersonator. At first glance, their keen concern for the
rules and purity seems to be evidence of the Sanskritization of the Pambada imper-
sonators; they seem to adopt both Sanskritic notions and the Brahmanical doctrine
to raise themselves up to a higher position in the caste hierarchy. However, the
notion of purity in the context of būta worship cannot be reduced to the dualistic
model in which purity is opposed to an impurity that endangers the former with
contamination. Rather, the notion of purity is closely connected with that of divine
power or śakti, which is transcendental and thus immune to pollution (T
, –, –; see also H ; F ). Dayananda Pambada’s
comments on the relationship between the purification of a būta impersonator and
ritual pollution provide a clue to understanding this point.
When a Pambada is selected as a būta impersonator, he should be purified
by a Brahman priest. This ritual is called kalaśa snāna. After this ritual, he
becomes immune to the pollution that occurs through either death or birth.
He becomes free from them. Even if his father dies, he is exempt from sūtaka
[deat
h-pollution]. (Dayananda Pambada,  June )
: būta     | 
Purified by a Brahman priest, a Pambada impersonator is enabled to access the
śakti of the būta, and thus to obtain sacredness, which frees him from ritual pol-
lution.13 In this sense, as F (, –) argues about rituals in a Hindu
temple, purity is not the end of the ritual but is only a means to access divine
power. Thus, the purity of a Pambada impersonator should be understood not as
evidence of Sanskritization but as a necessary condition for him to invoke śakti
within himself.
A number of fundamental characteristics of būtas are inextricably linked to
wildness and femininity, in a word, śakti. It is notable that in Tulu, śakti can refer
to power, the existence of the supernatural powers of the būta, or the būta itself
(U , : ). That the būtas are believed to be the spirits of local
heroes/heroines who met tragic deaths—or those of wild animals—contributes to
this sense of danger and feral power. This is further connected to a feminine aspect
of the būta, as in Perar, Balavandi is regarded as androgynous, and devotees oer
beautiful saris to the deity. As these elements show, the būtas embody the strong,
dangerous, and fertile śakti.14
If the relationship between Brammabermeru and Balavandi in the village shrine
should thus be understood not in terms of the dualistic principle of high-low or
purity-impurity, but in terms of the interrelation between purity and śakti, the
question to be answered next is why people need Brammabermeru to be enshrined
in a special position in the village būta shrine, despite their regard for Balavandi as
the principal deity. To investigate this, I will analyze the process of the nēma in the
village būta shrine.
T   śakti    
  
I investigate the dynamic circulation of divine power in the nēma in
Perar here, focusing on the interrelation between Balavandi and Brammaber-
meru.15 First, I examine the process of the confirmation or reproduction of rank
among the Pejattaya and sixteen guttus through the interaction between the būtas
and the heads of these families. Second, I investigate the creation of the transac-
tional network through the flow of various actors, substances, and śakti within and
beyond the ritual stage.
The nēma starts on the night of the full moon in the month of māi and is held
for three days and nights.16 It primarily consists of the rituals for Balavandi, Arasu,
and Pilichamundi. Each ritual consists of the same basic process. The first stage of
the ritual is called the gaggaradecci.17 The impersonator, wearing a heavy anklet
called a gaggara, stands in front of the altar, which is placed next to the guṇḍa,
on which the baṇḍāra (sacred treasure of the būtas) is enshrined. The body of the
impersonator begins to shake the moment the gaipattunār oers a prayer, and
the other guttu heads throw rice and flowers on him. The impersonator, possessed
by the būta, dances around the precinct and greets the Pejattaya and the guttu
heads one by one according to their rank.
 | Asian Ethnology / 
The second stage is the recitation of thedana by the impersonator in front
of the thousands of devotees thronging the shrine. In the third stage, called the
nēmadecci,18 the impersonator wears a big halo-like structure called an ai on his
back. The priests, heads of guttus, and main workers follow him, and together they
all march around the precinct. Then the possessed impersonator speaks to oracles
in front of all the guttus. He receives a young coconut from the gaipattunār,
pours its juice on the floor, and gives it back to the gaipattunār with blessings.
At the end of the ritual, the possessed impersonator touches the hands of each
guttu head with his sword and gives them blessings.
During the ritual, the heads of the main guttus communicate with the būta
through the Pambada impersonator. The most significant and repeated form of
their communication is the mutual gifting between the guttu heads and the būtas.
In the nēma, the guttus oer the būtas a part of their farm products, which are
regarded as being originally owned by the būtas. The būtas receive these oerings
and in return give them oracles and blessings to ensure the future prosperity of the
whole village. Finally, a part of these oerings are distributed as prasāda among
the heads of the guttus and other devotees, according to their ranks.
Through the above ritual process, the ranks of the Pejattaya and the main
guttus are not only confirmed but also constituted. All communication between
the guttu heads and the būtas, such as being called by their family names or being
given prasāda, happens according to their rank; and therefore it authorizes their
status in relation to the deity. In this sense, the ritual concerns the constitution of
the hierarchy and social-political power in the village society.19
At the same time, the mutual gifting activity between the būtas and the devotees
creates a transactional network to which all the actors involved in the ritual are linked,
and in which a finite set of substances and śakti flow and circulate. Or rather, the flow
of substances such as farm products and prasāda containing būtas’ sákti actualize the
nexus of various actors, of which social-political hierarchy is only one part
.
The contrastive positions of and interrelation between Brammabermeru and
the būtas Balavandi and Pilichamundi in the ritual performance oer a window
through which this point can be investigated further. In the nēma, all the main
deities except Brammabermeru always appear from outside the central shrine. For
example, the priests, accompanied by some of cākiridakuru, walk up to the shrine
on top of a hill and oer a pūjā to the deity Pilichamundi, as incarnated in the pos-
sessed Pambada impersonator, who comes down from the hilltop to the central
shrine. After the ritual for Pilichamundi inside the precinct is complete, the deity
is oered both vegetarian oerings and blood sacrifices right outside the shrine
building. In a similar way, Balavandi, incarnated in the Pambada impersonator,
also appears from outside the shrine as a half-naked, dangerous, and furious deity.
Contrary to these wild būtas, Brammabermeru is neither incarnated in an
impersonator nor appears outside the guṇḍa. The brahmaliga, the only mani-
festation of the deity, is permanently fixed inside the guṇḍa, which is regarded as
the sanctum sanctorum of the būta shrine. During the nēma, the asraṇṇa and sev-
: būta     | 
eral other Brahman priests from neighboring villages hired especially for the nēma
oer a pūjā to the deity.
The contrastive relation between Brammabermeru and Balavandi is vividly
crystallized in the last part of the ritual for Balavandi. The possessed imperson-
ator of Balavandi, wearing an ai and riding on a beautifully decorated wheeled
wooden horse, parades around the altar with the guttu heads, priests, and many
cākiridakuru. He stops just in front of the guṇḍa and starts a “battle” against
Brammabermeru, who remains inside the guṇḍa. The impersonator violently
swings his arms, pulls the mustache o his face, and throws it into the guṇḍa,
along with pieces of garlands pulled from his neck.
A
ccording to an informant from the Mundabettu guttu, this performance repre-
sents the mythical battle between Balavandi and Brammabermeru in the dana.
Apart from this, it also clearly shows the contrastive yet complementary relation-
ship between the two deities; namely, the wild, dangerous, and a
ndrogynous deity
 . Balavandi at the nēma in the village būta shrine,  March .
 | Asian Ethnology / 
Balavandi contacts and at the same time separates himself from Brammabermeru,
who is incarnated in the pure, still, masculine figure of the brahmaliga.
Through this ritual process, the śakti of Balavandi embodied in the possessed
impersonator flows from the outside into the central shrine, circulates within it,
and finally reaches its sanctum sanctorum. The dangerous, wild śakti of Balavandi
received in the guṇḍa is then transformed into a calmer, more controllable form
and finally distributed to the devotees as prasāda from Brammabermeru.20
This suggests why people need Brammabermeru to be enshrined in the special
position in the village būta shrine, despite their regarding Balavandi as the princi-
pal deity. The position of Brammabermeru is critical in the shrine, as is the trans-
actional network created by the ritual, as the “polar opposition” (F ,
) to Balavandi, as well as the device for receiving and transforming the wild śakti
into grace for the people. In this sense, the divinity of Brammabermeru is certainly
relational, as D (, –) discerns in his general thesis on Hindu divin-
ity. However, contrary to Dumont’s view, the essence of the relationship between
Brammabermeru and Balavandi cannot be fully understood through the dualistic
principles of superior and inferior statuses or purity and impurity. Rather, it must be
comprehended as a complementary relationship between purity and śakti. Here the
purity of Brammabermeru can be understood in the same way as that of the Pam-
bada impersonators: it is the necessary condition for gaining access to and receiv-
ing the būtas’ śakti, which transcends the binary opposition of purity-impurity.21
The findings of this study thus urge us to reconsider not only the Dumon-
tian interpretation, but also Fuller’s view that the “little” village deities symbolize
the caste system by linking themselves to the “great” deities through comple-
mentary hierarchical relationships (F, ; , ). According to
Fuller, worship of the village deities legitimates in religious terms the caste hierar-
chy whose summit is occupied by Brahmans. While Fuller denies the sociological
reductionism of, for example, D and P (), he also considers the
complementary relationship between the contrastive deities only in terms of the
Brahmanical caste hierarchy, which presumes the purity-impurity dichotomy and
the supremacy of Brahmans.
To the task of finding an alternative interpretation to those based on Brah-
manical caste hierarchy, the notion of a transactional network is critical. As we have
seen, the mutual gifting activity between the būtas and the devotees, as well as
the interaction between the contrastive deities, not only creates but actualizes the
transactional network to which all the actors involved in the ritual are linked, and
in which particular substances and śakti flow, circulate, and are distributed.
The actors entering into this network are not only people and deities as sover-
eign persons, but also various wild animals and spirits from the forests and moun-
tains that are worshipped as būtas. Through their entering into and acting in the
network as būtas, these wild animals and spirits, which represent the guḍḍe or wild
nature and embody its śakti, are personified as social actors who can communicate
and interact with human beings from within their mutually dierentiated, relative,
and often contrastive positions in the network.
: būta     | 
Thus the transactional network created by the būta ritual constitutes and medi-
ates not only the social, economic, and religious relationships among the people,
but also the complementary relationships among human beings, wild creatures,
and spirits, all of who are personified through interaction in the network. Here the
social-political hierarchy and authority of the people, which is confirmed through
the ritual process, becomes only a part of the nexus of substances and actors that
constitute the fluid network within and beyond the ritual stage.
C
Ritual practices of the low castes have often been considered through
concepts such as Sanskritization, consensus, and replication, or interpreted as resis-
tance against the dominance of the high castes. The tendency common to these
analyses is their interpretation of low castes’ ritual practices in terms of caste hier-
archy and power relations. In such analyses, the significance of the ritual is reduced
to a representation of the unequal caste relationship in the social-political sphere.
Contrary to the hierarchy-centered analysis of previous studies, this study has
attempted to provide a new perspective for understanding the ritual practices of
the non-Brahman, low-caste people not simply in terms of caste hierarchy, but of
a network and its relations. Focusing on the relational aspect of divinity and the
importance of śakti in ritual contexts, this study has oered an alternative per-
spective from which to see the complementary opposites in rituals not merely as
reflections of unequal caste relations, but as the basis of the transactional network
among the actors and substances linked to it.
Būta worship in Perar has been at the center of the system of entitlements
through which local products, land, and prasāda have been distributed among
the people in exchange for their services at the village būta shrine. In pre-colonial
Perar, this distribution process was handled by the Mundabettu guttu. However,
this system was partly undermined in the colonial situation through the penetra-
tion of its modern administration and jurisdiction, which presumed a Brahmanical
caste hierarchy and the supremacy of Sanskritic deities. As a result of a legal dispute
between the local asraṇṇa and the Mundabettu guttu, the village būta shrine was
ocially reorganized into a religious institution in which Brammabermeru, as a
quasi-Hindu deity, was given the supreme position and the Brahman priest came
to occupy a central role.
At first glance, this event and the people’s apparent admission of the Brahman-
ical doctrine and the caste hierarchy based on its dichotomies seem to provide
evidence of the Sanskritization of the būta worship in Perar. However, a close
investigation of popular narratives and ritual practices reveals that būta worship
can be fully interpreted by neither the notion of Sanskritization nor the Brahmani-
cal dualistic principle of purity-impurity. Rather, it should be understood as the
dynamic relationship between purity and śakti.
As seen in this article, the interrelations between Balavandi and Brammabermeru
show that their divinities are complementary and relational, as wild sacredness and
 | Asian Ethnology / 
its recipients. The relationship between these contrastive deities constitutes the
basis of the flow and distribution of śakti through the transactional network to
which are linked not only human persons and deities as sovereign persons but also
wild creatures and spirits personified as būtas. Thus, the transactional network not
only constitutes social-political relations among people but also creates and vital-
izes human-nature-spirit relations, in which the social-political is but a part of the
broader network.
As shown in the case of būta worship, it is insucient to analyze the ritual prac-
tices of the low-castes only in terms of the unequal caste relations in mundane
human society. Rather, it is necessary to broaden our analytical scope to see their
ritual practices both from within and beyond the social-political sphere. The rituals
then reveal themselves as the constitutive parts of the transactional network, which
links humans to deities, spirits, and wild creatures through the dynamic flow and
interaction of the mutually dierentiated substances, actors, and sacred power.
N
.
The native language of the region is Tulu. While bhūta means “ghost” in Sanskrit, in
South Kanara the term būta is generally used for local deities. This article follows U’s
(–) system of transliteration for all local terms.
.
The traditional occupation of the Billava caste is toddy-tapping and that of the Bant caste
is cultivation. While the Bants are regarded as the “dominant” caste in the area, most of the caste
groups in my research field are designated as “Other Backward Classes” in Karnataka State.
. On būta worship in general, see C (), G (), and B ().
. For example, see H () and U  (; ). This tendency cor-
responds to many anthropological works that analyze the practices of the marginalized as
“weapons of the weak” (S ) or “ritualized resistance” (C , ).
. On the mimetic ritual practices of Dalits, see also S ().
. On the fluid transactions of substances in Hindu societies, see M (), M-
 and I (), and D ().
. These two villages comprised a single village called Perar until they were administra-
tively separated in . The fieldwork on which this article is based was conducted from
May to September ; in March, August, and September ; and from December 
to January .
. The dana in this article is reconstituted by the author based on interviews with
Gangadara Rai, Baaleekrishna Shetty, and Dayananda Pambada. The interviews were con-
ducted on – July  and  August .
.
The brahmaliga is a small stone with a smooth surface that is worshipped as the sym-
bol of Brammabermeru.
. Because the village būta shrine was located in Padu Perar, the institution was called
“the Padu Perar institution” in the court record.
. Under the  Act, “hereditary” temples (that is, temples whose managers had not
previously been selected by government ocers) were seen as private institutions and were
thus relatively free from any direct outside control. The “excepted” temple category was
abolished in  (P , ).
: būta     | 
. This corresponds to Dumonts view of caste society in terms of the hierarchical ideol-
ogy of purity and impurity as the encompassing ideology of Hindu hierarchical society (see
S , ).
. This is the case in purification rituals in Sri Lankan Tamil society presented by T
(, ). According to Tanaka, priests wearing a sacred kāppu cord containing mantra
śakti become immune to death pollution. For the experiences of būta impersonators in ritual
practices, see I ().
. On śakti and femininity, see F (, –) and U  (, ).
On śakti’s relationship to tragic death and wildness, see U (, –).
. On the circulation of śakti, see M (, ) and T (, ).
. The month of māi in the Tulu calendar corresponds to  February to  March in the
solar calendar.
. Gaggaradecci is the initial dance performed by the impersonator wearing sacred anklets
(U , : ).
. This word originated from the phrase nēmada ecci, which refers to the shivering of the
būta impersonator’s body during the annual festival (U , : ).
. For the constitutive features of the redistributed prasāda of the deity, see A
(, ).
. T (, ) describes a similar process of the transformation of śakti in Hindu
rituals.
. On the dynamic relationship between two contrastive deities, see also S (,
).
R
A, Arjun

Worship and Conflict under Colonial Rule: A South Indian Case. Cambridge:
Cambridge University Press
.
A, Arjun, and Carol Appadurai B
 The South Indian temple: Authority, honour and redistribution. Contribu-
tions to Indian Sociology (N.S.) : –.
B, Brenda E. F.
 The goddess and the demon: A local South Indian festival and its wider
context. Puruārtha : –.
B, Heidrun
 On an Auspicious Day, at Dawn... Studies in Tulu Culture and Oral Litera-
ture. Wiesbaden: Harrassowitz Verlag.
C, Peter J.
 Spirit possession and spirit mediumship from the perspective of Tulu oral
traditions. Culture, Medicine and Psychiatry : –.
C, Jean
 Body of Power, Spirit of Resistance: The Culture and History of a South
African People. Chicago: The University of Chicago Press.
D, Valentine E.
 Fluid Signs: Being a Person the Tamil Way. Berkeley, CA: University of
California Press.
 | Asian Ethnology / 
D, Veena
 Language of sacrifice. Man (N.S.) : –.
D, Nicholas B.
 The Hollow Crown: Ethnohistory of an Indian Kingdom. Cambridge:
Cambridge University Press.
D, Louis

Religion/Politics and History in India. Paris and The Hague: Mouton
Publishers.
 Homo Hierarchicus: The Caste System and Its Implications. Chicago: The
University of Chicago Press (complete revised English edition).
D, Louis, and D. F. P
 Pure and impure. Contributions to Indian Sociology : –.
F, C. J.
 Gods, priests and purity: On the relation between Hinduism and the caste
system. Man (N.S.) : –.
 The Hindu pantheon and the legitimation of hierarchy. Man (N.S.) :
–.
 The Camphor Flame: Popular Hinduism and Society in India. Princeton and
Oxford: Princeton University Press. (revised and expanded edition)
G, Alfred
 Art and Agency: An Anthropological Theory. Oxford: Oxford University
Press.
G, K. Chinnappa
 The Mask and the Message. Mangalagangothri: Madipu Prakashana.
H, Edward B.
 Spirit possession and social structure. In Anthropology on the March: Recent
Studies of Indian Beliefs, Attitudes and Social Institutions, ed. B. Ratnam,
–. Madras: Book Center.
 Ritual pollution as an integrator of caste and religion. The Journal of Asian
Studies : –.
I, Miho
 Playing with perspectives: Spirit possession, mimesis, and permeability in
the Buuta ritual in South India. Journal of the Royal Anthropological Insti-
tute : –.
M, McKim

Hindu transactions: Diversity without dualism. In Transaction and Meaning:
Directions in the Anthropology of Exchange and Symbolic Behavior, ed. Bruce
Kapferer, –. Phil
adelphia: Institute for the Study of Human Issues.
M, McKim, and Ronald I
 Toward an ethnosociology of South Asian caste systems. In The New Wind:
Changing Identities in South Asia, ed. Kenneth David, –. Paris and
the Hague: Mouton Publishers.
M, Michael
 An Untouchable Community in South India: Structure and Consensus.
Princeton: Princeton University Press.
: būta     | 
N  , A. V., and Denis F, eds.
 The Devil Worship of the Tuluvas: From the Paper of the Late A. C. Burnell.
Mangalore: Karnataka Tulu Sahitya Academy.
P, Franklin A.
 Religion under Bureaucracy: Policy and Administration for Hindu Temples
in South India. Cambridge: Cambridge University Press.
R, Gloria Goodwin
 The Poison in the Gift: Ritual, Prestation, and the Dominant Caste in a
North Indian Village. Chicago: The University of Chicago Press.
S, William S.
 God of Justice: Ritual Healing and Social Justice in the Central Himalayas.
Oxford: Oxford University Press.
S, James C.
 Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven,
CT: Yale University Press.
S, Yasumasa
 Anthropology of Untouchability: “Impurity” and “Pollution” in a South
Indian Society. Osaka: National Museum of Ethnology.
S, M. N.
 Religion and Society among the Coorgs of South India. Oxford: Oxford
University Press.
S, Marilyn
 The Gender of the Gift: Problems with Women and Problems with Society in
Melanesia. Berkeley, CA: University of California Press.
 Cutting the network. The Journal of the Royal Anthropological Institute :
–.
T, Akio
 Recast(e)ing identity: Transformation of inter-caste relationships in postco-
lonial rural Orissa. Modern Asian Studies : –.
T, Masakazu

Patrons, Devotees and Goddesses: Ritual and Power among the Tamil Fisher-
men of Sri Lanka. New Delhi: Manohar Publishers
.
T, Edgar
 Castes and Tribes of Southern India. Vol. v. Delhi: Cosmo Publications.
(Originally published )
U, Yasushi
 Soil, self, resistance: Late-modernity and locative spirit possession in Kerala.
Puruārtha : –.
 Passions in the landscape: Ancestor spirits and land reforms in Kerala, India.
South Asia Research : –.
U, Uliyar Padmanabha, ed.
– Tulu Lexicon. Six vols. Udupi: Rashtrakavi Govind Pai Samshodhana
Kendra.
... According to Tanaka, priests wearing a sacred kāppu cord containing mantra śakti become immune to death pollution. 12 On the complementary relationship among Hindu deities, see also Fuller (1988) and Ishii (2015a). 1 For recent studies focusing on animism, in addition to the works of Viveiros de Castro and Willerslev examined in this chapter, see, for instance, Descola (1996), Bird-David (1999), and Pedersen (2001). 2 As mentioned in Chapter 1, recent studies on magical-religious phenomena, in cluding spirit possession, tend to analyse them not as a mere critique of moder nity, but as components of it (e.g. Behrend & Luig 1999;Comaroff & Comaroff 2002). ...
... This fact points to a Dravidian religious system quite different from the Brahminical one of North India (Claus 1978a&b). Even the Dravidian notion of purity is quite different from the Brahminical notion (Ishii 2015). 8 Remember John Rawls' (1971) 'original position' which forces the reflecting subject behind the 'veil of ignorance' to recon with the possibility to find himself in the position of the worst of society. ...
Article
Full-text available
All over India there are peculiar examples of dispute resolution where the responsibility for justice is completely delegated to spirits such as būtas or daivas, devtas or jinns. This paper will show that a spirit-based justice system such as this is not necessarily tied to romantic and orientalising ideas of ancient society. When the paper compares a spirit-based justice system with a modern state-based legal system, it treats both as coeval. The paper is thereby able to explore the implications of such a comparison for modern Political and Legal Philosophy. One of these implications is that it puts a question mark on the modern claim to an ethical neutrality of a law-based justice system. Both, the law-based justice system and the justice system based on the belief in spirits, are relying on a particular ethical background consensus–one spiritual, one secular. For the spirit-based system this poses no problem as long as lifeworlds continue to be enchanted and communicative action is not cut loose from ties of sacred authorities, or released from the bonds of archaic institutions. Moreover, the continued existence of spirit-based justice systems in much of the post-colonial world falsifies the modernist view that only small and relatively undifferentiated groups are able to integrate by regulating behaviour through archaic religious institutions like spirits. With the failure of modern law to make up for the legitimation crisis of liberal-democratic societies, postcolonial legality with its archaic institutions, "legal diversity" , and paradigms of "porous legalities" may offer new avenues of participation which were not conceivable within the blinds of the modern episteme. These may not be able to displace the normalised paradigms of modern, secular law-based regimes in practice, but they do shake up their resting position within liberal political theory. The paper is looking at Indian examples of a widespread phenomenon: the self-regulation of communities by way of a spirit-based justice system. Although such a system could in the legal pluralism approach be described as a system of law, there are good methodological reasons not to do so. However, the spirit-based system and the law-based system reconnect on the level of principles of justice. On this level, the spirit-based system appears as an alternative to systems based purely on law, whether customary or statutory. After a few methodological considerations, I will argue that these spirits are efficacious institutions of justice. To argue thus, I claim that we can leave aside the question whether these spirits are 'real' in the objective sense. It suffices to show that their existence is an inter-subjectively shared reality for the believers. Such beliefs, as part of the social imaginary, are efficacious just like any other social institution whose reality is grounded in the intersubjective "social world" rather than in the "objective world" of science or the "subjective world" of experience. The paper compares the two systems―one spirit-based; one based on statutory law―on the level of ideal-types. I thereby avoid the asymmetrical comparison of an ideal system on one hand with an operational system on the other. When comparing both systems on the level of ideas, it is apparent that one of them suffers from a major paradox which the other does not have to grapple with. The paradox afflicting the state-sponsored legal regime can only be resolved by resorting to the stipulation of a moral consensus, which is precisely what is said to be lacking under modern conditions. While a system based on state-sponsored law purports to be able to do without appeals to an integrated morality of its citizens, the spirit-based system is said to be less suited to modern conditions of plurality precisely because it rests on the assumption that everyone shares the same belief. One of the staple claims of modernist theories of law is that modern society cannot be integrated or stabilised based on spiritual beliefs that not everyone shares. While this may be true for Western societies, the claim seems less obvious in the case of societies like India where, in spite of their (different) modernity , people continue to believe in the force of supernatural beings like spirits and deities. To the modernist, 'premodern' societies (with the supposedly more uniform religious beliefs of their members) serve as examples of the allegedly lost unity of ancient times in which a self-standing conception of justice was not yet required.This paper will show that the belief in the spirits' efficaciousness as mediators of justice is shared by members of all religious denominations in spite of their internal differences.
Chapter
This essay explores how transnational migrants from a regional land-owning caste have retained and reinvented political roles in their hometowns and agrarian villages in coastal Karnataka, South India. Successful migrants have gained in wealth and prestige, translating their earlier position as landlords into ideals of leadership and entrepreneurship. They reconfigure political influence by engaging in and patronizing agrarian rituals, maintaining political connections, and refashioning themselves as development agents and local leaders. Based on ethnographic research in India and the UAE, the chapter outlines the regional and historical configurations of these connections and transformations.
Article
This paper investigates spirit (būta) worship in a special economic zone (SEZ) in India by considering practices of care around specific constellations of nature and infrastructure: fluid, contingent assemblages of the ‘natural’ and ‘artificial’ environments. Occult phenomena in modern settings have often been interpreted as metaphorical critiques of modernity by neophyte proletarians. In the SEZ, however, it is not workers but executives who undertake the primary role in būta rituals. In addition, the rituals’ main aim is towards not division but connection among modern technology, nature, and divinities. The SEZ management assumes the role of primary caretaker of an assemblage which constitutes both industrial plants and spiritual landmarks. The rituals enable the people to manage the entanglements of infrastructure with spirits and nature, which are not only modern but also untamed and divine. In the process of caring for these entanglements, people experiment with novel ontological arrangements.
Article
This book deals with ritual healing in the Central Himalayas of north India. It focuses on the cult of Bhairav, a local deity who is associated with the lowest castes, the so-called Dalits, who are frequently victims of social injustice. When powerless people are exploited or abused and have nowhere else to go, they often turn to Bhairav for justice, and he afflicts their oppressors with disease and misfortune. In order to end their suffering, they must make amends with their former victims and worship Bhairav with bloody sacrifices. Many acts of perceived injustice occur within the family, so that much of the book focuses on the tension between the high moral value placed on family unity on the one hand, and the inevitable conflicts within it on the other. Such conflicts can lead to ghost possession, cursing, and other forms of black magic, all of which are vividly described. The book includes a personal account of the author's own experiences in the field as well as descriptions of blood sacrifice, possession, exorcism, and cursing. The book begins with a straightforward description of the author' s fieldwork and goes on to describe the god Bhairav and his relationship to the weak and powerless. Subsequent chapters deal with the lives of local oracles and healers; the main rituals of the cult and the dramatic Himalayan landscape in which they are embedded; the moral, ritual, and therapeutic centrality of the family; the importance of ghosts and exorcism; and practices of cursing and counter-cursing. The final chapter examines the problematic relationship between ritual healing and modernity.
Article
Data collected in a major Hindu temple in South India show that the religion of the Hindu gods cannot be explained in terms of the social (or caste) order and must first be understood in its own right. The gods, whose crucial attribute is great power, are worshipped by Brahman priests. The entire temple environment should be kept very pure, and purity and pollution define and idiom by which priests (and devotees), the gods' servants, show them respect. The gods themselves cannot be polluted. Food-offerings, like other rituals, please the gods when made respectfully, but their pattern does not replicate inter-caste food exchange. The competence of Brahman priests, ranked below non-priestly Brahmans, is primarily defined within a religious tradition, rather than by caste rank. Purity and pollution embody more than high and low status, and have ethical and spiritual aspects, expressing the proper relationship between men and gods. In sum, the relation between man and god is not a simple transposition of hierarchical relations between men.
Article
Foreword 1. The Problem Defined: The Need for an Anthropology of Art 2. The Theory of the Art Nexus 3. The Art Nexus and the Index 4. The Involution of the Index in the Art Nexus 5. The Origination of the Index 6. The Critique of the Index 7. The Distributed Person 8. Style and Culture 9. Conclusion: The Extended Mind Bibliography Index
Article
New technologies have stimulated the rehearsal of old debates about what is new and what is old in descriptions of social life. This article considers some of the current uses to which the concepts of `hybrids' and `networks' are being put. It could be seen as following Latour's call for a symmetrical anthropology that gathers together modern and nonmodern forms of knowledge. In the process, the article reflects on the power of analytical narratives to extend endlessly, and on the interesting place that property ownership holds in a world that sometimes appears limitless.
Article
This article examines the interpretation of Vedic sacrifice in the Mimamsa school of ancient India. The sacrificial cult moves simultaneously on the axes of individual desire and cosmic order. The cosmos is recreated through sacrifice by both men and gods in common submission to the Vedic World. The sacrificial act par excellence is defined as that in which the person appears in an adjectival capacity in relation to the act. It is through the Logos that Vedic sacrifice seeks to bridge the discontinuity between individual desire and cosmic order, and between emotivism and submission to a moral imperative. It thereby unites modes of thought and codes of conduct, thought to be contradictory in Western conceptions of the sociology of religion.