Professional Documents
Culture Documents
Series Editors
VOLUME 155
Edited by
István Keul
LEIDEN | BOSTON
Cover Illustration: Hanuman image awaiting consecration (Banaras 1995, photo by István Keul)
Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface.
issn 0169-8834
isbn 978-90-04-31370-5 (hardback)
isbn 978-90-04-33718-3 (e-book)
Notes on Contributors vii
Index 391
Notes on Contributors
Marie-Luce Barazer-Billoret
is Senior lecturer (Maître de conferences) in Sanskrit and Indian religions at
the Université Sorbonne nouvelle—Paris 3. Her main field of research is ritual
in the Śaivasiddhānta. She has published critical editions and translations of
Śaiva treatises in the French Institute of Pondicherry Indological Series.
Marzenna Czerniak-Drożdżowicz
is professor in the Department of Languages and Cultures of India and South
Asia at the Institute of Oriental Studies, Jagiellonian University, Cracow.
Her main fields of interest are the Tantric traditions of India, especially the
Vaiṣṇava Pāñcarātra, its Sanskrit literature and its present-day religious centers
in South India, as well as the religious art of South India.
Ronald Davidson
received his PhD in Indian Buddhist Studies from the University of California,
Berkeley. He is Professor and Chair of Religious Studies, and the Director of the
Humanities Institute at Fairfield University in Connecticut. His primary inter-
ests include the social history of tantric Buddhism, the ritual systems of tantra
and dhāraṇī literature of the late Mahāyāna, and the sociology of knowledge in
both Gupta-Vākāṭaka and medieval India.
Shingo Einoo
received his PhD from Marburg University and has been professor of Sanskrit
Philology at the Institute for Advanced Studies on Asia, University of Tokyo
since 1991. He was Lecturer at Kyushu Tokai University and Associate Professor
at the National Museum of Ethnology. His main area of interest is the forma-
tion and development of ritual from the time of the latest Vedic literature until
the present.
Marko Geslani
completed his PhD in Religious Studies (Asian Religions) at Yale University
and is currently an assistant professor in the Department of Religion at Emory
University. His research centers on the early history of Hindu ritual, with a con-
cern for the enduring role of ritual in the study of religion. Current projects
investigate the ritual bases of image worship and temple infrastructure, and
the ritualization of astronomy/astrology in early India.
viii NOTES ON CONTRIBUTORS
Dominic Goodall
joined the École française d’Extrême-Orient (EFEO) in 2000 and, apart from
four years in Paris between 2011 and 2015, when he gave lectures at the École
pratique des hautes études on Cambodian inscriptions and on the history of
Śaivism, he has spent his career at the EFEO’s Pondicherry Centre. Among his
publications are editions and translations of works of poetry in Sanskrit and of
Śaiva scriptures and treatises, often in collaboration with colleagues in various
universities.
Ellen Gough
received her PhD in Asian Religions from Yale University and is currently an
assistant professor in the Department of Religion at Emory University. Her
work focuses on the religious traditions of India, in particular Jainism. At
present, her research centers on Jain Tantra, with other teaching and research
interests including ritual theory, Jain and Hindu festivals, astrology, medita-
tion, yoga, and material culture.
István Keul
is professor in the Study of Religions at the Department of Archaeology,
History, Cultural Studies and Religion (AHKR) at the University of Bergen. His
main areas of research include various aspects of the history and sociology of
religions in South Asia. He is currently engaged in comparative work on mod-
ern religious movements and is involved in a research project on the socio-
cultural dynamics of religious spaces in urban India.
Elisabeth Raddock
received her PhD in Sanskrit from the Department of South and Southeast
Asian Studies at the University of California, Berkeley in 2011. She is assistant
professor at the Department of Historical, Philosophical and Religious Studies,
Umeå University (Sweden) where she teaches courses on Hinduism and
Buddhism. She directs the program for teacher education in Religious Studies
within the School of Education at Umeå University.
S.A.S. Sarma
received his PhD from the University of Calicut and has been a researcher at
the Pondicherry Centre of the École française d’Extrême-Orient (EFEO) since
1989. After completing his post-graduation in Sanskrit from the University of
Kerala, S.A.S. Sarma joined the Adyar Library and Research Centre as a research
scholar. He is presently engaged in various projects at the EFEO, some of which
concentrate on the edition of Śaiva texts.
NOTES ON CONTRIBUTORS ix
Anna A. Ślączka
obtained her PhD in Indology from Leiden University, where she subsequently
worked. She was an associated fellow at the International Institute for Asian
Studies (IIAS) in Leiden and the EFEO in Pondicherry. Her research covers
Hindu ritual and art, and she is currently involved in a project on production
and casting technology of Chola bronzes. Since 2009 she has been the curator
of South Asian art at the Rijksmuseum, Amsterdam.
Annette Wilke
is professor in the Study of Religion and Head of Department at the University
of Münster. Besides systematic studies in religion and aesthetics (religion
and the senses), ritual studies, and mysticism, her major fields of research
are Hindu traditions: Advaita Vedanta, Tantra, sonic aspects of Hindu culture,
reform Hinduism, and Tamil Hindu temples in Germany. She is currently work-
ing on aesthetic communication and in a research project on Global Hinduism.
CHAPTER 1
Consecration rituals are key prerequisites for the building of temples and the
worship of temple images in South Asia. These complex rituals accompany
the construction process of sacral buildings, as well as the consecration and
installation of temple statues. In the latter case, which is the overarching
theme of the present volume, man-made sculptures are ritually transformed
into (containers of) deities by the rite of infusion with life, prāṇapratiṣṭhā or
simply pratiṣṭhā.1 The pratiṣṭhā belongs to a wide category of practices classi-
fied under the umbrella-term pūjā, ritual activities performed in connection
with the worship of iconic or aniconic ritual images in temples and domes-
tic contexts. These practices can be subdivided into several subcategories, the
most common of these being regular (nitya) rituals such as, for instance, the
worship offered on a daily basis by religious specialists and devotees alike at
public shrines and temples, or by laypersons in the privacy of their homes.
Rituals connected to particular occasions (naimittika) comprise another sub-
category. These, too, can be performed either regularly (for example as yearly
temple celebrations) or in connection with specific occasions. Image conse-
crations are exactly such occasional rituals that often constitute unique events
in the lives of their sponsors and are performed in connection with the con-
struction or renovation of shrines and temples.
The gradual transition from Vedic religion to a more personal and image-
centered religiosity that started in the last centuries BCE led to the develop-
ment of new rituals in both public and domestic contexts. The performance
of image consecrations in private settings is described in late (fourth–fifth
century) Gṛhya texts,2 and Varāhamihīra’s Bṛhatsaṃhitā (sixth century) pres-
ents the public installation (saṃsthāpana or sthāpana) of temple images.
Scholars such as Einoo and Takashima (2005) have suggested a development
of the pratiṣṭhā from the simpler, personal form to a much more complex (and
1 The pratiṣṭhā has received quite substantial scholarly attention over the years. A selection
of relevant works, both for image and temple consecrations, is as follows: Gonda 1954/75;
Gombrich 1966; Smith 1984; Bühnemann 1991; Barazer-Billoret 1993–4; Bentor 1996; Brunner-
Lachaux 1998; Einoo and Takashima 2005; Ślączka 2007; Colas 1989 and 2010.
2 See Einoo 2005: 13, 95–113.
3 ‘[I]nstallation as a social event had an import far beyond the image as a focus of devotion and
worship, for it involved such factors as the spread of religious communities of temple priests,
the religious donation and administration of lands, or the regulation of religious expansion
through the agency of economic and political powers’ (Colas 2010: 319).
4 See Kane 1941: 896–916 and Brunner-Lachaux 1998: vii–viii on the various types of
consecration.
An introduction 3
out that pratiṣṭhā (which he translates as ‘ground,’ ‘basis,’ ‘support’) was one of
the words used to convey the idea of a firm, stable, durable ground for humans
and the entire earth, a support for all individual and universal existence.
Interestingly, however, in spite of its apparent etymological unambiguousness,
the term and its morphological derivatives embarked on rather unexpected
semantic/metaphorical trajectories:
brahman, of the eternal dharma and of absolute bliss. In the medieval Śaiva
Bṛhajjābāla Upaniṣad, pratiṣṭhā is an ontological state associated with one of
Śiva’s five forms (Vāmadeva) and one of the five elements (water).6 Arriving
at the meaning that is particularly relevant for the purposes of this collection,
namely pratiṣṭhā as consecration, Gonda writes:
6 ‘It is a state of being like the supreme knowledge [vidyā] and the supreme “peace” [śāntyatītā],
it is an aspect of the final goal, the prospect of which is held out to the god’s devotees. It is
also systematically connected with the aspects of prosperity and ever-yielding abundance
and with the ritual means of attaining the god’s bliss’ (Gonda 1954/75: 370).
7 On pratiṣṭhā in early tantric works such as the Prapañcasāra and Śāradātilaka, the various
locations for infusing life in tantric ritual systems, and the personification/deification of
prāṇa as the goddess Prāṇaśakti, see Bühnemann 1991.
8 The word is also applied in connection with towns (establishment, founding), kingship
(establishment, accession), or with the meaning of ‘having found a firm resting place’ in
the sense of happiness or prosperity (a semantic variant found in the Mahābhārata), not to
mention other meanings in the areas of law, medicine, and warfare (372–3). Colas (2010: 321)
mentions the word’s ‘political shade of meaning’ as evidenced in Viśākhadatta’s seventh- to
eighth-century (or possibly earlier) play Mudrārākṣasa.
An introduction 5
9 Buddhist relics (śarīras, dhātus) are described in a great variety of Buddhist sources as
infused with morality and wisdom, and full of virtue (Schopen 1998: 261). On the instal-
lation and cult of Buddhist scriptures see chapter 2 in Schopen 2005. The discussion of
pratiṣṭhā in early Buddhism is based on Colas 2010.
10 See the discussions in Eck 1981; Fuller 1992; Davis 1997.
6 Keul
11 Davis 1997: 479. In addition to the Advaita Vedānta position, Davis offers in the first chap-
ter of his book an overview of some of the other influential Indian perspectives on the
worship of images over the centuries.
12 Examples may be found in Smith 1984: 54–63; Hikita 2005: 146–85; Takashima 2005:
116–25.
13 See von Rospatt’s (2010) discussion of the consecration ceremony in the eleventh- to
twelfth-century Kriyāsaṃgrahapañjikā, its employment of saṃskāras and its relation to
later rituals in Newar Buddhism.
An introduction 7
1 The Essays
mixtures, cow products, and various types of clay. In addition, Geslani points
out further parallels and similarities, such as the fire offering (homa) and the
observation of fire omens in both contexts, or the laying of the image on a bed/
pillow in the pratiṣṭhā and the dream divination in the royal yātrā rites. In a
second step, the author looks at possible Vedic antecedents for the pratiṣṭhā
ritual described in the Bṛhatsaṃhitā. The main focus of his inquiry lies on late-
Vedic śānti rites, and therein especially on the close connection between the
homa and the bathing sequences. According to Geslani, the yātrā and pratiṣṭhā
rituals described by Varāhamihira may have their origin in Atharvan śānti cer-
emonies. The sixth-century author, however, modifies the Vedic ritual struc-
ture decisively in several places. Perhaps the most important of these are the
alterations in the puṣyasnāna rite, but also in the pratiṣṭhā, his preference for
the pūjā, and his allowing for the agency of the deities, to the detriment of the
fire sacrifice and Vedic mantras. In the last part of his essay, Marko Geslani
shows how elements of Vedic ritual (mantras, homa, sacrificial dregs, bathing
techniques) as well as astrological motifs are preserved in later (late-Vedic and
Puranic) installation ceremonies.
Dominic Goodall discusses in his essay the changes undergone by the
liṅgapratiṣṭhā in the Mantramārga tradition by looking at the Niśvāsatattva
saṃhitā’s account of pratiṣṭhā and the differences in later works. The exami-
nation of the liṅga installation in the Niśvāsatattvasaṃhitā, one of the oldest
surviving works of Tantric Śaivism (c. 450–550 AD), reveals the lack of the
defining moments of pratiṣṭhā known from later accounts of similar rites,
such as the one provided in the eleventh-century Somaśambhupaddhati or in
South Indian temple āgamas from the twelfth century. In the Niśvāsatattva
saṃhitā’s account there is no instilment with life force, be it as jīvanyāsa
(‘imposition of spirit’) or prāṇapratiṣṭhā (‘installation of life-breath’), nor is
there an equivalent to the opening of the image’s eyes (nayanonmīlana) or
the showing of auspicious objects to the liṅga. Also, another moment charac-
terized as defining for the pratiṣṭhā in analyses of the Somaśambhupaddhati,
namely the (sexually connotated) lowering-down of the piṇḍikā or yoni over
the liṅga, is missing in the Niśvāsatattvasaṃhitā, in which there seems to
be no distinction made between the basis of the liṅga (pīṭha) and the yoni.
In later liṅgapratiṣṭhā accounts, the tripartite nature of the liṅga finds its
expression in the assigning of distinct mantras onto its three parts, and in the
division into three sections of other objects to be installed. Goodall points
to other developments as well, such as the changing descriptions of the ritual
specialists involved in the pratiṣṭhā, or the growing importance over the cen-
turies of the rites connected to the vāstumaṇḍala. To illustrate the differences
between early and later pratiṣṭhā texts, the author offers summaries of two
An introduction 9
deity into the image, at the beginning of the daily rites. For the Śrīvaiṣṇavas,
arcāvatāra was a central concept through which the fully-fledged presence
of the deity, not only in temple statues but also in movable images, was
established.
Elisabeth Raddock examines in her essay the first chapters of the eighth-
to ninth-century Hayaśīrṣa Pāñcarātra, another Vaiṣṇava Pāñcarātra text that
deals with both the building of temples and the production of temple sculp-
tures. The main focus of the text, however, is on ritual, only providing infor-
mation on architectural aspects when these are ritually significant. Raddock
discusses the various possibilities to categorize the text, one such possibil-
ity being to ascribe it to a group of texts that meet both the criteria of ritual
manuals and texts on architecture. Another question concerns the addressee
of these texts, with multiple (non-exclusive) options: the sponsor of the ritual,
the head priest, or other ritual specialists. The Hayaśīrṣa Pāñcarātra lists the
key actors in a temple-building project, giving a detailed description of the
necessary qualities that a potential supervising ritual specialist, ācārya, has to
fulfill. There are also passages in the text that prescribe the qualifications (and
undesired qualities) of the main artisan/architect (sthāpaka) in the temple-
building project.
Damaged images in temples have to be de-installed and replaced by new
ones. S.A.S. Sarma discusses in his contribution ritual texts from Kerala that
deal with jīrṇoddhāra, the renovation procedures for replacing damaged
images, including liṅgas. According to the various texts, these procedures are
to be performed when a temple is in a state of decay, the image is broken, has
the wrong measurements, is hollow, has been removed from its pedestal by
thieves, or has been washed away by a river. In certain cases, sculptures can be
repaired, using gold or the mixture of five metals (pañcaloha). In present-day
temples, the permission of the deity is required before initiating the replace-
ment process of the damaged image. On the basis of astrological calculations
(devapraśna, ‘the questioning of the deity’) the temple management decides
whether the image is to be replaced or not. Sarma describes a three-day ritual
comprising the extracting of life from the damaged image, the uninstalling of
this image from its pedestal, as well as the installation procedures for the new
image. Ritual texts also offer solutions for unforeseen events during the instal-
lation procedures, such as, for example, the breaking of the kalaśa. In that case,
the remains of the water from the pot are used together with ritually purified
water to fill the new pot. In exceptional situations, de-installation rituals were
also performed for sculptures that were not damaged but were in an immi-
nent danger of being destroyed. The author relates the case of the image from
12 Keul
the Guruvayur temple, whose ‘life force’ was ritually transferred from the main
image to the temple’s processional image and transported to the Ambalapuzha
temple during the armed conflict between Tipu Sultan of Mysore and Zamorin,
the ruler of Malabar. The movable image was later brought back to Guruvayur
and the main image was re-installed.
In his article, Shingo Einoo compares the installation of images with the
dedication of a garden and the planting of a tree. The latter rites belong to the
larger category of utsarga, the ritual dedication of certain objects, sites, etc.
for public use. The term pratiṣṭhā has been used in a number of texts to des-
ignate such ceremonies as well, where the central objects of the rituals were
not images of deities, as in the dedication of gardens, and the consecration of
pastureland, wells, and water tanks. The religious rewards for the planting of
trees are—according to the texts—numerous and comparable to some extent
to the merits of a pratiṣṭhā: destruction of evil; attaining heaven or the world
of Viṣṇu; becoming Brahmā, Viṣṇu, or Gaṇeśa. In some contexts, the merit
resulting from the planting of bushes and fruit trees or of an entire garden
is seen as equivalent to the merit acquired by performing śrauta rituals, such
as the godāna, agniṣṭoma, or even aśvamedha, respectively. Usually the texts
emphasize the usefulness of the dedicated objects in addition to the merits.
Einoo points out that one of the common underlying religious motivations for
performing utsarga rituals was connected to the śrāddha, ancestor worship,
and exemplifies this idea by discussing text passages related to vṛṣotsarga (the
setting free of a bull) and taḍāgapratiṣṭhā (the inauguration of a water tank).
The main differences between the pratiṣṭhā and utsarga rituals consist in the
meritoriousness of the dedicated object. According to the texts, the objects of
utsarga only give religious merit if they are useful for others. And, while image
consecrations and the building of temples produce merit for the sponsor and
provide often long-term opportunities of employment for priests, this is not
the case with utsarga rituals.
In her contribution, Ellen Gough looks at the historical development of
the sūrimantra in Jainism, an element of the eye-opening sequence in a Kānjī
Svāmī Panth installation ceremony that the author documented in 2008. Even
if the followers of this twentieth-century branch of the Digambara tradition
reject mendicancy, during this sequence the lay ritual specialist removed his
clothes before whispering the sūrimantra into the image’s ear, imitating thus
a Digambara monk. Through the examination of various medieval sources,
Gough shows how this central element in the installation developed between
the eighth and eleventh centuries and is the result of the interaction between
Digambaras and Śvetāmbaras and of the combination of their respective prac-
tices. The requirement that the performance of pratiṣṭhā should be undertaken
An introduction 13
and its main priest, before giving an account of the temple’s re-consecration in
2014, as well as an attempt at a prognosis regarding the future trajectory of the
temple and community.
The last essay in the volume deals with the contemporary practice of
pratiṣṭhā. In my own contribution I discuss the three-day consecration of a
Hanumān image, documented in the mid-1990s in Varanasi. After addressing
the issue of performative efficacy in rituals, and highlighting some of the more
important preliminary stages of the consecration, I offer a detailed description
of the ritual activities on each of the three days. In order to illustrate the com-
plexity and various morphological levels of the consecration procedures, both
the numerous auxiliary as well as the core parts of the ritual are given equal
attention. A team of five ritual specialists was in charge of the event, led by an
experienced ācārya, who was also the family priest of the temple owner, the
‘sacrificer,’ or yajamāna in the consecration. The latter’s relatives and friends
witnessed or participated in large parts of the ritual, thus transforming the
consecration into a social event. Based on the empirical material, in the final
section of the essay I reflect on the sequentiality and the constituent units of
the consecration, and the ‘ritual-grammatical’ rules according to which the
various components are connected together.
Acknowledgements
References
Barazer-Billoret, Marie-Luce. 1993–4. ‘L’installation des liṅga et images dans les tem-
ples selon les āgama śivaïtes.’ Bulletin d’Etudes Indiennes 11–12, 39–69.
Bentor, Yael. 1996. Consecration of Images and Stupas in Indo-Tibetan Tantric Buddhism.
Leiden: Brill.
An introduction 15
Smith, H. Daniel. 1984. ‘Pratiṣṭhā.’ In Āgama and Śilpa, ed. K.K.A. Venkatachari.
Bombay: Ananthacharya Indological Research Institute, 50–68.
Takashima, Jun. 2005. ‘Pratiṣṭhā in the Śaiva Āgamas.’ In From Material to Deity: Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New Delhi: Manohar,
115–4.
von Rospatt, Alexander. 2010. ‘Remarks on the Consecration Ceremony in Kuladatta’s
Kriyāsaṃgrahapañjikā and its Development in Newar Buddhism.’ In Hindu and
Buddhist Initiations in India and Nepal, eds Astrid Zotter and Christoph Zotter.
Wiesbaden: Harrassowitz, 197–260.
CHAPTER 2
It is well-known that the earliest firmly datable ritual instruction for the
Hindu image installation ceremony, or pratiṣṭhā, is to be found in chap-
ter 59 of Varāhamihira’s 6th century omen compendium, the Bṛhatsaṃhitā
(BS). Varāhamihira describes his brief text as a ‘general summary’ of the
installation,2 which, in his view, could be performed by a number of sectar-
ian groups, including Vaiṣṇavas, Śaivas, Śāktas, Sauras, Buddhists, and Jains.3
Accepting the possibility that numerous, widely divergent structures for this
rite may have existed in Varāhamihira’s era,4 I take the view here that his text
1 In addition to the NTNU conference on consecration rituals in October 2012, parts of this
paper were presented at the annual meeting of the American Oriental Society in Boston,
April 2012.
2 sāmanyam idaṃ samāsato lokānāṃ hitadaṃ mayā kṛtam |
adhivāsanasanniveśane sāvitre pṛthag eva vistarāt || 22 || BS 59 ||
‘I have made this general [instruction], summarily, for the good of all. The Adhivāsana
and the Sanniveśana (i.e. pratiṣṭhā) [are explained] separately and extensively in the Sāvitra.’
Unfortunately, the Sāvitra is no longer extant. The commentator Bhaṭṭotpala, who elsewhere
quotes extensively from Varāhamihira’s other works, provides little information about this
text beyond its name, describing it simply as a ‘treatise belonging to the Sun-god’ (saure
śāstre).
3 The various image-worshipping sects are mentioned in an oft-cited verse, BS 59.19. As
Varāhamihira says at the end of this verse, ‘the ritual should be performed for whichever god
[is to be installed] by the devotees of that god, according to their own procedure’ (ye yaṃ
devam upāśritāḥ svavidhinā tair tasya karyā kriyā | 19d | BS 59 |).
4 Interestingly, Bhaṭṭotpala adds in his gloss of BS 59.22 (above n. 2) that Varāhamihira has
written this brief text according to the ‘Vedic procedure’ (vaidikena vidhānena), whereas in
the Sāvitra, the adhivāsana and sanniveśana are explained separately for ‘each god’ (pra-
tyekasya devasya). It is possible that Bhaṭṭotpala understands his phrase vaidikena vidhānena
in contrast with the term svavidhinā, which Varāhamihira earlier used (59.19, previous note)
to describe the image installation procedures of the various sects. This raises the possibil-
ity that in Varāhamihira’s era a number of different procedures—besides his simple ‘Vedic’
structure--may have existed for different sectarian groups. Nonetheless, most subsequent
versions of the pratiṣṭhā ceremony do appear to have maintained the ‘Vedic’ procedure out-
lined in BS 59, despite adding a number of additional ritual elements to this structure.
5 Masahide Mori takes a similar view of BS 59 (2005: 200–201).
image installation as apotropaic consecration 19
The term ‘pratiṣṭhā,’ as has been well understood, may be translated as a ‘firm
establishment or support’ (Gonda 1975), and, as the name of a particular rit-
ual, refers to the physical act of installation: placing the image on a base or
plinth within the temple sanctuary.6 But while not immediately evident in this
titular notion of physical installation, pratiṣṭhā variants across Vedic, Purāṇic
and Āgamic texts recommend another ceremony called the ‘adhivāsana’ as a
necessary preliminary to the physical installation. The precise referent of this
term is less certain. In some versions of the ceremony it seems to refer to the
laying of the image in water overnight—a process often termed ‘jalādhivāsa.’7
Yet the BS, along with many late-Vedic and Purāṇic versions of the pratiṣṭhā,
omits the jalādhivāsa altogether, and instead understands the adhivāsana as
the aspersion of the image in a ‘bathing’ hut, or maṇḍapa—usually at least one
evening prior to the installation proper. Often, the image is laid on a bed or pil-
low, suggesting that it has been put to sleep for the night. In these versions of
the ceremony we might translate adhivāsana as the ‘overnight’ or ‘sleep-over
ritual’ (from the causative of √vas, ‘to dwell’).
It is this preliminary adhivāsana portion of the pratiṣṭhā that contains the
majority of ritual activity in Varāhamihira’s account, and which will concern
us here. A summary of this ritual text is given in Appendix 1 (with sanskrit
text in the notes for reference). The basic outlines of this ritual will be familiar
from standard accounts of the pratiṣṭhā.8 It includes a number of details
6 Other more interpretive and theological views of pratiṣṭhā can be had in Gonda (1975: 370),
Smith (1984: 50), and Brunner (1998: iv). See Hikita (2005: 143) for a summary. Interpretations
of the ceremony may alternatively emphasize the priestly view (empowering an object/
image), the devotional (recognizing an object of devotion, consecrating it as fit for worship),
or the theological (the god establishes him/herself in the object). An additional, symbolic
interpretation sees the placement of the image on its base as the sexual union between Śiva
and Śakti, or between Viṣṇu and Lakṣmī. All of these interpretations may of course apply
simultaneously with the more literal sense of a physical emplacement. Varāhamihira himself
employs the terms saṃsthāpana (59.15) and sanniveśana (59.22) as synonyms. Thus while
‘consecration’ is often used to translate pratiṣṭhā I will follow Einoo (2005) and Colas (1989,
2010) here in preferring the term ‘installation,’ in order to emphasize the physical act at the
heart of the ceremony.
7 For instance, the Vaikhānasagṛhyasūtra 2.10 employs the verb adhivāsayati to describe the
laying of the image in water over night, prior to its aspersion and installation. According
to Smith (1984: 54–56), the procedure of a Pāñcarātra text, the Kapiñjalasaṃhitā, includes
jalādhivāsa as the second of sixteen necessary steps in the correct performance of pratiṣṭhā.
8 Compare the summaries of pratiṣṭhā rites by Einoo (2005) (for the gṛhyapariśiṣṭas),
Takashima (2005) (śaiva āgamas), and Hikita (2005) (purāṇas).
20 Geslani
which are not found in earlier Vedic royal ceremonies such as the rājasūya or
aindra mahābhiṣeka, for instance, the setting of the ritual in the maṇḍapa
pavilion (step 1), the use of clay as a bathing ingredient (step 3), the observation
of fire omens (step 4), and the laying of the image on a bed (step 7). In order to
explain some of these novel ritual elements, I will argue that the adhivāsana
portion of the pratiṣṭhā redeploys three important elements found elsewhere
in Varāhamihira’s textual corpus, and which may derive ultimately from the
ritual sequence preceding the king’s yātrā or military expedition. Here I focus
in particular on the bath with plant infusions and clay (step 3), the homa
and observation of fire omens (step 4), and the laying of the image to sleep
(steps 6–7). By basing our analysis on the single ritual corpus of Varāhamihira,
which places special emphasis on the king, we may be able to propose a
different trajectory for the formation of pratiṣṭhā than might be accom-
plished by focusing on the ritual cycle of divine images alone, in Purāṇic or
Āgamic sources.
plakṣāśvatthodumbaraśirīṣavaṭasambhavaiḥ
kaṣāyajalaiḥ maṅgalyasaṃjñitābhiḥ sarvauṣadhibhiḥ kuśādyābhiḥ || 8 ||
dvipavṛṣabhoddhataparvatavalmīkasaritsamāgamataṭeṣu |
padmasaraḥsu ca mṛdbhiḥ sapañcagavyaiś ca tīrthajalaiḥ || 9 ||
pūrvaśiraskāṃ snātāṃ suvarṇaratnāmbubhiś ca sasugandhaiḥ |
nānātūryaninādaiḥ puṇyāhair vedanirghoṣaiḥ || 10 || BS 59 ||
It (the image) is bathed head to the East, with waters infused with plakṣa,
aśvattha, udumbara, śirīṣa and vaṭa [leaves]; with all herbs having auspi-
cious names, such as kuśa; with clay dug up by elephants and bulls, from
mountains, anthills, and the banks of confluent rivers; with clay found in
lotus ponds, with pañcagavya and with tīrtha water, with waters having
gold and jewels, and with perfumed waters—all accompanied by the
sounds of various drums, puṇyāha, and the sounds of the Vedas.
image installation as apotropaic consecration 21
nyagrodhaśirīṣāśvatthapatragandhāś ca kṛttikāsnāne |
bahubījapraśastatoyair jayārthino rohiṇīsnānam || 2 ||
muktākāñcanamaṇisaṃyuktena ambhasā mṛgāṅkarkṣe |
raudre vacāśvagandhāpriyaṃgumiśrair jalaiḥ kathitam || 3 ||
āditye gomayagoṣṭhamṛdbhir atha gauraśālibhiḥ puṣye |
siddhārthasahasradvayapriyaṃgumadayantikābhiś ca || 4 || YY 7 ||
For the bath under the Kṛttikās, the leaves and [?] scents of nyagrodha,
śirīṣa, aśvattha. Under Rohiṇī the bath for one seeking victory is done
with waters mixed with many excellent seeds. Under Mṛgaśiras, with
water containing pearls, gold, and jewels; in Ārdrā, it is recommended
with waters mixed with vacā, aśvagandhā, and priyaṅgu. Under Punarvasu
with cow’s dung and clay from a cow stall; under Puṣya with yellow
rice grains, two thousand yellow mustard seeds, and priyaṅgu and
madayantikā.
In this way the text recommends a different set of materials for bath-
ing under each of the twenty-seven nakṣatras. Each list is not entirely
unique; a number of species (for example priyaṅgu and madayantikā) occur
under more than one nakṣatra. Furthermore, while plant-derived ingre-
dients appear most frequently, ghee and honey are sometimes added to
the mixtures, while some baths are performed with clay, and one nakṣatra
(Mṛgaśiras, as seen in the quotation above) requires a mixture of pearls, gold
and jewels.
In the second part of YY 7, following this variable arboreal bath, the king is
to be ‘cleansed’ (śodhayet) with seven types of clay, gathered for instance from
a mountain peak, ant-hills, or river banks, or dug up with an elephant tusk and
bull’s horn. We may refer to it as the ‘alluvial’ bath:
22 Geslani
girivalmīkanadīmukhakūladvayaśakrapādamṛdbhir ataḥ |
dvipavṛṣaviṣāṇapārthivagaṇikādvārāhṛtābhiś ca || 13 ||
giriśikharān mūrdhānaṃ valmīkamṛdā ca śodhayet karṇau |
nadyubhayakūlasaṃgamamṛdbhiḥ prakṣālayet pārśve || 14 ||
indrasthānād grīvāṃ bāhū karivṛṣabhayor viṣāṇāt |
hṛdayaṃ ca nṛpadvārāt kaṭim api veśyāghṛhadvārāt || 15 || YY 7 ||
Next, [he does the following] with clay from mountains, anthills, both
banks of a river-mouth, the foot of Indra’s [banner], and with clays dug
with a elephant tusk, cow’s horn and with clay from the king’s door and
the door of a courtesan. He should purify the head [of the king] with the
clay from a mountain peak; the ears with the clay from an ant-hill. He
should clean the sides [of his body] with the clay collected from both
banks of a river at its mouth; the throat with the clay from Indra’s banner;
the arms with the clay from the tusk (or horn) of the elephant and bull;
the heart with clay from the king’s gate, and the hips with clay from the
gate of the courtesans’ dwelling.
9 For this text I have generally followed the edition of Devīprasāda Laṃsāla (1969), though
I have also occasionally referred to Pingree’s (1972) edition.
10 ‘The main stars beginning in the east are Puṣya, Hasta, Śravaṇa and Śraviṣṭhā, and also
Aśvinī in the north. Some say that Naiśakara (Mṛgaśiras), Tvāṣṭra (Citrā), Anurādhā, and
Pauṣṇa (Revatī) are second best’ (puṣyo ‘tha hastaḥ śravaṇaḥ śraviṣṭhā prācyādimukhyāny
udag aśvayuk ca | naiśākaraṃ tvāṣṭram athānurādhā pauṣṇaṃ ca madhyāni tathāhur eke
|| BY 4.4 ||). BY 4.21–25 [20–23 of Pingree ed.] provide bathing substances for all of these
nakṣatras, except for Citrā.
image installation as apotropaic consecration 23
priyaṅgusiddharthakanāgadānagorocanākṣaudraghṛtaiḥ sametaiḥ |
prāg ātmarakṣā praticakrapūtaiḥ snānonmukhasya āvanipasya kāryā || 8 ||
BY 19 ||
First the protection of the king seeking the bath is performed with
priyaṅgu, siddhārthaka, nagadāna, gorocana and ghee, [all of] which
have been purified by praticakra [mantra].
The seven types of clay mentioned here are likely the same as those given
above in YY 7, which, as we have seen, were used to purify different parts of
the king’s body. Here they are combined additionally with three cow products
(milk, curd and ghee) as well as an herbal, perfurmed conoction.
Despite their variations, Varāhamihira’s two yātrā texts present a fairly con-
sistent set of bathing practices, which combine a variable number of plant-
infused concoctions with a single ‘cleansing’ bath with clays collected from
various locations. These practices are deeply tied to the astrological process
of the yātrā, since the variable nakṣatra bath is determined by the specific
asterism (and the direction) under which the king is to depart. Meanwhile the
entire bathing process seems, according to its namesake, to assure the king’s
victory during the impending military campaign.
More importantly, this evidence raises a number of compelling similari-
ties between the pre-departure bath of the king and the complex bath of the
image in the BS 59. First, the nakṣatra-related ‘arboreal’ bath may have sup-
plied the plant materials used in the adhivāsana. In fact, the list of various fig
trees in BS 59.8 (plakṣa, aśvattha, udumbara, śirīṣa and vaṭa) resembles the
list of trees recommended for the first asterism, Kṛttikā, in YY 7.2 (nyagrodha,
śirīṣa, aśvattha). Second, the clay bath prescribed in YY 7.13–15/BY 19.10, closely
resembles the use of clay in BS 59.9. Four of the types of clay used for the king
(from a mountain top, anthill, bull and elephant horns, and river banks) are
included in the bath of the image.12 Third, BY 19.10 mentions other materials
included in the bath of the image at BS 59.9–10: milk, curd, clarified butter
(three of the six elements included in pañcagavya),13 and water containing all
herbs (sarvauśadhi) and perfume. Thus it seems plausible that the complex
bath of the adhivāsana derives from the Victory Bath of the yātrā ceremonies.
12 B S 59 includes clay from lotus ponds, whereas YY 7 adds clay from the king’s gate, a cour-
tesan’s dwelling, and Indra’s banner. Clay from a lotus pond or tank is mentioned for the
nakṣatra bath under Hasta (YY 7.6/BY 4.22 [20 in Pingree]).
13 On the pañcagavya, see Einoo 2005: 106–108. We may also note that at YY 7.4, cow’s dung
is included under the substances required for the bath under Punarvasu.
image installation as apotropaic consecration 25
A Brahmin should make an offering in the fire with mantras related to the
god who is to be installed. The fire-omens have been explained by me in
[the chapter on] raising Indra’s banner.
Here Varāhamihira references BS 42, the chapter on the festival of Indra’s ban-
ner (indradhvaja). Like the image installation ceremony, that ritual also pre-
scribes a fire offering, after which the omens of the fire are to be observed.14
The relevant verse of the indradhvaja reads:
A fire that has the shape of a desirable object,15 which is fragrant, soft,
thick and full of flames is auspicious. Otherwise it is inauspicious. [This]
has been described in detail in the Yātrā.
that Varāhamihira adapted this motif from the yātrā texts to a number of other
royal rituals in the BS, including the pratiṣṭhā, where it occurs in service of
the image.
The sthāpaka should lay the image on a well-strewn bed. Having kept the
sleeping image [there] over [night?] with [attendants] who stay awake
singing and dancing,16 he should fix it in place (i.e. complete the physical
installation) at a time indicated by the astrologer.
A similar overnight structure recurs in a number of other rites in the BS, most
prominently, in the puṣyasnāna the “Bath of Prosperity” (BS 47). Just like the
pratiṣṭhā, this ritual begins on the previous day, when the purohita, astrologer
and ministers depart the city in the eastern or northern direction to prepare
the ritual space, and to invoke the deities who will consecrate the king on the
following day. After performing these ritual duties, this group is advised to
spend the night at the ritual space:
Having performed the pūjā for those gods who have been invoked, they
should stay [there] that evening. [In order to examine] the signs of the
dream, whether good or ill, the instruction for the dreams has been men-
tioned in the Yātrā.
Note that the verb used for this direction is again √vas (in the optative
vaseyur)—the same verb at the root of the noun adhivāsana. Yet unlike in the
adhivāsana, in the case of the puṣyasnāna Varāhamihira indicates the purpose
of the overnight stay, saying, ‘[In order to examine] the dream omen, whether
good or bad, the instruction for dreams has been mentioned in the Yātrā’
(sadasatsvapnanimittam yātrāyāṃ svapnavidhir uktaḥ).17 This passage refers
to the ritual sequence known as the svapnanimitta, the observation of onei-
ric signs in order to determine the likely success of the impending undertak-
ing. As the commentator Bhaṭṭotpala notes, the full instruction for this ritual
sequence is to be found in the eighteenth chapter of the BY.18
When we turn to that text, we find again that the paradigmatic subject of
the dream divination sequence is in fact the king, who, prior to his ceremonial
departure for the military expedition, is instructed to spend a night at his pri-
vate temple and observe his dreams, along with his minister, astrologer, and
purohita:
As one of many ceremonial details of this dream divination, the purohita pre-
pares a pillow at the head of the king, surrounded by four pots of water. This
detail resembles a similr motif in Purāṇic-era pratiṣṭhā ceremonies, where the
image is laid on a pillow at the end of the preliminary adhivāsana sequence
17 Bhaṭṭotpala: teṣv āvāhiteṣu sureṣu pūjāṃ arcāṃ vidhāya tāṃ śarvarīṃ rātriṃ te
daivajñāmātyayājakā vaseyus tiṣṭheyuḥ | kim artham sadasatsvapnanimittam
śubhāśubhasandarśanārtham | yātrāyāṃ yajñeṣvaśvamedhīyāyāṃ svapnavidhiḥ
svapnaparīkṣaṇavidhir uktaḥ kathitaḥ |
yātrāyāṃ yajñeṣvaśvamedhīyāyāṃ must refer erroneously to the Yakṣyeśvamedhīyayātrā,
an alternate name for the BY.
18 Or chapter 16 in Pingree’s ed. See previous note.
28 Geslani
I would like now to take this analysis one step further, to explore the pos-
sible antecedents of Varāhamihira’s ritual program in Atharvan śānti rites.
19 It is perhaps because the image cannot report on its own dreams that the actual divina-
tion has been omitted.
20 ‘Having arrived at his own country, the king should confer bali and further favors
to the Pramathas (i.e. the Guhyakas), Asuras, Bhūtas, and gods, according to the
procedure [already] mentioned’ (svaviṣayam upagamya mānavendro ‘v[b]alim
upayācitakāni cādhikāni | nigaditavidhinaiva saṃpradadyāt prathama[matha]
gaṇāsurabhūtadaivatebhyaḥ || 5 || BY 34 ||). My emendations. This reference is from
Pingree’s edition, as the chapter is wanting in Laṃsāla’s edition.
image installation as apotropaic consecration 29
The structure of these late-Vedic rituals, I will argue, helps to explain both
the fire offering of Varāhamihira’s adhivāsana, as well as the second bath
or consecration that often follows the homa in early-medieval accounts of
pratiṣṭhā.
We have seen that a number of the royal rites described in the BS directly
refer to conventions set forth in Varāhamihira’s yātrā texts. Together, these ref-
erences suggest that the rites of the BS operate according to pre-established
ritual and astrological conventions, such as the observation of fire omens, the
observation of dreams, and the measurements of the altar. I have recently
studied a number of these conventions in the yātrā rituals (Geslani 2016) and
found that they overlap in great measure with rites found in the ancillary texts
of the Atharvaveda. A summary of my findings can be found in the following
table:
In short, although the precise sequence of rites varies, the preliminary ritu-
als of Varāhamihira’s royal journey correspond in large measure with the pre-
liminary rituals of the mahāśānti, as prescribed in the ŚK (and supplemented
by the AVPŚ). These common rituals consist in the worship of a group of
deities—the Lokapālas, the Guhyakas or Vināyakas, the Nakṣatras, and the
Planets or Navagraha. More relevant for present purposes, the ŚK, like BY/YY,
also contains a preliminary, purifying bath with clay and jewels, comparable
to the bath applied to the departing king and the divine image. Furthermore,
30 Geslani
both the yātrā ritual sequence and the mahāśānti culminate with a fire
offering.21
In order to understand further the possible relationship between Atharvan
śānti and pratiṣṭhā, I will outline Atharvan śānti rituals in brief. Beginning with
the ŚK, Atharvan śānti rites can be characterized by two core technical fea-
tures: first, mantra collections or ‘gaṇas’ and second, the saṃpātas or ‘dregs’
(Gonda 1968) of the fire offering. These two techniques are employed for the
preparation of a special type of consecration waters, already known to ear-
lier Atharvan texts as śāntyudaka.22 A mantragaṇa consists of a set of indi-
vidual verses from the Atharvaveda grouped together under a given heading
or theme, such as ‘longevity’ (āyuṣya) or ‘safety’ (abhaya). A group of man-
tras having the heading of śānti appears in the earlier Atharvan manual, the
Kauśikasūtra, and may have served as a model for generating the additional
gaṇas employed in the ŚK and in the Atharvan Pariśiṣṭas.23 Śānti rites take the
basic Vedic fire sacrifice, or darśapūrṇamāsa-iṣṭi, as their frame (ŚK 2.20.3),
and hence can be understood as versions of the simple domestic offering.
Yet in the place of the yājyā—the mantra customarily recited with the main
oblation—śānti rites instead prescribe the recitation of a series of the afore-
mentioned mantragaṇas from the Atharvavedasaṃhitā (the implication being
that only Atharvan specialists would be eligible to recite them). For instance,
the mahāśānti ritual described in the ŚK requires that no less than eighteen
gaṇas be recited at the main fire offering.24 Such recitations were quite lengthy;
the number of discrete Atharvan verses included among these eighteen groups
easily runs into the hundreds. What is truly distinctive about śānti rites is that
the “dregs” (saṃpātas) of these lengthy mantric offerings—which would seem
to contain the power of the mantras themselves—are mixed (ā√nī) into the
water pot used to bathe the ritual sponsor at the end of the ceremony. Thus the
21 In fact, I have argued that it is within this specifically Atharvan ritual structure that we
see the addition of distinctly astrological ritual activities, namely the examination of
dreams and fire omens. See Geslani 2016.
22 See Rotaru 2009.
23 Hence AVPŚ 31, the so-called gaṇamālā, lists thirty-one of such gaṇas to be employed
variously in the mature Atharvan ritual system.
24 ‘He should mix the dregs in the pot, offering with the following mantras’ (saṃpātān
ānayet kumbhe juhvan mantrair atho ‘ttaraiḥ | 2.22.5 cd |). The list of eighteen
mantra gaṇas from ŚK 2.23.1–24.2 is as follows: śānti, kṛtyādūṣaṇa, cātana, matṛnāman,
vāstoṣpatya, pāpmahan, yakṣmopaghāta, svapnāntika, āyuṣya, varcasya, svastyayana,
abhaya, aparājita, śarmavarman, devapurā, rudra, raudra, citrā. Specific mantras in
pratīka form can be found in the gaṇamālā (AVPŚ 31).
image installation as apotropaic consecration 31
A Brahmin priest should make an offering in the fire with mantras related
to the god who is to be installed. The fire-omens have been explained by
me in [the chapter on] raising Indra’s banner.
25
Nīlamatapurāṇa 841–842 (Kumari ed.); Ādipurāṇa (Ikari and Hayashi ed.), lines 2632–
2633; Viṣṇudharmottarapurāṇa 2.19.4–5. See Geslani 2012: 337.
32 Geslani
same five Atharvan mantragaṇas during a preliminary fire offering, and they
explicitly employ the saṃpātas of this offering to consecrate the new king, in
addition to the recitation of other mantras related to the gods (Viṣṇu, Indra,
Sūrya the Viśvedevas and Soma, according to Viṣṇudharmottara 2.19.3–4).
In the case of the pratiṣṭhā he has prescribed, either alternatively or in addi-
tion to the Atharvan gaṇas, mantras ‘related to (or belonging to) the deity who
is to be installed’ (yo devaḥ saṃsthāpyas tanmantraiḥ). Varāhamihira does not
specify further a further source for these mantras, and this may be in keep-
ing with the “non-sectarian” approach of his text.26 As we will see, one impor-
tant criterion for the selection of mantras in the image installation concerns
whether they derive from a ‘Vedic’ or ‘non-Vedic’ source.
The question of mantras may be closely related to the fact that, despite his
likely familiarity with Atharvan homa, Varāhamihira does not explicitly indi-
cate the use of saṃpātas in any of his rituals involving fire sacrifices, including
the pratiṣṭhā. We find the same situation in the puṣyasnāna (BS 47), an apotro-
paic consecration that represents the astrological counterpart to the Atharvan
puṣyābhiṣeka (AVPŚ 5). While both versions of this ritual share the same under-
lying logic of astrological protection through aspersion, Varāhamihira’s version
seems to de-emphasize the role of the fire sacrifice. In the Atharvan version, the
homa with mantragaṇas occurs at the outset, and—following the conventions
of Atharvan śānti—figures centrally in the preparation of the consecratory
waters. In the BS, however, the fire sacrifice does not feature in the prepara-
tion of the consecration mixture. It only appears at the end of the rite—after
the king’s aspersion—and merely offers an opportunity to observe the omens
in the fire (BS 47.77–78). In addition, Varāhamihira supplies his own mantra
for the central bath of this ritual (47.55–70). Unlike the five mantragaṇas of
the yātrā texts, this non-Vedic mantra uses “literal” speech to request the gods
directly to ‘consecrate’ (abhiṣiñcantu) the king. The efficacy of the rite thus
falls on the agency of the gods (and the pūjā offerings they have received in
exchange at BS 47.20–21), rather than in the power of mantras ‘mixed’ into
the consecratory waters through the saṃpātas (which are never mentioned).
In this way, Varāhamihira violates the bond between mantras, saṃpātas, and
aspersion that is central to the Atharvan ritual structure—a structure to which,
as we have seen, he may have been greatly indebted. He seems to prefer pūjā,
non-Vedic mantras, and the direct agency of the gods to fire sacrifice, Vedic
mantras, and Vedic priests.27
26 Bhaṭṭotpala for his part adds that these mantras should not only relate to the god in ques-
tion, but should be Vedic (tanmantrair taddaivatyair vaidikair mantrair).
27 I develop this distinction at length in a monograph in preparation.
image installation as apotropaic consecration 33
I would propose that a similar scenario is at play in the pratiṣṭhā. The fire
offering on the evening of the adhivāsana does not appear to function as a
necessary prelude to an act of aspersion, as it does in Atharvan śānti rites,
and hence no saṃpātas from this offering are applied—at least explicitly—to
the bathing of the image. In fact, it is unclear whether BS 59 even prescribes
a second bath at all, following the fire offering.28 In any case, it seems that
Varāhamihira is content that the fire offering should function in the traditional
‘Vedic’ sense: simply, to invoke the presence of the deity who is to be installed.
The further details of this interaction are left unspecified, but we are left to
presume that the god, who has been coaxed to the ritual space by the promise
of offerings, enters into the image of his (or her) own accord, to receive the
further supplication of devotees (steps 6 and 8).
I have argued that, from the perspective of Varāhamihira, the structure of the
pratiṣṭhā ceremony appears to mimic rituals designed for the king. There is
evidence, furthermore, that Varāhamihira’s corpus of royal rites—both in his
yātrā texts and in the BS—may have first developed in the context of Atharvan
śānti rituals. I would like to present in this final section some evidence from
late-Vedic and Purāṇic sources, to demonstrate the varied, but enduring influ-
ence of Atharvan śānti rituals on the structure of the adhivāsana. I focus spe-
cifically on the ‘second bath’ of typical versions of the adhivāsana, which is
often the culminating act of the ritual sequence. In short, in contrast with
Varāhamihira’s version of the adhivāsana, a number of these later sources do
apply the saṃpāta-dregs of the fire offering in the second bath of the image,
thus preserving the Atharvan ritual structure that combines mantra, fire offer-
ing, and bathing. In these instances, the saṃpāta-dregs regularly facilitate the
transferral of divine presence into the image—just as they facilitated the trans-
ferral of Atharvan mantras (or their protective powers) to the body of the king.
28 As we will see, most versions of the adhivāsana include a ‘second’ bath (following the
initial bath with leaves, clay and jewels), preceded directly by the main homa. In BS 59 it
is possible that this event is indicated by the adjective snātām (BS 59.14a), which directly
follows the fire offering (see steps 5 and 6). Bhaṭṭotpala, however, reads this term sim-
ply as a reference to ‘the image whose bath has been performed’ (tāṃ pratimāṃ snātāṃ
kṛtasnānām). At any rate, even if we depart from the commentary, there is no indication
that the dregs of the preceding fire sacrifice are used in the bathing of the image.
34 Geslani
Then the ācārya, taking the fire from the house of a śrotriya brahmana,
installing it in his own fire pit as mentioned, rekindling it, declaring the
intention, saying ‘I will sacrifice for the purpose of attaining the presence
of a portion of the god here,’ doing everything up to placing down the
praṇīta waters; establishing (the image) there, invoking the god with
mantras having his form or with the vyāhṛti syllables, he should make an
offering with palāśā, udumbara, aśvattha, or āpamārga kindling wood,
with ghee, rice, [or] sesame, 1008, 108, or 28 times.
Here the performer of the fire sacrifice explicitly states that its purpose is to
‘attain the presence of the god here’ (devasya atra kalāsaṃniddhyartham).
Immediately following this passage, the text continues, saying:
29 Not to be confused with the earlier Āśvalāyanagṛhyapariśiṣṭa. See the introductory
remarks of Aithal (1963) for the differentiation of the two texts.
image installation as apotropaic consecration 35
He should throw the dregs (saṃpāta) of the ghee offering into the śānti
pots (śāntikalaśa) that are set down at each fire pit. Then (the performer)
touching the feet of the image with (his) head, offering those offerings
beginning with the sviṣṭakṛt, he should offer the final oblation (i.e. the
pūrṇāhuti). Similarly the ṛtviks should make offerings in their own fire
pits. Then sprinkling (the image) with water placed in the pavillion, with
the four verses beginning ‘samudrajyeṣṭhā,’ and with the three verses,
beginning ‘āpo hi ṣṭhā,’ installing the image, bathing it with water from
the pot(s?) with the remainders, sprinkling it with pure waters with the
hymn ‘ambayo yanty adhvabhir,’ wrapping it with a pair of cloths, wor-
shipping it with the five services, he should perform the adhivāsana for
three nights, or one nights, or a single day.
In this scenario the dregs of the fire offering intended to obtain a portion
of divine presence are mixed into ‘śānti-pots,’ which are ultimately used to
bathe the image in the culminating act of the adhivāsana. As indicated by the
phrases praṇītasaṃsthāpanāntam kṛtvā (‘having performed [everything] up to
setting down the praṇīta waters’) and sviṣṭakṛdādi hutvā pūrṇāhutiṃ juhuyād
(‘having offered the sviṣṭakṛt and so forth he should pour the final oblation’),
it is clear that this is the main offering (yājyā) of a full fire sacrifice. After the
conclusion of the fire ritual, a pot called the saṃsrāvakalaśa, which seems to
be a reference to the pot with the saṃpātas (the śānti-pot mentioned above), is
used to bathe the image. Thus the ĀśvGPŚ reconstitutes the close relationship
between fire sacrifice and ritual bathing proposed by the Atharvan śānti rites,
while designating the pot of dregs as a ‘śānti-pot’ may betray an even more
direct influence.
Of course the fact that this late-Vedic installation rite deploys the struc-
ture and techniques of Atharvan śānti does not necessarily imply that the
image installation serves the same function of appeasement. The difference
between this application of the apotropaic consecration and its potential
Atharvan prototypes can be illustrated best by considering the mantras recited
during the fire sacrifice. ĀśvGPŚ prescribes two optional types of mantras:
tatprakāśamantrair vyāhṛtibhir vā, ‘mantras having the form of that [god] or
[mantras] with the vyāhṛti syllables.’ This option likely refers to a distinction
between Vedic and non-Vedic mantras. The ‘mantras resembling that god’ or
‘well known [to belong to] that god’ (tatprakāśamantra) likely refer to man-
tras such as—in the case of Viṣṇu—the puruṣasūkta (RV 10.90), or the verse
36 Geslani
that begins ‘Viṣṇu bestrode this world’ (idaṃ viṣṇur vicakrame; TS 1.2.13.e/RV
1.22.17). In the late and post-Vedic eras such mantras from the old saṃhitās
became associated with the current ‘Hindu’ gods, and were recited regularly
in their associated rites of pūjā and pratiṣṭhā.30 At the same time non-Vedic
mantras also appear in many of the same ritual contexts (Einoo 2005: 100–104;
Colas 1994). These mantras used literal, verbal commands to promt the deity,
by name, to action. For example they might invoke (e.g. āvāhayāmi) the god,
or solicit his or her acceptance of offerings (e.g. pratigṛhyatām). In the case
of the invocation, such literal speech was combined with the sacred vyāhṛti
syllables, as in the following specimen from the pariśiṣṭas of Bodhāyana
(2.13.18) and Hiraṇyakeśi (1.7.11): ‘Oṃ bhūḥ puruṣaṃ āvāhayāmi | Oṃ bhuvaḥ
puruṣaṃ āvāhayāmi | Oṃ suvaḥ puruṣaṃ āvāhayāmi | Oṃ bhūr bhuvaḥ suvaḥ
puruṣaṃ āvāhayāmi |.’ Such literal types of mantras present a more direct
link with the particular ritual action that they accompany. Hence the pur-
pose of the adhivāsana—to obtain divine presence—could be appropri-
ately indicated by the verb āvāhayāmi. Nevertheless, these literal, non-Vedic
mantras did not differ functionally from the older Vedic mantras, which
were also thought to attract the gods to the fire offering (however obliquely,
as hymns which refer to them by name, or describe their exploits). As the
instruction of Āśvalāyana implies, both types of mantras could suffice for
the purpose of the fire offering. Thus however the basic format of this type
of adhivāsana—which uses the saṃpāta in invoke the divinity in the second
bath—may have derived from the ritual format of śānti, it seems easily to have
been adapted to the separate problem of obtaining the presence of a god in
an image.
Comparable applications of the saṃpāta for the purpose of invocation
can be found in Purāṇic versions of the pratiṣṭhā ceremony. For example, the
Garuḍapurāṇa (1.48), which delays the second bath until the moment on the
second day when the image is installed in the temple, employs a ‘pot of dregs’
(saṃpātakalaśa) for this activity (1.48.95; Hikita 2005: 161). Nonetheless, there
are versions of the adhivāsana that incorporate saṃpātas in conjunction with
the second bath, and yet seem to associate this bath with the older apotropaic
functions. Take for instance the account of the Matsyapurāṇa (264–65; see
Appendix 2 for summary). In this version, prior to the second bath, an elabo-
rate fire offering is performed at five fire pits constructed in and around the
maṇḍapa. The main performer of the rite (the sthāpaka) is stationed at the
30 The selections from the pariśiṣṭa of Baudhāyana edited and translated by Harting (1922)
provide a number of excellent examples of this application of Vedic mantras to newer
rituals dedicated to Śiva, Viṣṇu, and other ‘Hindu’ gods.
image installation as apotropaic consecration 37
central fire pit, while attendant priests (called mūrtipas) along with represen-
tatives of each of the four Vedas are stationed at the kuṇḍas set in the cardinal
directions. Again, the dregs of these offerings are applied to the bath of the
image (steps 12–15):
The sthāpaka should offer a homa near the head [of the image] with such
mantras as those pertaining to śānti and puṣṭi, preceded by the vyāhṛti
syllables. . . . [Every priest] should offer oblations to these [image protect-
ing deities] with Vedic mantras appropriate to each, procure a śānti-pot,
and set it down at his respective fire pit. The full (i.e. final) oblation is
regarded [to occur] after one hundred or one thousand [oblations]. [Each
priest] should put [the pot?] down calmly, with his feet evenly spaced on
the ground. He should then place the dregs (saṃpāta) of the oblations
into the full water pots. He should bathe the god on the lower, middle and
upper limbs, with that [pot]. He should bathe it standing upright with the
water containing the dregs (saṃpāta) of the offerings.
the adhivāsana proper, in the pratiṣṭhā-half of the ritual, when the god is phys-
ically installed in the temple (Matsyapurāṇa 266.33–54). This direct method of
invocation obviates the need for transferring divine presence through a water
pot. And since the coveted moment of invocation is to be accomplished dif-
ferently, the second bath of the adhivāsana, has ‘reverted,’ so to speak, to its
earlier apotropaic function. Like the body of the king, the image is protected—
appeased and rectified—before it is (re-)enthroned in the kingdom. Given
that this purāṇic text was reproduced in the Pratiṣṭhākāṇḍa of Lakṣmīdhara’s
influential, 12th century digest, the Kṛtyakalpataru (pp. 68–89), the notion
of pratiṣṭhā as inclusive of apotropaic consecration seems therefore to have
remained current during the second millennium CE.
Conclusion
In this paper I have tried to indicate the somewhat complex historical relation-
ships between śānti and pratiṣṭhā that may be recovered from surviving ritual
manuals. I based my analysis on a reading of BS 59, the earliest datable ver-
sion of this ceremony, and a somewhat simple version that shares a number of
striking similarities with ceremonies of the royal yātrā described elsewhere in
the corpus of Varāhamihira. These similarities also coincide with several tech-
nical conventions applied broadly in the rituals of BS that are directly based
on precedents from BY/YY. Thus it would not suffice to say that Varāhamihira’s
image installation rite borrows from the royal rituals in a ‘thematic’ way; rather,
the pratiṣṭhā may in fact share a deeper structural affinity with the royal ritu-
als. Based on this affinity, it may be that the royal body, as figured in late-Vedic
and astrological texts, provided an important paradigm for the image in the
pratiṣṭhā.
We gain further insight into these paradigms in the texts of the Atharvaveda,
which link Vedic mantras, fire offerings, and bathing in a standard ritual struc-
ture through the application of sacrificial dregs (saṃpāta) to the bathing
waters. Despite the fact that Varāhamihira seems reluctant to use Vedic man-
tras and saṃpātas precisely in this way, late-Vedic and Purāṇic versions of the
pratiṣṭhā appear to preserve the Atharvan complex of mantras, fire sacrifice,
and sacrificial dregs during the second bath of the image in the adhivāsana. In
these later versions of pratiṣṭhā, Atharvan bathing techniques appear to have
been adapted to a new purpose: the placement of divine presence into the
image. Using the dregs of a sacrifice with Vedic or non-Vedic mantras, the god
is first invoked into a water pot, later used to bathe the image. If this hypo-
thetical reconstruction is correct, it might explain why, even without the use
image installation as apotropaic consecration 39
capacities—we may have to consider the ritual life of its post-Vedic, royal
prototype with greater scrutiny.
Adhivāsana
[1] construction of the adhivāsana-maṇḍapa and the materials for the image
(BS 59.1–6)32
[2] preparation of the sthaṇḍila and seating the image on a throne (7)33
[3] bathing the image with water mixed with sprouts and leaves, clay gathered
from different locations, pañcagavya, tīrtha water, and waters mixed with jewels
(8–10)34
[4] recitation of mantras in the east and south-east (11)35
[5] fire offering with mantras dedicated to the specific god who is being installed,
and the observation of the signs of the fire (12–13)36
[6] bathing [?], clothing, ornamenting and worshiping the image with flowers (14ab)
[7] laying the image on a bed (14cd)37
[8] tending the sleeping image with songs and dances (15ab)38
Pratiṣṭhā39
[9] worshipping with flowers, clothing, ointments, drums and conches (16ab)
[10] taking the image around the temple clockwise (16cd)40
[11] bali-offering (17a)
[12] worshipping brahmaṇas and assembly members (17b)
[13] placing a piece of gold in the base (17c)
[14] placing the image on the base (17d)41
[15] worshipping the performers involved in the rite (sthāpaka, astrologer, priest,
assembly, carpenter) (18ab)42
[11] bali-offering, placing the image protectors (mūrtipa) and gate keepers
(dvārapāla) in all directions (23–29)
[12] fire offering by the sthāpaka near the head of the image (30–32)
[13] construction of four fire pits around the maṇḍapa, worship of the presiding
deities of the image (33–36)
[14] fire offering at each of the fire pits, and placing the saṃpātas in the pots (37–45b)
[15] bathing the image with the saṃpāta-water (45–46b)
[16] incense and food offerings throughout the night (46c–47)
References
Primary Sources
Ādipurāṇa: Yasuke Ikari and Takao Hayashi (eds) ‘Ādipurāṇa.’ In A Study of the
Nīlamata, edited by Yasuke Ikari. Kyoto: Institute for Research in the Humanities,
Kyoto University, 1994.
Āśvalāyanīyagṛhyapariśiṣṭa [ĀśvGPŚ]: Āśvalāyanagṛhyasūtram. Edited by Vinayak
Ganesha Apte. Ānandāśrama Sanskrit Series 105. Poona: Anandashrama Press, 1936.
Atharvavedapariśiṣṭa [AVPŚ]: The Pariśiṣṭas of the Atharvaveda. Edited by George
Melville Bölling and Julius von Negelein. Leipzig: Otto Harrassowitz, 1909–1910.
Bodhāyanagṛhyaśeṣasūtra: Bodhāyanagṛhyasūtram of Bodhāyanamaharṣi. Edited by
L. Srinivasachar and R. Sharma Sastri. 3rd ed. Oriental Research Institute Series 141.
Mysore: Oreintal Research Institute, 1983.
Bṛhadyātrā [BY]: Aśvamedhīyayātrā. Edited by Devīprasāda Laṃsāla. Purātattva
Prakāśanamālā 46. Kāṣṭhamaṇḍapa: Rāṣṭriyābhilekhālaya, 1969 [saṃ 2026];
Bṛhadyātrā of Varāhamihira. Edited by David Pingree. Government of Tamil Nadu,
1972.
Bṛhatsaṃhitā [BS]: Bṛhat Saṃhitā by Varāhamihirācārya with the Commentary of
Bhaṭṭotpala. Edited by A.V. Tripāṭhī. Sarasvatī Bhavan Granthamālā 97. Varanasi:
Varanaseya Sanskrit Vishvavidyalaya, 1968.
Garuḍapurāṇa: The Garuḍa Mahāpurāṇam. Delhi: Nag Publishers, 1984.
Hiraṇyakeśigṛhyaśeṣasūtra: Sātyāṣāḍhaviracitam Śrautasūtram. Ānandāśrama
Sanskrit Series 53. Poona: Anandashrama Press, 1907–1929.
Liṅgapurāṇa: Liṅga Purāṇa of Sage Kṛṣṇa Dvaipāyana Vyāsa. Edited by J.L. Shastri.
Delhi: Motilal Banarsidass, 1980.
Matsyapurāṇa: The Matsyamahāpurāṇam. Delhi: Nag Publishers, 1983.
Nīlamatapurāṇa: The Nīlamata Purāṇa. Edited and translated by Ved Kumari. Delhi:
Motilal Banarsidas, 1968–1973.
Pratiṣṭhākāṇḍa: Kṛtyakalpataru of Bhaṭṭa Lakṣmīdhara. Vol. IX: Pratiṣṭhākāṇḍa. Edited
by K.V. Rangaswami Aiyangar. Gaekwad’s Oriental Series 167. Baroda: Oriental
Institute, 1979.
image installation as apotropaic consecration 43
Secondary Sources
Aithal, K.P. 1963. ‘Āśvalāyanagṛhyapariśiṣṭa, edited by K.P. Aithal.’ The Adyar Library
Bulletin 27, 217–287.
Brunner, Hélène. 1998. Somaśambhupaddhati. Quatrième Partie. Pondichéry: Institute
Français d’Indologie.
Colas, Gérard. 1989. ‘L’instauration de la puissance divine dans l’image de temple en
Inde du sud.’ Revue de l’Histoire des Religions 206, 129–150.
Colas, Gérard. 1994. ‘On the Baudhāyanagṛhyapariśiṣṭasūtra and the Vaiṣṇavāgamas.’
In Pandit N.R. Bhatt Felicitation Volume, eds P.-S. Filliozat, S.P. Narang, C.P. Bhatta.
Delhi: Motilal Banarsidass, 511–525.
Colas, Gérard. 2010. ‘Pratiṣṭhā: Ritual, Reproduction, Accretion.’ In Hindu and Buddhist
Initiations in India and Nepal, eds A. Zotter and C. Zotter. Wiesbaden: Harrassowitz
Verlag, 319–339.
Einoo, Shingo. 2005. ‘Notes on the Installation Ceremonies Described in the
Gṛhyapariśiṣṭas.’ In From Material to Deity: Indian Rituals of Consecration, ed. Shingo
Einoo and Jun Takashima. New Delhi: Manohar, 95–113.
Geslani, Marko. 2012. ‘Śānti in the Development of Purāṇic Rājyābhiṣeka.’ Indo-Iranian
Journal 55/4, 321–377.
Geslani, Marko. 2016. ‘Astrological Vedism: Varāhamihira in Light of the Later Rituals of
the Atharvaveda.’ Journal of the American Oriental Society 136/2, 305–323.
Gonda, Jan. 1968. ‘Atharvaveda II.7.’ In Mélanges d’Indianisme à la mémoire de Louis
Renou. Paris: E. de Boccard, 301–336.
Gonda, Jan. 1975. ‘Pratiṣṭhā.’ In Selected Studies 2. Sanskrit Word Studies. Leiden: E.J.
Brill, 338–374.
44 Geslani
Dominic Goodall
A Note of Introduction
Numerous publications of recent years have shaped our picture of the early
history of installation (pratiṣṭhā), beginning with Brunner 1998, Einoo and
Takashima 2005, Ślączka 2007, Willis 2009, Mills 2011*, and Colas 2010, recently
republished in French as part of Colas 2012. This paper is intended to highlight
the accounts of some still unpublished and some inadequately edited early
Śaiva sources that should be drawn upon for a fuller picture of the develop-
ment of installation-rites in the early Mantramārga (tantric Śaivism). In his
survey of a handful of Śaiva accounts of liṅga-pratiṣṭhā, Jun Takashima (2005)
opines that the practice of adopting an already existing liṅga (liṅga-parigraha)
must predate the practice of installation (pratiṣṭhā) in the Mantramārga and he
does so partly on the basis that the earliest scriptural source to which he refers
in his survey, the Svāyambhuvasūtrasaṅgraha (c. seventh-century), does not
give an account of pratiṣṭhā. But what of other early Mantramārgic accounts?
In this paper, some unpublished pre-tenth-century accounts of liṅga-
pratiṣṭhā will be examined, notably that given in the second chapter of the
Guhyasūtra of the Niśvāsatattvasaṃhitā.1 The Niśvāsatattvasaṃhitā appears
to be the earliest scripture of the Śaivasiddhānta to survive, and may also be
the earliest surviving work of the Mantramārga. Its account of pratiṣṭhā falls
in the fifth of the five books (sūtra) into which the work is divided, as transmitted
in one ninth-century Nepalese palm-leaf microfilmed by the NGMPP. That fifth
book is probably the latest layer of the work, but there are reasons to suppose
that even that late layer can be no later than the seventh century CE. The study
of the Niśvāsatattvasaṃhitā (supported for three years by a Franco-German
‘Early Tantra’ project cofinanced by the Deutsche Forschungsgemeinschaft
and the Agence Nationale pour la Recherche) has suggested some new criteria
1 The first volume of an edition and annotated translation of this important work has now
appeared (Goodall, Sanderson, Isaacson et al. 2015). Some of the discussion of pratiṣṭhā that
appeared there (pp. 60–66) has been reproduced in the following pages and expanded upon.
for judging the relative antiquity of pre-tenth-century Śaiva scriptures, and has
thrown up many new interpretative riddles. Its rituals of pratiṣṭhā are the only
rituals described in this extensive scripture in which Vedic mantras are used.
This suggests that, as we would expect, tantric installation rituals draw upon
earlier non-tantric models.
∵
In Jun Takashima’s conclusion to his study of pratiṣṭhā in Āgamic sources,
he speaks (2005: 142) of his ‘hypothesis that the earliest stage of the pratiṣṭhā
ritual is the liṅgaparigraha,’ in other words appropriation by a sādhaka of an
already existing liṅga. The earliest of the various sources that he consulted for
his study appears to be the Svāyambhuvasūtrasaṅgraha, and that text might
indeed support such an hypothesis, for, although it refers to the possibility
of fashioning, and therefore installing, a new liṅga, it gives no instruction for
doing so, giving instead the impression that the recommended option would
be to adopt an already existing liṅga (21:13–18).
• •
13a sunakṣatra°] My, Ed.; sunakṣatre Hṛdayaśiva 15a viprāḥ] Ed.; viprā
My; vidvān Hṛdayaśiva.
(Hṛdayaśiva includes this chapter of the Svāyambhuvasūtrasaṅgraha
as the tenth in his compendium the Prāyaścittasamuccaya, a tran-
scription of the oldest manuscript of which is produced as an appen-
dix to Sathyanarayanan 2015.)
His body pure, on a day made propitious by a good asterism and auspices,
having worshipped the god of gods he should take possession of a liṅga.
He should resort to a liṅga that has been installed before by sages and
On Image-Installation Rites in the Early Mantramārga 47
2 A liṅga that is kīlita could conceivably mean one that has been pierced and so damaged or
subjugated, but the list of divinities mentioned afterwards suggests that this might instead
refer to the staking out of the ground in which the liṅga installed in order to demarcate it
as sacred territory (see Goodall, Sanderson, Isaacson et al. 2015: 60, n. 79). The expression
upary upari liṅgam is unclear: could it refer to liṅgas that have multiple smaller liṅgas carved
into them in layers, in other words, to the sort of object that is known as a sahasraliṅga in
South India today? The Kiraṇatantra, in chapter 51, after giving an account of different types
of liṅgas that may be installed by sādhakas seeking different rewards, might seem to imply
this (51:68c–69b): liṅge śatasahasrākhye śatāyuḥ pūjito bhavet / uparyupari liṅge tu bhogaḥ
syād uttarottaram, ‘In the case of [installation] of a liṅga called ‘one-hundred-thousand’, [the
installer] will be respected and live for a hundred years: with each liṅga that is on top of
another, the benefits will be [correspondingly] greater and greater.’ It is possible, however,
that we should understand these two half-lines to refer to two different types of liṅga.
48 Goodall
3 For an overview of what survives of the canon of the Śaivasiddānta and how the various
scriptures may be dated relative to one another, see the preface to Goodall 2004.
4 Work has begun on editing parts of these materials, but so far it has touched just on the
themes of iconography in the Devyāmata (Ślączka 2016) and definitions of temple-types
(Mills 2011*).
On Image-Installation Rites in the Early Mantramārga 49
5 For a text, translation and discussion of this passage, see Goodall 2011: 238–239.
50 Goodall
1:108–9 Ananta, the throne, sits in the square section; the deity, whose
body is his mantras, sits on the octagonal section.
1:110–134 The cosmos can be mapped out in Sadāśiva’s body, starting with
Kālāgni in his big toe.
1:135–6 Thus all things are in the liṅga and no siddhi can succeed without
it, nor can liberation be attained.
2:1 Topics of chapter 2: characteristics of the installer (sthāpaka), of the
ground and of the building (prāsāda), of the piṇḍikā and the perfor-
mance [of installation].
2:2 The sthāpaka may be brahmacārin, gṛhastha or tāpasa (=vānaprastha?),
but not saṃnyāsin or belonging to another āśrama.
2:3–4 He should know that the liṅga contains everything.
2:5–10 Those with certain defects should not be sthāpakas.
2:11–16 The kind of ground that is required and that is not required.
2:17–19 Veneration of sthāpakas and craftsmen (śilpin) by maṅgala-
uttering brahmins and laying down the plan (sūtrapāta) to the accom-
paniment of music.
2:20 Proportions of liṅga and shrine.
2:21–37 Sixteen types of prāsāda (1 saurabha, 2 mandara, 3 meru, 4 kailāśa,
5 bhadraka, 6 nandana, 7 nandivardhana, 8 rājagṛha, 9 gaja, 10 vṛṣabha,
11 siṃha, 12 haṃsa, 13 garuḍa, 14 padma, 15, kumbha, 16 samudga).
2:39–40 Position of the temple in the town and the directionality of the
doors.
2:41–44 Four doors are named and assigned deities and asterisms:
6 Other accounts speak of just one; but perhaps one of this pair is the snānamaṇḍapa men-
tioned in 2:71 below. Note that the Sātvatasaṃhitā and later Pāñcarātra scriptures have 3
sheds: yāgamaṇḍapa, snānageha and nayanonmīlanageha (see Hikita 2005: 168).
7 These are herbs. Cf. Hikita 2005: 159.
8 These two are apparently also herbs; they also feature in the account of pratiṣṭhā found in
the Svāyambhuva-pañcarātra in 18:17b (Acharya 2015: xlvii).
9 This dismissal here, immediately after the last moment of sculpting is complete, might
seem to imply that the expression sthāpaka is used in this particular place to mean
sthapati or śilpin. But in SP4 (see p. xv), both the craftsman (śilpin) and the officiating
ācārya receive gifts here, the former because he is being given congé, and the latter pre-
sumably simply because a significant point in the ritual has been crossed.
10 The same preparation for the fire-rites appears, with identification of many of the same
mantras, in, e.g., the Garuḍapurāṇa’s account: see Hikita 2005: 160–1 (in which many of
the mantras are identified). Another parallel is, once again, the Svāyambhuva-pañcarātra
in chapter 7 (see Acharya 2015: xliv–xlv, in particular fn. 80).
52 Goodall
11 It is not clear what these are. Perhaps simply the brahmamantras and the vyāhṛtis.
But note that Einoo (2005b: 100–102) discusses various permutations of the vyāhṛtis
intermixed with names of Viṣṇu used as mantras for the āvāhana in accounts of pratiṣṭhā
in late Gṛhyasūtras. Could it be, therefore, that brahmavyāhṛti refers to a mantra
made of a mixture of the 3 or 7 vyāhṛtis with the names of Śiva that are the names of the
brahmam antras?
12 This seems to be in part a repetition of what was prescribed in 2:46. Was the earlier
passage an adumbration of what is described now in full here? Or was the earlier pas-
sage about finding a brahmasthāna in a different place (a maṇḍapa rather than in the
prāsāda)?
13 Ṛgveda 1.40.1.
14 Ṛgveda 1.90.6.
15 For a brief explanation, with text-references, to this practice of depositing jewels, see
TAK 3 s.v. navaratna.
On Image-Installation Rites in the Early Mantramārga 53
16 Perhaps this means that he should be unstinting in his use of the above three categories
of dhātu, oṣadhi and ratna?
17 For illustrations of such objects, typically in the form of embossed plaques, see for
instance Ślączka 2006, plates 31 and 32, and Le Thi Lien 2005.
18 We have provisionally assumed this to be an as yet unidentified Vedic mantra.
54 Goodall
Now of course every detail of this summary cries out for explanation or com-
mentary, but much of that will have to be postponed until we publish the
chapter in its entirety. For the moment, we may observe first of all that there
appears to be no definition of pratiṣṭhā in the Niśvāsatattvasaṃhitā and that
it would be difficult to say whether there is really any one defining moment
in the rituals. One might expect a moment in which the liṅga is instilled with
some sort of life-force to be identified as a central moment. But we find here
no jīvanyāsa (‘imposition of spirit’), as we find in some later accounts, such
as the eleventh-century Somaśambhupaddhati (SP4, 2:216), nor any mention
of insertion of prāṇa (‘life-breath’) as, for example, in some manuscripts of
the Kiraṇatantra,19 nor a rite of prāṇapratiṣṭhā (‘installation of life-breath’),
as in some twelfth- and post-twelfth-century South Indian Temple Āgamas
(see Tāntrikābhidhānakośa 3 ad loc.). Another animating moment in other
traditions is the nayanonmīlana, the opening of the eyes of the image with
a golden stylus. The liṅga of course does not have eyes, but this moment is
often equated (for example by Somaśambhu in SP4, 2:177) with the tracing on
the liṅga of the lines that suggest the glans of a phallus. Again, this is ritually
accomplished with a golden stylus. Now the tracing of the lines of the liṅga
is included in the Guhyasūtra (2:73c–75), as in many later accounts, but it is
noteworthy that such an equivalence is not alluded to, nor is it alluded to in the
Bodhāyanagṛhyaśeṣasūtra’s above-mentioned Rudrapratiṣṭhākalpa at a point
where the presentation of the ritual seems really to invite the mention of the
equivalence if it had been felt to exist (2.16.13–14):
19
Kiraṇatantra 56:52, as transmitted in a recent Nepalese manuscript, D, but not in the
(uninterpretable) reading of the old Nepalese MS of 924 AD or in that of the Southern
edition of Devakkoṭṭai.
On Image-Installation Rites in the Early Mantramārga 55
With the fire of gold he should open the eye, saying ‘You are fire.’
In the case of [the installation of a] liṅga, this is omitted, since there are
no eyes.
Nor do we find auspicious objects presented at this point before the liṅga
as though it had just now acquired vision, a feature we find, for example, in
Pūrva-Kāmika 64:63–64. Nor has the tracing of lines acquired such importance
that the rules laying down the proportions and procedure require paragraphs
of detailed instruction, such as we find, for example, in Kiraṇatantra 56:9–24.
The treatment in the Guhyasūtra is not only very brief, but actually gives little
indication of the shape and proportions of the lines to be traced, whereas later
accounts are quite specific: that of the Somaśambhu-paddhati (SP4 2:166–175),
for example, makes quite clear that the central nāla is to protrude up above the
level of the highest point of the 2 pārśvarekhās (see Brunner 1998: 136, Fig. 1).20
One might alternatively expect the kernel of the rite to be what Somaśambhu
identifies as its defining moment in SP4 1:1, namely the moment when the
spouted collar of stone called the piṇḍikā or yoni is lowered down over the
liṅga to rest upon the base, called the pīṭha. This is how Hélène Brunner, in her
copious annotation, shows us we must understand SP4 1.1cd, which defines
pratiṣṭhā as the union of the liṅga, which is Śiva, with the pīṭha, which is Śakti:
pīṭhaṃ śaktiḥ śivo liṅgaṃ tadyogaḥ sā śivāṇubhiḥ. This indeed echoes several
early remarks like this one, made at the very end of the Guhyasūtra’s account
of pratiṣṭhā (2:128ab): liṅge rudro umā vedī ekīkṛtya ca pūjayet, ‘Rudra is in the
liṅga, Umā [in] the vedī: after joining them as one, one should worship’. But
the Niśvāsatattvasaṃhitā makes no obvious distinction between pīṭha and
piṇḍikā, for it appears to use only the latter expression, even when it cannot
refer to the flanged and spouted collar of stone which Brunner identifies as
the piṇḍikā proper. Thus Guhyasūtra 2:220ff, for instance, give the dimensions
of bricks to be used for the piṇḍikā, which is therefore clearly not a one-piece
collar. Now in her discussion of the distinction between pīṭha and piṇḍikā
(1998: xviii–xix and 210–12, fn. 82), Brunner observes (p. xix) that more recent
works than Somaśambhu’s often fail to distinguish between the two; but in
fact Somaśambhu himself, as well as most older (pre-eleventh-century) litera-
ture, often makes no distinction either. So the Niśvāsatattvasaṃhitā is certainly
not alone here: the Mataṅgapārameśvara too only uses the expression piṇḍikā
(kriyāpāda 13:23 and 13:40). The Kiraṇatantra and the Mohacūḍottara appear
to use piṇḍikā, piṇḍī and pīṭha interchangeably. The Sarvajñānottaratantra, by
20 None of these works homologises the lines with Brahmā, Viṣṇu and Rudra, and that, as
Brunner observes (1998:137, fn. 371), appears to be a post-twelfth-century homologisation.
56 Goodall
contrast, appears only to use the expressions pīṭha and vedī (with one uncer-
tain use of uttaravedī, in 19:66, which might correspond to the upper collar).21
Brunner at one point suggests (1998: 212, fn. 82)22 that the various shapes pre-
scribed for the piṇḍikā can only refer to the upper-most part of the pedestal, in
other words to the collar that should alone properly be referred to as piṇḍikā.
(For an illustration of this sort of object, see Figure 1, a collar no longer in use
at Tirunāvalūr, Tamil Nadu, or see Planche XIV in Brunner 1998, which shows
photographs of a broken piṇḍikā both dismantled and resting in place around
a liṅga.) Guhyasūtra 2:121–4, for instance, list seven types (called Vāpī, Yakṣī,
Vedī, Vajriṇī, Padmasaṃsthānā, Maṇḍalā, Trikoṇā), and the Kiraṇatantra has a
similar list of eight (53:18ff), as does the Mohacūḍottara (3:13ff). But Brunner’s
conclusion on this point is not obvious to us. Why can these shape-names not
refer to the whole unit that surrounds a part of the liṅga above ground? Why
should the whole pedestal not be, for instance, round or triangular?
It seems to us that we should accept that, even if a base and a collar were
distinguished by the period of Somaśambhu, and even if pīṭha and piṇḍikā
were typically the terms privileged respectively for base and collar, we cannot
trace the terminological distinction back far.23 Moreover, it seems not impos-
sible that the Niśvāsatattvasaṃhitā simply doesn’t know about the flanged
and spouted collar of stone that is slid over the liṅga. The description of the
placing of base and liṅga (Guhyasūtra 2:116–19) is not clear and could be read
in two different ways, depending on whether one assumes or does not assume
its existence:
21 tatas tūttaravedyā vai pīṭhasya racanāṃ kuru / vidhinā śāstradṛṣṭena sarvamantrāṇi caiva
hi 19:66.
•
66a tatas tūttaravedyā vai] N; tatas tūttaravedyāñ ca M; tatas sūktaravedyāṃ ca Te;
•
tatasūttaravedyāṃ ca L 66b pīṭhasya racanāṃ kuru] MTeLpc; pīṭhasya racanāṃ kuruḥ
•
N; pīṣya racanāṃ kuru Lac (unmetrical) 66d sarvamantrāṇi caiva hi] N; sarvakarmāṇi
caiva hi ML; sarvakarmāṇi caiva hi Te.
‘Then construct the upper pedestal (?) of the pedestal, following the procedure taught
by scripture and [using] all the [appropriate] mantras.’
22 ‘Le Mohaśūrottara [. . .] donne pour cet objet huit formes possibles, qui ne concernent en
aucune façon le pīṭha.’
23 One might speculate, however, that piṇḍikā can only have come to mean ‘pedestal’
because it originally means ‘hub’ and that it therefore probably did originally mean the
wheel-like surround from which a liṅga emerges. But this speculation seems rather weak
in the absence of early textual support: other than the Niśvāsatattvasaṃhitā another
source of perhaps comparably early date that does not support such an understanding of
piṇḍikā is the Bṛhatsaṃhitā, which uses piṇḍikā several times (e.g. in 55:16) to denote the
base of a statue and not a collar for a liṅga.
On Image-Installation Rites in the Early Mantramārga 57
Figure 1 Shows a collar (piṇḍikā) that once encircled the central section of a liṅga but that is
now no longer in use at Tirunāvalūr, Tamil Nadu.
Photo: Dominic Goodall.
58 Goodall
sexual aspect of the liṅga cult according to the Saiddhāntika scriptures’ points
to these features in the Somaśambhupaddhati to shore up her thesis (1998b:
95–96).
Another moment that one might reasonably expect to be identified as cen-
tral is the insertion or application of mantras. And yet it appears that tantric
mantras are hardly used at all in the pratiṣṭhā described in Guhyasūtra 2.29 It
is instead Vedic mantras that are deployed, and not only for brahmaghoṣa (the
creation of auspicious ambient Vedic noise), but even for the nyāsa (though this
label is not used for this act of ritual placing) of mantras before the installation
of the cosmic hierarchy (prakriyā) on the liṅga. (One may contrast this with the
nyāsa that takes place in the Somaśambhu-paddhati.) Plainly the Niśvāsatattva
saṃhitā’s prominent use of Vedic mantras, although it uses none elsewhere
in any other part of the text (with the exception of the possibly Vedic forms
of the brahmamantras) suggests that these rituals are borrowed from earlier
non-Mantramārga material.30 Furthermore, as Diwakar Acharya has pointed
out, some of the Vedic mantras that are involved in the Niśvāsatattvasaṃhitā’s
account of pratiṣṭhā turn out to be the same as those used in the parallel
account given in an early Vaiṣṇava work the Svāyambhuva-pañcarātra.31
As well as the absence of a defining moment in the ritual, we may note that
the preparation of the maṇḍapa (or two maṇḍapas, as in the Niśvāsatattva
saṃhitā) in front of the projected prāsāda is related in almost all Śaiva sources
(the Sarvajñānottara, as we shall see below, is an exception) in surprisingly
elaborate detail. As Brunner observes (1998: xxxiii), we learn much, much more
detail about the temporary yāgamaṇḍapa and about the objects that are to be
made ready inside it in preparation for the pratiṣṭhā than we learn about the
32
Sarvajñānottara 19:88–89:
tato mūrtidharān aṣṭau sakalīkṛtavigrahān
japadhyānaratāṃ chāntān ardharātre praveśayet 19:88
uttiṣṭhateti viprendrā kalaśān vāripūritān
mantradravyasamāyuktān sampādayatha me ’dhvare 19:89
•
88a. °dharān aṣṭau] Lpc; °dharāny aṣṭau N; °dharāṣṭau L (unmetrical) 88cd. •
chāntān ardharātre praveśayet] L; °ratāṃ śārntā ardharātre pracocayet N 89a. •
•
uttiṣṭhateti viprendrā] conj.; uttiṣṭha * ti viprendrā N; uttiṣṭhaṃtteti viprendra L 89d.
me ’dhvare] N; me parān L.
Then he should cause eight mūrtidharas to enter [the ritual space] in the middle of
the night, each with his body transformed into Śiva by the imposition of mantras
(sakalīkṛtavigrahān), devoted to recitation and meditation, calm. [He should cause them
On Image-Installation Rites in the Early Mantramārga 61
that is to say of the first five or nine bricks (prathameṣṭakā) or stones, is indeed
also absent from the account of the Niśvāsatattvasaṃhitā, as are the mūrtipas
assigned to those first bricks or stones.
Another development that the discovery of the Niśvāsatattvasaṃhitā’s
account might seem to highlight is the growth in importance of a rite involv-
ing the vāstupuruṣa, the spirit occupying the place selected for construction.
The Niśvāsatattvasaṃhitā nowhere mentions the vāstumaṇḍala and the ven-
eration of the vāstupuruṣa as Brahmā, whereas we find an extremely detailed
treatment of this subject in SP4 and in virtually every other account from that
of the Bṛhatsaṃhitā onwards. We should note, however, that, while the SP4
gives a different grill, with different numbers of squares for placing each of
the divinities of the vāstumaṇḍala (always the same divinities) according to
whether one is preparing the site for a prāsāda, a house, a puṣkariṇī or a vāpī,
the Bṛhatsaṃhitā apparently only treats of the vāstumaṇḍala in the context
of house-construction (Brunner 1998: xxxvi). The Kiraṇa, however, does have
a detailed vāstumaṇḍala, with 81 squares for the house and 64 for a temple.
As Brunner notes with justifiable puzzlement (1998: xxxv–xxxvi), the number
seems to decrease in inverse proportion to the importance of the construc-
tion. We note that the vāstupuruṣa and his maṇḍala do not figure at all in the
account of the Sarvajñānottara, a trait that again suggests its archaic character,
to which we shall return below. In the Niśvāsatattvasaṃhitā, however, we find
what may be a trace of the vāstumaṇḍala: two of the three extant names of
the four doorways of the prāsāda allude to divinities who appear in adjacent
places in the vāstumaṇḍala, namely Puṣpadanta in the West and Bhalvāṭa in
the North (Guhyasūtra 2:41–44); one of the four names is lost, and the other,
Ānanda, does not appear to match a divinity in the vāstumaṇḍala.
Moving from ritual performance to the liṅga itself, we should underline that
the Niśvāsatattvasaṃhitā alludes to a pre-classical form in Guhyasūtra 1:78c–81
(this time with the readings of the two apographs K and W supplementing
those of the ninth-century manuscript N):
to enter] saying (iti) ‘Get up, o brahmins! Prepare me pots filled with water and fitted out
with mantras and [the requisite] substances for the rite!’
62 Goodall
The last part should be square; he should make the middle one octagonal,
the part in front of (/above) that should be hexadecagonal; and having
made it round [above that?], he should make the head like the egg of a
hen or in the shape of a parasol. On the right / South [of the head?] he
should make it [somewhat] higher: [it will thus be capable of] bestowing
the fruits of all desires. Once it has been polished on a grindstone with
cow-dung, sand, bristles and rope,33 he should also wash it thoroughly
with cow’s urine and water.
33 Or should this be understood to mean ‘sand and ropes of bristle’?
34 Some Puranic accounts, however, mention other forms. The Agnipurāṇa, for instance,
records a type (called vardhamāna) that has, starting from the bottom, a square, an
octagonal, a hexadecagonal, an icosidodecagonal, a sixty-four sided and a round part
(53:3–5).
On Image-Installation Rites in the Early Mantramārga 63
The archaeological record, however, shows that before this tripartite form
with square, octagonal and round sections became the virtually invariable
norm, there was considerable variation, and among the older types attested
are liṅgas that have extra multi-faceted sections. Figure 2, for example, shows
an old liṅga, no longer in worship, that now stands in front of the Pallava
rock-cut Atiraṇacaṇḍeśvara shrine, beside the tiger-mouth cave, just north of
Mamallapuram (Tamil Nadu). What is now above ground and visible begins
with a hexadecagonal section, followed by one with thirty-two facets, and then
a long round section, on which the thinly incised lines that suggest a stylised
glans are just visible.
The only other Mantramārga text known to us that prescribes anything
other than the classical tripartite shape with square, octagonal and round
sections for man-made liṅgas that have no faces is the Sarvajñānottara, and
once again this deviation from the norm seems archaic. The passage must be
quoted with an apparatus, for while the Niśvāsatattvasaṃhitā comes down to
us from one ninth-century Nepalese manuscript (and three twentieth-century
apographs of the old document), the Sarvajñānottara reaches us through sev-
eral disparate and incomplete sources. Collated here are: a fragmentary ninth-
century manuscript (N), a Telugu-script paper manuscript kept in Madras
(Te), a closely related Devanāgarī-script paper manuscript from Pondicherry
(L), and two Grantha-script manuscripts that transmit Aghoraśiva’s twelfth-
century commentary on the work, a palm-leaf one held in Trivandrum (T) and
a paper one held in Pondicherry (P). (For futher details, see the bibliography.)
• 52a • tatra sarvasamaṃ śreṣṭhaṃ] TP; omit. NML; tatra sarvatra samaśreṣṭha
Te • 52b suvṛttonnatamastakam] T; omit. NML; saṃvṛttonnatamastakam P;
supraktonnatamastakaḥ Te • 52c • °nirmuktaṃ] NTP Te L; °nirmuktaḥ M • 52f
aṣṭāśraṃ madhyataḥ sthitam] TP Te L; omit. N; aṣṭāśraṃ madhyamaṃ sthitam
M • 53a vṛttaṃ caiva tadūrdhvaṃ tu] TPM; omit. N; vṛttaṃ caiva tamūrdhvaṃ
Te L • 53b ṣaḍānana] TPM Te L; ṣaḍānanaḥ N
64 Goodall
Figure 2 Shows an old liṅga in front of the Pallava rock-cut Atiraṇacaṇḍeśvara shrine, beside
the tiger-mouth cave, just north of Mamallapuram (Tamil Nadu). The first visible
section above ground is hexadecagonal; the one above that is icosidodecagonal
(with thirty-two facets); at the base of the round section above that are the thinly
incised lines that suggest the shape of a glans.
Photo: Dominic Goodall.
On Image-Installation Rites in the Early Mantramārga 65
Among them, the best is one that is regular throughout, with a rounded
high head, devoid of all faults and equipped with all [required] charac-
teristics; he should make its lowest section square, {the one situated in the
middle octagonal, and [the one] above that round,} o Skanda.
• 53d • °samasya tu] N; °mayasya ca TPM Te L • 54a tathā kuryāc] TM; tathā
kuryā N; tataḥ kuryāc PL; tataḥ kuryā Te • 54b caturaśram adhaḥ punaḥ] conj.;
caturaśram adho punaḥ N; caturaśram ataḥ param T; caturaśram ataḥ punaḥ P
Te L; cāturaśram ataḥ punaḥ M • 54c paripūrṇataraṃ] NTMTe; paripūrṇatām P
(unmetrical); paripūrṇātaraṃ L • 54d adhordhvaṃ tu kṛśaṃ] N; ata ūrddhvaṃ
kṛtaṃśaṃ T (unmetrical); ata ūrdhvaṃ kṛśaṃ PTe; tvata ūrdhvaṃ tu śaṃ M;
ata ūrdhva kṛtaṃ L
But for a pure liṅga that is like a hen’s egg, he should accordingly make it
like a barley-corn in the middle: it should be square, once again, under-
neath, rather fuller in the middle and thin at the top and bottom.
In the above passage, we seem to have four types of liṅga described: the clas-
sical tripartite liṅga, with square, octagonal and round sections (19:52–53b),
for which see, for example, Figure 3; an egg-shaped liṅga with a square base
(19:53c–54), a form exemplified by (presumably pre-Angkorian) liṅgas in the
Ta Keo Museum (see Figure 4) and the Kampong Cham Museum (see Figure 5)
in Cambodia; a liṅga that is faceted from top to bottom (19:55), a type per-
haps exemplified by the polished, multi-faceted liṅgas that are today found in
shrines in several Pallava sites,36 for example in the Atiraṇacaṇḍeśvara shrine
(see Figure 6); and a truncated, and therefore perhaps flat-topped liṅga shaped
like a mṛdaṅgam drum (19:56). Another of the liṅgas in the Museum in Sambor
Prei Kuk has what might be considered a mṛdaṅgam-drum-like shape (see
Figure 7), but it does not have a particularly flat top, a feature that is typically
associated with relatively late liṅgas (as Gritli von Mitterwallner has shown
in her article on the ‘Evolution of the Liṅga’ (1984), there is, broadly speaking,
35 We have assumed here that 19:56cd continues the description of the liṅga called mṛdaṅga,
but it may after all be a description of a distinct type called chinnamastaka.
36 Even if one does not take an extreme position in the controversy about whether Pallava
caves originally had or did not house liṅgas at all (for references to discussions of this
question, see Francis 2009: 335–336, n. 96), one may doubt whether such faceted or pris-
matic liṅgas were installed in Pallava times, for their high polish and geometrical regular-
ity make them seem unrelated to the Pallava shrines in which they are found; but it should
be pointed out that in the Tamil-speaking South, as far as I am aware, such liṅgas are to be
found only in Pallava-period sites, and never, as one might expect if they dated from many
centuries later, in Cōḻa-period buildings. Besides the polished, multi-faceted liṅga that
now stands in the Atiraṇacaṇḍeśvara sanctuary, twelve other examples in Pallava-period
contexts may be mentioned: that of the Airāvateśvara in Kancheepuram (for a photo-
graph, see Francis 2009, Fig. 269 [= Francis 2013, fig. III.242]); that of the Talagirīśvara
shrine at Panamalai; that of the Kailāsanātha shrine in the Brahmapurīśvara temple-
compound at Tirupaṭṭūr in Lalgudi Taluk; that of the Guṇabharīśvara temple in Tiruvatikai
in Panruti Taluk; that of Taccur in Kallakuricchi Taluk; those in the third, fourth, seventh
and eighth of the miniature temples (counting from the left as one faces them) of Pallava
queens at the Eastern end of the Kailāsanātha temple in Kancheepuram (I am grateful to
my colleague Valérie Gillet for these details; for a photograph of the liṅga in the seventh
aedicule, see Francis 2009, Fig. 265 [= Francis 2013, fig. III.238]); the broken liṅga whose
shrine faces out towards the sea from the shore-temple at Mahabalipuram); and (as my
colleague Valérie Gillet has further reminded me) in the Mukteśvara and Mataṅgeśvara
shrines in Kancheepuram.
On Image-Installation Rites in the Early Mantramārga 67
Figure 3 Tripartite liṅgas, with square, octagonal and round sections, from the Museums
of Kampong Cham and Angkor Borei, Cambodia. This is the classical form, and
yet the full curves of the uppermost section that receives worship (the pūjābhāga)
have not yet been smoothed into geometrical abstraction.
Photos: Dominic Goodall.
• 2b saṃhāre nikhilāny ataḥ] Ed.; saṃsāre nikhilaṃ vaca N • 2c liṅgam iti Ed.;
liṅga iti N • 3c–4d] N; omit. Ed.
All creatures are dissolved (√lī) into it at the [cyclical] resorption [of the
universe], and so (ataḥ) it is therefore (tena) called liṅga. Because it is
subtle it is called liṅga. It is proclaimed to be of three kinds: the first is the
‘non-manifest’, the second is said to be the ‘manifest’ and the third is
On Image-Installation Rites in the Early Mantramārga 69
Figure 6 A polished, multi-faceted liṅga whose facets reach from top to bottom
(cf. Sarvajñānottara 19:55), of a type found today in shrines in several Pallava sites,
this one being in the Atiraṇacaṇḍeśvara shrine just north of Mahabalipuram, Tamil
Nadu. This prismatic liṅga may look centuries later than the shrine in which it is
found; but, as we have observed in a note above, it is worth remarking that in South
India this sort of highly polished faceted liṅga is commonly found in Pallava or
late-Pallava-period sites and not elsewhere.
Photo: Dominic Goodall.
70 Goodall
Figure 7
A liṅga with a shape somewhat like a
mṛdaṅgam-drum (cf. Sarvajñānottara
19:56) kept in the Museum in Sambor
Prei Kuk, Cambodia.
Photo: Dominic Goodall.
Figure 8 A four-faced mukhaliṅga from Badoh Pathari whose top is distinctly flattened
and marked off by an edge from the side of the liṅga, rather than continuing
smoothly from it as part of a single curve.
Photo: American Institute of Indian Studies.
On Image-Installation Rites in the Early Mantramārga 71
Figure 9 A four-faced liṅga now in the gallery around the courtyard of the
Vīraṭṭāṉeśvara temple in Tiruvatikai, in Panruti Taluk, Tamil Nadu, in
which we see that one face, the one turned to the camera, is plainly intended
to be Aghora because of its bulging eyes and fangs. This liṅga reflects the
Mantramārgic notion that Śiva’s five brahmamantras are his five faces
(the fifth and upper face being typically not represented). The date is
unknown; Nagaswamy (1989: 31 and caption to Fig. 1) suggests, without
strong supporting arguments (see Francis 2009*: 74), that it may have been
produced in the seventh century.
Photo: Dominic Goodall.
shows the outlines of four flat-topped and flat-sided liṅgas that correspond,
according to the prescriptions of the Ajitatantra (4:40ff), to the curvaceous
labels kukkuṭāṇḍa, chatra, ardhacandra and trapuṣa. Nothing, however, in the
Sarvajñānottara’s descriptions suggests such geometric abstraction.
Returning to what the Sarvajñānottara has to say about the classical tripar-
tite liṅga, two observations may be made. The first is that in the passage that
is quoted above, it is only in the South Indian witnesses that we find all the
three sections: in the ninth-century Nepalese manuscript, 19:52f–53a are not
transmitted (hence their inclusion in curly braces above), so that the Nepalese
text in fact specifies only that the bottom should be square. Now it may be that
this omission is simply an accident of transmission, but it might equally be the
case that the Southern sources reflect an altered text in which two extra pādas
have been interpolated in order to bring the text’s prescriptions in line with
the later even more thoroughly established norm. The second observation,
however, is that it is plain that the redactor(s) of the Sarvajñānottara knew
of the tripartite form, for it speaks of it elsewhere in the same chapter, e.g. in
19:117–118, as will be clear from the summary below. This does not mean that
19:52f and 53a must be authorial, since there is no reason that the tripartite
form should have been consistently referred to throughout, just as it is not con-
sistently referred to in the Guhyasūtra of the Niśvāsatattvasaṃhitā.
In later accounts we find a ritual element that is also absent in both the
Guhyasūtra’s and the Sarvajñānottara’s accounts that further underlines the
tripartite nature of the liṅga, and that is the superimposition onto its three
sections of mantras for ātmatattva, vidyātattva and śivatattva, the names of
three great tranches into which the universe is divided. These three tattvas do
appear elsewhere in the Guhyasūtra, in what may well be a later layer of text
(Guhyasūtra 9:183–184), and in the Sarvajñānottara (3:14), but they do not figure
in their treatments of installation. By the time of the Somaśambhu-paddhati,
however, almost all manner of thing that is to be installed (pratiṣṭhita) seems
first to be divided into three sections, presumably in imitation of the liṅga
itself, and subjected to nyāsa of the cosmos, variously grouped, but always
including the triad ātmatattva, vidyātattva and śivatattva: the first stones of the
temples (SP4 1:41ff and 53–54), the liṅga itself (SP4 2:203ff), a statue of Viṣṇu
(SP4 6:46–47), the doorway (SP4 7:6–7), the flag and flag-staff (SP4 9:13–14).
Since the Sarvajñānottara’s chapter devoted to pratiṣṭhā has not been
published, we present here, as we have done for the Niśvāsatattvasaṃhitā’s
account, a summary based on our preliminary edition, but with this caveat:
given the fragmentary nature of the sources (the old Nepalese manuscript,
for instance, does not give testimony for the whole chapter) and the poor
transmission even in the passages for which we have several witnesses, our
On Image-Installation Rites in the Early Mantramārga 73
37 For this bali-offering, made upon formally entering a piece of ground that has been
selected for a temple, see Tāntrikābhidhānakośa 3 s.v. digbali and praveśabali.
74 Goodall
19:22 Once the measuring [of the stone that will become the liṅga?] (sūtritam)
has been completed, worship of the astra should be performed, then he
should join the astra[-pot?] with astras (astrair astraṃ niyojayet).
19:23 Placing [of the stone that will become the liṅga?] in fragrant astra-
empowered water for one night.
19:24 Smearing honey on the blade of the tool that will be used to cut it.
19:25ab As one cuts, one should recite astra and meditate [on astra], but as
the blade enters(?) one should recite hṛdaya.
19:25cd After finding a suitable rock, another performance of worship.
19:26–27 bhūtabali; cutting the stone from the base of the rock.
19:28 Venerating it, he should lay the belly that is of the gods (jaṭharaṃ
devatātmakam) at the bottom(?) and the head in the NE (?).
19:29 Unclear instructions about directionality.
19:30ab The face (vaktra) [of the deity] should be [set to face] wherever the
door is placed.
19:30c–32 A shrine to the East and South of one’s settlement should face West;
a shrine to the West should face West; a shrine to the North should face East
or West: deities should not be set to face South or North (nottarābhimukhān
devān na kuryād dakṣiṇāmukhān).
19:33–34b The faces [of the liṅga?] should not be placed in the intermediate
directions, or famine, war and such will follow.
19:34cd Proportions [of door(?)].
19:35–36 A compound wall, which may be square or round, and should have
gates (gopurānvitam),38 is to be made. Inside it one should place eight temples
(mandirān) at the sides, which should face the main shrine but should not jut
into it. In them one should install the Lokapālas.
19:37 At the door outside are to be placed Durgā and Vighnavināyaka; in front
of Śiva, Nandin and Mahākāla.39
38 We cannot of course conclude from this that a South Indian temple is being described:
compound-walls with gate-ways are not unique to the South, and the huge gopurams now
associated with Southern temples do not appear before that of the Tanjore temple at the
beginning of the eleventh century.
39 dvāradeśe bahiḥ sthāpyau durgāvighnavināyakau
sthāpyau nandimahākālau śivasya purataḥ sadā 19:37
• •
37] om. T 37c sthāpyau nandimahākālau] PTeL; sthāpya nandimahākālau N; sthāpyau
•
nandimahākāla . . . M 37d śivasya purataḥ sadā] NPTe; . . . spuratas sadā M; śivasya
pura+ta+stathā L .
On Image-Installation Rites in the Early Mantramārga 75
19:38–39 Then the śivaliṅga in the middle, with full belly and sides
(sampūrṇodarapārśvam), round high head, etc, and with markings [in its
stone(?)] of lotus, half-moon, wheel, śrīvatsa [or] svastika.40
19:40–41b Dimensions for worst, middling and best types, ranging from 1 cubit
to 9 cubits.
19:41c–43 Ideal characteristics of the earth on the chosen site, depending on
whether the founder is brahmin, kṣatriya, etc.
19:44–45 Dimensions for jewel liṅgas (from 1 finger-breadth up to 1 vitasti),
which require no temple, but may instead be worshipped at home (gṛhe).
19:46 The liṅga and vedī should be of the same material and proportionate to
one another.
19:47 The vedī may be of gold, silver or pure copper for a jewel liṅga if the
appropriate jewel is not available.
19:48 Four shapes of vedikā: square, crescent-moon, full [viz. round?],
lotus-shaped.
19:49–50b After orienting the vedī correctly, one should insert the liṅga into its
middle, leaving 2 parts outside.
19:50cd?
19:51 One should carefully raise (uddhareta) the marks (lakṣaṇam) in the mid-
dle; having made 3 equal parts, he should leave the upper one.41
19:52abcd The best liṅga should be regular and devoid of all faults.
19:52e–53b One should make it square at the bottom {and, according to
Southern sources but not the old Nepalese MS, octagonal in the middle and
round above that}.
19:53c–56 Alternative liṅga-shapes: hen’s egg, octagonal, mṛdaṅga,
chinnamastaka.
19:57 Reinstallation of jewel-liṅgas in case of mistaken installation.
19:58 The pedestal (vedī) should incline to the North, should be as smooth as
the belly of a mirror; its spout (praṇāla) should be half a part to the twelve
parts that its mekhalā measures.
40 Cf. Guhyasūtra 1:91–94, which has a longer list that also contains a few auspicious symbols
that may be Vaiṣṇava: cakra, gada (metri causa for gadā?), lāṅgala, śrīvatsa and vanamālā.
41 This could be interpreted to mean that there are in fact 4 layers: above the square and
octagonal sections a long round one might follow whose lower part would fall more or
less in the middle of the liṅga if viewed as a whole; this round-sectioned part in the
middle of the liṅga would have the distinctive marks recalling a glans ‘raised’ upon it,
while the summit of the liṅga, still round in section, would be counted as the fourth part.
The ‘raising’ of the distinctive marks has already been mentioned above in 19:20 and will
be returned to below in 19:77–84!
76 Goodall
42 Note that this identification of the three sections of the liṅga with Brahmā, Viṣṇu and
Rudra is not in the account of the Guhyasūtra.
On Image-Installation Rites in the Early Mantramārga 77
19:119–121b All deities reside in the part that is worshipped; the liṅga is the
concentrated energy of all gods and entities (tattvānām); Umā, who has
the radiance of pure gold, is the pedestal (pīṭha).
19:121c–123 Having formed a throne using Dharma, etc., he should worship with
the praṇava, which belongs to all divinities, and so becomes the divinity on
whom one meditates.
19:124 If the liṅga slips (cyavate), one must perform a rite of appeasement
(śānti).
19:125–126b After homa, one should give a pair of cattle, then venerate Śiva in
the liṅga and the fire.
19:126c–127 Śiva in the liṅga receives worship; in the fire he receives oblations;
in the ācārya he bestows knowledge.
19:129–132b The ācārya should worship the weapons of the Lokapālas in the
pīṭha, the Lokapālas in the root [of the liṅga] (mūla), the Gaṇeśas in its middle,
and the Vidyeśvaras above that, ranged as on a wheel, and Śiva resting upon
Śakti and surrounded by the brahmamantras and the śivāṅga-mantras in
the top of the liṅga.
19:132c–134 By washing the liṅga with the astra-pot, the obstacles that bear
the form of Rudra (rudrarūpadharā vighnāḥ) are driven off; by washing after-
wards the pīṭha with water from the vardhanī-pot with the pāśupatāstra, the
obstacles that bear the form of Umā are driven off.43
19:135 Then he should venerate Śiva and dismiss him.
19:136–137 After venerating Caṇḍeśa with his five brahmamantras, one should
offer to him what had been offered to Śiva (nirmālya) with the right hand while
uttering dhunicaṇḍeśvarāya huṃ phaṭ svāhā, and while removing the
[previous] garland with the left hand.
19:138 The liṅga should never be left with nothing on its head; one should not
wave one’s hand above it.
19:139 The functions of the mudrās known as tarjanī, liṅgamudrā and niṣṭhurā
are mentioned.
19:140 Washing.
Holy fields of power, wherever God has established himself [as a sponta-
neously occurring liṅga], or [wherever there may be] any liṅgas that are
associated with powers and that have been used by sages and gods—
he should station himself in one of these sorts of places, or he should
carefully set up [his own] liṅga. I shall teach you that rite [of installation
next]: listen with focussed mind, my dear.
gokulākulabhūdeśe coraḍāmaravarjite
nissarppakīṭavalmīke ītibhiḥ parivarjjite 3:1
devadevam pratiṣṭhāpya susahā[yo] . . .
The subject of this third chapter is, in other words, the pursuit of siddhis by a
mantra-adept, helped by his ritual assistant (uttarasādhaka / sahāya), a sub-
ject that dominates much of the Guhyasūtra and that justifies the inclusion of
an account of the installation of a liṅga. From the perspective of the redactor,
the creation (or not) of a temple that might subsequently serve the purposes
of other worshippers seems to be entirely incidental.
To conclude, Jun Takashima’s intuition about the absence of pratiṣṭhā in the
early Mantramārga appears to be a good one: certainly it seems as though the
Niśvāsa’s Guhyasūtra may indeed have integrated and begun to adapt prescrip-
tions about pratiṣṭhā from an earlier non-Mantramārgic tradition that was
concerned with the installation of temples at a time when the Mantramārga
was not.
Abbreviations
References
Primary Sources
Agnipurāṇa. Agni Puranam By Shrimanmaharshi Vedavyas, with preface by
Manasukharāya Mora. Gurumandal Series No. XVII. Calcutta, 1957.
Ajita. Ajitamahātantram The Great Tantra of Ajita, ed. and trans. N.R. Bhatt, Jean
Filliozat, Pierre Sylvain Filliozat. 5 vols. Kalāmūlaśāstra Series 49. New Delhi: Indira
Gandhi National Centre for the Arts / Delhi: Motilal Banarsidass, 2005.
Atharvavedapariśiṣṭa. The Pariśiṣṭas of the Atharvaveda, ed. George Melville Bolling
and Julius von Negelein. Vol. 1, parts 1 and 2 (all published). Leipzig: Harrassowitz,
1909–1910.
Kāmika. Pūrvabhāga and Uttarabhāga. No editor accredited. Published by
C. Swaminatha Gurukkal. Madras: South Indian Archaka Association, 1975 and 1988.
Kiraṇatantra.
śrīmatkiraṇāgamamahātantram śrīgaruḍopaśrutam idaṃ śrībilvāraṇyasthalābhi-
jātaśrīmacchaivāgamābdhipārīṇaśrīrāmasvāmiśivācāryavaryavarasūnunā
śrīmanmāyūrasthasarvaśāstraviśāradaśrīvaidyanāthaśivācāryavaryavarāntevāsinā
śrīpañcanadasthāmmāḷagrahāraśrīśaivāgamapāṭhaśālāpradhānādhyāpanena
ca śrīmattatpuruṣaśivāparanāmakaśrīpañcāpageśaśivācāryavaryeṇa yathāmati
pariśodhitaṃ. devakoṭṭai[-]śivāgamasiddhāntaparipālanasabhādhyakṣaiḥ kumbha
ghoṇanagaravirājamānaśrīkomaḷāmbāmudrākṣaraśālāyāṃ mudritaṃ, ed. Ti. Rā.
Pañcāpageśaśivācārya. Śivāgamasiddhāntaparipālanasaṅghaprakāśitasaṅkhyā 16
(= Ed.). Devakōṭṭai, 1932.
Also consulted: National Archives Kathmandu (NAK) MS 5–893 (= N), NGMPP
Reel No. A 40/3, palm-leaf, Nepalese script, dated to 924 AD; NAK 5–4780 (= D),
NGMPP Reel No. B 172/21, paper, Devanāgarī script, dated to 1901 AD. See Goodall
1998 for further details.
Tāntrikābhidhānakośa 3. Tāntrikābhidhānakośa III. Ṭ–PH. Dictionnaire des termes tech-
niques de la littérature hindoue tantrique. A Dictionary of Technical Terms from Hindu
Tantric Literature. Wörterbuch zur Terminologie hinduistischer Tantren. Fondé sous la
direction de Hélène Brunner, Gerhard Oberhammer et André Padoux. Direction édito-
riale du troisième volume : Dominic Goodall et Marion Rastelli. Österreichische
Akademie der Wissenschaften Philosophisch-historische Klasse Sitzungsberichte,
839. Band. Beiträge zur Kultur- und Geistesgeschichte Asiens Nr. 76. Vienna:
Österreichische Akademie der Wissenschaften, 2013.
Niśvāsatattvasaṃhitā. The Niśvāsatattvasaṃhitā. The Earliest Surviving Śaiva Tantra
Volume 1. A Critical Edition & Annotated Translation of the Mūlasūtra, Uttarasūtra &
Nayasūtra edited by Dominic Goodall in collaboration with Alexis Sanderson &
Harunaga Isaacson with contributions of Nirajan Kafle, Diwakar Acharya & others.
Collection Indologie 128 / Early Tantra Series 1. Pondicherry: Institut Français de
On Image-Installation Rites in the Early Mantramārga 81
Studies
Acharya, Diwakar. 2015. Early Tantric Vaiṣṇavism: Three Newly Discovered Works of
the Pañcarātra, The Svāyambhuvapañcarātra, Devāmṛtapañcarātra and
Aṣṭādaśavidhāna, Critically edited from their 11th- and 12th-century Nepalese palm-
leaf manuscripts with an Introduction and Notes. Collection Indologie 129 / Early
Tantra Series 2. Pondicherry: Institut Français de Pondichéry / École française
d’Extrême-Orient / Asien-Afrika-Institut, Universität Hamburg.
Bhatt, Filliozat and Filliozat. 2005. See Ajita above.
Brunner Hélène, ed. and trans. 1963, 1968, 1977, 1998. Somaśambhupaddhati. 4 vols:
Première Partie. Le rituel quotidien dans la tradition śivaïte de l’Inde du Sud selon
Somaśambhu; Deuxième Partie. Rituel Occasionnels dans la tradition śivaïte de l’Inde
du Sud selon Somaśambhu I : Pavitrārohaṇa, Damanapūjā et Prāyaścitta; Troisième
Partie. Rituels occasionels dans la tradition śivaïte de l’Inde du Sud selon Somaśambhu
II : dīkṣā, abhiṣeka, vratoddhāra, antyeṣṭi, śrāddha; and Rituels dans la tradition siv-
aïte selon Somaśambhu. Quatrième partie : rituels optionnels : pratiṣṭhā. Publications
de l’IFI No. 25. Pondicherry: Institut Français d’Indologie.
Brunner, Hélène. 1998b. ‘The sexual Aspect of the liṅga Cult according to the
Saiddhāntika Scriptures’. In Studies in Hinduism II, Miscellanea to the Phenomenon
of Tantras, edited by Gerhard Oberhammer, 87–103. Vienna: Verlag der öster-
reichischen Akademie der Wissenschaften.
Colas, Gérard. 2010. ‘Pratiṣṭhā: Ritual, Reproduction, Accretion’. In Hindu and Buddhist
Initiations in India and Nepal, eds Astrid Zotter and Christof Zotter. Ethno-Indology:
Heidelberg Studies in South Asian Rituals 10, 319–339. Wiesbaden: Harrassowitz.
Colas, Gérard. 2012. Penser l’icône en Inde ancienne. Bibliothèque de l’École des hautes
etudes 158. Turnhout: Brepols.
Einoo, Shingo. 2005a. ‘The Formation of Hindu Ritual’. In From Material to Deity. Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New Delhi: Manohar
7–49.
Einoo, Shingo. 2005b. ‘Notes on the Installation Ceremonies described in the
Gṛhyapariśiṣṭas’. In From Material to Deity. Indian Rituals of Consecration, eds Shingo
Einoo and Jun Takashima. New Delhi: Manohar, 95–113.
Francis, Emmanuel. 2009*. Le discours royal. Inscriptions et monuments pallava
(IVème–IXème siècles). Thesis defended at the University of Louvain-la-Neuve in
2009.
Francis, Emmanuel. 2013. Le discours royal dans l’Inde du Sud ancienne. Inscriptions et
monuments pallava (IVème–IXème siècles). Tome I Introduction et sources.
Publications de l’Institut orientaliste de Louvain 64. Louvain-la-Neuve: Université
Catholique de Louvain.
Goodall, Dominic, ed. and trans. 1998. Bhaṭṭa Rāmakaṇṭha’s Commentary on the
Kiraṇatantra. Volume I: chapters 1–6. Critical edition and annotated translation.
On Image-Installation Rites in the Early Mantramārga 83
Takashima, Jun. 2005. ‘Pratiṣṭhā in the Śaiva Āgamas’. In From Material to Deity. Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New Delhi: Manohar,
115–142.
Willis, Michael. 2009. The Archaeology of Hindu Ritual. Temples and the Establishment
of the Gods. Cambridge, etc.: Cambridge University Press.
CHAPTER 4
Anna A. Ślączka
Introduction
Early Śaiva texts remain almost entirely unknown to scholars of Indian art
and ritual. The manuscripts of many of them have only recently been discov-
ered in libraries of North India and Nepal and they are, for the major part,
unedited and not translated. This situation is gradually improving, with sev-
eral treatises being currently edited by the École française d’Extrême Orient
and the French Institute of Indology in Pondicherry, but a lot still remains to
be done.1
The importance of the early Śaiva pratiṣṭhātantras lies in their early date.
They are, for instance, considerably earlier than other, much better known,
Sanskrit treatises on ritual and image making. Until present, scholars of Indian
art used to base themselves on the group of edited texts, such as the Mānasāra
(edited and translated already in 1934), the Mayamata (1970) and the trea-
tises studied by T.A. Gopinatha Rao in his ground-breaking Elements of Hindu
Iconography (1914) that comprised mainly the so-called South Indian tantric/
āgamic literature (the South Indian scriptures of the Śaiva Siddhānta branch
of Śaivism). The availability of these treatises made them a major, and often
the only source, for the study of Hindu iconography. This situation, however,
is not without pitfalls. For although the South Indian tantric/āgamic texts
certainly played an important role in the past and are still authoritative in
many contemporary Śaiva temples, especially in the Tamil country, they have
never been truly pan-Indian treatises, certainly not in the form that is available
to us today. Moreover, as proved by recent research, they appear to be much
later than previously assumed, and presumably not earlier than the 12th cen-
tury CE.2 In spite of this, they are frequently used as tools to interpret rituals
1 In the recent years, several Nepalese manuscripts have been studied and edited as a part
of the Franco-German project on early tantra, under the guidance of Dominic Goodall and
Harunaga Isaacson.
2 See e.g. the Preface to Goodall 2004.
and art from regions outside South India and to explain works of art and mon-
uments that predate the 12th century. It is therefore of great importance that
the demonstrably earlier treatises, such as the Devyāmata, are being edited.
In the first part of this article, I will concentrate on the date of the
Devyāmata and all the problems associated with it. Next, I will give an over-
view of Devyāmata’s chapters and, finally, I will discuss the ratnanyāsa ritual,
which I will bring in relation to similar passages in other texts.
Devyāmata must already have been well known at the time when the oldest
manuscript was written, i.e. in the mid-11th century. Other works quoting the
Devyāmata are the c. 12th-century Īśānaśivagurudevapaddhati;8 the treatise
on ritual and installation Pratiṣṭhālakṣaṇasārasamuccaya, whose oldest sur-
viving manuscript dates back to 1160 CE;9 and a compilation of āgamic texts,
the Śataratnasaṅgraha.10 Furthermore, passages very close to those of the
Devyāmata’s chapter on the iconography of the deities are found in a number
of texts of various genres dating from approximately the same period, namely
the 10th-century Mohacūrottara,11 the 11th- century Samarāṅgaṇasūtradhāra
and the eclectic Agnipurāṇa.12 The language of these passages is closely allied
to that of the Devyāmata and the authors utilize the same formulations and
vocabulary. These similarities are, of course, not as significant as the attributed
citations, but they prove the existence of a distinct iconographic tradition, not
only in art and architecture itself (as is well known), but also in the textual
sources, that flourished in North India during this period. This tradition is, on
several points, different from that of the tantric/āgamic texts transmitted in
the South.
Considering all of the above, the text should in any case predate the 11th
century. In fact, a recent study demonstrates that the Devyāmata’s chapters on
architecture are surprisingly close to, and sometimes seem even to predate, the
passages in the Bṛhatsaṃhitā, a 6th-century treatise on astrology that includes
information on architecture and image making. Furthermore, the 6th–7th-
century temples in northwest India, especially Gujarat, seem to reflect the
8 For a discussion on the date of this text, see Goodall 2004: cxi–cx.
9 See Acharya 2005: 216, note 26. In the Īśānaśivagurudevapaddhati we read in various
places: devyāmatāt samākṛṣya likhitaṃ leśato mayā, ‘written by myself having briefly
extracted (it) from the Devyāmata’.
10 For a discussion on its date, see Goodall 2004: cxv–cxix.
11 The date of the Mohacūrottara: Dominic Goodall, personal communication.
12 For these correspondences, see Ślączka 2011a and 2016.
88 Ślączka
13 Mills (PhD thesis, Oxford 2011). Moreover, in her paper ‘Dating and placing early Śaiva
texts through prāsādalakṣaṇa, the characteristics of temples’ (15th World Sanskrit
Conference, Delhi, January 2012) Mills pointed out that the architectural instructions in
the Devyāmata allow us to place the material as far back as the 5th century CE, which
would correspond with the proposed date for the earliest portions of the Niśvāsa.
14 These two texts have a description of Gaṇeśa in the chapter on iconography of the dei-
ties that resembles that of the Devyāmata. The descriptions are not identical, but it is
interesting to note that all these texts prescribe the proportions of the bodily parts of the
god (in the case of other deities, the proportions are not given). In the Devyāmata only
numbers are used, while the Pratiṣṭhālakṣaṇasārasamuccaya, and the Mohacūrottara use
both genuine numbers and the bhūtasaṃkhyās, which may be an indication that the pas-
sages given by them are later. For the lack of bhūtasaṃkhyās in the Devyāmata’s chapters
on architecture, see Mills 2016.
15 A large section of the Niśvāsa-corpus, the Niśvāsatattvasaṃhitā, is the subject of a major
ongoing study by Goodall, Sanderson et al. At the time of writing, the first volume of this
study, which has now appeared, had not yet been completed.
16 The colophons read: iti niśvāsākhye mahātantre bījabhede devyāmate or bījabhede
pratiṣṭhātantre devyāmate.
The Ratnanyāsa Ritual in the Devyāmata 89
Kdei, Preah Vihar, Phnom Sandak, Prasat Tor and Sdok Kak Thom.17 Yet, in
India, the Niśvāsa has gradually been supplanted in importance by works of
more developed theology. The fact that the author of the Devyāmata stresses
its connection with the Niśvāsa may point to an early date of the Devyāmata,
in the time when the Niśvāsa was still considered an important work. If this
attribution is correct, it would make the Devyāmata one of the earliest known
Śaiva texts.
On the other hand, there are several indications that the Devyāmata might
not be as archaic as some would hold it to be. Or at least, that a great por-
tion of the text has been heavily modified in more recent times (this of course
is valid for the majority of the texts, but the question is how much has been
modified). This certainly holds true for the aforementioned chapter on the
iconography of the deities. In the form available today, the chapter does not
appear earlier than the 9th century CE. This relatively late date is based on
the comparison with passages on image making in parallel texts and on the
analysis of the iconographic prescriptions.18 A good example is the iconogra-
phy of the dancing Śiva. The Devyāmata speaks of a specific ten-armed form of
the god, depicted dancing on the back of his vehicle, the bull, surrounded by
the gaṇas and other attendants. Visual representations of this form are found
in the northeast of the Indian subcontinent, the present West Bengal and
Bangladesh, and are, apart from a few images of a later date, restricted to that
area. The earliest representations date from the 9th, perhaps late 8th century
CE. No other text, with the exception of the Pratiṣṭhālakṣaṇasārasamuccaya,
which bases itself on the Devyāmata, describes this peculiar form, so the cor-
relation between text and images seems certain and it seems plausible in this
case that the descriptions postdate the images.19 Similarly, the chapter on the
ratnanyāsa—the very topic of this article—shows signs of being either heav-
ily reworked or compiled at a later date, certainly after the 5th–6th century
proposed for the chapters on architecture.
17 See: Sanderson (2001: 7, note 5 and 22, note 28). This connection makes the Devyāmata
not only a great source of information on Hindu art and ritual in India, but also, possi-
bly, an important document for the study of art and ritual in Southeast Asia during the
Angkorean period, the more so that no indigenous Khmer manuals on these topics have
survived.
18 For the edition of the chapter, see: Ślączka 2011a and 2016. The dating is tentative. For the
chronology of the images, see Ślączka 2015.
19 Śiva dancing on the back of the bull is also described in the 10th-century Mohacūrottara,
but there he has sixteen arms. Such images are only found in Nepal in a later period, for
example in the Tusa Hiti in Patan. See Ślączka 2015.
90 Ślączka
The Devyāmata, moreover, does not seem to have preserved the archaisms
concerning social religion, so prominently present in the Niśvāsa-corpus.
Social structures are not very well-developed in the Niśvāsa, for example the
roles of the ācārya, sādhaka, and so on, are not very well defined yet.20 Still,
they appear rather well-defined in the Devyāmata, as shown in sections on
ritual. For example, the text mentions specific helpers of the sthāpaka known
as mūrtipās—a term not occurring in early texts—, whose desired characteris-
tics are even listed in a separate short chapter.
It appears therefore that the text, taken as a whole and in the form avail-
able to us today, although certainly pre-11th century, cannot be as early as
the proposed 5th or 6th century. How much of the text still preserve the old
core, however, cannot be established or even guessed at before all the chapters
are transcribed, edited and analyzed.
The Devyāmata is written in the form of a dialogue between Śiva and the
Goddess. It begins with the praśnapaṭala, ‘the chapter on questions’, in which
Devī tells Śiva what topics she would like to learn about. The answers are pro-
vided in the following chapters, of which each is dedicated to one specific
subject. Consecration rituals as presented in the Devyāmata occur on two
points during temple construction: during the preparation of the terrain for
the future temple and, later on, but placed earlier in the text, during the liṅga
installation (liṅgapratiṣṭhā).
The Devyāmata in its most complete form consists of over a hundred
chapters (presumably 106).21 The list of chapters is given in the Appendix.
Manuscript C misses chapters 13 to 20,22 and 25 to 63, which amounts to almost
all chapters discussing the liṅgapratiṣṭhā and the iconography. It seems almost
a conscious decision to remove all the sections on ritual and, along with them,
the six chapters on iconography that directly follow the pratiṣṭhā, and to con-
centrate instead on the temple building alone.
As usual in this genre of texts, the chapters are grouped together around
a few broader topics. In the Devyāmata, these are the installation of a liṅga
(chapters 2 to [57]), iconography ([58]–[63]) and temple architecture (64–106).
23 Chapter numbers not given by C are my own addition and therefore written in square
brackets.
92 Ślączka
The following chapter (22) specifies a good astrological moment to begin the
ceremony. Next, we find instructions concerning the preparation of the ritual
space, such as the construction of various temporary structures (maṇḍapa)
and the sthaṇḍila (23); the preliminary ceremonies that take place there, such
as the adhivāsana of the liṅga and the piṇḍikā (24-[27]); the offering (bali) to
the ‘demons’ [28]; and the choice of a good astrologer [29]. These are followed
by a chapter describing ritual implements, presumably as a preparation for the
homa (sruksruvalakṣaṇa, [30]).25 Next, the adhivāsana is explained in detail
[31], followed by the establishment of the brahmashtāna ([32]; the location in
the centre of the garbhagṛha where one should fix the liṅga) and the instal-
lation of the Brahmā-stone (brahmaśilā [33–35]). Then the liṅga is brought
to the garbhagṛha (liṅgapraveśana, [36]) and the ceremony of placing the
gems is performed (ratnanyāsa, [37]). Following chapters [38–49] deal with
the visualization of Sadāśiva and his entourage, placing the śivatattvas and the
ādhāraśakti. The entire section dedicated to liṅgapratiṣṭhā ends with listing
the possible guṇas and doṣas, the appeasement of the lords of the geographi-
cal directions, the placing of the piṇḍikā, the worship of Caṇḍeśa, and the
usual reward for the officiating priest and his helpers by the patron (dakṣiṇā
and gurupūjā), [50–57].
Guessing from the preceding chapters, the ratnanyāsa takes place after the
preliminary ceremonies have been completed and the liṅga has been brought
to the garbhagṛha, but—as can be expected—it is not installed yet. The
objects are presumably deposited in small cavities incised on the surface of
the, already installed, brahmaśilā.28
6–9. [The main ‘ingredients’ of the ritual are enumerated]: precious stones
(ratna), minerals (dhātu), grains (bīja) and fragrant herbs (oṣadhi).32 They
should be placed clockwise, beginning with the east and ending with the
northeast,33 with the recitation of the tantramantra, the praṇava, the īśāna,34
the hṛdaya, the upacāra, or an ‘universal mantra’.
10–11. The cavities should be anointed with a mixture of bdellium and the
products of the cow or bdellium and milk. Having gathered all the ingredients,
one should begin placing the precious stones in the cavities on the brahmaśilā.
12–15. [Placing the eight precious stones (ratna)]: a diamond should be
deposited in the direction of Indra (east), a pearl in Agni (SE), a cat’s eye in
Yāma (S), a conch-shell in Nairṛti (SW), a crystal in Varuṇa (W), a topaz in Anila
(NW), a sapphire in Kubera (indranīla; N), a sunīla in Īśāna (NE). When there
is no pearl, a coral should be placed instead, a rājāvarta instead of a cat’s eye, a
ruby instead of a sapphire.
16–18. [Placing the eight minerals (dhātu)]:35 yellow orpiment (haritāla),
śilā (?), collyrium (añjana), talk/mica (abhraka), kāśīma/kāśīsa,36 vermillion
32 The text reads: gandhaś cauṣadhayas, but no ‘fragrant substances’ (gandha, sometimes
constituting a separate ‘category’ in similar lists) are mentioned in the following verses.
33 Manuscript B (A misses the first pāda because of manuscript damage; C, as noted above,
lacks the entire chapter) reads: prāsāde svakrameṇaiva yāvad īśāna gocaraḥ. Dīptāgama
20 (between verse 244c and 244d, in a verse given only by manuscripts C2 and D), how-
ever, has in this place prāgādyanukrameṇaiva, which seems a more plausible reading. Cf.
also Guhyasūtra 2.107ab: pūrvādārabhya nyastavyā yāvadīśānagocaram.
34 Or: [while reciting] for each [ingredient] the mantra of the Lord of the [appropriate]
geographical direction. Manuscript A gives: svadīśeśena mantreṇa; in B the text is missing
because of damage to the manuscript. Similar expressions are found in a number of cog-
nate texts. According to the Somaśambhupaddhati, the bījas of the lokapālas, beginning
with Indra are: lūṃ, rūṃ, śūṃ, ṣūṃ, vūṃ, yūṃ, sūṃ and hūṃ (Brunner 1998: 38, n. 121; 194,
n. 22).
35 Devyāmata ratnanyāsa, 16:
haritālaṃ śilā caiva añjanam abhrakam tathā /
kāśī[s]aṃ caiva siṃdūraṃ mākṣikaṃ gaurikaṃ tathā //
An almost identical list, including the mysterious kāśī[s]a, is given by
Pratiṣṭhālakṣaṇasārasamuccaya 16.38:
haritālaṃ śilāṃ caiva añjanam cābhrakam tathā /
kāśīvam atha sindūraṃ mākṣikaṃ gairikaṃ nyaset //
Dīptāgama 20.247cd–248ab has a similar list, but substitutes sīsa (lead) for śilā and
gandhika for kāśīsa. Guhyasūtra 2.106 gives, among others, śilā, gandhaka and kāsīsa.
36 It is not clear what is meant here. Kāśī[s]a was, obviously, problematic as guessed by
numerous variants of it given by several texts. Cf. Pratiṣṭhālakṣaṇasārasamuccaya 16.38
(in the note above), Matsyapurāṇa 266.11: kākṣīkāśī and Guhyasūtra 2.106: kāsīsa.
The Ratnanyāsa Ritual in the Devyāmata 95
(sindūra), pyrites (mākṣika), and red chalk.37 When there is no collyrium avail-
able, one should place gandhaka; when there is no vermilion, pādara;38 when
there is no pyrites, one should place saurāṣṭī; when there is no red chalk, one
should place rocanā.39 One should place these in due order, beginning with
the east.
19–21. [Placing the eight grains (bīja)]: wheat, barley, sesame, muṅga,40 wild
rice, millet (śyāmaka), mustard and rice. One should place these in due order,
beginning with the east. In the lack of muṅga one should use canaka; when
there is no canaka, one should use pulse (māṣa).
22–23. [Placing the eight herbs (oṣadhi)]: one should place, clockwise, san-
dalwood, red sandalwood, dark aloe wood, añjanamūlika, uśīra, viṣṇukrāntā,
sahadevā and lakṣaṇā.41
24–26. In the absence of [prescribed] precious stones, one should place a
diamond; in the absence of diamonds, gold; in the absence of minerals one
should place yellow orpiment; in the absence of grains, one should place every-
where barley; in the absence of herbs, the wise one should place sahadevā.
One should not use excessively much grain; an excess or want of grain will
result in disease, sorrow and fear.42
27–30. A lotus, a bull, the Meru or the Earth (pṛthivī) endowed with all
characteristics43 should be placed on the brahmaśilā, in the middle. The lotus
37 Manuscript A gives gaurika, ‘white mustard’ in 16d (in B this half-verse is missing), and
in 17d we find gorika (in both manuscripts), but one would rather expect a ‘mineral’ here,
most probably gairika, ‘red chalk’, which given by several similar lists.
38 Probably wrong for pārada, quicksilver, that is common in such lists. The same mistake
occurs in verse 32. Cf. Pratiṣṭhālakṣaṇasārasamuccaya 16.39 and 63, verses almost identi-
cal to those given by the Devyāmata, but reading pārada.
39 Probably for gorocanā, yellow orpiment.
40 Presumably mudga-beans.
41 Sandal wood, uśīra, viṣṇukrāntā and sahadevā are commonly found in similar lists.
Lakṣaṇā seems more problematic. Cf. Pratiṣṭhālakṣaṇasārasamuccaya 16.44d: sahāṃ ca
lakṣaṇaṃ kvacit; Dīptāgama 20.50d: sahalakṣmīṃ ca vinyaset, Matsyapurāṇa 266.14ab:
vaiṣṇavīṃ sahadevīñ ca lakṣmaṇañ ca tataḥ param.
42 A prescription found in various texts, especially concerning the grains. See, for instance,
Viṣṇusaṃhitā 18.32cd and Paramasaṃhitā 19.47ab.
43 Manuscript B (A shows a number of lacunae due to damage) reads here:
padmaṃ vṛṣo tha vā meru pṛthivīlakṣaṇānvitā /
eṣām anyatamam madhye nyaset brahmaśilotpari //
Pratiṣṭhālakṣaṇasārasamuccaya 16.51 is very similar, but does include the tortoise (men-
tioned by the Devyāmata in the next verse):
kūrmākṣmāṃ merum ukṣāṇaṃ padmaṃ vā lakṣaṇānvitam /
eṣām ekatamaṃ madhye nyasen brahmaśilāvaṭe //
96 Ślączka
Guhyasūtra 2.110cd mentions padma, vṛṣa, kūrma and vedi; Dīptāgama 253cd-254ab
pṛthivī, meru, padma and vṛṣabha; Matsyapurāṇa 266.16 kūrma, dharā and vṛṣa.
44 The text gives here plural: brahmasthāneṣu, which does not sound convincing. The verse
is illegible in both A and B because of damage to the manuscripts (but with the ending—
eṣu well-visible). It might be wrong for brahmasthāne tu, a common mistake. Another
possibility would be to understand that the lotus and the metals: silver, gold and copper
should be deposited in different places, but it seems less likely.
45 A golden tortoise that should face the entrance and the Earth, in that case ‘endowed with
all jewels’, are also mentioned in the chapter brahmasthānasādhana verse 32. It seems
that the text now simply repeats what should be done, but in more detail. Yet, one would
have to edit all the chapters concerning the brahmasthāna to determine what is really
happening here.
46 The reconstructed reading on basis of A and B is: dhātubhiḥ sarvaratnaiś ca bījaiś
coṣadhibhis tathā /padmādīnāṃ nyased ekabrahmasthāneṣu mantravit, with eka- perhaps
wrong for evam.
47 Or: vajra and tāla: diamond and palmyra palm leaf, but it seems unlikely.
48 Kṛsara, according to Monier-Williams (1899), is a ‘dish consisting of sesamum and grains’
(Pratiṣṭhālakṣaṇasārasamuccaya 16.35 gives here kṛśarā). However, Somaśambhupaddhati
IV.III.14cd has a very similar half-verse with variant readings kṛsara / kesara, the latter
being accepted by Brunner (1998: 194) in her edition and translated as ‘étamines d’or’.
49 It is not clear what is prescribed here: pāda 36c is illegible in both manuscripts.
The Ratnanyāsa Ritual in the Devyāmata 97
is established in this way the people rejoice. Together with the tortoise one
should place earth taken from a mountaintop, a river or a shore, silver, and
barley grains.
41. One should place [these?] on the tortoise-stone (kūrmaśilā)50 on the
brahmasthāna. There will be a constant prosperity for the kingdom and hap-
piness for the yajamāna.
42–46. The ratnanyāsa is thus completed. One should place all ingredients
so that they will not move. The cavities filled with ingredients . . .51 The sthāpaka
should fill in the cavities holding the precious stones with liquidized bdellium
or with cow-milk and should make the surface of the brahmaśilā even. Then
he should worship the Lord and the throne (āsana) with mudrās and mantras.
Thus ends the chapter on the placing of gems.
50 The kūrmaśilā occurs here for the first time in this chapter, but it was already mentioned
in the chapter brahmaśilānyāsa, verse 2, in the passage dealing with the establishing of
the brahmasthāna. In some texts it is a synonym of the brahmaśilā. In the Devyāmata,
however, it seems that there are two separate stones (as in Matsyapurāṇa 266) as both are
mentioned in chapter 85 (mūlapādapīṭhalakṣaṇa).
51 The verse is illegible in both manuscripts.
52 For instance, the (main) image of Viṣṇu or the Goddess, installed in the garbhagṛha.
98 Ślączka
53 Only prathameṣṭakānyāsa and its variant śilānyāsa appear equally popular.
54 Garbhanyāsa, for instance, is prescribed mainly (with a few exceptions) by South
Indian texts, mūrdheṣṭakānyāsa only in South India texts (but not in those from Kerala),
ṣaḍādhāra only in the Tantrasamuccaya (a text from Kerala) and the Śilparatna.
55 Again, along with the prathameṣṭakānyāsa.
56 For the date of the Guhya, see Goodall 2015.
57 See Bṛhatsaṃhitā 59.17 (ed. Dwiwedi; 60.17 ed. Bhat), Garuḍapurāṇa 48.81–82,
Matsyapurāṇa 266, Viṣṇudharmottarapurāṇa III.110. Agnipurāṇa describes the
ratnanyāsa in two places: once, in chapter 60 (vāsudevapratiṣṭhāvidhi, perhaps inspired
by the Hayaśīrṣapāñcarātra), and second time in 97 (śivapratiṣṭhākathana), the latter
being an almost exact copy of Somaśambhupaddhati IV.3 (śivaliṅgapratiṣṭhāvidhi). The
dates assigned to purāṇas are uncertain. There is, however, an agreement in placing the
Viṣṇudharmottarapurāṇa somewhere in the 7th century (Kramrisch 1946, note 10).
The Matsya is also considered a relatively early work and is dated by Kane (1946: x)
between 300 and 600 CE, but it is impossible to tell when exactly the passages on ritual
have been included.
58 See, for example, Ajitāgama 18, Kamikāgama I.66 and I.70, Kāraṇāgama I.59.172cd–
185ab, Rauravāgama 28.69–70ab, 30.40–46, 36.12, 37.25, 38.15, 51.21, Dīptāgama 20,
Kumāratantra pp. 24–25, Cintyāgama 20, Īśānaśivagurudevapaddhati IV.39.92, IV.47,
Somaśambhupaddhati IV.2–3, Atrisaṃhitā 18, Kāśyapajñānakaṇḍa 45, Kriyādhikāra 8,
Hayaśīrṣapāñcarātra 38, Pādmasaṃhitā 11.30, 13.55–58, 28, Jayākhyasaṃhitā 20.306ff,
Paramasaṃhitā 19.40ff, Marīcisaṃhitā 15.5, Viṣṇusaṃhitā 18.22ff.
The Ratnanyāsa Ritual in the Devyāmata 99
and the Śilparatna59 that base themselves on the South Indian āgamas, and in
the āgama-śilpaśāstra Kāśyapaśilpa.60
59 Mayamata 34 seems to allude to the ritual, but does not explain it, and does not use the
term ratnanyāsa. Dealing mainly with architecture and iconography, these three texts do
not discuss liṅga- and pratimāpratiṣṭhā in detail. Still, they do describe in detail the con-
secration rituals connected with temple building proper, such as the placing of the first
and crowning bricks and the embryo-deposit (garbha). The North Indian Aparājitapṛcchā
does mention a ‘ratnanāsa’ and items of various ‘categories’ including a figure of a tor-
toise, but it is not clear where exactly the objects should be placed (see 148.28, 149).
60 The Kāśyapaśilpa is a text on architecture and iconography, but at the same time it is
closely connected with the South Indian śaivāgamas, particularly the Aṃśumad, being
presumably its upāgama (see Ślączka 2007: 11–16). The text describes placing the sup-
port stone for the liṅga, called ādhāraśilā, an equivalent of the brahmaśilā (transcript T1,
chapter 53; in the edition by Vajhe it is chapter 59 and there are considerable differences
in both texts), but the placing of gems and other objects seems not to be mentioned.
As the text is not edited critically, and the variant readings of different manuscripts are
numerous, it is possible that the ratnanyāsa is mentioned elsewhere, even if only in a
single sentence.
61 For instance, the Viṣṇudharmottara and the Garuḍa where the descriptions of the
ratnanyāsa are still very concise. It is used in the Agnipurāṇa and the Matsyapurāṇa.
62 Guhyasūtra 2.105a where the term may occur is given in the uppermost line on folio 47
verso. The top of the palm leaf is partly damaged, with as a result a following reading:
ratna /damage /tataḥ kuryāt. The remaining lowermost part of the ligatures, however,
strongly suggests the reading nyāsaṃ, which would result in: ratna[nyāsaṃ] tataḥ kuryāt.
The unpublished transcription of the Guhya by Goodall, which he kindly shared with me,
gives -sthānaṃ in the damaged part. As noted by Goodall, it is a reading provided by the
scribe of the Kathmandu apograph (NGMPP A 159/18 / msK) of the Nepalese palm leaf
manuscript of the Guhya (NGMPP A 41/14). It might have been either a conjecture or a
reading of akṣaras that are only partly legible. To me, however, the shape of the only vis-
ible ligature, suggests rather -nya- than—stha-. The term ratnanyāsa seems here to refer
to the entire ritual rather than to the depositing of precious stones only.
100 Ślączka
to any consecration ritual with terms ratna- and garbhanyāsa often used
interchangeably.63
63 Goudriaan (1965: 137) and Dagens (2004: 204) use the term garbhanyāsa while referring
to a deposit for an image. Yet, there is no basis for it in the texts translated by them, the
Kāśyapajñānakāṇḍa and the Dīptāgama respectively, or in any other Sanskrit treatise
dealing with these rituals. The only exception known to me is Viśvakarmavāstuśāstra
73.10, which does use garbhanyāsa in reference to the deposit for an image.
64 The Mohacūrottama (f. 35v–36r) mentions the ceremony very briefly: gems and so on
should be put pīṭhagarbhe, and in another verse piṇḍikāgarte. Ratnanyāsa might also be
mentioned elsewhere, but I did not have the possibility to read the entire text (which still
awaits edition).
65 Images of tortoises are prescribed for all types of temples and for various consecration rit-
uals, see for example (for the garbhanyāsa): Dīptāgama 4, Īśānaśivagurudevapaddhati 27,
Kāraṇāgama 6; (for the prathameṣṭakānyāsa): Atrisaṃhitā 6, and so on. Tortoises made
of gold, gold foil or stone were found in consecration deposits of temples of all affilia-
tions, while those of bulls mainly in Hindu temples dedicated to Śiva. One of the few texts
where the image of a tortoise is clearly associated with Viṣṇu is Tantrasamuccaya 1.150
(Viṣṇu is invoked in the stone tortoise buried in the middle of the foundation pit).
The Ratnanyāsa Ritual in the Devyāmata 101
66 The exceptions are: the Marīcisaṃhitā (mentions the ratnanyāsa in a single sentence),
and, possibly, the unedited Aṃśumadāgama and the previously mentioned Kāśyapaśilpa.
The Kumāratantra mentions only precious stones.
67 The Guhya manuscript has a lacuna here, with as a result only four precious stones men-
tioned, but one should assume that originally there were eight. See: Guhyasūtra 2.108–109.
68 About four verses in total, but scattered over the entire chapter. It should be stressed
that the passages from the Devyāmata found here are different from those found in the
Pratiṣṭhālakṣaṇasārasamuccaya. No verses occur in all three texts. Similarities with the
Dīpta are also found in the passages concerning the installation of the brahmaśilā.
69 This also is the case of some verses of Devyāmata brahmasthānasādhanapaṭalaḥ.
102 Ślączka
Taking into account the arguments presented above, the ratnanyāsa passages
in the Devyāmata cannot be considered very archaic. The tendency is that
the later the text, the more complex the description, with the peak reached
in the ritual texts of the Pāñcarātras and the Vaikhānasas, and the South Indian
70 Cf. Devyāmata 16cd and Matsya 266.11cd, Devyāmata 19 and Matsya 266.12; Devyāmata
22 and Matsya 266.13. Further, the precious stones are the same in both texts, and given
in the same sequence, but the Matsya enumerates them in a single verse, which strongly
resembles Pratiṣṭhālakṣaṇasārasamuccaya 16.33. On the other hand, one should remem-
ber that the lists of objects to be placed, especially the precious stones, often resemble
each other, with only minor variations.
71 Although not in one and the same section. Part IV.II.102 in a section explaining for the
first time the items to be deposited during the ratnanyāsa mentions a tortoise, the Earth,
a bull and a lotus; Part IV.III.11–14 speaks of a tortoise, a bull, the Meru and the Earth.
72 Placing a lotus leaf and a living tortoise in the first layer of bricks of the Vedic fire altar is
already prescribed in the Śatapathabrāhmaṇa (see Śatapathabrāhmaṇa VII, 5, 1, 1ff.; VII,
4, 1, 15ff., etc).
73 Apart from the Somaśambhupaddhati, the Meru is mentioned in the Dīptāgama (the
only āgama prescribing four of the five items found in the Devyāmata, namely the Earth,
the Meru, the lotus and the bull). The Earth is prescribed by all of the above and by
Matsyapurāṇa 266 and Jayākhyasaṃhitā 20. Hayaśīrṣapañcarātra 38 prescribes all five
items. Kāraṇāgama I.59 (as quoted by Bhatt 1972: 43, note 13) prescribes the Earth and the
Meru, but they are part of the list of maṅgala signs.
The Ratnanyāsa Ritual in the Devyāmata 103
74 One of the earliest truly complex descriptions of the ratnanyāsa is given by the Pāñcarātra
Jayākhyasaṃhitā, which has been dated to about the 8th century CE (Rastelli 1999;
Matsubara 1994). Relatively early is also the Hayaśīrṣapañcarātra (c. 9th CE as given by
Dasgupta 1989 and Smith 1978), whose description of the ritual shows some (small) simi-
larities with that of the Devyāmata (see the previous note).
75 It would be interesting to look at other unedited early texts on ritual in the future (see
note 64).
76 For example, between the chapter on iconography of the deities and the description of
Sadāśiva in ādisṛṣṭipaṭala.
77 It would be especially interesting to compare prescriptions on the same topic that occur
both in chapters on architecture and in those on liṅgapratiṣṭhā, such as those concerning
the brahmaśilā. It would also be useful to analyze the śilānyāsa (placing the [first or foun-
dation] stones) chapter, which is part of the section on architecture, and compare it with
similar chapters in other texts. Moreover, it would be desirable to make a ‘manuscript
tree’ for the available manuscripts. Could manuscript C, which omits almost all sections
dealing with installation rites and iconography, be based on an older source (older than
manuscript A) concerning mainly architecture?
104 Ślączka
iconography correspond with the statuary of Bengal. Given the current state of
our knowledge, it would, however, be imprudent to draw any firm conclusions,
and the same is true for the speculations concerning the date of the Devyāmata
in general. In order to determine the date, even approximately, it would be nec-
essary not only to edit the entire text, but also to compare it with its supposed
source, Niśvāsatattvasaṃhitā, and with other early pratiṣṭhātantras, such as
the Mohacūrottara, the Mayasaṃgraha and the Piṅgalāmata, which still await
edition. Information gained from these texts may also alter the development
pattern of the ratnanyāsa drafted here.
It should be added that the way the ratnanyāsa developed and the date of
the earliest texts mentioning it are important for yet other reason. Traces
of consecration rituals, in form of specifically placed stones and small objects
of symbolic value, have been discovered in temples all over South and
Southeast Asia.78 Many of them show, to a degree, correspondences with the
ratnanyāsa as described in the textual sources. In the course of my research,
I could list about twenty images of tortoises and twelve bulls, not mention-
ing the innumerate lotus flowers and other items, often made of gold or gold
leaf, discovered on various places in the foundation or below the pedestals
of images. Even though one cannot assume that the ritual performers always
strictly followed the prescriptions given by the texts, an attempt to compare
textual and archaeological data may provide interesting insights as it is often
easier to attribute a date to a temple than to a Sanskrit text.
Until the present, however, the earliest archaeological remains of consecra-
tion rituals were predating the texts that describe them.79 This has changed
with the discovery of the Niśvāsa-corpus, especially the Guhyasūtra, which
presumably is the earliest known text that speaks of depositing figures other
than precious stones and gold, namely a tortoise, a bull, a lotus, and the mys-
terious ‘vedi’, somewhere below the pedestal of an image, and may therefore
be the textual basis for consecration rituals actually performed in South and
Southeast Asia around the 7th century CE. It should be remembered that the
Niśvāsa-corpus was known in Cambodia during the Angkorean period where
several objects testifying to the performance of consecration rituals have been
found. On the other hand, the Devyāmata, a text certainly later than the Niśvāsa
78 See Ślączka (2007: 221ff and Appendix IV). It should be noted that the majority of the
objects were found in Southeast Asia. For possible reasons of this, see Ślączka, idem,
254–260.
79 For instance, the gold leaves of various shapes found in Go Thap, Vietnam (not later than
7th century; see Ślączka 2011b) or the golden tortoise and lotus flower found in Suan Por
Iad, Thailand (c. 7th century; see Ślączka 2007: Appendix IV).
The Ratnanyāsa Ritual in the Devyāmata 105
(on which it may or may not be based) and with a more elaborate description
of the ratnanyāsa, could perhaps function as the missing link between the ear-
liest Saiddhāntika treatises and the later, 11th and post-11th century Śaiva texts,
such as the Somaśambhupaddhati and the South Indian āgamas and tantras.
Appendix
80 Only in C the chapters are numbered (B only gives numbers of three chapters). Chapters
not included in C, those whose colophons in C are illegible, and those (a rare case)
included in C but without a number, are numbered by myself, with numbers given in
square brackets. Some chapter names are illegible or there are small differences among
the manuscripts, and at times it is difficult to guess whether we have to do with a proper
chapter ending or with an ending of a section: both may end with the usual ‘flourish’
symbol and some chapters omit the usual iti . . . paṭala ending. The ‘matching’ of the three
manuscripts is time-consuming, partly because the photographs (of B and C) often not
follow the sequence of the palm-leaves.
106 Ślączka
– māsādiparīkṣā 22
– maṇḍapalakṣaṇa 23
– adhivāsanakarmakrama 24
– brahmarekhālakṣaṇa [25]
– liṅgādhivāsanakra[ma] [26]
– piṇḍikādhivāsanavidhi [27]
– bhūtabalipradāna [28]
– daivajñaparigraha [29]
– sruksruvalakṣaṇa [present only in manuscript B] [30?]
– adhivāsana [31?]
– brahmasthānasādhana [32?]
– brahmaśilānyāsa [33?]
– brahmasthānaparityāge anyadeśaguṇadoṣalakṣaṇa [34?]
– brahmaśilāmadhyadik[c]alanadoṣani[r]deśa [35?]
– liṅgapraveśana [36?]
– ratnanyāsa [37?]
– mudrālakṣaṇa [38?]
– anantāsana [39?]
– paramatattvavicā[r]a [40?]
– śaktyāvatāra [41?]
– nādamūrtyavatāra [42?]
– bindvākhyaparijñāna [43?]
– [the sign marking the end of a chapter in A and B, but no iti . . .] [44?]
– [the sign marking the end of a chapter in A and B, but no iti . . .] [45?]
– ādisṛṣṭipaṭala [46?]
– sadāśivatattvā[y/p]avarṇanā [47?]
– śivasvarūpavicārokti [48?]
– anantāsanapraśnavicāra [49?]
– sthāpyamān[e] liṅgaguṇadoṣa [50?]
– dikśānti [51?]
– piṇḍikāsthāpana [52?]
– devyāma . . . paṭalaḥ [53?]
– dakṣiṇā [54?]
– pratiṣṭhākāle karmapaṭalaṃ parisamāptaṃ [55?]
– caṇḍayāga [56?]
– pratiṣṭhātantre gurupūja [57?]
– mukhaliṅgalakṣaṇa [58?]
– sāmānyapratimālakṣaṇa [59?]
– sūrāṇām vividhapratimālakṣaṇa [60?]
– dānavādīnāṃ samudāyalakṣaṇa [61?]
The Ratnanyāsa Ritual in the Devyāmata 107
– lepyabandhalakṣaṇa 101
– bhittibandhana 102
– yojyāyojyalakṣaṇa 103
– bandhalakṣaṇa 104
– dvāravedhadvāralakṣaṇa [105?] [no chapter number given in C]
– gṛhavāstu [106?] [no chapter number in C]
References
Manuscripts
Aṃśumadāgama (Aṃśumadbhedāgama, Aṃśumattantra). Institut Français de
Pondichéry. Nos. T3, T158, T273, T889, T957 and T1070. Paper transcripts in
Devanāgari.
Devyāmata, msA: Devyāmata (Niśvāsākhyamahātantra), NAK MS 1–279, NGMPP A41/15.
Palm-leaf, Newari. Complete. Total number of folios in the bundle: 124; number of
folios giving the text of the Devyāmata: 121.
Devyāmata, msB: Devyāmata (Niśvāsākhyamahātantra), NAK MS 5–446, NGMPP
A41/13. Palm-leaf, Newari. Incomplete. 113 folios.
Devyāmata, msC: Devyāmata (Niśvāsākhyamahātantra), NAK MS 1.1003, NGMPP, reel
number B 27/6. 30.5x4.5cm. Palm-leaf, Newari. 1136 CE. Incomplete. 103 folios.
Dīptādyāgamasya paṭalāḥ. Institut Français de Pondichéry RE 26313. Palm-leaf,
Grantha.
Guhyasūtra of the Niśvāsatattvasaṃhitā. NAK MS MS 1–127, NGMPP A 41/14. Palm-leaf.
Newari.
Kāraṇāgama. Institut Français de Pondichéry T313a. Paper transcript in Devanāgarī.
Kumāratantra. Institut Français de Pondichéry T675. Paper transcript in Devanāgarī.
Mohacūrottara NAK MS 5–1977, NGMPP Reel No. A 182/2. Paper; Devanāgarī; copied
from an old Nepalese palm-leaf manuscript.
Niśvāsatattvasaṃhitā. NAK MS 1–127, NGMPP A 41/14. Palm-leaf. Newari.
Printed Sources
Agnipurāṇa of Mahaṛṣi Vedavyāsa, ed. Baladeva Upādhyāya. Varanasi: Chowkhamba
Sanskrit Series Office, 1966.
Ajitāgama, ed. N.R. Bhatt. 3 vols. Pondicherry: Institut Français d’Indologie, 1964, 1967
and 1991.
Aparājitapṛcchā, ed. Popatbhai Ambashankar Mankad. Baroda: Oriental Institute,
1950.
Bṛhatsaṃhitā of Varāhamihira with the commentary of Bhaṭṭotpala, ed. Sudhākara
Dvivedī. 2 vols. Benares: E.J. Lazarus and Co., 1895–97.
The Ratnanyāsa Ritual in the Devyāmata 109
Secondary Sources
Acharya, Diwakar. 2005. ‘The role of Caṇḍa in the early history of the Pāśupata cult and
the image on the Mathuā pillar dated Gupta era 61.’ Indo-Iranian Journal 48,
207–222.
Banerjea, Jitendra Nath. 1985. The development of Hindu iconography. Delhi: Munshiram
Manoharlal (orig. pub. 1941).
Bhat, Ramakrishna M. 1981. See: Bṛhatsaṃhitā.
Bhatt, N.R. 1964 and 1967. See: Ajitāgama.
Bhatt, N.R. 1972 and 1985. See: Rauravāgama.
Brunner-Lachaux, Hélène. 1963, 1968, 1977 and 1998. See: Somaśambhupaddhati.
Bühnemann, Gudrun 2003. The Hindu pantheon in Nepalese line drawings: Two manu-
scripts of the Pratiṣṭhālakṣaṇasārasamuccaya. Varanasi: Indica.
Dagens, Bruno and Marie-Luce Barazer-Billoret. 2000. Le Rauravāgama: Un traité
de ritual et de doctrine śivaïtes. 2 vols. Pondicherry: Institut Français de
Pondichéry.
Dasgupta, Kalyan Kumar. 1989. ‘The Pāñcarātra tradition and Brahmanical iconogra-
phy.’ In Shastric traditions in Indian arts, ed. Anna Libera Dallapiccola with Christine
Walter Mendy and Stephanie Zingel-Avé Lallemant. Stuttgart: Steiner, 71–91.
Eggeling, Julius. 1989. The Śatapatha-Brāhmaṇa. According to the text of the
Mādhyandina School. Part III: Books V, VI and VII. Delhi: Motilal Bandarsidass (orig.
pub. 1894).
Goodall, Dominic. 1998. Bhaṭṭarāmakaṇṭhaviracitā Kiraṇavṛttiḥ / Bhaṭṭa Rāmakaṇṭha’s
commentary on the Kiraṇatantra. Pondichéry: Institut Français de Pondichéry;
École française d’Extrême-Orient.
The Ratnanyāsa Ritual in the Devyāmata 111
Ślączka, Anna A. 2007. Temple consecration rituals in ancient India: Text and archaeol-
ogy. Leiden: Brill.
Ślączka, Anna A. 2011a. ‘The iconography of the Hindu deities in the Devyāmata, an
early Śaiva pratiṣṭhātantra.’ In Interrelations of Indian literature and arts, ed. Lidia
Sudyka. Cracow: Ksiegarnia Akademicka, 181–261.
Ślączka, Anna A. 2011b. ‘The brick structures of Go Thap—tombs or temples?’ Bulletin
of the Indo Pacific Prehistory Association 31, 108–116.
Ślączka, Anna A. 2015. ‘Dancing Śiva images from Bengal.’ In Studies in South Asian heri-
tage: Essays in memory of M. Harunur Rashid, ed. Mokammal H. Bhuiyan. Dhaka:
Bangla Academy, 125–155.
Ślączka, Anna A. 2016. ‘The two iconographic chapters from the Devyāmata and the art
of Bengal.’ In Tantric studies: Fruits of a Franco-German project on early tantra, eds
Dominic Goodall and Harunaga Isaacson. Pondicherry: Institut Français de
Pondichéry / Ecole française d’Extrême-Orient / Asien-Afrika-Institut, Universität
Hamburg, 181–246.
Smith, H. Daniel. 1978. The Smith Āgama collection: Sanskrit books and manuscripts
relating to Pāñcarātra studies: a descriptive catalog. Syracuse, NY: Maxwell School of
Citizenship and Public Affairs, Syracuse University, 1978.
CHAPTER 5
Marie-Luce Barazer-Billoret
My essay is based on the core of the āgamic corpus, that is to say the twenty-
eight main Śaivasiddhānta āgamas (mūlāgama)—as far as they have been pre-
served up to now—as well as the more or less two hundred ancillary treatises
(upāgama) which supplement them. I will use mainly these two categories of
texts, leaving aside the various manuals (paddhati) which, dealing with similar
topics, may be contemporary with some treatises while being far more recent
than others.
In theory each āgama is to comprise four parts dealing with doctrine, rites,
practices and yoga respectively. As for upāgamas, they are supposed to clarify
some specific topics regarding doctrine, rites, etc. But it happens that they look
like full-fledged treatises (see for instance the well-known Mṛgendrāgama said
to be an upāgama of Kāmikāgama). From a practical point of view, it appears
that it is mostly the parts of the treatises dealing with rites that have been kept.
Without entering here in the whys and wherefores of such a situation and of
the supposed four-section organisation of āgamic treatises, we shall notice that
the care taken to keep the ritual sections emphasizes the importance given to
rites by Śaivasiddhānta followers.
To clarify things it is worth to evoke the context in which these treatises
have been composed and in which they are used up to now. This context,
marked by a religious practice centred on temple worship, leads to a more
or less continuous trend of temple foundations. This is especially true in the
Tamil country, where most of the preserved treatises have most probably
been written. Such temples are to be built and to be used according to very
precise rules. This is well shown by the self-deemed ‘Installation treatises’ or
‘Treatises for installations’ (pratiṣṭhātantra) which are dealing exclusively with
the building of a temple, its consecration and the various ceremonies to be
performed in it. It happens also that a treatise kept in several manuscripts may
be found in some of them, reduced only to the chapters dealing with the topics
listed above on pratiṣṭhātantras (see for instance several examples in
Dīptāgama often deemed in colophons as a pratiṣṭhātantra). Thus, a full
1 See for instance, Dīptāgama Vol. II, Introduction p. 12, and ch. 54.13cd and 32.96cd.
The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas 115
Asia have shown that intermediary deposits were regularly placed at various
levels of the temple elevation, between the foundation and crowning depos-
its, which suggests that corresponding ceremonies were performed. However,
such deposits and ceremonies are not mentioned in the treatises.2
In these treatises most of the images in the installation ceremonies are intended
for temples. They may be iconic or symbolic, that is to say images proper or
symbols such as the liṅga. Amongst them, we find images set for worship in
the sanctum of the temple as well as in its shrines and sub-shrines, and also
the mobile ones used for festivals and processions. Another important group
of images comprises those deemed as ‘decorative,’ placed upon the façades of
the temple and its annexes. In addition, some statues may be located outside
the temple enclosure, as it is sometimes the case of, for instance, the Seven
Seers or Durgā. Thus the goddess Durgā may be placed in a king’s palace or
in a private house. The statue of the king himself may be installed in his own
palace with the same ceremony than the one performed for a deity. Let us add
that installation ceremonies similar to those performed for images may also
concern important worship structures such as the offering altar (balipīṭha). In
this case the Bhūtas, who are in a sense the recipients of the offerings, are only
rarely represented on the altar.
The treatises we are using here are Śaivite. It means that the only truly
essential installation ceremony concerns the main image placed in the main
sanctum of the temple, that is to say a liṅga, which may be a liṅga-with-image,
a Mukhaliṅga. As a matter of fact, the Rauravāgama seems to be the only
treatise saying that an iconic representation, Someśvara, is placed in the sanc-
tum of a Śiva’s temple behind the liṅga. The pattern it describes roughly cor-
responds to what may be seen in Pallava temples, in which the image is that
of Somāskanda, and in Cālukya and Rāṣṭrakūṭa temples, where it is that of a
three-headed Maheśa.3
In all treatises the liṅga installation ceremony is the most important one,
and it is used as a reference for all other installation ceremonies. As we shall
see later, the liṅga installation can be seen as the paradigm of all other simi-
lar installations. Among these other installations we find those of all of Śiva’s
2 See for instance temples of Śriśailam Dam area in Andhra Pradesh, which have been disman-
tled, transferred and rebuilt or various rebuilt or renovated temples at Angkor (Cambodia).
3 See our translation of the Rauravāgama, Introduction pp. xlvi–il.
116 Barazer-Billoret
as well as very technical ones regarding the precise placing of the liṅga in the
sanctum. Such rules show that the material process of building a temple and
of installing its divine inhabitants is very minute and its rules are strictly to be
followed.
4 See Rauravāgama Kp.57.117 in N.R. Bhatt edition, Vol III, p. 114 and ibidem p. 115, fn 16, a quo-
tation of Aghoraśivācārya’s Śivapratiṣṭhāvidhiḥ which follows the same pattern (see also
Brunner 1998: 162).
118 Barazer-Billoret
The first pattern we just mentioned, where the turning of the image into
a god occurs on the installation day, seems to be followed only in the most
ancient treatises. The second one, where that change occurs during the sojourn,
appears in more recent works that are also marked by other differences in the
description of the installation ceremony. Thus they propose various ways to
install the god in the image (see my 1993–1994 paper), which can be perceived
as chronological clues; one may also notice differences in the placing of the
various deities in vases as well as in the performing of fire ritual and of corre-
sponding purifications. Such variations, which are many, may be used to locate
the time period in an āgamic text and, what’s more, to appreciate its internal
coherence. However such indications are not to be taken as an absolute crite-
rion: for instance a quite recent text may prescribe a ritual pattern as described
in older texts (see for example below, Dīptāgama about images). That being
said, the basic pattern of installation is always similar for the liṅga.
than those to be applied to the installation of the liṅga. However such ‘sim-
plification’ may also lead us to be cautious about the notion of ‘pattern’ or of
‘model’ that we have just presented: in other terms, the question is to know if
the differences we may note between the installation of a liṅga and that of the
images allow or not to maintain that the liṅga ceremony is the basic model
for the ones performed for images. We shall clarify our remarks by using the
Dīptāgama dual presentation of those installations as an example.
According to the same Dīptāgama (chapter 20), during the ‘sojourn’ cer-
emony the liṅga is to lie upon a couch with the head to the East; there sev-
eral mantras are imposed upon it: first brahmamantras, aṅgamantras and
kalās, then twice the mantras of the three tattvas then those of the faces of
the five mūrtis (mūrtivaktra), and lastly Netramantra and the ‘limbs of Science’
(vidyāṅga). Then is performed the classical list of rites directing the god to
come into the liṅga, invitation, presence, etc. The liṅga being now turned into
the god is offered holy waters, a flower, incense, cooked food and betel leaves.5
The next day, of the installation itself, the imposition of mantras is repeated,
and the germs (bīja) are transferred from the vases to the liṅga.
According to the numerous chapters of the Dīptāgama dealing with Śiva
manifestation installations, it appears that like the liṅga, during the sojourn
ceremony, each image is lying on a couch, with its head facing upward and to
the East; it is then dressed with a red cloth, but nothing more is done, espe-
cially no imposition of mantras. On the next day, the image is taken up from
the couch and placed upon the bathing altar where the installation proper will
take place. On the bathing altar the image is put in the vat and it stands facing
east; there the thirty-eight kalās as well as the brahma° and aṅgamantras are
imposed upon it. Then the image is worshipped and it is at that time that its
installation (sthāpana) is performed: the germ-syllable (bīja) of Śiva is trans-
ferred from the Śiva vase to the heart of the image while the germ-syllable of
the Goddess is transferred from her vase to the pedestal of the image, and lastly
the water contained in the vases is poured upon the image which has now
been turned into Śiva. This Śiva is then placed in the pavilion where he is wor-
shipped and, later on, an utsava is performed which—due to the ambiguity of
the term utsava—may be a festival as well as just a procession.
To summarize the teaching of Dīptāgama, let us say that, as far as the liṅga
is concerned, the transformation into Śiva starts during the sojourn ceremony,
while that ceremony, when it comes to an image, is no more than a sanctifica-
tion of that image. That seems to make quite a difference and one may wonder
what to conclude from it. Does it mean that the treatise has been written in
5 Dīptāgama I.20.130ab–133cd.
120 Barazer-Billoret
two different periods? Or else that the most important ceremony regarding the
liṅga, its turning into a god, is extended by making it to start a day earlier, while
such an extension is considered as unnecessary for the image installation? It is
difficult to decide, as the treatise looks very homogeneous regarding other ele-
ments of the ceremony such as for instance mantras or oblations. Whichever
is the case, the possible co-existence of two models—one for the liṅga and the
other for images—leads us to wonder about the unity of the treatise and may
be a useful tool when trying to understand its making-of.
Lastly, even if our essay is mainly based upon canonical literature, the
paddhatis cannot be excluded: as a matter of fact, amongst the patterns that
underline an obvious evolution of installation ceremony rites we can found
the one proposed by the Somaśambhupaddhati and also used in some āgamas,
for instance Kāmikāgama.6
The Somaśambhupaddhati was ascribed by Mrs. Brunner to the end of the
11th century (1095), and when a little later Aghoraśivācārya wrote a treatise on
installations, he borrowed the Somaśambhupaddhati’s pattern. It is that pat-
tern which nowadays is used in South India for performing installations. In fact
it is very likely that āgamas which in their present stage follow the very same
pattern have borrowed it from either Somaśambhu’s or Aghoraśivācārya’s
works, which adds one more element when looking at the relative chronology
of the āgamic literature.
References
Sanskrit Texts
Ajitāgama, édition critique par N.R. Bhatt, 3 vol. Pondichéry (PIFI n°24), 1964, 1967,
1991.
Kāmikāgama I, ed. C. Svāmināthaśivācārya, Madras 1975.
Kiraṇāgama, Devakottai 1932.
Mṛgendrāgama (Kriyāpāda et Caryāpāda) avec le commentaire de Bhaṭṭa
Nārāyaṇakaṇṭha, édition critique par N.R. Bhatt (PIFI n° 23), Pondichéry 1962.
Rauravottarāgama, édition critique par N.R. Bhatt, Pondichéry 1983 (PIFI n°66).
122 Barazer-Billoret
Ronald M. Davidson
Introduction
In distinction to most of the papers in this volume, which examine the place-
ment of permanent images in well-established and affluent temples by
high-caste priests, the object of analysis in this paper is one iteration of the
phenomenon of temporary altars and folk rituals, as found in the Buddhist
archive. For some time I have been interested in the fourth to seventh century
Buddhist ritual texts, those that contain the dhāraṇī rituals, which preceded
the formation of actual Buddhist tantrism.1 The dhāraṇī texts may be either
relatively simple or quite complex, and their complexity becomes understand-
ably more robust as the centuries progress through the sixth to eighth century,
after which we see a diminution of the genre, as it becomes marginalized by
tantric Buddhist scriptures from the ninth century forward. However, from
the fourth to the seventh century, dhāraṇī Buddhist texts constitute our best
record concerning the involvement of Buddhist monks and laity with the spec-
trum of Indian ritual systems, even if dhāraṇī texts must be supplemented by
statements elsewhere inside and outside the Mahāyāna and Vinaya records.
The association of Buddhist communities with the nāga cult and serpent
worship has already been detailed in the work of many scholars over the previ-
ous two centuries, since the pioneering work of Burnouf in his translation of
the Saṃgharakṣitāvadāna (1844: 315–335) and of Fergusson in his study of the
archaeology of Sāñcī and Amarāvatī (1868). The more recent efforts of schol-
ars have improved on previous work and provided us with a good series of
models on the interrelationship between the monastic institutions and Indian
snake divinities (Vogel 1926, Härtel 1976, Rawlinson 1986, Bloss 1973, Shaw
2013, Strauch 2014). An increasing body of work on the larger sphere of such
autochthonous divinities has revealed that the nāgas were a specific itera-
tion of a larger interaction, in which Buddhist monastic communities placed
1 My interest has yielded studies concerning dhāraṇī semantics, Davidson 2009, pragmatics,
Davidson 2014a, and canon modeling, Davidson 2014b.
Taishō Tripiṭaka no. 1007 is a medium-length text, but its length belies its dif-
ficulty, in great part because of the nature of the translation terminology and
irregular phonological transcription systems employed. It is a complex text,
with several small ritual procedures invoked, which is commonly found in
dhāraṇī literature of the period, but goes beyond others in its richness and
detail. Neither the title nor the translation dating is entirely secure, but both
may be proposed with some degree of confidence. What makes the text inter-
esting is, like other anonymously translated texts attributed to the Liáng period
(502–557 CE), it incorporates into the work instructions not found in other ver-
sions of the same text. A careful comparison to the Gilgit fragments, Bodhiruci
II’s 706 CE translation and the Tibetan imperial translation indicate that the
anonymous Indian paṇḍita who brought the earliest text to China included
performative explanations into the translation, the kind that would have been
passed down outside the text and articulated as part of the actual instruction
in the ritual. These performative instructions are commonly communicated
informally as person-to-person directions in Buddhist ritual, but texts whose
translation is attributed to the Liáng period are somewhat exceptional in
including those instructions in the translation. Their testimony suggests that
other surviving medieval ritual texts may in fact be shorthand summaries of a
larger ritual program that was of necessity more orally explicit and extensive.
That is, contrary to what we see in other forms of Buddhist literature, this
text is longer and more robust in its directions than later translations of the
same material. Where we see the overlap with the Gilgit manuscript, it is
126 Davidson
relatively clear that the other translators were more attentive to the task of
translation per se, whereas our text incorporates material from other sources
into the Chinese rendering. Some of the supplementary material simply rep-
resents elaborations on the ritual, to make the directions more complete.
Other parts of the translation incorporate other textual or oral materials
not necessary for the rituals involved, but which constitute general or back-
ground information. In this respect, the *Mūlamantra is similar at least to the
Saptabuddhaka said to have been translated around the same time. As I have
shown elsewhere (Davidson 2014b), the Liáng Saptabuddhaka is more com-
plete than any of the other translations of that work, and frequently provides
sufficient ritual instruction so that individual rites can actually be performed.
This is in distinction to the other translations, which—doubtless mirroring the
abbreviated nature of their originals—do not provide sufficient instruction.
As it is, the *Mūlamantra is an extremely important moment in dhāraṇī
texts. Not only does it prescribe the earliest complete nāga rain ritual—as
opposed to simply praying for rain—it is reputedly the first Buddhist text to
specify mudrā instructions, and provides the earliest description of certain
other ritual systems as well. It provides a sufficiently elaborate painting pro-
gram that Matsumura, the editor of the Gilgit fragments, described it as a “text
on esoteric iconography,” which is not an adequate description as so many
other ritual events are discussed. The *Mūlamantra demonstrates the con-
struction of an idiosyncratic, early maṇḍala, discusses an elaborate description
of signs and omens, especially concerning the homa ritual. Related to the nāga
rain ritual, it is the earliest Buddhist text to describe the ritual propitiation of a
vināyaka, and describes a ritual first attested in the avadāna literature on con-
trolling a nāga. It also demonstrates a specific kind of abhiṣeka ceremony, one
that ties it firmly to the domestic ritual syllabus (gṛhyasūtra). The extensive
ritual program requires a complete translation and study of the text, which is
under preparation.
For all of its having initiated new directions, the actual name of the text
is something of a curiosity. As mentioned above, in the modern period it has
been recently identified as the Mahāmaṇivipulavimāna-viśvasupratiṣṭhita-
guhyaparama-rahasya-kalparāja-dhāraṇī, but this identification is in need
of reexamination. Mahāmaṇivipulavimāna is the title of the long, primarily
mythological, work attached in India mostly to the front of the *Mūlamantra
prior to its retranslation as an aggregate work in 706 CE by Bodhiruci II
(T. 1006). The aggregate text was again translated by Amoghavajra, probably
between 746 CE and 774 CE (T. 1005a), and he included siddhamātṛka tran-
scriptions of most of the mantras. The Gilgit fragments edited by Matsumura
indicate that some version of the text circulated in the fifth-sixth century,
A Nāga Altar in 5th century India 127
2 Vira and Chandra, Gilgit Buddhist Manuscripts 1724.5, 1726.5, 1727.2, 1730.4; see Matsumura
for an edition; cf. T. 1005a.19.(transcribing maṇḍala as 曼荼羅): 625b20, 627b21–2, 630a18,
630b20, 632b17, 632b24, 633a19, 633b7, 633c10; T. 1006.19.(transcribing 壇/壇場): 639c20,
641b17–23, 641c15, 642b12–17, 642c19–643a8, 643b27–644a15, 644c9, 646a16, 646b6, 646b25,
646c7, 646c26, 647a23, 647b12, 648b1, 648b26, 649a6, 649b26; T. 1006.19.(道場): 643b8–10,
644a16–25, 646a2, 647b4, 647c29; T. 1007.19.(壇): 658b2–23, 659b28-c2, 660a11, 660b7, 660c6,
661b28, 663a8, 663a18, 664c4, 665a1, 667a25–668b12.
3 Amoghavajra, T. 1005a.19.624a19: 根本陀羅尼品; Bodhiruci II, T. 1006.19.640c5: 根本呪品;
To. 506.fol. 300a3/Tog. 468, fol. 449b3: rtsa ba’i rigs sngags. A similar gloss is given in the
Fānfànyǔ, T. 2130.54.1008b7, 1020c11.
4 Mahāmāyūrī-vidyārajñī, p. 9.12; T. 987.19.480a4. We note that there are other instances of
牟梨 being used in mantra transcription, most notably in the mantras of Krakucchanda,
which may also transcribe mūle; see Qīfǒ bāpúsà sǔoshūo dàtúolúoní shénzhòujīng,
七佛八菩薩所說大陀羅尼神呪經, T. 1332.21.537a28–b3, *Dhāraṇīsaṃgraha, 陀羅尼雜集,
T. 1336.21.581c1–5, 601b20–22; see Davidson 2014b.
128 Davidson
5 Pulleyblank 1991, pp. 186, 188, provides 梨 and 利 approximately the same value, li.
A Nāga Altar in 5th century India 129
6 On the problem of the Chinese canonical catalogues, see Tokuno 1990.
7 Storch 2016, pp. 114–117 discusses the position of dhāraṇī literature in Liáng catalogs. I thank
Professor Storch for our consultation on this question.
130 Davidson
homa, etc.; these belong to the larger Indian literature on signs and must be
considered in conjunction with those many other sources.
i. [658b1] 若欲祈雨。取青牛糞泥作一龍。一身三頭朱染脊背金裝胸
臆。
If you want to pray for rain, take the manure from a blue cow [and] mud
and make [an image of] a nāga: it has a single body with three heads,
vermillion on the back and gold ornamented on the thorax.
First make a square platform. If there is a high tower with a level area,8
then make it [there] at an appropriate time. Furthermore, paint this plat-
form with blue-green [cow dung], and on top of the platform place the
nāga [image]. In each of the four corners of the platform place a water
jug and an incense brazier. In one burn frankincense (Boswellia sacra), in
one sandal incense, in one storax resin incense (śaileya, Styrax officina-
lis), and in one bdellium incense (guggulu, Commiphora wightii).
Then on the outside of that platform but connected to it, make another
platform [surrounding the first, the mekhalā], four cubits square, and
smear it over with cow dung. In the four corners [of this second plat-
form], place pots. One contains water mixed with milk, one contains
water mixed with curds, one contains water mixed with milk-rice, and
one contains water mixed with fat.
Then place in the center platform eight burning oil lamps. Again, with
jiăo-jí-shí-lí 9 and grain, decorate this platform. Now in each of the cor-
ners, insert an arrow. Then attach a five-colored thread to the top of each
arrow, [making a perimeter to the maṇḍala]. Then make five-colored
banners suspended from the tops of the arrows.
v. [658b11] 又將七種穀散其壇内又用五彩莊飾供設壇上。又以當時所
有種種花果盡設壇内。
Again, take seven varieties of grain and strew them on the inside plat-
form, and employ five[-colored] variegated ornaments set out as offer-
ings arranged on top of the platform. Then with the various kinds of
flowers and fruit appropriate to the season, set them completely [cover-
ing] the inner platform.
9 Uncertain identity. The interlinear gloss (紅藍花) indicates the safflower (Carthamus tinc-
torius), which is normally equivalent to kusumbha in Sanskrit, but this clearly is a poor fit
with the phonology. Could this be the kiṃśuka flowers (Bhutea frondosa) mentioned in
Vārāhagṛhyapuruṣa 4.1, as part of the sarpabali?
132 Davidson
If it has rained too much, then sanctify white mustard seeds [saying the
mantra] a full eight hundred times and throw them into a body of water
wherein a nāga resides. The rain will immediately stop. At this time you
are not permitted to speak [ordinary speech] . If while you are praying for
rain there should occur an unusually swift cloud, a violent wind, a painful
mist or unexpected hail, then make a peg out of khādira wood and insert
it into the side of a nāga ravine (*śvabhra). Then all the obstacles and dif-
ficulties will entirely be scattered away.
If, while one is praying for rain, and even having averted obstacles, still
the rain does not descend, then in the middle of a platform, one should
draw a vināyaka. Then one should take eight hundred grains of white
mustard seed. Saying a mantra for each mustard seed, he should hit that
vināyaka. Having completed the full eight hundred, he is able to ensure
that all those [spirits] creating obstacles will themselves be bound and
no longer can cause obstruction, so that the rain will descend. With the
matter concluded, the vidyādhara should take milk and pour it over the
image, which will make the vināyaka very happy. From this time on,
whatever one seeks for, will happen according to one’s desire. The
vidyādhara’s body and clothing are always all to be pleasing and clean.
10 The mantra is reconstituted from the following texts: Siddhamātṛka T. 1005a.19.624b3–7,
T. 1005b.19.634b24–28; Derge Tibetan transcription To. 506, fol. 299a6–7; Chinese tran-
scriptions are T. 1005a.19.624a22–28, T. 1006.19.640c9–14, T. 1007.19.658c2–7. The Tibetan
A Nāga Altar in 5th century India 133
The other three complete texts have variations on this rain ritual, albeit none
so extensive.11 Part of the difference in the ritual treatment is simply editorial:
where the *mūlamantra is introduced and how it relates to what precedes or
follows. In T. 1007, the mantra is placed as above, at the conclusion of the sec-
tion identified as concerning the practice of this mantra (*mūlamantravidhi).
In distinction, the aggregated 706 CE translation of Bodhiruci II and the
Tibetan of Vidyākaraprabha and dPal-brtsegs both introduce the *mūlamantra
prior to the section in which it is to be explained, and in neither translation
is the mantra even identified as the *mūlamantra, thus leaving the “practice
of the *mūlamantra” without a mantra to be practiced, a peculiar situation
to say the least.12 The situation is handled better in Amoghavajra’s transla-
tion (T. 1005a.19.624a22–27), where the *mūladhāraṇī is identified directly
as such, and is included in the section specified as conveying its practice
(根本陀羅尼品: as if *mūladhāraṇīpaṭala). Moreover, Amoghavajra’s text also
includes (in the Taishō Tripitaka) the mantra in siddhamātṛka characters after
the Chinese section, so that the mantra is presented in two positions, one ear-
lier in the text in Chinese transcription and one siddhamātṛka text in a posi-
tion similar to T. 1007; additionally, Amoghavajra’s supplemental work T. 1005b
contains siddhamātṛka versions of the mantras in his translation. The other
difference between the translations, however, is in the detail of the ritual, and
as we explore the background of the text, some of the major differences of the
respective translations will be identified.
In constructing this ritual text, the Buddhists were not spinning their perfor-
mances of whole cloth, but—like virtually all of Indian religion—this work
is a pastiche of pre-existent systems which were modified, hybridized, and
brought into conformity with each other under the aegis of the specialist, in
our case a Buddhist ritualist. Through this process, they were addressing the
Indian horizon with specific requirements in mind, so that the modification
of the Buddhist ritual syllabus was a comprehensible response to the needs
of their time. Searching to construct their institutions within the tableau of
Indian life, they made a strong place for Buddhist ritual systems in select areas,
and nāga-related weather rituals constituted one of the more important of
these procedures. In the event, the Buddhists brought together at least three
other Indian streams of religious behavior, and formulated them in the evolv-
ing environment of the Buddhist dhāraṇī literature. We will examine them in
order: A. local or folk cultic sources, B. Gṛhyasūtra sources and C. vidyādhara
sorcerer sources. Following these, we will examine the path not taken by
Buddhists, brahmanical rain and nāga rituals, and finally revisit the Buddhist
antecedents and their contribution to the *Mūlamantra formulation.
於池中取瞿摩夷和土為泥揑作一龍。腰已上為菩薩身。作菩薩面。於
其頭上出三蛇頭。腰已下為蛇身。於池中盤屈。其龍遍身以黃丹塗令
作赤色。以金薄貼龍心上。T. 1005a.19.625a9–13.
In the center of that [painted] pond, take cow dung (gomaya) and mix it
with earth to make mud, molding and fashioning it into a nāga. From the
waist up it is to have the body of a bodhisattva, and give it a bodhisattva’s
face; above the head, emerge three snakeheads. From the waist down,
make a snake body, coiled into the [painted] pond. The entire nāga body
is colored red by being daubed over with massicot [paint], with gold foil
over the nāga’s heart.
So, while some of the details are different, none identify this as anything other
than a temporary image, distinct from the stone images of nāgas so often dis-
cussed in the archaeological record. The survival of unbaked mud images—
not to speak of cow dung images—is virtually impossible in India, especially
A Nāga Altar in 5th century India 135
after several monsoon seasons in the Gangetic Valley. Somewhere within this
area may be the presumed source site of the text, for a variety of historical
reasons: Buddhist history, artisan traditions, our knowledge of texts from this
period, the gṛhya traditions found in the text, etc.
We find verification of the early use of mud images in the delineation of an
seasonal rite in the Nīlamata, where one is to worship piśācas by making an
image of mud, or mud with leaves:
Other prescriptions for mud images, generally of folk sources, are occasionally
met in Indian documents. The Brāhmaparvan of the Bhāviṣyapurāṇa mentions
the construction of thumb-size images of Gaṇapati, first made from various
kinds of wood, and then out of clay taken from the tusk of an elephant, the
horn of a bull or from an anthill.13 We find mention of images of wood and
mud in the Bṛhatsamhitā, which attributes them diverse benefits.14 Similarly,
mud images are specified in some earlier Buddhist texts, like the Śrāvakabhūmi,
which mentions images of wood, stone and mud.15
These descriptions are not entirely clear, however, since we are generally
not informed whether the statues are small or large, baked in the sun or fired
in a kiln, or whether the ‘mud’ is worked in a combination of quick-lime and
sand into stucco. Such distinctions are extremely important, for Varma’s exam-
ination of the various stucco formulae and manufacturing techniques suggest
that stucco requires a greater resource base and the involvement of dedicated,
professional artisans (Varma 1983, 1987). In distinction, rudimentary mud or
13
Bhāviṣyapurāṇa 1.30.1: hastidaṃtamṛttikāmayam aṃguṣṭaparvamātraṃ gaṇapatiṃ
kārayet . . . vṛṣabhaśṛṃgamṛttikāṃguṣṭamātraṃ . . . valmīkamṛttikāṃguṣṭamātraṃ. Similar
statues are described in Bodhāyanaśeṣasūtra II.13, and following.
14
Bṛhatsamhitā 59.4: dārumayī mṛnmayī tathā pratimā. A somewhat different situation is
found in the manufacture of a dust or sand image (pratikṛtiṃ . . . pāṃsumayīṃ) of a per-
son for destructive rites in Ṛgvidhāna 3.81–82.
15
Śrāvakabhūmi, p. 416.7: kāṣṭhāśmaśādakṛt.
136 Davidson
16 Kramrisch 1939, pp. 95–97; Gupta 1965, pp. 223–4; Agrawala 1948, pp. 26–27; Ray 2006,
pp. 84–89; Kala 1980, pp. 66–68; Lal 1993, p. 129; Härtel 1993, pp. 120, 125, 155, 163, 195.
17 See also Mañjuśriyamūlakalpa, Śāstri 62.1: manuṣyākārārddhasarpadeha.
18 Kala 1980 pp. 66–67, nos. 347–349; To. 506, fol. 301a3: lus ni li khris byug par bya’o.
19 See Einoo 1994 for his editions; most of his references hold across editions, but
Bhattacharya’s edition of the Garuḍapurāṇa has them at 1.129.23–28.
A Nāga Altar in 5th century India 137
The day especially devoted to its (nāga) worship is the fifth day of the
bright half of Shrāvan, which is called Nāg panchami. In some places Nāg
panchami is observed on the 5th day of the dark half of Shrāvan. On this
day in image of a snake is made of cowdung or earth, or its picture is
drawn on the wall. (Enthoven 1914: 138)
After the Diwali in Kangra, a festival is held to bid good-bye to the snakes,
at which an image of the Naga made of cowdung is worshipped. (Crooke
1896, 2:138)
In Garhwal, the ground is freely smeared with cow dung and mud, and
figures of five, seven, or nine serpents are rudely drawn with sandal-wood
powder or tumeric; rice, beans, or peas are parched; lamps are lighted
and waved before them; incense is burnt and food and fruit offered.
(Atkinson 1882–86, 2: 836)
Many other instances could be cited, but in none of these situations is a brah-
man necessary, nor is a temple visit required, nor are any of the other brah-
manical caste-related functions invoked. Other modern domestic festivals,
such as Lakṣmīpūjā or Ganeś Cathurtī, also have a folk-based component, and
make use of the availability of small clay images that are procured for the tem-
porary construction of an altar.20 In such instances, the procedures are acces-
sible to all irrespective of caste.
20 Some editions of the Garuḍapurāṇa (e.g., Nirnaya Sagar Press ed. 2.12.77) identify the
manufacture of a golden image of a nāga in the case of a person dying from snakebite, to
be given to a brahman; see Abeg 1956, p. 168 for the section. However, in other editions
(e.g., Bhattacharya’s) the verses are not traceable, and so the authority of this offering is
questionable.
138 Davidson
The position of non-orthodox priests in charge of nāga shrines was also the
observation of Vogel (1926: 248–49, 252) in the shrines he surveyed, and like
many such temples, these shrines feature individuals answering questions in a
state of possession, in this case by the nāga spirit.
21
Mānavagṛhyasūtra 2.16.6: tūṣṇīm api kṣudrā prakṣālitapāṇiḥ; Dresden translates, ‘Even a
lesser caste woman could perform the sarpabali, if she washes her hands and remains
silent (i.e., does not recite the mantras).’
A Nāga Altar in 5th century India 139
expectations of the twice born, so that, even if their status was low, nonethe-
less their presence was ubiquitous and their influence profound.
As is well known, the texts of the gṛhyasūtras prescribe the daily, monthly
or annual obligatory rituals (nitya), the intermittent rituals of passage (naimit-
tika), and optional rites (kāmya) as needed by the family or clan. Because this
last category, however, is as elastic as the concept of desirability, various sup-
plemental rituals were integrated into the optional rubric as these appeared
advantageous to the ritualist or his patron. Some of these additional rituals
were simply reductions of rites already handled at the śrauta level—the con-
secration of a king, the performance of the evening agnihotra, etc. But many
others were entirely new, and represented the ritualization of many aspects of
ordinary existence: the blessing of cattle or horses, the production of personal
benefit, birth of sons, ritual for remarriage, the pacification of a problem, to
mention but a few. Included in these became the propitiation of various gods
and local spirits, employing the newly emerging strategies of pūjā offerings,
whether strewn (bali) or fire (homa). Einoo has correctly argued that these
were developed in part from the madhuparka guest rituals and in part from
elements outside the Vedic tradition.22 We might also observe that guest ritu-
als change from locale to locale and evolve over time, so that various regional
styles eventually became formalized into specific pūjā practices as well; this is
what we see in the later gṛhya rituals generally, which lacked the absolute uni-
formity often envisioned for the Vedic tradition, a uniformity it did not actually
display.
In our instance, we find many of the items specified in the later gṛhya texts,
and the altar on which the nāga is positioned is the most important. While it is
called a ‘maṇḍala’ in Buddhist works generally, it is developed out of the older
gṛhya altar, most often called a sthaṇḍila, although the designations maṇḍala
or vedika are sometimes used.23 Curiously, the earliest Buddhist maṇḍalas
show affinities with the sthaṇḍilas of the Yajur and Sāmaveda gṛhya traditions,
for that is where similar locations of pots and altar layout occurs, but Ṛgveda
supplementary ritual procedures are also apparent. A few examples may suf-
fice, although they could be multiplied extensively.
22 Einoo (1996: 80–81) indicates six stages in the development; see also Bühnemann 1988:
32–34.
23 Einoo (2005: 24–41) has argued for the relationship between maṇḍalas and the sthaṇḍila,
and we can see in our data that this is true of rudimentary Buddhist maṇḍalas as well,
although other factors come into play at various stages.
140 Davidson
Having formed [an altar] in the shape of a vedi to the north of the village,
having covered it with branches and unwashed cloths, having filled pots
with all the flavors, let them be put into the directions and leaves filled
with all the seeds placed in the intermediate directions.
Dresden’s translation: After having gone out, before sunset, to the north
or to the east of the village and having prepared on a clean spot, under an
Aśvattha (Ficus religiosa) or Nyagrodha (Ficus indica) tree, or in the
neighbourhood of water, a kind of altar, and having made ready on this
(altar), which should be quadrangular and provided with branches of for-
est trees, which should be hung with strips of (coloured) cloth, which
should be full of perfumes, wreaths and garlands, and should be provided
with a multitude of untwisted (?) white garlands, upon which should be
placed, facing the four quarters of the earth, jars (filled with water) and
baskets filled with a mixture of grains (of rice, barley, etc.) and pieces of
gold, which should be full of flour-cakes, layers (of grass), baked grain,
pastries (?), auspicious objects (amulets, etc.), fruits and unhusked bar-
ley-grains, all kinds of perfumes, all kinds of juices, all kinds of herbs, all
kinds of jewels . . .
The section continues on, but the robust nature of the offerings and the layout
of the altar may be favorably compared with both our own nāga ritual and
other rites evinced in classical Mahāyāna dhāraṇī texts.
The earliest recorded Buddhist nāga rites demonstrate the Buddhist appro-
priation of the sthaṇḍila altar from the fourth century on, if the early fifth
century translations are any guide. Perhaps the earliest were the altars in the
very short ritual instructions that followed the actual text of the Mahāmāyūrī-
vidyārājñī, where we find in a 402–412 translation attributed to the Kuchean
monk Kumārajīva a very brief description:
24 Since so many gṛhya texts specify a sthaṇḍila and since many identify them with the
maṇḍala form, it is a bit difficult to choose, but important statements are also found in
Bodhāyanagṛhyaśeṣasūtra 1.23.6; Āgniveśyagṛhyasūtra 2.6.7; Vaikhānasagṛhyasūtra 2.1.
142 Davidson
孔雀王呪場。用牛屎塗地。用散七色華。幡四十九枚刀四枚鏡四枚箭
一百枚弓一張瓨七枚盛漿。黒羊毛繩十六尋。薄餅二十五番。然七油
燈。酪一器糗漿一器。飯一器薄餅一器。安石榴一器華一器。
Other examples could be multiplied, but there is little in any of the early
Buddhist ritual instances that stands outside of or apart from the contem-
porary use of sthaṇḍilas in the domestic context within rites of passage or
optional rituals, ones that confer status or other benefit on the patron. In all
of these, the altar is rudimentary, positioned on the ground but often raised a
few inches, smeared with cow dung, squarish/round in shape, and decorated
with pots of water, lamps, flags and related items. During the Gupta era, ritual
offerings were amplified in number and in kind, but seldom in performative
function—they lend authority and validity to the altar. The only item regularly
included in Buddhist offerings that is seldom seen in the standard domestic
sthaṇḍila are images, since image use came to be employed rather late in brah-
manical ritual systems. The primary exception to this were the small images of
the planets, made of various materials, found in many rituals on the pacifica-
tion of the grahas, the seizers who possess and bring disease (Yano 2004).
25 Taishō 988.19.484c4–8, attributed to Kumārajīva, 402–412 CE; reading 屏 píng for 餅 bĭng
in 484c6. Other analogs: T. 984.19.458c13–26 [Samghabhara 502–520 CE]; T. 985.19.477b15–
29 [Bodhiruci II 705 CE]; T. 982.19.439b4–11 [Amoghavajra 746–774]); in some Chinese
translations, this is a rite for the transmission of the mantra/text rather than for the culti-
vation of the mantra as here.
26 The most important contribution on vidyādharas to date is Graf 2001; see also van
Buitenen 1958.
A Nāga Altar in 5th century India 143
27
Bhaiṣajyavastu pp. 124–130; Divyāvadāna introductory story to avadāna xxx, pp. 435–442;
Bodhisattvāvadānakalpalatā chapter 64, the Sudhanakinnaryavadāna, studied by Straube
2006; part of this story is translated from the Divyāvadāna in Vogel 1926: 184–86. On the
overall textual distribution of the story and its Borobudur connection, see Jaini 1966.
144 Davidson
did not provide the necessary moisture for crops. The cruel king is informed
that a snake-charmer (āhituṇḍaka) who, as a vidyāmantradhārin, is capable of
this by means a ritual of binding. The description of the ritual is connected to
our own.
He went to prepare for the ritual of food and strewn offerings, and on the
seventh day will return. Having returned, he will stick in pegs of khadira
wood into the banks of the [nāga] pool in the four directions, and, having
tied them up with strings of different colors, he will invoke the mantras.
(Bhaiṣajyavastu 127.17–19)
Unlike this one, our nāga-control ritual does not involve the construction of
a border (sīmā), as this kind of binding came to be called, but otherwise the
use of a khadira (acacia catechu) peg (kīlaka) is similar in its magical ability to
control the nāga. And like ours, the Janmacitra in this story retains traces of
the vernacular landscape, for Janmacitra is called a nāgapota, a Prakritic form
of ‘nāga son’ (nāgaputra), and the overall nature of the story is one of magical
sensibilities outside of the dynamics of elevated religion. We also see the divi-
sion of powers between the nāga as a bringer of rain in times of drought and
a rain god who is to provide rain on a regular basis, responding to the virtue of
the ruler; thus, the myth articulates two sources of rain, a point to which we
will return below.
Other rituals involved in domination employ a khadirakīlaka, and these may
be mentioned. Of particular interest is the killing ritual found in various sources,
such as at the end of the Gaṇapati propitiation rite in the Bhāviṣyapurāṇa
(1.39.32).
Now if you want to kill, then having made a peg of khadira wood, and
having imagined the man or woman, then pierce [that person] in the
heart, and in a moment, he will die.
The brevity of the rite as given here is interesting, since the sādhaka is not
even required here to invoke the god, say a mantra, or do much of anything
A Nāga Altar in 5th century India 145
at the time of the performance. We would even expect some kind of effigy,
but that is not part of the instruction. And we have little sense that this is
other than a rite appended to Gaṇapati as a floating rite that serendipitously
was written down. That perception is validated by the use of a khadira peg
(khādiram śaṅkum) stabbed into an effigy of dust or sand in the Ṛgvidhāna
(3.112–113) to destroy one’s enemy. We also find that the Śuklayajurvidhāna
(p. 4.4) instructs the reader to make twelve-inch pegs of either iron or khad-
ira and stab or bury it in the house of the enemy for his assured destruction.
The relationship between iron and khadira is also specified in the Ṛgvidhāna
(3.78), where one wards off rivals by throwing iron pegs into a khadira homa.
The Amoghapāśahṛdaya additionally contains a khadira-kīlaka ritual—much
more irenic as it is dedicated to the pacification of an area—and there the pegs
are employed to perform a protection boundary (sīmā); the pegs are joined
together by string of various colors, similar to the circling of Janmacitra’s
pond.28 Indeed, the Karmapradīpa provides a description of such stakes:
Our ritual further describes (viii.) the domination and satisfaction of the
vināyaka demon if the nāga ritual is not effective. If the nāga propitiation is in
part based on Smārta pūjā—itself grounded in the guest-greeting rituals and
manners of various areas and schools—is not necessarily true of many of the
rites associated with the agency of the vidyādhara sorcerers. Some of the rituals,
like the control rite we just discussed but unlike the pūjā to the nāga, are con-
cerned with forceful domination. In this vināyaka instance, there is little sense
of our vidyādhara propitiating the deity in advance. The spirit is summoned,
forcibly if necessary, told what to do or forced to abide by the vidyādhara’s will,
and only propitiated with milk after his successful performance.
In terms of a demonic group that assisted the eventual formation of the
Gaṇeśa persona, this is evidently the earliest description of a ritual for con-
trolling a vināyaka in Buddhist literature. The Buddhist version is a survival
28
Amoghapāśahṛdaya, Meisezahl 1962: 325. I have translated this ritual and discussed some
of its import in Davidson 2011. We note the magical properties of khadira may also be
beneficial, and Ṛgvidhāna 4.83–85 provides a prescription for a patient suffering from
consumption (yakṣma) to drink powdered khadira in water, with honey and ghee.
146 Davidson
And here is the curiosity, one that is important for the study of classical and
medieval Buddhist ritual. By the time our ritual is translated into Chinese, the
vināyakas appear to have already been amalgamated elsewhere into the rap-
idly evolving Gaṇeśa persona, so that the Yajñavālkyasṃrti—while acknowl-
edging that there are four vināyakas—affirms that there are in fact only one
who is ordained as the lord of the gaṇas by Brahma and Rudra to create and
disperse obstacles (I.270: vināyakaḥ karmavighnasiddhyarthaṃ viniyojitaḥ |
gaṇānām ādhipatye ca rudreṇa brahmaṇā tathā). This is a well-known classi-
cal brahmanical strategy: the declaration of unity amidst bewildering plurality,
facilitated by assigning multiple local divinities to one of the pan-Indian dei-
ties of the Sanskritic archive. But that is not the only interesting feature of our
ritual. Perhaps the most important aspect is that, when earlier religious tradi-
tions are incorporated into Sanskritic literature, we observe that they do not
suddenly cease or evaporate from the religious landscape, as is so often pre-
sumed. The presumption is informed by the idea that high-status cults always
displace, appropriate or redefine low-status ritual systems, whereas we actu-
ally observe a complex series of community processes too often glossed with
the hazy and imprecise Indological discourse of ‘absorption’. Because these
traditions begin in specific areas at folk or popular level, most simply continue
on in situ, at least for a while until their function is otherwise articulated, their
sites are appropriated, or they cease to be important (if that ever happens). In
the vināyakas’ instance, by the seventh century they are aggregated into, and
occluded by, Gaṇapati/Gaṇeśa among the Buddhists as well, for Gaṇeśa may be
both the author of obstacles (vighneśvara) and their remover (vighnahartā).31
For our purposes, we note that few of the brahmanical texts agree with each
other on vināyaka details, except to affirm that there are one or more vināyakas
who either possess a person or make obstacles in the pursuit of any endeavor.
Consequently, if a person dreams of any number of disturbing images—water,
or an outcaste, or a person with a shaved head, or one wearing a red robe—or
if a brahman cannot become a teacher, or a prince obtain a kingdom, a young
lady cannot find a husband or a young wife cannot conceive, a farmer cannot
cultivate the soil or a merchant ply his trade: any or all of these and more are
indicators of possession by a vināyaka, who creates impediments to any pur-
suit. We note in most of the early literature, our vināyaka has nothing to do
with one or another of the great gods. He seems more like the theological thug
at the crossroads, one who must be tamed.
31 Duquenne 1988 explores the somewhat later Gaṇapati rituals in the Chinese archive, and
mentions Bodhiruci II’s statement that a vināyaka ritual was translated between 581–589,
apparently after our own text.
148 Davidson
And the purpose of the *Mūlamantra ritual is exactly that. The vidyādhara
draws a vināyaka image on his altar, and hits the vināyaka image on top of
his head with sanctified mustard seeds (sarṣapa), eight hundred times. The
vināyaka is not treated nicely until the obstacle is removed, and only then is
he offered milk so that he will become the vidyādhara’s eternal servant. This
is not the supplication via various offering that we find for a vināyakavidhi in
brahmanical literature, where the vināyaka/Gaṇeśa is offered various sweets,
or alcohol and meat, or various kinds of standard strewn offerings (bali) or
homa fire offerings, as in the case of the rice/milk congee. These offerings in
the orthodox texts are done so that the vināyaka will either not establish prob-
lems, will remove problems, or will facilitate the success of the venture.
While not all vidyādhara rituals enact violence, coercion or compulsion on
the god/spirit, many of them do, especially when the object to be addressed is
a well-known hostile presence. The following treatment of Mahākāla is indica-
tive of these kinds of rites, from the vidyādhara’s anger to the supernormal
benefit accrued to him. Mahākāla appears to be represented as a yakṣa, unsur-
prising as some of the names eventually associated with Śiva or his retinue
seem to have been understood by Buddhists as applicable to yakṣas, particu-
larly notable in the case of the yakṣa Maheśvara.32 The following is from the
fragmentary text of the Amoghapāśa-mahākalparāja, first translated in 707 CE
by Bodhiruci II, about a century and a half after ours. It is consequently a bit
more specific about the ritual, but there is little here that would be out of place
in our text, with the exception of the use of the ‘Krodharāja’ mantra.
Krodharāja onto your right hand seven times, the vidyādhara should slap
the statue. Then Mahākāla will shriek, but the vidyādhara should not be
afraid, but again with anger he is to be conquered (with another slap).
The tears will fall from the statue and the vidyādhara, having taken up the
tears, will make a tilaka (in his forehead), so that he will become invisible
and be able to see all kinds of hidden treasure.
Then the vidyādhara should slap Mahākāla again, so that he begins to
vomit blood. When the vidyādhara touches it, he will be able to fly into
the sky and will become the Overlord of Yakṣas. Having offered Mahākāla
bdellium gum incense, and having recollected the Krodharāja(mantra),
Mahākāla will come and stand before the vidyādhara. Whatever the
vidyādhara desires, Mahākāla will do, and all the retinue of the Mothers
will be under the vidyādhara’s control, so that they will do whatever is to
be done according to the desires of the vidyādhara.
The same pattern seen in the treatment of the vināyaka is also expressed here:
the vidyādhara punishes the image until it either cries or bleeds, preferably
both since each confers a different power. Only then is Mahākāla treated well,
offered incense, and made an inseparable servant.
Perhaps the most distinctive aspect of this ritual is its purpose: the invocation
of nāgas to control the various facets of wet weather, including fog, hail, and
other water-related weather events. As we have seen above in the myth of the
nāga Janmacitra, the association of nāgas with every variety of precipitation
has been an enduring theme in Buddhist lore, reiterated in Buddhist literature
in virtually every language into which the tradition was translated. Inscribing
that mythology into a performative expression in nāga rain rituals certainly
extended from the earlier mythology of nāgas and their control over waters,
vapors and atmospheric effects.
In distinction, the rituals for rain in brahmanical sources—with one insig-
nificant exception so far as I have been able to determine—do not address or
invoke the nāgas for assistance with weather.34 While the association with and
placement of snake spirits within water, rain, mist, fog and thunder has been part
of the brahmanical mythic landscape in India since Indra’s conquering of Vṛtra
34 I was interested to learn that Professor Einoo, with his much greater experience in such
matters, had independently arrived at the same conclusion, and had presented a paper on
the topic, but it had not been published as of the time of the conference.
150 Davidson
in Ṛgveda 1.24 onward, that does not mean that nāgas were specifically invoked
by brahmanical ritualists during the optional rain rites (kāmyavṛṣṭikarma)
when drought was a problem. As others have argued, when Indra conquered
Vṛtra, the goal was in part to control the waters, and the mythological dimen-
sion may be understood in some sense an effort to control resources and the
ritual authority associated with those resources. Consequently, in brahmanical
literature, one or another god is most often considered the deity responsible
for rain and to whom one is to pray for rain—Indra, Parjanya, Varuṇā, Sūrya,
the Maruts and Mitra, to name but the most important.35 The primary rain
ritual in the śrautasūtras is the kārīrīṣṭi, although a vṛṣṭi or other rites may be
alternatively employed.36 The gṛhyasūtras, their supplements and related rit-
ual extensions in vidhāna and puranic literature, similarly rely on these gods.37
Occasionally, a liturgical phrase simply indicates that the unnamed god rains,
or that one wishing the god to rain should perform a ritual.38 Nonetheless,
I can find scant record of the nāgas granted authority equivalent to that the
gods themselves receive in orthodox ritual texts, and the only occasional
notices I have found of the association of nāgas and rain in the modern period
are with rural or tribal groups not dependent on brahmins (e.g., Oldham
1901: 471).
The reason for this state of affairs is relatively straightforward. Sending rain
down in the proper amount and for the proper duration was part of the overall
exchange with the gods. Brahmanical sacrifice required that the divinities of
the orthodox tradition control the rains, for the gods received sacrifice so that
35 For a theology of Sūrya as rainmaker, see Vāyupurāṇa 51.14–52, Matsyapurāṇa 124.
36 The Śrauta sources for this ritual are identified in Dandekar 1958–95, Śrautakośa I/2: 567–
72, 627–28.
37 A partial listing of the rituals for rain in the gṛhya literature would include Kauśikasūtra
41.2–7, Āgniveśyagṛhyasūtra 2.5.10, Gautamagṛhyapariśiṣṭa 1.22, 2.22, 2.23, 2.24,
Atharvavedapariśiṣṭa 36.22, 65.1, 65.3, 70b.17, Sāmavidhānabrāhmaṇa 3.4.10, Ṛgvidhāna
2.41, 2.90–92, 2.156, 4.127. Among the Sāmaveda traditions, the mahānāmnī verses can be
used to cause rain: Gobhilagṛhyasūtra 3.2.28–29, Jaiminigṛhyasūtra 1.17. We note that the
Āgrahāyaṇa ceremony, which marks the conclusion of sarpabali, occasionally has rain
mentioned, but is not specifically a rain ritual, for it is a general request for many benefits,
including rain: Pāraskaragṛhyasūtra 3.2, Śāṅkhāyanagṛhyasūtra 4.17–18. Many rain rites
are found in the purāṇas: Nāradapurāṇa 2.14.68, Bhāviṣyapurāṇa 1.29.30, Nīlamatapurāṇa
755, Agnipurāṇa 127, 258.53, 259.51, etc. Gonda 1980, mentions rain rituals, mostly in pass-
ing: 44, 110, 119, 132, 134, 255, 308, 398–9.
38 Śuklayajurvidhāna p. 4: devaṃ varṣayitukāmaḥ. Even in the earlier Buddhist
Sthaviragāthās, as in the case of Subhuti’s, it is a god who rains, as he likes: Theragātha 1.1:
vassa deva yathāsukhaṃ.
A Nāga Altar in 5th century India 151
While the orthodox theology of sacrificial acts has traditionally been domi-
nated by the question of the debts the brahmins pay to the gods, the ṛṣi and the
ancestors (e.g., Malamoud 1996: 92–108), here we find a separate narrative, one
that applies to all humans. The relationship is reciprocal and based on a for-
mality of exchange. Humans provide sacrificial offerings and receive rain and
the objects of their desires as a consequence. In this ritual narrative, were the
brahmins to pray to nāgas for rain, it would interrupt the cycle of offering and
divine bequest—a quintessential theology of exchange—and as such would
interrupt the relationship between the community and their priests, for no one
needs a priest to make offerings to a nāga.
We are even instructed in orthodox literature that, when nāgas need rain,
they themselves must appeal to the gods, despite their association with mists
and watery abodes. In the Mahābhārata, the Ādiparvan (21–22) contains a
short, Icarus-like story that becomes elaborated later, as in the Brahmapurāna.39
It describes the continuing conflict between the two daughters of Prajāpati
who become the wives of Kāśyapa, Vinatā the mother of Garuḍa and Kadrū
the mother of the nāgas. At one point in the story, Kadrū compels Garuḍa
to carry her nāga children to an ideal island for pleasure. But they were car-
39
Brahmapurāṇa 159; the Mahābhārata vulgate version is mentioned by Vogel, p. 51. A very
helpful discussion of the conflict between the two sisters and the Indo-Iranian roots to
the narrative is found in Knipe 1967, pp. 345–49.
152 Davidson
ried too close to the sun, so that their nāga bodies were scorched and they
lost consciousness. Consequently, Kadrū resorted to praising Indra in order to
bring rain to her sons to revive them, which he did, sending down masses of
blue clouds, blotting out the sun and moon, flooding the area with rain and
reviving the nāgas. Thus, in brahmanical lore, the nāgas may live in moist
environments, but do not appear to be able to control the various forms of pre-
cipitation, including rainfall. They must—with the rest of the world—appeal
to divinity for precipitation.
This is not to say that there are no ritual appeals and invocations to nāgas in
brahmanical literature, but those instances wherein we find brahmanical ophi-
olatry, it is for purposes other than rain.40 Probably because of the nāgas’ con-
nection with the subterranean realm or their control of gems, these objectives
may include the search for gold. “ ‘May there be homage to the snakes!’ Let there
be [a homa of] ghee and milk-rice gruel at a nāga site, and gold will be found.”
(Śuklayajurvidhāna Annāśāstri: 15.8–9: namo ‘stu sarpebhya iti ghṛtapāyasaṃ
nāgasthāne juhuyāt | suvarṇam udpadyate || Śarmā 2.39). Here, the goal is
control of the snakes and retrieval of gold by means of the Sarpasūkta of the
Vājasaneyi-śuklayajurveda (Vājasaneyisaṃhitā 13.6 = Taittirīyasaṃhitā 4.2.8.g-
1), which pays homage to the snakes of the earth, the intermediate space and
the heavens. Other subterranean goals are occasionally mentioned, and the
relationship between the human and serpentine sphere is intertwined with
myths of serpent maidens (nāgī), powers and treasure obtained in the under-
world where they live (pātāla), to mention but a few (Vogel 1926).
And the single instance I have found of a brahmanical text associating
offerings to nāgas with rainfall is in two sentences appended to this part of
the Śuklayajurvidhāna. After mentioning gold, the recensions of the text also
append variously phrased non-sequitur instructions, “If rain is desired, also sac-
rifice peacock feathers and rain will occur; if common flax flowers are offered,
then great rain will come.”41 Yet in this instance, it appears that an eyeskip mis-
copying from another rain ritual elsewhere in the chapter (Annāśāstri 27.5–6;
Śarmā 2.45–46) yielded this misleading occurrence. The late medieval com-
mentary, the Yajurmañjarī (p. 164), makes no mention of rain in conjunction
with the ritual of the Sarpasūkta. Nor do the commentaries to the Sarpasūkta
40 Einoo 1994 surveys the goals found in the later nāgapañcamī rituals, none of which invoke
rainfall.
41 Śuklayajurvidhāna, Annāśāstri 28.3–4 (vṛṣṭyarthe śikhaṇḍyādīñ juhuyāt vṛṣṭir bhavati |
atasīpuṣpair mahāvṛṣṭir bhavati |; cf. Śarmā ed. 2.39, p. 64, and the Śikṣāsaṅgraha text
(pp. 350.21–351.5) employs this not just for rain, but with the goal of fearlessness.
A Nāga Altar in 5th century India 153
in the Vājasaneyisaṃhitā mention rain in this context, but only seek protec-
tion through these powerful beings.
In distinction to this probable error in copying, the most significant brah-
manical ophiolatrous ritual, well studied by Winternitz, was the sarpabali,
found in many of the domestic texts and sometimes enumerated as one of
the seven lesser domestic rites (saptapākayajña; e.g., Gautamadharmaśāstra
8.18). The sarpabali was generally to take place on the full moon in the month
of śrāvaṇa (July-Aug) in the midst of the monsoon, and was thus often sim-
ply called the śravaṇā rite. As portrayed in the Mānavagṛhyasūtra, the moti-
vation was fear of snakes (Mānavagṛhyasūtra 2.16.1: sarpebhyo bibhyat).
Consequently, the procedure of the ritual is generally to ward off snakes from
coming into a perimeter established by the ritualist, either outside of his
house (usual) or the larger community, left to the discretion of the patron or
performer. The rite was most often done by making an offering of flour and
water, the scattered offering (bali), via a special implement, the darvī spoon,
which is also used in several other gṛhya and śrauta rituals.42 Sometimes the
perimeter is incised, and in the invocation of the perimeter, the deity “White”
(Śveta) is often invoked, to keep nāgas away with his foot.43 Winternitz has
convincingly shown that this divinity is a horse figure drawn from lore of Indra
in both the Ṛgveda and the Atharvaveda.44 The invocation usually given was
a modification of an Atharvaveda mantra; the modified non-Vedic form is
found in the Mantrapāṭha, a collection of extra-canonical mantras formally
affiliated with the Āpastamba branch of the Taittirīya-Kṛṣṇayajurveda school.45
Many of the sarpabali rites also employ the same Sarpasūkta verses from the
Vājasaneyisaṃhitā (16.13) or Taittirīyasaṃhita (4.2.8.g-1) we saw used to obtain
gold. Sarpabali rituals also often prescribe an homage to bhauma, the genis
loci, even when other gods like Viṣṇu or Agni may additionally be invoked.
If other goals are mentioned—wealth, for example—the central nature of
the sarpabali is not lost, for it also may involve removing of domestic bed-
ding for the next two months until the Āgrahāyaṇa ceremony is done after the
42 Karmapradīpa 2.5.15 and Bhāradvājagṛhyasūtra 1.1 offer descriptions of the darvī ladle.
43 In Jayarāma’s commentary to the Pāraskaragṛhyasūtra (p. 291.33) it is noted that the
incised line is understood to have the quality of Śveta’s foot: sarpagamanamārge
śvetarekhāyā dūśyamānatvāt tasya śvetapadatvam.
44 Winternitz 1888: 49–52: Ṛgveda 1.117.7, 1.118.7, 1.119.10, 7.71.5, 9.88.4, 10.39.10.
45 Atharvaveda 10.4.3, Mantrapāṭha II.17.27. The variant or a recension of it is found in
Āśvalāyanagṛhyasūtra 2.3.3, Pāraskaragṛhyasūtra 2.14.4, Kātyāyanagṛhyasūtra 2.17.4,
Hiraṇyakeśigṛhyasūtra 2.16.8, and Bhāradvājagṛhyasūtra 2.1.9, Śāṅkhāyanagṛhyasūtra
4.18.1, and the Mānavagṛhyasūtra 2.7.1. We note that, anomalously, the Śāṅkhāyanagṛhya
sūtra really has two sarpabali chapters: 4.15 and 4.18.
154 Davidson
monsoon is over, at which time the bedding is returned to its normal position.
Perhaps we may perceive a ritual uncertainty in this procedure—there is no
reason to do this if the anti-snake perimeter holds—with the practical attitude
of ‘better safe than sorry’ implicated. Nonetheless, the primary ritual employ-
ing the Sarpasūkta has nothing to do with rain, which would be peculiar as the
problem in śrāvaṇa is not so often a lack of rain but flooding. Praying for rain
at this time would generally be counterproductive. Rain is required in the hot
season or (optionally) in the fall if the monsoon fails, but this is not an optional
rite; it is an obligatory annual rite (nityakarma), hence its status as one of the
seven domestic sacrifices (pākayajña).
Turning to the establishment of nāga images (nāgapratiṣṭhā) in orthodox
literature, this consecration rite is also done for purposes other than rainfall
and in quite a different manner: they appear to involve the temporary place-
ment of the anthropomorphic image in water (jalādhivāsa), the use of homa
besides the kinds of pūjā offerings familiar in our ritual, the aspersion of the
image with the elaborate contents of vases, and the invocation of mantras from
their particular archives. Despite the widespread observation of nāgapratiṣṭhā
rituals in South Indian temples today, ancient brahmanical sources are sur-
prisingly scarce for the placement of images.46 We will consider nāga statue
placement rituals in the Gautamagṛhyapariśiṣṭa and the Rauravāgama, and
only the latter has been studied in any detail.47
Gautamagṛhapariśiṣṭa 2.12 is arguably the more interesting, as it seems
earlier, invoking (2.12.14–15, 28) the nāga as Śeṣa, Ananta, Saṃkarṣaṇa and
Balabhadra, all resonant with the presence of Viṣṇu in the sarpabali of the
gṛhyasūtras. But there are also invocations of Īśāna and Aghora (2.12.19), and
this range of invocation is consistent with Smārta practice generally. While
a contemplation of the image of a nāga is described, its actual composition
is not mentioned. Nonetheless, it is also to be painted: red body, with yellow
clothing (2.12.11). Many mantras are invoked, some traceable to the Sāmaveda
tradition of Gautama with which this text is affiliated, but the work describes
(as does ours) a mūlamantra, in this case more convincingly applicable to
46 Mahalingam 1965, p. 56. Besides these three texts, I have found instances of nāgapratiṣṭḥā
rather incidentally; e.g., Muktabodha T0120 contains a Nāgapratiṣṭhā (pp. 1–79) and a
Rauravāgamapaddhati that involves nāgapratiṣṭhā.
47 Dagens and Barazer-Billoret 2000: xxii–xxiii for a comparison of the two chapters of the
text; N.R. Bhatt in his notes to the text, vol. 1, pp. 137–142, provides quotations from other
Śaivāgamas. I do not have access to the full Kāraṇāgama or the other Śaivāgamas quoted
by Bhatt in his note, so I cannot discuss nāgapratiṣṭhā rites found in either this or the
other texts he identifies.
A Nāga Altar in 5th century India 155
nāgas: oṃ nāgāya namaḥ, which was probably employed for nāga worship in
a relatively wide area. In the end, the benefits are specified: ‘By this [ritual]
he [the patron] will obtain eight sons, complete dominion, and whatever he
wishes he will receive’ (Gautamagṛhyapariśiṣṭa 2.12.28: anenāṣṭaputrān lab-
hate samastaiśvaryaṃ labhate yaṃ yaṃ kāmaṃ prārthayate taṃ tam āpnoti).
All told, this ritual reveals the devapratiṣṭhā background of the rite, since the
nāga is identified as deva at one point (2.12.25), and it seems that the editor or
scribe saw little reason to change the nomenclature.
The Rauravāgama expresses the establishment of a nāga statue in two
chapters (38 and 57), both of which associate the nāga with the installation
of a goddess and both of which mention a nāgamūlamantra, unsurprising as
there are various mūlamantras scattered throughout the text. The earlier chap-
ter specifies that the statue is to be made of either iron, stone, wood, mud or
even of an amalgam of expensive substances (dhātuja), and may be either a
five-headed nāga or an iconic statue with a five-headed (or one, two or three-
headed) nāga atop it; the later chapter only specifies that the primary image
is to be of one or more faces (up to five) and a second image for worship in a
pavilion is to be made of gold.48 Both require the establishment of pavilions
where much of the ritual activity is carried on prior to the final placement
of the image. The relatively sparse directions in the earlier chapter (38) are
elaborated extensively in the later chapter (57), which also shows an increasing
concern with Vedic mantras. Indeed, we find in the later chapter the specifica-
tion of a nāgagāyatrī (57.112–113), one of the many instances of the adaptation
of the gāyatrī to a specific theological persona. Besides formally integrating
the cult of the eight nāgarājas, the longer chapter provides a list of benefits
(57.128c–130): whatever is desired will be obtained—husbands, sons, power,
victory, wealth, so that one enjoys life like a god in heaven (modate divi deva-
vat). Overall, the Rauravāgama chapters attempt to move the direction of the
discussion away from popular rituals to a platform for brahmanical authority
and patronage, on a par with brahmanical temple rituals generally.
The emphasis in all such rites is on the gods and their powers, to which
the nāgas become subordinated, formally so in the case of the two chapters
of the Rauravāgama, as the images are to be placed in devī temples. We see
verification of the practice in the British Museum Stone Inscription of the
Time of Kaṇiṣka, where the image is that of a nāga and nāgī is being dedicated
at the temple of the village goddess (Lüders 1907–08, p. 240: devī gramasya).
Perhaps that is one reason for the relative paucity of nāga rituals beyond those
48
Rauravāgama 38.10: lohajaṃ śailajaṃ dārujaṃ vā prakalpayet | mṛṃmayaṃ vātha kar-
tavyam dhātujaṃ vā viśeṣataḥ.
156 Davidson
associated with protection from snakes and the monsoon: the functions attrib-
uted to the nāgas in many instances were those proper to the gods. So, given
the anthropological data on the nāga rituals throughout India, their relative
absence in the received archive suggests their liminal status in the Sanskritic
ritual discourse. We even see the reverse of the nāgas’ placement within tem-
ples of the goddess by intrusion of the deities into a nāga’s site in the Nīlamata-
purāṇa, which explains that Rāma set up a statue of the Sārṅgin (Viṣṇu) at a
site of the nāgarāja Ananta:
Not far from that site of Puṇyoda, the palace of the great soul
Ananta, the King of Nāgas, [Rāma] performed dreadful penance,
And established the [image] of the god Sārṅgin.49
Setting this into perspective, despite the dedication of the Nīlamata to nāga
lore, this early purāṇa does not mention a single instance of the installation of
a nāga image, in contradiction to its outline of piśāca image worship, as previ-
ously described.
Whatever the widespread applicability of the process of site appropriation
on behalf of orthodoxy, there can be little question that nāga-focused rituals
for the purposes of weather change are scarcely recognized in brahmanical
literature overall. There and elsewhere, the consistent emphasis on the gods
as the controlling authorities, and on the placement of nāgas in subordinate
relationship to the gods, goddesses, or other emblems of religious hierarchy,
serve to emphasize the primacy of brahmanical communities as the chief
arbiters of ritual benefits. It suggests that the nāgas were icons of alternate
authority, ones with a lengthy pedigree, however they are treated in Sanskritic
literature.
49
Nīlamata p. 98, de Vreese reads 1189c tasmād adūre puṇyodām, but p.98n1 notes the gloss
puṇyodām adure puṇyodāsamīpe ity arthaḥ, suggesting the original line read puṇyodād,
which is how I read the verse. Amarakośa 1.1.19 (or 1.1.14, depending on the numbering)
lists Śārṅgin as one of the names of Viṣṇu, and probably meaning Kṛṣṇa, as Śārṅga is
Kṛṣṇa’s bow in the Mahābhārata.
A Nāga Altar in 5th century India 157
In distinction, Buddhists monks have invoked nāgas for the purposes of bring-
ing rain at least since the fourth century, approximately two hundred years
prior to our translation. Even before that, the archaeology of Buddhist locales
supports a model of the early Buddhist communities living in close proxim-
ity to nāga sites. As early as the second century BCE site of Maṇiyār Maṭh in
Rājagṛha we find both formal and informal nāga ritual enclosures in proxim-
ity to important Buddhist congregations; other major sites included Sonkh,
Sāñchī and Mathurā.50 However, there is only the most tenuous indication
of the rituals of these sites, in their few dedicatory inscriptions. With no sur-
viving epigraph on their construction, however, it is difficult to determine
with precision their purposes or patronage. The fact that they were erected
in stone—with its concomitant requirement of an advanced resource base—
suggests official patron(s), either a metropolitan organization, governor or royal
patron being the most likely. Moreover, it is not apparent that the Buddhist
monks initially involved themselves with the activity of these proximate sites,
and the literature supports an early ambivalence to nāga communities, if
the multiple mythological instances of conflict with nāga worshippers—like
Urubīlva Kāśyapa—are of any indication.
The association of nāgas and rainfall may have extended from their overall
control of water and water sources, one of the reasons that, even today, nāga
sites are located at springs and tanks in North India. Possibly our earliest exam-
ple of monastic involvement in a nāga water rite is Fotudeng, a Central Asian
Buddhist monk, perhaps from Kucha but trained in Kashmir, who arrived in
China in 310 CE. His biography in the Gāosēngzhūan (T.2059) was studied and
translated by Wright, who renders the nāga episode as follows:
The source of the water for the moats of the city of Hsiang-kuo was
located below the T’uan-wan shrine, five li northwest of the city. This
water had suddenly dried up. Lo inquired of Teng (i.e., Fotudeng), ‘How
shall I get the water to flow?’ Teng replied, ‘We should now get a dragon to
come.’ Lo, whose style was Shih-lung (世龍 Dragon of the Age), thought
Teng was mocking him and replied, ‘It is precisely because the dragon
50 On nāga sites generally, see the helpful discussion in Härtel 1993: 426–7. The major exca-
vation reports on Rajgir are Marshall 1909, and Nazim 1940. On Sāñchī nāga sites, see
Shaw 2004, 2013, who seems hesitant to appreciate that Buddhist proximity does not nec-
essarily mean monastic ritual involvement or to acknowledge that the ritual agendas of
the communities might have diverged.
158 Davidson
could not make the water flow that you were asked.’ Teng said, ‘This is a
serious statement and not a jest. The source of the stream certainly has a
holy dragon living in it. Now if I go and speak commandingly, water will
surely be obtainable.’ Then with a number of disciples such as Fa-shou,
he went up to the source of the stream. At its source the old streambed
had been dry for a long time and was cracked so as to look like cart tracks.
His followers were doubtful and afraid that water would be difficult to
get. Teng sat down on a corded bench, burned Parthian incense, chanted
an invocation of several hundred words. When he had done like this for
three days, water seeped out a few drops at a time. There was a small
dragon, about five or six inches long, which came out with the water. Teng
said, ‘That dragon is poisonous. Do not go near it.’ In a little while the
water came in abundance, and the dry moats were all filled.51
The ritual is hard to understand here, but apparently Fotudeng sat on a high
seat of some variety, and the ‘corded seat’ of the biography (shéng-chuáng
繩床) can represent mañca or mañca-pīṭha, a raised platform seat, with per-
haps some kind of string or chord wrapped around it to denote its sacred status.
He burned guggulu (安息香, Commiphora wightii) incense and pronounced
a lengthy text of some variety (呪願數百言). The description does not allow
us any certainty on its source, but two texts—the Mahāmeghasūtra and the
Mahāmāyūrīvidyārājñī—were translated rather early into Chinese, and some
early version of one or both are our best guesses as to his recitation.
Neither of these works provide much ritual direction, however, until we get
to the early fifth century, when we begin to see the development of actual rain
rituals. So far as I have been able to detect, the earliest indication of this is in a
translation of the Mahāmegha-sūtra attributed to Dharmakṣema, 414–421 CE,
which provides a short ritual instruction, maddening in its lack of clarity:
若有國土欲祈雨者。六齋之日其王應當淨自洗浴。供養三寶尊重讃嘆
稱龍王名。善男子。四大之性可令變易。誦持此呪天不降雨。無有是
處。T.387.12.1094b24–28
If there is a head of a state who wishes to cause rain, [after] six days of
fasting, the king should clean and wash himself. Then he should offer
worship to the three gems, praising them humbly. Then he should invoke
51 Translation Wright 1990, p. 49; this is a reprint of the article ‘Fo-T’u-Teng: A Biography,’
Harvard Journal of Asiatic Studies 11/3–4 (1948): 321–371. The section of the Gausengzhuan
is T.2059.50.384a1–a12.
A Nāga Altar in 5th century India 159
the name of the nāgarāja. O son of good family! There will be a transfor-
mation in the nature of the four mahābhūta elements. For one reciting
and retaining this mantra, there is no place that heaven will not send
down rain.
Translated a little more than a century before our ritual, its paucity of liturgi-
cal information should pique our curiosity as to the circumstances that moved
Buddhist communities from simply offering prayers for rain to constructing
altars, making images and offering substances to the images on behalf of oth-
ers. Here we see that the king, as head of state, is to perform his own offering
and calling on the nāgarāja. This is vaguely in line with the understanding of
the duty of kings in early India—as kṣatriyas they are responsible for kṣatra,
authority over the security and protection of the people—but equally it may
represent the tailoring of the rite to the Chinese paradigm, where the king of a
country is also its chief ritualist.
Still, it is not hard to see these rituals as being on an irregular continuum
with our own text, and we may observe the conduct of monks in India contem-
porary with the ritual Dharmakṣema translated. In his 417–21 CE memoir of his
travels in India, Fǎxǐan observed a relationship between monks in a monastery
at Sāṅkasya and a nāga shrine.
住處有一白耳龍。與此眾僧作檀越。令國內豐熟雨澤以時無諸災害。
使眾僧得安。眾僧感其惠。故為作龍舍敷置坐處。又為龍設福食供
養。眾僧日日眾中別差三人到龍舍中食。每至夏坐訖龍輒化形作一小
蛇。兩耳邊白。眾僧識之。銅盂盛酪以龍置中。從上座至下座行之。
伏若問訊。遍便化去。每年一出。
Dwelling there was one nāga, *Śvetakarṇa, acting as the patron of the
Saṃgha. He caused those within the country to experience increase of
rain and full water tanks, all timely and without any misfortune. This
caused the Saṃgha to live peacefully. The Saṃgha was moved to express
its gratitude, and thus constructed a nāga house (*nāgagṛha), with a
small bed (*niṣadyā) placed within. And on behalf of the nāga they set
out divine food as an offering (*naivedyapūjā). Each day the Saṃgha
would deputize three monks to go to the *nāgagṛha and set out the food.
Each time the Saṃgha approached the completion of the rains’ retreat
(varṣoṣita), the nāga would send out an image of itself, made to look like
a small snake, with the two ears in white on the sides. The Saṃgha recog-
nized it. They then took a copper basin and filled it with curds, and put
the snake on top. [It was taken] to visit the monks, from the most senior
160 Davidson
to the most junior, and it bowed down as if asking a question. Each time
it would make as if to go away, and each year it again returned.52
The use of various color terms to identify snakes has already been noted by
Vogel (1926: 191, 226, 263), and we may also observe that ‘ear’ terms are con-
spicuous in nāga names, probably an interpretation of the cobra’s hood as ears.
One of the names for a nāgarāja found in the Mahāvyutpatti is ‘White one’
Śvetaka (no. 3326), as is Hastikarṇa (‘elephant ears’ no. 3313), so this predispo-
sition is evident here as well; there is a minor descendent of Pārikṣita named
Śvetakarṇa in the Harivaṃśa (114.5–7 = Brahmapurāṇa 13.127–129), so the name
is otherwise known. And in modern times, ephemera—posters, block prints,
etc.—for the nāgpañcamī festival often show nāgas with curious ears, shaped
rather like a dog’s. The nāga seems to have been set in a bath of curds (lào 酪,
generally dadhi) rather than milk, as others have sometimes interpreted this
term, perhaps in light of the modern propensity to try bathing nāgas in milk.
Following the curds-and-nāgas theme, three of the more important inscrip-
tions concerning the nāga cult at Mathura concern dedications to Nāgendra
Dadhikarṇa, one whose ears are like curds.53
Irrespective of the lack of clarity on these terms, and despite the uncertain
nature of the cult and the offerings that Fǎxǐan’s three monks made on a daily
basis, generally they would seem to be aligned with the kinds of offerings we
have seen in Kumārajīva’s altar for the Mahāmāyūrī or Dharmakṣema’s altar for
the Mahāmegha, as mentioned before. In the event, the folk-based mud/dung
images of snakes found within our text could have been used when the nāga
did not make his physical appearance to receive his offerings. And it would
be unlikely that such an offering system would have been set up by Buddhist
monks without some physical item representing the nāga in question, if only
to the point of using a jug of water (kalaśa) as was often employed to personify
the deity or spirit in question.
We do not know, however, if in such cases there was actually any equivalent
of pratiṣṭhā as employed in elaborate temple programs elsewhere. The sup-
position has sometimes been that Vārāhamihira’s observation on Buddhist
pratiṣṭhā must have been true somehow. In Bṛhatsaṃhitā 59.19, Vārāhamihira
indicated that Viṣṇu’s statue, the statue of the sun, the liṅga of Śiva, the images
52 T.2085.51.860a6–13; Deeg 2005, pp. 281–83 is not so helpful here as he is elsewhere.
53 Lüders 1961, pp. 62, 70, 127. Cunningham discovered the tank of a nāga at modern Sankisa,
dedicated to Kārewar, the ‘black one’, who was honored on nāgpañcamī and is one of the
few still associated with rainfall; Cunningham 1871: 274.
A Nāga Altar in 5th century India 161
of the mātṛ-s, of Brahmā, of the Buddha and of the Jina, each should be placed
in their respective temples by the sectarian principals involved. Bhaṭṭotpala’s
comment specifies that the Buddhists would do this by the procedure of the
pāramitā. Unfortunately, I know of no Buddhist statement that validates this
procedure at the time of Vārāhamihira, even if Prajñāpāramitā texts were so
employed at a later date, and this may be the basis for Bhaṭṭotpala’s observa-
tion. Given the ritual texts from the late Gupta period actually at our disposal,
it would seem that the Buddhists were much less invested in elaborate place-
ment rituals than the orthodox systems, and that they tended to reformulate
their rites with some regularity.
Be that as it may, nāga-focused rain rituals were an important and continu-
ing part of the Buddhist ritual menu, and even in the *Mūlamantra a shorter
rite employing the hṛdayamantra is elsewhere described in passing.
若呪白芥子滿八千遍。散於空中能令雨下。一切那伽咸當敬順.
*Mūlamantra 659a12–14
Similar rituals,54 found throughout both the Chinese and Tibetan canons, are
rearticulated in indigenous ritual manuals, and it would be out of the bounds
of this article to delineate them all, with their attendant individual sites and rit-
ual programs from Afghanistan to Japan and Sri Lanka to Mongolia. We do see,
however, in this instance and the longer ritual above, the vacillation between
the exchange of offerings and the coercive attitude of this one. Eventually, the
idea of pūjā, with its fundamental vocabulary of reciprocity, will become the
dominant theme, and the Buddhists will emulate the guest-reception behav-
iors that marked polite society among the twice-born.
Conclusion
The nāga rain ritual, as found in the *Mūlamantra, was but one moment in
an important facet of Buddhist ritual enterprise: the cultivation of divine ser-
pents for the purpose of bringing rain to communities of their patrons. The
54
Hevajra-tantra I.ii.20, Amoghapāśamahākalparāja fol 29b5 are later examples.
162 Davidson
Acknowledgments
This work was carried out in part with the generous support of the National
Endowment for the Humanities, the American Council of Learned Societies
and Fairfield University. I thank István Keul for organizing the conference and
editing the volume, and my many friends and colleagues for their encour-
agement and assistance, especially Charles Orzech, Victor Mair, Katherine
Schwab, Jacob Dalton, Richard Payne and Glen Hayes. Any errors in this article,
of course, are mine alone.
A Nāga Altar in 5th century India 163
Bibliography
Abbreviations
sTog Skorupski, Tadeusz. 1985. A Catalogue of the Stog
Palace Kanjur. Tokyo: The International Institute for
Buddhist Studies.
T. Taishō Shinshū Daizōkyō. Takakusu Junjirō and Watanabe Kaikyoku (eds)
1924–35. Tokyo: Daizōkyōkai.
To. Ui, Hakuju, et al, (eds) 1934. A Complete Catalogue of the Tibetan Buddhist
Canons (BKaḥ-ḥgyur and Bstan-ḥgyur). Sendai:
Tôhoku Imperial University.
Primary Sources
Agnipurāṇa. Ānandāśramasthapaṇḍitas. 1909. Agnipurāṇam. Bombay: Ānandāśrama.
Āgniveśyagṛhyasūtra. Ravi Varma, L.A. 1940. Āgniveśyagṛhyasūtra. Trivandrum:
University of Travancore.
Amoghapāśahṛdaya. Meisezahl, R.O. ‘The Amoghapāśahṛdaya-dhāraṇī: The Early
Sanskrit Manuscript of the Reiunji Critically Edited and Translated.’ Monumenta
Nipponica 17: 265–328.
Amoghapāśamahākalparāja. Kimura Takayasu, ed. ‘Transcribed Sanskrit Text of the
Amoghapāśakalparāja.’ Taishō Daigaku Sōgō Bukkyō Kenkyūjo Nenpō (大正學綜合
佛教研究所年報) vol. 20 (1998): 1–58; vol. 21 (1999): 81–128; vol. 22 (2000): 1–64;
vol. 23 (2001): 1–76; vol. 26 (2004): 120–183; vol. 32 (2010): 170–207; vol. 33 (2011):
32–64. To. 686; T.1092, Trans. Bodhiruci II 707 CE.
Āśvalāyanagṛhyasūtra. Vidyáratna, Rámanáráyana and Vedántávágísa, Anandachandra.
1866–69. The Grihya Sútra of Áswaláyana, with the Commentary of Gárgya Náráyana.
Calcutta: The Asiatic Society.
Atharvaveda. Whitney, William Dwight and Nag Sharan Singh. 1987. Atharva-Veda-
Saṁhitā: Text with English Translation, Mantra Index and Names of Ṛṣis and Devtas.
2 vols. Delhi: Nag Publishers.
Atharvavedapariśiṣṭa. Bolling, George Melville and Julius von Negelein. 1910. The
Pariśiṣṭas of the Atharvaveda. Leipzig: Otto Harrassowitz.
Bhaiṣajyavastu. Dutt, Nalinaksha. 1947. Gilgit Manuscripts, vol. 3/1. Kashmir: Research
Department.
Bhāradvājagṛhyasūtra. Salomons, Henriette J.W. 1913. Bhāradvājagṛhyasūtram: The
Domestic Ritual According to the School of Bhāradvāja. Leyden: Brill; rpt. New Delhi:
Meharchand Lachhmandas Publications, 2004.
Bhaviṣyapurāṇa. Sharma, Rajendra Nath. 1993–95. Śrībhaviṣyamahāpurāṇaṃ. 3 vols.
Delhi: Nag Publishers.
164 Davidson
Ṛgvedasaṃhitā. van Nooten, Barend A. and Gary B. Holland. 1994. Rig Veda: A Metrically
Restored Text with an Introduction and Notes. Cambridge, MA: Department of
Sanskrit and Indian Studies, Harvard University.
Ṛgvidhāna. Bhat, M.S. 1987. Vedic Tantrism: A Study of Ṛgvidhāna of Śaunaka with Text
and Translation. Delhi: Motilal Banarsidass.
Sāmavidhānabrāhmaṇa. Sharma, B.R. 1964. Sāmavidhānabrāhmaṇa with
Vedārthaprakāśa of Sāyana and Padārthamātravivṛti of Bharatasvāmin. Tirupati:
Kendriya Sanskrit Vidyapeetha.
Śāṅkhāyanagṛhyasūtra. Sehgal, S.R. 1987. Śāṅkhāyana Gṛhya Sūtram (Belonging to the
Rgveda). 2nd ref. ed., Delhi: Sri Satguru Publications.
Saptabuddhaka. To. 270, 512, 852. Tog 472. T. 1127, 1333, 1334.
Satyāṣāḍhagṛhyaśeṣasūtra. Āgaśe, Kāśīnāthaśāstrī. 1907–32. Satyāṣāḍhaviracitaṃ
Śrautasūtram. 10 vols; vol. 8, pp. 1–123. Puṇyākhyapattane: Ānandāśramamudraṇāyala.
Śikṣāsaṅgraha. Vyāsa, Paṇḍit Yugalakiśora. 1893. A Collection of Śikshás by Yájñavalkya
and Others, With Commentaries on Some of Them. Benares Sanskrit Series, nos. 35,
40, 42, 44 & 46. Benares: Braj B. Das & Co.
Śrāvakabhūmi. Shukla, Karunesha. 1973. Śrāvakabhūmi of Ācārya Asaṅga. Patna: K.P.
Jayaswal Research Institute.
Śuklayajurvidhāna. Annāśāstri, K.P. 2001. Śuklayajurvidhānasūtram Vāgharakula-
bhūṣaṇamahārājaśrīkālanāthapraṇītayajurmañjaryākhyavyākhāsamalaṅkṛtam, pp.
1–50 Varanasi: Chaukhamba Vidyabhavan. Śarmā, Rameśacandra Dāśa. 2014.
Yajurvidhānasūtram. New Delhi: Mānyatāprakāśanam. Also Śikṣāsaṅgraha, pp.
327–361.
Suvarṇabhāsottamasūtra. Nobel, Johannes. 1937. Suvarṇabhāsottamasūtra: Das
Goldglanz-Sūtra: Ein Sanskrittext des Mahāyāna-Buddhismus. Leipzig: Otto
Harrassowitz.
Taittirīyasaṃhitā. Weber, Albrecht. 1871–72. Die Taittiîya-Saṃhitâ. Indische Studien
11–12. Śástrī, A. Mahádeva and Rangacharya, K. 1894–98. The Taittiríya Saṃhitá of
the Black Yajur-Veda with the Commentary of Bhaṭṭabháskaramiśra. 12 vols. Mysore:
Government Branch Press.
Theragātha. Oldenburg, Hermann and Richard Pischel. 1883. The Thera- and Therî-
Gâthâ: (Stanzas Ascribed to Elders of the Buddhist Order of Recluses.). London: Pali
Text Society.
Vaikhānasagṛhyasūtra. Caland, W. 1927. Vaikhānasagṛhyasūtram and
Vaikhānasadharmasūtram [Vaikhānasasmārtasūtram]. Calcutta: Asiatic Society of
Bengal.
Vājasaneyisaṃhitā. Shastri, Jagdishlal. 1978. Vājasaneyi-Mādhyandina-Śukla-Yajurveda-
Saṁhitā. Delhi: Motilal Banarsidass.
Vārāhagṛhyapuruṣa. Rolland, Pierre. 1971. Un Rituel Domestique Védique: Le
Vārāhagṛhyasūtra. D’aix-en-Provence: Publications Universitaires de Lettres et
Sciences Humaines.
A Nāga Altar in 5th century India 167
Secondary Sources
Agrawala, V.S. 1948. Terracotta figurines from Ahichchhatrā. Delhi: Prithivi Prakashan.
Agrawala, P.K. 1991–92. ‘The Kumbhāṇḍas in Indian Art and Literature.’ Prāgdhārā—
Journal of the U. P. State Archaeological Organisation 2, 97–102.
Atkinson, Edwin T. 1882–86. The Himalayan districts of the North-western Provinces of
India (The Himalayan Gazetteer). 3 vols. in 6 parts. Allahabad: North Western
Province and Oudh Government Press.
Bloomfield, Maurice. 1913. ‘The Character and Adventures of Muladeva.’ Proceedings of
the American Philosophical Society 52, 616–650.
Bloss, Lowell W. 1973. ‘The Buddha and the Nāga: A Study in Buddhist Folk Religiosity.’
History of Religions 13/1, 36–53.
Bolling, G.M. 1913. ‘The Çāntikalpa of the Atharvaveda.’ Journal of the American Oriental
Society 33/3, 265–78.
Brown, Robert L. 1991. Ganesh: Studies of an Asian God. Albany, NY: State University of
New York Press.
Bühnemann, Gudrun. 1988. Pūjā: a study in smārta ritual. Vienna: Publications of the
De Nobili Research Library.
van Buitenen, Hans. 1958. ‘The Indian Hero as Vidyādhara.’ Journal of American Folklore
71, 305–311.
Burnouf, Eugène. 1844. Introduction à l’histoire du Buddhisme indien. Paris: Imprimerie
Royale.
Cohen, Richard S. 1998. ‘Nāga, Yakṣiṇī, Buddha: Local Deities and Local Buddhism at
Ajanta.’ History of Religions 37/4: 360–400.
Courtright, Paul B. 1985. Gaṇeśa: Lord of Obstacles, Lord of Beginnings. New York and
Oxford: Oxford University Press.
Crooke, W. 1896. The Popular Religion and Folk-lore of Northern India. 2 vols.
Westminster: Archibald Constable & Co.
Cunningham, Alexander. 1871. Archaeological Survey of India: Four Reports Made
During the Years 1862–63–64–65. Simla: Government Central Press.
Dandekar, R.N. 1958–1995. Śrautakośa: English Section. 2 vols. in 5 parts. Poona: Vaidika
Saṁśodhana Maṇḍala.
Davidson, Ronald M. 2009. ‘Studies in Dhāraṇī Literature I: Revisiting the Meaning of
the Term Dhāraṇī.’ Journal of Indian Philosophy 37, 97–147.
168 Davidson
Jaini, Padmanabh S. 1966. ‘The Story of Sudhana and Manoharā: An Analysis of the
Texts and the Borobudur Reliefs.’ Bulletin of the School of Oriental and African
Studies 39, 533–558.
Kala, Satish Chandra. 1980. Terracottas in the Allahabad Museum. New Delhi Abhinav
Publications.
Knipe, David M. 1967. ‘The Heroic Theft: Myths from the Ṛgveda IV and the Ancient
Near East.’ History of Religions 6/4, 328–360.
Kramrisch, Stella. 1939. ‘Indian Terra-Cottas.’ Journal of the Indian Society of Oriental
Art 7, 89–110.
Lal, B.B. 1993. Excavations at Śṛiṅgaverapura (1977–86), vol. 1. New Delhi: Archaeological
Survey of India.
Lüders, Heinrich. 1907–08. ‘Three Early Brâhmî Inscriptions.’ Epigraphia Indica 9,
239–248.
Lüders, Heinrich. 1961. Mathurā Inscriptions—Unpublished Papers Edited by Klaus L.
Janert. Abhandlungen der Akademie der Wissenschaften in Göttingen Philologisch-
Historische Klasse Dritte Folge, Nr. 47. Göttingen: Vandenhoeck & Ruprecht.
Malamoud, Charles. 1996. Cooking the World: Ritual and Thought in Ancient India.
Translated from the French by David White. Delhi: Oxford University Press.
Mahalingam, T.V. 1965. ‘The Nagas in Indian History and Culture.’ Journal of Indian
History 43/1, 1–69.
Marshall, John H. 1909. ‘Rājagṛha and its Remains.’ Annual Report, Archaeological
Survey of India, 1905–06, 86–106
Matsumura, H. 1983. ‘A Text on Esoteric Iconography from the Gilgit Manuscripts.’
Mikkyo Zuzo 2, 71–79.
Nazim, M. 1940. ‘Excavations at Rajgir.’ Annual Report of the Archaeological Survey of
India, 1936–37, 45–47.
Oldham, C.F. 1901. ‘The Nāgas. A Contribution to the History of Serpent-Worship.’
Journal of the Royal Asiatic Society of Great Britain and Ireland July, 1901, 461–73.
Pulleyblank, Edwin G. 1991. Lexicon of Reconstructed Pronunciation in Early Middle
Chinese, Late Middle Chinese, and Early Mandarin. Vancouver, B.C.: University of
British Columbia Press.
Rawlinson, Andrew. 1986. ‘Nāgas and the Magical Cosmology of Buddhism.’ Religion 16,
135–153.
Ray, Himanshu Prabha. 2006. ‘The Archaeology of Bengal: Trading Networks, Cultural
Identities.’ Journal of the Economic and Social History of the Orient 49/1, 68–95.
Sander, Lore. 1968. Paläographisches zu den Sanskrithandschriften der Berliner
Turfansammlung. Wiesbaden: Franz Steiner.
Schopen, Gregory. 2002. ‘Counting the Buddha and the Local Spirits In: A Monastic
Ritual of Inclusion for the Rain Retreat.’ Journal of Indian Philosophy 30, 359–388.
170 Davidson
Marzenna Czerniak-Drożdżowicz
* The research on South Indian temple cult is being conducted in the frame of the research
grant of the Polish National Centre of Science, decision number UMO-2011/03/B/HS2/02267.
1 See for example Colas 1996: 302.
2 I have already treated some aspects of the pratiṣṭhā ceremony in my 2011 book (Czerniak-
Drożdżowicz 2011) and in the text presented at the CEENIS (Central and East European
Network for Indian Studies) Conference in Warsaw in November 2011 (to be published
shortly). On the relation of Pāñcarātra and Śrīvaiṣṇava see e.g. Young 2007.
The five modes/kinds of god’s presence in the world are: para, vyūha,
vibhava, arcāvatāra and antaryāmin. First references to them as a group can be
found in such Pāñcarātrika sources as the Īśvarasaṃhitā and Śrīpraśnasaṃhitā,
which are both probably not earlier than the 13th c. AD,3 therefore it is possible
that the idea of these five gathered in a group developed over time.4
Para is the highest form, sometimes called nityavibhūti—the eternal mani-
festation. It is everlasting, existing above this universe in the highest abode
of Viṣṇu called parama pada, indentified with Viṣṇu’s heaven Vaikuṇṭha or
described as dwelling above or outside it. This form is not accessible to devo-
tees and often called Nārāyaṇa or Vāsudeva. It is described, for example, in the
Sāttvatasaṃhitā 1.25–26. The text reads:
Śrībhagavān uvāca
ṣāḍguṇyavigrahaṃ devaṃ bhāsvajjvalanatejasam |
sarvataḥ pāṇipādaṃ tat sarvato ‘kṣiśiromukham || 1.25
param etat samākhyātam ekaṃ sarvāśrayaṃ prabhum |
‘Śrībhagavān said:
The god has a form of six qualities, bright as a shining sun,
with limbs everywhere, with eyes, head and face [present] everywhere.
They called Him the highest (para), the only one, being a dwelling/refuge
of everything, the Lord.’
One of the crucial features of this form is its possessing the six divine charac-
teristics (guṇa)5 in their highest/upmost degree.6 The para form has particular
iconographical representations and is often depicted as four-armed and lying
on the Śeṣa snake. Such a form, having snake’s seven hoods above his head, can
be called ādimūrti.7 Para is also represented as accompanied by Garuḍa and
other deities such as Viṣvaksena, who form an usual group of attending gods.
3 In the Śrīvaiṣṇava tradition probably the earliest references to these five forms can be found
in the Varadarājastava of Kūreśa, the 9th c., the pupil of Rāmānuja (Young 2007: 196).
4 Īśvarasaṃhitā 20.263cd–264 (see appendix 2.4).
Śrīpraśnasaṃhitā 2.54cd–55ab (see appendix 7).
5 These are: knowledge—jñāna, strength—bala, lordship/divinity—aiśvarya, bravery—vīrya,
glare/light—tejas and power—śakti.
6 The definition of these qualities can be found for example in the Ahirbudhnyasaṃhitā
2.56–61ab.
7 According to the Ahirbudhnyasaṃhitā (9.31ab), in this form fully possessing the six divine qual-
ities, the god is accompanied by Lakṣmī (devyā lakṣmyā samāsīnaṃ pūrṇaṣāḍguṇyadehayā).
The form is also identified with Ādimūrti, imagined as the one sitting or lying on the Śeṣa
Image Consecration In The Pāñcarātrika Sources 173
serpent, which has his seven hoods forming a kind of umbrella above the head of god.
The right leg of the sitting figure is usually stretched, while the left one is folded; the right
hand touches the snake and the left one is placed on the left knee. In the other two hands
the four-armed Viṣṇu holds a conch (śaṅkha) and discus (cakra). He is depicted in the
coral colour, on his right side there is Bhṛgu and Brahmā, on the left Mārkaṇḍeya and Śiva.
See for ex. Rao 2009: 68–79.
8 In the Vaikhānasa tradition these emanations are called differently: Puruṣa, Satya, Acyuta
and Aniruddha.
9 More elaborate information concerning the concept of vibhūti and antaryāmin one can
find in the works by G. Oberhammer concerning the philosophical school of Rāmānuja;
see for example Oberhammer 1998 and Oberhammer 2000.
10 The Sanskrit text of the Upaniṣad after Srinivasa Chari 2004: 225, n. 22.
174 Czerniak-Drożdżowicz
The fifth mode, arcāvatāra is in fact usually listed as the fourth, before
antaryāmin. It is the most interesting one from the point of view of our topic
because it refers to god present in his representations. Among these represen-
tations some belong to the category of svayaṃvyakta, in which god manifests
himself of his own will, and as such one understands for example the main
mūrtis from Srīraṅgam, Melkoṭe and Tirupati-Tirumala Vaiṣṇava temples.
Some of these representations belong to the divya/daiva category, consecrated
by gods, which means that due to requests of devotees god comes down to
earth in the form of his idol. To this category one can count for example the
main mūrti in the Varadarāja temple in Kāñcīpuram. The third category is
known as saiddha, which refers to the idols consecrated by wise men, when
god, in response to their austerity (tapas), incarnates himself in the form of an
idol. The category known as mānuṣa refers to the representations created by
men in which god appears through the consecration ceremony described in
āgamas, and most of the temple idols belong to this type.11 The arcāvatāras,
concrete representations, very attractive for the devotees, were mentioned by
the Āḻvārs and Śrīvaiṣnava ācāryas.12 A particular accessibility (saulabhya) of
god in his idols, despite of the devotees‘ social and religious status, decides in
favour of the popularity of this mode of god’s existence.
The form can be imagined or material, but due to the god’s presence, its
shape and the way it is worshipped is regulated by prescriptions. God in his
material forms could reside only in particular places, which are his shrines and
temples, constructed and consecrated according to the rules.13
In the context of the Pāñcarātra and Vaikhānasa temples, usually five types
of representations are mentioned.14 The main unmovable idol is known
as dhruvabera/mūlabera and is installed in the main chapel (garbhagṛha).
11 For example in ĪS 19.50 (see appendix 2.2); see for ex. Srinivasa Chari 2000: 207–232; Rao
2005, ch. 5 ‘The Deity and its Modes.’
12 See for ex. Narayanan 1985, where the author writes that in the texts by the Āḻvārs one can
find religious hymns directed to 108 forms of god, connected with the particular places in
which images are established.
13 Among necessary ceremonies for establishing the temple are those of prathameṣṭakānyāsa,
garbhanyāsa and murdheṣṭakānyāsa. See for example Ślączka 2007 or Czerniak-
Drożdżowicz 2011.
14 For these five types of images in the Vaikhānasa tradition see for example Colas 1996:
296, Colas 1986:70–75 and Hikita 2005:146, f. 14. These types are mentioned for example in
the ĪS 19.763–769. Nevertheless, for example the PādS in its kp (19.1–2, see appendix 4.2)
speaks about six kinds of idols.
The Īśvarasaṃhitā (17.238cd–239ab, see appendix 2.1) also mentions 6.
Image Consecration In The Pāñcarātrika Sources 175
15 These three types of idols are often accompanied by goddesses Śrī and Bhū. If the idol is
not accompanied by them, it is called ekabera, meaning single, lonely image.
16 Bera does not appear in the Monier-Williams and probably is a deformation of the
Sanskrit word vera meaning the body (see Colas 1986: 71).
17 For more information on images in the Vaiṣṇava tradition see for ex. Colas 1986: 70–75,
Narasimhan 2007.
18 For the meaning of the word, different kinds of this ceremony, detailed description of its
elements and some examples from Pāñcarātrika sources (mostly SāS, ĪS, PārS and JāyS)
see Hikita 2005.
Vedic literature, and especially early texts of vedāṅgas, mostly those concerning ritual
(kalpa), do not supply information about specific buildings in which the god’s idols were
worshipped, since they refer mostly to the fires distributed on the sacrificial ground on
a specially constructed altars, but not in a building. Only in later texts connected with
the ritualistic literature, for example gṛhyasūtrapariśiṣṭa, or in the commentary on the
Kauśikasūtra entitled Kauśikapaddhati (11th c. AD), does the term maṇḍapa appear with
the meaning of the pavilion in which the ritual is to be performed.
19 It is one of the four pādas, namely topics treated by the Tantric texts. The other three are:
jñānapāda, concerning theological and philosophical issues, caryāpāda concerning ritu-
als meant for individual devotee and yogapāda, concerning meditation practice.
20 These subjects are strictly connected with the religious ideas therefore the process of
creation of the idol and construction of the temple is not the result of an independent
decision of a sculptor or architect. Among the earliest Tantric sources that refer to these
ceremonies there is a Śaiva text Rauravāgama dated to around the 8th century AD, in
which the description of liṅgapratiṣṭhā can be found. On Raurava texts see for ex. Goodall
2004: XLIV–XLVI. Then it can be found in other early Śaiva works, also in the XXVII āhnika
of the Abhinavagupta’s Tantrāloka (10th/11th c. AD).
176 Czerniak-Drożdżowicz
21 The Pāñcarātrika text Pādmasaṃhitā (cp 13.101cd–105ab, see appendix 4.1) speaks about
five types of installation in connection with the types of god’s representations: sthāpanā,
āsthāpanā, saṃsthāpanā, prasthāpānā and pratiṣṭhā. A similar description can be found
in the Pāñcarātrika text the Viṣṇusaṃhitā (15.2–3 see appendix 9.1), though the last type is
not called pratiṣṭhā, but pratiṣṭhāpanā.
22 I will present the installation ceremony making use of selected texts, mostly the
Paramasaṃhitā, but descriptions of the pratiṣṭhā and sthāpana can be found in
other Pāñcarātrika texts as well, for ex. Sātvatasaṃhitā ch. XXV entitled ‘rule concern-
ing installation’—pratiṣṭhāvidhi, Jayākhyasaṃhitā ch. XX, ‘rule concerning instal-
lation’—pratiṣṭhāvidhi, Pādmasaṃhitā kp ch. XIII ‘rule concerning installation of
the trident/flag’—śūlasthāpanavidhi, XIV—‘rule concerning the shape/colour of
the deities’—devatāvarṇavidhir, XV—‘description of different sitting and standing
[poses]’—sthityāsanādibhedanirṇaya, XXVIII—‘rule concerning installation’—
pratiṣṭhāvidhi; XXIX—‘rule concerning installation of the fish and other [avatāras]’—
mīnādipratiṣṭhāvidhi; Īśvarasaṃhitā ch. XVI—‘rule concerning installation, the
temple etc.’ prāsādādipratiṣṭhāvidhi, XVIII—‘rule concerning installation’—
pratiṣṭhāvidhāna, XIX—‘rule concerning reparation/expiation’—prāyaścittavidhi;
Viṣvaksenasaṃhitā ch. X—‘rule concerning the characterisctics of the image’—
pratimālakṣaṇādividhi, XI—‘rule concerning disposition of the weapon and colours’—
varṇāyudhavinyāsavidhi, XVI—‘rule concerning instalation’—pratiṣṭhāvidhi,
XVIII—‘rule concerning installation of goddess etc.’ devīsthāpanādividhi; Viṣṇusaṃhitā
ch. XVIII—‘installation’—pratiṣṭhā.
23 Interesting material concerning construction, installation and consecration of the
divine images in the Pāñcarātra tradition can be found in the Devāmṛtapāñcarātra, newly
discovered in Nepal by Dr. Diwakar Acharya (Oxford). The text has been recently pub-
lished, but I was provided with the passage before the publication. I would like to present
some of the beginning verses (according to the numbers used by D. Acharya) from the
critical edition prepared by Dr. Acharya (Acharya 2015): Devāmṛtapāñcarātra 1.7–17 (see
appendix 1).
24 ParS 3.5–11ab (see appendix 5.1). Similarly reads the Viṣṇusaṃhitā (29.55cd–57, see appen-
dix 9.2). The Paramasaṃhitā also differentiates forms in accordance with the goal to be
achieved by a particular ritual: apara, which is the eight-armed one, is worshipped for
worldly goods (bhoga), while para, which is four-armed, is worshipped for emancipation
(mukti).
Image Consecration In The Pāñcarātrika Sources 177
a presence of god.25 After that ceremony the object is not the same anymore—
the installation being a moment of transition from the material structure
to the residence of a god, who in his idol becomes an object of worship.
The description of the installation and consecration, for example in the
ParS (ch. 18 and 19) usually begins with preliminaries called adhivāsana,26
which includes the description of founders and performers; the right star
constellation is also considered. The pavilion for the preliminary ceremony
(adhivāsamaṇḍapa) is built in the vicinity of the temple. In this pavilion, on
the pedestal (snānapīṭha) is performed the bath of the idol.27 The idol, whose
size is correlated with the measurements of the temple, is left in the water for
at least one day. The ceremony is performed by the priest, whose body is made
holy trough the imposition of mantras (nyāsa).
On the second day the idol is taken out of the water, brought into the pre-
viously prepared pavilion and placed on the cushion. The teacher opens the
eyes of the idol (unmīlana), using a sharp device, and the craftsman (śilpin)
places clarified butter and honey on the pupils of the idol. The idol is bathed
with water, dredged with grains and rice and cleaned with five cow products
(pañcagavya). All these activities are performed with the accompaniment of
the five mantras (pañcamantra).28 The idol is smeared with earth brought
from holy pilgrimage places and mountains’ peaks, dug with buffalo or boar
horn, or elephant tusk. If such earth is not available, he uses earth consecrated
with mantras. Sandal paste, water from holy places and other substances are
applied. New cloths and the sacred thread are then offered to the idol, as well
as sandal paste, oils and ornaments, lamps, an umbrella and garlands. Above
25 As for example Venkatachari points out (Venkatachari 2005: 33): ‘The function of pratiṣṭhā
rituals is to charge a physical object with spiritual presence or to change its perceived
nature in such a way so that it is ‘seen.’ After its sanctification, the object is essentially
different from what it was before. These rituals represent, as it were, a juncture, a tran-
sitional moment when a structure of stone or wood becomes a sacred shrine; when a
piece of granite or bronze or plaster becomes the palpable presence of the divine plenum;
when the ordinary becomes extraordinary; and when the profane becomes venerable. In
other words, the material object made by the contractors, carpenters and craftsmen—the
takṣakas and śilpins—is transformed into a spiritual one by the pratiṣṭhā rituals done by
the ācārya. It is this radical transformation that animates the image, giving life to the pūjā,
and eliciting temple honors for the image-form of God.’
26 Hikita (2005: 148) writes, following Smith (1984: 54–63), that one of the late Pāñcarātrika
texts, Kapiñjalasaṃhitā, (not available to me) enumerates 18 steps for the image installa-
tion, although Smith (1984) mentions in this context 16. See footnote 35 below.
27 Though the best place for the ritual is the river.
28 ParS 18.50cd–52ab (see appendix 5.2).
178 Czerniak-Drożdżowicz
the idol the canopy is placed and fans are given to him. After that the idol is
taken back to the pavilion and laid down on the comfortable bedding with
his head to the south. The teacher should reconstruct his body, which is done
through the imposition of mantras. Then begins the next ceremony, the invo-
cation of the god into his idol (āvāhana). After worshipping god in the idol the
teacher asks him to be permanently present in it. Since the god is treated as a
guest or king, arghya and pādya waters, as well as food (nivedya) are given to
him. The next ceremony is the fire offering (agnikārya) performed in the three
sacrificial altars (kuṇḍa).29 The Brahmins who participate in the ceremony
imagine a lotus in the fire, having deities on the petals and stamens, and they
should worship these deities. Money is given to the Brahmins while the god is
again provided with water, clothes, a canopy and then completely covered to
be unseen. Then the brahmaśilā stone, which is to be installed at the base of
the idol, is worshipped and the night is spent wakeful with dances and singing.
On the next day the place in which the installation will take place is again
thoroughly checked.30 Also the idol itself should not be damaged and should
not have any blemish. Then, after a bath, the teacher circumambulates the
pavilion and recreates his body using the mantras. He cleans the temple,
places flags on the gates and vessels at the entrance and collects all the sub-
stances needed for offerings as well as jewels, minerals and seeds; then come
also musicians. Bad omens are pacified, which is done by offerings of clari-
fied butter. Further the teacher circumambulates the idol, scatters seeds and
holy grass, cleans the whole place with water consecrated with the protective
(astra) mantra and deposits the substances with proper mantras. Butter is
offered in the fire to pacify the powers of the place (vāstuśānti) and demons.
The teacher also assigns the directions within the garbhagṛha.31 First the ped-
estal with the brahmaśila stone is installed in the middle of the garbhagṛha,
then the stone is worshipped and 9 jewels and other substances are deposited
29 In them the offering is done 12, 8 or 4 times. In the middle kuṇḍa clarified butter is offered
for all deities and in the side one for the guardians of months (mūrtipāla). Havis offering
is given to the god in the central kuṇḍa.
30 For example, whether there is enough water in the vicinity of the temple, whether it
is free from bad people and illnesses; it should be a sunny day and the stars should be
auspicious.
31 This is done with a string anointed with the juice of the sandalwood to precisely establish
the point of intersection of the lines assigning directions. The other substances are: dia-
mond (vajra), ruby (padmarāga), cat’s eye (vaiḍūrya), sapphire (nīla), pearl (mauktika),
topaz (puṣparāga), mother of pearl (śaṅkha), emerald (marataka) and crystal (sphaṭika).
Image Consecration In The Pāñcarātrika Sources 179
in the 9 holes (garta) situated in all directions.32 Afterwards the priest prepares
the runoff for the water (pūjāvāripatha) and when the pedestal is ready, the
idols of accompanying deities are prepared for installation.33 Then the idol is
worshipped and the god is again invoked into it. The priest offers all required
substances to the god present in the idol and places the idol in the palanquin
in which it is taken around the temple with sounds of conches, drums, songs
and with umbrellas, canopies, flags, lamps, incenses etc. Then the idol is taken
inside the garbhagṛha, in which it is installed on the pedestal with its face to
the east. To position the idol in the right way, the peg on the lower part of the
idol should come exactly into the hole of the pedestal. This is done with the
use of a rope hanging above the pedestal. The pedestal is seen as a throne of
god, and when the god in his idol is properly established there, all the sacrifi-
cial substances are again offered to him. After covering the idol with cloths, the
priest installs the idols of the god’s attendants. Gifts should be given to all the
participants in the installation, usually food and drinks, cloths and perfumes.
For the first three days after installation, the god in his idol is not worshipped
and the regular service begins on the fourth day, after the bath of the idol.
The cult directed to the so installed idols is then thoroughly described in the
Pāñcarātrika sources.34
The general schedule of this ceremony presented in different texts is similar,
though the descriptions could differ in some details.35 Thorough descriptions
can be found also in the literature of the Vaikhānasas.36
32 These jewels are: diamond (vajra), ruby (padmarāga), cat’s eye (vaiḍūrya), sapphire (nīla),
pearl (mauktika), topaz (puṣparāga), mother of pearl (śaṅkha), emerald (marataka) and
crystal (sphaṭika).
33 In the case of Viṣṇu also Garuḍa and Ananta. There are also places for guardians of
the directions assigned. One of the important deities, in case of the Vaiṣṇava temples,
is Viṣvaksena. To this deity are given the remnants from the offerings. Pedestals for bali
offerings are also prepared.
34 For example Pādmasaṃhitā chapter 5.
35 Smith for example presents 16 subsequent elements of the ceremony according to
the Kapiñjalasaṃhitā (only in Telugu script): 1. maṇḍapa, 2. jalādhivāsa, 3. snapana,
4. śayanādhivasa, 5. agnisaṃskāra, 6. brāhmaṇabhojana, 7. śāntihoma, 8. sparśana,
9. gehaśuddhi, 10. ratnanyāsa, 11. pūrṇāhuti, 12. sthāpana, 13. prokṣaṇa-balageha,
14. mantranyāsa, 15. parivāra, 16. dakṣiṇā; see Smith 1984: 54f.
36 Among the most important of their texts is the Vimānārcanākalpa attributed to Marīci,
thus known also as the Marīcisaṃhitā; see Colas 1986. The presentation of the Vaikhānasa
texts dealing with the idols, their creation and consecration in the temple one can find in
the three volumes of the Vaikhānasāgamakośa, published in Tirupati (Rashtriya Sanskrit
Vidyapeetha) in 1991–2004.
180 Czerniak-Drożdżowicz
37 Damages can be caused by different factors: atmospheric circumstances, but also by break-
ing of some parts, scratches, leakage of water, fungus, insects and animals. Desecration of
the holy object or space can be also caused by an inappropriate behaviour of people; see
for example Narasimhan 2005.
38 Viṣṇusaṃhitā 24.3: dehaṃ dehī yathā jīrṇaṃ tyaktvā dehāntaraṃ vrajet / tyaktvā jīrṇaṃ
tathā bimbaṃ devo ‘pi bhajate navam //—‘As, abandoning the old body, the inhabitant of
the body moves to the other body, in the same way, abandoning the old idol, the god uses
the new one.’
39 See for example Viṣṇusaṃhitā 24.36–70.
Image Consecration In The Pāñcarātrika Sources 181
arcāvatāra that is crucial for understanding of the god’s presence in this world
in his representations.
There are not too many and direct references to this issue in the Pāñcarātrika
texts.40 Here one can refer to the works of, for example, Phyllis Granoff, who
supposes that at first the cult of images was a domain of earlier Bhāgavatas
and only then was taken up and appropriated by the Pāñcarātrikas (Granoff
2006). More about arcāvatāra can be read in the literature of the Śrīvaiṣṇavas,
being immediately associated with the Pāñcarātra. Nevertheless, some clues
can be found for example in such Pāñcarātrika texts as the Jayākhyasaṃhitā
(circa 8th–9th c. AD), the Pādmasaṃhitā (probably not earlier than 12th c. AD)
the Viśvamitrasaṃhitā (after the 11th c. AD), the Pārameśvarasaṃhitā (12th–
13th c. AD); there are some clues also in the Viṣṇudharmottarapuraṇa (dated
from 5th c. AD to even 11th c. AD), which is closely connected with the Pāñcarātra
(Granoff 2006). Actually, the two earlier texts, namely JayS and VDhP present
some skepticism towards the cult of idols and the process of ensuring god’s
presence in his images. VDhP reads that, in fact, if god is omnipresent and does
not need anything, so what is the reason for a special invocation—3.108.05cd:
āvāhanena kiṃ kāryaṃ tasya sarvagatasya tu //, what is the reason of offer-
ings to the god, who is fully and eternally satisfied—3.108.28ab: nityatṛpto na
tṛptyartham upahāraṃ pratīcchati [Ed.; prayacchati] /—‘One who is always
satisfied does not receive/want a gift/oblation for satisfaction’; what are
offerings to his idol for—3.108.18ab: nityatṛptasya kiṃ tasya kāryaṃ bhavati
cārcayā /—‘For eternally satisfied one what is the activity/ritual [done] for
with [the usage of] the image?
The answer given by this text is that it is done in fact for a pleasure of
worshippers: tasyārcākāraṇam viddhi kartṛprītiprayojanam (3.108.20ab) and
also for accomplishing one’s own favour: evam āvāhanaṃ tasya tathā pūjā ca
yādava / prabuddhaiḥ kriyate yatnād ātmānugrahakāraṇāt //—‘In this way the
invocation of him, as well as [his] worship, O Yādava should be done with the
effort by the enlightened ones with the aim of [one’s] own favour.’ (3.108.27).41
The doubts expressed in this text are also connected with the need of an addi-
tional invocation (āvāhana) of gods already installed—āvāhitāḥ sannihitās
tu devā bhavanty avaśyaṃ nṛpa mantrayuktyā / svatuṣṭaye devavarasya viṣṇor
āvāhanaṃ vajra budhaiḥ pradiṣṭam (3.108.31) //—‘[Already] Invoked, gods
are certainly fixed/present with the usage of mantras, O Lord. For one’s own
PādS is one of three rūpas of the Highest God.48 They can appear as mental
images adored in meditation, as mantras encompassing god’s potencies and
used for worship, and also as arcās—the images worshipped in temples. There
are many of them, each one having its characteristics and presenting manifold
presence of god in the real world.
The PādS stresses that it is the mantra that has a decisive role in bringing
god near to his devotees and this feature is clearly visible in the rituals of instal-
lation and consecration of the images. Through mantras god enters the image,
makes it alive and divine. The image is then charged with god’s presence and
becomes a real and living manifestation of him. PādS gives also a description
of a meditation by the ācārya performing the consecration. Through such a
meditation the powers of god are transmitted into the idol.49 Viṣṇu in his high-
est form (nityodita rūpa) manifests himself in the mūlamantra and the ācārya
draws him first into his heart and then transmits him into the image—all due
to the power of mantras he uses. Such a process brings about the descent
(niveśayet = avatāra) of god into the idol (pratimāyāṃ = arcāyāṃ), though in
the PādS one does not find the term arcāvatāra.
As the mūrtyutpatti passage of the PādS (PādS, jp, 2, 6c–40) states, the god’s
idols are divine mūrtis, the forms appearing in the process of creation, which
then are present in the temples incarnated in their idols. Every mūrti is a mani-
festation of the Highest God and every arcā is charged with his presence. Not
only the main arcā, but every one of them is pervaded by the highest form of
god.50 Mūrti is a condensation of divinity, taking shape of a particular divine
figure. Arcā is then pervaded by this divinity which enters the material shape
which, duly prepared, already exists. Through his meditation the ācārya is able
to transfer the god into his heart and then into the idol, so the ācārya himself
is a medium serving for transmission in this process.51
The Pārameśvarasaṃhitā, classified as a vyākhyā (elucidation/elaboration)
of the Pauṣkarasaṃhitā, clearly states the status of the image and the reason
for which god incarnates in it:
48 These three are Highest Vāsudeva (Para Vāsudeva), puruṣa and prakṛti.
49 PādS, kp, 28, 52–61 (see appendix 4.4).
50 For example, Vāsudeva is invoked to enter the images of his attendants: PādS, kp, 28,
34–38b (see appendix 4.3).
51 The god enters the image not directly, but via his devotee’s body, which, due to this fact, is
seen also as a kind of a mūrti of god. Interesting remarks concerning the issue of different
forms of god one can find in the PhD thesis of Silvia Schwarz, submitted at the Vienna
University in 2012 and published in 2014. See also Schwarz 2006.
184 Czerniak-Drożdżowicz
The god named Vāsudeva, who has a form consisting of six guṇas,
for the benefit/favour of the practitioners having approached/attained
the temple, being [his] solid/material body, rich in all knowledge and
ritual, being the abode of all tattvas; he, having entered into the right
form, gives his favour.52
The idea of arcāvatāra was elaborated especially by the Śrivaiṣṇavas, for whom
it was one of the crucial concepts. They gave a prominence to this particular
form, since for them through arcāvatāra god was fully present in the temple in
the mūlavigraha as well as in festival images, in the same manner as in his heav-
enly abode. Therefore, the temple was treated as a piece of heaven on earth and
the most important among these places was Srīraṅgam temple called Bhūloka
Vaikuṇṭha—‘Heaven on Earth’ (Narayanan 1994: 130–131). Also the substance
from which the idol was made had a special status in the Śrīvaiṣṇavas’ views.
The idol was not a body made of stone, but it was made of a non-earthly tran-
scendental substance called śuddha sattva, not connected with guṇas (p. 54).
Interestingly, as Narayanan (p. 62) observes, while Vedānta Deśika many times
mentions the fact that arcā is made of śuddha sattva, Pillai Lokācārya consid-
ers deliberations concerning the nature of arcāvatāra as vulgar and improper.
Praising the role of arcāvatāra Pillai Lokācārya says that its glory comes
from the fact that it represents at the same time the superiority of god but
also his accessibility, so the temple and the arcā are even better than heaven.
He also points out the mutual relationship of god and his devotees, since also
God himself needs devotees to be worshipped and served (Narayanan 1985:
64). He wrote: ‘Viṣṇu’s incarnation as an Inner Controller resembles water hid-
den deep in the ground. His incarnation as Transcendent Deity resembles the
distant water of the oceans surrounding the earth. His incarnations in the form
of emanations are like the inaccessible milk ocean. His glorious incarnations
in human form resemble rivers that periodically flood, then dry up. But his
incarnation in an image is like the full, deep pools in those rivers where water
is always available’ (Śrīvacanabhūṣaṇa 24).53 While, as Richard H. Davis (1999:
31) writes: ‘Other modes of Viṣṇu’s manifestation may be unattainable, irreg-
ular, or overwhelming, but Viṣṇu’s incarnation in an image is a calm, stable,
and easily reached as a pool of water. By rendering God physically present in
a particular fixed location, icons enable the whole liturgical system of temple
transactions between God and his human worshipers, which the Vaiṣṇava and
Śaiva texts prescribe in glorious detail.’
52
Pārameśvarasaṃhitā 11.236–237 (see appendix 6).
53
Śrīvacanabhūṣaṇa translated in Davis 1997: 31.
Image Consecration In The Pāñcarātrika Sources 185
Bibliography
Sanskrit Sources
Ahirbudhnyasaṃhitā of the Pāñcarātrāgama. 1986. Edited by M.D. Ramanujacharya.
Vol. I and II. First Reprint. Madras: Adyar.
Īśvarasaṃhitā. 1923. Prativādibhayaṅkarānantācāryaiḥ saṃśodhitā. (Śāstramuktāvaĺī
45). Kāñcī.
Īśvarasaṃhitā. 2009. Critically edited and translated in Five Volumes by Lakshmi
Tathacharya. Introduction by V. Varadachari and G.C. Tripathi. Delhi: Motilal
Banarsidass.
Jayākhyasaṃhitā: Jayākṣarasaṃhitā. Ms No. pra 49 vi vaiṣṇavatantram 2. Reel No. B
29/3, palm-leaf. Nepal German Manuscript Preservation Project (NGMPP), National
Archives, Kathmandu (NAK), Nepal. (photocopy courtesy to A. Sanderson).
186 Czerniak-Drożdżowicz
Secondary Sources
Acharya, Diwakar. 2015. Early Tantric Vaiṣṇavism: Three Newly Discovered Works
of the Pañcarātra. The Svāyambhuva-pañcarātra, Devāmṛtapañcarātra and
Aṣṭādaśavidhāna. Pondichéry: EFEO/IFP/Asien-Afrika-Institut, Universität
Hamburg.
Āgama Suṣamā. Glimpses of the National Seminar on Āgamas. 2005. Eds
Lakshminarasimha Bhatta, K. Hayavadana Puranik, Haripriya Rangarajan. Tirupati:
Rashtriya Sanskrit Vidyapeetha.
Image Consecration In The Pāñcarātrika Sources 187
Hikita, Hiromichi. 2005. ‘Consecration of Divine Images in the Temple.’ In From Material
to Deity. Indian Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New
Delhi: Manohar. 143–197.
Hudson, Denis. 2007. ‘The Vyūhas in Stone.’ In Studies in Hinduism IV. On the
Mutual Influences and Relationship of Viśiṣṭādvaita Vedānta and Pāñcarātra, eds
Gerhard Oberhammer and Marion Rastelli. Wien: Österreichische Akademie der
Wissenschaften, 123–178.
Lester, Robert C. (ed. and transl.) 1979. Śrīvacana Bhūṣaṇa of Pillai Lokācārya. Madras:
Kuppuswami Sastri Research Institute.
Narasimhan, K.K.C. Lakshmi. 2005. ‘Renovation and Consecration of Temples and
Icons (Jīrṇoddhāraṇa/Navīkaraṇa).’ In Āgama Suṣamā. Glimpses of the National
Seminar on Āgamas, eds Lakshminarasimha Bhatta, K. Hayavadana Puranik,
Haripriya Rangarajan. Tirupati: Rashtriya Sanskrit Vidyapeetha, 202–216.
Narasimhan, K.K.C. 2007. A Study of Vaikānasa Iconography. Mumbai: Ananthacharya
Indological Research Institute.
Narayanan, Vasudha. 1985. ‘Arcāvatāra: On Earth as He Is in Heaven.’ In Gods of Flesh,
Gods of Stone. The Embodiment of Divinity in India, eds J.P. Waghorne and N. Cutler.
Chambersburg: Anima, 53–66.
Narayanan, Vasudha. 1994. The Vernacular Veda: Revelation, Recitation, and Ritual.
Columbia, SC: University of South Carolina Press.
Oberhammer, Gerhard. 1998. Materialien zur Geschichte der Rāmānuja-Schule IV. Der
‘Innere Lenker’ (antaryāmī). Wien: Österreichische Akademie der Wissenschaften.
Padoux, André (ed.) 1990. L’image divine: culte et meditation dans l’hindouisme. Paris:
Editions de CRNS.
Rangachari, Dewan Bahadur K. 1931. The Sri Vaishnava Brahmans. Madras: Bulletin of
the Madras Government Museum (reprint 1986).
Rao, S.K. Ramachandra. 2005. The Agama Encyclopedia (Revised Edition of Āgama-
Kosha). Vol. IV Pāñcarātrāgama, vol. IX Consecrations. Delhi: Sri Satguru
Publications.
Rao, T.A. Gopinatha. 1997. Elements of Hindu Iconography. Vols I and II. Delhi: Motilal
Banarsidass (repr.)
Rastelli, Marion. 2002. ‘The Āsana according to the Pārameśvarasaṃhitā or a method of
writing a saṃhitā.’ In Studies in Hinduism III. Pāñcarātra and Viśiṣṭādvaitavedānta,
eds Gerhard Oberhammer and Marion Rastelli. Wien: Österreichische Akademie
der Wissenschaften, 9–32.
Rastelli, Marion. 2003. ‘Der Tempel als Mythisierung der Transzendenz.’ In
Mythisierung der Transzendenz als Entwurf ihrer Erfahrung. Arbeitsdokumentation
eines Symposions, eds Gerhard Oberhammer and Marcus Schmücker. Wien:
Österreichische Akademie der Wissenschaften, 313–348.
Image Consecration In The Pāñcarātrika Sources 189
Schwarz Linder, Silvia. 2006. ‘Remarks on the Doctrine of the Jīva in the Pādmasaņhitā.’
In Tantra and Viśiṣṭādvaitavadānta, ed. M. Czerniak-Drożdżowicz. Cracow
Indological Studies 8, 109–132.
Schwarz Linder, Silvia. 2014. The Philosophical and Theological Teachings of the
Pādmasaṃhitā. Vienna: Österreichische Akademie der Wissenschaften.
Ślączka, Anna. 2007. Temple Consecration Rituals in Ancient India. Text and Archaeology.
Leiden, Boston: Brill.
Smith, H. Daniel. 1984. ‘Pratiṣṭhā.’ In Agama and Silpa. Proceedings of the Seminar held
in December, 1981, ed. K.K.V. Venkatachari. Bombay: Ananthacharya Indological
Research Institute.
Srinivasa Chari, S.M. 2000. Vaiṣṇavism. Its Philosophy, Theology and Religious Discipline.
Delhi: Motilal Banarsidass (repr.)
Srinivasa Chari, S.M. 2004. Fundamentals of Viśiṣṭādvaita Vedānta: a study based on
Vedānta Deśika’s Tattva muktā-kalpā. Delhi: Motilal Banarsidass.
Takashima, Jun. 2005. ‘Pratiṣṭhā in the Śaiva Āgamas.’ In From Material to Deity. Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New Delhi: Manohar,
115–143.
Vaikhānasāgamakośa, vol. I. 1991. eds N.S. Ramanuja Tatacharya, U. Shankara Bhatta,
Lakshminarasimha Bhatta, K.A. Balasubrahmanian, T.G. Anantasubrahmanyam,
M. Anantapadmanabha Bhatta; vol. II (Beranirmana Prakaranam, Part I) and
vol. III (Beranirmana Prakaranam, Part II). 2004. ed. Lakshmi-narasimha Bhatta,
Hayavadana Puranika. Tirupati: Rashtriya Sanskrit Vidyapeetha.
Valpey, Kenneth Rusell. 2006. Attending Kṛṣṇa’s Image. Caitanya vaiṣṇava mūrti-seva as
devotional truth. Abingdon, Oxon: Routledge.
Venkatachari, K.K.A. 2005. ‘Pratiṣṭhā Rituals with Special Reference to Pāñcarātra
Āgamas.’ In Āgama Suṣamā. Glimpses of the National Seminar on Āgamas, eds
Lakshminarasimha Bhatta, K. Hayavadana Puranik, Haripriya Rangarajan. Tirupati:
Rashtriya Sanskrit Vidyapeetha, 33–41.
Young, Katherine K. 2007. ‘Brāhmaṇas, Pāñcarātrins, and the Formation of
Śrīvaiṣṇavism.’ In Studies in Hinduism IV. On the Mutual Influences and Relationships
of Viśiṣṭādvaita Vedānta and Pāñcarātra, eds Gerhard Oberhammer and Marion
Rastelli. Wien: Österreichische Akademie der Wissenschaften, 179–261.
190 Czerniak-Drożdżowicz
1 Devāmṛtapāñcarātra 1.7–17
‘I would like to hear about the installation of the [idol of] god in detail.
Of which material one should create his pure image,
what should be the [proper] measurements for the main and less important limbs/
parts?
What should be the measure of the body marked with [proper] characteristics?
Of what measure should be makuṭa [crest/tiara] and the head?
Then, what should be the measure of the forehead,
what kind of nose and eyes, O great father?
What should be the measure of the cheeks and ears?
How to create lips and what should be the shape of the chin like?
What are the measures of arms, fingers and neck,
Image Consecration In The Pāñcarātrika Sources 191
how the torso should look like, how the chest, and how to make the circle of the navel.
What are the measures of the belly, hips and penis, 12
laps, knees, shinbones, ankles and feet,
finger nails, thumbs and the shapes of everything. 13
Then his cart should be made, O Janārdana [Viṣṇu].
What kinds of different weapons of god one should create? 14
How one should fashion Viṣṇu with all [required] characteristics?
How one should build the temple and beautiful pavilion? 15
How should the pedestal (piṇḍikā) look like, and what should be the measure of the
sanctuary (garbhasūtra)?
Furthermore, how does one decide about the choice of the day according to position
of the planets and constellations? 16
How should be the pavilion divided and what should be the shape of the altar?
What are the features of the priest who installs (sthāpaka) and who are those known
as mūrtidhara (assistants).
All this, O Lord, tell me shortly.’ 17
I am grateful to Dr. Acharya for sharing the text with me and enabling me to use the
verses presented above.
2 Īśvarasaṃhitā
2.1 17.238cd–239ab
karmārcā cotsavārcā ca balyarcā tīrthakautukam ||
nimittasnapanārcā ca śayanārceti ṣaḍvidhāḥ |
2.2 19.50
svayaṃvyakte tathā divye saiddhe cārṣādyamānuṣe |
jīrṇoddhāravidhiṃ vakṣye śṛṇudhvaṃ munisattamāḥ ||
2.3 20.16cd–17
kṣetreṣu puṣkarādyeṣu svayam eva jagatpatiḥ || 20.16
arcārūpeṇāvatīrya vyaktaiś cakrādilāṃchanaiḥ <cakrādilāñchanaiḥ> |
samanvitaḥ sannidhatte <saṃnidhatte> bhaktānāṃ hitakāmyayā || 20.17
‘In the places such as Puṣkara and others, the Lord of the world himself
having descended in the form of arcā with beautiful signs such as cakra and others
is marked [and resides] in the vicinity for the benefit of devotees (. . .).’
192 Czerniak-Drożdżowicz
2.4 20.263cd–264
namas te pararūpāya namas te vyūharūpiṇe ||
namo vibhavarūpāya namas te tv antarātmane |
namas te ‘rcāvatārāya nānākārāya te namaḥ ||
3 Jayākhyasaṃhitā
3.1 20.207cd–208
athādivāsanaṃ kuryād vidhidṛṣṭena karmaṇā ||
dhyānākhyaṃ niṣkalaṃ śuddhaṃ yena sannihitaḥ <saṃnihitaḥ> sadā |
mantro hy arcāgato vipra syāt paṭastho ‘thavā punaḥ ||
3.2 20.331
ārādhito ‘si bhagavan sādhakānāṃ hitāya ca |
tvayā ‘py anugrahārthaṃ ca vastavyam iha sarvadā ||
‘O, Lord! I have worshipped you for the welfare of the sādhakas and now in order to
show us your favour, you should reside here forever.’
4 Pādmasaṃhitā kp
4.1 13.101cd–105ab
sthāpanāsthāpane caiva tathā saṃsthāpanā ‘pi ca || 13.101
prasthāpanā pratiṣṭheti pratiṣṭhāpañcakaṃ smṛtam |
yā pratiṣṭhā bhavet sthāne sthāpanā sā prakīrtitā || 13.102
yā pratiṣṭhāsane proktā sā cāsthāpanasaṃmitā |
śayane yā pratiṣṭhā ca sāca saṃsthāpanā bhavet || 13.103
yāne ca yā pratiṣṭhā sā nāmnā prasthāpanā bhavet |
pratimā yārcanāpīṭhe karmārceti prakīrtitā || 13.104
tasyāṃ yā ca kriyā proktā sā pratiṣṭheti kīrtitā |
Installation which is applied in the case of the standing [image] is called sthāpanā,
For the standing one is called asthāpanā,
For the lying one, the one [called] saṃsthāpanā,
For the moving one this called prasthāpanā.
For the image which is on the pedestal and is called karmārcā
the installation performed for it is called pratiṣṭhā, this is declaired.’
4.2 19.1–2
Śrībhagavān
karmārcādipratikṛteḥ prakāraḥ kathyate ‘dhunā |
karmārcā cotsavārcā ca balyarcā ca tathaiva ca || 19.1
snānatīrthobhayārcā ca svāpotthānārcayā saha /
pratimāḥ ṣaḍ vidhātavyāḥ pūjāyām uttamā bhavet // 19.2
4.3 28.34–38ab
āvāhayet tato devaṃ tamasaḥ param avyayam |
ānandaṃ sarvagaṃ nityaṃ vyomātītaṃ parātparam || 28.34
marīcicakramadhyasthaṃ vāsudevam ajaṃ vibhum |
mūlamantreṇa viśveśam āvāhya gurur ātmavān || 28.35
brahmarandhreṇa bimbāntaḥ praviṣṭaṃ paricintayet |
kumbhasthaṃ paramātmānaṃ pratimāyāṃ niveśayet || 28.36
mantranyāsaṃ tataḥ kuryān mūlamantreṇa deśikaḥ |
yāceta devaṃ sāṃnidhyaṃ parivārān prakalpayet || 28.37
samārādhya tato devaṃ pāyasānnaṃ nivedayet |
4.4 28.52–61
purato mūlaberasya baddhapadmāsanasthitaḥ /
prāṇān āyamya śuddhātmā dhyāyed brahma sanātanam // 28.52
vāsudevam ajaṃ śāntam ujvalaṃ santatoditam /
anādimadhyanidhanam ekaṃ cāpy acalaṃ sthiram // 28.53
cidghanaṃ paramānandaṃ tamasaṃ param avyayam /
jnānaśaktibalaiśvaryavīryatejaḥsamanvitam // 28.54
āpādapāṇim aspṛśyam ācakṣuḥśravaṇādikam (em.; acakṣuḥśravaṇādikam Ed.) |
sarvatra karavākpādaṃ sarvato ‘kṣiśiromukham || 28.55
gatāgatavinirmuktaṃ ravikoṭisamaprabham |
caitanyaṃ sarvagaṃ nityaṃ vyomātītaṃ tadadbhutam || 28.56
citsāmānyaṃ jagat yasmin mūlamantrātmakaṃ param |
evaṃ vidhaṃ sadā viṣṇum āhlādaṃ praṇavātmakam || 28.57
taṃ niyujya mahāviṣṇuṃ mataṃ viṣṇau niveśayet |
viṣṇuṃ ca hṛdaye padme samāvāhyārcayet kramāt || 28.58
samāvāhya śriyā (corr.; sriyā Ed.) paścān mantreṇa puruṣātmanā |
añjalisthe tataḥ pīṭhe dhyāyed viśvātmanā guruḥ || 28.59
194 Czerniak-Drożdżowicz
5 Paramasaṃhitā
5.1 3.5–11ab
mūrtimān eva pūjyo’ sāv amūrter na tu pūjanam |
kāryārthaṃ mūrtayas tasya lokānugrahahetavaḥ || 3.5
ataḥ puruṣarūpeṇa kalpayitvā tam acyutam |
abhyarcya parayā bhaktyā siddhiṃ gacchanti mānavaḥ || 3.6
nirākāre tu deveśe nārcanaṃ saṃbhaven nṛṇām |
na ca dhyānaṃ na ca stotraṃ tasmāt sākāram arcayet || 3.7
Image Consecration In The Pāñcarātrika Sources 195
‘He should be worshipped with a form, there is no worship for the formless [God].
In ritual his forms are the sources of the emancipation for people.
Therefore, having fashioned Acyuta in the human form,
having worshipped with the highest devotion people approach accomplishment.
For men there is no worship of god without a form,
there is no meditation, no hymn, therefore one should worship [the god] with a form.
The worship, praise and meditation for the one possesing a form
should be done according to the rules seen in the śāstras, O God.
Being attracted by the yogins with the highest devotion, this God of gods
for their favour accomplished the four-armed form.
Therefore in such a form one should worship the God of gods,
and his worship, according to the [difference of] fruits is described in śāstras as twofold.
The first fruit is earthly good, the highest fruit is emancipation.’
5.2 18.50cd–52ab
prāṅmukhaṃ snānapīṭhasthaṃ snāpayet tam udaṅmukhaḥ /
siddhārthataṇḍulaiḥ piṣṭaiḥ pañcagavyaiḥ krameṇa ca // 18.51
ācchādya snāpayet paścāt pañcamantraiḥ samantataḥ /
‘[The teacher] facing north should bath an image facing the east and standing on the
pedestal for bathing,
[using] mustard seeds and rice, flour and five cow products in due order.
After covering [with a cloth] he should bathe [the idol] thoroughly while reciting five
mantras.’
6 Pārameśvarasaṃhitā 11.236–237
7 Śrīpraśnasaṃhitā 2.54cd-55ab
8 Viṣṇudharmottarapurāṇa 49.1–5
vajra uvāca:
rūpagandharasair hīnaḥ śabdasparśavivarjitaḥ /
puruṣas tu tvayā proktas tasya rūpam idaṃ katham // 1
mārkaṇḍeya uvāca
prakṛtir vikṛtis tasya dve rūpe paramaatmanaḥ /
alakṣyaṃ tasya tadrūpaṃ prakṛti sā prakīrtitā // 2
sākārā vikṛtir jñeyā tasya sarvaṃ jagat smṛtam /
pūjādhyānādikaṃ kartuṃ sākārasyaiva śakyate // 3
svatas tu devaḥ sākāraḥ pūjanīyo yathāvidhi /
avyaktā hi gatir duḥkhaṃ dehabhṛdbir avāpyate // 4
ato bhagavatānena svecchayā yat pradarśitam /
prādurbhāveṣv athākāraṃ tad arcanti divaukasaḥ // 5
9 Viṣṇusaṃhitā
9.1 15.1–3
atha vakṣyāmi saṃkṣepāt pratiṣṭhāpañcakaṃ kramāt /
parameṣṭhyādibhir mantraiḥ kartavyaṃ tu viśeṣataḥ // 15.1
sthāpanāsthāpanā caiva tathā saṃsthāpanā punaḥ /
prasthāpanā ca pañcoktāḥ pratiṣṭhāpanayā saha // 15.2
sthitāsīnaśayānānāṃ yānagasya calasya ca /
yā kriyā pañcadhā proktā sā pratiṣṭheti kīrtitā // 15.3
‘I will present to you shortly the fivefold installation in due order, one should perform
it especially with parameṣṭhin and others mantras [together with five mantras known
as pañcopaniṣad].
sthāpanā, āsthāpanā as well as saṃsthāpanā,
prasthāpanā—five are described together with pratisthāpanā.
For standing, sitting, lying, riding in a cart and for the movable [idol]
Five-fold ceremony is known as pratiṣṭhā.’
9.2 29.55cd–57
na ca rūpaṃ vinā devo dhyātuṃ kenāpi śakyate || 29.55
sarvarūpanivṛttā hi buddhiḥ kutrāsya tiṣṭhati |
nivṛttā glāyate buddhir nidrayā vā parīyate || 29.56
tasmād vidvān upāsīta buddhyā sākāram eva tam |
asti tasya parokṣaṃ tad iti kiñcid anusmaret || 29.57
Elisabeth Raddock
1 Introduction
At the end of the first chapter of the Hayaśīrṣa Pañcarātra, the god Brahmā
asks Viṣṇu the above questions, which are then answered in the rest of the
work. This article focuses on the first few chapters of the Hayaśīrṣa Pañcarātra,
a text of around the 8th-9th century CE that sets down the rules and rituals
that govern the construction of that most dramatic, physical representation
of religion in southern Asia, the Hindu temple. The very first chapter gives us
the story of Viṣṇu as Hayaśīrṣa and the subsequent three chapters outline the
characteristics of various employees in the temple project, notably the ācārya.1
In this article I want to explore three topics. First, is the text a śilpa śāstra or a
ritual text? Second, for whom might this text have been written? Third, which
are the various roles discussed in the text and what are important characteris-
tics of the employees in the temple project according to the text?
1 Chapter 2 of the Hayaśīrṣa Pañcarātra starts with a list of texts and then continues with a
discussion of the characteristics of the ācārya.
2 The Text
3 Historical Context
The Hayaśīrṣa Pañcarātra was composed in the early stages of what schol-
ars, including Ronald Inden (1979) and Richard Davis (1991, 1997), have called
‘Temple Hinduism,’ a period which begins with something of an explosion of
temple construction and continues for several hundred years. Following these
scholars I will use the term ‘Temple Hinduism’ for the ideological setting and
historical formation that became the dominant religious and political order
in South Asia from the 7th–8th centuries CE and the following, approximately
five hundred years.3 Temple Hinduism consisted of a large number of schools
2 Note that the dating of architectural and ritual texts is far from precise. The topic is further
explored in chapter 4.1 of my dissertation ‘Listen how the wise one begins construction of
a house for Viṣṇu: vijānatā yathārabhyaṃ gṛhaṃ vaiṣṇavaṃ śṛṇv evaṃ Chapters 1–14 of the
Hayaśīrṣa Pañcarātra.’ Here follows a summary of the arguments on the date of the Hayaśīrṣa
Pañcarātra. First, the discussion on script in the text indicates that the text was produced
shortly before the 10th century. This also agrees with the date of the Agni Purāṇa, which
borrowed extensively from the Hayaśīrṣa Pañcarātra. Thus the latter must be older than the
former, which is before the 10th century. Second, the scholarly consensus points to 800 CE.
Third, to establish a lower limit for the date I have looked at the Viṣṇudharmottara Purāṇa
as well as some ‘internal evidence’. As our text has, possibly, borrowed sections from the
Viṣṇudharmottara Purāṇa, it must be younger than that text, that is 7th century. The ideas of
vyūhas discussed by Das Gupta (1989) and the lokeśas discussed by Wessels-Mevissen (2001)
both argue for an earlier rather than later date. Thus the 8th–9th century seems to be an
acceptable working hypothesis, the reference to the scripts may suggest a date later, rather
than earlier in that period.
3 Note that the ideologies and practices that ‘Temple Hinduism’ incorporates did not vanish
with the end of this period. However, they were no longer the dominating orders in Northern
India.
200 Raddock
4 Arts may have ritual elements incorporated as well (see Thielemann 2002: 9).
Choosing an ācārya for temple construction 201
5 See Rajan 1952: 87–101, Stein 1960: 163–76, Stein 1980, Appadurai and Breckenridge 1976: 187–
211, Dirks 1987.
6 See for example Meister 1979: 204–19.
202 Raddock
4 Audience—is the Text Written for the Ācārya, the Yajamāna or the
Priest?
While architects clearly followed the rules laid down in texts, such as the
Hayaśīrṣa Pañcarātra, it is not certain that the texts were written for them, or
even by them. George Michell (1977/94: 78) states that the texts that make up the
vāstu śāstra are more likely ‘the theoretical writings of theologians, the learned
Brāhmaṇas, than manuals of architectural and artistic practice.’ The Hayaśīrṣa
Pañcarātra, which clearly focuses on ritual, and only discusses architectural
features that are of ritual significance (such as proportion, the inner sanctum
and doors) provides strong evidence for Michell’s statement. Michell (1994: 78)
continues, saying that the śilpa śāstras’ ‘true function’ is ‘as a collection of rules
that attempts to facilitate the translation of theological concepts into architec-
tural form.’ At first glance, this formulation sounds similar to Coomaraswamy
1934/94, who attempts to articulate the spiritual truths symbolized by architec-
tural forms. Michell’s formulation, however, isolates three moments: theologi-
cal concept, śāstra and architectural form. The notion of theological concept
in Michell is better grounded in the Hindu tradition than Coomaraswamy’s
formulation about eternal ideas. So too, Coomaraswamy moves directly from
architectural form to theosophical idea, while Michell stresses the interme-
diary of ritual and rule, which is central to the genre of śilpa śāstra and bet-
ter articulates the process of temple construction. Coomaraswamy (1934/94:
24–25) writes: ‘Art [in India and elsewhere, and especially hieratic art] is by
definition essentially conventional (saṃketita) . . . Conventionality [in art] has
nothing to do with calculated simplification . . . or with degeneration from rep-
resentation.’ That is, though art follows rules this does not imply that the art
is degenerate—it is a form of art different from what we think of as modern
art where breaking with tradition, innovation and individuality are features
that are highly praised. In the architectural tradition in which the Hayaśīrṣa
Pañcarātra was composed, certain rules are not optional: proportion was (and
is still) seen as an essential characteristic of beauty. In fact when temple and/
or sculpture lack the right proportions, the gods will not settle there. Thus
while the artist or architect had a fairly large amount of freedom, certain rules
needed to be followed.7 At the same time certain traditions were carried on
without texts, as part of the artists’ tradition, for example the specific styles of
various artistic periods such as the Pāla period sculpture. On the other hand
Meister (1990: 395–400) tells us that ‘[s]uch ‘scientific’ texts [śāstras] were writ-
ten as much to provide a ritual validation for the construction of temples—as
7 See chapter 10.6 of my dissertation, as well as Sinha 1996: 382–399 and Sinha 2000.
Choosing an ācārya for temple construction 203
8 The directions, which in the Hindu tradition are divine (see Wessels-Mevissen 2001).
9 See, for example, Meister 2007.
204 Raddock
10 Note that this is not a traditional category but a creative usage of the term.
11 Pāñcarātra texts should, traditionally, have four parts: jñāna (knowledge), yoga (concen-
tration), kriyā (making), and caryā (doing) (Schrader 1916/1973: 23). However, the only
text that conforms to this pattern is the Pādma Saṃhitā. Most of the later Saṃhitās deal
only with kriyā and caryā. The word kriyā means action, work, deed, etc. Kriyā, in the
Saṃhitās, covers ritualistic actions beginning with ploughing and ending with consecra-
tion (Varadachari and Tripathi 2009: 144). Thus, the Hayaśīrṣa Pañcarātra can also be
called a work dealing with (mainly) kriyā.
Choosing an ācārya for temple construction 205
12 Alberti cited in Freedberg 1989: 44–47. Compare Eco 1988: 85.
206 Raddock
To build a temple, ‘the person desiring to construct a house for the gods’
(yajamāna)14 hires certain people to perform the work on his behalf. The text
implies that the yajamāna cannot be just anyone: he needs enough money
to support the project. The Hayaśīrṣa Pañcarātra, like many other texts both
ritual and śilpa, explicitly specifies the qualities that the various workers need
to have. The Hayaśīrṣa Pañcarātra, like most Pāñcarātra works, is, however,
by no means precise in the way it uses various titles for workers it associates
with the temple construction. Various categories of people, or titles referring to
people of different positions in the construction of the temple, are discussed.
Often the differences and relationship between these categories of workers is
unclear. The following discussion is an attempt to define the key positions in
the construction process.
The Yajamāna—The yajamāna is the patron of the temple construction.
The same term is used in relation to the Vedic sacrifices. The yajamāna is the
one who wants to make a temple and who gains the merit from construction.
13 The Agni Purāṇa also tells us that the person who hoards his money and does not spend
it, is ignorant and fettered even when alive (38.22–6).
14 Hayaśīrṣa Pañcarātra, 11.2.
Choosing an ācārya for temple construction 207
The text states that the one who desires to build a temple should search for an
ācārya:
This verse is important because it indicates that the text, or at least this por-
tion of the text, is written for the person who is seeking an ācārya. Though the
Hayaśīrṣa Pañcarātra does not explicitly use the word yajāmana, here, it seems
that he is the intended referent. If one considers a large temple complex, where
construction continued for several generations, many different donors are not
unexpected. In terms of the temple proper, inscriptions normally name only
a single donor. Occasionally, however, one person is said to have constructed
the temple in honor of a second. For example, at Pattadakal, two queens of
the 8th century Chalukya king Vikramāditya II both dedicated temples in his
honor (the Virūpakṣa and the Mallikārjuna).16 Most texts, however, including
the Pādma Tantra and the Hayaśīrṣa Pañcarātra, speak of the yajamāna as an
individual. It is therefore possible that the Hayaśīrṣa Pañcarātra was written
for the yajamāna. That intent would explain the emphasis in the first few chap-
ters on the qualities and disqualifications of an ācārya, and other main agents
in the construction staff. These characteristics would be important chiefly for
the one who will hire the ācārya (and others), the yajamāna.17 Apart from hir-
ing, the yajamāna only needs to be present at a few of the rituals that mark
important stages in the temple construction. The focus on the role of the
yajamāna in some texts, such as the Pādma Saṃhitā, and (though not quite
as explicit) the Hayaśīrṣa Pañcarātra, illustrates the difference in outlook
between these texts and purely śilpa śāstra works, such as the Mayamatam
or Mānasāra, that do not pay any attention to the yajamāna. The later texts
are technical in nature, while the former focus on the devotional and ritual
15 Hari-bhakti-vilāsa (HBV) 19.93. Large portions of the Hayaśīrṣa Pañcarātra have been
quoted in the HBV see discussion in chapter 6 of my dissertation. The verses cited are the
verses in the HBV corresponding to those quoted above from the Hayaśīrṣa Pañcarātra.
16 The Virūpakṣa temple was founded in 745 CE by Queen Lōka Mahādēvī. She was the
senior Queen of Vikramāditya. The Mallikārjuna temple wad founded by Queen Trailōkya
Mahādēvī and it is called Trailōkeśvara in the Paṭṭadakal pillar-inscription (755 CE).
The queens and their inscriptions are discussed in Meister & Dhaky 1986: 78–90, plates
220–282.
17 Hayaśīrṣa Pañcarātra, 2.11. and Pādma tantra, kriya. 1.19.
208 Raddock
aspects of a temple and its construction. The yajamāna is sometimes called the
kāraka—the one who causes the architect, the kartṛ, to do the work.18
The Ācārya—The ācārya directs the building of the temple from the selec-
tion of the site to the final installation of the deities. Like the priests in the
Vedic yajña, he acts on behalf of the yajamāna throughout the construction of
the temple. As is obvious in reading the Hayaśīrṣa Pañcarātra, extra care must
be taken in selecting the man who will be the ācārya. The text spends nearly
three whole chapters discussing his qualifications and disqualifications.19 The
text’s detailed account of the desired and undesired qualities of an ācārya,
include his learning, lineage and physical condition.
18 Samarāṅganasūtradhāra, 56.303, quoted in Kramrisch (1946/2007, vol 1. p 9, note 18). See
also Pārameśvara Saṃhitā 8.178–194.
19 Chapters 2–4.
20 HBV 19.94 reads hi for tu. Compare Smith 1963, Pādmasaṃhitā, p. 7, note 21.
21 HBV 19.96.
22 The 1952 commentary answers the question as to how one can be a gṛhastha and a
Brahmacārya in the same time. If one, as a gṛhastha, avoids sleeping with ones wife on
certain days the one stays Brahmacārya—there are only two permitted days during each
moon cycle. Manu Smṛti (3.50) says ‘Regardless of the order of life in which a man lives,
if he avoids women during the forbidden nights and during the other eight nights, he
becomes a true celibate’ (Olivelle 2004: 47). Olivelle explains ‘The rule is simple, a man
who has sex with his wife only to produce offspring and not for lust should be considered
a celibate.’ Thus a couple should not have sex when the wife is infertile or unclean (i.e.
menstruating, Olivelle 2004: 243, note to 3.50). In reference to Rāmāyaṇa 1.47.18, the story
of Ahalya, Robert Goldman discusses the ṛtukāla, ‘the period of fertility during which sex
was uniquely countenanced, even mandated, in traditional India’ (Goldman 1978: 391).
Thus, if the couple only has sex during the woman’s ṛtukāla the husband will continue to
be a Brahmacārya.
23 Kacchadeśa, Kāverī, Koṅkaṇa, Kāmarūpa, Kaliṅga, Kāñchī, Kāśmīra, Kośala, explained in
verse 3.4.
24 H BV 19.99abcd.
Choosing an ācārya for temple construction 209
The brāhmaṇa of all the varṇas is the one who is learned in Pañcarātra,
set free from anger and greed, without faults and jealousy. 2.1327
He should be a worshipper of the same deity.28 He should avoid food
from śūdras. When it is not [possible to] obtain a brāhmaṇa, a kṣatriya
can [be the authority] for vaiśyas and śūdras. 2.1529
When not obtaining a kṣatriya, a vaiśya can be arranged for the śūdras.
But not at any time is a śūdra allowed to be an ācārya. 2.1630
25 That is the ruler of the country will sponsor the temple through performing his sacrifices
there, which then means that the officiating priests will get paid and that the temple will
have a steady source of patronage. This mention of royal patronage indicates a developed
stage of the cult and is an index—like the brāhminical view of caste articulated in verses
12–17—of the brāhminicization of the cult relative to its antinomian, heterodox, tantric
origins.
26 ab = HBV 19.99ef.
27 H BV 19.95. This verse makes it sound like all that is needed to be (or become?) a Brāhmin
is knowledge of the Pāñcarātra texts. However, this is contradicted by verse 15 and 16
where sūdras clearly are not in any way able to take part in the construction of the temple,
thus, they cannot become Brāhmaṇas. One may also understand the verse as saying that
Brāhmins are the only ones who have access to and receive proper training regarding
these texts and traditions.
28 Presumably this means the same deity as that of the yajamāna.
29 HBV 19.97.
30 HBV 19.98 has śūdrasya for śūdras tu and naivācāryaivam for na cācāryatvam.
Smith takes this verse to mean that a śūdra also can be an ācārya, arguing that this
would mean that the Hayaśīrṣa Pañcarātra conforms more to the earlier Pāñcarātra
view of initiation in which caste is not an important factor (Smith 1963: 6, n. 19). I do
not agree with Smith’s understanding of the text, however. It seems to me that the text
clearly says that a śūdra cannot, under any circumstances be an ācārya, not even for other
śūdras.
210 Raddock
The ācārya is learned, though the Hayaśīrṣa Pañcarātra specifies certain texts
he ought not study:
Even if he is a brāhmaṇa, one who knows words and sentences, and logi-
cal proofs, and is completely conversant with the Vedas, if he is delighted
by the paśuśāstras35 he is not an ācārya and not a teacher.36 3.1537
If one does hire a person who possesses any of the disqualifying characteristics
the temple, as well as the god enshrined inside, will not give the merit that one
hoped for.
31 gaṇḍamālī.
32 Manuscript B and C has veda and vedāṅga while the other manuscripts read veda and
aṅga.
33 nāstika—see 3.7 above.
34 sarvajña.
35 These are presumably the texts of the Pāśupatas, a Śaiva sect. Compare Helene Brunner
who notes that ‘[a] good master is naturally: śivaśāstra-samāyuktaḥ paśusāstra-
parāṇmukhaḥ (Su ch 1, 54p; almost the same line in Vāyavīyasaṃhitā quoted in KD, p 30)’
(Brunner 1992: 42). Gonda notes that the Pāśupatas were for ‘a certain period’ ‘the most
formidable rivals’ of the Pāñcarātrins (Gonda 1970: 93). These verses certainly seem to
indicate competition between Vaiṣṇava and Śaiva tantrikas.
36 deśika.
37 = HBV 19.117 reads paravāk for padavākya, like ms A (not chosen by the editor as the main
variant, though he otherwise generally uses the reading of ms A.)
Choosing an ācārya for temple construction 211
However the next chapter (chapter 4) discusses the fact that a twice-born lib-
erated person who is able to guide others across “the ocean of saṁsāra” is a
teacher, a guru even without the positive marks. The most important charac-
teristics of an ācārya is that he knows the Pañcarātra tradition.
A twice born who knows that ātman is superior everywhere, even though
he is without all these marks he is a guru. Here there is no doubt. 4.639
The foremost of twice-borns who knows this catuṣpātsaṃhitā,40 even
though he is without all these marks, he is worthy to make sacrifice. 4.7
He who is learned in the Pañcarātra [texts], who is a knower of the
truth of the established view, even without all the auspicious marks, he is
distinguished as an ācārya. 4.841
When a guide arises who is not without knowledge, O sinless one, that
guru who is a fully enlightened seeker of truth is acknowledged42 as a
guide. 4.9
For whom there is the highest devotion to Viṣṇu, [and in that manner
of devotion of Viṣṇu] so also [devotion] of the Guru. Indeed, he alone is
to be regarded as a sthāpaka. This is the truth, I tell you. 4.1043
But he who, deluded by his greed for a temple, would not perform the
consecration, without doubt will go to a frightful hell with his pupils. 4.1144
The achievement and success of each step of the construction and consecra-
tion of the temple depend on the ācārya.45 Perhaps the Hayaśīrṣa Pañcarātra
and other similar Pāñcarātra texts, such as the Pādma Saṃhitā, were com-
posed for the ācārya as well. The rituals, many of which are described in detail
(certainly much more so than in śilpa śāstra texts such as the Mayamatam),
seem to be described from the ācārya’s point of view (for example, chapter 8
discusses the search for the śalya). The use of prescriptive grammatical forms,
such as viddhi līn and gerundive, generally translated in my text as ‘he should’,
seem to in most cases refer to the ācārya. The text prescribes where he should
either perform a ritual task himself or have others do the more practical things,
such as digging holes.
According to śilpa texts (for example, the Śilparatna 1.29–4246), the ācārya
is first and foremost an architect. The Hayaśīrṣa Pañcarātra, however, tells us
that the primary qualification is that the ācārya be a Pāñcarātrin. If an initiated
Brāhmaṇa cannot be found then an initiated kṣatriya can be employed. If the
latter cannot be found, a vaiśya Pāñcarātra may be employed. A śūdra may not
be an ācārya (Hayaśīrṣa Pañcarātra, 1.2.15–16). That normative prescription
is more restrictive than the Pādma Saṃhitā (krīya pāda 1.16–17), which states
that the ācārya can be a brāhmin, kṣatriya, vaiśya or a śūdra of the anulomaka
sector.47 The Pādma Saṃhitā is more in keeping with the early Pañcarātra
view that anyone can enter/ participate in the sect. The fact that the Hayaśīrṣa
Pañcarātra excludes śūdras is typical of the development of tantric sects
that become progressively more Sanskritic and Brāhminical, conforming to
Brāhminical/twice-born Hindu norms.
The ācārya should worship the same deity as the yajamāna (2.15). He should
be free from various kinds of diseases and deformities (listed in chapter 3). The
third chapter discusses the characteristics of an ācārya that should be avoided.
Many of these are diseases that either are deadly or cause deformation of the
body (such as leprosy). The text also emphasizes the importance of avoiding
people who worship a different deity, including Śiva. The ācārya ‘should not
despise tantra’ (3.6). He should know the Vedas and vedāṅgas. He should know
the temple building techniques described in tantric texts, and he may not be a
nāstika (one who denies the Vedas, a non-believer, 3.7, 3.14, 5.1–2). Chapter four
describes the positive characteristics of the ācārya. Essentially it states that as
long as he is a twice-born man who knows the Pāñcarātra texts and the truth,
even without the auspicious marks, he is an ācārya.
The ācārya is further defined by his function. The ācārya has an executive
function, but he also carefully observes omens in order to perform propitiatory
rites. It seems likely that the ācārya represents a distinct professional group.
The sthāpaka—main architect/ artisan—works as a liaison between the
other artisans and the ācārya. Pāñcarātra works are by no means precise in the
way they designate workers associated with the ācārya in the temple build-
ing activities. For example, in śilpa śāstra texts (such as the Mayamatam and
the Mānasāra) subtleties of special tasks and hierarchical status are observed.
The rathakāra, for instance, refers to a special kind of artisan. In Pāñcarātra
texts, rathakāra may refer to the chief śilpin and is referred to as sthapati, śilpin,
rathakāra, takṣaka and sthāpaka.48 Chapter four mentions that the sthāpaka
has the highest devotion for Viṣṇu and the guru (4.10). While the Hayaśīrṣa
Pañcarātra and other Pāñcarātra texts are far from precise on this topic, other
texts are. The broader tradition has a hierarchy of professions within the
field.49 The sthapati is also commonly referred to as sūtradhāra (‘the one
who holds the strings,’ that is the one in charge of planning the layout of
the temple—the architect). The sthapati/sūtradhāra is the highest of the roles
within the hierarchy of artists.50
Other texts define the sthapati in much clearer terms. The Mānasāra tells
us that the sthapati is the guru of the sūtra-grāhin (draftsman), vardhaki
(designer), and takṣaka (carpenter). In turn, the sūtra-grāhin guides the other
two. The vardhaki, in turn, guides the takṣaka (Mānasāra 2.18–21). However,
the Mānasāra also tells us (2.31) that the sthapati has the qualifications of an
ācārya, which indicates that we should take the word ācārya perhaps more as
a ‘director’ rather than as a specific appointment.
The Mānasāra refers to the sthapati, together with the sthāpaka (the princi-
pal assistant or actual builder) as the master of the house-opening ceremonies
of a dwelling (Mānasāra 37. 7, 14–7, 58, 73–4, 83, 85). These include making
offerings to the Vāstu deities, in which the sthapati and sthāpaka both take
part.51 The Mahābhārata seems to know the sthapati as a learned person.
The sthāpati [should be] perfect in his discernment (or intellect), [and]
skilled in the science of vāstu. (Mahābhārata 1.47.14ab52)
48 For śilpa śāstra texts on the qualifications of the chief artisan and his relation to the
ācārya see Śilparatna 1.1.30–41 in Kramrisch 1946: 10 and Smith 1960: 15ff.
49 See Mishra (2009) especially chapter three which deals with the śilpin’s organization.
50 Mishra 2009: 92.
51 This is discussed more in detail in chapter 10 of my dissertation regarding the offerings to
the Vāstu deities.
52 Sthāpatir buddhisampanno vāstuvidhyāviśārada (Mahābhārata 1.47.14ab).
214 Raddock
The Hayaśīrṣa Pañcarātra lists the disqualifications for a sthāpaka, which are
many, ranging from physical deformity or decease, place of origin as well as
knowledge and ways of living. For example:
53 A Naiṣṭhikaḥ is ‘a perpetual religious student who continues with his spiritual precept
even after the prescribed period and vows lifelong absence and chastity’ (Apte).
54 Mahāpātakacidhitaḥ, the term Mahāpāta refers to a person who has committed one of
the five cardinal sins.
55 HBV 19.106.
56 Leprosy—kuṣṭī.
57 The text reads śvitrī. Apte tells us that svitram is white leprosy. According to Wikipedia,
white leprosy is the vitiliginous sort which attacks the face (http://en.wikipedia.org/wiki/
Alphito).
58 Consumption—kṣayī. On consumption see Zimmermann (1987) where he mentions con-
sumption and various terms used for this condition (consumption = rājayakṣman (p. 175),
kārśya (cachexy, p. 161), and snehakṣaya (p. 177).
59 Kacchadeśa is the name of a place in the south. Kaccha refers to a marshy area, a bank or
bordering region (Apte). Therefore Kacchadeśa seems to be a specific place, however, it
could also refer to a wasteland.
60 I am reading ungata as udgata (Agni has that reading in verse 39.6cd which is otherwise
the same as 3cd.).
61 a = HBV 19.107a.
62 Modern Assam.
63 H P 3.3d-3.4ab= Agni 39.6d–7a 1/2b.
64 A dealer in antidotes or magical things.
65 bcd= HBV 19.107bcd.
Choosing an ācārya for temple construction 215
Many of these undesired qualities are rather obvious. Of course one would not
want a violent architect or someone who hates the ācārya, because that would
make their cooperation rather difficult. Others represents exclusionary brah-
minical views of society, such as not being allowed to be a śudra, disabled,
widow’s bastard, one living on theater or very dark. Others are not entirely
clear and we may only speculate of their meaning; for instance the places men-
tioned from which the sthāpaka may not come (Kacchadeśa, Kāverī, Koṅkaṇa,
Kāmarūpa, Kaliṅga, Kāñchī, Kāśmīra, Kośala and Mahārāṣṭra). The Hayaśīrṣa
Pañcarātra was clearly not conscious of ‘India’ as modern geographers know
66 Punarbhū can refer to a ‘re-existence’ or ‘a (virgin) widow remarried.’ Punarbhū could
then possibly refer to the son of a remarried widow.
67 svayambhū is a name of Śiva in Pāñcarātra texts (MW), which could indicate the exclusion
of followers of Śiva. Svayam means self and bhū, born, thus selfborn or born from one self
i.e. not created by anyone else. Stella Kramrisch discusses svayambhū as a name for Śiva
in her book The Presence of Siva (1992: 226).
68 Nāstika—one who denies the authority of the Vedas (Gupta, 1983: 70), this includes not
only groups within Hinduism but also Buddhists and Jains. The incorporation of Vedic
elements, and stressing the importance of the vedas is another, relatively, late develop-
ment in the Pāñcarātra tradition, indicating its brahmanicization.
69 hīnāṅga—missing a limb.
70 vyasanī—unlucky.
71 Presumably physicians were excluded because they dealt with impure objects.
72 sthāpaka—see chapter 8.1 for discussion of this term.
73 It seems to me that the discussion about the disqualifications of the sthāpaka ends here
and that we return to a discussion of the ācārya—it is, however, likely that they are dis-
qualifications for both groups of people—who would want a hypocritical employee or
someone who is completely lacking skills?
216 Raddock
it. The term ‘India’ is never used here. It is clear from the text that the country
the text approves of lies somewhere in the north central/ north east of what we
today call India. In his paper ‘Epic Journeys: Travel by Land, Sea and Air in the
Literature of Ancient India,’ Robert P. Goldman has discussed the various ways
traveling is depicted in Sanskrit literature. Here Goldman indicates that place
one should not travel outside, as defined by Manu.
The area between the Himalayas (in the North) and the Vindhya moun-
tains (in the South) and lies to the east of the Sarasvatī River and west
of the confluence of the Ganges and the Yamunā River is known as
the ‘Middle Country’ (madhyadeśa). The wise know the land that lies
between the eastern and the western oceans and between the aforemen-
tioned mountain ranges as the ‘Land of the Āryans’ (āryavārta). This
land, the natural range of the black antelope, whose skin is used in vari-
ous rituals, alone is said to be as fit for sacrifice. What lies beyond it is
the land of the barbarians. People of the higher social classes (the ‘twice-
born’ dvijātayaḥ) should make strenuous efforts to live there. Those of
the servant class, under pressure of making a living, may live anywhere.74
Though the geographical range defined by the Manusmṛti and the Hayaśīrṣa
Pañcarātra are not the same, they suggest a similar idea, and delineate approx-
imately the same area. The common idea is that living and acting outside this
area is not in accordance with the divine prescriptions for yajña sacrifices
and that the inability to properly perform Vedic yajña will negatively affect
your connection with the gods. The Baudhāyana Dharmasūtra (1.2.14–16) pre-
scribes purifications for Brāhmaṇas who have traveled to certain regions, such
as Avanti and Kaliṅga.75 The Hayaśīrṣa Pañcarātra also expresses a concern
with Brāhmaṇas coming from outside this middle country, indicating that one
ought not only not travel outside, but also that one also ought not to mix with
people from other areas, at least not in a ritually significant way.
Deśika—teacher—the one who points out or instructs. He is of particular
importance in chapters five and six of the Hayaśīrṣa Pañcarātra where prepa-
ratory rituals, such as selecting, taking possession of and preparing the plot,
are described. In these chapters of the Hayaśīrṣa Pañcarātra the deśika is the
one who performs the rites.
Then the Deśika should perform the caruhoma, along with the
mūlamantra. Subsequently, the knower of mantras should give the
pūrṇāhutiṃ, ending with the vauṣaṭ exclamation. 6.19
Here the text specifically states that the deśika should perform the caruhoma
and mūla mantra. The relationship between the ācārya and the deśika is
unclear. Different terms may be used referring to the same person:
Even if he [is considered to have] a bad behavior, and [be] without auspi-
cious signs, he is nevertheless a teacher (deśika), who is a guide (tāraka)
[who crosses] over the ocean of saṃsāra. 4.576
The ācārya, deśika and a tāraka77 may be titles for different persons or, per-
haps more likely, different terms for ācārya. The terms deśika and tāraka are
standard forms of praise for a guru. The Hayaśīrṣa Pañcarātra does not provide
definite definitions of the ācārya, the sthāpaka and the deśika. Chapter four
begins by saying that we shall hear about the characteristics of the ācārya. In
verse five it uses the terms deśika and tāraka, verse eight again uses the term
ācārya, while verses six and nine use the terms guru and tāraka respectively.
Verse ten switches to the sthāpaka, though the text seems to refer to the same
ācārya.
Now I will again speak specifically of the auspicious signs of an ācārya. 4.1
A twice born who knows that ātman is superior everywhere, even
though he is without all these marks he is a guru. Here there is no doubt.
4.678
He who is learned in the Pañcarātra [texts], who is a knower of the
truth of the established view, even without all the auspicious marks, he is
distinguished as an ācārya. 4.879
76 H BV 19.101.
77 Tāraka is translated as guide and deśika as teacher in the translation of the text.
78 H BV 19.100.
79 H BV 19.102. This verse makes it clear that the text places the knowledge of the Pāñcarātra
tradition foremost.
218 Raddock
When a guide arises who is not without knowledge, O sinless one, that
guru who is a fully enlightened seeker of truth is acknowledged80 as a
guide. 4.9
For whom there is the highest devotion to Viṣṇu, [and in that manner
of devotion of Viṣṇu] so also [devotion] of the Guru. Indeed, he alone is
to be regarded as a sthāpaka. This is the truth, I tell you. 4.1081
Śilpin—The śilpin is a stone carver, the one who carves the deity (and, in all
likelihood, other decorative carving as well). He is probably of a low-caste
background.82 The artisans in general, though possibly given more respect
than most low-caste people, seem to have belonged to the śūdra class. The
Viṣṇu Smṛti 2.4–14 says that śūdras should serve the other classes and practice
art. While the term śilpin is used in the Hayaśīrṣa Pañcarātra to refer to a stone
carver, this is not commonly the usage of the term. As R.N. Mishra (2009: 2–3,
123–4) has shown, the term śilpin usually refers to a more general craftsman or
artist group. He uses the term to refer to all workers at construction of build-
ings. In Mishra’s view the term śilpin does not include the supervising function
or the architect. It is broader than the restrictive stone carver definition given
in the Hayaśīrṣa Pañcarātra, however.
While there are many other categories of workers involved in the construc-
tion of a Hindu temple, these are the categories that the Hayaśīrṣa Pañcarātra
discusses in the first kānda.83 The text does not, at this point, tell us which cat-
egory of worker is responsible for which part of the temple construction and
planning nor does it give us insight in to the interrelationship between these
categories of people. What it does tell us, as we have seen, in great detail, is the
employees’ character and physical qualifications a view of caste and persons
that is normative according to a brahmanical perspective. It does not necessar-
ily represent the view of all the artisans.
To conclude, the Hayaśīrṣa Pañcarātra is a primarily ritual text where śilpa
sāstra material is included only when it has ritual significance. The text seems
80 The term iṣyate (root √iṣ) has several meanings, such as the passive meaning acknowl-
edge or accept used here, it can also mean laid down or prescribed which is the meaning
used in verse 5.9 below.
81 H BV 19.103.
82 Agni Purāṇa 36.11 and 55.6 mentions that faults (doṣa) of the śilpin may be removed by
immersing an image in water. Doṣa here probably refers to mistakes that the śilpin might
have made.
83 One ought to refer to Mishra (2009: 86ff), especially chapter three, for more in-depth dis-
cussion on the various titles, their interdependence, organization, etc., as described both
in inscriptions and texts.
Choosing an ācārya for temple construction 219
at first to be written primarily for the yajamāna due to the many details of
the various employees’ qualifications and disqualifications. However, the rest
of the text provides details, for example, with respect to soil quality, beyond
what the yajamāna could be expected to use, details which would be essential
for the ācārya.
Bibliography
Hayaśirṣa Pañcarātra
Sānkhyathīrtha, Bhuban Mohan, with a foreword by Dines Chandra Bhattacharya and
an Introduction by Kshitis Ch. Sarkar. 1952. The Hayaśirṣa Pañcarātra, an ancient
treatise on architecture and consecration of images, vol. 1, ādikāṇḍa, paṭalaḥ 1–14.
Rajashahi: Varendra Research Society.
Shastri, Kali Kumar Dutta (ed.) 1976. Hayaśirṣapañcarātra ādi-kāṇḍa. Calcutta: The
Asiatic Society.
Mahābhārata
Mahābhārata: Critical Edition, 24 vols. 1933–1970. With Harivamsa. Critically edited
by V.S. Sukthankar et al. Poona: Bhandarkar Oriental Research Institute.
Mayamatam
Dagens, Bruno. 1994. Mayamatam. New Delhi: Indira Gandhi National Center for the
Arts.
Pādma Saṃhitā
Smith, Daniel. 1963. Pañcarātraprādādaprasādhanam, chapters 1–10 of the kriyapada
of the Pādma Saṃhitā, a Pañcarātra text on Temple Building. Madras: Published by
the author.
220 Raddock
Pārameśvarasaṃhitā
1953. Śrīpāñcarātrāntargatā Śrīpārameśvarasaṃhitā. Śrīraṅgam: Śrī Vilasam Press.
Viṣṇudharmottara Purāṇa
Shah, Priyabala. 1958/2002. Viṣṇudharmottara Purāṇa, Third Khaṇḍa, Volume 3., Delhi:
Parimal Publications.
Mukherji, Parul Dave. 2001. The Citrasūtra of the Viṣṇudharmottara Purāṇa. Delhi:
Motilal Banarasidass Publishers.
Sugita, Mizue, digitalized version, October 20, 1998, based on the edition of Shah,
Priyabala (ed.) 1958. Viṣṇudharmottarapurāṇa, Third Khaṇḍa, Vol. 1, Gaekwad’s
Oriental Series No. 130. Baroda: Oriental Institute.
Kramrisch, Stella. 1928. The Vishnudharmottara Part III: A Treatise on Indian Painting
and Image-Making. Second Revised and Enlarged Edition. Calcutta: Calcutta
University Press.
Secondary Sources
Appadurai, Arjun and Carol A. Breckenridge. 1976. ‘The South Indian Temple: Authority,
Honour and Redistribution.’ Contributions to Indian Sociology 10, 187–211.
Apte, Vaman Shivaram. 1992. Practical Sanskrit-English Dictionar. Rinsen Book Co.
Brunner, Hélène. 1992. ‘Jñāna and Kriyā: Relation between Theory and Practice in the
Śaivāgamas.’ In Ritual and Speculation in Early Tantrism, ed. Teun Goudriaan. New
York: SUNY Press, 1–60.
Coomaraswamy, Ananda K. 1934/1994. The Transformation of Nature in Art. Delhi:
South Asia Books (orig. publ. Cambridge Harvard University Press).
Das Gupta, Kalyan K. 1989. ‘The Pāñcarātra tradition and Brahmānical Iconography.’ In
Sastric Traditions in Indian Arts, ed. Anna L. Dallapiccola. Wiebaden: Steiner, 71–92.
Davis, Richard. 1991. Ritual in an Oscillating Universe. Princeton, N.J.: Princeton
University Press.
Davis, Richard. 1997. Lives of Indian Images. Princeton, N.J.: Princeton University Press.
Dirks, Nicholas B. 1987. The Hollow Crown: Ethnohistory of an Indian Kingdom,
Cambridge: Cambridge University Press.
Eco, Umberto. 1988. The Aesthetics of Thomas Aquinas, transl. Hugh Bredin. Cambridge
Mass.: Harvard University Press.
Freedberg, David. 1989. The Power of Images: Studies in the History and Theory of
Response. Chicago: Chicago University Press.
Goldman, Robert P. 1978. ‘Fathers, Sons, and Gurus: Oedipal Conflict in the Sanskrit
Epics.’ Journal of Indian Philosophy 6, 325–92.
Goldman, Robert P. 2007. ‘Epic Journeys: Travel by Land, Sea and Air in the Literature
of Ancient India.’ invited talk at a conference on travel in antiquity held at the
University of Pennsylvania in March, 2007.
Choosing an ācārya for temple construction 221
Gonda, Jan. 1970/1996. Viṣṇuism and Śivaism, Jordan lectures in comparative religion,
University of London, The Athlone Press, republished by Munshiram Manoharlal
Publishers, New Delhi.
Gupta, Sanjukta. 1983. ‘The Changing Pattern of Pāñcarātra Initiation. A Case Study in
the Reinterpretation of Ritual.’ In Selected Studies on Ritual in the Indian Religions,
ed. Ria Kloppenborg. Leiden: Brill, 69–91.
Inden, Ronald. 1979. ‘The Ceremony of the Great Gift (Mahādāna): Structure and
Historical Context of Indian Ritual and Society.’ In Asie du Sud, Traditions et
Changements. Paris: Editions du Centre national de la recherché scientifique, 131–36.
Jones, Lindsay. 2000. The Hermeneutics of Sacred Architecture: Experience, Interpretation,
Comparison. Cambridge, MA: Harvard University Press.
Kane, Panduranga V. 1968. History of Dharmaśāstra. Poona: Bhandarkar Oriental
Research Institute.
Kramrisch, Stella. 1946/2007. The Hindu Temple, Delhi: Motilal Banarasidass Publishers.
Kramrisch, Stella. 1992. The Presence of Siva. Princeton, N.J.: Princeton University Press.
Meister, Michael W. 1979. ‘Maṇḍala and Practice in Nāgara Architecture in North India.’
Journal of the American Oriental Society 99, 204–19.
Meister, Michael W. 1985. ‘Measurement and Proportion in Hindu Temple Architecture.’
Interdisciplinary Science Reviews 10/3, 248–258.
Meister, Michael W. 1990. ‘De- and Re-Constructing the Indian Temple.’ Art Journal
49/4, New Approaches to South Asian Art, 395–400.
Meister, Michael W. 2007. ‘Early Architecture and Its Transformations: New Evidence
for Vernacular Origins for the Indian Temple.’ In The Temple in South Asia, vol. 2., ed.
Adam Hardy. London: British Association for South Asian Studies, 1–19.
Meister, Michael & M.A. Dhaky. 1986. Encyclopaedia of Indian Temple Architecture.
Vol I. AIIS.
Michell, George. 1977/94. The Hindu Temple. Chicago: University of Chicago Press.
Mishra, R.N. 2009. Śilpa in Indian Tradition—Concept and Instrumentalities. Shimla:
Indian Institute of Advanced Study.
Olivelle, Patrick. 2004. The Law Code of Manu. Oxford: Oxford University Press.
Raddock, Elisabeth. 2011. Listen how the wise one begins construction of a house for
Viṣṇu: vijānatā yathārabhyaṃ gṛhaṃ vaiṣṇavaṃ śṛṇv evaṃ Chapters 1–14 of the
Hayaśīrṣa Pañcarātra, PhD dissertation, University of California, Berkeley
Rajan, K.V. Soundara. 1952. ‘The Kaleidoscopic Activities of Medieval Temples in
Tamilnad.’ Quarterly Journal of the Mythic Society 42, 87–101.
Schrader, Otto. 1916. Introduction to the Pāñcarātra and the Ahirbudhnya Saṃhitā.
Madras: The Adyar Library and Research Centre, 5–12 and 24.
Smith, Daniel. 1960. The Temple-Building Activities of the Śrī-Vaiṣṇavas in South India
according to available extant Pāñcarātrāgama texts with special reference to the
Pādma Tantra. Dissertation, Yale University, unpublished manuscript.
222 Raddock
Sinha, Ajay J. 1996. ‘Architectural Inventions in Sacred Structures: The Case of Vesara
Temples of South India.’ Journal of the Society of Architectural Historians 55/4,
382–399.
Sinha, Ajay J. 2000. Imagining Architects: Creativity in the Religious Monuments of India,
Newark, DE: University of Delaware Press.
Ślączka, Anna. 2007. Temple Consecration Rituals in Ancient India, Text and Archaeology,
Leiden: Brill.
Stein, Burton. 1960. ‘The Economic Function of a Medieval South Indian Temple.’
Journal of Asian Studies 19, 163–76.
Stein, Burton. 1980. Peasant State and Society in Medieval South India. New Delhi:
Oxford University Press.
Thapar, Romila. 2002. Early India, from the Origins to 1300. Berkeley: University of
California Press.
Thielemann, Selina. 2002. Divine Service and the Performing Arts in India. New Delhi:
A.P.H. Books.
Varadachari, V. and G.C. Tripathi. 2009. Īśvarasaṃhitā, vol 1. New Delhi: Indira Gandhi
National Centre for the Arts.
Wessels-Mevissen, Corinna. 2001. The Gods of the Directions in Ancient India: Origin
and Early Development in Art and Literature (until c. 1000 A.D.), Berlin: Dietrich
Reimer Verlag.
Zimmermann, Francis. 1987. The Jungle and the Aroma of Meats—an Ecological Theme
in Hindu Medicine Berkeley: University of California Press.
CHAPTER 9
As we know, most of the early ritual manuals, such as the early Śaiva text, the
Niśvāsatantra (Goodall 2015), deal with rituals related to ātmārthapūjā, or rit-
uals performed for one’s self. They do not generally deal with rituals related
to temples. Among the several topics discussed in these early texts are also
included details about replacing or re-installing a liṅga used in the rituals per-
formed for one’s self, in a chapter often called jīrṇoddhāra. It is natural that a
liṅga used for personal worship will require replacement, when for instance,
it is dropped, destroyed, burnt, stolen or even taken away by birds or animals.
For example, the Prāyaścittasamuccaya (119cd–124) of Trilocanaśiva, a 12th-
century Śaiva manual on expiations says:
The early Śaiva manuals, such as the Kiraṇa (Chapter 57), Sarvajñānottara
(Chapter 20), Pratiṣṭhālakṣaṇasārasamuccaya (Chapter 21), Somaśambhu
paddhati (4:10.1–27, Vol. IV, pp. 357–377), Jñānaratnāvalī (Madras R. 14898,
pp. 565–570), Prāyaścittasamuccaya (119–124) etc., also discuss the de-
installation of a damaged liṅga, with the re-installation of a new one. As we
discussed above, most of these concern the liṅga used for rituals for one’s self
and not a liṅga installed in a temple.
According to Kiraṇa (57: 13–14), if a liṅga in worship is damaged, after an
offering of pāyasa mixed with ghee in the fire one hundred times is made, the
damaged liṅga, if it is made of stone, should be dropped in water while the
Vāmadeva mantra is recited, and offered in fire if it is made of wood while
the aghora mantra is recited. Once this is done, a new liṅga should be installed.
Sarvajñānottara, in its 20th chapter provides a detailed description for re-
installing a new liṅga in the place of a liṅga damaged by fire (dagdhe) or worn
away (kṣīṇe), and the text also includes methods for re-installing a liṅga which has
been installed with the wrong measurements. Pratiṣṭālakṣaṇasārasamuccaya,
in its chapter 21 in 46 verses, discusses in detail the re-installation of liṅga in
the place of damaged ones with all the necessary ritual details including the
anujñā or obtaining permission from the deity to replace the liṅga. This chap-
ter also indicates the difficulties one may face in worshipping a damaged liṅga
and proposes the installation of a new one in its place:
pratiṣṭhāpya bhavec chuddhiḥ piṇḍikāyāṃ tathaiva ca /
hastāt tu patite liṅge jale vahati vā sthite // 121 //
tatrāpi lakṣam āvartya punaḥ saṃskāram arhati /
tryayutād agninā līḍhe yadi kāntir na hīyate // 122 //
vyaṅgitaṃ dagdhavicchāyaṃ sevitaṃ śvapacādibhiḥ /
patitaṃ ghoragartādau hṛtaṃ rājādibhir yadā // 123 //
parityajet tadā liṅgaṃ japtvā ghoraṃ daśāyutam /
liṅgāntaraṃ pratiṣṭhāpya viśuddhim adhigacchati // 124 //
‘ Re-Installation ’ of Idols Replacing Damaged Ones 225
2 Manuscripts of these short texts are available in public as well as private manuscript librar-
ies: Kalījīrṇoddhāra (Tarananallur Mana, Kerala, Ms No. 199C); Kṣetrapālajīrṇoddhāra
(Oriental Research Institute, Trivandrum, Ms. No. 4468; Kuttalakattu Mana, Kerala,
Ms No. 55); Jīrṇadhvajoddhāraprakāra (Tarananallur Mana, Kerala, Ms No. 176C);
Jīrṇāṣṭabandhoddhāravidhi (Oriental Research Institute, Trivandrum, Ms No. 25399);
Jīrṇoddhāra (Oriental Research Institute, Trivandrum Ms No. 25400); Jīrṇoddhārakālanirṇaya
(Oriental Research Institute, Trivandrum, Ms No. 25401); Jīrṇoddhārakrama (Oriental
Research Institute, Trivandrum, Ms No. 5892–93); Jīrṇoddhāravidhi (Oriental Research
Institute, Trivandrum Ms No. 25402–05).
226 Sarma
(cont.)
At this stage, I would like briefly to discuss what actually happens these days
when the priest of a Kerala temple notices any damage to the idol in worship.
The priest informs the temple management and any action with regard to the
replacement of the idol takes place only with the permission of the deity which
is obtained through an astrological process known as devapraśna or ‘question
concerning the deity’ or ‘question put to the deity.’ Devapraśna6 is conducted
by an astrologer, by making astrological calculations and ascertaining whether
to replace the damaged idol or not. If it is to be replaced, then the details for
the re-installation, such as the time for the installation, expiatory rituals to be
performed during the re-installation process, etc. will be discussed. If it is not
to be replaced, then the reasons for this will also be explained to the manage-
ment of the temple as well as to the public. These days it is quite common
to see Kerala temples undergoing a detailed and public astrological scrutiny,
particularly when unfortunate events associated with the temple take place or
when modifications of the shrines or of the rituals are required.
According to the Kerala ritual texts there are two types of jīrṇoddhāra, namely
niṣkramaṇa and saṅkoca:
When the liṅga or the idol is damaged, the niṣkramaṇa type of jīrṇoddhāra is
to be performed by replacing the damaged liṅga or idol with a new one. But
when the prāsāda or temple needs renovation, saṅkoca type of jīrṇoddhāra is
to be performed. The niṣkramaṇa again is further classified into two types, one
performed with kalaśa and the other with bālālayapratiṣṭhā.7
texts as well as in the South Indian āgama texts, such as the Kāmika, Kāraṇa, Suprabheda,
Ajita, Kumāratantra etc., it is missing from the demonstrably pre-twelfth-century āgamas.
8 The source for these verses is not known. But most of the Kerala ritual text editions include
these verses in their manuals, not as part of the text, but as additional verses for recitation,
ex. Puṭayūrbhāṣā (1988: 253–254), Tantrasamuccaya (1992: 140–141), while Viṣṇusaṃhitā
(1991: 189) includes some of the verses as part of the text.
230 Sarma
Figure 1 Obtaining the idol for installation from the craftsman (bimbaparigraha).
Photo Ajithan.
‘ Re-Installation ’ of Idols Replacing Damaged Ones 231
Then, after the necessary rituals meant for the new idol such as obtaining the
idol for installation from the craftsman (bimbaparigraha) [fig. 1], opening of
the eyes (netronmīlana ) [fig. 2], cleaning of the idol (śodhana), placing the idol
in water for a stipulated time (jalādhivāsa) [fig. 3], pouring of water from a pot
that is worshipped for purifying the idol (bimbaśuddhikalaśa) etc. have been
performed, the new idol should be kept in a bed facing the sanctum.
Next, the worshipping of the tattvakalaśa pot takes place followed by a
fire ritual named saṃhāratattvahoma. Then the playing of the pāṇi, a musi-
cal drum in the saṃhāra-krama takes place. Then the previously mentioned
tattvakalaśa pot needs to be taken inside the sanctum and installed in a
svastimaṇḍala in front of the idol. After worshipping four brahmins with gold,
garments etc. the priest should pray to them thus:
Thereupon the priest enters the sanctum and prays to all the subordinate gods
so that they be absorbed into the tattvas (meaning in the tattvakalaśa) and then
does the ablution of the tattvakalaśa upon the liṅga or idol. After the ablution
the pot used for the tattvakalaśa is kept upside down on the seat (pīṭha) and
the bundle of darbha grass (kūrca) on top of the pot, facing the south, which
is unusual in temple rituals. Then the priest sits facing the idol while the door
of the sanctum is closed and imagines that his own iḍā, piṅgalā and suṣumnā
and those of the idol are one. Then he visualizes himself seated in the heart of
the god and imagines the prāsāda and the temple complex to be two lotuses.
Then through the prescribed rituals the priest should spread the caitanya
existing in the idol, from its mūlādhāra through the suṣumnā up to the tower
(stūpa) of the temple and then further spread it up to the enclosure of the
temple (prākāra) and visualize the entire temple complex filled up completely.
He also imagines that all of the subordinate gods present there are absorbed in
this caitanya as fish caught in a net.
Next he ritually brings back (saṃhārarūpeṇa) the caitanya that had been
spread over the temple complex, first to the sanctum and then to the seat
(pīṭha) and idol and finally to the entire idol. Then that caitanya is absorbed in
the heart in the form of lotus (hṛdayakamala) of the god. Then the priest imag-
ines the God as `cinmātra’ and does the removal of the connection of channels
(nāḍīviccheda), and then opens the sanctum.
7 Jīvakalaśapūjā
Then in front of the idol a pot meant for the jīvakalaśa is installed and wor-
shipped with necessary rituals.
8 Jīvodvāsana
Next the priest, holding flowers in both hands and with the prescribed man-
tras, causes the sthūla-sūkṣma-parā forms of God to be absorbed from the
idol by the pot filled with water (kalaśa). Then he holds the conch filled with
water, following the necessary rituals, and moves it close to the feet of the idol,
towards mūlādhāra, while reciting the mantras of Kṣetrapāla. While plac-
ing it near the mūlādhāra he recites the thirty-nine tattvas, pañcatattvas and
pañcabrahmamantras in the prescribed format and moves towards the heart,
while reciting the above mantras again, and then further to the point between
the eyebrows (bhrūmadhya). There too he recites the prescribed mantras
and the caitanya is absorbed by the conch from all these three parts; then
he pours the water stored in the conch into the jīvakalaśa while reciting the
five tattvas in the order of creation. Then after the necessary rituals meant for
the worship of the jīvakalaśa, the pot is covered with silk and garlands. The
idol is covered with cloth tied with a rope made of grass. Then the priest lifts
and hold the jīvakalaśa on his head, and moves towards the area where the
adhivāsa will be performed and places it there.
234 Sarma
9 Bimboddhāra
Figure 4 The damaged idols are kept on the southern side of the temple.
Photo Ajithan.
236 Sarma
Like the late South Indian Śaiva ritual manuals, the ritual texts of Kerala
describe expiations related to temples, some of which are particularly asso-
ciated with installation rituals. For example, if the jīvakalaśa falls or breaks,
whatever water remains in the broken pot needs to be collected and used, along
with the necessary amount of purified water to fill the new pot, which will be
used and worshipped as jīvakalaśa. The texts also provide expiations for occa-
sions when the leaves used for decorating the kalaśa fall off. If the jīvakalaśa
becomes polluted then an ablution needs to be performed on a piece of gold,
on which the saptaśuddhi ritual is then performed so that the caitanya gained
by this ablution is transferred to the new jīvakalaśa. Thus we see, that the
Kerala ritual manuals also provide detailed descriptions of expiations which
give the priests solutions for almost any kind of unfavorable circumstance.
We may now see whether the ritual of de-installation can be performed for
an idol when it is not damaged. The priests of Kerala seem to have used this
ritual very cleverly to protect an ancient idol. Guruvayur, the famous Krishna
temple in Kerala, has undergone several attacks due to foreign invasions. In
1788, Haider Ali’s son and successor, Tipu Sultan, the ruler of Mysore, came to
the Malabar area where the temple is located, leading a powerful army intent
on defeating Zamorin, then king of Malabar. Guruvayur faced possible total
destruction and temple authorities were worried about the temple and espe-
cially, its precious main-idol (mūlavigraha) installed in the temple.
The priests thought of a way to protect the idol: the caitanya from the main-
idol was absorbed into the processional image and the main idol was hidden
within the temple, probably buried. The processional image invested with
the caitanya was moved to safety in the Ambalapuzha’s Sri Krishna Temple in
Travancore, nearly 150 kilometers from Guruvayur. It was rescued just in time.
Tipu’s army arrived, plundered the temple, destroyed the subsidary shrines and
set the complex on fire. In March 1792, the forces of Zamorin combined with
the British, the latest foreign power to cast its eye on the region, drove Tipu out
of Malabar and, by September 1792, the processional image was brought back
from Ambalapuzha. The main idol which was kept buried for several years was
re-installed in the sanctum and the caitanya was re-absorbed back into the
main idol, from the processional image. Even now, a special ritual is performed
every day in the Ambalapuzha temple for the deity of Guruvayur, since this
deity was worshipped there during the years when the processional image of
Guruvayur was kept there.
Conclusion
We find that the early texts of ritual literature treat only rituals related to
ātmārthapūja (doing ritual for one’s own benefit) whereas the later ritual texts,
including the texts of Kerala, elaborately discuss parārthapūjā (temple rituals).
Among the ritual texts of Kerala, while texts such as the Śaivāgamanibandha
and the Prayogamañjarī do not elaborate upon rituals pertaining to the reno-
vation of temples, those such as the Tantrasamuccaya not only describe them
in detail but also provide rituals for different kinds of renovations. It is worth
noting that, particularly in Kerala, the pāṇi or drum plays a very important role
during the installation rituals. While the ritual manuals of other regions limit
the use of Vedic mantras, we find that the Kerala texts employ Vedic mantras
238 Sarma
References
Shingo Einoo
This volume mainly deals with the pratiṣṭhā or consecration rituals of images
of deities.1 The utsarga or the dedication of some item to the public use is a
related subject with the pratiṣṭhā. The vṛṣotsarga (the rite of letting loose a
bull) and the taḍāgapratiṣṭhā (the consecration of a reservoir) belong to the
category of utsarga.2 Rites for the planting of trees and the dedication of a gar-
den are included into the utsarga, and in this article I first analyse the descrip-
tions of the planting of trees and the dedication of a garden and compare them
with the pratiṣṭhā so that different characteristics of them become clear.
The ritual texts that treat the rites of tree planting and the dedication of a
garden are of two groups: the first group describes only the merits of the tree
planting, while the other group describes ritual procedures in some detail. The
texts dealing only with merits of the ceremonies are limited in number. They
are as follows:
1 I have edited a volume dealing with the image pratiṣṭhā. See Einoo and Takashima 2005.
2 I also have published articles on the vṛṣotsarga and the consecration of a reservoir. See Einoo
2004 and Einoo 2002.
3 HirGŚS 1.7.3 [96,22] athāto vṛkṣāropaṇavidhiṃ vyākhyāsyāmaḥ, ‘And from now we will
explain the rite of planting trees.’
4 Mbh 13.99.22cd ata ūrdhvaṃ pravakṣyāmi vṛkṣāṇām api ropaṇe /22/ ‘Hereafter I will explain
(merits) of the planting of trees.’ See Padmapurāṇa 6.27.13–18.
5 This section begins as follows: Bhaviṣyapurāṇa 2.1.10.35 yas tu vṛkṣaṃ prakurute
chāyāpuṣpaphalopagam / pathi devālaye cāpi pāpāt tārayate pitṝn / kīrtiś ca mānuṣe loke
pratyabhyeti śubhaṃ phalam /35/ ‘One who plants a tree having shadows, flowers and fruits
on a road or in a temple liberates one’s ancestors from bad condition; he wins fame among
the people; he obtains good results.’
Padmapurāṇa 1.58.1–56.6
Padmapurāṇa 6.27.13–18.7
Viṣṇudharmottarapurāṇa 3.297.1–31.8
Many more texts are dedicated to the description of ritual procedures. They
are as follows:
ŚāṅkhGS 5.3.1–5.9
ĀśvGPŚ 4.10 [180,2–10].10
HirGŚS 1.7.4 [98,26–100,3].11
Agnipurāṇa 70.1–8 (see Kane 2: 896).12
Bhaviṣyapurāṇa 2.1.10.1–29.13
Bhaviṣyapurāṇa 2.1.10.30–34.14
6 Padmapurāṇa 1.58.1 śākhinām eva sarveṣāṃ phalaṃ vakṣyāmi yādṛśam / tac chṛṇudhvaṃ
mahābhāgā ropaṇe ca pṛthak pṛthak /1/ ‘I will explain which results the planting of all
kinds of trees brings one by one. Listen to that, you highly distinguished ones.’
7 Padmapurāṇa 6.27.13 athaiteṣāṃ tu vṛkṣāṇāṃ ropaṇe ca guṇāñ śṛṇu / ‘And listen to the
merits of planting various trees.’
8 Viṣṇudharmottarapurāṇa 3.297.1ab vṛkṣasaṃropaṇān martyo mahat phalam upāśnute /
‘Man obtains a great reward from the planting of trees.’
9 That it deals with a rite in an ārāma or a garden is suggested by the beginning sūtra,
ŚāṅkhGS 5.3.1 athārāme ‘gnim upasamādhāya /1/ ‘And after having kindled a fire in a
garden.’
10 The beginning sentence is athārāmeṣv apy evam, ‘And likewise even for gardens’; in the
preceding chapter an inauguration ceremony of a reservoir is described.
11 The opening verse is as follows: athāto vṛkṣodyāpanavidhānaṃ vyākhyāsyāmaḥ, ‘And from
now we will explain the rite of accomplishing the (planting of) trees’ and the chapter
ends as follows: evaṃ kṛte vidhāne ca pippalodyāpanābhidhe // samagraṃ labhate kartā
phalam āropaṇodbhavam // iti //4// ‘Thus having performed the rite called the accom-
plishing the planting of pippala trees, the performer obtains a full profit from planting
trees.’
12 The title of the chapter is atha vṛkṣādipratiṣṭhākathanam, ‘And a description of the con-
secration ritual of trees and others” and the chapter begins pratiṣṭhāṃ pādapānāṃ ca
vakṣye ‘I will explain the consecration ritual of trees.’ Agnipurāṇa 70.1–8 partly corre-
sponds to Matsyapurāṇa 59.1–20 (59.1a, 3a pādapānāṃ vidhiṃ) and Padmapurāṇa 1.28.1–
22ab (1.28.1a, 2c pādapānāṃ vidhiṃ).
13 This chapter deals with introductory rite for the sowing seeds of any plants, see
Bhaviṣyapurāṇa 2.1.10.24b and 27cd tato bījaṃ suśodhayet / . . . mantreṇa bījam āropayet
tataḥ /27/ ‘After that one should purify seeds well; . . . With a mantra one should sow seeds
then.’
14 It describes the planting of tulasī: 30a tulasyā bījam ādāya, ‘Having taken seed of tulasī.’
The Planting Of Trees And The Dedication Of A Garden 243
Bhaviṣyapurāṇa 2.3.1.1–36ab.15
Bhaviṣyapurāṇa 2.3.2.69+–74.16
Bhaviṣyapurāṇa 2.3.3.1–7 =/ Bhaviṣyapurāṇa 2.3.6.1–7.17
Bhaviṣyapurāṇa 2.3.3.8–10.18
Bhaviṣyapurāṇa 2.3.5.7–33.19
Bhaviṣyapurāṇa 2.3.7.1–5.20
Bhaviṣyapurāṇa 2.3.8.1–13.21
Bhaviṣyapurāṇa 2.3.9.1–4.22
Bhaviṣyapurāṇa 2.3.10.1–11.23
Bhaviṣyapurāṇa 2.3.14.1–6.24
Bhaviṣyapurāṇa 2.3.15.1–18.25
Bhaviṣyapurāṇa 4.128.1–45.26
15 It concerns the dedication of a garden: 1a ārāmādau viśeṣo yo vakṣyate, ‘Special rules con-
cerning a garden and others are explained.’
16 A garden is suggested in the dedication mantra given in 2.3.2.69:. . .ārāmaṃ
vanaspatidaivataṃ supūjitaṃ . . ., ‘a garden which is dedicated to Vanaspati and well
honoured’; the description has only the concluding part.
17 Both texts describe the consecration ritual of a small garden: 1ab kṣudrārāmapratiṣṭhāṃ
ca vakṣye.
18 It begins as follows: 8ab ekādivṛkṣaṃ vṛkṣāṇāṃ vidhiṃ vakṣye, ‘I explain the rite of plant-
ing one tree or many trees.’ Some beginning parts are lacking.
19 Topics of the chapter are shown as follows: 7 ārāmasya vidhiṃ vakṣye pratiṣṭhāvidhivistaram
/ hīnārāmasya ca tathā ekavṛkṣasya ca dvijāḥ /7/ ‘I will explain the rite of a garden, the rite
of its consecration ritual in detail, of a small garden and of one tree as well, o Brahmins.’
20 It begins with 1ab ekādivaravṛkṣāṇāṃ vidhiṃ vakṣye, ‘I will explain the rite of planting one
tree or many trees.’
21 The subject of this chapter is the consecration ritual of aśvattha trees as shown 1a
athāśvatthapratiṣṭhāyāṃ, ‘And in the consecration ritual of aśvattha trees’; Bhaviṣyapurāṇa
2.3.8.1–9ac corresponds to Bhaviṣyapurāṇa 2.3.4.1–9ab which describes the consecration
ritual of a reservoir of water.
22 The rite of planting a vaṭa tree is expressed as 1a vaṭasthānam atho vakṣye, ‘I will explain
further the fixing of vaṭa trees.’
23 Here the consecration ritual of a bilva tree is discussed: 1a vakṣye bilvapratiṣṭhāṃ.
24 The consecration ritual of a flower garden is the theme of this chapter: 1a
puṣpārāmapratiṣṭhāṃ tu vakṣye.
25 The consecration ritual of a tulasī plant is introduced as follows: 1ab jyeṣṭhāṣāḍhe tulasyāś
ca pratiṣṭhāṃ vidhivac caret, ‘One should perform the consecration ritual of a tulasī plant
duly in the month of jyaiṣṭhya or āṣāḍha.’
26 The subject of this chapter is shown in the beginning śloka: 4.128.1 vṛkṣāropaṇamāhātmyaṃ
vada devakinandana / udyāpanavidhiṃ caiva sarahasyaṃ samāsataḥ /1/ ‘You may teach,
o son of Devakī, the particular virtue of planting trees and the rite of accomplishing the
planting together with secret meanings thoroughly.’
244 Einoo
The word pratiṣṭhā might have primarily been used for the pratimāpratiṣṭhā
or the installation ceremony of a deity image, but secondarily it was used also
for the group of rites of the tree planting and some other similar ceremonies.
Among the passages showing the subject of each chapter, which are men-
tioned in the previous notes 3 to 26, some Purāṇas such as Agnipurāṇa 70.1a
pratiṣṭhāṃ pādapānāṃ, Bhaviṣyapurāṇa 2.3.3.1a =/ Bhaviṣyapurāṇa 2.3.6.1a
kṣudrārāmapratiṣṭhāṃ, Bhaviṣyapurāṇa 2.3.5.7b pratiṣṭhāvidhivistaram,
Bhaviṣyapurāṇa 2.3.8.1a athāśvatthapratiṣṭhāyāṃ, Bhaviṣyapurāṇa 2.3.10.1a
vakṣye bilvapratiṣṭhāṃ, Bhaviṣyapurāṇa 2.3.14.1a puṣpārāmapratiṣṭhāṃ and
Bhaviṣyapurāṇa 2.3.15.1ab tulasyāś ca pratiṣṭhāṃ, use the word pratiṣṭhā for
the items other than the deity images. It is, however, to be noted that the texts
which are seemingly older than the Purāṇas, such as ŚāṅkhGS 5.3.1–5, ĀśvGPŚ
4.10 and HirGŚS 1.7.3 and 4, do not use the word pratiṣṭhā for the rite of tree
planting and dedication of a garden. On the other hand, these older texts use
this word pratiṣṭhā and verbal forms of the root prati-ṣṭhā- in the installation
rite of a deity image.27
27 See the following examples. AgnGS 2.4.10 [71.20–72,1] atha cet pratimāṃ kurvan . . .
śaṅkhacakragadādharaṃ caturbhujaṃ kṛtvā agāre vā vimāne vā pratiṣṭhāpya, ‘And if one
will make an image (of Viṣṇu), one makes a four-armed one bearing a conch-shell, a discus
and a club and sets it up firmly in a house or a palace.’ VaikhGS 4.10 [62,13–14] tasmād gṛhe
paramaṃ viṣṇuṃ pratiṣṭhāpya sāyaṃ prātar homānte ‘rcayati, ‘Therefore, having estab-
lished in one’s house the highest Viṣṇu, one worships him at the end of a fire-oblation in
the evening and in the morning.’ VaikhGS 4.10 [62,14–15] ṣaḍaṅgulād ahīnaṃ tadrūpaṃ
kalpayitvā pūrvapakṣe puṇye nakṣatre pratiṣṭhāṃ kuryāt, ‘One should make his image of
not less than six aṅgulas height and perform the consecration in the bright half month
on the day of any auspicious nakṣatra.’ See also VaikhGS 4.11 [64,9], BodhGŚS 2.7.3, 6 and
HirGŚS 1.3.17 [35,28–30] and [36,4] where a verbal form pratiṣṭhāpayāmi appears in the
mantra used in the consecration rite. In a long description of the consecration rite of an
image ĀśvGPŚ uses the noun pratiṣṭhā and its verbal forms many times: ĀśvGPŚ 4.2 [176,7;
8; 9]; 4.4 [176,13; 24]; 4.5 [177,26]; 4.6 [177,29; 178,5]; 4.7 [178,18–19]; and 4.8 [179,3; 8–9].
But this text uses the word pratiṣṭhā also for other items than the deity image, see e.g.
ĀśvGPŚ 4.2 [175,19–21] of the sacrificial fire, 4.2 [175,25] of a maṭha and others, 4.5 [177,1] of
a kalaśa and 4.6 [177,28] of the sacrificial fire. There is an interesting example in PārGSPŚ
[404,1] athāto vāpīkūpataḍāgārāmadevatāyatanānāṃ pratiṣṭhāpanaṃ vyākhyāsyāmaḥ,
‘And from now we will explain the establishment of a reservoir of water (vāpī), a well, a
tank, a garden and a temple.’ Here the word is applied also to the inauguration ceremony
of reservoirs of water. In the Purāṇas the pratiṣṭhā are seemingly applied to a reservoir of
water, e.g. in Agnipurāṇa 64.1ab and Bhaviśyapurāṇa 2.2.17.1ab. Such an example is not
found in the older texts such as ŚāṅkhGS 5.2.1–9, KāṭhGS 71.12–13, MānGS 2.10.8, AgnGS
The Planting Of Trees And The Dedication Of A Garden 245
Among the ritual texts that treat the rites of tree planting and the dedica-
tion of a garden, the oldest one may be ŚāṅkhGS 5.3.1–5. There are two exam-
ples from the Gṛhyasūtrapariśiṣṭas. As many as sixteen passages are derived
from the Bhaviṣyapurāṇa and, again, most of the examples originate from the
Bhaviṣyapurāṇa, Book 2, Part 3. In order to know the characteristics of this part
of the Bhaviṣyapurāṇa, I show the table of contents of Bhaviṣyapurāṇa 2.3.
2.4.3 [62,1–16], AVPŚ 39, ĀśvGPA 29 [261,6–263,7], ĀśvGPŚ 4.9, BodhGŚS 4.4 and HirGŚS
1.7.1 [93,17–96,7].
28
Gocarman is here a particular measure of land defined as follows in Bhaviśyapurāṇa
2.3.2.25ab gavāṃ śataṃ vṛṣaś caiko yatra tiṣṭhaty ayantritaḥ, ‘that land which a hun-
dred cows with one bull occupy without being closely packed together’ (Kane 1974: 859,
n. 2021). For different definitions of gocarman, see Kane 1974: 859, n. 2021 and Monier
Monier-Williams, A Sanskrit-English Dictionary, s.v. gocarman.
246 Einoo
Apart from the ceremonies of the tree planting and dedication of a garden, a
variety of consecration rituals of various things such as a dam, pastureland,
a temple, a place for supplying water, a well, a reservoir of water and so on
are described here. Bhaviṣyapurāṇa 2.3.19.1–29 is wholly dedicated to the
pratiṣṭhā ceremonies of various deities. Thus we can characterize this part of
the Bhaviṣyapurāṇa as a collection of ceremonies categorized as utsarga and
pratiṣṭhā. R.C. Hazra says that the Bhaviṣyapurāṇa, Book 2, namely the mad-
hyamaparvan, is a later interpolation.
As a matter of fact, the three Parvans—Madhyama, Pratisarga and
Uttara—are comparatively late appendages. Of the three, the Madhyama
Parvan, which is not mentioned in Bhav I,2,2–3 speaking of five Parvans, viz.,
Brāhma, Vaiṣṇava, Śaiva, Tvāṣṭra and Pratisarga, is full of Tantric elements,
recognises the authority of the Tantras, and mentions the Yāmalas, Dāmaras
etc. Moreover, none of the numerous verses quoted from the ‘Bhaviṣya-p.’ or
‘Bhaviṣya’ by the comparatively early commentators and Nibandha-writers like
Bhavadeva, Jīmūtavāhana, Vijñāneśvara, Aparārka, Devaṇabhaṭṭa, Ballālasena,
Aniruddhabhaṭṭa, Hemādri, Madanapāla, Mādhavācārya and Śūlapāṇi is
found to occur in this Pravan though it is full of Smṛti materials. So, it can
hardly claim to have come from an early date (Hazra 1940: 169). And on the
basis of the fact that Raghunandana in his Smṛtitattva quoted some verses
from Bhaviṣyapurāṇa 2.3.18 Hazra concludes that the Bhaviṣyapurāṇa 2.3 can
be dated earlier than 1500 A.D. (See also Rocher 1986: 153 with n. 92).29
I have mentioned above the texts that describe only the merits of the tree
planting. This kind of information is not limited to these texts. The texts that
mainly describe the ritual procedures also sometimes say about the merits or
effects of these rituals. I collect these data and analyze them and I come to
know that these data are roughly of two groups: while the one group of verses
or passages simply says what kinds of utilities trees have or how trees are use-
ful, the other group exalts the ritual act by mentioning the merits or favorable
effects attained by performing it.
29 Both R.C. Hazra and L. Rocher characterize the Bhaviṣyapurāṇa 2 that it contains many
tantric elements. But the word tantric is seemingly misleading. It must rather be said that
this part of the Bhaviṣyapurāṇa contains ‘Smṛti materials’ as Hazra says in the quotation
above, and that ‘Smṛti materials’ mainly consist of ritual descriptions.
The Planting Of Trees And The Dedication Of A Garden 247
At first I show the utilities of trees. Mbh 13.99.30ab simply says that trees
satisfy people with flowers and fruits.30 Viṣṇudharmottarapurāṇa 3.297.20
enumerates leaves, flowers, fruits and shadows.31 Some texts mention also ben-
eficiaries. According to Mbh 13.99.28 and Padmapurāṇa 6.27.15cd–16ab trees
worship the devas with flowers, the pitṛs with fruits and the atithis with shad-
ows.32 Viṣṇudharmottarapurāṇa 3.297.13cd–14a assigns flowers to the devas,
fruits to the human beings and shadows to the atithis,33 Viṣṇusmṛti 91.5–8
teaches that he who plants trees satisfies the devas with flowers, the atithis
with fruits and travellers with shadows and add further that he satisfies the
pitṛs with water when it rains.34 Bhaviṣyapurāṇa 4.128.4–5ab enumerates many
useful items of trees: Good trees having a dense covering of leaves satisfy liv-
ing beings with shadows, barks and sprouts, the devas with flowers, the pitṛs
with fruits and the human beings (thus certainly to be supplemented) with
flowers, leaves, fruits, shadows, roots, barks and wood.35 There are texts that
mention only the pitṛs as beneficiary. According to Padmapurāṇa 1.58.6 leaves
fallen in the water are like piṇḍas for the pitṛs,36 Viṣṇudharmottarapurāṇa
3.297.15cd–16ab says that water dropped from leaves of a tree satisfies a dead
ancestor.37 Birds and insects eat fruits.38 Some texts emphasize the usefulness
of shadows.39
Thus useful parts of a tree are enumerated such as flowers, leaves, fruits,
shadows, roots, barks and wood. Even drops of water from leaves are men-
tioned. Beings who gain benefits are gods, ancestors, human beings, travellers,
birds, insects and cows. Among the beneficiaries of the products of trees the
pitṛs or the ancestors are mentioned most often and it means that the impor-
tance of the pitṛs as beneficiaries is emphasized in this context.
Now I turn to the effects of the tree planting. Effects of the tree planting
are variously mentioned in the texts describing the tree planting. They can be
divided into a number of groups. In the following I collect them according to
these groups. First, some phalas or fruits, and kāmas or wishes are said to be
obtained. HirGŚS 1.7.3 [98,5–15] mentions various trees to be planted in differ-
ent places and ends the statement by saying ‘having planted them according
to rule a wise man obtains infinite fruit.’40 A pippala tree and a vaṭa tree, when
planted in various places, grant all wishes.41 One who performs the planting of
trees obtains all wishes and infinite fruit.42
It is well understandable that the tree planting increases fame of one who
plants trees, as HirGŚS 1.7.3 [97,16] says: ‘The planting of trees such as dhātrī,
kapittha and bilva increases fame.’43 One who plants trees obtains fame not
only in this world, but also in the yonder world.44 And the tree planting not
only increases fame but also destroys evil.45 The Agnipurāṇa also confirms that
evil will perish46 and according to the Bhaviṣyapurāṇa even evil deeds com-
mitted in the past three births perish and the heavenly world will be attained.47
48 For the period in which thirty thousand Indras change, ‘indrāyutatrayam’ in Matsyapurāṇa
59.18 and Padmapurāṇa 1.28.19cd–20ab or for one billion and a thousand million kalpas
‘kalpakoṭisahasrāṇi kalpakoṭiśatāni ca’ in Padmapurāṇa 1.58.31–33ab and Bhaviṣyapurāṇa
2.1.10.55.
49 Padmapurāṇa 1.58.11b and 6.27.19a svargān na hīyate.
50 Padmapurāṇa 1.58.9d svargo bhavati cākṣayaḥ. See also Mbh 13.99.27cd, Padmapurāṇa
6.27.15ab and Viṣṇudharmottarapurāṇa 3.297.29–30.
51 In the post-Vedic ritual texts the aśvattha tree has a close relationship with Viṣṇu. The
firewood of aśvattha is used to obtain Viṣṇu’s world (AVPŚ 30.4.3c). In a rite for one desir-
ing a son aśvattha is used as firewood and Viṣṇu is worshipped (Ṛgvidhāna 3.144–146
(28.1–3), see Gonda 1980: 110). An aśvattha tree is worshipped as having Viṣṇu’s form in an
expiatory rite for averting evil at the birth of a child in the gaṇḍānta conjunction (HirGŚS
1.5.13 [63,9–10]). P.V. Kane, while describing the śrāddha at Gayā, teaches a mantra in
which Viṣṇu is worshipped as an aśvattha tree (Kane 1973: 664, n. 1502; a similar mantra is
handed down in Padmapurāṇa 1.58.16).
52 HirGŚS 1.7.3 [97,14–15] also says that a nimba tree brings cure of disease to its performer.
Nimba trees, on the other side, are regarded as ominous. According to Bhaviṣyapurāṇa
2.1.10.48a the planting of nimba brings damages to domestic animals (paśuvināśana) and
Bhaviṣyapurāṇa 2.3.15.16 counts it among trees the consecration of which is not to be
performed.
250 Einoo
of the agniṣṭoma and the pauṇḍarīka53 are gained respectively and one who
dedicates a garden (ārāma) obtains the reward of the aśvamedha.
One who plants various trees does not see hell. P.V. Kane quotes a śloka from
the Bhaviṣyapurāṇa and translates it: ‘he who plants either one aśvattha or one
picumanda or one nyagrodha or ten tamarind trees, or the three trees i.e. kapit-
tha, bilva and āmalaka or plants five mango trees would not see hell’ (Kane
1974: 895). Almost the same verse is handed down in Bhaviṣyapurāṇa 4.128.11
and a corresponding verse is found in HirGŚS 1.7.3 [96,23–24].54 Matsyapurāṇa
59.19cd and Padmapurāṇa 1.28.21ab say in the same wording that one obtains
the highest accomplishment of not being born again.55 It means that people
can obtain mukti or final liberation. HirGŚS 1.7.3 [96,28–97,1] says as follows:
Among the eighteen kinds of trees the aśvattha tree is the best, also the vaṭa
tree and picumanda tree. When in the bright fortnight of caitra month white
petals appear of the tree one plants, it may be a tree of Brahmin caste that
gives final liberation.56 Agnipurāṇa 70.1ab begins the description as follows:
pratiṣṭhāṃ pādapānāṃ ca vakṣye ‘haṃ bhuktimuktidām / ‘I will explain the
consecration of trees that gives enjoyment and liberation.’ Reward of enjoy-
ment and liberation may be mentioned in so many places in the post-Vedic
literature that it may not be necessary to give examples. HirGŚS 1.7.3 [97,12–13]
says that the vaṭa tree in all parts of which twisted hairs grow like roots gives
enjoyment and liberation like Śiva in person.57
Effects and rewards of the tree planting just discussed concern the per-
son himself who plants trees and dedicates gardens. Some texts hand down
a reward that concerns the ancestors of the person who plants trees. Mbh
53 A kind of soma sacrifice that lasts for eleven days, see for example Caland 1931: 584–585.
54 HirGŚS 1.7.3 [96,23–24] aśvattham ekaṃ picumandam ekaṃ nyagrodham ekaṃ
daśa tintiḍīś ca / kapitthabilvāmalakītrayaṃ ca pancāmravāpī narakaṃ na paśyet //
Viṣṇudharmottarapurāṇa 3.297.14b suggests this effect: nārakyaṃ nāsti pādape.
55 Matsyapurāṇa 59.19cd, Padmapurāṇa 1.28.21ab paramāṃ siddhim āpnoti
punarāvṛttidurlabhām.
56 HirGŚS 1.7.3 [96,28–97,1] teṣv aṣṭādaśabhāreṣu kunjarāśana uttamaḥ / tathāiva vaṭavṛkṣaḥ
syāt picumando ‘pi tādṛśaḥ // śuklapakṣe madhau māse yasya śukladalodbhavaḥ / dṛśyate
sa dvijātiḥ syād vaptur vai muktikārakaḥ // The Bhaviṣyapurāṇa mentions mukti as reward
of planting a very large number of trees, Bhaviṣyapurāṇa 2.1.10.38cd–39ab kāmena ropayed
viprā ekadvitriprasaṃkhyayā /38/ muktihetuḥ sahasrāṇāṃ lakṣakoṭīni yāni ca /
57 HirGŚS 1.7.3 [97,12–13] sarvāngeṣu jaṭā yasya prarohanti ca mūlavat / sa vaṭaḥ śaṃkaraḥ
sākṣād bhuktimuktiprado bhavet // As the aśvattha tree is regarded as Viṣṇu, the vaṭa
tree is considered as Śiva, for example in HGŚS 1.7.5 [101.9] āropite vaṭe nṛṇāṃ sākṣāc
chaṃkaravigrahe, Padmapurāṇa 6.115.22c rudrarūpī vaṭas and Skandapurāṇa 2.4.3.38cd
vaṭarūpī śivo yataḥ.
The Planting Of Trees And The Dedication Of A Garden 251
13.99.26 says that one who plants trees liberates his past and future paternal
ancestors.58 Matsyapurāṇa 59.19ab reads bhūtān bhavyāṃś ca manujāṃs
tārayed drumasaṃmitān, ‘(One who plants trees) liberates the past and
future peoples as many as the number of trees he plants.’59 According to
Bhaviṣyapurāṇa 2.1.10.35d one liberates his ancestors from pāpa, namely evil
or the bad situation in the yonder world. A similar idea was expressed con-
cerning the four good forms of marriage such as brāhma, prājāpatya, ārṣa and
daiva as suggested by P.V. Kane in his History of Dharmaśāstra, vol. 2, part 1, on
pp. 524f.60 He refers there to ĀśvGS 1.6.1, GautDhS 4.24–27, Manusmṛti 3.37–
38, Yājñavalkyasmṛti 1.58–60. Apart from the praise of the four forms of good
marriage, later sūtra texts hand down the same idea in the exaltation of vari-
ous rites and religious acts.61 Further, the Epic and Paurāṇic texts also teach
the purification of one’s ancestors by doing some meritorious acts.62 From the
58 Mbh 13.99.26 atītānāgate cobhe pitṛvaṃśaṃ ca bhārata / tārayed vṛkṣaropī ca tasmād
vṛkṣān praropayet /26/ Padmapurāṇa 6.27.13cd–14ab and Bhaviṣyapurāṇa 2.1.10.36 are
similar to it.
59 The corresponding Padmapurāṇa 1.28.20cd runs as follows: bhūtān bhavyāṃś ca
manujāṃs tārayed romasaṃmitān. The last word, romasaṃmitān, ‘as many as the num-
ber of hairs,’ seems to be strange, because trees have no hairs. Such an idea to express a
large number by saying as many as the number of hairs on the body of an animal appears
sometimes in the praise of donation of a certain animal such as a milk cow, for example,
in Mbh 13.70.32 dattvā dhenuṃ suvratāṃ kāṃsyadohāṃ kalyāṇavatsām apalāyinīṃ ca /
yāvanti lomāni bhavanti tasyās tāvadvarṣāṇy aśnute svargalokam /32/ and 72.42 dattvā
dhenuṃ suvratāṃ sādhuvatsāṃ kalyāṇavṛttām apalāyinīṃ ca / yāvanti lomāni bhavanti
tasyās tāvanti varṣāṇi vasaty amutra /42/ See also AVPŚ 1.50.4, Bhaviṣyapurāṇa 4.150.13,
Garuḍapurāṇa 2.4.28–29ab, Padmapurāṇa 6.64.50–51ab and Viṣṇudharmottarapurāṇa
1.147.19.
60 See also Caland 1929: xvii and Tokunaga 1997: 264 note on Bṛhaddevatā 6.110.
61 ĀśvŚS 2.12.6 (pavitreṣṭi), MānŚS 11.9.2.13 (śrāddha), BaudhŚS 28.2 [347,14–348,2] (pavitreṣṭi),
KāṭhGS 4.23 (aṣṭācatvāriṃśatsaṃmitavrata), JaimGS 2.8 [33,13–15] (anaśnatsahitākalpa),
BodhGS 3.3.29 (aṣṭācatvāriṃśatsaṃmitavrata), AVPŚ 11.2.4 (tulāpuruṣa), Ṛgvidhāna
3.25ab (3.5.2ab), ĀśvGPŚ 3.18 [174,16–17] (vṛṣotsarga), BhārPŚS 201 (pavitreṣṭi), BodhGŚS
1.24.10 (śatābhiṣeka), 3.16.12 (vṛṣotsarga), 4.4.18 (taḍāgapratiṣṭhā), HirGŚS 1.2.15 [19,21–24]
(prajākāmasyopadeśa), 1.6.28 [93.10–11] (anaśnatpārāyaṇavidhi), BaudhDhS 2.9.16.9
(prajākāmasyopadeśa), 3.9.17 (anaśnatpārāyaṇavidhi), GautDhS 9.74 (snātakadharma),
27.17 (cāndrāyaṇa).
62 Mbh 3.82.84 (stay in Gayā), 3, App. 13, 9–10 (snāna in some tīrthas), 13.83.27 (donation
of gold), Bhaviṣyapurāṇa 2.1.9.59–65ad (utsarga), 4.131.15ab (vṛṣotsarga), Kālikāpurāṇa
65.57–61ab (worship of Mahādevī), Mahābhāgavatapurāṇa 73.7–8ab (snāna in the
Ganges), Nāradapurāṇa 114.34ab (nāgapañcamī), Padmapurāṇa 6.27.51–52ab (donation
of land), Vāyupurāṇa 83.45, 48 (vṛṣotsarga).
252 Einoo
time of the later Vedic texts the care of one’s ancestors’ wellbeing in the other
world was seemingly constant concern for the people.
But the most striking effect of the tree planting is that the planted trees
become the sons of one who planted them. Mbh 13.99.30–31 says as follows:
‘Trees which are bearing flowers and fruits gladden human beings here in this
world and they liberate him who plants trees, like sons, in the other world.
Therefore one who desires felicity should always plant trees on the bank of a
pond and protect them like one’s sons; they are said to be sons from the reli-
gious point of view.’63 It is still a matter of the yonder world, as also Viṣṇusmṛti
91.4 says: ‘In the other world trees become sons of one who plants them.’64 But
Mbh 13.99.27ab and Padmapurāṇa 6.27.14cd seemingly say as a matter of this
world.65 Trees that become one’s sons do a duty as sons.66 And the expected
duty of the son is to perform the worship of the ancestors.67 It is even said
that one tree is much better than one hundred thousand sons.68 Thus trees are
expected to perform the worship of ancestors instead of the son.69 This idea is
expressed even in a mantra ‘O tree, you are chosen on the assumption as a son.
You may do the duty as the son for me.’70
When I compared the descriptions of the consecration of an image of a
deity and the rite for the planting of trees, I got an impression that the descrip-
tion of the tree planting very often exalted the usefulness of the trees them-
selves and recorded a variety of religious merits one can expect to obtain from
performing these ceremonies. I examined the pratimāpratiṣṭhā in nine texts71
and I found that the number of examples was limited. The consecration of an
image of a deity is for the sake of dharma, kāma, artha and mukti (Liṅgapurāṇa
2.47.5d). It brings performer’s growth (abhivṛddhi) (Viṣṇudharmottarapurāṇa
3.110 [372b,17]) or growth of happiness (mahābhāgyavṛddhi) (ĀśvGPŚ 4.3
[176,9]) or promotes growth (vṛddhikara), and no diseases happen for seven
births (Bhaviṣyapurāṇa 1.136.67cd–68). Performer’s protection is requested
in a mantra (Viṣṇudharmottarapurāṇa 3.110 [372b,18]). It annihilates evil
(pāpavināśahetu) (Matsyapurāṇa 266.69), removes all evil (sarvapāpahara),
gives long life (āyuṣya), and invincibility (aparājita) (Viṣṇudharmottarapurāṇa
3.116 [374b,1–2]). One obtains all wishes (sarvakāmas) and goes to Viṣṇuloka
(Viṣṇudharmottarapurāṇa 3.116 [374b,2]) or to Sūryaloka, when an image of
the Sun god is concerned (Bhaviṣyapurāṇa 1.136.72, 79–80). One who watches
the pratiṣṭhā goes to heaven (Viṣṇudharmottarapurāṇa 3.116 [374b,3–4],
Bhaviṣyapurāṇa 1.136.70). All the plants and fire used in the pratiṣṭhāvidhi
are celebrated in the Viṣṇuloka.72 The performer attains the same results as
ten aśvamedhas and one hundred vājapeyas bring (Bhaviṣyapurāṇa 1.136.71)
or effects of the Vedic rituals (kratuphala) in every month.73 He becomes
Parameśvara.74 But when the pratiṣṭhāvidhi is not performed, bad results
such as all distress, poverty, disease, infamy, blame, short life and sonless-
ness will happen (ĀśvGPŚ 4.3 [176,7–9]). This kind of difference between the
pratimāpratiṣṭhā and the tree planting or in other word this kind of peculiarity
of the tree planting ceremony originates from the difference of the nature of
70 Bhaviṣyapurāṇa 4.128.38cd tvaṃ vṛkṣa putraparikalpanayā vṛto ‘si kāryaṃ sadaiva bhavatā
mama putrakāryam /38/.
71 VaikhGS 4.10–11, BodhGŚS 2.13.1–39, ĀśvGPŚ 4.1–8 [174,30–179,10], Bṛhatsaṃhitā 59.1–22,
Agnipurāṇa 66, Bhaviṣyapurāṇa 1.134–135, Liṅgapurāṇa 2.47–48, Matsyapurāṇa 264–266
and Viṣṇudharmottarapurāṇa 3.110.
72 Viṣṇudharmottarapurāṇa 3.116 [374b,3] oṣadhyaḥ pavanaṃ vṛkṣā yat kiṃ cid upayujyate /
pratiṣṭhāyāṃ hi tat sarvaṃ viṣṇuloke mahīyate /.
73 Bhaviṣyapurāṇa 1.136.74 sthāpitvā raviṃ bhaktyā vidhidṛṣṭena karmaṇā / māse māse
kratuphalaṃ labhante nātra saṃśayaḥ /74/.
74 Liṅgapurāṇa 2.47.49–50 ya evaṃ sthāpayel liṅgaṃ sa eva parameśvaraḥ / tena devagaṇā
rudrā ṛṣayo ‘psarasas tathā /49/ sthāpitāḥ pūjitāś caiva trilokyaṃ sacarācaram /50/.
254 Einoo
the objects of the two ceremonies, namely the nature of the image of a deity
and that of trees or a garden. And we can also include a bull of the vṛṣotsarga
and a reservoir of water of the taḍāgapratiṣṭhā into the group of trees and a
garden. It means that there is a clear difference between the pratiṣṭhā and the
utsarga.
The image of a deity itself has some power or divinity, the value of which is
commonly acknowledged by many members of the society of medieval India
and even today. Some old, big trees, spreading many branches and covered
with many leaves, evoke some feeling of respect and they sometimes become
objects of worship. Bulls and big ponds are also treated with respect. But they
are secular or worldly things, they do not have originally any divine character.
But in the ordinary lives of the people these secular trees are useful in many
respects. A bull is also inevitable to maintain a herd of cows, a reservoir of
water is very important for agriculture. So it was necessary for the Indian soci-
ety to invent some religious motivations for inspiring rich people such as kings,
ministers and landowners to plant trees continually, to release a bull repeat-
edly and to dig a tank in a proper place. The religious idea thus selected for the
motivation for such meritorious acts was the ideas underlying the śrāddha, the
periodical worship of one’s own ancestors.
In the utilities of the tree I emphasized the statement that the pitṛs were
satisfied with fruits or flowers of a tree. Among the effects of the tree plant-
ing I pointed out two ideas: he who plants trees liberates his ancestors and a
planted tree becomes his son. The reason why this idea is important is that
principally only a son has a right and duty to perform the śrāddha for the sake
of his departed father and other ancestors.75 There are several statements that
explain the so-called etymologies of the word putra or son; a putra rescues
his father from hell called put or pud. P. Olivelle, in his The Āśrama System, on
p. 46 treats this problem. He refers to GB 1.1.2, Nirukta 2.11, Mbh 1.68.38, 220.14,
Rāmāyaṇa 2.115.12, Manusmṛti 9.138 and Viṣṇusmṛti 15.44. According to him,
Yāska in Nirukta 2.11 gives two possible etymologies; one is that ‘the son pro-
tects much by offering (to his ancestors),’ putraḥ puru trāyate niparaṇād vā,
and the other is that in which the name of hell, ‘put,’ appears, punnarakaṃ
tatas trāyata iti vā: ‘There is hell called put, he rescues (his father) from there.’
GB 1.1.2 [2,9–10] reads: punnāma narakam anekaśatatāraṃ tasmāt trātīti putras
tat putrasya putratvaṃ, ‘There is hell called put that has hundreds of banks. In
that the son rescues from there, the son is called so.’ Mbh 1.68.38 punnāmno
75 In the tradition of the Dharmaśāstra there were a variety of opinions about who were
entitled or obligated to perform the śrāddha when the dead person had no sons. For these
discussions, see Kane 1973: 256–261.
The Planting Of Trees And The Dedication Of A Garden 255
narakād yasmāt pitaraṃ trāyate sutaḥ / tasmāt putra iti proktaḥ svayam
eva svayaṃbhuvā // ‘In that the son rescues the father from the hell named
Put, therefore he is called the son by Brahmā himself.’ Viṣṇusmṛti 15.44 and
Manusmṛti 9,138 are very similar to it.76 Mbh 1.220.14, an example which
Olivelle mentions, is seemingly not meant for the etymological explana-
tion: punnāmno narakāt putras trātīti pitaraṃ mune / tasmād apatyasaṃtāne
yatasva dvijasattama, ‘As the son thus rescues the father from the hell named
put, therefore you should endeavour to raise offspring, o best Brahmin.’77 In
the story of Vena and Pṛthu some Purāṇas transmitted a very similar half-verse:
Vena was rescued from the hell named Put by his son, Pṛthu.78 The word putra
is explained even without referring to hell named put or pud. Vāmanapurāṇa,
Saromāhātmya, 26.31cd putraḥ sa kathyate loke yaḥ pitṝṃs trāyate bhayāt, ‘One
who rescues fathers from distress is called the son in the world.’79 Above I men-
tioned an effect of the tree planting that he who planted trees liberated his
ancestors and the same idea was thus expressed in the explanations of the
word putra. It means that the thought whether one’s ancestors are free from
misery and distress was a great concern in the Sanskrit religious texts.
The vṛṣotsarga was also brought into a close relationship with the ancestor
worship in the post-Vedic ritual texts.80 There are four Gṛhyasūtras that prescribe
the vṛṣotsarga,81 but they do not show even any hints of the ancestor worship.
As the post-Vedic texts I consulted the following texts: BodhGŚS 3.16 [321–322],
HirGŚS 1.8.1 [117,22–118,13], AVPŚ 18c, ĀśvGPA 26 [257,11–258,9], ĀśvGPŚ 3.18
76 We can add some other examples of the etymology such as BodhGŚS 2.1.15, Mbh 1.147.5,
Mbh 1.1622* (note on Mbh 1.147.5) and Harivaṃśa 66.20.
77 We can add similar examples. BodhGPbhS 1.2.5; HirGŚS 1.4.9 [44,6–7] pud iti narakasyākhyā
duḥkhaṃ ca narakaṃ viduḥ / pudi trāṇāt tataḥ putram ihecchanti paratra ca // ‘Pud is the
name of hell. They consider hell as misery. As the son means liberation in pud, they wish
the son here and hereafter.’ This is very similar to Mbh 1.845*.1–2 (note on Mbh 1.80.18)
and Harivaṃśapurāṇa 3.73.29cd–30ab.
78 See Harivaṃśa 5.24cd samutpannena kauravya satputreṇa mahātmanā / trātaḥ sa
puruṣavyāghra punnāmno narakāt tadā, Brahmāṇḍapurāṇa 1.36.151ab, Vāyupurāṇa
2.1.129cd and Viṣṇupurāṇa 1.13.42.
79 See also Garuḍapurāṇa 2.34.9bd . . . putras trātā yamālaye / tārayet pitaraṃ ghorāt tena
putraḥ pravakṣyate // and Skandapurāṇa 7.1.336.185cd putraḥ sa kathyate loke pitaraṃ
trāyate tu yaḥ /185/
80 The vṛṣotsarga was handed down in two versions, one in the śrauta version and the other
in the gṛhya and later version. For the analysis of the śrauta version and the gṛhya version,
see Einoo 2004.
81 They are ŚāṅkhGS 3.11.1–16, KauṣGS 3.6, KāṭhGS 59.1–6 and PārGS 3.9.1–10. See also ĀśvGS
4.8.36–39.
256 Einoo
82 The number of the other effects that are common to other rituals is rather limited. The
vṛṣotsarga gives wealth and grain (Matsyapurāṇa 207.15cd). The performer attains all
wishes and imperishable worlds (AVPŚ 18c.1.12) or all kinds of better states (ĀśvGPŚ 3.18
[174,23]) or infinite rewards (Devīpurāṇa 60.11). It is related with heaven (svargya) or pros-
perity of cattle (paśavya) (ĀśvGPŚ 3.18 [174,18–20]), it gives the result of the aśvamedha
(Devīpurāṇa 60.1ab, 10) or of the thousand Vedic rituals (sahasrakratu) (Devīpurāṇa
60.12–13) and also final liberation (Matsyapurāṇa 207.41, Viṣṇudharmottarapurāṇa
1.146.59).
83 ĀśvGPŚ 3.18 [174,16–17] sa yat pibati khādati lāṅgūlaṃ codakpūrṇam udasyati tena devān
ṛṣīn pitṝṃś ca prīṇāti vaṃśyāṃś cāsaptamam ubhayataḥ parāvarān uddharati. The read-
ing codakpūrṇaṃ may be changed into codapūrṇaṃ following ĀśvGPA 26 [258,7]. See
also Bhaviṣyapurāṇa 4.131.15ab vṛṣotsarge punāty eva daśātītān daśāparān.
84 ĀśvGPŚ 3.18 [174,18] pretārtham utsṛṣṭaṃ mahato narakād uttārayati tasmād ekādaśe ‘hni
pretāya vṛṣam utsṛjet.
85 BodhGŚS 3.16.7 yūthasya mukhyāś catasro vatsataryaḥ. . ./6/ avadhūnuyur jalabindūn
pītvā tṛptā yāntu pitaraḥ iti /7/ See AVPŚ 16.2.3 aśvamedhaṃ vṛṣotsargaṃ gosahasraṃ
ca yaḥ sutaḥ / dadyān madīya ity āhuḥ pitaras tarpayanti hi /3/ ‘When the son performs
the aśvamedha or the vṛṣotsarga or the gosahasravidhi, the fathers say ‘My son gives it,’
because it satisfies (them).’
86 Garuḍapurāṇa 2.41.9d, 11b preto mokṣam avāpnuyāt and Garuḍapurāṇa 2.41.13cd evaṃ
vidhiḥ samāyuktaḥ pretamokṣaṃ karoti hi.
87 Viṣṇusmṛti 86.19–20 utsṛṣṭo vṛṣabho yasmin pibaty atha jalāśaye / jalāśayaṃ tat sakalaṃ
pitṝṃs tasyopatiṣṭhati /19/ śṛṅgeṇollikhate bhūmiṃ yatra kvacana darpitaḥ / pitṛgaṇān
annapānaṃ tat prabhūtam upatiṣṭhati /20/ See also Bhaviṣyapurāṇa 4.131.15cd–18ab, 22.
The Planting Of Trees And The Dedication Of A Garden 257
Second, the vṛṣotsarga is gradually connected with the rite for the dead per-
son. The vṛṣotsarga performed for a dead on the day of his death is beneficent,88
or it is performed on the eleventh day after the funeral rite.89 According to the
context of Garuḍapurāṇa 2.41.12ab tataḥ śrāddhaṃ samuddiṣṭam ekoddiṣṭaṃ
yathāvidhi “after that the śrāddhaṃ intended especially for the dead person
is declared to be done”, the vṛṣotsarga is to be performed before the ekoddiṣṭa
śrāddha. Padmapurāṇa 1.10.20cd vṛṣotsargaṃ ca kurvīta deyā ca kapilā śubhā
/20/ ‘One should perform the vṛṣotsarga. A good brown cow is to be given’ is
given just at the end of the description of the ekoddiṣṭa śrāddha. The vṛṣotsarga
is to be performed for a man killed by snakebite after one year.90
In Gayā there is a banyan tree named akṣayavaṭa or undecaying banyan
tree, because what is given to the ancestors there does not decay.91 ‘It would
be desirable to have many sons; if even one son goes to Gayā, where there is
that banyan tree, well-known in the worlds, which makes undecaying.’92 Mbh
13.88.14 is part of gāthās sung by the pitṛs, but it is to be noticed that this verse
does not have any words that suggest the connection with the vṛṣotsarga. In
Mbh 3.85.7cd the situation is the same; the half verse is appears in the context
in which the akṣayavaṭa is praised.93 It is in Mbh 3.82.85 that a famous śloka
appears in full. Mbh 3.82.84–85 reads as follows:
Here the greatness of Gayā is not due to its akṣayavaṭa, but because that by stay-
ing there for a month seven generations of one’s ancestors are purified or liber-
ated from distress. And the merit of the aśvamedha and that of the vṛṣotsarga
are regarded as equal to the greatness of Gayā. That the setting free of a bull
or the vṛṣotsarga is included in this śloka presupposes that the effect of the
vṛṣotsarga for the benefit of the ancestors has been acknowledged widely in
the society of those days. Thus thirdly, the relationship of the vṛṣotsarga with
the ancestor worship was made clear by introducing this famous śloka to the
description of the vṛṣotsarga.94 And this śloka was also brought into the eulogy
of the śrāddha.95
The taḍāgapratiṣṭhā or the inauguration ceremony of a reservoir shows
also concern for the ancestors. Besides many effects of the performance of this
ceremony as regards the performer’s happiness in this world and in the other
auspicious Lake of Brahmā, which the Thirty Gods and the seers adore. It is for its sake,
tiger among men, that the ancient declare that one should wish for many sons, if one
may go to Gayā alone. There are also the rivers [sic] Mahānadī and Gayāśiras, prince sans
blame, and the brahmins celebrate the banyan tree Akṣayyakaraṇa, where the food given
to the ancestors becomes inexhaustible, O lord’ (Buitenen 1975: 399).
94 See Bhaviṣyapurāṇa 4.131.20, Garuḍapurāṇa 2.6.12 and Viṣṇudharmottarapurāṇa 1.146.58.
The śloka given in Devīpurāṇa 60.6cd-8ab begins with jāyeran instead of eṣṭavyāḥ.
Matsyapurāṇa 207.40 and Brahmapurāṇa 220.32 have the different third pāda: gaurīṃ
cāpy udvahet kanyāṃ or gaurīṃ vāpy udvahet kanyāṃ “and (or ‘or’) if he takes a white
girl as a wife” and the third pāda of BodhGŚS 3.16.13 and HirGŚS 1.8.1 [118,10–11] reads as
gaurīṃ vā varayet kanyāṃ, ‘or if he chooses a white girl as a wife.’ I do not find the reason
why a marriage with a white girl is meritorious.
95 One may find a good collection of passages of this śloka in Kane’s History of Dharmaśāstra,
vol. 4, p. 539 with note 1213 in the context of the vṛṣotsarga. But this collection might be
misleading, because all other passages than Matsyapurāṇa 22.6 come from the context
of the śrāddha, even if he discusses here the vṛṣotsarga. See also Kane vol. 4, pp. 652–653
with note 1477 in the description of Gayā.
The Planting Of Trees And The Dedication Of A Garden 259
world,96 some texts claim that the ancestors of the performer will be liberated
from misery and distress. Mbh 13.99.16 says, ‘When cows and virtuous men
always drink water from the reservoir dug by a man, he liberates all his family.’97
According to BodhGŚS 4.4.18 the performer purifies the twenty-one genera-
tions of his family, ten past generations, ten future generations and himself.98
Towards the end of the ceremony a cow is caused to cross over the pond.99
When the cow begins to cross over the pond, ĀśvGPA 29 [262,14–15] lays
down the recitation of a mantra which runs as follows: idaṃ salilaṃ pavitraṃ
kuruṣva śuddhāḥ pūtā amṛtāḥ santu nityam / tās tarantī sarvatīrthābhiṣiktam
lokālokaṃ tarate tīryate ca // ‘May you make this water purifying. The waters
should always be clean, purified, causing immortality. The cow crossing over
these waters cross over the Lokāloka Mountain that was sprinkled over with
waters from all holy places. And it is crossed over.’ The Lokāloka is the moun-
tain which stands at the end of the world within which the sun shines (Kirfel
1920: 121, 126). The second half of the mantra seems to express the desire to
cross the border to the yonder world safely, when the performer dies. Then, the
performer himself crosses over the pond while he touches the end of the tail of
the cow.100 Somaśambhupaddhati interprets this as follows: tatra saṃtaraṇāt
tena tīrṇā vaitaraṇī nadī, ‘By crossing over the pond by him together with
the cow the Vaitaraṇī River will be crossed over by him.’ As the Vaitaraṇī is a
river on the border to the other world, which a dead person should cross over,
known from the Mahābhārata onward, here we have an example of the con-
cern for the dead person (Einoo 2002: 713–712).
The idea that the dead father or fathers are cared and worshipped prop-
erly has been very important from the Vedic time onward and the śrāddha,
a form of the ancestor worship systematized towards the end of the period
of the Gṛhyasūtras, became a very important ritual. The śrāddha has been
References
Sanskrit Texts
Agnipurāṇa, The Agnimahāpurāṇam, Delhi: Nag Publishers, 1985.
ĀgnGS, Āgniveśyagṛhyasūtra, Edited by L.A. Ravi Varma, Trivandrum Sanskrit Series,
No. CXLIV, Trivandrum: University of Travancore, 1940.
101 Subject-Concordance given in the critical edition of the Kūrmapurāṇa on p. 791 mentions
24 text passages. According to my search the Purāṇas have more than twenty passages in
addition.
102 List of Works on Dharmaśāstra in Kane’s History of Dharmaśāstra, vol. 1, part 2, pp. 989–
1158 mentions as many as 148 titles of the Dharmanibandhas that begin with the word
śrāddha (pp. 1124–1131). As other Dharmanibandhas, for example, the Antyeṣṭipaddhati
of Nārāyaṇabhaṭṭa, treat also the śrāddha, the number of the Dharmanibandhas dealing
with the śrāddha may increase further.
The Planting Of Trees And The Dedication Of A Garden 261
Secondary Literature
Bhat, M.S. 1987. Vedic Tantrism, A Study of Ṛgvidhāna of Śaunaka with Text and
Translation. Critically Edited in the Original Sanskrit with an Introductory Study
and Translated with Critical and Exegetical Notes. Delhi: Motilal Banarsidass.
van Buitenen, J.A.B. 1975. The Mahābhārata, Translated and Edited, 2 the Book of the
Assembly Hall, 3 the Book of the Forest, Chicago and London: The University of
Chicago Press.
Caland, W. 1929. Vaikhānasasmārtasūtram, The Domestic Rules and Sacred Laws of the
Vaikhānasa School Belonging to the Black Yajurveda, Translated by Dr. W. Caland,
Calcutta: The Asiatic Society of Bengal.
Caland, W. 1931. Pañcaviṃśa-Brāhmaṇa, the Brāhmaṇa of Twenty Five Chapters,
Calcutta: The Asiatic Society of Bengal.
Bodewitz, H.W. 1999. ‘Yonder World in the Atharvaveda.’ Indo-Iranian Journal 42,
107–120.
Einoo, S. 1994. The Nāgapañcamī. Journal of the Japanese Association for the South
Asian Studies 6, 1–29.
Einoo, S. 2002. ‘Notes on the Inauguration Ceremony of a Water Reservoir.’ In Hakase
Kanreki Kinen Ronshu: East Asian Buddhism: Its Genesis and Development, ed.
Kimura Kiyotaka. Tokyo: Shunju sha, 718–703.
Einoo, S. 2004. ‘Notes on the Vṛṣotsarga.’ In The Vedas: Texts, Languages & Ritual:
Proceedings of the Third International Vedic Workshop, Leiden 2002 = Groningen
Oriental Studies, Vol. XX, eds Arlo Griffiths & Jan E.M. Houben Groningen: Egbert
Forsten, 35–48.
Einoo, S. and Takashima, J. 2005. From Material to Deity: Indian Rituals of Consecration,
New Delhi: Manohar Publishers.
Gonda, J. 1980. Vedic Ritual: The Non-solemn Rites. Leiden-Köln: E.J. Brill.
Hazra, R.C. 1940. Studies in the Purāṇic Records on Hindu Rites and Customs. Dacca.
Reprint: 1975, Delhi: Motilal Banarsidass.
Kane, P.V. 1973. History of Dharmaśāstra (Ancient and Mediaeval Religious and Civil
Law), Vol. IV, Second Edition. Poona: Bhandarkar Oriental Research Institute.
Kane, P.V. 1974. History of Dharmaśāstra (Ancient and Mediaeval Religious and Civil
Law), Vol. II, Part I and Part II, Second Edition. Poona: Bhandarkar Oriental
Research Institute.
Kirfel, W. 1920. Die Kosmographie der Inder nach den Quellen dargestellt. Bonn und
Leipzig: Kurt Schröder. Reprint: 1990, Hildesheim, Zürich, New York: Georg Olms
Verlag.
Olivelle, P. 1993. The Āśrama System: The History and Hermeneutics of a Religious
Institution. New York, Oxford: Oxford University Press.
264 Einoo
Rocher, L. 1986. The Purāṇas, a History of Indian Literature, Vol. II, Fasc. 3. Wiesbaden:
Otto Harrassowitz.
Tokunaga, M. 1997. The Bṛhaddevatā, Text Reconstructed from the Manuscripts of the
Shorter Recension with Introduction, Explanatory Notes, and Indices. Kyoto: Rinsen
Book Co.
CHAPTER 11
Ellen Gough1
For eight days in December of 2008, hundreds of Jains from both image-
worshiping sects of Jainism, Digambara and Śvetāmbara, gathered at the Jain
Center of Greater Phoenix in Arizona, U.S.A., to celebrate the installation of
twenty-six Jain temple images (mūrti, bimba, pratimā, etc.).2 During these
eight days of celebration, both Digambara and Śvetāmbara icons were installed
(pratiṣṭhita) in the temple at the Center, but only the Digambara mūrtis were
consecrated, or instilled with the energy of the Jina.3 Because the Śvetāmbara
practitioners at the Center believe that only a fully initiated monk has the power
to transform a temple image into the presence of the Jina, they had ordered the
consecration of their mūrtis—the eye-opening ceremony (añjanaśalākā)—to
be performed by a monk in Mumbai before the icons were flown to the United
States. Monks have taken a vow to only travel by foot, so they could not attend
the ceremony themselves. Many of the Digambaras at the Center, however,
are followers of a new branch of Jainism founded in the second half of the
1 Fieldwork for this essay was undertaken from January to October 2013 in Hastinapur, Jaipur,
Aligarh, Ahmedabad, Surat and Mumbai under the auspices of a Fulbright-Hays Doctoral
Dissertation Research Grant. I thank Phyllis Granoff for comments on an earlier version of
this paper.
2 For an overview of this ceremony, see ‘Phoenix Pratishtha Mahotsav: A Jain Temple Rises in
Valley of Sun,’ Jain Digest: A Publication by the Federation of Jain Association in North America
28/1 (Winter 2009), 11–12. Details not included in this article come from a telephone conver-
sation with a participant in the festival, Dr. Kirit Gosalia, on 27 September, 2012. I thank him
for his help in the research for this essay.
3 John E. Cort (2006) has discussed the theological problem of the Jain consecration ceremony,
which explicitly calls the Jina, a liberated soul completely detached from the material world,
into a Jain temple icon. While I recognize that many Jains consider Jina icons simply as sym-
bols, the mantras used in the pratiṣṭhā ceremony explicitly call the Jina into the icon, and lay-
people and mendicants repeatedly told me the icon either ‘becomes’ the Jina (bimb bhagvān
ban jātā hai) or embodies some of the energies of the Jina that remain in the world after its
liberation. Thus, for the purpose of this essay, I will assume that the pratiṣṭhā ceremony in
some way transforms an icon into the physical presence of the Jina.
twentieth century, the Kānjī Svāmī Panth, which has rejected the need for
mendicancy, advocating for a lay path to liberation.4 Because Kānjī Panthīs do
not require a mendicant to consecrate images, they could witness for them-
selves the full image consecration ceremony.
To undertake this ceremony, the Digambaras invited a ritual special-
ist (pratiṣṭhācārya) from India to perform the pañcakalyāṇaka-pratiṣṭhā in
which the icons to be consecrated are made to reenact the five (pañca) auspi-
cious events (kalyāṇaka) in the life of the Jina: conception (garbha/cyavana),
birth (janma), renunciation (dīkṣā/tapas), enlightenment/omniscience
(kevalajñāna), and death/liberation (mokṣa). Today, the pañcakalyāṇaka-pūjā
is common to the pratiṣṭhā ceremonies of Śvetāmbaras and Digambaras, but,
as we have seen, this Kānjī Svāmī version of the ceremony differs from other
Jain varieties in that the main officiant of this pūjā is not a monk, but a
layperson.
There is, however, one component of the Kānjī Svāmī pratiṣṭhā ceremony in
which these lay ritual specialists are required to imitate Digambara monks by
removing their clothes. This rite occurs during the reenactment of the fourth
kalyāṇaka, enlightenment (kevalajñāna). At this time, when the key moment
of the enlivening ceremony, the eye-opening, occurs, a sheet is held around
the mūrti that is to be consecrated, and behind the sheet, the pratiṣṭhācārya
removes his clothes before whispering a potent invocation, the sūrimantra,
into the ear of the mūrti.
Within Jain temple ritual culture, this rite is absolutely exceptional. I know
of no other ceremony in which a Jain layperson must remove all his clothes,
and it highlights a tension in the Kānjī Svāmī tradition between respecting the
ideal of full mendicancy and maintaining the belief that renunciation is not
required for liberation. Indeed, these rites draw attention to questions that
have been at the heart of Jainism since its formation: Who is a mendicant?
Who is a layperson? And what actions define these roles?5
The following essay looks at the roots of this ritualized lay nudity, examin-
ing the history and present-day uses of the Digambara sūrimantra. Of all the
components of the image consecration ceremony, why do Kānjī Panthī ritual
specialists only have to remove their clothes at the moment they recite the
sūrimantra? My hypothesis is that this practice emerged because of medieval
Digambaras’ adoption of tantric modes of image installation in which only
6 I follow the concise explanation found in Goodall 2004: xxi: ‘The central fact that character-
izes these tantric cults is that they are private cults for individuals who take a non-Vedic
initiation (dīkṣā) that uses non-Vedic (as well as Veda-derived) mantras and that is the means
to liberation.’
7 Michael Carrithers (1990: 154) recognizes this in his description of Digambara initiation
(dīkṣā) from the end of the last century: ‘Nor is anything passed on which might form a
bond, such as the mantra which is part of many Hindu ascetics’ dīkṣā.’ For the rites involved
in present-day Digambara mendicant initiation (dīkṣā) and promotion (sthāpanā), see
Syādvādamatī n.d.: 442–452.
268 Gough
occurs at the outset of the rite, when the disciple sits on a throne established
on an elaborate five-colored diagram representing the Jina’s preaching assem-
bly (samavasaraṇa).9 The guru, meditating on the ācāryamantra, performs the
ritual ablution (abhiṣeka) of the disciple (Nirvāṇakalikā, p. 8b).10
In the section on the consecration of an icon, the ācārya/sūrimantra
is employed at several points during the two main sections of the ritual:
(1) the adhivāsana rite, in which the Jina is called into the icon and the icon is
then put to sleep in a bed, and (2) the pratiṣṭhā ceremony, in which the icon
is established in the temple. The many rites involved in adhivāsana occur the
night before the final day of the image installation ceremony. The following
day, at sunrise, the sūri begins the installation (pratiṣṭhā). He ritually opens
the eyes of the icon (performs the añjanaśalākā) by reciting the ‘mantra of
the enlightened one’ (arhat-mantra) and applying a paste of jewels to the eyes
(Nirvāṇakalikā, p. 22b).11 The icon is then moved from the pavilion (maṇḍapa)
where the adhivāsana occurred and is taken to the temple. After the icon
has been established in the temple shrine, the sūri should sprinkle scented
sandalwood powder (vāsakṣepa) on the head of the icon while reciting the
ācāryamantra. The ācārya should then form his hands into the cakramudrā12
and ritually place the ācāryamantra onto the image, touching the icon (per-
forming nyāsa) and reciting the ācāryamantra three, five, or seven times
(Nirvāṇakalikā, pp. 23a–23b). The text also gives the option of reciting a differ-
ent formula, the pratiṣṭhāmantra, in place of the ācāryamantra, presumably
if a monk of a lower rank than ācārya performs the ceremony (Nirvāṇakalikā,
p. 23b).13
While the ācārya recites the sūrimantra at several points before the pratiṣṭhā,
this description of the placement (nyāsa) of the ācāryamantra on the icon
9 For some of the early textual accounts of the samavasaraṇa, in which a newly enlight-
ened tīrthaṅkara sits on a divinely-made throne, surrounded all the beings of the universe
seated in concentric circles, see Shah 1955/1998: 85–95 and Balbir 1994. For a recent dis-
cussion of the samavasaraṇa in both Digambara and Śvetāmbara art, see Hegewald 2010.
10 Nirvāṇakalikā citations refer to the Jhaverī 1926 edition.
11 The instructions for the eye-opening are translated in Dundas (2009: 4n11): ‘[A]fter utter-
ing the arhat mantra the monk ‘should open the eye of knowledge of the image with a
gold stick which has had a sweet substance put on it with a silver brush.’
12 For photos of the cakramudrā and other Śvetāmbara ritual gestures (mudrā), see
Somasundaravijaya 2007–2008: 18.
13 The pratiṣṭhāmantra provided here is identical to the vardhamānavidyā, the ritual for-
mula imparted to mendicants between the ranks of muni and ācārya. On this mantra, see
Shah 1941. See also footnote 47 below.
270 Gough
after it has been established in the temple appears to be the key moment when
the mantra is transferred to the icon, as the text at this point gives the contents
of the ācāryamantra. The Prakrit ācāryamantra the text provides begins with
the five lines of the most popular Jain mantra, the pañcanamaskāra-mantra
praising the five supreme beings (pañcaparameṣṭhī) of Jainism, and then
continues to praise other types of advanced practitioners who have achieved
certain powers (labdhi):
This set of praises to Jain ascetics with certain special powers such as clairvoy-
ance comprises the first section (pīṭha) of the different sūrimantras of compet-
ing Śvetāmbara mendicant lineages (gaccha) that began to be recorded in ritual
manuals (kalpa) by at least the thirteenth century, when Siṃhatilakasūri’s
Mantrarājarahasya was composed, and continue to be used in the promo-
tion of an ācārya and in image consecration to this day (Dundas 1998). The
Mantrarājarahasya records a number of different sūrimantras of different
mendicant lineages. They do not all contain the same number of sections, but
the majority of them contain five different sections, with the first made up of
these Prakrit praises to ascetics with superhuman powers, the second contain-
ing a Prakrit praise to the son of the first tīrthaṅkara, Bāhubali, and ‘a string of
adjectives expressing beauty and attractiveness in the feminine vocative case,’
and the third, fourth, and fifth sections comprising strings of syllables that are
not entirely translatable, even though, as we will see below in a discussion of
the fifth section, they have been interpreted to represent various Jain ideals
(Dundas 1998: 41–42).
Jains traditionally attribute the Nirvāṇakalikā to Pādaliptasūri, who is said to
have lived in the first half of the first millennium C.E., so they thus claim that
this text provides the earliest recording of the transmission of the sūrimantra
in both the promotion of an ācārya and in image consecration.15 However,
scholars have long held that the style of this work suggests a much later dating,16
and Alexis Sanderson (2011, 2011a, 2015) has recently demonstrated that key
sections of the daily rites (nityakarman), the image installation (pratiṣṭhā)
rites, and the initiation (dīkṣā) ceremony outlined in the Nirvāṇakalikā
were copied from a Śaiva manuscript of a text that dates to the first half of
the eleventh century, the Siddhāntasārapaddhati. Despite some strategic
changes such as the term ‘śiva’ being replaced with ‘jina,’ large chunks of the
Nirvāṇakalikā are word-for-word the same as the Siddhāntasārapaddhati.
Therefore, since the earliest reference to the Nirvāṇakalikā is found at the end
of the twelfth century, this text should be dated between the beginning of the
eleventh and end of the twelfth centuries.17 By this point, as noted above, we
already have the sūrimantra mentioned in Jain texts.
In addition, the Nirvāṇakalikā itself likely references earlier sources for the
sūrimantra that are no longer available. The section on bimbapratiṣṭhā in this
text quotes upwards of 70 Prakrit verses, many of which are said to come from
canonical scriptures (āgama) (Jhaverī 1926: 8a). This is the only section of the
text to do this, likely because it builds upon an earlier tradition of Jain image
consecration to which we unfortunately have little access today. While none
of these verses can be found in the 45 texts now designated as the Śvetāmbara
āgamas,18 several of the Prakrit instructions are nearly identical to those
found in the earliest known Jain outline of pratiṣṭhā, the eighth chapter of
15 On traditional understandings of Pādaliptasūri and the Nirvāṇakalikā, see Jhaverī 1926:
9b–19b.
16 See, for example, Dundas 2007: 186n86, who cites Dhaky 1994: 41n8.
17 ‘Siddhāntasārapaddhati is ascribed to the Mahārājadhirāja Bhojadeva of Dhārā, who
ruled during the first half of the eleventh century (earliest manuscript dated in 1072 CE).
The earliest dateable citation of the Nirvāṇakalikā is from 1191 AD, in Siddhasena’s com-
mentary on the Pravacanasāroddhāra of Nemicandra’ (Sanderson 2011a: 3).
18 On the āgamas, a problematic category that now generally refers to a ‘canon’ of 45 scrip-
tures, but has designated different groupings of texts throughout history, see Folkert 1993:
53–112.
272 Gough
19 The date of the Pañcāśakaprakaraṇa is disputed. Robert Williams (1965: 104) places
it in the sixth century. However, the second chapter of this text, on initiation (dīkṣā),
prescribes the initiand to be blindfolded and to throw a flower into a ritual diagram in
a similar manner to Buddhist and Hindu tantric initiations whose descriptions do not
occur as early as the sixth century (see Shinohara 2014: 205–225 on the earliest known
flower-throwing rite of this type in Buddhist sources, Atikūṭa’s Dhāraṇīsaṃgraha of 654
CE, and the introduction to Array 1988 on the dating of the presumed earliest Śaiva source
of this rite, the Svaccandatantra, to the eighth century). Unless this is the earliest known
example of this flower-throwing rite, this text, or at least this portion of the text, does not
belong to the sixth century, and could belong to the Haribhadrasūri who has been dated
to the eighth century. Paul Dundas (2002: 23) has also questioned the early dating of the
Pañcāśakaprakaraṇa.
20 Jhaverī 1926: 26bn2,3 notes that Pañcāśakaprakaraṇa 8.36 and 8.41 are quoted in
Nirvāṇakalikā, p. 26a. See also Pañcāśakaprakaraṇa 8.48 and 8.49 quoted in Nirvāṇakalikā,
pp. 26b and 27a.
21 tathācāgamaḥ | sadasanavadhavalavattheṇa chāiuṃ vāsapupphadhūpeṇaṃ | ahivāsijja
tinni vārāo sūriṇo sūrimanteṇa || (Nirvāṇakalikā, p. 21a). ‘Having covered [the image] with
a new white cloth with a fringe, the mendicant head (sūri), with perfume, flowers, and
incense, should put the deity to sleep by [reciting] the sūrimantra three times.’
22 Textual and inscriptional evidence suggests that Jain ācāryas were consecrating temple
icons as early as the tenth century. Two inscriptions in the town of Osiyā, Rajasthan, for
example, record ācāryas’ consecrations of temples in 954 C.E. and 976 C.E., respectively
(Nāhar 1918, inscriptions 134 and 792). In addition, Śāntisūri suggests that sūris had been
in charge of image consecration ceremonies for some time before the eleventh century
sūrimantra and tantricization of Jain image consecration 273
2 Tantricization
This ritual use of the ācārya’s mantra in pratiṣṭhā is one of the many examples
of the ‘tantricization’ Jainism underwent in the medieval period. While early
Jain texts recognize that the ācārya initiates disciples, they do not mention
that he should consecrate temple images (Deo 1956: 145–146). The idea that the
person who leads the pratiṣṭhā ceremony should have undergone the highest
level of initiation, becoming an ācārya, seems to have developed within tantric
traditions. By the eleventh century,23 this understanding was commonplace in
tantric sects. H. Daniel Smith’s English summary of the section on pratiṣṭhā in
the Jayākhyasaṃhitā, a later text from the tantric Vaiṣṇava tradition Pāñcarātra,
describes how the ācārya leads the ceremony (Smith 1984: 55–63). Yael Bentor’s
study of a variety of Indo-Tibetan tantric Buddhist texts similarly confirms
that a vajrācārya should be the main ritual actor in tantric Buddhist pratiṣṭhā
ceremonies (Bentor 1996: 50).24
Later Śaiva āgamas also agree that the performance of pratiṣṭhā is the
ācārya’s responsibility. Hélène Brunner has noted that in several āgamas,25
‘at the end of the chapter on ācāryābhiṣeka’ the guru ‘enumerates the duties, or
privileges,’ of an ācārya to his promoted disciple:
when he mentions in the Ceiyavaṃdaṇamahābhāsa that ‘there are many icons conse-
crated by ācāryas’ (bahuvihabimbāiṃ sūrīhiṃ paịṭṭiyāiṃ), trans. in Cort 2010: 291–292n35.
23 Jun Takashima has noted that the requirement that the ācārya perform pratiṣṭhā seems
to be a later development in the Saiddhāntika sources. In earlier texts, ‘pratiṣṭhā was in
fact a concern of the sādhaka,’ the initiate devoted to the acquisition of special powers
(siddhi) who is ranked below that of the ācārya (Takashima 2005: 137). This seems to be
the case for Vaiṣṇavas, as well. While in the earlier Pāñcarātra text the Jayākhyasaṃhitā,
‘the sādhaka leads the ceremony under the guidance of the preceptor,’ the later
Kapiñjalasaṃhitā has the ācārya head the rite (Hikita 2005: 187–188). This shift from
sādhaka to ācārya as the main officiant could align with tantric cults’ increasing emphasis
on public consecration of large, state-supported temples over the personal shrines of the
sādhaka. Jains may have been influenced by this development.
24 See also Mori 2005: 201–228.
25 Brunner (1990: 23n27) cites Somaśambhupaddhati, part 3, pp. 486 and 496, and Mṛgen
drāgama Kriyāpāda 8, 214b–216.
274 Gough
26 To my knowledge, this is the only Jain text on ācāryapada-sthāpanā to have the guru
make this announcement, suggesting that it was not part of the Jain ceremony and may
have been copied from the Śaiva text.
27 See the editions by Jinendravijaya 1981 and in Bhattacharyya 2010: 297–325.
28 See, for example, Shah 1987 and Tiwari et al. 2010.
29 Cort (2006: 82n6) notes that he consulted the Nirvāṇakalikā, along with the Pratiṣthākalpa
and a twentieth-century compilation, in his study of present-day Śvetāmbara image con-
secration. Dundas 2009 more understandably references the Nirvāṇakalikā, as his study
focuses on a contemporaneous Śvetāmbara text related to image consecration.
30 Sādhvī Saumyaguṇā, interviewed by author, Jaipur, May 7, 2013.
sūrimantra and tantricization of Jain image consecration 275
suggest that the Nirvāṇakalikā played a marginal role in the development of Jain
mendicant promotion and image consecration. Instead, texts by the twelfth-
century Ācārya Candrasūri, the fourteenth-century Ācārya Jinaprabhasūri, and
the fifteenth-century Ācārya Vardhamānasūri should be seen as more founda-
tional sources. Unlike the Nirvāṇakalikā, all these texts model the central act of
image consecration on the promotion of a mendicant leader.
We can first support this claim by looking at present-day ācāryapada-
sthāpanā rites. Today, different Śvetāmbara lineages follow different texts.
Members of the Kharatara Gaccha, most popular in Rajasthan, follow the
instructions given in the ritual manual Vidhimārgaprapā composed by
Jinaprabhasūri in 1306. Members of the five other extant Śvetāmbara mendi-
cant lineages,31 today more populous in Gujarat, follow the instructions given
in various bṛhadyogavidhi texts, manuals on ascetic rites composed by the
leaders of the different groups (samudāya) within each lineage.32
When I asked monks and nuns to name some earlier texts upon which
these modern bṛhadyogavidhi manuals are based, two sources mentioned were
Jinaprabhasūri’s Vidhimārgaprapā and the ritual manual Ācāradinakara, com-
posed in 1411–12 by the ācārya Vardhamānasūri. Manuscript collections suggest
that these texts were more widely used than the Nirvāṇakalikā in the pre-
modern period. For example, H.D. Velankar’s register of Jain manuscripts from
collections throughout India lists 20 different manuscripts of the Ācāradinkara,
29 of the Vidhimārgaprapā, and only three of the entirety of the Nirvāṇakalikā,
with seven other manuscripts devoted to that text’s section on pratiṣṭhā
(Velankar 1944: 261, 22, 356, 214).
In addition, in the accounts of ācāryapada-sthāpanā in the Ācāradinakara,
Vidhimārgaprapā, and bṛhadyogavidhi texts,33 the central act of ordination—
when the sūrimantra is transferred from guru to disciple—differs considerably
from the account in the Nirvāṇakalikā. In the texts in use today and those that
influenced them, unlike in the Nirvāṇakalikā, the moment when the sūrimantra
is imparted to the disciple does not occur during a ritual ablution (abhiṣeka).
The Nirvāṇakalikā’s lengthy description of the ablution of the disciple is not
31 Peter Flügel (2006: 317) has outlined the six extant Śvetāmbara image-worshiping
gacchas, with their tradition dates of founding in parentheses: ‘(1) the Kharatara Gaccha
(1023), (2) the A(ñ)cala Gaccha or Vidhi Pakṣa (1156), (3) the Āgamika or Tristuti Gaccha
(1193) and (4) the Tapā Gaccha (1228), from which (5) the Vimala Gaccha (1495), and
(6) the Pārśvacandra Gaccha (1515) separated.’
32 For a list of the different samudāya, see Flügel 2006, table 12.1.
33 I have access to two different recensions from the Tapā Gaccha: (1) Devendrasāgara
1943, of Ānandasāgarasūri’s samudāya, Sāgara Śākhā, and (2) Khāntivijayagaṇi 1926, of
Mohanalālasūri’s samudāya, Vijaya Śākhā.
276 Gough
Figure 1 Ācārya Premasūri recites the sūrimantra after decorating the right ear of his
disciple, the soon-to-be ācārya Hemacandrasūri, with sandalwood powder.
7 May, 2006.
Photo taken from a calendar in Hemacandrasūri’s collection.
Devendrasāgara 1943. In it, the right ears of the monks to be promoted were decorated
with sandalwood powder, oil, and a silver decoration (bādlā).
38 This is confirmed by Cort 2006: 82n6.
39 Sakalacandragaṇi also names the manuals of Haribhadrasūri, Hemācārya, Śyāmācarya,
and Guṇaratnākarasūri. See Kalyāṇavijayagaṇi 1956: 11b.
40 Jinaprabhasūrisūri, in Vidhimārgaprapā, p. 111, recognizes that he has taken a large chunk
of his pratiṣṭhāvidhi verbatim from Candrasūri.
41 Vardhamānasūri claims that he also drew upon the earlier texts of Āryanandi,
Kṣapakacandranandi, Indranandi, and Vajrasvāmī (Ācāradinakara, 150). This ‘Indranandi’
likely refers to the tenth-century Digambara tantric, whose Pratiṣṭhāpāṭha has been pre-
served in manuscript form (Velankar 1944: 261), but I do not have access to it. On the
Digambara influence on the Ācāradinakara, see Jain 2007.
42 Jinaprabhasūri does reference Pādaliptasūri in his outline of ācāryapada-sthāpanā,
but only in order to disagree with one of his prescriptions. In Vidhimārgaprapā, p. 67,
line 27, he notes that upon the disciple’s promotion to the rank of ācārya, ‘according to
Pādaliptasūri,’ the guru should give the disciple a cloth band used for meditative postures
(jogapaṭṭaya) and a khaḍiya. Jhaverī (1926: 5b) understands ‘khaḍiya’ to mean a ‘pen,’
while Sādhvī Saumyaguṇā (2005: 197) understands it to mean ‘slippers.’ This confusion
over what, exactly, this term khaḍiya means highlights how no other Jain text prescribes
this gifting, and these items are certainly not given to new ācāryas today. Jinaprabhasūri’s
mentioning of Pādaliptasūri is also a bit odd, since Jinaprabhasūri does not attribute
any other part of the ceremony to other authors of ritual manuals. As no other Jain text
prescribes this gifting, Jinaprabhasūri likely attributed this prescription to Pādaliptasūri
because he did not want to claim it himself, finding it out of place with Jain praxis.
278 Gough
In the Nirvāṇakalikā, the many components of the adhivāsana rite occur the
night before the pratiṣṭhā ceremony. The eye-opening rite is then performed
the next day, at the outset of the pratiṣṭhāvidhi, when the icon still remains
in the pavilion (maṇḍapa) constructed for the adhivāsana rites. After the icon
has been taken from the adhivāsana-maṇḍapa and established in the temple
shrine, the ācārya then sprinkles scented sandalwood powder on the icon and
performs nyāsa of the ācāryamantra. Other texts, such as Jinaprabhasūri’s,
which is modeled on Candrasūri’s, agree on this sequence of rites.45
Contemporary Śvetāmbara consecrations, however, following Sakalacan
dragaṇi, have mapped the life story of the Jina onto existing consecration
rites and have thus had to rearrange components of the ceremony to make
sense in terms of the life of the Jina. Therefore, modern consecrations have
ācāryas perform the adhivāsana, añjanaśalākā, and nyāsa of the sūrimantra
back-to-back in the middle of the night before the pratiṣṭhā as part of the
reenactment of the Jina’s enlightenment, or obtainment of omniscience
(jñāna-kalyāṇaka). As one scholar-monk described to me, the purpose of
the adhivāsana is to use mantras to instill in the icons the energies of the
tīrthaṅkaras that remain in the universe even when their souls are liberated
and inaccessible. The eye-opening and nyāsa of the sūrimantra, on the other
hand, fill the icon with the power of omniscience. Ideally, an ācārya should
undertake these rites in solitude, but if help is needed, a ritual specialist can
be present, and if an ācārya is not available, a mendicant of a lower rank can
stand in for an ācārya. Few mendicants of lower ranks perform this rite today,
however, and many specialists believe that the consecration is only effective if
performed by an ācārya in complete solitude. John Cort’s simplified account
of the adhivāsana, eye-opening and sūrimantra nyāsa is worth quoting
at length:
45 A close comparison of the differences in the adhivāsana rites of each of these texts will
have to be left to a later study. Nirvāṇakalikā, p. 22b, is the only text to require placing the
image in a bed overnight.
280 Gough
right ear of the image, and then rubs a small amount of paste made from
sugar,46 camphor and saffron into the ear (Cort 2006: 73–74).
When an ācārya conducts the ceremony, the ‘special mantra’ here refers to the
sūrimantra, making this portion of the rite a clear model of the promotion of
an ācārya.47 After reciting the sūrimantra into the ear of the icon, the ācārya
then forms his hands to make the cakramudrā and touches the different limbs
of each of the new icons while reciting the pratiṣṭhāmantra.48
The present-day nyāsa of the sūrimantra therefore differs from the account
in the Nirvāṇakalikā, which does not mention the decoration of the ear, and
only mentions the performance of the cakramudrā and nyāsa of the icon.
For the Nirvāṇakalikā, neither the ordination of an ācārya nor the conse-
cration of an image involves the recitation of the mantra into the ear of the
object of promotion. Candrasūri (Subodhāsamācārī, p. 17b), Jinaprabhasūri
(Vidhimārgaprapā, p. 102, line 16), and Vardhamānasūri,49 on the other hand,
all instruct the ācārya to transmit the ācāryamantra to the icon by decorating
its right ear and pronouncing the mantra. Thus, because the Nirvāṇakalikā’s
description of the promotion of an ācārya does not align with any other
Jain description, and its outline of the consecration of a temple image also
does not seem to have been very influential, with resepct to these two cer-
emonies, we can understand this text as a largely insignificant, if interesting,
medieval document.
4 Śvetāmbara Controversies
tantric ritual adepts, but instead mapped tantric promotions onto existing
mendicant ordinations and promotions. Thus, to become a Śaiva, Buddhist, or
Vaiṣṇava ācārya who can consecrate images, one does not have to be a men-
dicant; one only has to receive a tantric promotion in which one worships a
maṇḍala and receives an esoteric mantra. Medieval Jains, on the other hand,
required the ritual adept in charge of image consecration ceremonies—the per-
son who had received a tantric ordination by worshiping a maṇḍala and receiv-
ing the sūrimantra—to be a fully initiated monk. This combination of a tantric
and mendicant promotion created some controversies, since Jain mendicants
are not supposed to be involved in any type of physical worship of an image.
While the majority of Śvetāmbara manuals on image consecration com-
posed after the twelfth century have monks lead the image consecration, not
all Śvetāmbaras agreed on this practice. One Śvetāmbara mendicant lineage
founded at the beginning of the twelfth century, the Paurṇamīyaka Gaccha,
argued that mendicants were morally required to avoid image consecrations.
According to the Paurṇamīyaka Gaccha, a renunciant performing añjanaśalākā
violated three of the five great vows of a mendicant: the subtle levels of
violence inherent in icons result in the mendicant violating his vow of
total non-harm (ahiṃsā); the physical transactions and operations in the
ritual violate the mendicant vow of non-possession (aparigraha); and the
use of fragrant and colourful substances in the ritual violate the vow of
total chastity (brahmacarya) (Cort 2010: 281).50
50 While no texts of this now-extinct lineage remain, Paul Dundas (2009) has summarized
a twelfth-century Tapā Gaccha polemic that systematically rejects the above arguments
associated with the Paurṇamīyaka Gaccha.
51 Cort (2010: 279–281) has discussed an early thirteenth-century A(ñ)cala Gaccha manual
that rejects mendicant participation in image consecration.
52 Carcarī, v. 20. I thank Phyllis Granoff for directing me to this.
282 Gough
not members of the Tapā Gaccha should worship icons consecrated by mem-
bers of the A(ñ)cala Gaccha.
Other prominent teachers rejected these divisions, however (Cort 2010:
283), and today image-worshiping Śvetāmbaras all agree that ācāryas should
perform the eye-opening ceremony, the añjanaśalākā. The A(ñ)cala Gaccha
persists to this day, but the two seated ācāryas of this lineage do participate
in image consecrations, with Ācārya Kalāpaprabha having performed the
añjanaśalākā of over 50 icons. Indeed, in conversation, Ācārya Kalāpaprabha
insisted that only an ācārya can perform the eye-opening ceremony and was
unaware that the A(ñ)cala Gaccha had ever rejected this practice.53 Thus,
while ācāryas of all the image-worshiping Śvetāmbara lineages today partici-
pate in pratiṣṭhā ceremonies,54 this has not been the case throughout history,
and the unease over monastic participation in these ceremonies highlights the
ostensible contradictions between many of the practices and ideologies of Jain
mendicants. Reconciling tantric image installation headed by a senior monk
with Jain ideologies of ascetic non-action was never without debate.
In the last few decades, the Digambara community has contested the role of
the ācārya in pratiṣṭhā ceremonies in a manner remarkably similar to medi-
eval Śvetāmbaras. And these similarities are no coincidence. While scholars
have only studied the Śvetāmbara sūrimantra, more comparative scholarship
needs to be done to understand the full history of Jain ritual culture, since it
appears that Digambaras and Śvetāmbaras both tantricized their image con-
secration in the exact same way, and may have even used the same esoteric
mantra to do so.
Today, many Digambara mendicants are split into two main traditions:
(1) the Terāpanth, which predominates in North India, and (2) the Bīspanth,
which is more popular in Maharashtra and the southern states (Flügel 2006:
341–344). Along with some differences in ritual conduct, Bīspanthīs, unlike
Terāpanthīs, have maintained the role of bhaṭṭārakas, or clothed, celibate clerics
who oversee libraries and temple complexes and have taken modified mendi-
cant vows that allow them to ‘occupy an intermediary status between the naked
munis and the common laity’ (Flügel 2006: 344). From around the beginning of
the thirteenth century to the revival of the naked muni tradition at the outset of
the twentieth century, bhaṭṭārakas, who wear orange robes, essentially replaced
naked monks as the heads of Digambara communities. No bhaṭṭāraka seat
remains in North India today, but fourteen bhaṭṭārakas hold seats in the South,
where they are often called on to perform image consecrations.55 Bhaṭṭārakas
are also often invited to perform pratiṣṭhā ceremonies overseas, where naked
munis are not allowed to travel.56 Bīspanthīs and Terāpanthīs are thus the two
main ‘types’ of Digambaras, along with the Kānjī Panthīs discussed at the outset
of the essay, who self-identify as Terāpanthīs. While many Jains simply identify
as ‘Digambara,’ and do not distinguish between the three branches, these dis-
tinctions are important for this paper, as the debates over the contents of the
Digambara sūrimantra and whether or not a layperson or a mendicant should
pronounce it seem to relate to their development.
Much of the debate amongst Digambaras stems from drastically differ-
ent sūrimantras in use by ritual specialists and monks, competing modern
sources on pratiṣṭhā, and disagreements over pre-twentieth-century pratiṣṭhā
practices. The earliest Digambara text on pratiṣṭhā that has been pub-
lished, Vasunandi’s Pratiṣṭhāpāṭha, was most likely composed in the twelfth
century.57 Two slightly later texts, Nemicandra’s Pratiṣṭhātilaka, which has
been dated to around 1200 CE,58 and Pandit Āśādhara’s Pratiṣṭhāsāroddhāra,
which dates to the thirteenth century, have also been published. Unlike
their Śvetāmbara counterparts, these earliest known Digambara manu-
als are already organized around the reenactment of the five auspicious
events in the life of the Jina; as noted above, Śvetāmbaras may have adopted
this practice from Digambaras. None of these early texts, however, make any
mention of a sūrimantra.
Of the pre-modern sources that have been published, only one, the
Pratiṣṭhāpāṭha by Jayasena, alias Vasubindu, mentions the sūrimantra, but it
does not give the contents of the mantra (Pratiṣṭhāpāṭha, p. 284), and the date
of this text has not been confirmed. In the conclusion of the text, Jayasena
55 For brief biographies of each of these bhaṭṭārakas, see Jain Tīrthvandanā 2012: 26–47.
56 Bhaṭṭāraka Cārukīrti, head of the maṭha at Moodbidri, Karnataka, is particularly active
overseas and has participated in more than 50 image installations ceremonies at home
and abroad (Jain Tīrthvandanā 2012: 42).
57 Vasunandi’s Pratiṣṭhāpāṭha has been published in Kunthusāgara 1992.
58 On the provenance of this text, see R. Nagaswamy, ‘Jaina Temple Rituals: Pratishtha tilaka
of Nemicandra—a study.’
284 Gough
explains how he, disciple of Kundakunda, composed the ritual manual for
a king named Lalāṭṭa to aid in the consecration of a temple dedicated to the
eighth tīrthaṅkara, Candraprabha, in the south of India (Pratiṣṭhāpāṭha,
vv. 924–925). The editor of this text, published in 1925, thus claims that
Jayasena is the direct disciple of Kundakunda, the famous philosopher who
is traditionally dated to the first century C.E. (Dośī 1925). However, like the
Nirvāṇakalikā, the contents of the Pratiṣṭhāpāṭha suggest a much later dating,
and this note likely indicates that Jayasena is merely in Kundakunda’s lineage.
Sādhvī Saumyaguṇā (2006: 427) provides the most reasonable date for this text,
estimating Jayasena’s time to be around the fourteenth-fifteenth centuries.
Present-day Digambara image consecrations, which do employ the recita-
tion of the sūrimantra, do not necessarily follow a single text, and instead com-
bine a variety of traditions. In February 2013, in Jaipur, Rajasthan, I attended
a seven-day Digambara pañcakalyāṇaka-pratiṣṭhā festival attended by Ācārya
Vibhavasāgara and his disciples (saṅgha). As part of the ceremony, dozens of
temple images were consecrated for a Terāpanthī temple. To contextualize the
use of the sūrimantra within the larger ceremony, I will provide a summary
of the program, starring the use of the mantra on the sixth day. For the rites
specific to the pañcakalyāṇaka-pratiṣṭhā such as the eye-opening outlined in
more detail below, the ritual specialists followed the instructions provided in
Pratiṣṭhāratnakara, a late twentieth-century pratiṣṭhā handbook compiled by
the Terāpanthī ritual specialist Pandit Gulābcandra ‘Puṣpa’ from various medi-
eval Digambara image installation manuals, including those listed above.59
The consecration ceremony combined the rites recorded in this manual with
daily morning and evening pūjās, lectures by the monks, popular recitations
such as the Bhaktāmara Stotra,60 and the dramatizations of the events of the
life of the Jina. The ritual specialists, under the guidance of the monks, chose
to reenact the life of the first of the twenty-four tīrthaṅkaras, Ṛṣabha.61 In
the reenactment of the Jina’s life, one of the icons being consecrated became
Ṛṣabha, and members of the congregation and professional actors filled the
other roles. Each day’s activities started early in the morning and ran late into
the evening:
59 Gulābcandra (n.d.: 48) claims that the text is mostly based on Jayasena’s Pratiṣṭhāpāṭha,
but in total he consulted about 95 texts, including Pratiṣṭhāsāroddhāra and Pratiṣṭhātilaka.
60 On the Bhaktāmara Stotra, see Cort 2005.
61 An experienced ritual specialist will have recorded in his notebook the rituals required
to act out the life of each of the tīrthaṅkaras, but for Digambaras, it is most common to
reenact the life of Ṛṣabha. Pandit Vijaykumar Jain, interview with author, Mumbai, 19 July
2013.
sūrimantra and tantricization of Jain image consecration 285
62 The Terāpanthī version of the worship of the yāgamaṇḍala given in Pratiṣṭhāratnakara
rejects the worship of certain gods and goddesses found in the Bīspanthī version. For the
Bīspanthī version of the yāga mandala, see Jñānamatī 1999).
63 On the sixteen dreams seen by the mothers of the tīrthaṅkaras in the Digambara tradi-
tion, see Wiley 2009: 47.
64 For an English translation of this seventh-century philosophical treatise, see Modi 2010.
286 Gough
During these seven days, ritual specialists (pratiṣṭhācārya) and laypeople per-
formed most of the physical (dravya) pūjās. Ācārya Vibhavasāgara and the
monks in his saṅgha only participated in the physical worship of the icons at
two points in the ceremony. The first time occurred on the fifth day, during the
performance of dīkṣā, when the ācārya performed for the icon the same dīkṣā
65 On the war between Ṛṣabha’s sons over their father’s kingdom, see Jaini 2004, 53–56.
66 Invitation card for the February 2013 Paṃcakalyāṇak Pratiṣṭhā Mahotsav of the Pārśvanāth
Digambar Jain Mandir, Sector 3, Pratāpnagar, Sanganer (Jaipur). For an illustrated descrip-
tion of a Digambara image consecration ceremony in Gujarat in 1974, see Jain and Fischer
1978: 4–14.
sūrimantra and tantricization of Jain image consecration 287
67 Standing on the platform in front of the congregation, Ācārya Vibhavasāgara performed
abhiṣeka on the icon, then removed its clothes, picked cloves off of the head of the icon to
symbolize the removal of the disciple’s hair, wrote 48 different seed syllables (bījākṣara)
with sandalwood paste on different parts of the icon to represent the saṃskāras, or sacra-
ments, monks initiate themselves into, and presented the mūrti with a broom made of
peacock feathers (piñchi), a scripture (śāstra) and water pot (kamaṇḍalu), the three pos-
sessions of a Digambara monk.
68 The jñāna-kalyāṇaka ceremony Jain and Fischer (1978: 12) witnessed corresponded more
with Śvetāmbara practices, though a monk was not present: ‘In the middle of the night, in
seclusion and secrecy, mantras [were] muttered by the paṇḍit in front of the image.’
69 For the ‘Svastyayana,’ see Gulābcandra n.d.: 375.
70 Gulābcandra (n.d., 377) lists three other mantras that should be recited 108 times at
this time: sūryakalāmantra, candrakalāmantra, and prāṇapratiṣṭhā mantra. In the con-
secration I witnessed, however, these mantras were recited earlier in ritual, during the
adhivāsana, by ritually purified laypeople. The addition of these mantras seems to be a
more recent development.
71 On the types of knowledge, see Wiley 2009, 112.
288 Gough
72 After the adhivāsana, Pratiṣṭhāpāṭha, vv. 120–121, only outlines the mouth and eye-
opening rites, while Pratiṣṭhātilaka, p. 260, only mentions the recitation of ‘Svastyayana’
and the eye-opening. Pratiṣṭhāsāroddhāra, vv. 176–184, does include all three rites—the
mouth-opening, recitation of ‘Svastyayana,’ and eye-opening—but makes no mention of
the sūrimantra .
73 . . . . netronmīlanaṃ kuryāt | tataḥ sadyaiva sūrimaṃtreṇa sarvajñatvopalaṃbhanaṃ
vidadhyāt (Pratiṣṭhāpāṭha, p. 284). ‘. . .[The ācārya] should perform the eye-opening.
Then he should bestow omniscience [to the icon] by means of the sūrimantra.’
74 ācāryeṇa sadā kāryaḥ kriyāṃ paścāt samācaret | śrī mukhodghāṭane netronmīlane
kaṃkaṇojjhane || sūrimaṃtraprayoge cādhivāsane ca mukhyataḥ | . . . (Pratiṣṭhāpāṭha,
p. 117, vv. 378–379). ‘The ācārya should always perform [the painting of syllables on the
icon mentioned in the previous lines]. After that, he, chiefly, should perform the rites with
respect to the mouth-opening, eye-opening, removal of the ritual bracelet, adhivāsana,
and pronunciation of the sūrimantra.’
75 prāṇapratiṣṭhāpyadhivāsanā ca saṃskāranetrocchṛtisūrimaṃtrāḥ | mūlaṃ jinatvā’dhigame
kriyā’nyā bhaktipradhānā sukṛtodbhavāya | (Pratiṣṭhāpāṭha, p. 108). ‘The foundational
rites in the [icon’s] obtainment of the status of the Jina are the sūrimantra, eye-opening,
saṃskāras (applied in the dīkṣā kalyāṇaka), adhivāsanā, and prāṇa pratiṣṭhā. The other
rites, characterized by devotion, are for the production of merit.’
76 As the contemporary Ācārya Kunthusāgara explains, ‘Jab tak sūri mantra nahiṃ diyā jāye,
tab tak us pratimā meṃ prāṇ nahīṃ hai’ (Kunthusāgara 1992: 212). ‘Until the sūrimantra is
given, there is no life (prāṇa) in the icon.’
77 Śvetāmbara disagreements over the content of the sūrimantra have been documented in
Jambūvijaya 1968 and 1977. These differences, however, are often over one or two lines;
there are no divergences to the extent of those between the Digambara sūrimantras out-
lined in this paper.
sūrimantra and tantricization of Jain image consecration 289
do not agree on who should pronounce this mantra. Their sūrimantras also
drastically differ from one to the next.
I have collected upwards of a dozen different Digambara sūrimantras,
though more undoubtedly exist. The sūrimantra Ācārya Vibhavasāgara used
in the ceremony outlined above is effectively a combination of two well-
known Prakrit mantras, the pañcanamaskāra-mantra translated above, and
the cattāri-maṅgala, or the mantra in which the reciter takes refuge in the
four ‘auspicious entities’ (maṅgala) of enlightened beings (arhat), liberated
beings (siddha), monks (sādhu) and the doctrine (dharma) taught by the
omniscient ones (kevalin).78Along with that formula, the text used in the cer-
emony in Jaipur, the Pratiṣṭhāratnakara, provides another option, a Sanskrit
mantra that praises the supreme Brahman, begs for happiness, victory, purity,
and an increase in merit and auspiciousness, and requests the soul of the Jina
to reside in the icon forever.79 In the introduction to the Pratiṣṭhāratnakara,
Pandit Gulābcandra “Puṣpa” insists that one of these two sūrimantras should
only be given by an ācārya or muni, i.e. the classes of Digambara mendicants
who do not wear clothes.80 Ācārya Vibhavasāgara explained to me that though
monks are normally forbidden from worshiping with physical substances, they
are required to perform the dīkṣā and jñāna-kalyāṇaka ceremonies because
(1) only a naked monk can initiate a monk, and (2) only a naked monk has the
purity and power required to properly transmit the sūrimantra.
Kānjī Panthīs, on the other hand, follow another twentieth-century
text, Pandit Nāthūlāl Jain Śāstrī’s Pratiṣṭhāpradīpa, which contains a differ-
ent sūrimantra that combines the Prakrit namaskāra-mantra with a series
of seed-syllables and consonant clusters and a Sanskrit plea for the icon to
78 Without modification, the mantra reads: oṃ hrīṃ [namaskāra mantra], cattāri maṃgalaṃ
arihaṃtā maṃgalaṃ siddhā maṃgalaṃ sāhū maṃgalaṃ kevali-paṇṇato dhammo
maṃgalaṃ | cattāri loguttamā arihaṃtāloguttamā siddhā loguttamā sāhū loguttamāṃ
kevali paṇṇaṃto dhammo loguttamā | cattāri saraṇaṃ pavvajjāmi arihaṃta saraṇaṃ
pavvajjāmi siddheśaraṇaṃ pavvajjāmi sāhūsaraṇaṃ pavvajjāmi kevali paṇṇattaṃ
dhammaṃ saraṇaṃ pavvajjāmi krauṃ hrīṃ svāhā | (Pratiṣṭhāratnakara, p. 377).
79 The mantra reads, without modification: oṃ paramabrahmaṇe namo namaḥ svasti
svasti jīva jīva naṃda naṃda vardhasva vardhasva vijayasva vijayasva anusādhi anusādhi
punīhi punīhi puṇyāhaṃ puṇyāhaṃ māṃgalyaṃ māṃgalyaṃ vardhayet vardhayet evaṃ
jinabimbe ātmaghaṭaṃ vāyuṃ pūraya pūraya āgaccha āgaccha saṃvauṣaṭ tiṣṭha tiṣṭha
ṭhaḥ ṭhaḥ cirakālaṃ nandatu vajramayāṃ pratimāṃ kuru kuru grauṃ grauṃ svāhā |
(Pratiṣṭhāratnakara, p. 377).
80 Gulābcandra, Pratiṣṭhāratnakara, ‘Apnī bāt,’ 30, 44.
290 Gough
have limitless energy.81 While Śāstrī’s text also insists that an ācārya must give
the sūrimantra, as we will see below, Kānjī Panthīs have interpreted ‘ācārya’
to mean pratiṣṭhācārya, or ritual specialist. As described in the introduction to
this paper, these ritual specialists are required to remove their clothes before
imparting the sūrimantra. The Kānjī Panthīs with whom I spoke, however,
gave no coherent reason why the moment of imparting the sūrimantra, and
no other components of the ceremony, should require nudity. One prominent
Kānjī Panthī ritual specialist explained that in an image consecration, when
the image becomes God, one should take off one’s clothes to worship it. He
reasoned that in the jñāna-kalyāṇaka ceremony, the icon is effectively trans-
formed from a renunciant to the Jina, so one should remove one’s clothes to
pronounce the sūrimantra. This answer is not consistent with regular temple
practice, however, as laypeople always keep their clothes on when performing
pūjā and abhiṣeka to consecrated temple images.
The situation becomes even more complicated when we examine the
practices of Bīsapanthīs, who most commonly follow the instructions of
Nemicandra’s Pratiṣṭhātilaka, which, as noted above, does not contain a
sūrimantra. As such, until recently, the sūrimantra was not imparted to icons
in pratiṣṭhā ceremonies in South India. According to one ritual specialist, it is
only with the influence from North India that ācāryas have begun the practice
in the South in the last few decades, and some ceremonies do not include the
recitation of the sūrimantra at all.82 The edition of Pratiṣṭhātilaka most often
used today was published in 2006 by Gaṇinī Āryikā Jñānamatī Mātā, who is
stationed in Hastinapur, outside Delhi. In this edition, the sūrimantra is not
mentioned in the root text, but the same sūrimantra as found in Nāthulāl
Śāstrī’s manual has been included in an appendix (Pratiṣṭhātilaka, p. 357).
In the introduction to the text, Jñānamatī’s disciple, Āryikā Candanāmatī
Mātā, recognizes that the original Pratiṣṭhātilaka makes no mention of a
sūrimantra, but she explains that the mantra has been provided in the appen-
dix because its use has become popular in the present day (Pratiṣṭhātilaka,
p. 14). She makes no rule about when or by whom the sūrimantra should be
recited. Indeed, Jñānamatī’s saṅgha in Hastinapur has become somewhat lib-
eral about the recitation of the mantra. They are the only saṅgha, to my knowl-
edge, to allow nuns (āryikā) to give the sūrimantra, and when I observed a
81 The mantra reads, without modification: oṃ hrāṃ hrīṃ hrūṃ hrauṃ hraṃ a si ā u sā
arhaṃ oṃ hrīṃ rmlvyūṃ jmlvyūrṃ tmlvyūrṃ lmlvyūrṃ vmlvyūrṃ pmlvyṛūṃ bhmlvyūrṃ
kṣmlvyūrṃ [pañcanamaskāra mantra] anāhataparākramaste bhavatu te bhavatu te
bhavatu hrīṃ namaḥ.
82 Pandit Pradeepkumar, interview by author, Mumbai, 9 August 2013.
sūrimantra and tantricization of Jain image consecration 291
Figure 2 The sūrimantra is recited into the ear of a Jina icon during an image consecration
ceremony in Hastinapur, North India. January 21, 2013.
83 On the gāyatrīmantra, see Lal 1971. Interestingly, in the Āśvalāyana Gṛhyapariśiṣṭa’s
account of image consecration, the recitation of the gāyatrī-mantra into the ear of the
icon is required after the icon has been established in the temple (Einoo 2005: 99).
292 Gough
84 oṃ hrīṃ oṃ śrīṃ oṃ bhūḥ oṃ bhuvaḥ oṃ svaḥ oṃ māḥ oṃ hāḥ oṃ māḥ oṃ tatsavitu
vareṣyaṃ garbho devaya dhī mahī dhimo miḥ a si ā u sā ṇamo arahaṃtāṇaṃ anāhata
parākramaste bhavatu bhavatu oṃ ayaṃ mahānubhāvaḥ sākṣāt tīrthaṅkaro bhavatu
eṣo ’rhata sāksātavatīrṇo viśva pātviti svāhā. Taken (without modification) from Pandit
Pradeepkumar’s notebook, Mumbai, 9 August 2013.
85 Without modification, the mantra reads: oṃ hrīṃ kṣūṃ hrūṃ suṃ suḥ krauṃ hroṃ aiṃ
arhaṃ namaḥ sarva arahanta guṇabhāgī bhavatu svāhā (Kunthusāgara n.d.: 236).
sūrimantra and tantricization of Jain image consecration 293
should read another short Sanskrit mantra five times. This mantra, which con-
tains mostly seed syllables, praises the jinas.86 Finally, he should read another
longer Sanskrit87 mantra seven times. This mantra honors the ‘three jewels’ of
right faith, knowledge, and conduct, the original disciples of the tīrthaṅkaras
(gaṇadhara), the ritual diagram that praises the superhuman powers of these
disciples (gaṇadharavalaya), and the ācāryas. It also requests that the icon
remain fixed and that the teachers of Jain dharma who have knowledge of
tantra and yantras prosper.88
6 Digambara Controversies
We thus have several different versions of the sūrimantra and several differ-
ent interpretations of how it should be imparted. Some Digambaras do not
require the recitation of the sūrimantra, others instruct a clothed ritual spe-
cialist or lay person to recite it, others require the ritual specialist to remove
his clothes before reciting it, and others are adamant that only a naked monk
can pronounce the formula. Because of these disagreements, in recent years,
Digambara scholars from both sides of the debate over the role of the ācārya in
pratiṣṭhā ceremonies have published their understandings of the history and
proper use of this mantra.
Members of the Kānjī Swāmī Panth seem to have been especially invested
in this debate, as their icons at times have been deemed unworthy of worship
in ways reminiscent of medieval Śvetāmbara developments. One leader in
the Kānjī Svāmī community recounted to me a time when a Digambara muni
refused to take darśan in the temples at Maṅgalāyatan, the Kānjī Svāmī Panth
pilgrimage site situated just outside of Aligarh, Uttar Pradesh (est. 2002). The
monk, rejecting an invitation from the Maṅgalāyatan community, had claimed
that their icons were unworthy of worship (apūjya) because a naked monk had
not imparted the sūrimantra to them.
86 oṃ hrīṃ śrīṃ klīṃ hrāṃ hrauṃ śrīṃ śrauṃ jaya jaya drāṃ kali drā[ṃ] kṣa sāṃ mṛṃjaya
jinebhyoḥ oṃ bhavatu svāhā (Kunthusāgara n.d.: 236).
87 Only a small component of this mantra, a section of the pañcanamaskāra-mantra, ‘praise
to the ācāryas’ (ṇamo āiriyāṇaṃ), is in Prakrit.
88 Without modification, the mantra reads: oṃ hrīṃ krauṃ samyagdarśana jñāna
cāritrātara gātrāya caturaśiti guṇa gaṇadhara caraṇāya aṣṭhacatvāriṃśata gaṇadhara
valāya ṣaṭtriṃśata guṇa saṃyuktāya ṇamo āiriyāṇaṃ haṃ haṃ sthiraṃ tiṣṭha tiṣṭha ṭhaḥ
ṭhaḥ cirakālaṃ naṃdatu yaṃtra guṇa taṃtra guṇaṃ vedayutaṃ anaṃta kālaṃ vardhay-
antu dharmācāryā huṃ ruṃ kuru kuru svāhā svādhā (Kunthusāgara and Vijayamatī n.d.:
237).
294 Gough
to Nāthūlāl Jain Śāstrī, there should be no reason to think that a layperson who
has removed his clothes can impart the sūrimantra.
In actuality, it is likely that the true interpretation of this line from Jayasena’s
Pratiṣṭhāpāṭha lies somewhere in between the claims of these two ritual spe-
cialists. Jayasena’s text likely does instruct the ācārya to remove his clothes,
but instead of referring to a ritual specialist, he probably addresses a clothed
bhaṭṭāraka. If we accept Saumyaguṇa’s dating of Jayasena’s Pratiṣṭhāpāṭha to
around the fifteenth century, then it was composed when bhaṭṭārakas had
essentially replaced naked monks. Ravindra K. Jain (1999: 33) explains that
bhaṭṭārakas would remove their orange robes ‘in honour of the ancient ideal
[of asceticism] only when eating and when initiating another bhaṭṭāraka.’
Jayasena’s text suggests that consecrating images could be added to this list of
occasions when the bhaṭṭārakas would remove their clothes. As bhaṭṭārakas
were also termed ‘ācārya,’ Jayasena’s disputed instruction for the ācārya to
‘become naked’ could address an orange-robed bhaṭṭāraka. But was this removal
of clothes also ‘in honor of the ancient ideal’? Before the rise to prominence
of clothed bhaṭṭārakas around the thirteenth century, did naked Digambara
ācāryas, like their Śvetāmbara counterparts, employ the sūrimantra in image
consecration ceremonies? And was this sūrimantra imparted to them at the
time of their promotion to the rank of ācārya?
Kṣetra Beḍiyā in western Gujarat.92 Neither a date nor an author of these rites
is provided, but in the rite of the promotion of a bhaṭṭāraka (bhaṭṭārakapada-
sthāpanā-vidhi) described in this text, the guru is required to impart the
sūrimantra to the bhaṭṭāraka-to-be.93 This imparting of the sūrimantra is not
part of any other promotion of a mendicant—even an ācārya. Thus, receiv-
ing the sūrimantra seems to have been a key ritual action in distinguishing
between an ācārya and a bhaṭṭāraka.
Indeed, while I have only found seven different pre-modern Digambara
references to what is either termed the ‘sūramantra’ or the ‘sūrimantra,’ in all
the sources apart from Jayasena’s text on image consecration, ‘receiving the
sūramantra’ goes hand-in-hand with ‘becoming a bhaṭṭāraka.’94 All of these
references date from the seventeenth to early nineteenth centuries. For exam-
ple, in the early seventeenth century, a disciple of Bhaṭṭāraka Kumudacandra,
Gaṇeśa, composed a hymn praising his teacher (gurustuti) that reads:
92 For a discussion of this text, see Jain 2009: 116. I thank Tillo Detige for pointing me towards
it. Detige has also collected several undated manuscripts on bhaṭṭārakapada-sthāpanā
from the Sonāgiri Bhaṭṭāraka Granthālāya in the pilgrimage site of Sonagiri, Madhya
Pradesh, that confirm that bhaṭṭārakas in North India were imparted the sūrimantra
upon their promotions. Tillo Detige, email to author, 30 December 2013.
93 tataḥ śṝiguruḥ tasmai tatpadayogyaṃ paramparāgataṃ sūrimantraṃ dadyāt (Jain
2009: 117).
94 For another example, see the description of the promotion of Bhaṭṭāraka Dharmakīrti
from an undated piece of writing whose language places it in the eighteenth-nineteenth
centuries: ‘Sakalakīrti . . . śrīnogāme saṃghe padsthāpan karīne sāgvāḍe jaīne potānā
putrakne pratiṣṭhā karāvī pote sūrmaṃtra dīdho te dharmakīrtie varṣ 24 pāṭ bhogavyo’
(Johrapurkar 1958: 236, number 330). ‘Sakalakīrti, having established a bhaṭṭāraka seat in
the town of Śrīnogāma, having gone to Sagwara (Rajasthan), having established his son
[in the seat of the bhaṭṭāraka], gave his own sūr[i]mantra to Dharmakīrti, who remained
in that position of bhaṭṭāraka for 24 years.’
95 Kāslivāl (1967: 237) records two other verses from a gurustuti by Gaṇeśa that describe how
Kumudacandra received the ‘sūramantra’ when he became a bhaṭṭāraka.
sūrimantra and tantricization of Jain image consecration 297
Oh, brother! This excellent, enchanting monk was the light of the
Sarasvatī Gaccha. He became Kumuducandra Bhaṭṭāraka and all pious
people were enchanted.
Here, we can see that Kumuducandra received the sūrimantra when he became
a bhaṭṭāraka of the Sarasvatī Gaccha.
Other sources suggest that bhaṭṭārakas belonged to a special class of
Digambara ācāryas who had received the sūrimantra and thus, unlike regu-
lar ācāryas, were able to consecrate images. One popular story related to the
sūrimantra, which is mentioned in several different paṭṭāvalīs, or listings of
the succession of bhaṭṭārakas of different lineages, most convincingly illus-
trates this point. This story describes the promotion of the fourteenth-century
Bhaṭṭāraka Padmanandi. According to the tale, in Saṃvat 1375 (1318–19 C.E.), a
wealthy patron wished to have a pratiṣṭhā performed in Gujarat, so he invited
the Bhaṭṭāraka Prabhācandra to undertake the ceremony. Prabhācandra,
however, was stationed outside of Gujarat at the time, and could not attend
the ceremony. Therefore, one of the ācāryas in Prabhācandrasūri’s lineage
who was in Gujarat, Padmanandi, was given the sūrimantra and promoted
to the rank of bhaṭṭāraka so that he could perform the pratiṣṭhā.96 In this story,
the key distinction between an ācārya and a bhaṭṭāraka—between someone
who can perform pratiṣṭhā and someone who cannot—is the sūrimantra.
But when, exactly, this practice emerged must be an inquiry for fur-
ther research. I have not found a Digambara mention of a sūrimantra that
pre-dates the fifteenth century (if we date Jayasena’s Pratiṣṭhāpāṭha to this
time and believe the instruction was originally in the text), and all known
pre-modern references to the Digambara sūrimantra, including the manu-
script on bhaṭṭāraka ordination Premī references, come from North India, in
Rajasthan and Gujarat, where Śvetāmbaras have been populous for centuries.
Could Digambaras have adopted this practice from Śvetāmbaras? The parallels
between contemporary Śvetāmbara and Digambara image consecration prac-
tices, in which a monk pronounces an invocation termed ‘sūrimantra’ into the
ear of the icon at the time of the jñāna-kalyāṇaka ceremony, certainly suggest
communication between the two sects, but the lines of influence are not clear.
Further research, ideally by specialists of South Indian languages, is required
to answer this question.
96 Two accounts of this story, both from paṭṭāvalīs from the early nineteenth century, are
translated into English by Hornle 1892: 57–84. Two other paṭṭāvalīs that record this story
are published in V.P. Johrapurkar 195: 91 and in Kāslivāl 1997: 453.
298 Gough
101 For another outline of the gaṇadharavalaya as used in Digambara pratiṣṭhā, see
Nemicandra, Pratiṣṭhātilaka, pp. 329–331.
102 See also Jambūvijaya’s (1977: 299–304) discussion of other Śvetāmbara examples of the
gaṇadharavalaya.
300 Gough
the final two syllables stand for the fruit of renunciation, honor in the three
worlds, and eventual liberation.103
This Digambara manuscript thus describes the contents of the fifth section
(pīṭha) of the Śvetāmbara sūrimantra, as described in the first section of this
essay. While this section differs slightly from mendicant lineage to mendicant
lineage, and some versions begin with kirimeru, not irimeru, other versions do
begin with irimeru.104 Unfortunately, how this single folio of the sūrimantra,
written on the back of an entirely different text, found its way to this temple
in Jaipur remains a mystery, but it does provide further proof that there was
at least some textual exchange or concordance between Śvetāmbaras and
Digambaras regarding the sūrimantra, though modern members of these two
sects are completely unaware of these pre-modern exchanges. While this man-
uscript is not more than a couple hundred years old, the mantra recorded in
it may have been in use for much longer—it may have been used by naked
Digambara monks to consecrate temple images. More research needs to be
done on the exact contents of the pre-modern Digambara sūrimantra.
Concluding Remarks
103 Śrī Digambar Jain Candraprabhujī Mandir, Aṅkroṃ kā Rāstā, Kiśanpol Bāzār, Jaipur, man-
uscript no. 24. I thank Phyllis Granoff for help in reading this manuscript.
104 For examples of Śvetāmbara sūrimantras whose fifth sections correspond to the one
found in this manuscript, see Jambūvijaya 1977: 320, 322 and 338.
sūrimantra and tantricization of Jain image consecration 301
105 In another example of this ‘tantricization’ of image consecration, Koichi Shinohara
(2014b) has shown how Chinese Esoteric rites of image consecration were modeled on
the abhiṣeka of the ācārya.
302 Gough
References
Cort, John E. 2006. ‘Installing Absence? The Consecration of a Jina Image.’ In Presence:
The Inherence of the Prototype within Images and Other Objects, eds Robert Maniura
and Rupert Shepherd. Aldershot and Burlington: Ashgate, 71–86.
Cort, John E. 2010. ‘World Renouncing Monks and World Celebrating Temples and
Icons: The Ritual Culture of Temples and Icons in Jainism.’ In Archaeology and Text:
The Temple in South Asia, ed. Himanshu Prabha Ray. Oxford: Oxford University
Press, 267–295.
Deo, Shantaram Bhalchandra. 1956. History of Jaina Monachism from Inscriptions and
Literature. Poona: Deccan College Postgraduate and Research Institute.
Devendrasāgara, Muni, ed. 1943. Śrī Bṛhadyogavidhi. Ujjain: Śrī Ṛṣabhdevjī
Chagnīrābhjīnī Peḍhī.
Dhaky, M.A. 1994. ‘The Nāgendra-gaccha.’ In Dr. H.G. Shastri Felicitation Volume, eds
P.C. Parikh et al. Ahmedabad: H.G. Shastri Felicitation Volume Committee, 37–42.
Dośī, Hīrācand Nemcand. 1925. Introduction (‘Prastāvanā’) to Pratiṣṭhāpāṭha Saṭīka, by
Jayasena. Edited by Hīrācand Nemacand Dośī. Sholapur: Śeṭh Hīrācand Nemacand
Dośī.
Dundas, Paul. 1998. ‘Becoming Gautama: Mantra and History in Śvetāmbara Jainism.’
In Open Boundaries: Jain Communities and Cultures in Indian History, ed. John E.
Cort. Albany: State University of New York, 31–66.
Dundas, Paul. 2002. ‘Haribhadra on Giving.’ Journal of Indian Philosophy 30, 1–44.
Dundas, Paul. 2007. History, Scripture and Controversy in a Medieval Jain Sect. London
and New York: Routledge.
Dundas, Paul. 2009. ‘How Not to Install an Image of the Jina: An Early Anti-
Paurṇamīyaka Diatribe.’ International Journal of Jaina Studies (Online) 5/3,
1–23.
Einoo, Shingo. 2005. ‘Notes on the Installation Ceremonies Described in the
Gṛhyapariśiṣṭas.’ In From Material to Deity: Indian Rituals of Consecration, eds
Shingo Einoo and Jun Takashima. New Delhi: Manohar, 95–114.
Fischer, Eberhard and Jyotindra Jain. 1976. Art and Rituals: 2500 Years of Jainism in
India. New Delhi: Sterling Publishers.
Flügel, Peter. 2006. ‘Demographic Trends in Jaina Monasticism.’ In Studies in Jain
History and Culture, ed. Peter Flügel. London: Routledge, 312–398.
Folkert, Kendall. 1993. Scripture and Community: Collected Essays on the Jains. Edited by
John E. Cort. Atlanta: Scholars Press.
Goodall, Dominic, ed. and transl. 2004. The Parākhyatantra: A Scripture of the
Śaiva Siddhānta. Pondichéry: Institut Français de Pondichéry/École Française
D’Extrême-Orient.
Gough, Ellen. 2015. ‘The Ṇamokār Mantra’s Forgotten Brother.’ Jaina Studies: Newsletter
of the Centre of Jaina Studies 10, 32–34.
304 Gough
Gulābcandra, Pandit ‘Puṣpa.’ n.d. Pratiṣṭhā Ratnakar. Edited by Darbārī Lāl Koṭhiyā and
Jaykumār ‘Niśānt.’ Delhi: Prati Vihār Jain Samāj.
Ṣoḍaśakaprakaraṇa of Haribhadrasūri. Jalor: Śrī Ramjanavijayajī Jain Pustakālay, 1995.
Hegewald, Julia A.B. 2010. ‘Visual and Conceptual Links Between Jaina Cosmological,
Mythological and Ritual Instruments.’ Journal of Jaina Studies (Online) 6/1, 1–20.
Hikita, Hiromichi. 2005. ‘Consecration of Divine Images in a Temple.’ In From Material
to Deity: Indian Rituals of Consecration, eds Shingo Einoo and Jun Takashima. New
Delhi: Manohar, 143–197.
Hoernle, Rudolf A.F. 1891. ‘Two Pattavalis of the Sarasvati Gachcha of the Digambara
Jains.’ Indian Antiquary 20, 341–361.
Hoernle, Rudolf A.F. 1892. ‘Three Further Pattavalis of the Digambaras.’ Indian
Antiquary 21, 57–84.
Jain, Jyotindra and Eberhard Fischer. 1978. Jaina Iconography. Part One. Leiden: E.J.
Brill.
Jain, Ratancandra. 2009. Jainparamparā aur Yāpanīyasaṃgh, Part 2. Agra: Sarvoday
Jain Vidyāpīṭh.
Jain, Ravindra K. 1999. The Universe as Audience: Metaphor and Community among the
Jains of North India. Shimla: Indian Institute of Advanced Study, 1999.
Jain, Sāgarmal. 2007. Introduction (bhūmikā) to Ācāradinakara, Tṛtiya Khaṇḍa:
Pratiṣṭhā, Śāntikakarma, Pauṣṭikakarma evaṃ Balividhāna, by Vardhamānasūri.
Translated into Hindi by Sādhvī Mokṣaratnā. Edited by Sāgarmal Jain. Śājāpur,
Madhya Pradesh: Prācya Vidyāpīṭh.
Jain Tīrthvaṃdanā: Bhāratvarṣīya Digambar Jain Tīrthkṣetra Kameṭī kā Mukhpatra 2/11
(May 2012).
Jaini, Padmanabh. 2004. ‘Ahiṃsā and ‘Just War’ in Jainism.’ In Ahiṃsā, Anekānta and
Jainism, ed. T. Sethia. Delhi: Motilal Banarsidass, 47–61.
Jambūvijaya, Muni, ed. 1968, 1977 (contin. pagination). Sūrimantrakalpasamuccaya. 2
vols. Bombay: Jain Sāhitya Vikās Maṇḍal.
Jhaverī, Mohanlāl Bhagavāndās. Introduction to Nirvāṇakalikā, edited by Mohanlāl
Bhagavāndās Jhaverī, 1–20. Bombay: Seth Nathamalji Kanayyalaji Ranka, 1926.
Johrapurkar, V.P. 1958. Bhaṭṭāraka Saṃpradāya: A History of the Bhaṭṭāraka Pīṭhas
Especially of Western India, Gujarat, Rajasthan and Madhya Pradesh. Sholapur: Jaina
Samskriti Samrakshaka Sangha.
Jñānamatī, Āryikā. 1999. Yāgmaṇḍal Vidhān. Hastinapur: Digambar Jain Trilok Śodh
Saṃsthān.
Jñānamatī, Āryikā. 1997. ‘Muslim Yug ke Jainācārya.’ In Praśammūrti Ācārya Śāntisāgara
Chāṇī Smṛti Graṃth, ed. Kapūrcand Jain. Muzaffarnagar: Shri Mahavira Tyre
Agencies, 451–471.
Kalyāṇavijayagaṇi. 1956. Kalyāṇakalikā. Edited by Muni Bhadraṃkaravijaya. Jālor
Mārvāḍ, Rajasthan: K.V. Śāstrasaṃgrahasamiti.
sūrimantra and tantricization of Jain image consecration 305
Kāslivāl, Kastūrcand. 1967. Rājasthān ke Jain Sant: Vyakritva evaṃ Kṛtitva. Jaipur: Śrī
Digambar Jain Atiśay Kṣetra Mahāvīrjī.
Khāntivijayagaṇi, Paṅnyāsa, ed. 1926. Śrī Bṛhadyogavidhi. Jhīṃjhuvāḍā: Ūmedkhānti
Jain Jñānmandir.
Krishna Lal. 1971. ‘Sāvitrī—From Saṁhitās to Gṛhyasūtras.’ Annals of the Bhandarkar
Oriental Research Institute, 52, 1/4, 225–229.
Kunthusāgara, Gaṇadharācārya, comp. 1992. Pratiṣṭhā Vidhi Darpaṇ. Jaipur: Kunthu
Vijay Granthmālā Samiti.
Kunthusāgara, Gaṇadharācārya and Gaṇinī Āryikā Vijayamatī, comp. n.d. Laghu
vidyānuvād. Jaipur: Kunthu Vijay Granthmālā Samiti.
Mantrarājarahasyam of Siṃhatilakasūri. 1968. In Sūrimantrakalpasamuccaya,
Volume I, edited by Muni Jambūvijaya. Bombay: Jain Sāhitya Vikās Maṇḍal,
1–74.
McGregor, Ronald Stuart. 1984. Hindi Literature from its Beginnings to the Nineteenth
Century. Wiesbaden: Otto Harrassowitz.
Modi, Manish (transl.) 2010. Acarya Pujyapada’s Ishtopadesha: Beneficial Teachings.
Mumbai: Hindi Granth Karyalay.
Mori, Masahide. 2005. ‘The Installation Ceremony in Tantric Buddhism.’ In From
Material to Deity: Indian Rituals of Consecration, eds Shingo Einoo and Jun
Takashima. New Delhi: Manohar, 199–240.
Nāhar, Pūraṇcandra, comp. 1918. Jain Lekh Saṃgrah. Pratham Khaṇḍ. Calcutta: Jain
Vividh Sāhitya Śāstramālā.
Nagaswamy, R. ‘Jaina Temple Rituals: Pratishtha tilaka of Nemicandra—a study.’
http://tamilartsacademy.com/journals/volume2/articles/jain-temple.html (accessed
15 August 2013).
Nirvāṇakalikā. 2010. Appendix C of Historical Development of Jaina Iconography by
A.K. Bhattacharyya. Delhi: Bharatiya Kala Prakashan.
Nirvāṇakalikā. 1926. Edited by Mohanlāl Bhagvāndās Jhaverī. Bombay: Śeṭh Nathamaljī
Kanhaiyālāljā Rāṃkā.
Nirvāṇakalikā. 1981. Edited by Pūjyapaṅnyāsa Jinendravijaya. Devlālī, Rajasthan: Śrī
Bhuvansudarśan Jain Granthmālā.
Pañcāśakaprakaraṇa of Haribhadrasūri. 1997 Edited by Sāgarmal Jain and
Kamleśkumār Jain. Translated into Hindi by Dīnānāth Śarma. Varanasi: Pārśvanāth
Vidyāpīṭh.
Parmār, Alpnā. 2010. ‘Bhaṭṭārakoṃ kā Pañc Kalyāṇak Pratiṣṭhāoṃ meṃ Yogdān.’ In
Bhaṭṭārak Paramparā, ed. P.C. Jain. Jaipur: Jain Anuśīlan Kendra, Rājasthān
Viśvavidyālay, 176–193
‘Phoenix Pratishtha Mahotsav: A Jain Temple Rises in Valley of Sun.’ Jain Digest: A
Publication by the Federation of Jain Association in North America 28/1 (Winter 2009),
11–12.
306 Gough
Wiley, Kristi L. 2012. ‘Supernatural Powers and their Attainment in Jainism.’ In Yoga
Powers: Extraordinary Capacities Attained Through Meditation and Concentration,
ed. Knut A. Jacobsen. Leiden: E.J. Brill, 145–194.
Williams, Robert. 1963. Jaina Yoga: A Survey of the Mediaeval Śrāvakācāras. Delhi:
Motilal Banarsidass.
Williams, Robert. 1965. ‘Haribhadra.’ Bulletin of the School of Oriental and African
Studies, University of London 28/1, 101–111.
CHAPTER 12
It was a great feast for the Tamil migrants from Sri Lanka, and for continental
Europe a novelty altogether, to witness in July 2002 the consecration of the
first Hindu temple in the classical style of South Indian architecture in the
industrial area of Hamm-Uentrop, North-Rhine Westphalia (Germany). My
article presents a fieldwork study of traditional Śaiva-Āgamic consecration
rites in a new cultural environment, and explores performance aspects and
social effects. This includes circumstances and conditions, i.e. the wider con-
text, starting with the question: Who was able to achieve this little wonder
in a recent migrant community whose first generation belongs to the lowest
income groups,1 and moreover, in a country which witnessed much dissent
regarding the building of Muslim mosques in traditional regional styles? I am
going to start with this question to build a contextual frame for the detailed
description of the consecration rites and subsequent developments.
My approach is local sociology and ethnography of religion (regular field-
work at the temple since 1999). The period covered in this article ranges from
the consecration of the Sri Kamadchi Ampal Temple, Europe e.V.,2 in Hamm-
Uentrop in 2002 (June 30 to July 7) to its renewal—the re-consecration after
12 years—in 2014 (April 27 to May 4). Both these impressive events and the
twelve years in-between, which mark great changes and amazing develop-
ments, will be considered.3 There is a wide range of questions to be discussed:
What happened in 2002 during the eight days of extensive ritual action—and
in the preparatory and concluding rites? How did faithfulness to tradition
connect with re-modelling of tradition? What adaptations were made and
what exchanges took place with the new homeland culture? What local,
national and transnational dynamics did the consecration trigger off—regard-
ing the Tamil community, Hinduism in Germany, and effects on the majority
society? What difficulties does the temple encounter despite its great success?
How did the re-consecration in 2014 look like—in continuity and change to
the ceremony 2002 regarding rituals, actors, and participants?
These questions are not rhetorical and their suggestiveness is being inten-
tional. The consecration of the Kamadchi Temple marked the new visibility of
a largely invisible migrant religion. The advent of ‘ethnic Hinduism’ is a nov-
elty in Germany, and the Kamadchi Temple and its festivals played a major
role in public representation and perception. This temple was pivotal in the
successful rooting of ethnic Hinduism in German public space. Nowadays, the
temple is also internationally known and attracts up to 20.000 visitors from
all over Europe at the yearly temple festivals and big-style chariot processions
(in Tamil shortly tēr, ‘chariot,’ indicating the great procession and day). It is so
far the largest Hindu temple on the European continent and developed to be
a Hindu pilgrimage place not only for Tamil Hindus, but also for local natives
who enjoy finding India next door. The temple’s consecration in 2002 was the
starting point and trigger of these developments. From today’s perspective,
one can say, the Mahā-Kumbha-Abhiṣeka4 ceremony of 7 July 2002 was the
birthday of a new pilgrimage place of transnational reach and of other pro-
found changes.
The aim is a) to give detailed account of the set of actions that constituted
the consecration and inauguration of the Kamadchi Temple, and their recent
renewal, and b) to shed light on the local, national and international dynamics
and exchanges which followed the consecration ceremony, i.e. the construc-
tion and ritual sanction of a sacred building which looks like a real traditional
South Indian temple. Already from far it is discernable as such with its red and
white stripes and richly sculptured gopura and vimāna towers. As it turned
out, the consecration of a ‘real’ temple was very important for the politics of
4 This Sanskit term denotes the ‘great’ (mahā) ‘ablution’ (abhiṣeka) of the temple towers and
interior shrines with ‘jars’ (kumbha) of ritually empowered water, which marked the end
and peak of the consecration rites of the previous days. Concerning consecration, I am using
the Sanskrit terms, as the Ᾱgama Śāstras are mainly in Sanskrit and a strong tendency of
Sanskritization is detectable in the Kamadchi temple. The ceremony was announced by the
Sanskrit term in Tamil script and adapation (mahā kumpāpiṣēkam), and in German shortly
called ‘Einweihungszeremonie’ (‘cerermony of consecration’). In 2014, also the Latin translit-
eration ‘Maha Kumbhabhisekam’ was used to announce the event.
Āgamic Temple Consecration Transnationalized 311
identity and recognition of the Sri Lankan Tamil Hindu refugees in Germany,
and as well crucial for local, nationwide and international dynamics.
In terms of theory, I suggest a contact zone approach and an interactional
perspective. I am going to show that the Kamadchi Temple included in the
consecration rituals deliberate elements of integrating the new surroundings
and society. Indeed, the temple itself turned into a contact zone after its erec-
tion and consecration. The temple and its splendid yearly festival became not
only a major meeting place for Sri Lankan Tamils both across the country and
beyond, but had as well an impact on religious identities of German locals since
2002. Migration and adaption to the new country is not a one-way street. While
addressing ritual reconstruction, a focus on transfer is equally important. The
change and multifarious exchange in the new homeland must be considered,
too. When discussing the dynamics triggered off by the Kamadchi Temple, I try
to keep a marked interactional perspective regarding local, national, and trans-
national cultural flows. It is of interest that these dynamics had also impact on
other Hindu communities in Germany.
The ethnography presented will lead to some theoretical conclusions on the
transnationalization of ethnic Hinduism, aspects of change, and the role of
the aesthetics of sacred architecture and traditional rituals. The final remarks
will also address the contingency of the rapid developments and an uncertain
future of the Kamadchi temple despite its great success over the past 12 years.
5 At least 42.000 Sri Lankan Tamil Hindus remained in the country by 2013 and about half of
the Sri Lankan migrants (c. 33.000) obtained German citizenship.
312 Wilke
in North-Rhine Westphalia, where more than 50% of the migrants have settled.
Accordingly, most places of worship were established here—among them the
Sri Kamadchi Ampal Temple, e.V. Europe, of Hamm-Uentrop, on which I focus
in this article.
Initially, the ‘Europe’ in the title hinted at ambitious dreams of the initiative
chief priest Sri Paskaran Kurukkal who is the founder of the temple—nowa-
days it is much closer to reality. ‘Kamadchi’ indicates that the priest views his
German temple as a replica of the famous Kāmākṣī temple in Kanchi(puram),
Tamil Nadu. This South Indian pilgrimage place is closely related to the
Śaṅkarācāryas of Kanchi and to Śrīvidyā philosophy and ritual. Likewise, the
German temple is based on Āgamic (Śaiva Siddhānta) Hinduism in its exoteric
ritual (regular temple service thrice a day, performed by one of the two to four
regular priests6), and includes the performance of esoteric Śrīvidyā worship
(śrīcakra-pūjā and homa) (on special days; sole privilege of the chief priest).
There are several Śaṅkara icons on display. The temple also incorporates strong
elements of (Vedic-Purāṇic) Smārta Hinduism and Tamil bhakti. Unlike the
orthodox Brahmanic blueprint of Kanchi, but less uncommon in the ritual life
of Jaffna (Sri Lanka), the German Kamadchi temple moreover includes popu-
lar or ‘folk’ religion—most visibly in self-afflicting vows and kāvaṭi dance at the
yearly temple festival.7
The dominant stream in Sri Lanka/Jaffna and also in Germany is Śaiva
Siddhānta und popular Śaivism. It is therefore a noteworthy exception that
the pious founder and chief priest of the Kamadchi temple of Hamm-Uentrop
is neither a Brahmin nor belonging to mainstream Śaiva Siddhānta, but a Vīra
Śaiva. He comes from a Vīra Śaiva family of priests in Jaffna, some of whom
serve nowadays in the German temple founded by him. He took to Śrīvidyā
in Tamilnadu, where he migrated first before coming to Germany. The priest
attributes his success in establishing a real temple in Germany to the esoteric
Śrīvidyā rites (śrīcakra-pūjā and homa). Wall pictures, showing him performing
6 The number differs due to visa problems, and the like. However, at least two priests (Sri
Paskaran and his brother in law) will always be there. Both were already serving in the previ-
ous ‘temple’ (a simple one-family house) across the street, i.e. even before the inauguration
of the present structure took place. Since 2002 at least three regular priests were the average.
7 Cf. Wilke 2003: 125–168; 189–222. During 2003 and 2004 even ‘hook-swinging’ (‘bird kāvaṭi’)
was seen in Hamm-Uentrop. The practice was abandoned, however, after an accident with a
broken swinging beam. Sri Paskaran says that he was glad that the practice stopped. Although
he allows violent practices, because the participants want to perform them, he does not hap-
pily sanction them. He prefers a more spiritual kind of Hinduism, and argues that he only
allowed hook-swinging for vows that serve to avert great dangers, such as to prevent impris-
onment or being sent back to war-torn Sri Lanka.
Āgamic Temple Consecration Transnationalized 313
them and being graced by the vision of the goddess, so to speak support this
claim. The existence of such public pictures of self-staging may hint at the fact
that a non-Brahmin establishing an Āgamic temple needs strong legitimacy—
not less than authorization by the goddess herself. This is being supplemented
by the refuge to a very orthodox Brahmanic tradition (Śaṅkara’s Vedānta), a
value for renunciation (the priest is always orange-clad despite of being mar-
ried and having children), magically powerful Tantric worship (śrīcakra-pūjā),
and strictness in ritual purity (rules for anyone entering the temple ground—
no alcohol, no meat, no cigarettes, no menses—were publically announced
after the consecration ceremony). The wall-pictures hint also at the fact that
not only pious fervor, devotion and ritual proficiency (which are undoubt-
edly there) made Sri Paskaran successful, but also his charisma, self-staging,
ever new innovative ideas, ambitious plans, and managing skills. He had to be
inventive, as no public funding is available: All temples in Germany are pri-
vately sponsored.
Sri Paskaran was the first who started public processions at the yearly tem-
ple festival already in 1993, which remained an important financial resource.
And again, he was the first who managed to erect a traditional structure and
get enough funds for it from smaller and bigger donations. Altogether, the
costs of temple building ultimately amounted to almost three million Euro—
covering the surface (edifice and embellishment of the vimāna and gopura
towers) and the interior, including new, black granite deity images from
Mahabalipuram (except for the Navagraha) and colorful painted shrines for
the divinities. In addition to the central shrine (garbhagṛha) of Kamadchi Devī
and the vasantha-mantapa for the procession icons (utsava-mūrti), there are
a number of side-shrines harboring all major deities of South India (Gaṇeśa,
Śiva(liṅga), Murugan, Lakṣmī-Nārāyaṇa, Somāskanda, Navagraha, Ayappaṉ,
Bhairava, Caṇḍeśvara) and a separate shrine of Śani (the planet Saturn) out-
side the temple. It was very crucial to procure better-off sponsors for each of
the shrines. These sponsors—from different German cities—had important
functions at the consecration and re-consecration, and been supportive all the
years. Starting with the big-style consecration, the temple interior, festivals,
and temple grounds were made ever more beautiful and alike the great Āgama
temples in Sri Lanka.
In size and visibility as well as in organisational structure, profile and per-
sonnel, the Kamadchi temple differs from other (Tamil) Hindu temples in
Germany. It is still a big exception, as most of the temples do not owe this
name in the proper sense of the term. Many are ‘invisible,’ being located in
cellars, workshops, and industrial buildings. In the past years, however, there
was great strive to signal public presence and augment, enlarge and beautify
314 Wilke
the places of worship, even though most Tamil Hindus work in very poorly
paid jobs. The Kamadchi temple of Hamm-Uentrop was the most successful
regarding this common wish and strife. It managed to become the first and
foremost in visibility, public representation and perception. The Kamadchi
temple is the most striking example of a newer phase of temple building that
involves traditional architectural elements. The richly embellished vimāna and
gopura towers can be seen already from the highway and the temple festivals
are performed on each of the fourteen days (nowadays even sixteen days) with
great splendor.
The present Kamadchi temple is already the fourth version. It started hum-
bly in 1989 in a basement in the city area of Hamm. Already in 1993–94, how-
ever, it attained transregional significance among the Tamil Hindus for being
the first temple in Germany to establish the great cart procession (tēr) at the
yearly temple festivals. In 1996, the temple was closed down (in part due to
neighbourhood problems during procession that attracted already 4000 par-
ticipants from near and far), but in 1997 it re-opened in a newly built house
in the industrial area of Hamm-Uentrop where the processions can take
place more splendidly and be augmented by a festival market. Since 2001, the
year when the first large report appeared in the popular German magazine
Stern, 10.000–20.000 participants have been regularly present at the major
processional day (tēr). The years following consecration showed a constant
augmentation—Sri Lankan Tamils coming even from neighbouring countries,
but also increasingly native Germans. This is but one example of the transfigu-
rations and token of success since 2002.
So far, the Kamadchi temple of Hamm-Uentrop remains unrivalled. Its
amazing success has more than one reasons (starting with the scarcity of
Brahmin priests), but the most decisive reasons are to my assessment two
lucky circumstances: a) a highly initiative and zealous priest who acts in one
person as chief priest, guru, manager, president of the temple committee, and
chairman of the German advisory board, and b) the German board itself, sup-
porting him as well as checking him (if ambitious dreams flew too high). Not
many temples have a German advisory board, in fact I do not know of any
other one. Such a board is an invaluable social capital and diminishes pos-
sible conflicts. It brings in the know-how about the German situation and the
language skills which otherwise are often lacking. The German advisory board
of the Kamadchi temple includes influential citizens of Hamm. They help to
negotiate with the authorities and banks. The German board members had—
besides priest Sri Paskaran Kurukkal—a major share in the temple’s rise and
successful exchange and cooperation with the majority culture. They devote
Āgamic Temple Consecration Transnationalized 315
quite a bit of leisure time and honorary work, including active help and coor-
dination in the yearly temple festivals and guided tours for German groups vis-
iting the temple. No doubt, they were also instrumental in making everything
go smoothly around the temple’s erection and consecration rites. I suppose, it
was mainly them who cared to keep the German locals informed and included
in the process, e.g. by the signboard ‘Hier entsteht ein Hindu-Tempel’ (‘A Hindu
temple is being built here’). It is a remarkable—and for Sri Paskaran a very
happy coincidence—that the German board includes the architect, a (former)
member of the city council and the representative of the insurance company
who is responsible for the temple. They became not only members of the advi-
sory board, but also deeply devoted to the temple and its chief priest.
It is enough to mention as an example of happy interaction the good coop-
eration of priest and architect. The architect was found by the priest in an ora-
cle-like fashion (by randomly opening the phone-book) and it was the start of
a special friendship. The architect immediately showed interest in the project,
and Sri Paskaran took him to his first trip to India to show the architect the
big South Indian temples in Tamilnadu. The erection of the Kamadchi temple
became a joint venture between the German architect who built the raw struc-
ture, and the South Indian sthāpati M.G. Nagaraj and his team of artisans, who
spent nearly two years in Germany creating the surface of the temple towers,
and building and painting the interior shrines.
On July 7, 2002, the eight days of consecration reached their peak and pub-
lic conclusion: It was the day of mahākumbhābhiṣeka, the sprinkling of the
temple towers, shrines, and images with the ‘heavenly’ holy water of the
jars representing the different temple deities, that had been prepared and
charged with sacred power by means of fire sacrifices and mantra chants
during the previous days. Since Vedic times, ablution (abhiṣeka) has been
a ritual of bestowing prestige, status, splendour, and cosmological power,
and retained such essential functions in the later installation/consecration/
inauguration rites of images and temples (pratiṣṭhā), aiming at establishing
divine embodiment.8 The Kamadchi temple, i.e. Sri Paskaran, clearly regarded
8 Tsuchiyama 2005: 51–56, 64, 93. The textual tradition of pratiṣṭhā rites attributes great merits
and achievements to one who accomplishes an installation. The old Raurāgama, for instance
316 Wilke
the great ablution (abhiṣeka) on the final eighth day as the central event of
consecration. It was performed—like all rites of the previous days—by a large
number of priests (eleven had come from South Asia only for the function)
and assisted by the temple’s sponsors in the presence of many honorary guests,
much press and a large crowd of several thousand participants from all over
Germany and Europe, including even a few guests from overseas. Indeed, the
ceremony infused the temple with prestige and fame, i.e. sanctioned symboli-
cally its unique status. As already mentioned, it was so far the largest temple
in continental Europe, and the first one in the style of sacral architecture and
consecrated by the traditional rules of the medieval Śaiva Āgamas. In an indus-
trial no-man’s-land a holy place was created that would soon spread its fame in
Germany even among non-Hindus and far across the German borders among
Sri Lankan Tamil refugees.
As István Keul points out in this volume, consecration or pratiṣṭhā is not one
discernable discrete act, but a series of entangled ritual elements. It is more
the concerted ensemble than a single ritual which makes up consecration.
This cluster contains many symbolically powerful ritual elements. As scholar
of religion one may take, for instance, prāṇapratiṣṭhā as most crucial: the infu-
sion of vital air or breath into the image, whereby the stone transmutes into
a mūrti, a living icon full of sacred power, a real embodiment of the divinity.9
This corresponds to the major function of pratiṣṭhā to divinize the material
stone—be it image or temple, and give the immaterial ideas of the divine a
material shape. This process necessitates various acts of divinization, among
which abhiṣeka, prāṇapratiṣṭhā, and adhivāsa (preparatory rites in a ritual
tent) take a prominent, but not unique place.
My case study in Germany is not an exception to these observations, not-
withstanding the importance given to mahākumbhābhiṣeka on the eighth day.
This day was advertised as major one. It witnessed the biggest crowd and larg-
est presence of press and television—who in fact had a major share in creat-
ing the temple’s fame by spreading the news nationwide in big newspapers
states that the performer and his relatives obtain up to 21 generations liberation in Śiva’s
heaven (Śivaloka) (Hikita 2005: 151). Other texts, including Vaiṣṇava, promise fulfilment of all
desires (kṛtkṛtya), worldly happiness (bhukti), final liberation (mukti) and even supernatural
capacities (siddhi) (Hikita 2005: 188; see also 166). This is often also connected with mantra-
sādhana, daily pūjā and japa, and selflessness (Takashima 2005: 132, 134).
9 Historically, prāṇapratiṣṭhā does not seem to be as central. Takashima (2005: 116–125) men-
tions adhivāsa (the mantra and homa practices in the sacrifical hut preparing the water jars)
and abhiṣeka (the ritual bath) as most important elements of pratiṣṭhā rituals in the ancient
Śaiva Āgamas (ibid. 125). In the oldest Vaiṣṇava Pañcarātra-Saṃhitās prāṇapratiṣṭhā appears
sometimes and sometimes not (Hikita 2005: 173, 184).
Āgamic Temple Consecration Transnationalized 317
and national tv-channels, and even beyond the German borders. However,
important rituals took place during all the eight days (June 30 to July 7, 2012).
Moreover, the consecration rites did not end on July 7. All the eleven priests
from South Asia, whom Sri Paskaran Kurukkal had invited (ten from Sri Lanka
and one from South India) to perform the consecration rites along with him and
his brother-in-law Arikaraputhira Mathivani Iya, stayed on. They continued
for another 45 days (or a ‘maṇḍala’ of 48 days) after the mahākumbhābhiṣeka
to perform special pūjās, mantrajapa, and collective arcanas with the Lalitā-
Sahasranāma, and the Lalitā-Triśati to imbibe the new image and shrine with
enduring sacred power, and energize the scented water in 1008 decorated
shells. The final act was again an abhiṣeka ritual: Sri Kamadchi’s bath with the
sanctified water from 1008 shells on August 21, 2012.
Furthermore, pratiṣṭhā or consecration had started, strictly speaking, already
three years back with the traditional vāstupūjā and the ceremony of setting the
foundation stone on May 10, 1999. This was supplemented one year later by
the non-traditional topping-out ceremony on August 25, 2000. Both auspicious
dates had been astrologically decided and the rituals involved the śrīcakra as
major element. This is remarkable, as the topping-out festival (‘Richtfest’) is a
German custom not known in South Asia, whereas the laying of the founda-
tion stone is a special event in both cultures (in German: ‘Grundsteinlegung’).
I insert an account of these preliminary rites and the new elements which can
be discerned in the diaspora context, before coming back to a more detailed
description of the eight days ceremony in 2002.
10 Discussed in Einoo 2005: 31–32, For the following observations see also Luchesi 2003:
231–233 and Wilke 2003: 142–144, who offer a more elaborated account.
318 Wilke
11 Takashima 2005: 124 describes this ritual only in regard to image installation in the
already existing garbhagṛha (based on the ancient Rauravāgama). Setting the founda-
tion stone of the pīṭha (garbhagṛha and pedestial on which the image is placed) includes
the deposition of nine kinds of jewels, medicinal substances, and seeds. An equivalent
ritual occurred in Hamm-Uentrop in 2002, but included in addition again the śrīyantra
(see below).
Āgamic Temple Consecration Transnationalized 319
There are not many Brahmins or priestly specialists, much less professionals
in temple building, in Germany and Switzerland. Typically, in Trimbach priests
from England and other places were invited for the temple consecration. In
Hamm-Uentrop, Sri Paskaran invited eleven priests—one from India and ten
from Sri Lanka—to perform the ceremony. Part of the Sri Lankan priests were
relatives, four were Brahmins, and only one had been in Germany before—Sri
Paskaran’s learned ‘uncle’ Sri Prabuhudevan, a specialist in big ceremonies and
temple festivals.
Temple consecration in diaspora contexts poses a novel situation and chal-
lenge for both the new society and the Hindu priests. The rites may turn into an
open ground for creative adaptions and negotiations. However, an even more
common pattern is minute adherence to traditional regulations to ensure
authenticity. Both were happening in Hamm-Uentrop. The consecration rites
of 2002 described below included all the important elements prescribed by
the Śaiva Āgamas, as far as I can see. They largely conformed to the synopsis of
major features summarized by Hiromichi Hikita (2005: 151) and other data in
the volume From Material to Deity: Indian Rituals of Consecration (Einoo and
Takashima 2005). This book was very helpful to me. As I am not a researcher
of Śilpaśāstra, but a scholar of religion and cultural anthropologist, I was
never quite sure whether the rituals I observed, were traditional, conflated, or
new. Takashima et al., on the other hand, used for their description a great
deal of ritual texts on pratiṣṭhā which developed from the 6th to the 12th and
later centuries to its present form.12 Despite historical changes, it is remark-
able that many features were kept alive through the centuries and nowadays
even migrated into foreign lands. The following common elements of several
ancient texts were also found in Germany:
12 Most Āgamas which deal with pratiṣṭhā (of images) can be dated to the tenth or eleventh
century, and certainly developments occurred from the early Āgamas to the complex ritu-
als of post-twelfth century. It is well known that early texts do not yet include temple con-
secration. They are restricted to image consecration intended for the individual Sādhaka
(seeker) who pursues personal accomplishment, whereas the later temple installation
and worship is primarily for the public. Major difference of earlier and later times can be
explained by the different users and growing complexity. For these notes and my ‘synop-
sis’ below see the articles of Jun Takashima (115–141) and Hiromichi Hikita (143–197) in
Einoo and Takashima 2005, particularly 115, 117–118, 121–127, and 151, 153, 158–184, 191–194
Takashima deals almost exclusively with image pratiṣṭhā and Hikita with texts of the
Vaiṣṇavite tradition. There are many common features which reappeared in Germany.
Āgamic Temple Consecration Transnationalized 321
(maṇḍapa [also called yāgaśāla]) and the necessary rituals such as ablu-
tion of the image, the fire-offering to be performed therein . . . [After ablu-
tion] the image is laid horizontally on a bed to sleep over night. These
phases are all referred to by them as adhivāsa or adhivāsanam. Next
morning, the image is raised and carried in a car to the sanctum sanc-
torum in the temple. It is then raised on the pedestal thereafter jewels
have been put into its cavity. Lastly, the priests beginning with the main
worshipper are fed and given the honorarium for the ceremony. (Hikita
2005: 151)
The ritual texts also show peculiarities, and partially a different order of cer-
tain elements, e.g. concerning the pradakṣiṇa of the fully decorated image in
a car around the temple and the village, ablution with water and grains, and
putting the image to sleep (performed in Germany on the fifth day in the order
mentioned). Some texts have additional features, such as the eye-opening cer-
emony (tracing the sculpted lines), which was also witnessed in Germany, and
likewise the setting of the water jars in the special shed (in Germany a big tent
with the traditional four entrances). The setting of water jars (representing the
temple deities, the largest of them Sri Kamadchi) was a very important act
in Hamm-Uentrop, as the ritually empowered water would be needed for the
different consecrating ablutions (the abhiṣeka of the Kamadchi image before
the infusion of vital air, prāṇapratiṣṭhā, and the great ablution of the temple
towers, the cupolas of the interior shrines, and all the temple deities on the last
day). As already known in older texts, the empowering acts in the ritual tent
were invoking in the jars the deity’s dynamic power by mantras and repeated
fire-sacrifices (homa). There was a mixture of tantric and Vedic mantras,
such as the five brahmamantras. The different homas were being ended with
pūrṇāhuti. They were performed on fire-altars (agnikuṇḍa) in various shapes13
(square, half moon, circular, triangular, lotus shaped, pentagonal, heptagonal)
spread in different directions of the compass (including the priests heading
at different sides). In the centre of the tent was the main altar of the god-
dess surrounded by eight fire pits (agnikuṇḍa) and a sub-altar. This altar and
Kamadchi’s jar on it were bigger than the additional surrounding nine altars
(platforms with jars) in front of which stood a singular agnikuṇḍa. All of this
constituted very traditional features—including bringing a cow to the cere-
monial site (the empty garbhagṛha, and the ritual shed, where she remained
constantly present).
13 There seemed to me all or almost all variants present which are discussed by Einoo
2005: 23.
322 Wilke
The following account of the ritual sequence observed during the eight
major days (June 30–July 7, 2002) includes also important ritual elements not
found in Einoo and Takashima’s (2005) volume, such as the fixing of copper
vessels (kalaśa) on the temple roof, the insertion of yantras into the pedestals
on which the images were placed (being part of setting the foundation stone
in the pīṭhas of the individual images), and the oil-bath of the deities by lay
worshippers. As far as I can see, none of these extra features is modern or dias-
pora-specific, but rather established by a long and later Śilpaśāstra tradition.
However, better experts in Śilpaśāstra may possibly discern transformations or
elements which slipped my attention or where my account has failed.
14 My description is partly based on my own participant observation (first, and final
four days), the description given by Luchesi 2003: 239–244, and pictures taken by the
photographer R. Bruse. I was not able to note or record the mantras used during the fire
rituals.
Āgamic Temple Consecration Transnationalized 323
pots sealed with coconuts and mango leaves, indicating the presiding deities
of the directions, were placed. A long ceremony followed, including exten-
sive mantra recital by several priests and vāstuhoma. Leaving the altar, all the
eight directions around the temple building were sanctioned by a sequence
of individual sacrifices which involved the cutting of a watermelon and the
breaking of a coconut on which red vermillion was applied. Each sacrifice was
performed in front of a movable trident representing the goddess.
Meanwhile, the priest from South India prepared in the old temple across
the street the new yantras (which would in the following days be plastered in
the pedestal of the image), charging them with sacred power by mantras and
hand gestures. As also the old old yantras (removed from the base of the old
temple images) were laid out in front of him, I suppose the ritual included the
transference of power from the old into the new (śrī)yantras and visarjan of
the old.
The chief priest Sri Paskaran summarized as major action of the first day the
protection of the temple and temple grounds, the city of Hamm and the whole
country, as the rites warded off danger and inimical forces.
15 Cf. Hikita 2005: 191–194. The ceremony is known by several names. The following descrip-
tion of the ceremony at Hamm-Uentrop by Luchesi 2003: 239 mentions only the functions
of the artisan (śilpin), whereas Hikita’s sources give the worshiper (preceptor or adept)—
which would be Sri Paskaran—the major role and distinguish his acts strictly from those
of the artisan.
324 Wilke
and brows, so that indeed one could get the impression that the previously
‘blind’ image looked now as though awake and alive. He related that the first
look of Kamadchi should fall on him who was her ‘mother,’ the second look
on Sri Paskaran, whereas the assembly of lay worshippers had to wait until
another ritual was over. This ritual, called daśadarśana, was aiming at the
transformation of the goddess’ powerful fierce gaze into a mild and acces-
sible one by distributing the gaze on different objects and beings placed in
front of her: a young man, a young woman, a priest, a cow, camphor light, a
water vessel, a mirror, etc. Thereafter, all the worshippers could see the goddess
without harm.
The very same day the goddess was taken for a ride on a car. The (traditional)
processional route (pradakṣiṇa) around the village was realized by moving the
fully decorated, heavy image on a truck halfway to the city of Hamm and back
(from the industrial area and village of Hamm-Uentrop to the Maximilian Park,
Hamm, a large shopping centre at the periphery of the city). The programme-
brochure explained this was done, so that the goddess could get to know the
surroundings, the people and the city, and vice versa could the German neigh-
bourhood get familiar with her. This nice idea did not really work out. As it was
raining, the procession remained a lonely affair of the goddess, some priests
and a few lay worshippers. But clearly it was seen as an interactional ritual.
Returning to the temple grounds the sun came out—just in time for the
extensive ritual bath (abhiṣeka) of the undressed (i.e. with a loose cloth barely
dressed) image in front of the temple. This copious ablution lasted several
hours and involved many substances and actors. The chief priest and precep-
tor Sri Paskaran would always start pouring extensively water, thereafter grains
and other substances over the icon. Then came his family, the other priests,
the sponsors, the artisans and some German guests pouring water, grains, etc.
over the Kamadchi icon. After the bath the image was wrapped and laid down
horizontally to sleep inside the new temple near the entrance, along with the
other divinities under the same big bedspread.
bound with white strings, and sealed and decorated with coconuts, mango
leaves and colourful brocade cloth were placed as representations of the
shrines and temple deities. The spatial placement of the altars and pots cor-
responded to the placement of the new shrines in the temple, with Kamadchi
Devi in the center—represented by the largest altar and the largest kumbha.
In front of each altar a fire-pit in a different shape had been established and
around the altar were further pots with scented water. The priests began with
self-purification (nyāsa, japa and meditation) and worship of the ‘doorways’
and quarters around the altars. Sitting down they recited mantras to invoke
the deities and performed long hours of fire sacrifice (homa) [Fig. 1], ending
with nyāsa of the pots/deities and pūrṇāhuti. The lightening of the fire(s)
was achieved by the sunlight caught in a mirror. Mantra recital and homas
had the objective to transfer divine presence and power into the scented
water-pots.
Very late at night the images were brought into their shrines in the new tem-
ple and fixed on their pedestals by the artisans with a special paste that had
been carefully prepared by married women. Before fixing took place (starting
with Gaṇeśa), the priests had prepared the pedestals, i.e. installed the founda-
tion stone at each of the pīṭhas by depositing the deity’s sanctified yantra and
326 Wilke
nine kinds of jewels (navaratna)16 in the cavity of the pedestal. On this base
the images were fixed. Those images which had already been in the old temple
(the Navagraha) got first a good cleaning with a lot of soap.
Most difficult was the transport of the very heavy new icon of Kamadchi
up the stairs of her sanctum. It needed ten men and a lifting gear to get her
there und place and fix her on the pedestal. All were relieved that everything
went well.
16 Luchesi 2003: 242. The same action is described already in the Rauravāgama for the liṅga,
but without mentioning a yantra, and instead medicinal substances and seeds in addi-
tion to the ‘nine jewels.’ Cf. Hikita 2005: 124. Another early Āgama mentions eight kinds of
jewels, various minerals, grains, herbs, weapons, and a tortoise made of gold in the centre
(Hikita 2005: 131). I noticed at the re-consecration 2014 medical herbs and other items
being included.
Āgamic Temple Consecration Transnationalized 327
Figure 3 Transference of sacred energy by the priests touching a cord from the yajñaśāla to
the newly built temple and the ‘womb chamber’ of the goddess Kamadchi.
photo July 7, 2002, copyright A. Wilke.
Āgamic Temple Consecration Transnationalized 329
It was already two or three o’clock in the morning, when the magic eve-
ning reached its climax: prāṇapratiṣṭhā, ‘the establishment of breath’ in
the image of Kamadchi. Maybe more than any other ritual action, this item
(infusing vital air or the root-mantra into the image) is supposed to transform
a statue into a mūrti, embodying divinity in quite a real way. The infusion
of life-breath into the image of Kamadchi was performed by Sri Paskaran in
secret under a golden and orange cloth, while the doors of the shrine were
kept open [Fig. 4]. So this very intimate encounter of priest and goddess was at
the same time a public event, staged in the presence of all the priests and the
little crowd who had stayed on for this highly auspicious and thrilling moment.
When Sri Paskaran reappeared removing the cover and slowly walked towards
the temple entrance in a very still and contemplative mood, many lay worship-
ers would fully prostrate at his feet (in daṇḍa-like fashion) and treat the priest
himself as a holy person or divinity. Sri Paskaran declared in an interview that
prāṇapratiṣṭhā is a very special and exceptional thing and that a person is sup-
posed to perform it only once in his lifetime.
Figure 4 Prāṇapratiṣṭhā, the ‘establishment of breath,’ in the new icon of Kamadchi Devi by
the head priest Sri Paskaran Kurukkal.
photo July 7, 2002, copyright A. Wilke.
Āgamic Temple Consecration Transnationalized 331
re-opening and the ‘vision’ (darśan) of the beautifully decorated and richly
dressed deities, most splendid of all Kamadchi Devi. A long solemn pūjā cere-
mony with 64 offerings (upacāra) and the waving of many lamps was following.
Besides these traditional religious actions, performed by the priestly spe-
cialists and in part also the Sri Lankan Tamil sponsors, some of the festive inci-
dents of the day also encompassed what I have called the ritualization of the
social bonds with the majority society. This included a speech by Sri Paskaran
and the mayor of Hamm, and in the evening a Bharatnāṭya dance program with
a Sri Lankan and a German dancer. All the way through the eight days there
were clever openings toward the majority society, starting with the first day’s
rites of protection for the city of Hamm and the German nation. Similarly, at
the yearly temple festival through the years the motto has been ‘May all people
of the world be happy.’
Except for two researchers who were also present during several days of con-
secration (Brigitte Luchesi and myself), very few non-Indians would visit the
temple and festivals before 2002. However, one could notice a sudden change
almost right from the final day of consecration onwards. Already the previ-
ous temple festival in 2001 had attracted a photographer and journalist from
the Stern. The report about the festival and the erection of an edifice with the
traditional looks of Tamil Śaiva temple, was exceedingly positive, and included
beautiful colour pictures. At the consecration in 2002, a large body of press and
TV was present at the last day of mahākumhābhiṣeka. It was good public rela-
tion that spread the news nationwide, and eventually even worldwide.
3 Developments of Glocalization
The consecration rites were the very start of various new developments,
including ongoing beautification of the interior and exterior of the temple and
the temple grounds. Purity rules were now communicated explicitly. A leaf-
let informed how to behave in a temple. Among other things menstruating
women are asked to refrain from entering it—a purity rule followed thereafter
even by some native German women. A clear strategy was to render the temple
festivals each year more beautiful and make them resemble high-class Śaiva-
Siddhānta Āgama Hinduism back home. In the following notes the social
transformations will be of primary interest.
Germany is not a classical country of South Asian diaspora. The advent of
‘ethnic Hinduism’ is a novelty, as already indicated, and the Kamadchi Temple
played a major role in public representation and perception. It was a story
332 Wilke
the Kamadchi temple is credited to be the most decisive factor and dynamic
force in this endeavour and in having positive effects on the integration of the
high-rate migrant population of the city. The temple is often used as a podium
of exchange—from occasional interreligious functions to, almost daily, guided
tours. Whole busloads of visitors are nowadays no more rare. They bring
German school classes, senior citizens, political parties etc., for whom it is
often the first time to visit a Hindu temple. On weekends, frequently as many
native German visitors as native Sri Lankan Tamil can be discerned. Quite a
few Germans come more or less regularly. Some signal by their praying ges-
tures that they are more than tourists. Indeed, interviews conducted have
shown that part of the German visitors do not see themselves as outsiders, but
rather insiders. Temple and festivals also attract spiritual seekers who deliber-
ately want to cross religious boundaries.
A big event—both for the temple and the city of Hamm—are the big-style
chariot processions (tēr) that attract up to 20.000 participants from Germany
and all over Europe. The sensory aesthetics of the temple and its festivals have
not only been amplifiers of Tamil home culture, but also agencies of its public
display and motors of cultural and social change. Some fifteen years ago, the
processions—then within residential quarters of the city of Hamm—aroused
public dissent. Nowadays, they are greatly enjoyed by Tamil migrants as well
as members of the majority society. On the Tēr-day host and guest society are
inverted, and over the past fourteen years the Tēr-day has become an impor-
tant meeting-point of the trans-nationally spread Tamil community and its
young generation. It is not only a religious event, but a social one as well.
One of the most striking developments is that the temple is not anymore
only an interethnic space, as it used to be in its previous versions, but also
an intercultural one since the time of consecration. Due to the media reports,
which made it known nationwide, it has attracted an impressive number of
native Germans since 2002. As already mentioned it is nowadays not rare to
come across whole buses of tourists and school classes who want to know
more about Hinduism. Since 2014 there was hardly a day without at least one
guided tour (offered by a member of the German advisory board). Certainly,
most German guests come out of sheer curiosity, educational interest and also
in search of an exotic experience, or they appreciate finding India in their own
neighborhood. But some identify with the religious aspect and an amazingly
growing number started to participate in the rituals (pūjā and arcana) like
faithful Hindus.
Even a ‘German parish’ started to develop. Most members stem from Hamm,
but a few joined from other places. It is socially a very mixed loose network,
encompassing some distinguished citizens of Hamm, e.g. the architect and
insurance agent of the temple, but also unemployed contributing free labour
334 Wilke
to the temple, and a number of spiritual seekers and India lovers. They are
not converts, but more the type of spiritual nomad who lives a postmodern
religious life, combining different religious cultures. This German parish or
network has good relations to the Hindu ritual community, without however
mingling with it. The Sri Lankans can be called a festival community, whereas
the ‘German parish’ forms a spiritual communitas.
17 The first citation is taken from a sample of unpublished interviews conducted in 2008 by
the ethnographers Pablo Holwitt und Thomas John, Munster University; the second state-
ment (heard at Divali 2005) was raised by my own field-work. I heard a similar statement
from a lady in 2013, a manager, who had recently moved to Hamm and was giving guided
Āgamic Temple Consecration Transnationalized 335
can account for a temple deserving this designation, an edifice imbibed with
divine presence, and for the German natives the Kamadchi temple became
the major representation of Hinduism, precisely due to being built in tradi-
tional architectural style. The temple’s role as public representative was put
into action right at the moment when the media reported nationwide about
the copious consecration rites.
A point of interest is that the chief priest has meanwhile gained the reputa-
tion of a holy man and miracle-worker. A great number of Sri Lankans (and
also some native Germans) perceive him as someone very special, someone
who must be holy and full of the goddess’ powers, because he achieved such
a fabulous temple. In 2006, a group of Tamil Hindus from Krefeld, Essen, and
Nuremberg conferred the honourable title ‘pirathistha’ on him, which was
explained to mean an outstanding person and religious leader who made the
impossible possible.
It is noteworthy that a section of both the native German and the native
Hindu temple visitors see the temple and the processions as imbibed with a
particularly strong spiritual power and energy. Some regard the founder and
chief priest even as their personal guru. The priest has his own strategies
to stimulate and maintain this sacred aura, but no doubt, his authority and
charisma is not primarily due to self-ascription, but rather due to the attri-
bution of such qualities by others. Strikingly, a special weekend of Vedānta
teaching at the temple in 2013, organised by a German yoga group, announced
the ‘Shankaracharya Sri Paskaran’ being the preceptor, and thus attributed a
highly orthodox title of an elite of Brahmin monks in postmodern fashion to
the priest of the German temple. The event was mainly visited by Germans.
While these observations relate to lived and practised religion, some of
which is only applicable to a tiny minority of the German population, the
temple (community) was also successful in the collective and political sphere
regarding representation, recognition, and legal participation in equal reli-
gious rights. The temple community (helped by a Jewish lawyer) filed in 2005
an application to gain the status of a body of public law (‘Körperschaft des
öffentlichen Rechts’). The motion was lost at first, but the temple’s lawyer
raised a protest against the state of North-Rhine/Westphalia who had denied
the status. It was the first instance in Germany that a religious community
would go as far, and it was successful. Indeed, in 2013 the lawsuit against the
state was won and the cherished recognition as an institution of public law was
gained. This status means to be on equal footing with the two great Christian
tours in the temple. She says she had always wanted to go to India, but found now in the
temple Hamm-Uentrop all what she was looking for—‘phantastic vibrations.’
336 Wilke
churches—which was in most German states denied even to the (much larger
population of the) Muslims so far.18 Concretely it means to be able to raise
church taxes and offer religious education in public schools, etc.
So the temple of Hamm-Uentrop has achieved since its consecration a cer-
tain monopoly in representing Hinduism in Germany. This is not unchallenged,
however. Among the Tamil Hindu temples there is competition and rivalry, the
temples have different profiles and each seeks enlargement and more approxi-
mation to home tradition (Wilke 2003; 2006; 2013b). The Kamadchi temple is
nowadays not anymore the only one who received Āgamic consecration rites.
The same happened in Hannover 2009 and Trimbach (Switzerland) 2013; more-
over, part of the rites were performed in Muenster (Germany) 2012 and Berlin
2013. Besides careful observation to realize authentic reconstruction of home
culture, one can witness Āgamic tradition in transition. The Navacakti temple
of Muenster is an example of this: The measures of the goddess image do not
conform to traditional rules, the temple building is not traditional architec-
ture, but a former nursery, the priest is from a non-Brahmin lower caste-group
and received the help of ‘big brother’ Sri Paskaran.
A further interesting development for which the imposing Kamadchi tem-
ple was likely the major trendsetter and motor, was a sudden more visible and
explicit pluralization and ethnicization of Hinduism in Germany, visible in
a number of new temple projects after 2002. The Afghani Hindus publically
appealed to sponsor a new Hindu cultural centre in Köln and opened a new
temple in the city in 2004. The Indian Hindus started an Āgamic temple project
in Berlin (foundation stone ceremony in 2007) in addition to the existing Tamil
Hindu place of worship in the city—a Murugan temple, one of the oldest in
Germany, and critical towards Hamm-Uentrop and its priest. The Balinese
Hindus erected a small traditional style pagoda temple in Hamburg in 2010
and the Balinese founder explained that Hamm-Uentrop was not a place to
go for her, as it was so different from Balinese Hinduism. Bielefeld, too, wit-
nessed in 2013 a second temple in the city in addition to the Tamil Hindu one.
It was an Indian Krishna temple and the priest announced to the press that
18 Even for the Hindu temple, it may not be the last word, as the state might also make
use of its right of veto. However, juridical knowers of the situation suspect that the final
decision will be in favour of the Hindu community, as recently also the Bahá’i and the
Alevites had been granted the status of a body of public law. It was expected that also
Muslims would finally be recognized. In principal, other Hindu temple communities can
also apply when they fulfil certain injunctions like public representation, active religious
life, duration, and a substantial number of memberships. The latter was downplayed in
case of the Kamadchi temple due to its national and international outreach.
Āgamic Temple Consecration Transnationalized 337
yoga and Indian cooking classes would be offered soon. The Kamadchi temple
reacted with its own ambitious new expansion plans. Since 2009 it announced
an ‘International Hindu Cultural Centre,’ initially scheduled for 2014, but fun-
draising is still going on in 2016. The plan is to include a large marriage hall, a
library, a museum, and classes in language (Tamil and Sanskrit), music, dance,
cooking, meditation, yoga, and Hindu religion. This cultural centre intends to
address two major groups of receivers, having as objective (a) to guarantee the
transmission of Hindu tradition to the young generation born in Germany, and
b) to foster interreligious dialogue and knowledge about Indian culture among
the Germans.
19 I counted at least 30 kāvaṭi dancers, and 30 rollers, and among the ladies 28 who carried
burning camphor-pots and 25 milk-pots.
20 That the year 2014 was less frequented had possibly to do with the re-consecration cer-
emony around two months earlier and the constant special rituals taking place thereafter.
A board member told me that many people came spread over the two months, and prob-
ably therefore abstained from visiting again the temple festival.
340 Wilke
impediment was that the extension of the visa for the Indian temple artisans
was denied, so that only one temple tower could be colourfully painted for
the re-consecration. The temple festival in 2015 witnessed both the towers in
colour, but not the building—let alone opening—of the cultural centre (as
announced in 2014). It is not very likely that it can be realized in near future,
although a professionalization in soliciting for sponsorships and loans was
detectable in 2016.
Also in terms of status and reputation problems linger on. The story of suc-
cess of priest and temple did not go unchallenged. It is true, that at present,
the Kamadchi temple and/or its leading priest managed to access the status of
an official representative of Hinduism in Germany, or at least in North-Rhine/
Westphalia. This representative status was attested even by the Sri Lankan
embassy—which in Tamil eyes is of course not a thing to boast about. The
priest of the Kamadchi temple got criticized for his political abstinence and
disinterest, his reluctance to sympathize with and support the LTTE (Liberation
Tigers of Tamil Eelam) and aspirations of national autonomy, and worst of all,
he reaped grudge and enmity for having performed a pūjā ceremony in the Sri
Lankan embassy—a slap in the face of Tamil freedom fighters.
So the priest is on the one hand much admired, and on the other also harshly
criticized. He is not at all everybody’s darling, and possibly a streak of jealousy
and competition explain as much certain ambivalent feelings towards him
as do resentment and disappointment. Caste and attitude are debated sub-
jects. The non-Brahmanic status of the main priest of the Kamadchi temple
remained a touchy issue. Sri Paskaran’s non-Brahmin (yet priestly) background
is seen as not befitting to the major priest of a (prestigious) Āgamic temple.
There is a feeling among some that a temple consecrated according to Āgama
rules must be served by a Brahmin (an Ayer or Śarma). Besides caste, lacking
education, too much interest in money, name and fame, etc., have been addi-
tional grounds for critique. Some think the temple is only good for socializing
and big events, whereas quietude is only found and personal devotional prayer
better performed in other places, smaller places of worship, or even Christian
churches.
There is also a certain tiredness about the constant expectation of sponsor-
ing, and some reluctance regarding the amount of ārcana prices deemed to be
too high. Most are not aware about the great financial pressure of a temple in
this size, and suspect rather riches than void treasure boxes and empty bank
accounts.
All the positive and critical features show clearly that not only the temple
became famous, but also the priest became an influential public person who
is surrounded—as often in the case of ‘very important people’—with some
ambivalence.
Āgamic Temple Consecration Transnationalized 341
4 Re-Consecration 2014
festival,21 the re-consecration rites in a more narrow sense (mirroring the ones
in 2002) lasted eight days—from April 27 to May 4, 2014.22 The last day was
reserved again for kumbha-abhiṣeka and advertised as the most important one
(‘Haupt-Einweihungszeremonie’). Like in 2002, the head priest Somashkandak
Kurukkal from Canada/Australia was present, but not taking part in the
rituals. Moreover, two monks, Swami Brahmananda (a Śrīvidyā guru from
the Bhuvanesvara-Pitha, Tamilnadu, who now lives in the US)23 and Swami
Vishvananda (from the Babaji- and Yogananda lineage whose Ashram is in the
Taunus, near Frankfurt),24 were present, supposedly to enhance the sacred-
ness of the occasion. Swami Vishvananda was even advertised on a large poster
at the entrance, and he had brought his group of devotees to offer bhajans to
the Goddess after the mahākumbha ablutions were over.
Despite the fact that more priests than in 2002 were present, there was a
smaller crowd of lay-persons—except maybe for the oil-bath. This possibly
had to do with the spread of so-called inauguration-rituals throughout a period
of more than two months (27.04.–06.07.2014), which produced a dispersion of
visitors instead of a concentration within one or a few days. The re-consecra-
tion in the more narrow sense was badly advertised in barely a few catchwords
on a signboard and in the last minute on the Internet. Mahākumbhābhiṣeka of
the temple towers was already over at 9 a.m., which means a big crowd from
outside of Hamm was not to be expected. Although all the rituals in the tem-
ple and the yajñaśāla were aesthetically as luxurious as in 2002, everything
was less spectacular and joyous—an impression maybe having to do with the
reduced crowd and the uncosy, cold and rainy weather. The priests suffered a
lot from the cold, I was told, and during three nights there was a heavy thun-
derstorm that caused the backside of the temple stand in water, so that circum-
ambulation around it became difficult. Below I sketch a summary of the ritual
sequences and only go into more detail where I noticed change or gained more
insight about elements that were there already in 2002. Much of it mirrored the
initial consecration in 2002. As I was present only on 1, 3 and 4 May, these final
21 At least, it was publically announced this way on a sign-board in the temple. The board
made also known that from 22 June onwards there was daily a ‘special inauguration
ceremony’ (‘spezielle Einweihungszeremonie’).
22 I was also told that a whole year of yantra-pūjā would follow.
23 Unfortunately, I could not see and interview him, as he was not well and therefore not
present at the rituals, but I learned that he had already been present at the consecration
in 2002.
24 This Swami and his devotees were also present at Sri Paskaran’s 50th birthday on
September 13, 2013.
Āgamic Temple Consecration Transnationalized 343
days will be given preference. The announcement on the board marked them
also as major days. It mentions only:
Obviously the first four or five days (27.04.-01.05.14; Sunday to Thursday) were
meant for preparation, and mainly devoted to purification and protection rites,
and performances of renewal. There was of course no more eye-ceremony, but
most other elements were repeated. The pots in the yajñaśāla were set for wor-
shiping the deities in them by meditation, mantra rites and homa. Inside the
temple, the images were de-installed and re-installed to deposit in the pedes-
tal at their base new yantras and a paste of eight herbs, etc. This was done in
the night of the fourth and in the morning of the fifth day (May 1) and called
‘aṣṭabandana consecration.’ Moreover, the gods were awakened (āvāhana) in
the morning in the yajñaśāla, where again extensive fire-sacrifices (homa) took
place, for instance, on May 1, for several hours after 5 p.m. [Fig. 1].
I observed on this day and again when being present on May 3, that the
extensive ritual actions in the yajñaśāla included partially the (gṛhastha) spon-
sors of the shrines and were in many ways very close to descriptions based on
ancient Āgamas (cf. Takashima 2005: 122–125; 130). This resemblance included
also the rituals in the ‘womb chamber.’ In ethno-indological manner, i.e. the
combination of participating observation and the textual tradition, I traced on
May 1 and 3 similar patterns:
In the pots near the fire pits, the priests were worshiping the deities in the
yajñaśāla. This was partly a concerted action, partly several priests assisted
Sri Paskaran at Kamadchi’s altar (with the bigger, central pot representing
the chief goddess), two of them reciting mantras. There was performance
of nyāsa of the mūlamantra upon the pot of the chief goddess, followed by
dhyāna, homa, (100?) oblations with the mūlamantra, and at the end a homa
with the 50 letters of the alphabet. The priests passed long stretches of the
night with recitation, before performing pūrṇāhuti.—In the garbhagṛha,
25 The German announcement was explained in English: ‘Washing of Idols with sesame oil
before [sic!] the Devotees.’ It meant clearly ‘by’ the devotees who brought their own oil, or
purchased it in the temple shop.
344 Wilke
the installation of the foundation stone of the pīṭha was being renewed with
nine kinds of jewels, medicinal substances, and seeds.—The major fire sacri-
fice in the yajñaśāla was performed by the ‘preceptor’ Sri Paskaran and 16 of
the mūrtipas were Gṛhasthas (assisting and being assisted by the 16 priests).
Homas with the pañca-brahma-mantra und the mātṛkas were performed,
and each day japa and pūjā occurred.—For the transfer of energy (night
May 3–4, see below) the pots with scented water for ablution were placed
in front of the image in the garbhagṛha just as they were set before in the
sacrificial shed.
to Kamadchi’s image in the garbhagṛha, and three times the priests would
symbolically lead energy from the sacrifical shed to the image, keeping con-
stant contact with the cord with their ritual sticks and sacrificial spoons. First,
the cord was strung around the base of Kamadchi’s ‘naked’ image (I suppose
the yoni of the sitting goddess), second, around her trunk, i.e. her breasts,
and third, the cord was coiled up and placed on her head like a hat (the cord
resembled at this stage a yogi’s matted headdress). In a single action all the
16 priests would empty their spoons of ghee over the head of the goddess
and thereafter touch her body parts—yoni, breast, and face respectively—(in
nyāsa-fashion) with the ritual sticks. Thereafter, again in a concerted collective
action prāṇapratiṣṭhā was performed by the priests at once, i.e. unlike the first
consecration not only by Sri Paskaran and not under a cover [Fig. 5].
Figure 5 Prāṇapratiṣṭhā, the ‘establishment of breath,’ in the goddess Kamadchi by all the
priests at the ceremony of re-consecration.
photo May 4, 2014, copyright A. Wilke.
346 Wilke
My paper spans from the consecration rites in 2002 to their recent renewal in
2014 which aimed at a new charge of sacred energy, culminating in the Mahā-
Kumbha-Abhiṣeka on May 4, 2014, the ‘sacred bath’ of the temple towers, shrines
and images, after having transferred vital air into Kamadchi’s icon in the night
before. It was surprising to me that (unlike eye-opening) prāṇapratiṣṭhā had
also to be renewed. Besides a summary of the suggestive ritual actions, my
focus was on the years between consecration and reconsecration—the story
of success and present difficulties.
The past two decades witnessed an immense pluralization and ethniciza-
tion of Hinduism in Germany and the transnationalization of Āgamic culture
and temple consecration. The Kamadchi temple was a decisive driving force
to initiate and fuel these processes. Its feats of success can be traced locally,
nation-wide, and internationally. At the local level, the Kamadchi temple man-
aged to integrate into the social life of Hamm and gain enormous public rec-
ognition. At the national level, it attained a certain monopoly in representing
Hinduism in Germany and a church-like status as an institution of public law
(‘Körperschaft des öffentlichen Rechts’). Great effervescence was also gener-
ated at the international level, the temple being known nowadays as a new
European Hindu pilgrimage place even overseas. As I have shown, these three
aspects are entangled and make a good case of ‘glocalization.’
This final chapter aims at theorizing and a prognosis of future trajectories.
The ethnography presented leads me to some theoretical conclusions on the
transnationalization of ethnic Hinduism, aspects of change, and the role of
the aesthetics of sacred architecture and traditional rituals. However, the final
Āgamic Temple Consecration Transnationalized 347
remarks must also address the contingency of the rapid developments and an
uncertain future of the Kamadchi temple despite its great success in the past
14 years.
not that strictly orthodox, sometimes giving enthusiasts and charismatics new
chances, or bringing groups that were at the periphery in their homeland to
the centre in the diaspora. This situation is conducive for ritual change and
creating new professional elites. The charismatic Vīra-Śaiva Sri Paskaran,
for instance, made his way as ritual specialist, professional in temple build-
ing, spiritual teacher and guru full of sacred power, tapping from Āgamic and
non-Āgamic Hindu sources. Elements of novelty are also new transnational
networks of consecration professionals. More ambivalent are possible viola-
tions of Āgama rules as in the case of Muenster’s temple. Occasionally, even
completely newly created rituals may appear, like the topping-out ceremony
at Hamm-Uentrop. Religions are not static entities, but of fluid nature. Even
rituals, although being more stable than other aspects, are dynamic. Migration
processes lead to more rapid changes. The new situation poses challenges and
necessitates flexibility. It may lead to heavy configurations. At worst there can
be loss of tradition. But the new situation provides also new chances. It can
lead to upward mobility, for instance. The rapid rise of the Kamadchi temple
and the success of its zealous priest are an example.
present the temple towers in colourful paint. That only the vimāna could
‘make it,’ reveals some of the difficulties of Āgama culture going abroad.
References
Ballard, Roger. 1994. Desh Pradesh. The South Asian Presence in Britain. London.
Baumann, Martin, Brigitte Luchesi and Annette Wilke (eds). 2003. Tempel und Tamilen
in zweiter Heimat. Hindus aus Sri Lanka im deutschsprachigen und skandinavischen
Raum. Würzburg: Ergon.
350 Wilke
Baumann, Martin & Andreas Tunger-Zanetti. 2014. Der Hindutempel in Trimbach. Von
der Idee bis zur Einweihung. Lucerne: Zentrum für Religionsforschung.
Einoo, Shingo and Jun Takashima (eds) 2005. From Material to Deity. Indian Rituals of
Consecration. Delhi: Manohar.
Einoo, Shingo. 2005. ‘The Formation of Hindu Ritual.’ In From Material to Deity. Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. Delhi: Manohar, 7–49.
Hikita, Hiromichi. 2005. ‘Consecration of Divine Images in a Temple.’ In: From Material
to Deity. Indian Rituals of Consecration, eds Shingo Einoo and Jun Takashima. Delhi:
Manohar, 143–197.
Hobsbawm, Eric and Terence Ranger (eds) 1984/1992. The Invention of Tradition. Cam
bridge: Cambridge University Press.
Luchesi, Brigitte. 2003. ‘Hinduistische Sakralarchitektur und Tempelgestaltung in
Hamm Uentrop.’ In Tempel und Tamilen in zweiter Heimat. Hindus aus Sri Lanka
im deutschsprachigen und skandinavischen Raum, eds Baumann, Martin, Brigitte
Luchesi and Annette Wilke. Würzburg: Ergon, 223–274.
Marla, Sandhya. 2015. Diaspora-Religiosität im Generationenverlauf: Die zweite
Generation srilankisch-tamilischer Hindus in NRW. Würzburg: Ergon.
Takashima, Jun. 2005. ‚Pratiṣțhā in the Śaiva Āgamas.’ In From Material to Deity. Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. Delhi: Manohar,
115–141.
Tsuchiyama, Yasuhiro. 2005. ‘Abhiṣeka in the Vedic and post-Vedic Rituals.’ In
From Material to Deity. Indian Rituals of Consecration, eds Shingo Einoo and Jun
Takashima. Delhi: Manohar, 51–93.
Wilke, Annette. 2003. ‘ “Traditionenverdichtung” in der Diaspora: Hamm als Bühne der
Neuaushandlung von Hindu-Traditionen.’ In Tempel und Tamilen in zweiter Heimat.
Hindus aus Sri Lanka im deutschsprachigen und skandinavischen Raum, eds Martin
Baumann, Brigitte Luchesi and Annette Wilke. Würzburg: Ergon, 125–168.
Wilke, Annette. 2006. ‘Tamil Hindu Temple Life in Germany: Competing and
Complementary Modes in Reproducing Cultural Identity, Globalized Ethnicity
and Expansion of Religious Markets.’ In Religious Pluralism in the Diaspora, ed.
P. Pratap Kumar. Leiden: Brill, 235–268.
Wilke, Annette. 2013a. ‘Tamil Temple Festival Culture in Germany: A New Hindu
Pilgrimage Place.’ In South Asian Festivals on the Move, eds Ute Hüsken and Axel
Michaels. Wiesbaden: Harrassowitz, 369–395.
Wilke, Annette. 2013b. ‘Ein Hindu-Pilgerort in Westfalen. Zur gesellschaftspolitischen
und sozio-religiösen Bedeutung des Kamadchi-Tempel von Hamm-Uentrop als
neue kulturelle Kontaktzone.’ In: Religion und Politik im gegenwärtigen Asien.
Konvergenzen und Divergenzen, eds Edith Franke and Katja Triplett. Münster: Lit,
13–52.
CHAPTER 13
István Keul
The old woman sitting on the bench muttered quick sentences, her eyes
fixed on the image in the shrine. She spoke Bhojpuri, fast, with a sense of
urgency, and she had no teeth or dentures, which rendered her words
almost entirely incomprehensible to me. The woman did not seem to be
praying, nor was she reciting any of the popular hymns dedicated to the
simian deity. Initially I thought that she was talking to the shrine owner
who used to perform his elaborate morning worship sitting on the cool
marble tiles, half-hidden behind the shrine’s entrance. Passing by the old
woman I noticed that there was no one else there: her words were directed
to Hanumān. She spoke softly, beseechingly, and every now and then she
paused as if waiting for an answer, only to resume the imaginary conver-
sation after a few moments of listening. The image standing in the shady
depths of the small room had half-closed eyes. With the corners of the
mouth turned slightly upwards, the deity’s statue looked as if it was smil-
ingly encouraging this faithful devotee to continue speaking. Dressed in
color-coordinated, exquisite-looking silk clothing, Hanumān stood in a
relaxed tribhaṅga posture, leaning gracefully on a silver club. The shining
ornaments on his upper arms and ankles and his bejewelled crown
underlined his royal appearance. The old woman’s intimate conversation
partner seemed to exude an air of benevolent might.
This scene took place in 2014, on one of the southernmost ghāṭs of Vārāṇasī.
I had witnessed the ritual installation of this Hanumān statue almost two
decades earlier, and had been following the site’s history closely ever since.
From the moment the image was carried into its abode on a torrid July after-
noon in 1995, a small but constant stream of devotees had been visiting the
shrine regularly over the years. The performance of the three-day installa-
tion ritual had obviously generated the desired effects, transforming an artful
sandstone sculpture of a popular deity into a ‘living’ image worthy of wor-
ship. The consecration of temple images belongs—together with life-cycle
rituals, initiations and ordinations—to a category of complex South Asian
rituals where efficacy is not only highly relevant but also immediately pal-
pable: wedding rituals generate lasting unions; through initiations, initiands
become initiates into a certain lineage of doctrines and practices; ordinations
transform aspiring novices into fully fledged members of monastic commu-
nities. In a similarly straightforward manner, an image consecration is per-
formed in order to bring about a permanent, materially contained presence of
a deity. In the case of this small Hanumān temple there seems to have been no
doubt in the minds of the god’s worshippers that the procedure of infusing life
into the image (prāṇapratiṣṭhā) was successful, ‘this’ Hanumān becoming thus
for them—along with hundreds of ‘other’ Hanumāns in Vārāṇasī alone—a
seat of the deity’s all-encompassing power and a focal point for ritual activities.
In the academic study of religions, the issue of efficacy has been a major con-
cern in theoretical approaches to ritual. In both classical and more recent works
of anthropology and sociology, religious/ritual actions have been described in
terms of a fundamental dichotomy between mere symbolic-expressive acts
and others that are (functionally) effective.1 Such approaches have often been
complemented by the differentiation between an emic and an etic perspective
on ritual work and efficacy (Sax 2010). More recent approaches have attempted
to move the question of ritual efficacy to the background, emphasizing the
non-rationality (Goody 1961), meaninglessness (Staal 1979), or intentionality
(Humphrey and Laidlaw 1994) of ritual activity. However, for certain instances
of ritual, such as the installation of temple images, the efficacy question is con-
stitutive and needs therefore to be addressed explicitly.
A highly useful point of departure for analyzing image consecrations
and questions related to their efficacy is the approach proposed by Stanley
Tambiah (1979 and 1985), who states that rituals ‘achieve a change of state, or
do something effective’ by virtue of their performance, if enacted felicitously,
‘under the appropriate conditions’ (1985: 79). Tambiah relies on J.L. Austin’s
(1962) linguistic theory and his postulate of an ‘illocutionary force’ in com-
municative acts. Weddings, initiations, ordinations, and image consecrations
are thus ‘performatively efficacious’ rituals if performed correctly and if they
conform to certain other felicity conditions that are closely tied to their social
dimension: According to Tambiah (1979: 127), in order to be efficacious, ritu-
als are required to conform to certain social conventions, something that is
also a decisive factor for their legitimacy.2 Here too, Tambiah is in tune with
speech act theory: regarding the force of speech acts, Austin specifies that
1 For recent discussions of ritual efficacy see Podemann Sørensen 2006, Sax et al. 2010, Quack
and Töbelmann 2010, Töbelmann 2013.
2 See also Bell 1992: 41.
notes on the structure of an image installation 353
3 For example, over the centuries, public installation rituals were often part of larger socio-
economic and political schemes that included land donation and administration, as well as
religious expansion (Colas 2010: 319).
354 Keul
1 Preparations
When I arrived in Vārāṇasī in the fall of 1994, the shrine introduced in the
beginning of this essay was closed, the image within flawed, the deity vul-
nerable in spite of its otherwise widely believed omnipotence.5 A few weeks
earlier, the seven-foot-tall heavy stone image had all of a sudden cracked, and
the slightly displaced upper part was on the verge of falling, posing thus an
imminent danger to anyone entering the shrine. According to the ritual texts,
the damaged image was not to be worshipped anymore: undesired, malignant
spirits and other superhuman, potentially threatening entities were now able
to take up residence in a material representation that had entered a state of
ritual ambiguity.6 With Hanumān somehow still there, to be ritually released
only several months later and the old image immersed immediately afterwards
in the Ganges, his might and potency manifest in this image was believed to
be clearly limited. Given the nature of the damage, but also the opportunity to
creatively reshape the temple space, there seemed to be no doubt in the mind
4 Among other issues, Tweed (2015: 371f.) raises pertinent questions regarding the location,
duration and nature of the ‘everyday’.
5 For an earlier detailed description of the antecedents, of the preparations leading up to the
consecration, as well as of the event itself, see Keul 2002. I am grateful to de Gruyter, the pub-
lisher of that monograph, for allowing me to use the material presented there in this essay,
in a slightly modified and extended English version. A brief overview is also included in Keul
2015.
6 On the various criteria for the unsuitability (for worship) of liṅgas and other temple images,
and on the ritual procedures for their disposal or ritual recovery, see, for example, Davis
(1997: 252f.). On jīrṇoddhāra, see Czerniak-Drożdżowicz 2014 and Sharma in this volume.
notes on the structure of an image installation 355
of the owner about what was to be done: to purify, renovate and redecorate the
structure, and install a new mūrti. He commissioned several sculptors to pro-
duce drafts of the new statue, giving them a number of specifications regard-
ing the image’s appearance.7 Their task turned out to be a rather ungrateful
one, with only one of the designs looking in the end promising enough to the
owner, who is an art graduate. Work on the sculpture began in February 1995
with the fashioning of a clay model made of material brought by workers in
baskets from a precisely indicated spot on the banks of the Ganges. Before the
sculptor forcefully threw the first fistful of clay onto the board with a rope fig-
ure fixed on it (the clay model’s ‘skeleton’), he sprinkled Ganges water onto
7 According to the temple owner, a certain form of Hanumān started appearing in his dreams
shortly after the shrine had to be closed. The image differed somewhat from more con-
ventional representations in that it carried the sun in the upheld hand, a reminder of the
Rāmāyaṇa-episode in which Hanumān as a hungry child reaches for the sun with the inten-
tion of swallowing it, having mistaken it for a ripe fruit. The temple owner’s vision lacked
the apocalyptic elements found in the myth, where Indra intervenes and injures Hanumān,
and—in retaliation—the simian deity’s father, the wind god Vāyu, deprives the world of
air and breath (Keul 2002: 9, 52; see also Lutgendorf 2007: 131–3). In the owner’s dreams,
Hanumān exhibited all the splendor of the powerful, mighty god that he was, appearing at
the same time as a playful, carefree, mischievous child. The combination of might and mis-
chief, playfulness and power needed to be captured adequately in the new image.
356 Keul
both the board and the heap of clay, and recited a brief prayer. He worked
quickly and skillfully, and completed a rough outline of the figure’s body
in a few hours. It was also not long before the head of the model started to
look simian.
In the course of several days he fashioned a number of versions in clay, aim-
ing to follow closely a drawing and a calendar image hung on the wall right
before him, both of which he received from the temple owner. The sculptor’s
‘studio’ was the inner courtyard of the owner’s property, who attentively fol-
lowed the work’s progress, commented on the various versions, and advised
the sculptor. Several times he asked for complete remakes, or more or less
radical alterations. Then, after ten days of concentrated work, the clay model
passed his scrutiny and the craftsman prepared a two-part mold from a mix-
ture of cement, plaster of Paris, and glue. The casting of the plaster model for
the image took another week, and when the statue finally emerged from the
mold, immaculately white, it looked like the negative of the dark-grey clay
model destroyed in the process. On the basis of this plaster figure, in the course
of three months the sculptor created the final image, made of fine-grained,
buff-colored sandstone quarried from the Chunar hills situated not far south
of the city. Even though the successive clay, plaster, and sandstone images had
exactly the same shape and dimensions, the differences in appearance were
considerable. While the plaster model made overall a rather well-rounded,
putto-like impression, the sandstone statue8 seemed much more delicate,
with a dynamic figural composition, a strong corporeal presence and a distinct
facial expression.
Satisfied with the outcome of the work, the owner decided to proceed to
the next stage, the de-installation of the damaged image. In the second half of
June, the owner’s family priest (purohit) performed a brief ritual for the dam-
aged image, in the course of which he expressed his gratitude and asked the
deity to leave the statue. Several men were needed to carry the heavy stone
image down the steps to the Ganges, where they loaded it on a boat, rowed
out into midstream and released it there into the river (visarjana). Renovation
work on the shrine’s structure began immediately afterwards and continued
over several weeks, with the new roof completed after some delay, after the
new image’s installation.9
8 The image was a high relief with a back slab, approximately 150 cm high.
9 While I was fortunate to witness considerable parts of the preparatory process described
above, in the meantime I had left India. The description of the visarjana is based on conver-
sations with the purohit and the temple owner.
notes on the structure of an image installation 357
On the basis of astrological calculations, the family priest had already deter-
mined several auspicious time periods for the installation in the beginning of
that year, and—as the sandstone image was nearing completion—a date was
set for the ritual procedures. Invitation cards were sent out to the temple own-
er’s relatives, to friends and acquaintances,10 and to some important person-
alities in the city, such as the deputy police commissioner and the influential
mahant of the city’s most important Hanumān temple (Sankat Mochan). In
the last days of June, the temple owner’s house was buzzing with preparations.
It was being thoroughly cleaned, and some of the rooms repainted. Carpenters
were building wooden pedestals for the ritual diagrams (maṇḍalas and yan-
tras) to be used in the ritual. A mason prepared a low brick wall around a small
perimeter in the large room in which a sacrificial area (yajñamaṇḍapa) had
10 Thankfully, aware of my great and still growing interest in all things Hanumān-related,
the temple owner sent me one, too. I consulted with my academic teachers who urged me
to return to Vārāṇasī and document the consecration. I am greatly indebted to Heinrich
von Stietencron and Vasudha Dalmia, who extended their generous advice and constant
support not only in connection with this project, but also in later stages of my academic
trajectory, as well as to Shashank Singh for his trust and friendship over many years.
358 Keul
been marked off. Situated to the west of the yajñamaṇḍapa, this wall delimited
the so-called bathing area (snānamaṇḍapa), which was to become the setting
for a large number of ritual sequences that often involved the pouring of fluids
(water, milk, oil, honey) over the image in various stages of the consecration.
To control these fluids, the mason also built an elaborate drainage system for
the bathing area. The family priest inspected the entire location repeatedly,
reflecting on the correct dimensions of the sacrificial space and indicating
the precise positioning of the ritual diagrams and other implements. He also
handed the temple owner a long list of items required for the installation.11 In
the final days before the event the temple owner’s relatives arrived, many of
them from the neighboring state of Bihar, and were accommodated in various
parts of his large house.
The day before the consecration began, two ritual specialists laid out
multi-colored rice diagrams as prescribed in the ritual handbooks,12 position-
ing them in the corners and along the east side of the sacrificial space. Areca
nuts were distributed over every rice diagram, again following the ritual pre-
scriptions. The nuts represented individual deities, or—as in the case of the
main maṇḍala—groups of gods and goddesses. A clockwise enumeration
of the corner diagrams in the yajñamaṇḍapa, starting from the northwest
corner, shows the following order: kṣetrapālamaṇḍala, navagrahamaṇḍala,
yoginīmaṇḍala, and vāstumaṇḍala.13 The other focal points of ritual activities
11 The list contained 101 items, including colored powders, flower garlands, fruit, cereals
(rice, wheat, barley), pulses (chickpeas, black lentils), seeds (sesame, mustard, lotus),
spices (cloves, cardamom), various metal and earthen vessels, textiles, fluids (different
types of water: from a crossing, from the confluence of rivers, rainwater, Ganges water;
milk, cow urine, essential oil, rosewater), betel nuts, and leaves (birch, mango, banyan,
ḍhāk, pīpal, rose apple, aśok). For the complete list, with indications of quantity for most
items, see Keul 2002: 255–7.
12 The handbook used in this consecration was Pratiṣṭhāmahodadhiḥ, ed. by Agninārāyaṇ
Miśra, of which the purohit had a well-used early edition. My own copy was the sixth
edition, published in Vārāṇasī at Master Khelāḍī Lāl, Śītalprasād Saṅskṛt Pustakālay,
1986. The priest also mentioned another reference handbook that he occasionally used,
Sarvadevapratiṣṭhāprakāśaḥ (the edition in my possession was published in Bombay,
1994).
13 Nine squares of black-colored rice served as the seat for forty-nine guardian deities on the
kṣetrapālamaṇḍala (the diagram of the field or area guardians) placed in the northwest-
ern corner of the ritual enclosure. On the square-shaped navagrahamaṇḍala (the dia-
gram of the nine heavenly bodies, the grahas, in the northeastern corner) the priests laid
out symbolic figures of the nine grahas, using red-, yellow- and black-colored rice. The
yoginīmaṇḍala in the southeastern corner was rectangular, and it also included—in addi-
notes on the structure of an image installation 359
along the eastern side of the sacrificial space (between the navagraha and the
yoginīmaṇḍala) were a sarvatobhadramaṇḍala, two heaps of colored rice (one
serving as the seat for Gaṇeśa and one for the pūṛṇakalaśa), a mātṛkāmaṇḍala,
and a śrīyantra.14 A large number of vessels, coconuts, various powders, rice,
and other materials were placed in their proximity. Previously, the sacrificial
area had been purified and screened off from the rest of the room, with no one
being allowed inside, apart from the ritual specialists and the temple owner
in his function as sponsor and main (human) protagonist of the consecration.
2 Pratiṣṭhā: Day 1
At five o’clock in the morning of the first day, the priests and the temple owner,
who was the patron (or sacrificer, yajamāna) of the ritual, entered the sacri-
ficial space and sat down at their designated places. The team of five ritual
experts comprised the family priest (purohit), who acted during the consecra-
tion as the main ritual specialist or ācārya; two experienced officiating priests
(ṛtviks), who closely assisted the ācārya and also performed ritual sequences
independently; and finally, two young priestly assistants who worked under
the close supervision of their senior colleagues. During most part of the
tion to the blue, yellow, white, and red squares for the sixty-four female goddesses—three
triangular fields (black, red, and white) for Kālī, Lakṣmī, and Sarasvatī. The vāstumaṇḍala
(the diagram of the homestead) was again square-shaped, positioned in the southwestern
corner of the yajñamaṇḍapa, and its pedestal was wrapped in green cloth. The wooden
pedestals of the other three maṇḍalas mentioned here were also surrounded by cloth: red
for the navagraha-, yellow for the yoginī-, and black for the kṣetrapālamaṇḍala. For a defi-
nition of and a differentiation between the terms maṇḍala and yantra (in Śaivasiddhānta
texts), as well as on the vāstumaṇḍala as an example of balimaṇḍala, see Brunner 2003.
For detailed discussions of maṇḍala and yantra in various Hindu traditions see the con-
tributions in Padoux 1986 and Bühnemann et al 2003.
14 The sarvatobhadramaṇḍala (the all-auspicious diagram) was square/shaped and sym-
metrical, with the basic pattern of 19 × 19 square fields that were grouped in larger geo-
metrical patterns in the colors yellow, red, white, blue, and black, with an eight-petalled
lotus in the centre. It served as the seat for fifty-seven deities (or groups of deities) and
was the largest in the ritual enclosure, its pedestal surrounded by red cloth. In the course
of the procedures, it was also canopied in the same color. The rectangular mātṛkāmaṇḍala
was the seat of the sixteen female goddesses called the Mothers (mātṛkās), with fields laid
out in red, yellow, and blue rice. In its direct vicinity, the śrīyantra (the diagram of Śrī) was
painted with red dots on a white cloth. The dots served as seats for (groups) female god-
desses, along with the main goddess of the diagram, Śrī (Keul 2002: 15–16).
360 Keul
consecration’s first day, when viewed from the eastern side of the yajñamaṇḍapa,
the yajamāna was positioned in the front and centre of the ritual enclosure. To
his right sat the main priest, who would guide and closely follow the yajamāna’s
actions. The two ṛtviks were further behind, joining in the recitation of the
Vedic hymns and handing the main priest and the yajamāna—whenever
needed—implements for ongoing ritual sequences. And still further back in
the sacrificial space, the two young assistants would compete with each other
in demonstrating their knowledge of the recited texts, as well as their skills in
the preparatory work with the ritual materials.
After a moment of concentration, the preliminary phases of the pratiṣṭhā
began with an abundant purificatory sprinkling of Ganges water performed
by the yajamāna and aimed at the entire sacrificial area, including the ves-
sels, implements, and protagonists. An additional item meant to reinforce and
guarantee the purity of the sacrificer was a ring of kuśa grass that he received
from the head priest and wore on the ring finger of his right hand during
most of the proceedings. These first ritual sequences (as well as many other
later ones) were accompanied by the recitation of verses from the Yajurveda,
including the Śāntipātha. A rite of internal purification and expiation followed
(sarvaprāyaścitta), during which the yajamāna regulated his breathing several
notes on the structure of an image installation 361
times (prāṇāyāma),15 repeated the relevant verse after the purohit, and drank
pañcagavya16 from a small clay cup. Introduced by the head priest as godāna,
the giving away of a cow, the next preliminary sequence was described more
adequately in the verse accompanying the ritual, where the term goniṣkraya
was used, the giving away of the ‘equivalent of a cow,’ which was in this case a
silver coin. The purohit then announced the place and time (deśakāloccāraṇa),
as well as method and purpose (pradhānasaṃkalpa) of the ritual that was to be
performed over the next three days.17 In this and in many other instances dur-
ing the event, he acted as the yajamāna’s mouthpiece, speaking on his behalf.
It was no surprise that the first of the many detailed worship sequences was
addressed to Gaṇeśa in his quality as remover of obstacles, who was invited
to the pratiṣṭhā and asked to take his seat on an earthen vessel placed on a
heap of red-colored rice. The yajamāna then performed a ṣodaṣopacārapūjā
(the elaborate pūjā with sixteen services)18 for the deity. An earthen pot was
15 On prāṇāyāma in various ritual contexts see, for example, Einoo 2002: 25f., including
a discussion of some of the classical Indological literature on the topic (Kane, Bühler,
Caland, and others).
16 Pañcagavya is the mixture consisting of the five—direct and derivative—products of the
cow: milk, curd, ghee, cow dung and cow urine. (For a discussion of pañcagavya in the
context of consecration rituals see Einoo 2005: 106–108.) The verse recited was:
yat tvagasthigataṃ pāpaṃ dehe tiṣthati māmake / prāśanāt pañcagavyasya dahaty
agnir ivendhanam //
‘The evil that has entered my skin and bones and is in my body, burns through drink-
ing pañcagavya in the same way as fire consumes firewood.’
This essay focuses in its theoretical considerations primarily on the consecration’s perfor-
mance and internal structure. In order not to overburden the text with excessively many
notes, the quotation of verses recited during the event will be kept at a minimum, except
when it comes to sequences that will resurface in the analysis, as well as some of the
decisive moments on the third day of the installation. As will be pointed out repeatedly
in the event’s description, nearly every ritual act was accompanied by the recitation of
specific Vedic or Puranic verses. Much of the structure of this consecration (including
individual ritual sequences and the texts that were recited) is similar to or has parallels in
the material discussed in the contributions by Einoo, Hikita, and Takashima in Einoo and
Takashima 2005, whose work is based on pratiṣṭhā descriptions in early ritual texts.
17 See Bühnemann 1988: 114 for an elaborate example of an announcement of place and
time (deśakāloccāraṇa) and of a declaration of intention (saṃkalpa). On saṃkalpa as
ritual commitment and intentio solemnis see Michaels 2005.
18 In the course of the three-day installation ceremony, this form of worship occurred
numerous times in connection with the invocation of the large number of deities imag-
ined to be present on the sacrificial site. In other cases the abbreviated version with
362 Keul
five services (pañcopacāra) was chosen. For instance, when the deities on the various
maṇḍalas were invoked individually, they were often offered the pañcopacāra.
19 On pūrṇakalaśa see Gonda 1980: 131f., Bühnemann 1988: 45–6.
20 On punyāhavācana in consecration rituals see Einoo 2005: 104–106.
21 On ācāryapūjā see Bühnemann 1988: 197; on the significance of the brahmins’ hands see
Gonda 1980: 445.
notes on the structure of an image installation 363
and a nail hammered by one of the ṛtviks into the pedestal of the vāstumaṇḍala,
accompanied by carefully executed hand gestures by the yajamāna, who once
again followed the lead of the priests, trying to emulate the mudrās shown to
him as closely as possible. Several repetitions were necessary before the super-
vising ṛtvik manifested his approval. This rite was aimed at the nāgas, the ser-
pent deities of the netherworld and their leader, Nāgarāja. Rice and Ganges
water were thrown in all directions, meant as offerings for the nāgas. Then, the
deities at the vāstumaṇḍala were invoked and elaborately worshipped: first in
groups, with the rtvik slowly going line by line over the maṇḍala and draw-
ing imaginary lines with a bundle of kuśa grass, connecting (and touching)
the rows of areca nuts, and then individually, with one of the ṛtviks applying
sandalwood paste on each nut. This sequence illustrated once again the team-
work required for the smooth and efficient performance of most of the ritual
activities in the consecration: the purohit was reciting the relevant text pas-
sages, including the names of the individual deities, one of the rtviks indi-
cated the exact areca nut into which the deity was invoked (the order of the
names in the text did not correspond with the order displayed in the lines on
the maṇḍala), the other ṛtvik touched them with the ring finger of his right
hand, and the yajamāna offered rice in the direction of (or directly on) the
concerned nut. The main deity, Brahmā Vāstoṣpati, was invited to take his seat
in the metal pot (kalaśa) on the maṇḍala, containing water, rice, flowers and
dūrvā grass, on which a metal plate (tāmrapātra) with rice and a coconut was
placed. Here, and in the case of some of the other maṇḍalas, the main deity
was also offered clothing (vastra: the same type and size of yellow fabric as
offered to the priests on the previous day), a red protective cord, and a sacred
thread (yajñopavīta). And finally, the priests surrounded the entire ritual
ground with a cord, performing a clockwise maṇḍapapradakṣiṇa and offering
libations to the protectors of the cardinal directions (dikpālas) invited to take
their seat in clay pots (containing water and leaves) with trays on them, posi-
tioned around the yajñamaṇḍapa. The yajamāna followed the priests, bend-
ing down and touching each pot. Then, a pañcopacāra worship was offered to
all dikpālas. All the ritual sequences described here were accompanied by the
constant recitation of Vedic verses.
The installation procedures continued at the largest and most important
diagram, the sarvatobhadramaṇḍala. The invocation and extensive wor-
ship of its numerous deities was followed by one of the culminating ritual
sequences on the event’s first day, with Hanumān for the first time at the cen-
ter of attention. The god was invited to take his seat in the middle of the dia-
gram, where a large kalaśa was placed, topped by a coconut. Before that, the
notes on the structure of an image installation 365
25 The approximate size of this image was 2.5 × 3.5 cm.
366 Keul
After a few minutes the yajamāna performed the first part of the eye-opening
rite, following the indications of the purohit and tracing the contours of the
statue’s eyes and eyeballs, and holding up a mango in front of its face. This part
of the ritual sequence remained hidden from the spectators’ eyes, as it was
done under the aforementioned cloth. The first phase of the eye opening usu-
ally requires the craftsman’s presence: according to the texts, it is the śilpin who
performs the ritual. However, the craftsman was not present at the consecra-
tion and was substituted in this sequence by the temple owner. The cloth was
then temporarily removed and the image sprinkled with Ganges water. The
purohit indicated to the yajamāna to fold his hands (praṇām kariye!), which
he did, at first standing and then sitting in front of the statue. Bells were rung,
and the ṛtviks offered a pūjā to Hanumān, during which the materials offered
(several offerings of rice, pañcāmṛta, flowers) were thrown into the large pot
placed before the image and containing water from the kalaśas. The statue was
then again covered with the white cloth and the yajamāna decorated it with a
yellow lotus flower.
Meanwhile, in the yajñamaṇḍapa the two assistants were busy setting
up a canopy for the sarvatobhadramaṇḍala. A śānti recitation to all deities
(viśvedevāḥ) followed. The ritual activities continued at the maṇḍala of the
Yoginīs. The sixty-four goddesses were invited individually to the consecration
and offered the worship with five services. In addition to the eight times eight
notes on the structure of an image installation 367
square fields with colored rice, on the diagram there were also three larger tri-
angular fields with areca nuts for Sarasvatī, Kāli, and Lakṣmī. They too received
an elaborate worship. The diagram of the protectors of the field (kṣetrapāla)
followed next, with the invocation and worship of further forty-nine deities.
These and most of the other sequences of the first day were watched by a vary-
ing number of relatives and friends sitting on the ground on matresses covered
with white linen placed along the wall near the ritual arena, the yajamāna’s
mother being among the most avid of the spectators.
For the two following sequences, the team of specialists and the yajamāna
returned to the bathing place, where—after the removal of the cloth covering
the statue and a brief Hanumān pūjā—the yajamāna poured various types of
water over the image from eight small, earthen vessels, accompanied by the
recitation of Vedic hymns. Together with the assistants the sacrificer then dried
the statue, applied red paste and yellow rice to its forehead, and ‘crowned’ it
with a flower garland, before covering it again, this time with a piece of yellow
silk brocade with gold threads. Hidden under the brocade, the yajamāna and
a priest performed the second part of the eye opening (netronmīlana), using a
golden needle and a mirror. With the golden needle the temple owner traced
the contours of the image’s eyes, while the priest held up the mirror in order to
neutralize the deity’s first gaze, believed to be highly loaded with energy and
potentially destructive. According to the priest, another function of the mirror
was similar to the use of the mango in the first part, namely to offer the deity
an auspicious and pleasant object to behold as his first sight.
In the next sequence, the purificatory consecration rite with food
(annādhivāsana), the Hanumān statue was covered with wheat and chickpeas.
In order to limit somewhat the quantity of cereal and gram needed for the
procedure, a kind of box was set up around the image, with a framed glass
front and a back made of close-meshed wire. Verses and hymns were recited,
and one of the ṛtviks and the yajamāna poured alternately chickpeas from a
basket and wheat (with the help of a winnowing fan) over the image, creat-
ing an almost regular pattern of brown and yellow layers at the glass front
of the box. But then, even though the box seemed to have been well thought
out, its wire part bulged, and the receptacle swallowed much larger quanti-
ties of chickpeas and wheat than initially estimated and prepared, and the
yajamāna and the team of ritual specialists had to wait for some time until
more gram and wheat was provided. This involuntary break did not seem to
be entirely unwelcome: almost twelve hours into the consecration—and given
the July pre-monsoon heat—every moment of relaxation was appreciated
by officiants and participating spectators alike. However, the specialists
never lost focus, and used such breaks between sequences to discuss among
themselves—and sometimes briefly with the young foreign observer, too—
368 Keul
details concerning the installation procedures. After about ten minutes more
wheat was brought, and the sculpture’s shoulders and head disappeared under
the cereal. Covered with the cloth used earlier in the jalādhivāsana sequence,
and having been offered a brief pujā including flower garlands, it spent the
night this way. The first day of the consecration was concluded with ārtī, the
offering of light, and a circumambulation of the yajñamaṇḍapa. Facing first
the covered image of the deity and then the sarvatobhadramaṇḍala (the seat of
Hanumān in the yajñamaṇḍapa), the yajamāna waved repeatedly a lamp with
five flames, followed by the same procedure in the direction of all the other dia-
grams. The priests held their hands over the flame and touched them to their
heads. The lamp was then taken out of the ritual enclosure, and relatives and
friends of the temple owner who waited in anticipation rushed to touch the
flame and receive the blessings of the assembled deities. Meanwhile, the recita-
tion of devotional hymns continued at the sarvatobhadramaṇḍala, culminating
in exclamations such as ‘bajrāṅgabalī kī jai!’ ‘Hail to the mighty Diamond-
limbed One!’ The team then walked clockwise around the yajñamaṇḍapa. The
yajamāna bowed respectfully in front of the sarvatobhadra, was given a tilak
and a mango by one of the ṛtviks, and touched the purohit’s feet, who then ani-
mated everyone present in the room to say once again loudly ‘mahāvīr svāmī kī
jai!’ ‘Hail to the Great Hero!’ ending thus the ritual procedures of the day.
3 Day 2
At half past five in the morning, the team of ritual specialists and the yajamāna
again took their places in the yajñamaṇḍapa. Around 20–25 spectators (rela-
tives and friends of the family) sat around the ritual arena and watched the
second day of the installation begin with a worship of all the deities invoked
the day before. However, unlike the preceding day, the divine guests were not
summoned one by one, the sixteen services being offered collectively to the
maṇḍalas or groups of deities in the same order as on the first day. Gaṇeśa and
the kalaśa were worshipped after the deities of the śrīyantra, followed by the
vāstumaṇḍala, the dikpālas, the sarvatobhadra (with Hanumān), the Yoginīs,
and the kṣetrapālas. The ritual specialists were clad in some of the garments
gifted to them the previous day: the assistants wore both the lower and upper
garment, and the ṛtviks and the purohit placed the folded red chequered scarf
on their shoulders.26
26 However, the purohit’s scarf was put to an entirely different use after only a few minutes
into the morning recitations, when the wooden block to be used later in the kindling of
notes on the structure of an image installation 369
Figure 7 Day 2, early morning. In the background the image under cereals, covered.
After the yajamāna had once again performed the formal appointment of the
priests (ācāryavaraṇa), the Hanumān image was uncovered, accompanied by
the sounds of the conch (śaṅkhanāda) and the ringing of bells. Immediately
the purohit indicated that the sculpture’s face was to be cleaned first. Helpers
collected the gram and wheat, which were to become at the end of the ritual—
along with other items used in the consecration—part of the main priest’s
ritual payment (dakṣinā). The image was bathed with Ganges water from a
small brass pot, with the mixture of the five cow products (pañcagavya), and
then again with water, and then dried before being covered completely—
under the continuous recitation of Vedic hymns and verses—with garlands
made of white flowers and green leaves, and decorated with one large yellow
and several red lotus flowers, until only the face remained visible. This was
the sequence called puṣpādhivāsana, the purification and consecration with
flowers. Also red paste was applied on the sculpture’s forehead. During this
part of the consecration helpers positioned a large painted canvas depicting
jungle scenery behind the image. Alluding to the deity’s simian background,
the yajamāna remarked that this would make Hanumān feel at home in the
the fire arrived. The priest wrapped it carefully in the named scarf and placed it on the
ground close to himself.
370 Keul
renovated shrine.27 Interestingly, although the image could not yet count as
fully ritually enlivened at this stage, it nevertheless received the highly respect-
ful, at times tender and playful, personalized treatment that often character-
izes the interaction between devotees and their deities, especially in domestic,
private contexts. Near the end of the puṣpādhivāsana sequence, the temple
owner applied red color onto the statue’s lips and dark color on the eyes, and
most of those present at the consecration agreed with him that the statue,
adorned with flowers and make-up, seemed to smile contentedly.
In the yajñamaṇḍapa, preparations had meanwhile been made for the
kindling of the sacrificial fire (agnimanthana). After offering a pūjā with
sixteen services to the block and the pole of wood used in the churning
of the fire (araṇipūjā), as well as to the wooden ladles and other implements
for the fire ritual (homa), the priests and the yajamāna produced the fire in
the traditional way by placing the tip of the pole made of udumbara-wood
(the upper rubbing element, or uttarāraṇi) into one of the cavities on the rub-
bing block (the lower rubbing element, or adharāraṇi, aśvattha-wood), pulling
vigorously the ends of the rope wrapped around the pole and swivelling the
tip inside the cavity.28 As soon as smoke arose, the purohit took the block and
swung it briefly but energetically to and fro, supplying air to the faintly glow-
ing pieces of camphor in the cavity until—visibly satisfied with the result—
he declared the advent of Agni, the fire god. The priests welcomed Agni with
Vedic recitations, transferred the glowing camphor from the lower block onto
a plate with dry cow dung, ghee, and cotton, and, after some additional blow-
ing and swinging, flames started flickering visibly between the dried cowpats.
At this point one of the ṛtviks and an assistant went out of the yajñamaṇḍapa
into an interior courtyard to prepare the fire ritual. They first purified with
Ganges water the square altar (vedī) that had been built the previous day. Water
was also sprinkled on the various wooden ladles and receptacles, as well as the
bundles of grass and wooden sticks which they arranged in a row according
to ritual prescriptions. The high relevance of the correct placement of imple-
ments and instruments in a ritual context became clear once again when
the second, more experienced, ṛtvik briefly joined the outside team to clarify
details regarding the proper layout of the grass bundles. Meanwhile inside, the
yajamāna and the remaining ritual specialists performed a detailed invocation
and worship of the navagrahas at their maṇḍala. Compared to the procedure
at the other diagrams, the ritual here showed one significant difference: the
kalaśa with the tāmrapātra (with rice) and the coconut was installed not on
the maṇḍala but was placed near it. When the navagrahapūjā was done, all
ritual specialists and the yajamāna gathered around the fire altar and started a
homa reciting Vedic mantras to Agni, Indra, and Soma and offering rice (white
and black) and ghee into the fire. Mango wood and ghee was used to fuel
the flames. The purohit consulted from time to time his handbook to make
sure that every one of the numerous deities invited to the consecration was
invoked by name and given her or his offering in a sequence that took over an
hour to complete. Each verse was concluded with the exclamation svāhā!, and
the throwing of sacrificial material into the fire. With the help of a sacrificial
ladle the yajamāna poured ghee on the fire each time a deity’s name was
called, and one the ṛtviks and the two assistants made rice oblations for most
of the deities, or—in the case of the navagrahas—offered the aforementioned
grass bundles, twigs, and other pieces of wood. As the central deity of the
28 The yajamāna pulled the rope first, while one of the ṛtviks applied the necessary pressure
on the pole, then the second ṛtvik took over from the yajamāna, with the first ṛtvik still
responsible for the pressure. Then the ṛtviks changed places. In the last stage, one of the
young assistants pulled the rope and ṛtvik number two assisted him. The whole procedure
took about 20 minutes. For recent descriptions of agnimanthana rituals see, for example,
Tachikawa et al. 2001: 56–8 and Shulman 2009: 171.
372 Keul
consecration ritual, Hanumān was the adressee twice here: once in the context
of the main oblation (pradhānahoma), with 108 so-called ‘silent offerings,’ and
a little later in the sarvatobhadrahoma directed to the deities of the eponymous
maṇḍala, of which Hanumān was the presiding deity. Overall, the activities at
the vedī were rather static, monotonous, and lengthy, with scorching heat com-
ing both from the July sun and the sacrificial fire, complemented by copious
amounts of smoke from the latter. But the team stayed focused throughout,
seemed to offer the sacrificial material at the right moments and corrected
immediately whatever glitches had occurred in the recitation by repeating
the entire verse in question. The only (minor) compromise due to the exces-
sive heat was a vestimentary one: the elderly, bald purohit tied a handkerchief
around his head, thus probably deviating somewhat from the dress code pre-
scribed in the ritual handbooks for officiating prānapratiṣṭhā specialists.
After a brief break, the consecration continued inside the house with the
purification of the image with butter (ghṛṭādhivāsana). Hanumān was greeted
with the sounds of the śaṅkha and bells, the garlands were carefully removed
from the sculpture, and the yajamāna applied clarified butter (ghṛṭa) onto it.
As in the case of the jalādhivāsana, the stipulated length of purification time
was reduced here, too, from one night to a considerably shorter interval: the
notes on the structure of an image installation 373
ghee-smeared statue was covered with a white cloth, and the sequence was
over after a few minutes. Another round of conch-blowing and bell-ringing
was followed by several affusions with water. This sequence inaugurated the
elaborate bathing of the image (mahāsnāna), a composite component of the
consecration that included a generous application of curd (dadhisnāna), a
small quantity of cow urine (gomūtra), baths with the mixture of the five cow
products (pañcagavya), the five nectars (pancāmṛta), the rubbing of the entire
image with a flour mixture made of seven types of grain (saptadhānya), and
the application of a small amount of cow dung (gobar) onto the figure’s ‘hair-
line.’ Successive treatments with several other ingredients followed: ghee, a
powder made of herbs (sarvauṣadhi), honey (a ‘honey bath,’ madhusnāna), the
mixture of seven types of earth collected from different places (saptamṛttikā),
butter, brown sugar (with its abrasive, peeling qualities), and essential oils.
The mahasnāna ended with the pouring of various kinds of water on the
image: from a ford (tīrthajala), rainwater (meghajala), from the confluence
of two rivers (saṅgamjala), and finally Ganges water (gaṅgājala). The entire
bathing sequence was accompanied by recitations, but also by more or less
hushed conversations between a relatively large number of onlookers from
outside the sacrificial ground. From time to time, there was a slight confusion
regarding the content of the more than twenty earthen vessels grouped around
the sculpture, but everything was quickly sorted out and there seemed to be
374 Keul
no doubt in the minds of the experienced priests about the correct order of
the numerous baths and applications of substances. The various parts of the
elaborate bathing were carried out by the yajamāna and the two ṛtviks, the two
assistants taking charge at the end of the sequence, rubbing the image dry.
Then, a group of five musicians (shehnai, drums, and cymbals) started play-
ing in an adjacent room. The Hanumān statue in the main hall was dressed in
red silk, and richly adorned with golden, fish-shaped ear ornaments, a neck-
lace, a crown, bracelets and anklets, and finally garlanded. The priests and the
yajamāna performed a detailed pūjā of the image, with sprinkling of Ganges
water, offerings of flowers and fragrances, and ārtī. The pūjā ended with the
spectators (more than 50 of them) morphing into participants and calling out
enthusiastically well-known slogans: ‘Hail to the mighty Diamond-limbed One,
to the Great Hero, to the Lord who removes sorrows!’ (bajrāṅgabalī, mahāvīr
svāmī, saṅkaṭ mocan bhagvān kī jai!), an enumeration of some of Hanumān’s
best-known epithets. For a few minutes, all ritual activities stopped, and those
present rejoiced in the devotional atmosphere and the music.
Dressed, adorned and worshipped, the Hanumān image was now ready for
the next stage of the consecration, the circumambulation of his domain/city
(nagarapradakṣiṇa). Or—one should rather state—would have been, for in
notes on the structure of an image installation 375
the meantime, almost unnoticed by the larger public, the priestly assistants
and one of the yajamāna’s closest friends had prepared in a back room a plas-
ter stand-in for the valuable and heavy sandstone statue:29 it was the model
used in the production process of the stone image. It, too, was dressed in yel-
low silk and garlanded, then placed on a bier and taken out into the crowd in
the atrium. The purohit called out hanumān ki jai!, the music started playing,
conch horns were blown, and four bearers carried the plaster image into the
street, where they waited for a few minutes until a large number of guests gath-
ered behind them, forming a procession.
Meanwhile it was late afternoon and still very hot outside. The festive pro-
cession that took the Hanumān image around parts of Assīghāṭ did therefore
not have many spectators, but the few people who were on the streets stopped
and folded their hands respectfully when the image carried by the sweating
bearers approached standing upright on the bier. The musicians were in front,
together with a tall young man (a nephew of the yajamāna) who forcefully
waved a red flag with a gold border, heralding a high-standing member of the
Hindu pantheon. Also one of the ṛtviks walked ahead of the image carrying
the vessel from the sarvatobhadramaṇḍala on his shoulder. Accompanying his
colleague, one of the young priestly assistants blew at regular intervals into a
conch horn. Two men flanked the Hanumān image, one holding a red umbrella
with golden fringes over the statue’s head, the other fanning it rhythmically
with a whisk of yak tailhair. The yajamāna was in the procession as well, occa-
sionally rushing ahead and stopping to watch from the margins. Despite the
heat, many of his relatives and friends present during the days of the conse-
cration joined the ‘circumambulation of the city,’ which was in fact a tour of
the immediate neighborhood and took around half an hour to complete. The
pace was rather slow and the itinerary included a brief stop at the neighbor-
ing pañcāyatana temple. After climbing its steep and narrow flight of steps
and spending some time in its main sanctum, the procession returned to the
yajamāna’s house, into the atrium with the ritual arena. Hanumān’s name was
called out loudly again, and the plaster model disappeared into the back room
from where it had emerged.
The remaining ritual procedures of the second day were quite brief. A bed
was brought into the room, with pillows and coverlets, and the (sandstone)
Hanumān image was symbolically put to rest (śayanādhivāsana). The yajamāna
then performed, under the guidance of the purohit, the rite of aṅganyāsa,30
placing mantras on various parts of the image. After a recitation of sixteen
Figure 12 Aṅganyāsa.
Puruṣasūkta-verses and holding a mango in his left hand, he touched the desig-
nated parts with a long reed of kuśa grass, while the priests read the bījamantras
(mantras with seed syllables). Day two ended with an elaborate pūjā and ārtī,
the sounding of the śaṅkha and the ringing of bells, addressed both to the
sculpture and to the Hanumān installed on the sarvatobhadramaṇḍala. Once
again, everybody rushed to touch the flame, placing small amounts of money
on the tray with the lights and uttering verses in praise of Hanumān.
4 Day 3
The ritual procedures started at five in the morning with recitations and pujās
at all maṇḍalas, in the same order as on the previous days, this time includ-
ing the navagrahas as well. The Hanumān image stood covered with a red silk
cloth, the silver club, sceptre, and ornaments lying on the bed in front of it, on
spotlessly white linen between two roll pillows and beside a colorful brocade.
When the priests sounded the bell and blew the conch, and the yajamāna
uncovered the statue, its eyes were silver foils with dark eyeballs painted on
them, and on its forehead it had a Vaiṣṇava-looking tilak. An elaborate pūjā
followed and the chanting of devotional verses for Hanumān.
notes on the structure of an image installation 377
The image’s transfer to its final location was imminent. The patron’s relatives
and friends participated in the preparations, with one of the uncles erupting
periodically in bajrāṅgabalī kī jai-calls, which then prompted others to join in
as well. Judged by the reactions of the purohit and the yajamāna, such sponta-
neous outbursts of Hanumān devotion—although unscripted—did not seem
to be unwelcome at all. The atmosphere on this last day was festive overall,
but also in some moments suspenseful to a certain extent. Musicians started
playing again, and the image was tied with strong ropes to thick bamboo poles
and carried out of the room, down the stairs through the backyard past the
fire altar and again down a few more stairs to the temple. Maneuvering the
heavy sandstone block through narrow doors was a challenge, but an enthu-
siastic group of men (uncles, cousins, and friends of the yajamāna) man-
aged to carry the image unharmed to its destination where it was set upon a
pedestal.31 While the task was physically extremely strenuous, everybody
seemed particularly happy to participate in one way or another, including
women and children. Many of those who were not directly involved in giving
advice or lifting, rang bells, played mañjīra cymbals, or rattled ḍamaru drums,
adding an acoustic dimension to the commotion, to the effect that from the
ghāṭ a small crowd of onlookers gathered in the small space in front of the
temple, falling almost instantly into a celebratory mood.
After the back slab of the sculpture was attached to the wall, the priestly
assistants began with the application of vermillion (sindūr), starting as was the
case with the many other anointing sequences on the preceding days with the
statue’s face and head. As soon as it was covered entirely in the red colouring,
the image was adorned with silver ornaments (a crown, earrings, a necklace,
anklets, bracelets, a nose ornament), a yajñopavīta, and a silver club. A few
minutes later, at the astrologically auspicious moment, recitations of Vedic
verses began, with the family priest and the yajamāna entering the sanctum.
The core prāṇapratiṣṭhā sequences followed, with the main formulas and text
passages recited by the purohit and repeated (as accurately as possible) by the
yajamāna. A special variant of the Gāyatrī verse, modified to include the name
of the simian deity, was recited first,32 followed by an emphatic formulaic wish
31 Earlier that day, the five gems (pañcaratna) and a number of other items were placed into
a casket in the image’s pedestal, in the ratnanyāsa sequence. The sanctum walls had been
whitewashed the day before.
32 oṃ tatpuruṣāya vidmahe vāyuputrāya dhīmahi tanno hanumat pracodayāt. ‘We acknowl-
edge Tatpuruṣa, we meditate on Vāyuputra (Son of the Wind = Hanumān), may Hanumān
inspire us.’ The text in the handbook continued here with: iti mantrena pratiṣṭhāpya etc.
378 Keul
addressed directly to the deity.33 Both times, the yajamāna touched the image,
first the feet, then the region of the heart. After an offering of incense and flow-
ers, (the deity’s) puruṣa was invited into the sculpture.34 The recitation of the
Dhruvasūkta35 was followed by bhūtaśuddhi and ṣaḍaṅganyāsa, the placing
of bīja (seed) syllables on six different parts of the statue’s body.36 Again, the
yajamāna touched various parts of the image according to the prescription in
the ritual handbook. Bījamantras were recited, and the so-called dhyāna or
meditation verse dedicated to the Prāṇaśakti, the deification of life breath.37
The purohit asked the yajamāna to repeat after him three times a succession
of seed syllables in order to confirm the establishment of (life) breath (prāṇa,
here in the plural form), soul or life (jīva), and all sense organs (sarvendriyāṇi)
respectively, and to ask once again for an everlasting and happy (sukhaṃ
33 oṃ supratiṣṭhito varado bhava śāśvato bhava. ‘May you be well-established, may you be a
granter of wishes. May you be permanent.’
34 oṃ bhūr puruṣam āvāhayāmi / oṃ bhuvaḥ puruṣam āvāhayāmi / oṃ svaḥ puruṣam
āvāhayāmi / om bhūr bhuvaḥ svaḥ puruṣam āvāhayāmi // On the use of this mantra in
various ritual traditions and contexts, see Einoo 2005: 100–4.
35 Ṛgveda 10.173.2–6.
36 For a detailed discussion of nyāsa see Bühnemann 1988: 121f.
37 On dhyānas see, for example, Bühnemann 1992.
notes on the structure of an image installation 379
ciraṃ) presence of the deity in his new material embodiment.38 The yajamāna
performed a pūjā of the deity, which included the enumeration of twelve well-
known names/attributes of Hanumān.39 The yajamāna then bowed and the
purohit asked on his behalf for forgiveness in case any mistakes were made
during the installation. Once again, the Gāyatrī verse was recited, this time
devasya karṇe, into the deity’s ear: both the purohit and the yajamāna lowered
thereby their voices to a minimum and leaned as close as possible towards the
image. The twelve names were then repeated as well, before the priest wel-
comed Hanumān, reminding him at the same time that his presence here was
expected to last forever.40
The yajamāna and the purohit returned to the inner courtyard, where
the remaining ritual specialists had in the meantime prepared the altar, the
implements, and the sacrificial material for the second, concluding homa.
Ten dīpa-lights were placed around the altar for the guardians of the direc-
tions (dikpālas). Another light with four flames (caturmukhadīpa), for the
guardians of the field (kṣetrapālas), was placed on a large leafplate, which was
taken later along with rice and other offerings by the family barber (nāpit) to
a street crossing and left there. This homa was only slightly shorter than the
previous one, and it ended with the offering into the fire of a full sacrificial
ladle (pūrṇāhuti) and a continuous stream of melted butter (vasordhārā). The
priests then applied some ash from the sacrificial fire onto the yajamāna’s
forehead, before the latter circumambulated the altar.
38
oṃ āṃ hrīṃ krauṃ yaṃ raṃ laṃ vaṃ śaṃ ṣaṃ saṃ haṃ saḥ hanumatdevasya prāṇā iha
prānāḥ / oṃ āṃ hrīṃ krauṃ yaṃ raṃ laṃ vaṃ śaṃ ṣaṃ saṃ haṃ saḥ hanumatdevasya
jīva iha sthitaḥ / oṃ āṃ hrīṃ krauṃ yaṃ raṃ laṃ vaṃ śaṃ ṣaṃ saṃ haṃ saḥ hanumat-
devasya sarvendriyāṇi vāṅmanaścakṣuḥśrotrajihvāghrāṇāpāṇā ihaivāgatya sukhaṃ ciraṃ
tiṣṭhantu svāhā //
39
hanumate namaḥ / añjanāsūnave namaḥ / vāyuputrāya namaḥ / mahābalāya namaḥ /
rāmeṣṭāya namaḥ / phālguṇasakhāya namaḥ / piṅgākṣāya namaḥ / udadhikramaṇāya
namaḥ / sītāśokavināśāya namaḥ / lakṣmaṇaprāṇadātre namaḥ / daśagrīvadarpahantre
namaḥ // ‘Adoration to Hanumān, to the son of Añjanā, to the son of Vāyu, to the mighty
one, to the yellow/eyed one, to the crosser of the ocean, to the remover of Sītā’s sorrow, to
Lakṣmaṇa’s life-giver, to the destroyer of Daśagrīva’s pride.’
40
svāgataṃ devadeveśa madbhāgyāt tvam ihāgataḥ / sānnidhyaṃ sarvadā deva hanu-
man parikalpaya // yāvaccandrāvanīsūryās tiṣṭhanty apratighātinaḥ / tāvat tvayātra
sthātavyaṃ svecchābhaktyānukampayā // ‘Welcome, divine lord of the gods, who came
here to please me! Make your presence last forever. As long as the moon, the earth and the
sun exist unaltered (undefeated), as long may you stay here, voluntarily, out of devotion
and compassion.’
380 Keul
Back in the yajñamaṇḍapa, the yajamāna and the priests performed a brief
pūjā for the deities of all the maṇḍalas. An elaborate ārtī ceremony followed:
first in front of the sarvatobhadramaṇḍala, then in the shrine in front of the
newly installed image. The yajamāna carried the fire with the five flames
into the house again, where he was already eagerly awaited by a crowd of
relatives and friends who then rushed to touch the flames. The next ritual
sequence was the payment of the priests’ ritual salaries (dakṣinā), followed
by the sprinkling of water from all vessels onto the participants. All the deities
were ritually dismissed (visarjana), and consecrated food items were distrib-
uted. The third day of the event was concluded with the ritual feeding of the
priests (brāhmaṇabhojana).
structure. The assembly parts41 for such rituals come from various levels, such
as the material, the kinetic, and the acoustic (including the linguistic) levels:
objects (vessels, flowers, incense, fruit, textiles, etc.) were deployed during
most ritual sequences; courses of action with specific movements and gestures
were followed, for example in the purificatory sequences (of the sacrificial
area, the fire altar, the ritual instruments, the actors, the image), in the liba-
tions to deities and ancestors, in the worship with the five or sixteen services,
and in the fire ritual; and formal decisions (saṃkalpa) were extensively articu-
lated at the beginning of the consecration, mantras were recited during many
of the ritual sequences, music was played, bells rung, and conches blown.
A brief note on the relative/relational hierarchy and the centrality of ritual
units that form the complex ritual event described above is in order. While
some of the installation’s (groups of) ritual units had a largely auxiliary—but
no less essential—function, ensuring, for example, the purity of the sponsor
or of the sacrificial area, or the presence of a large number of deities at the
consecration, others were intended to contribute more directly to the transfor-
mation of the statue into a divine image. Examples for the latter, more central
units are the two stages of the eye opening and the ritual instilment of divine
power into the various parts of the deity’s body (aṅganyāsa). In the descrip-
tion of the installation I did not differentiate between main and auxiliary ritual
units for two reasons: I assumed that a step-by-step presentation of the event
as practiced in contemporary India will be a welcome addition to analyses of
(often abbreviated, sometimes cryptic) prescriptions in ritual texts and man-
uals. There, the emphasis lies on the more central pratiṣṭhā units, while the
auxiliary ones are in many cases only alluded to or hinted at, thus obscuring
somewhat the complexity of this ritual. Secondly, the detailed description
offers a better look at the ritual’s many facets and its complex structure, as well
as numerous opportunities for exemplifications when it comes to rules accord-
ing to which ritual units are connected together.
After the early influential analyses of sequential phases in rituals by
Mauss, van Gennep, Leach, and others, it was—once again—Tambiah who
described rituals as ‘patterns and ordered sequences of words and acts’42 and
41 In the categorization of morphological levels I follow Oppitz 1999 who aptly titled his
article ‘Montageplan von Ritualen,’ ‘Assembly Plans of Rituals’.
42 ‘Ritual is a culturally constructed system of symbolic communication. It is constituted of
patterned and ordered sequences of words and acts, often expressed in multiple media,
whose context and arrangements are characterized in varying degree by formality (con-
ventionality), stereotypy (rigidity), condensation (fusion), and redundancy (repetition)’
(Tambiah 1979: 119).
382 Keul
who pointed out that all complex rituals contain visible rules of sequenc-
ing (Tambiah 1985). In addition, in a programmatic essay published in 2004,
Burkhard Gladigow presented a systematic analysis of specific constellations
that occur when complex rituals are sequenced into as many elements as pos-
sible. According to Gladigow, ritual elements are defined by their repeatability,
be it in the same or in another ritual. It is this repeatability that renders them
discernible and recognizable, delimitable and combinable. Several ritual ele-
ments connected together form a ritual sequence, a composite building block
of complex rituals. Also ritual sequences have to be identifiable as coherent
units when found in different ritual environments.43
A large number of such discrete sequences can be discerned without much
difficulty in our ethnographic instance of ritual, some rather straightforward,
others more or less complex in themselves, opening up multiple morpho-
logical levels. Here are some examples for the former, more linear type: the
eye-opening sequence consisted of the covering of the image, the drawing or
tracing of the image’s eyes with the help of a golden needle, and the subse-
quent showing of a mango or a mirror. Similarly, the elaborate bathing of the
image (mahāsnāna) included a large number of discernible components or
ritual elements delimited by the manipulation of different substances used in
the process. The elaborate ārtī ceremony conducted at the end of each day and
attended by a large number of invited guests is also such a sequence with mul-
tiple and discrete elements that are recognizable to the audience, as they are
frequently performed in homes and encountered during temple visits.
Some examples for the more complex type should be mentioned as well.
Many of the ritual components observed during the three days of the installa-
tion have discernible constitutive parts that are ritual sequences in their own
right, comprised in turn of several ritual elements. For example, the activities
at the various mandalas, which would constitute ritual sequences accord-
ing to Gladigow, can be subdivided into smaller but still composite compo-
nents, quite familiar and comprehensible for the non-specialist participants.
These, too, occur in the daily worship of deities in private homes, shrines and
temples, the most common being the pūjā with five and respectively sixteen
services performed for the deities invited to take their seats on the mandalas.
A brief look at the constitutive parts of the ritual elements leads to the ques-
tion regarding the number of morphological levels discernable in a complex
ritual. The ṣoḍaṣopacāra comprises sixteen so-called services (upacāras), such
43 Gladigow 2004: 59–60. The author, who used for his model predominantly examples from
ancient Greece, has discussed ritual complexity in a number of seminal articles. See, for
instance, Gladigow 2006.
notes on the structure of an image installation 383
as the welcoming of the deity, the offering of water for washing the feet and
for rinsing the mouth, the offering of flowers and of a yajñopavīta, and so on.
Each of these actions consists of bodily movements, includes the manipula-
tion of objects, and is accompanied by the recitation of texts. Similar examples
for smaller, but still complex ritual components are: nyāsa, with the reciting
of bījamantras and the touching of various parts of the image with kuśa grass;
homa, where the recitation of verses goes hand in hand with the libations and
offerings of rice made into the fire; nagarapradakṣiṇa, which again combined
elements from several different morphological levels. The latter was once
again an easily identifiable sequence for the non-specialist participant, with
many temples having movable images (calapratimā) that are taken out on cir-
cumambulations and processions, often in connection with yearly festivals.
Returning to the issue of connective rules in complex rituals mentioned in
the beginning of this essay, the analogy between the internal structuring of
such practices and the organization of language is upheld to a certain extent.
Patterns in ritual activity have been analyzed by Frits Staal (1989), Lawson and
McCauley (1990), and others in the same way as the ordering of language in
a sentence.44 Staal’s well-known and controversially discussed claims regard-
ing the semantics of ritual is not our primary concern at this point, the focus
lying on the usefulness of his ‘syntactic’ approach. In this line of analysis he
looked at various possibilities of combining ritual sequences and discussed,
for example, the ‘syntactical’ relations (or rules) of embedding (or framing),
serialization, insertion, mirroring, and others. Analyzing the Agnicayana
ritual, Staal found, for example, that the construction of a layer of the altar
was framed by (or embedded in) two ritual sequences, pravargya (a milk
offering to the Aśvins) and upasad (a ritual fight against demons) (1989: 85f.).
Another set of formative rules for a ritual have been put forward by Lawson
and McCauley (1990). Their examples of such rules include repetition, redupli-
cation, substitution, fusion, transfer, and intermission. Reduplication is quite
common in complex rituals and concerns the doubling of ritual sequences or
elements. Substitution involves the replacement of one element by another
of equal value. Fusion means the merging of different elements, and in the
case of transfer, ritual sequences are ‘quoted’ in other contexts. Intermissions
are planned cesurae between sequences. Through the internalization of such
interdependencies, ritual competence develops: participants (actors and spec-
tators) learn to understand rituals, in a way similar to that in which the syntax
of a language is learned.
44 See on the following also Michaels 2010a and 2012. Axel Michaels has written extensively
on the morphology, syntax, semantics and pragmatics of rituals in India and Nepal.
384 Keul
had doubts about whether the heavy image could be carried on a bier safely
through the narrow lanes and over the slippery steps of the planned route,
and was looking for an alternative solution. The priest suggested loading the
statue on a lorry or a small tractor-trailer, which the yajamāna found com-
pletely inappropriate for the occasion and categorically declined. He in turn
brought up the idea of using the plaster image as a stand-in double, and now
it was the purohit who expressed his disapproval. In the end, it was the priest
who complied. So, while in the sacrificial area the main sandstone image was
apparently being prepared for the procession, in an adjoining room the much
lighter—and dispensable—plaster model received the same treatment. At
the decisive moment, it was this image that was lifted on a bier and carried
around the locality. Frequent occurrences of ritual adaptation in South Asian
contexts notwithstanding, a situation such as this one might question the
credibility of the sequence, or even of the installation procedures as a whole
for that matter. However, it seems that the legitimacy and plausibility of the
nagarapradakṣiṇa performed by the surrogate image was safeguarded in sev-
eral ways. Firstly, the use of the plaster statue was sanctioned—even if some-
what reluctantly—by the head priest, who was the highest religious authority
in the context of that event. Secondly, the stand-in was accompanied by two
entities that were considered fully imbibed with the deity’s presence: the
pitcher (kalaśa) from the sarvatobhadramaṇḍala into which Hanumān had
been ritually invited earlier, and the consecrated little golden Hanumān image
within that vessel. After the tour the plaster image was carried unceremoni-
ously into a back room, the pitcher was placed back on its mandala, and the
consecration continued seamlessly.45
I propose to end this essay with two brief notes on temporal and spatial
aspects of the imagined divine presence in the sandstone statue, based again
not on the reading of prescriptive ritual texts for image consecrations but
on the observation of an actual ethnographic instance of this type of ritual,
and on the (authoritative) feedback from the main ritual specialist conduct-
ing it. The first note concerns the various stages through which an artifact
was ritually transformed into a (permanent receptacle of a) deity. A number
of sequences were directly aimed at the ritual enlivenment of the image: the
two phases of the eye opening, the purifying consecrations with various sub-
stances (adhivāsanas), the—crucial—infusion with life (prāņapratiṣṭhā), and
45 After the installation of the sandstone statue, the plaster image was kept for a few months
in one of the side niches of another, larger temple owned by the yajamāna’s family. When
a few months later it started showing the first signs of deterioration, it was immersed in
the Ganges.
386 Keul
the nyāsa sequences. However, while the process was supposed to culminate
on the last day with the statue’s installation in the temple and its final transfor-
mation into a fully fledged, worshipable divine image, in several instances dur-
ing the three-day event it received a treatment that clearly anticipated its later
status. Three pūjās were performed for the statue on the first day: one after
the first part and one before the second part of the eye opening, as well as one
after the annādhivāsana sequence. During the second day’s puṣpādhivāsana,
the image was adorned with the care and respect that one usually observes in
temple rituals. Another, more detailed pūjā was performed for the sandstone
image on the same day, preceding the circumambulation of the city. In light of
these practices, a discussion of the exact moment(s) of the ritual crossing-over
of the material image into an imagined divine realm seems almost obsolete.
The transformation process, as observed during this complex, socio-technical
event (Keul 2015), seemed to be as much gradual as it was entangled, with con-
siderable latitude for individual ritual/devotional and aesthetic alterations
and emphases.
The second note refers to the rather contradictory answers of the purohit to
my—admittedly provocative and simplistic—inquiry regarding the simulta-
neous presence of ‘several’ Hanumān-entities during the consecration, as well
as to the question of a possible horizontal transfer of divine ‘life-force’ between
the yajña- and the snānamaṇḍapa. For there seemed to be, on the one hand,
‘one Hanumān’ imagined as present on the sarvatobhadramaṇḍala, his valence
reinforced (actually: reduplicated) by the small Hanumān figure that was con-
secrated and placed in the kalaśa, and then there was the gradually emerging,
somehow principal, ‘other Hanumān’ in (the form of) the statue. A few years
after the consecration, I had the fortunate opportunity to visit the priest in his
home. He graciously listened to my query, and replied without hesitation (and
slightly amused) that there was no question of ‘several Hanumāns’, the god
having obviously the ability of being present integrally and simultaneously in
all his splendour and omnipotence in every and any locality or container he
considered adequate. But only, the priest added after a brief pause, if the deity
chose to do so. He touched here, of course, on an often-neglected, eminently
important issue that deserves a detailed treatment of its own, namely the idea
of agency and autonomy of superhuman protagonists in consecration rituals.
In fact, the priest’s explanation seemed to have already covered the second
question I had prepared, the one concerning Hanumān’s translatio from the
pitcher on the sarvatobhadra into the image. But then I decided to pose it any-
way, and the purohit looked at me with a slightly mischievous smile saying that
there was no question at all of any transfer of life-substance (prāṇa) between
the two areas, the Hanumān in the yajñamaṇḍapa area being only temporarily
notes on the structure of an image installation 387
present, as were the other deities invited to the event. The god’s role there, he
said, was that of a witness (sākṣī), albeit a key one, of the installation proce-
dures. At the end of the event ‘this Hanumān’ was ritually dismissed (visarjana)
together with the other deities, so that his seat, the sarvatobhadramaṇḍala,
could be disposed of, and his material receptacle, the kalaśa on the mandala,
emptied and washed. The ultimately important Hanumān was ‘the other
one,’ the one that inhabited/animated the image. Having pointed this out, the
purohit leaned back with a satisfied look and rested his case.
References
Austin, John Langshaw. 1962. How to Do Things with Words. Oxford: Clarendon Press.
Bell, Catherine. 1992. Ritual Theory, Ritual Practice. New York and Oxford: Oxford
University Press.
Brunner, Hélène. 2003. ‘Maṇḍala and Yantra in the Siddhānta School of Śaivism:
Definitions, Description and Ritual Use.’ Leiden, Boston: Brill, 153–78.
Bühnemann, Gudrun. 1988. Pūjā: A Study of Smārta Ritual. Vienna: Institute for
Indology.
Bühnemann, Gudrun. 1992. ‘The “Dhyana Collections” and their Significance for Hindu
Iconography.’ Journal of Indian and Buddhist Studies 40/2, 1080–1086.
Bühnemann, Gudrun et al. (eds) 2003. Maṇḍalas and Yantras in the Hindu Traditions.
Leiden: Brill.
Colas, Gérard. 2010. ‘Pratiṣṭhā: Ritual, Reproduction, Accretion.’ In Hindu and Buddhist
Initiations in India and Nepal, eds Astrid Zotter and Christoph Zotter. Wiesbaden:
Harrassowitz, 319–39.
Czerniak-Drożdżowicz, Marzenna. 2014. ‘When Gods Get Broken: The Theory and
Practice of the jīrṇoddhāra and navīkaraṇa in the Pāñcarātrika Sources.’ Cracow
Indological Studies 16, 51–86.
Davis, Richard H. 1997. Lives of Indian Images. Princeton: Princeton University Press.
Einoo, Shingo. 2002. ‘Two Ritual Topics in the Yajnavalkya Smrti: The Tirthas in the
Hand and the Pranayama.’ Minamiajiakenkyu 14, 20–39.
Einoo, Shingo. 2005. ‘Notes on the Installation Ceremonies Described in the
Grhyapariṣiṣtas.’ In From Material to Deity: Indian Rituals of Consecration, eds
Shingo Einoo and Jun Takashima. Delhi: Manohar, 95–113.
Einoo, Shingo and Jun Takashima (eds) 2005. From Material to Deity: Indian Rituals of
Consecration. Delhi: Manohar
Gladigow, Burkhard. 2004. ‘Sequenzierung von Riten und die Ordnung der Rituale.’ In
Zoroastrian Rituals in Context, ed. Michael Stausberg. Leiden and Boston: Brill,
57–76.
388 Keul
Quack, Johannes and Paul Töbelmann. 2010. ‘Questioning “Ritual Efficacy.” ’ Journal of
Ritual Studies 24/1, 13–28.
Sax, William. 2010. ‘Ritual and the Problem of Efficacy.’ In The Problem of Ritual Efficacy,
eds William S. Sax, Johannes Quack, and Jan Weinhold. New York: Oxford University
Press, 3–16.
Sax, William S., Johannes Quack, and Jan Weinhold (eds). 2010. The Problem of Ritual
Efficacy. New York: Oxford University Press.
Searle, John Rogers. 1969. Speech Acts: An Essay in the Philosophy of Language.
Cambridge: Cambridge University Press.
Shulman, David. 2009. Spring, Heat, Rains: A South Indian Diary. Chicago: University of
Chicago Press.
Staal, Frits. 1979. ‘The Meaninglessness of Ritual.’ Numen 26/1, 2–22.
Staal, Frits. 1989. Rules Without Meaning: Ritual, Mantras and the Human Sciences. New
York and Bern: Peter Lang.
Tachikawa, Musashi, Bahulkar, Shrikant, and Kolhatkar, Madhavi. 2001. Indian Fire
Ritual. Delhi: Motilal Banarsidass.
Takashima, Jun. 2005. ‘Pratiṣṭhā in the Śaiva Āgamas.’ In From Material to Deity: Indian
Rituals of Consecration, eds Shingo Einoo and Jun Takashima. Delhi: Manohar,
115–142.
Tambiah, Stanley J. 1979. ‘A Performative Approach to Ritual.’ Proceedings of the British
Academy 45, 113–69.
Tambiah, Stanley J. 1985. Culture, Thought and Social Action: An Anthropological
Perspective. Cambridge, MA: Harvard University Press.
Töbelmann, Paul. 2013. ‘Wirksamkeit.’ In Ritual und Ritualdynamik: Schlüsselbegriffe,
Theorien, Diskussionen, eds Christiane Brosius, Axel Michaels, and Paula Schrade.
Göttingen: Vandenhoek & Ruprecht, 222–8.
Tweed, Thomas. 2015. ‘After the Quotidian Turn: Interpretive Categories and Scholarly
Trajectories in the Study of Religion since the 1960s.’ Journal of Religion 95/3,
361–85.
Welbon, Guy R. 1984. ‘Mahāsaṃprokṣaṇa 1981. Āgama and Actuality in a Contemporary
Temple Renovation.’ In Āgama and Śilpa, ed. K.K.A. Venkatachari. Bombay:
Ananthacharya Indological Research Institute, 69–102.
Index