Consecration Rituals in South Asia

Table of contents :
Notes on Contributors
Consecration Rituals in South Asia: An Introduction
2 Vedic Astrology and the Prehistory of Varāhamihira’s Pratiṣṭhāpanādhyāya: Image Installation as Apotropaic Consecration
3 On Image-Installation Rites (liṅga-pratiṣṭhā) in the Early Mantramārga
4 The Ratnanyāsa (Placing of Gems) Ritual in the
Devyāmata, an Early Śaiva Pratiṣṭhātantra
5 The Importance of Pratiṣṭhā Ceremonies in the
6 Studies in Dhāraṇī Literature IV: A Nāga Altar in
5th Century India
7 At the Crossroads of Art and Religion:
Image Consecration in the Pāñcarātrika Sources
8 Choosing an ācārya for Temple Construction and
Image Installation
9 ‘Re-Installation’ of Idols Replacing Damaged Ones, with Special Reference to the Ritual Literature of
10 The Planting of Trees and the Dedication of a
Garden: A Comparison with the Image Pratiṣṭhā
11 The Sūrimantra and the Tantricization of Jain
Image Consecration
Āgamic Temple Consecration Transnationalized
13 Ritual Observed: Notes on the Structure of an
Image Installation

Citation preview

Consecration Rituals in South Asia

Numen Book Series Studies in the History of Religions

Series Editors Steven Engler (Mount Royal University, Calgary, Canada) Richard King (University of Kent, UK) Kocku von Stuckrad (University of Groningen, The Netherlands) Gerard Wiegers (University of Amsterdam, The Netherlands)


The titles published in this series are listed at

Consecration Rituals in South Asia Edited by

István Keul


Cover Illustration: Hanuman image awaiting consecration (Banaras 1995, photo by István Keul) Library of Congress Cataloging-in-Publication Data Names: Keul, István, editor. Title: Consecration rituals in South Asia / edited by István Keul. Description: Leiden ; Boston : Brill, 2017. | Series: Numen book series :  studies in the history of religions : ISSN 0169-8834 ; Volume 155 |  Includes bibliographical references and index. Identifiers: LCCN 2016052323 (print) | LCCN 2017003055 (ebook) | ISBN 9789004313705 (hardback : alk. paper) | ISBN 9789004337183 (e-book) Subjects: LCSH: Rituals—South Asia. | Religious articles—South Asia. Classification: LCC BL603 .C65 2017 (print) | LCC BL603 (ebook) | DDC  203/.8—dc23 LC record available at

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: issn 0169-8834 isbn 978-90-04-31370-5 (hardback) isbn 978-90-04-33718-3 (e-book) Copyright 2017 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner.

Contents Notes on Contributors vii 1 Consecration Rituals in South Asia: An Introduction 1 István Keul 2 Vedic Astrology and the Prehistory of Varāhamihira’s Pratiṣṭhāpanādhyāya: Image Installation as Apotropaic Consecration 17 Marko Geslani 3 On Image-Installation Rites (liṅga-pratiṣṭhā) in the Early Mantramārga 45 Dominic Goodall 4 The Ratnanyāsa (Placing of Gems) Ritual in the Devyāmata, an Early Śaiva Pratiṣṭhātantra 85 Anna A. Ślączka 5 The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas 113 Marie-Luce Barazer-Billoret 6 Studies in Dhāraṇī Literature IV: A Nāga Altar in 5th Century India 123 Ronald M. Davidson 7 At the Crossroads of Art and Religion: Image Consecration in the Pāñcarātrika Sources 171 Marzenna Czerniak-Drożdżowicz 8 Choosing an ācārya for Temple Construction and Image Installation 198 Elisabeth Raddock 9 ‘Re-Installation’ of Idols Replacing Damaged Ones, with Special Reference to the Ritual Literature of Kerala 223 S.A.S. Sarma

vi 10


The Planting of Trees and the Dedication of a Garden: A Comparison with the Image Pratiṣṭhā 241 Shingo Einoo

11 The Sūrimantra and the Tantricization of Jain Image Consecration 265 Ellen Gough 12

Āgamic Temple Consecration Transnationalized 309 Annette Wilke


Ritual Observed: Notes on the Structure of an Image Installation 351 István Keul 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 interests 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 formation 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 concern 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.



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, meditation, 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 modern religious movements and is involved in a research project on the sociocultural 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.



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 working on aesthetic communication and in a research project on Global Hinduism.


Consecration Rituals in South Asia: An Introduction István Keul 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 classified under the umbrella-term pūjā, ritual activities performed in connection with the worship of iconic or aniconic ritual images in temples and domestic 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 subcategory. These, too, can be performed either regularly (for example as yearly temple celebrations) or in connection with specific occasions. Image consecrations are exactly such occasional rituals that often constitute unique events in the lives of their sponsors and are performed in connection with the construction or renovation of shrines and temples. The gradual transition from Vedic religion to a more personal and imagecentered religiosity that started in the last centuries BCE led to the development 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) presents 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; BrunnerLachaux 1998; Einoo and Takashima 2005; Ślączka 2007; Colas 1989 and 2010. 2  See Einoo 2005: 13, 95–113.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_002



public) ritual system. In his introduction to the volume on Indian rituals of consecration that he jointly edited with Jun Takashima, and summing up the findings of several of the volume’s contributors, Shingo Einoo points out that the pratiṣṭhā gradually gains in complexity through absorbing ‘various kinds of ritual elements,’ and that ‘in its fully developed stage the pratiṣṭhā ceremony consists of a great number of ritual acts’ (Einoo 2005: 2–3). Analyzing in the same collection the description of liṅga installations in early Śaiva āgamas, Jun Takashima identifies several stages, from the ritual appropriation of an existing liṅga to the simple installation in a small temple for private worship, to the construction of a temple and the more complex consecration of a liṅga for public worship (Takashima 2005: 136–7). Colas (2010: 319f.) has argued more recently against a diachronic development leading from private and less elaborate to public and more complex forms suggesting that there are reasons to assume that public and private pratiṣṭhās existed contemporaneously in the first centuries CE, and that the ritual procedures described in the fifthcentury Gṛhyapariśiṣṭa texts might as well be reduced versions for domestic use of more elaborate public installation rituals that were already common in South Asia. Colas points—among other examples—to Buddhist epigraphic records from the first four centuries CE that mention public pratiṣṭhās of relics and images, emphasizing at the same time the importance of political, social, and historical factors for the development and transformation of image consecration rituals over the centuries.3 In addition to the aforementioned texts, descriptions of consecration rituals of different types4 can be found in a large number of Purāṇas and other theological and liturgical works from the Śaiva, Vaiṣṇava, and Śākta traditions, not to mention the numerous ritual handbooks (compiled from the eighth–ninth centuries onwards) meant to serve as manuals for officiating religious specialists. The polysemic term pratiṣṭhā occurs frequently in various contexts, even before the ritual consecration of private and temple images became common. A brief inquiry into its various meanings in Vedic and post-Vedic text genres may prove useful when it comes to tracing semantic (and ritual) continuities in the history of consecration rituals. In the most extensive examination of the term’s use in Vedic and post-Vedic literature, Jan Gonda (1954/75) points 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


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: In contradistinction to some other terms for related ideas the etymology and ‘basic’ meaning of the word pratiṣṭhā are perfectly clear [. . .]. Deriving from prati—‘towards, near to, against, down upon, upon, on’—and sthā—‘to stand (firmly), to take up a position, to stand still, to stay, to abide’, pr.[atiṣṭhā]—must, theoretically, have been a very suitable term for any foot or base in a literal sense. Yet, there is no sufficient evidence of those metaphorical applications which have developed in connection with words for similar ideas in other languages: ‘foot, bottom, or basis of almost every object, foot as a measure of length’ etc. It had, from the beginning, other connotations. (Gonda 1954/75: 338–9) In the Atharvaveda, pratiṣṭhā means an ability located in the feet of a living being, while in the Aitareya Āraṇyaka it also means the hold one has of the ground one is standing on.5 In the same text, the stomach is characterized as the pratiṣṭhā of food. Less surprisingly, the verb pratitiṣṭhati is used in the Śatapatha Brāhmaṇa for ‘standing firmly on one’s feet,’ and the causative form, pratiṣṭhāpayati, to express the placing of an object. Pratiṣṭhā can mean the back part or support of the body (the tail in the case of a sitting bird), and also the last constituent of an object or action, or the concluding part of verse. (These and the following selections are examples from Brāhmaṇa texts.) The Vedic sacrifice is described as standing firmly on (pratitiṣṭhati) or being supported by four pillars (the Vedas and the corresponding priests). The earth is a pratiṣṭhā or foothold for both deities and sacrificers. The latter are ‘well-supported,’ having performed ritual offerings and chanted the (metrically three– footed and therefore firmly established) gāyatrī verse. Gonda lists numerous other examples for the figurative use of both the verbal and the substantive forms (342f.), which often occur directly connected to or in the wider context of the Vedic sacrifice. The semantic range of pratiṣṭhā widens further in post-Vedic texts (368f.). In the Bhagavadgītā it is used to express a stable, detached mental state (BhG 2.57). In verse 14.27, Kṛṣṇa declares himself to be the pratiṣṭhā of the immortal 5  On this and the following, see Gonda 1954/75: 339f.



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: If a Hindu makes or purchases an image of a deity it is his invariable practice to perform certain ceremonies, called prāṇapratiṣṭhā, ‘the establishment or instalment of vital breath, of life, endowment with animation’. It has often been said that by going through this process of ‘consecration’ the nature of the images changes, that they are no longer merely the materials of which they are constructed, but become containers of life and supranormal power. (371) Relevant text examples invoked here by Gonda are—among others—the Bṛhatsaṃhitā and the Vaikhānasasmārtasūtra. In addition to image consecration (including their re-consecration, punaḥpratiṣṭhā), the term is also used (in Gṛhyapariśiṣṭasūtras, Purāṇas, Tantras) in connection with the foundation and construction of temples, as well as of wells, water reservoirs, or parks, for sites that are dedicated to the public. Turning to tantric texts (such as the Tantrarājatantra and others), Gonda mentions that even though these works distinguish between the ritual placing of divine power into an image, into the body of a human practitioner, or in a yantra or a vessel (kumbha, kalaśa), all these actions are called pratiṣṭhās.7 Numerous other ‘shades of meanings’ from various post-Vedic text genres conclude Gonda’s survey, copiously illustrating the term’s versatility.8

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


There are interesting observations to be made with regard to the contexts in which the term pratiṣṭhā (or its Prakrit equivalent) is used to designate consecrations in early Buddhism. The installation of items believed to possess special qualities, such as relics or scriptures,9 does not seem to imply a ritual transformation of these—already venerated—objects, but is prescribed in the Vinaya literature for the sanctification of the earth. Such (utmost meritory) early Buddhist pratiṣṭhās were most probably performed at auspicious moments in the presence of monks and laypeople, ‘as a public process of stabilizing sacred articles on the surface of the earth and an ideology of terrestrial expansion of the Buddhist religion and message’ (Colas 2010: 322). The same seems to apply in the case of man-made architectural elements of stūpas (vedī, chattra, and others) or early Buddhist images. Also, in some early Pāñcarātra texts, the ritual installation of objects such as stones marked with footprints emphasizes the ‘placing’ (nyāsa), the material fixation/stabilization of these items (already considered sacred), not their transformation. This, however, is the exception. Most Hindu ritual texts prescribe rules for consecrating new artifacts specially produced in order to be ritually converted into temple images (322–3). Rituals specially designed to convert statues of stone, metal, wood, or other materials into (receptacles of) deities raise fundamental questions regarding the connection between deity and image in South Asia.10 The centuries-old existence and present-day persistence of consecration rituals can readily be interpreted as evidence for devotees’ conception that images in the region’s temples and shrines are actually deities, permanent as in the case of temple images or temporary as with festival images. After all, in consecration rituals ‘profane’ sculptures, man-made lifeless artifacts, are ritually infused with divine life force, becoming thus the focus of prayer, worship, and other religious activities. A glance at some of the numerous Sanskrit terms used to designate (iconic) images is helpful to some extent here as they express: (1) the sculptures’ likeness of the deity they present (pratikṛti, pratimā); (2) their embodying or taking the form of the deity (mūrti); (3) their offering a means or a form for the devotee to grasp the essence or nature of the deity (vigraha) (Eck 1981/98: 38). This latter term points to other traditional conceptions of the relationship between image and deity, versions of which 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 installation 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.



present-day students of South Asia’s religions will probably encounter in the field. Interlocutors may point out—depending on their religious affiliation and/or philosophical inclination—that one worships the deity the image represents and not the image itself; that images only serve as a visual means to establish a relationship with a transcendent god; that the divine is formless and incomprehensible; and that the worship of images is in fact inappropriate for those who aspire to higher knowledge.11 However, to observers of contemporary everyday religious practices the relationship between deities and their images in Hindu temples, shrines, and homes appears usually to be twofold. The devotees’ worship is addressed to the deity as image and to the deity whose power is in the image (Fuller 1992: 61). In the course of elaborate, multi-stage ritual processes, sculptures or other man-made objects change into entities considered worthy of being worshiped. Primarily on the grounds of the imagined metamorphosis undergone by their main objects, image consecrations can serve as examples par excellence of transformative rituals. The change is not sudden but gradual, even if some of the constitutive ritual sequences seem to bring about a much more substantial progress than others in the image’s development towards (being the container of) a deity. The number of constituent sequences varies according to the different schools and traditions.12 Various types of purifications and affusions, the imposition of mantras and powers, the ritual opening(s) of the eyes, and other elements contribute to the transformation, a process that has sometimes been interpreted (for example in texts of the Pāñcarātra, Vaikhānasa, Digambara Jain traditions, but also in Buddhist texts13) as similar to the gradual refinement of a person’s body through lifecycle rituals (saṃskāras). Zooming out to a much wider, more theoretical and systematical angle, and looking at what happens to the ritual’s main object, an image consecration is what, in the terms of Don Handelman (2006: 47), could be called ‘an event that models.’ Pratiṣṭhās combine features defined as characteristic for such events. They are specialized microcosms, are teleological with a specified direction, and they resolve a conflicting state of existence (lifeless/animated). Of course, the intricate ritual 11  Davis 1997: 479. In addition to the Advaita Vedānta position, Davis offers in the first chapter 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


choreographies affect not only the sculptural representations of deities, but more or less everyone and everything involved in the ritual. Ritual sponsors and officiating priests perform roles according to culturally established rules and follow closely detailed scripts; interested audiences observe and occasionally join in the activities; divine agents are invoked and directed towards prescribed locations; spaces, receptacles, and persons are purified; substances and materials manipulated and symbolically altered. Familial or communitarian bonds are reinforced, statuses confirmed, and social visibilities enhanced. Image consecrations are profoundly relevant and potentially transformative on a number of levels, such as the social, the experiential, or the performative. The essays in this volume address major issues connected with the pratiṣṭhā in Śaiva, Vaiṣṇava, Buddhist, and Jain contexts, offering insights into the consecration procedures in various traditions and historical periods. The collection is thematically and methodically broad, with detailed textual studies and contributions that discuss contemporary performances of installation rituals. The sources examined include pratiṣṭhā-related chapters in an early encyclopedic text, early Śaiva treatises, Āgamas, a late-classical Buddhist text, Vaiṣṇava Pāñcarātra sources, and Jain manuals. Historical continuity and change, and the transformational moments in the pratiṣṭhā, are overarching motives in a number of contributions, while others have more specific themes such as the connection between art and religion, the qualifications of the ritual specialist, the re/installation of damaged images, and the dedication of a garden. The ethnographic accounts of consecration rituals include a Jain pratiṣṭhā in Jaipur, a temple and image consecration in Hamm-Uentrop (Germany), and the installation of a Hanumān image in Varanasi. 1

The Essays

A text of central importance for the installation of images in South Asia is chapter 59 of Varāhamihira’s encyclopedic work (sixth century), the Pratiṣṭhāpanādhyāya, in which the author succinctly offers a general description of the installation ritual, which he holds valid for virtually every major iconocentric tradition of his time, including the Buddhists and Jains. In his contribution, Marko Geslani first places the rite described in this chapter of the Bṛhatsaṃhitā in the wider context of Varāhamihira’s other writings on rituals, and shows how the adhivāsana (‘overnight’) sequences of the pratiṣṭhā contain references to rites performed for/on the king prior to royal military campaigns. While in the pratiṣṭhā the image is aspersed with various materials, in the yātrā rites the king is bathed in a similar way, with plant-based



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 lateVedic 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 ceremonies. The sixth-century author, however, modifies the Vedic ritual structure 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 examination 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 characterized 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 centuries 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


unpublished accounts (Sarvajñānottara and Niśvāsatattvasaṃhitā) from different periods. Anna Ślączka’s essay focuses on one of the central ritual components in the consecration of images, the ratnanyāsa, ‘the placing of the gems.’ This ritual consists of the depositing of small objects (often precious or semiprecious stones) below the pedestal of an image. The ratnanyāsa is mentioned in varying detail in a large number of texts from different periods, Śaiva and Vaiṣṇava. Some of the relevant passages (such as those in the Bṛhatsaṃhitā or certain Purāṇas) are brief, while other ritual texts include elaborate descriptions of the objects and their positioning. Ślączka especially examines one text, an early (heterogeneous pre-eleventh-century CE) Śaiva treatise on the consecration of temples, titled Devyāmata. After discussing in detail its possible dating and structure, and offering a synopsis of the ratnanyāsa ritual included in the treatise, the author looks at the development of the ritual by comparing the accounts in various textual sources across the centuries and genres, pointing also to the potential that lies in comparing texts with archaeological findings in an attempt to trace the history of the ratnanyāsa ritual. She concludes that the ratnanyāsa account in the Devyāmata can be dated between the late seventh and early eleventh centuries, which would place this text in an intermediary position between the earliest Śaivasiddhānta treatises discussed in the previous contribution by Dominic Goodall (the fifth- to sixth-century Niśvāsa texts) and later Śaiva texts such as the Somaśambhupaddhati or (South Indian) āgamic texts, which form the source material for the next essay. The āgamic sources on pratiṣṭhā analyzed by Marie-Luce Barazer-Billoret often start with a Vedic formula to be used in the first stages of temple consecration rituals. The author points out that the Indian subcontinent has been characterized by a continuity of ritual practices over the centuries. The focus of Barazer-Billoret’s essay lies on the patterns found in installation rituals of liṅgas and other images, and on the question regarding the exact moment of transformation from a statue into an entity thought of as divine. In one pattern, the image’s/the liṅga’s transformation takes place on the last day of the lengthy ritual procedures, namely the day of the installation itself, through the transfer of mantras from the ritual receptacles into which the deities were invoked, to the image (Śiva) and its pedestal (the Goddess). However, there is no general agreement in the āgamic sources regarding this issue. While the older texts seem to conform to the pattern outlined above, there are also sources that do not allow for any conclusive statement when it comes to the decisive transformative moments. And in other, more recent, texts the metamorphosis occurs a day before the installation proper, namely during the performance of the adhivāsa ceremonies in the sacrificial pavilion.



Drawing on a wide range of texts, but focusing in particular on a lateclassical, pre-tantric Buddhist source tentatively called Mūlyamantra-sūtra, Ronald Davidson examines in his essay the installation of a Buddhist nāga altar and the rain rituals connected to it. He points to three possible areas as sources for these ceremonies and the complex, aggregate ritual and doctrinal system contained in the text. The first of these areas is the non-Sanskritic worship of nāgas (snake deities) in North Indian agricultural communities, as attested in descriptions of the use of mud images in some South Asian texts, in the survival of terracotta figurines representing nāgas at a number of sites in Uttar Pradesh, Bihar, and Bengal, as well as in ethnographies describing nineteenth-century nāga-related sites and practices. Second, a number of the items mentioned in the Mūlyamantra can be found in the later Gṛhya texts, especially the layout of the altar. In the third area, some of the practices of local sorcerers called vidyādharas, practices characterized by the control and domination of a hostile being, for example of a demonic vināyaka or of the yakṣa Mahākāla, can be found in the nāga ritual of the Mūlyamantra. Davidson points out that the worshiping of nāgas for rainfall is not continued in brahmanical literature where the snake divinities are invoked in various other contexts, such as the sarpabali rite, or in instances where nāgas are associated with the underworld and its treasures, or in the placement of nāga statues (nāgapratiṣṭhā) in connection with the installation of deities. On the other hand, the practice of nāga-related rituals and their connection to rain went on for several centuries in Buddhist contexts over large parts of Asia. In her essay on image consecration in the Vaiṣṇava Pāñcarātra, Marzenna Czerniak-Drożdżowicz emphasizes the close connections between art and religion/ritual reflected in the sources of this tradition. Among the five modes of divine presence in the world conceptualized by the Pāñcarātrikas, arcāvatāra refers to the presence of the deity in its various forms of representation: selfmanifested, consecrated by deities, consecrated by siddhas, or man-made, which have to be installed according to rules laid down in the āgamic texts. The author presents the consecration procedures based on selected Pāñcarātra texts, with the main focus on the Paramasaṃhitā account, and looks also at a pratiṣṭhā-related area discussed in ritual treatises and texts on art, the repairing of damaged images (jīrṇoddhāra, navīkaraṇa). Another main theme of the essay is an inquiry into the development of the idea of arcāvatāra. Two of the early Pāñcarātrika texts, Jayākhyasaṃhitā and Viṣṇudharmottarapurāṇa, express some reservations when it comes to image installation and worship. Even if it states that through consecration the image is transformed into the deity, the Jayākhyasaṃhitā still prescribes āvāhana, the invocation of the

An introduction


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 eighthto ninth-century Hayaśīrṣa Pāñcarātra, another Vaiṣṇava Pāñcarātra text that deals with both the building of temples and the production of temple sculptures. The main focus of the text, however, is on ritual, only providing information on architectural aspects when these are ritually significant. Raddock discusses the various possibilities to categorize the text, one such possibility 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 templebuilding 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 replacement 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 installation 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 imminent danger of being destroyed. The author relates the case of the image from



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 designate 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 practices. The requirement that the performance of pratiṣṭhā should be undertaken

An introduction


by initiated ācāryas seems to have developed in the aforementioned period in a number of other traditions as well (Śaiva Siddhānta, Vaiṣṇava Pāñcarātra, Indo-Tibetan tantric Buddhism). In Śvetāmbara texts, the consecration sequence in question replicates the recitation of the sūrimantra by the guru into the (ritually decorated) ear of the disciple, conferring upon him the status of an ācārya (or sūri). After examining a number of texts relevant for the rituals of contemporary Śvetāmbaras and Digambaras, and pointing out distinctions, mutual influences, and controversies regarding the role of ascetics in image consecrations, Gough gives a summary of a seven-day Digambara pañcakalyāṇaka pratiṣṭhā festival observed in 2013 in Jaipur. The sūrimantra sequence took place on the sixth day, being described in the manual Pratiṣṭhāpāṭha as one of the main components for the ritual transformation of the image. Overall, however, there are a number of different versions of the sūrimantra today and there is disagreement among the Digambaras on how it may be used in the pratiṣṭhā. Annette Wilke discusses in her essay major events and developments connected to the Kamadchi temple in the industrial area of Hamm-Uentrop, a mid-sized town in Westphalia, Germany. The temple, built in the South Indian classical style, was inaugurated in the summer of 2002 and re-consecrated twelve years later. A Hindu temple has existed in Hamm since 1989, first in a basement and later at another location in a newly built house. As a result of the initiative of an energetic priest supported by a German advisory board, it was possible to build and consecrate the new temple in the traditional architectural style. Wilke offers a detailed outline not only of the initial consecration procedures but also of the rituals of renewal. The core sequences of the 2002 pratiṣṭhā were performed over eight consecutive days. However, this was preceded by a number of other rituals such as the vāstupūjā and the ceremony of setting the foundation stone performed three years earlier, as well as a nontraditional German topping-out ceremony, ‘Richtfest,’ in 2000. The main rites included the worship of various divinities, the opening of the image’s eyes, the circumambulation or procession, the ritual bathing of the image, a fire sacrifice, the depositing of the gems in the pedestal, the installation of the image in the temple, the instilment of life (prāṇapratiṣṭhā), and, on the final day, the great ablution, at which the press and television were also present. Wilke then goes on to examine various developments on different levels that followed after the consecration, such as the evolution of the new temple into a Tamil Hindu pilgrimage and festival site, with participants not only from Germany but also from abroad, or the development of the Kamadchi temple into the most important, representative site for Hinduism in Germany. She also touches on some negative aspects and events in connection with the temple



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 complexity 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 The volume originated in a conference organized in the fall of 2012 at the Department of Archaeology and Religious Studies, the Norwegian University of Science and Technology (NTNU), Trondheim. I am grateful to the Department and to NTNU’s Faculty of Humanities for financially supporting the event. I thank the editors of Brill’s Numen Book Series for including this collection into the series. An anonymous reader at Brill has offered valuable suggestions for improvement. I would also like to thank Maarten Frieswijk, Stephanie Paalvast, Giulia Moriconi and Renee Otto for their support in guiding the project smoothly and patiently through the publication process. References Barazer-Billoret, Marie-Luce. 1993–4. ‘L’installation des liṅga et images dans les temples 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


Brunner-Lachaux, Hélène. 1998. Somaśambhupaddhati: rituels dans la tradition śivaïte selon Somaśambhu. 4me partie, rituels optionnels: pratiṣṭhā. Pondicherry: Institut français d’indologie. Bühnemann, Gudrun. 1991. ‘The Ritual for Infusing Life (prāṇapratiṣṭhā) and the Goddess Prāṇaśakti.’ Zeitschrift der Deutschen Morgenländischen Gesellschaft 141/2, 353–65. Colas, Gérard. 1989. ‘L’instauration de la puissance divine dans l’image de temple en Inde du Sud.’ In Revue de l’histoire des religions 206/2, 129–50. 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. Davis, Richard H. 1997. Lives of Indian Images. Princeton, NJ: Princeton University Press. Eck, Diana. 1981/1998. Darśan: Seeing the Divine Image in India (3rd ed.). New York: Columbia University Press. 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, 95–113. Einoo, Shingo and Jun Takashima (eds). 2005. From Material to Deity: Indian Rituals of Consecration. New Delhi: Manohar. Fuller, Christopher J. 1992. The Camphor Flame: Popular Hinduism and Society in India. Princeton, NJ: Princeton University Press. Gombrich, Richard. 1966. ‘The Consecration of a Buddhist Image.’ Journal of Asian Studies 26/1, 23–36. Gonda, Jan. 1954. ‘Pratiṣṭhā.’ Studia Indologica Internationalia 1, 31–54; reprinted in J. Gonda. 1975. Selected Studies II, Sanskrit Word Studies. Leiden: Brill, 338–74. Handelman, Don. 2006. ‘Conceptual Alternatives to “Ritual.” ’ In Theorizing Rituals: Issues, Topics, Approaches, Concepts, eds Jens Kreinath, Jan Snoek, and Michael Stausberg. Leiden: Brill, 37–49. 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. New Delhi: Manohar, 241–82. Kane, Pandurang Vaman. 1941. History of Dharmaśāstra, Vol. II, Part II. Poona: Bhandarkar Oriental Research Institute. Schopen, Gregory. 1998. ‘Relic.’ In Critical Terms for Religious Studies, ed. Mark Taylor. Chicago: University of Chicago Press, 256–68. Schopen, Gregory. 2005. Figments and Fragments of Mahāyāna Buddhism in India: More Collected Papers. Honolulu, HI: University of Hawai’i Press. Ślączka, Anna Aleksandra. 2007. Temple Consecration Rituals in Ancient India. Leiden: Brill.



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.


Vedic Astrology and the Prehistory of Varāhamihira’s Pratiṣṭhāpanādhyāya: Image Installation as Apotropaic Consecration1 Marko Geslani It is well-known that the earliest firmly datable ritual instruction for the Hindu image installation ceremony, or pratiṣṭhā, is to be found in chapter 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 sectarian 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’ (pratyekasya 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 possibility 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

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_003



represents an early, pan-sectarian version of pratiṣṭhā, laying out a general structure that might offer a baseline for comparison with variants from later Purāṇic and Āgamic sources.5 In this paper I would like to situate BS 59 in the context of Varāhamihira’s broader ritual program. Recently I have attempted to explore the construction of some important rituals in the Bṛhatsaṃhitā by reference to two other works of Varāhamihira’s corpus, the Bṛhadyātrā (BY) and the Yogayātrā (YY), and in conjunction with the late ritual texts of the Atharvaveda, namely, the Śāntikalpa (ŚK) and Pariśiṣṭas (AVPŚ) (Geslani 2016). These sources develop a number of new ritual forms attending to the special aims of the royal office (rājadharma), placing the king himself in the role of the Vedic sponsor (yajamāna). It is not surprising then that the royal body figured as the central object of attention in such rites, subject to a number of ritual activities—most frequently bathing, but also such acts as fasting, adornment, and the binding of amulets. While it might easily be supposed that the divine image in the pratiṣṭhā ceremony is loosely based on the king of Vedic royal consecration rites, I will argue here— following a line of inquiry suggested by Michael Witzel (1987: 428–30)—that it is this particular royal body, ritualized within an Atharvan-astrological regime, which provides the template for the divine image (the ‘body’ of the god) in the pratiṣṭhā ceremony of BS 59. My argument relies on two interrelated sets of evidence. In the first place, I will try to show that the adhivāsana portion of the pratiṣṭhā in BS 59 recapitulates some of the distinctive structural motifs and discrete elements found in the preliminary rites of Varāhamihira’s texts on the yātrā (or military expedition)—in particular, the bath of clay, the observation of fire omens (agninimitta) and the observation of dreams (svapnanimitta). Second, having demonstrated the correspondence between the image and the king in astrological sources, I will suggest how this astrological ritual structure (especially its use of homa) may be viewed as an elaboration of ritual patterns found in śānti rites prescribed in the texts of the Atharvaveda. Together this evidence suggests that Varāhamihira’s version of the pratiṣṭhā may—at least in terms of ritual form, if not in terms of ritual meaning or purpose—represent a variant of what I call the ‘apotropaic consecration,’ an aspersion ceremony based on Atharvan śānti paradigms, and originally designed to counteract inauspicious forces. This conclusion is supported independently by versions of the pratiṣṭhā found in both the Vedic gṛhyapariśiṣṭas and several purāṇas, all of which adopt Atharvan ritual techniques to the installation of images. versions of the pratiṣṭhā ceremony do appear to have maintained the ‘Vedic’ procedure outlined 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



The Structure of Varāhamihira’s Pratiṣṭhāpanādhyāya

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 ritual, 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 pillow, 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).



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 accomplished by focusing on the ritual cycle of divine images alone, in Purāṇic or Āgamic sources. 2

The Pratiṣṭhāpanādhyāya (BS 59) in Light of the Yātrā Rites

2.1 The ‘First’ Bath as Victory Bath Perhaps the most distinctive ritual action in the adhivāsana portion of the pratiṣṭhā is the complex bath of the image with a series of materials, including waters infused with tree-leaves, clay drawn from various locations, pañcagavya, tīrtha water, and jeweled water (step 3): 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 auspicious 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


A strikingly similar bath is prescribed as part of the ritual prelude to the king’s departure for the military expedition (yātrā). Varāhamihira records two slightly different versions of this royal bath, YY 7, the “Nakṣatra Victory Bath” (nakṣatravijayasnāna), and BY 19, more simply titled the “Victory Bath” (vijayasnāna). Let us review both versions of the text. YY 7 consists of two separate bathing segments. First, verses 1–12 prescribe a specific set of bathing ingredients for each of the 27 nakṣatras. In this “arboreal,” nakṣatra-related segment of the bath, most of the ingredients are drawn from different species of plants. Here, for example, are three verses of the text: 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 bathing 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 ingredients 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:



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. In this way the account of YY 7 emphasizes two distinct bathing sequences, one relying on plant-based mixtures and the other on clay gathered from different locations. Note that while the logic behind the nakṣatra/arboreal bath is not explained, the alluvial bath uses the verbs meaning ‘to purify’ or ‘wash’ (√śudh; pra√kṣal), suggesting that it functions to purify the body of the king. The BY9 separates the arboreal and alluvial portions of the bath. The nakṣatra-specific arboreal bath is dealt with in BY 4, whereas the bath of clays appears in BY 19. BY 4 provides detailed information for relating the direction of the king’s departure with a specific nakṣatra. Verses 21–25 of this text supply bathing substances for the primary and secondary nakṣatras that are genearlly recommended for travel in each of the four directions.10 These verses closely parallel the instructions for the nakṣatra bath that we have 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


seen above in YY 7.1–12. Perhaps because it omits the nakṣatra-specific directions, which are described elsewhere, BY 1911 is simply titled the ‘Victory Bath’ (vijayasnāna)—rather than the ‘Nakṣatra Victory Bath.’ No specific instruction is given to relate BY 4.21–25 with BY 19, but it is possible that the text assumes that the nakṣatra-specific bathing instructions of BY 4 are carried out in conjunction with the general instructions of BY 19. Another possibility is that the variable nakṣatra bath has been replaced with a more general procedure. The central ritual action of BY 19 begins as follows: 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]. It is possible that the ātmarakṣā (‘bodily protection’) procedure mentioned here, using priyaṅgu, siddharthaka, nāgadāna, gorocanā, honey, and ghee, is meant as a general alternative to the variable nakṣatra baths of YY 7/BY 4, since these substances are commonly found among the ingredients for the nakṣatra baths. In this case we might infer that the nakṣatra bath has the purpose of ‘protecting’ the body of the king. Directly following this ‘bodily protection,’ the ‘Victory Bath’ itself is prescribed as follows: śvetasya babhror athavā vṛṣasya carmāstare vyāghramṛgendrayor vā | tatsthasya kuryān manujeśvarasya jayābhiṣekaṃ vidhivat purodhāḥ || 9 || kramān mahīrūpyasuvarṇakumbhaiḥ kṣīrasya dadhno haviṣaś ca pūrṇaiḥ | snāyāc ca toyaiḥ saha saptamṛdbhiḥ paścāc ca sarvauṣadhigandhatoyaiḥ || 10 || BY 19 || [9] The purohita should, according to instruction, perform the jayābhiseka of the king stationed there, on the skin of a white or tawny bull, or of a tiger or lion. [10] He should bathe the king, in order, with pots made of earth, silver or gold, filled with milk, curd, and clarified butter; and with water containing seven kinds of clay, and finally with water mixed with all herbs and scents. 11  Chapter 17 in Pingree’s edition.



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 consistent set of bathing practices, which combine a variable number of plantinfused 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 similarities 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 supplied 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. 2.2 Fire Omens Varāhamihira’s reliance on conventions set forth in his yātrā texts is even more direct in the case of the observation of fire omens during the adhivāsana of the image installation ceremony. The observation of these omens occurs in conjunction with the homa offering (step 4):

12   B S 59 includes clay from lotus ponds, whereas YY 7 adds clay from the king’s gate, a courtesan’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


yo devaḥ saṃsthāpyas tanmantraiś cānalaṃ dvijo juhuyāt | agninimittāni mayā proktānīndradhvajotthāne || 12 || BS 59 || 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 banner (indradhvaja). Like the image installation ceremony, that ritual also prescribes a fire offering, after which the omens of the fire are to be observed.14 The relevant verse of the indradhvaja reads: iṣṭadravyākāraḥ surabhiḥ snigdho ghano ‘nalo ‘rciṣmān | śubhakṛd ato ‘nyo ‘niṣṭo yātrāyāṃ vistaro ‘bhihitaḥ || 31 || BS 42 || 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. Thus Varāhamihira refers us again to another source, which he calls simply the yātrā. In fact, as we will see, Varāhamihira makes frequent reference to this text in the ritual instructions of the BS, such that the yātrā functions as a crucial source of practical conventions, especially as they apply to divination sequences occuring in ritual contexts. In our case, the commentator Bhaṭṭotpala understands the above instance of the term ‘yātrā’ as a refernce to YY 8 (corresponding to BY 20–21), which describes a homa ceremony dedicated to the planets, to be performed prior to the king’s departure. This sacrifice, however, has more than a propitiatory function, but it in fact offers yet another occasion for divination. Both yātrā texts supply a series of omens that may be derived from the condition of this fire at the moment of offering the main oblation. The same omens—Varāhamihira seems to be suggesting—apply to the fire offering in the pratiṣṭhā of BS 59 (and in the indradhvaja of BS 42) as in the original yātrā rite. We will return to this text, and its homa format later, but here I would stress that in the royal setting, the reading of the fire omens forms one ritual motif preceding the departure of the king. It appears 14   sitavastroṣṇīṣadharaḥ purohitaḥ śākravaiṣṇavair mantraiḥ | juhuyād agniṃ sāmvatsaro nimittāni gṛhṇīyāt || 30 || BS 42 || . 15  For iṣṭadravya, Bhaṭṭotpala gives ātapatrādi, that is, auspicious objects such as a parasol.



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. 2.3 The Sleeping Image and Dream Divination The final indication of the contribution of yātrā-based ritual motifs has to do with the laying of the image to sleep. As is apparent in the schematic summary of the text, Varāhamihira’s pratiṣṭhā follows a two-day format. That is, after the series of rites that comprise the first half of the ceremony (steps 1–5) the sthāpaka (the image handler), lays the image on a bed (step 6). Then the image, now sleeping (suptām), is attended through the night by worshippers who stay awake singing and dancing (step 7): pratimāṃ svāstīrṇāyām śayyāyām sthāpakaḥ kuryāt || 14cd || suptāṃ sagītanṛtyair jāgaraṇaiḥ samyag evam adhivāsya | daivajñasampradiṣṭe kāle saṃsthāpanaṃ kuryāt || 15 || BS 59 || 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: āvāhiteṣu kṛtvā pūjāṃ tāṃ śarvarīṃ vaseyus te | sadasatsvapnanimittaṃ yātrāyāṃ svapnavidhir uktaḥ || 22 || BS 47 || 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 mentioned in the Yātrā. 16  Utpala glosses jāgaraṇa as asvāpaniṣevaṇas, ‘waking (un-sleeping) attendants.’ Incidentally, it is at this point that Varāhamihira employs the verb adhivāsya, suggesting that he understands the term ‘adhivāsana’ as related to an overnight vigil.

image installation as apotropaic consecration


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 oneiric signs in order to determine the likely success of the impending undertaking. 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 private temple and observe his dreams, along with his minister, astrologer, and purohita: brāhmīṃ sadūrvām atha nāgapuṣpīṃ kṛtvopadhānaṃ śirasi kṣitīśaḥ | pūrṇān ghaṭān puṣpaphalābhidhānān āśāsu kuryāc caturaḥ krameṇa || 3 || yaj jāgrato dūram udaiti daivam āvartya mantrān prayatas trir etān | laghv aikabhug dakṣiṇapārśvaśāyī svapnaṃ parīkṣeta yathopadeśam || 4 || BY 18 || Having made a pillow of brahmī (leaves) or nāga flowers with dūrva blades, at the head of the king, he should place four full pots filled with flowers and fruits in the directions. He carefully repeats the mantra “yaj jāgrato dūram udaiti daivam” (VS 34.1) three times. Lying on his right side, having eaten a small meal, he quickly observes the dream according to instruction. As one of many ceremonial details of this dream divination, the purohita prepares 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.



(the corresponding moment in Varāhamihira’s text can be found at step 7). I would argue then, that whatever may have been the original form of the adhivāsana, the motif of laying the image to sleep in Varāhamihira’s version of the rite may have been adapted from dream divination sequences originally centering on the king.19 So far I have argued that the adhivāsana of BS 59 repackages a number of ritual motifs found in other ceremonies described in Varāhamihira’s texts— motifs which ultimately derive from the ritual program of the royal military campaign or yātrā. Hence I would see this astrological, royal figure as the model for the image, which recapitulates in the adhivāsana the same ritual narrative that precedes the king’s departure. This interpretation may explain two additional details common to pratiṣṭhā ceremonies. First, we may be able to interpret the ‘opening of the eyes’ of the image (netronmīlana)—a sequence that common to among variants of the pratiṣṭhā, but not mentioned in the BS—as the awakening of the king after his night sojourn. Second, on the following day, the image is taken clockwise around the temple before it is installed (or enthroned) in the temple. This pradakṣiṇa journey, which might otherwise be linked to the Vedic chariot drive of śrauta consecrations, seems to reenact in miniature the king’s yātrā, or digvijaya conquest. After his successful conquest, according to BY, the king returns to his own country (here the temple) and grants bali and further requests to the various beings who safeguarded him during the military campaign.20 Thus according to our interpretation, the physical installation of the image refers not (or not only) to the king’s coronation, but his triumphant return from conquest. 3

Vedic Astrology and the Pratiṣṭhā

I would like now to take this analysis one step further, to explore the possible 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 divination 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

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 references 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: Table 1

Ritual correspondences between varāhamihira’s yātrā and Atharvan mahāśānti

Ritual/Deity Group

Bṛhadyātrā (ch)

Yogayātrā (ch)

Śāntikalpa (ch)

Lokapāla Guhyaka/Vināyaka dream divination clay bath Nakṣatra/Vijaya-snāna (Nava)graha homa (mantra-gaṇas) fire divination

15.4cd 15 16 4.19–23 17 18 19.1–7 19.8–10

6.1–18 6.19–28 6.29 7.13–15 7.1–12 6.1–18 8.1–7 8.8–19

2.14 1.3–9 AVPŚ 68 1.5–6 2.1–12/AVPŚ 1 1.10–18 2.24 AVPŚ 24/29a

a See Bosch (1978).

In short, although the precise sequence of rites varies, the preliminary rituals of Varāhamihira’s royal journey correspond in large measure with the preliminary 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,



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 features: 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 earlier Atharvan texts as śāntyudaka.22 A mantragaṇa consists of a set of individual verses from the Atharvaveda grouped together under a given heading or theme, such as ‘longevity’ (āyuṣya) or ‘safety’ (abhaya). A group of mantras 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 aforementioned 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


entire ceremony culminates in a ritual bath with an elaborately prepared water mixture produced with the dregs of a fire offering. The close relation in Atharvan śānti ceremonies between the fire offering and the act of bathing or consecration is crucial for our purposes. This relation may help us explain yet another common motif of many early-medieval versions of the adhivāsana, which feature a second bathing ritual that follows a fire offering. In fact, Varāhamihira’s fire offering most decisively demonstrates his debt to Atharvan ritual conventions. Recall the homa sequence in his adhivāsana of BS 59 (step 5): yo devaḥ saṃsthāpyas tanmantraiś cānalaṃ dvijo juhuyāt | agninimittāni mayā proktānīndradhvajotthāne || 12 || BS 59 || 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. As we have seen, the chapter on Indra’s banner (indradhvaja) also prescribes an offering in which the omens of the fire are to be observed, and which refers us to a fuller description of the subject in the ‘yātrā.’ As I have detailed elsewhere, the corresponding chapters in YY/BY—in addition to detailing the expected fire omens—prescribe that the requisite fire offering should be made with five mantragaṇas found in Atharvan śānti rituals: āyuṣya, abhaya, aparājita, śarmavarman, and svastyayana (Geslani 2016), Thus Varāhamihira’s fire offering in the image installation appears to be but one of several variations of a sacrifice that was likely based on an Atharvan prototype. Varāhamihira recommends the observation of fire omens in at least three other important royal ceremonies besides the image installation, including the indradhvaja (BS 42.30d), nīrājana (BS 43.14), and the puṣyasnāna (BS 47.78). It is possible to speculate whether this instruction functions as kind of ritual paribhāṣa or ‘convention’ within his texts. According to this hypothesis, at each occurrence, the text would intend the performance of a fire sacrifice using the five Atharvan mantra gaṇas, which could then be supplemented by other mantras specific to the given ritual context, for example mantras dedicated to Indra and Viṣṇu in the case of the indradhvaja. Such a convention appears at least to operate in the Purāṇic royal coronation ceremony, or rājyābhiṣeka, surviving in the Nīlamata and Viṣṇudharmottara purāṇas.25 These texts prescribe the 25   Nīlamatapurāṇa 841–842 (Kumari ed.); Ādipurāṇa (Ikari and Hayashi ed.), lines 2632– 2633; Viṣṇudharmottarapurāṇa 2.19.4–5. See Geslani 2012: 337.



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 addition 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 keeping with the “non-sectarian” approach of his text.26 As we will see, one important 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 indicate 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 apotropaic consecration that represents the astrological counterpart to the Atharvan puṣyābhiṣeka (AVPŚ 5). While both versions of this ritual share the same underlying 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 preparation 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 question, 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


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). 4

The ‘Second’ Bath of the Classical Adhivāsana

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 influence of Atharvan śānti rituals on the structure of the adhivāsana. I focus specifically 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 offering, and bathing. In these instances, the saṃpāta-dregs regularly facilitate the transferral of divine presence into the image—just as they facilitated the transferral 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 simply 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.



Among Vedic versions of the pratiṣṭhā in pariśiṣṭa-level literature, the clearest example of the use of saṃpātas during an adhivāsana bath can be found in a late text, the Ṛgvedic Āśvalāyanīyagṛhyapariśiṣṭa (ĀśvGPŚ 4.4–6),29 which Einoo has summarized (2005: 98–100). This highly elaborate version of the ceremony includes the first ‘purifying’ bath examined above, including clay, herbs, jewels, etc. (ĀśvGPŚ 4.5 [p. 177, lines 4–11]). This first bath appears to take place in a separate pavillion (snānamaṇḍapa), constructed beside another, primary maṇḍapa, where image is carried after its first bath, and in which an elaborate fire sacrifice is performed at five fire pits (kuṇḍa). The purpose of this fire sacrifice is made abundantly clear in the following passage: athācāryaḥ śrotriyāgārād agnim āhṛtyoktavat svakuṇḍe pratiṣṭhāpyānvā­dhāya devasyātra kalāsaṃnidhisiddhyarthaṃ yakṣya iti saṃkalpya praṇītāsthāpanāntaṃ kṛtvā tatra pratiṣṭhāya devam āvāhya tatprakā­śamantrair vyāhṛtibhir vā palāśodumbarāśvatthāpāmārgasamidājyacarutilair aṣṭasahasram aṣṭaśatam aṣṭāviṃśatiṃ vā juhuyāt || ĀśvGPŚ 4.6 || 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: tatrājyamāhutisaṃpātaṃ pratikuṇḍaṃ sthāpitaśāntikalaśeṣu nikṣipet | tataḥ śirasy upari pādayoḥ pratimāṃ spṛṣṭvā sviṣṭakṛdādi hutvā pūrṇāhutiṃ juhuyād evam ṛtvijo ‘pi svayaṃ kuṇḍe juhuyuḥ | atha maṇḍapasthakalaśodakena samudrajyeṣṭhā iti catasṛbhir āpohiṣṭheti tisṛbhir abhiṣicya pratiṣṭhāya devamantreṇa saṃsrāvakalaśodakena saṃsnāpyāmbayo yanty adhvabhir iti sūktena śuddhodakakalaśenābhiṣicya 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


vastrayugmenācchādya pañcopacāraiḥ saṃpūjayed ity adhivāsanaṃ trirātram ekarātraṃ vā sadyo vā kuryāt || ĀśvGPŚ 4.6 || 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, worshipping 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 structure 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 mantras such as—in the case of Viṣṇu—the puruṣasūkta (RV 10.90), or the verse



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 purpose of the adhivāsana—to obtain divine presence—could be appropriately 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 elaborate 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


central fire pit, while attendant priests (called mūrtipas) along with representatives 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): śiraḥsthāne tu devasya sthāpako homam ācaret | śāntikaiḥ pauṣṭikais tadvanmantrair vyāhṛtipūrvakaiḥ || 30 || . . . etebhyo vaidikair mantrair yathāsvaṃ homam ācaret | tathā śāntighaṭaṃ kuryāt pratikuṇḍeṣu saṃnyaset || 43 || śatānte vā sahasrānte saṃpūrṇāhutir iṣyate | samapādaḥ pṛthivyāṃ tu praśāntātmā vinikṣipet || 44 || āhutīnāṃ tu saṃpātaṃ pūrṇakumbheṣu vai nyaset | mūlamadhyottamāṅgeṣu devaṃ tenāvasecayet || 45 || sthitaṃ ca snāpayet tena saṃpātāhutivāriṇā | 46ab | Matsyapurāṇa 265 | 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 protecting 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. This version of the second bath is roughly identical to that of Āśvalāyana, insofar as it takes place in a maṇḍapa equipped with five fire pits. The central offering takes place near the head of the image, which has been laid ‘to rest.’ But rather than the expected mantras related to a specific deity, the main offering is performed ‘with such mantras as those pertaining to śānti and puṣṭi, preceded by the vyāhṛti syllables.’ Whatever specific mantras may be intended here, it is clear that they are not simply mantras of invocation, but encompass the effects of our earlier śānti rites. Note once again that the pots containing the remainders appear to be called ‘śānti pots.’ In fact, what appears to have happened here is that this version of the pratiṣṭhā has accomodated another mode of invocation. Using direct visualization and literal mantras, the ‘moment of instantiation’ seems to occur after



the adhivāsana proper, in the pratiṣṭhā-half of the ritual, when the god is physically 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 differently, 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 relationships 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 version 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 technical 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 rituals. 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 structure through the application of sacrificial dregs (saṃpāta) to the bathing waters. Despite the fact that Varāhamihira seems reluctant to use Vedic mantras 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 hypothetical reconstruction is correct, it might explain why, even without the use

image installation as apotropaic consecration


of saṃpātas, a number of early medieval groups continue to invoke the god into a water pot—mentally through visualization and/or orally through the direct recitation of mantras—prior to bathing the image.31 The ritual lineage of śānti thus may have contributed one ‘method’ of divine instantiation to the pratiṣṭhā, a ritual whose medieval variants became increasingly diverse complex. It may eventually be possible to differentiate this method from others (e.g. nyāsa, dhyāna), with the goal of developing a typology of different forms of instantiation. So it seems that Atharvan techniques for apotropaic consecration were incorporated in settings far beyond what I have proposed were their earlier Vedic-astrological contexts. I close, however, with two ‘complications’ that seem to arise from the historical trajectory that I have proposed. First, the application of śānti to divine instantiation was, as we have seen, very much dependent on the selection of mantras. The Atharvan śānti rituals employed the mantragaṇas, collections of Vedic verses that encapsulated certain qualities or physical states (such as appeasement, safety, or longevity), which could be transferred to the king’s body through the saṃpāta. Analagously, in the invocation of gods into water pots with saṃpātas, what was invested in the physical image—strictly speaking—was not the presence of the god per se, but a mantra that merely invoked, or requested the god’s presence. Perhaps this ritual problem formed one setting for the theological question of the relationship between gods and speech, for which both Vedic and ‘tantric’ theologies would offer ample response (Malamoud 1998, Padoux 1990). Second, aside from these theological considerations, we have also seen that the historical relationship between śānti and pratiṣṭhā cannot simply be understood in terms of a straightline development, from appeasement to invocation. I suspect that the Vedic and astrological roots of pratiṣṭhā were not entirely erased in image installation ceremonies. As the evidence of the canonical account from the Matsyapurāṇa suggests, some versions of the ceremony could be quite explicit regarding the apotropaic functions of the second adhivāsana bath. And even apart from the question of ritual resemblance, the general atmosphere of astrological concern in pratiṣṭhā ceremonies is hard to deny. The endurance of Atharvan and astrological motifs in image installation rites should be noted, especially considering the political functions of temple worship in the post-Gupta era. To understand the function of the image installed in the temple—beyond its already well-recognized devotional

31  Cf. Vaikhānasagṛhyasūtra 4.11; Rauravāgama 28.45–50 (Takashima 2005: 123); Liṅgapurāṇa 2.47.



capacities—we may have to consider the ritual life of its post-Vedic, royal prototype with greater scrutiny.

Appendix 1: Pratiṣṭhā According to Bṛhatsaṃhitā 59


[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 32  [step 1] diśi yāmyāyāṃ kuryād adhivāsanamaṇḍapaṃ budhaḥ prāg vā | toraṇacatuṣṭayayutaṃ śastadrumapallavacchannam || 1 || pūrve bhāge citrāḥ srajaḥ patākāś ca maṇḍapasyoktāḥ | āgneyyāṃ diśi raktāḥ kṛṣṇāḥ syur yāmyanairṛtyoḥ || 2 || śvetā diśy aparasyām vāyavyāyāṃ tu pāṇḍurā eva | citrāś cottarapārśve pītāḥ pūrvottare kāryāḥ || 3 || āyuḥśrībalajayadā dārumayī mṛṇmayī tathā pratimā | lokahitāya maṇimayī sauvarṇī puṣṭidā bhavati || 4 || rajatamayī kīrtikarī prajāvivṛddhiṃ karoti tāmramayī | bhūlābhaṃ tu mahāntaṃ śailī pratimāthavā liṅgam || 5 ||  śaṅkhūpahatā pratimā pradhānapuruṣaṃ kulaṃ ca ghātayati | śvabhropahatā rogān upadravāṃś ca kṣayam kurute || 6 || 33  [step 2] maṇḍapamadhye sthaṇḍilam upalipyāstīrya sikatayātha kuśaiḥ | bhadrāsanakṛtaśīrṣopadhānapādāṃ nyaset pratimām || 7 || 34  [step 3] plakṣāśvatthodumbaraśirīṣavaṭasambhavaiḥ kaṣāyajalaiḥ | maṅgalyasaṃjñitābhiḥ sarvauṣadhibhiḥ kuśādyābhiḥ || 8 || dvipavṛṣabhoddhataparvatavalmīkasaritsamāgamataṭeṣu | padmasaraḥsu ca mṛdbhiḥ sapañcagavyaiś ca tīrthajalaiḥ || 9 ||  pūrvaśiraskāṃ snātāṃ suvarṇaratnāmbubhiś ca sasugandhaiḥ | nānātūryaninādaiḥ puṇyāhair vedanirghoṣaiḥ || 10 || 35  [step 4] aindryāṃ diśīndraliṅgā mantrāḥ prāgdakṣiṇe ‘gniliṅgāś ca | vaktavyā dvijamukhyaiḥ pūjyās te dakṣiṇābhiś ca || 11 || 36  [step 5] yo devaḥ saṃsthāpyas tan mantraiś cānalaṃ dvijo juhuyāt | agninimittāni mayā proktānīndradhvajotthāne || 12 || dhūmākulo ‘pasavyo muhur muhur viphuliṅgakṛn na śubhaḥ | hotuḥ smṛtilopo vā prasarpaṇaṃ cāśubhaṃ proktam || 13 ||

image installation as apotropaic consecration


[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] [10] [11] [12] [13] [14] [15]

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

worshipping with flowers, clothing, ointments, drums and conches (16ab) taking the image around the temple clockwise (16cd)40 bali-offering (17a) worshipping brahmaṇas and assembly members (17b) placing a piece of gold in the base (17c) placing the image on the base (17d)41 worshipping the performers involved in the rite (sthāpaka, astrologer, priest, assembly, carpenter) (18ab)42

Appendix 2: Adhivāsana According to Matsyapurāṇa 264–265 construction of the maṇḍapa and bathing maṇḍapa (MtP 264.13–26) bringing the image into the maṇḍapa (27–28) tracing the eyes (29–34) dividing the image into parts (35–40) bathing the image in the bathing maṇḍapa with pañcagavya, clay or ashes (265.7–8) honouring the image with perfume and unguents and covering it with cloth (9–11b) bringing the image (to the main maṇḍapa?) on a cart (11c–12) laying the image on a bed, with a water pot placed at its head (13–15) anointing the image with ghee, honey, and mustard seeds; worshipping with incense and flowers (16–17) offering clothing, accouterments and food (18–22)

37  [steps 6–7] snātām abhuktavastrāṃ svalaṅkṛtāṃ pūjitām kusumagandhaiḥ |  pratimāṃ svāstīrṇāyām śayyāyām sthāpakaḥ kuryāt || 14cd || 38  [step 8] suptāṃ sagītanṛtyair jāgaraṇaiḥ samyag evam adhivāsya | 15ab |. 39  The categorical distinction between the adhivāsana and the pratiṣṭhā is justified by the following instruction: “He should perform the installation of the image at a time indicated by the astrologer” (daivajñasampradiṣṭe kāle saṃsthāpanaṃ kuryāt || 15cd ||). 40  [steps 9–10] abhyarcya kusumavastrānulepanaiḥ śaṅkhatūryanirghoṣaiḥ |  prādakṣiṇyena nayed āyatanasya prayatnena || 16 || 41  [steps 11–14] kṛtvā baliṃ prabhūtaṃ sampūjya brāhmaṇāṃś ca sabhyāṃś ca | dattvā hiraṇyaśakalaṃ vinikṣipet piṇḍikāśvabhre || 17 || 42  [step 15] sthāpakadaivajñadvijasabhyasthapatīn viśeṣato ‘bhyarcya | 18ab |



[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)


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


Rauravāgama: Rauravāgama. Edited by N.R. Bhatt, 2 vols. Pondichéry: Institute Français d’Indologie, 1961–1972. Śāntikalpa: George Melville Bölling, ed. ‘The Śāntikalpa of the Atharva-Veda,’ Transactions and Proceedings of the American Philological Association 35 (1904): 77–127 [=ŚK 2]; ‘The Śāntikalpa of the Atharva-Veda,’ JAOS 33 (1913): 265–278 [=ŚK 1]. Vaikhānasagṛhyasūtra: Vaikhānasasmārtasūtram. The Domestic Rules of the Vaikhānasa School Belonging to the Black Yajurveda. Edited by W. Caland. Calcutta: Asiatic Society of Bengal, 1941. Viṣṇudharmottarapurāṇa: The Viṣṇudharmottarapurāṇam. Delhi: Nag Publishers, 1985. Yogayātrā [YY]: Yogayātrā. Edited by Rāmacandra Jhā. Mahārājādhirājakāmeśvarasiṃhagranthamālā 23. Darabhaṅga, 1986; H. Kern, ed. ‘Die Yogayātrā des Varāhamihira.’ Indische Studien 10 (1868): 161–212; 14 (1876): 312–358; and 15 (1878): 167–184;

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.



Harting, P.N.U. 1922. Selections from the Baudhāyana-Gṛhyapariśiṣṭasūtra. Amersfoort: J. Valkhoff & Co. 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. Malamoud, Charles. 1998. ‘Bricks and Words: Observations on the Bodies of the Gods in Vedic India.’ In Cooking the World: Ritual and Thought in Ancient India, transl. David White. Delhi: Oxford University Press, 207–225. 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. Padoux, André. 1990. Vāc: The Concept of the Word in Selected Hindu Tantras, transl. J. Gontier. Albany: State University of New York Press. Rotaru, Julieta. 2009. ‘The Śāntyudakavidhi in the Atharvavedic Tradition.’ In Śrīnidhiḥ: Prof. Shrikant Shankar Bahulkar’s Gratitude Volume, ed. Shripad G. Bhat. Pune: Saṁvidyā Institute for Cultural Studies, 162–204. Smith, H. Daniel. 1984. ‘Pratiṣṭhā.’ In Agama and Silpa: Proceedings of the Seminar held in December, 1981, ed. K.K.A. Venkatachari (Anantacharya Indological Research Institute Series, no. XVI). Bombay, Anantacharya 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, 115–142. Witzel, Michael. 1987. ‘The Coronation Rituals of Nepal, with special reference to the coronation of king Birendra.’ In Heritage of the Kathmandu Valley, eds Niels Gutschow and Axel Michaels. St. Augustin: VGH Wissenschaftsverlag, 417–467.


On Image-Installation Rites (liṅga-pratiṣṭhā) in the Early Mantramārga 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 development 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 liṅgapratiṣṭhā 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_004



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). śuddhakāyaḥ sunakṣatranimittopacite dine sampūjya devadeveśaṃ kuryāl liṅgaparigraham 13 sthāpitam ṛṣibhiḥ pūrvaṃ sarvalakṣaṇalakṣitam āśrayet siddhidaṃ liṅgaṃ svayaṃ vā parikalpitam 14 kīlitaṃ varjayed viprāḥ skandaviṣṇvindramātṛbhiḥ upary upari liṅgaṃ ca mukhaliṅgaṃ tathaiva ca 15 atha vā lakṣaṇair yuktaṃ sthaṇḍilaṃ parikalpayet sthaṇḍile ’pi hi mantrāṇāṃ siddhir āśu pravartate 16 śreṣṭhaṃ paścānmukhaṃ liṅgam abhāve prāṅmukhaṃ smṛtam yāmyakauberavaktraṃ tu na siddhāv iṣyate budhaiḥ 17 parigṛhyepsitaṃ liṅgaṃ liṅgakalpoditaṃ śubham pūjayet satataṃ mantrī matvā deham aśāśvatam 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 transcription of the oldest manuscript of which is produced as an appendix 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


that is possessed of all required characteristics and bestows success, or to one that he has prepared himself. He should avoid, o brahmins, one that [belongs to a sacred area that] has been staked out (kīlitam) by Skanda, Viṣṇu, Indra or the Mothers, and also one that is above another (?),2 or one that has a face or faces emerging from it. Alternatively he may prepare a bare patch of ground with the requisite characteristics, for on a patch of bare ground too the mastery of mantras may also be quickly attained. The best is a liṅga that [is in a shrine that] faces West. In the absence [of that], an East-facing one is recommended. But one that faces South or North is not held to be suitable for siddhi by the wise. Having adopted the liṅga that he wishes to take, that is [as] described in manuals on liṅgas, that is auspicious, the mantrin should constantly venerate [it], bearing in mind that the body is impermanent. There is, however, a detailed account of liṅgapratiṣṭhā in what may be the earliest scripture of the Mantramārga to survive, the Niśvāsatattvasaṃhitā. Now there is a Rudrapratiṣṭhākalpa in the Bodhāyanagṛhyaśeṣasūtra, which may therefore be the earliest surviving account of what might be described as a Śaiva pratiṣṭhā, but the norm that the author of that text had in mind was the installation of a statue, and, as Einoo (2005b: 96) has observed, that Rudrapratiṣṭhākalpa so closely follows a Viṣṇupratiṣṭhākalpa given earlier in the same text that it may be assumed to be simply calqued upon it. Before I now launch into a discussion of liṅgapratiṣṭhā, I wish to mention my indebtedness to Dr. Diwakar Acharya, not only for his earlier presentation on the subject in 2007, in which he pointed out similarities with the account of the Vaiṣṇava Svāyambhuvapāñcarātra, but also for his numerous emails on the subject, passing on materials and making suggestions and corrections to the text, etc. 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.



By examining various Śaiva sources, we can try to judge where pratiṣṭhā fits into the Mantramārga tradition and in what ways its rôle changes. This is of course exactly what Jun Takashima attempted, but of the handful of Śaiva scriptures he chose, the Svāyambhuva­sūtrasaṅgraha has no account of pratiṣṭhā, that of the Mṛgendra is largely lost, that of the Mataṅga is rather too short to be the basis of much extrapolation, that of the Kiraṇa is very different in the South Indian edition, whose text is often uninterpretably corrupt, from in the Early Nepalese manuscript that transmits it, and that of the Raurava is a post-twelfth-century account produced for the religious world of South Indian temples.3 In other words, we are missing a large part of the puzzle here, and what we would really have to do to examine the changing form and function of pratiṣṭhā through time is to study the surviving manuscripts of the Kiraṇa, Devyāmata, Sarvajñānottara, Piṅgalāmata, Mayasaṅgraha, Mohacūḍottara, Bṛhatkālottara. Since each of these accounts needs to be read from a range of manuscripts and critically edited,4 this paper is intended, like Jun Takashima’s interesting contribution, to advance our understanding of the genesis and evolution of pratiṣṭhā rites somewhat further, but it makes no claim to be definitive. Besides the excellent volume edited by Professors Einoo and Takashima (in which Jun Takashima’s article appears, as well as other articles mentioned in these pages), another obvious starting point for trying to understand liṅgapratiṣṭhā is the richly annotated edition of Hélène Brunner of the eleventh-century Somaśambhupaddhati, whose entire fourth volume is devoted to our subject. When preparing this paper, I switched back and forth between this celebrated eleventh-century account and that of the Niśvāsa­tattva­saṃ­ hitā noting the many points on which they differed. Since the Niśvāsa­tattva­ saṃhitā’s account of pratiṣṭhā is still unpublished, it may be useful to offer first a summary that the reader may wish to consult from time to time, after which we shall draw attention to some points in which the prescriptions of the Niśvāsa­tattva­saṃ­hitā differ from those of later works.

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


S ummary of the Account of pratiṣṭhā Given in the Guhyasūtra of the Niśvāsa­tattva­saṃ­hitā


1:29–35 List of siddhikṣetras where one should set up a liṅga. 1:36–55 liṅgas of 5 classes of materials (metal, gemstones, clay, stone, wood). 1:56 Without a liṅga, no siddhi is possible. 1:57–72b liṅgas of various materials, with measurements. 1:72c–78b Eight varieties of stone liṅga (differentiated by colour). 1:78c–81 The shape of the liṅga: square at the bottom, octagonal above that, then hexadecagonal, then round, with the top rounded like a hen’s egg or umbrella. 1:82–4 Best and worst dimensions. 1:85–90 Things to avoid, in particular round marks (maṇḍala) of certain colours that are visible when the stone is cut, which bear odd names, typically of animals:

 1. yellow (māñjiṣṭha)   2. & 3. [text damaged]   4. kapila   5. red (aruṇa)   6. molasses-coloured (guḍavarṇa)   7. grey (kāpota)   8. white   9. mottled (citra) 10. black (nīla)  11. tilacitra

— — — — — — — — — —

[frog] [text damaged] rat chameleon rock (pāṣāṇa) gecko sand (vāluka) bees (bhramara) flies (makṣikā) crow, vulture, dove

1:91–7 Other special characteristics and signs. 1:98–102b Devī recapitulates that the liṅga should be square at the bottom, octagonal in the middle and round at the top, then asks about the nirvacana, about the dimensions and form, about the pedestal (piṇḍikā), and about the installation. 1:102c–6 The pedestal (pīṭha) [of the throne of worship] is square [and is] Ananta. There are 4 legs (Dharma, Jñāna, Vairāgya and Aiśvarya) and 4 cross-struts (the 4 yugas).5 1:107 Above these is the lotus with the 9 śaktis [on its 8 petals and calyx]. 5  For a text, translation and discussion of this passage, see Goodall 2011: 238–239.



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 performance [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 maṅgalauttering brahmins and laying down the plan (sūtrapāta) to the accompaniment 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:

East: Ānanda, presided over by Indrāgni, linked with Viśākhā and Rikṣa; South: ??, presided over by Kālamṛtyu, and linked with ??; West: Puṣpadanta, presided over by Aryamādeva, linked with Uttarā & Phālgunī; North: Bhalvāṭa, presided over by Saumya, and linked with Mṛgarikṣa. 2:45 Dimensions of the doorway. 2:46–7 lines are laid from NE to SW and from SE to NW to determine and then avoid the brahmasthāna and a pit is dug where the brahmaśilā is placed. 2:48 The ‘cruel’ days of Tuesday, Sunday and Saturday should be avoided. 2:49–60 Best and worst days for installation. 2:57c–60 Inauspicious events after which installation should be avoided.

On Image-Installation Rites in the Early Mantramārga


2:61 In front of the prāsāda, 2 maṇḍapas6 are constructed out of sacrificial woods. 2:62 A new canopy and standard should be furnished [for each?], and pillars (?) and 8 pots placed in each. 2:63–65 The pots, ideally of gold (or silver, copper or earthernware) should be variously filled: 2 with pañcagavya, 2 with jewel-water, 2 with mixed fruit, 2 with scented water, 2 with gold and jewels, 2 with siṃhī and vyāghrī,7 2 with flowers, 2 with ṛddhi and vṛddhi.8 2:66–67b One should prepare a bathing stand (snānapīṭha), a bed, a whisk, umbrella, bell, mirror and archway[s]. 2:67c–68 The Eastern archway should be of Palāśa wood, the Southern of Udumbara, [the Western of Aśvattha, and the Northern of banyan]. 2:69–70 Standards with bright flags should be raised for the 8 Lokapālas. 2:71 Bringing the liṅga into the pavilion for bathing (snānamaṇḍapa) on a cart and setting it on the bathing-stand. 2:72–73b Bathing the liṅga with various substances. 2:73c–75 Marking lines on the liṅga with a golden stylus. 2:76 One [Who?] should give a cow, bangles, ornaments, cloths to the Sthāpaka.9 2:77–80 Verses to be recited by Bahvṛca, Adhvaryu, Chāndoga and Atharvan.10 2:81–82 The liṅga should be placed beside the fire; the life-cycle rites (saṃskāra) for the fire should be performed: digging, removing [earth?], sprinkling, purification with 5 saṃskāras, after inserting the fire and installing the Brahmā. 6  Other accounts speak of just one; but perhaps one of this pair is the snānamaṇḍapa mentioned 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 presumably 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).



2:83–5 Fire rites with the brahmamantras (pavitraiḥ), with the mahāvyāhṛtis and with the virūpākṣa. 2:86 The hands [of the officiant?] are warmed in the fire and on the bottom, middle and top (the base of the palm, base of the fingers and tips of the fingers?) he touches them with the brahmavyāhṛtis.11 2:87 Laying on (the term nyāsa is not used) of Vedic mantras [on to the liṅga] by touching, beginning with the ‘feet’ and shanks. 2:88–95b Then knees, genitals, waist, navel, belly, hands, skin, orifices (?), sandhis, neck, lips, teeth, tongue, nose, ears, 3 eyes, brows, forehead, head, buddhi, ahaṃkṛti. 2:95c–96b He should install (vinyase[t]) the cosmos from Kālāgnirudra up to Śiva [on to the liṅga]. 2:96c–99b He should worship the infinite sakala Śiva according to the intructions of earlier books of the text (pūrvasūtravidhānena), lay [the liṅga] on a bed and offer all manner of offerings. One should remain awake while it rests (śayyāgate liṅge).12 2:99c–101 Laying threads from NE to SW and from SE to NW he should make a cross, to mark the brahmasthāna, which he should avoid, and make the piṇḍikā there: he should put the brahmaśilā in the pit, then join? (sandhayet) a square (caturasraṃ). 2:102–104b . . . he should wake the god with uttiṣṭha brahmaṇaspate,13 set him upright, anoint the square [piṇḍikā?] with milk-rice with honey to the accompaniment of madhu vātā ṛtāyate.14 2:104c–105b He should smear the pit with pañcagavya and sprinkle it; then he should make a place for the jewels and set them starting from the East.15

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 passage 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


2:105c–107b He should place 8 minerals (dhātu), beginning in the E and ending in the NE. 2:107c–108b He should place 8 types of grain in the same manner. 2:108c–109 Jewels named [text damaged], to be placed with the mantra of the appropriate Lokapāla. 2:110–111 He should be unstinting with haritāla, barley and gold.16 In the middle he should place a lotus, bull, [and(?)] turtle or altar (vedi) fashioned out of gold.17 2:112–114a After a homa using the śivasaṃhitā, he should pick the god up with the mantra rathe tiṣṭhan nayati and take him, at an auspicious moment, on a circumambulation of the town (pura), having feasted brahmins and given them dakṣiṇā. 2:114c–115b An argha offering of 8 constituents is offered and the deity is caused to enter [the prāsāda]. 2:115c–116b At the right moment he should recite viśvacakṣuḥ kṛtām18 and, with vāmadeva cause [the liṅga] to join with Gaṇāmbikā [viz. the piṇḍikā]. 2:116c–118b The yoni is Viṣṇu; the liṅga is Śiva: he should insert [them] at once (sadya?) in the pit, reciting imā rudrāya. If Śiva moves on the ground or shakes or noisily slips or falls then he should begin a seven-day śānti rite. 2:118c–119b In the NE, the liṅga . . .(?) he should then join (bandhayet) the piṇḍikā reciting trātāram indram. 2:119c–120b The lower part [of the liṅga] he should place in the pit; the middle part is within the piṇḍikā; the upper part, duly marked in the middle, should be worshipped. 2:120c–126b Bricks one hand in length and 12 finger-breadths across and 4 in height should be made for the piṇḍikā. Seven shapes of piṇḍikā are described (Vāpī, Yakṣī, Vedī, Vajriṇī, Maṇḍalā, Trikoṇā and Ardhacandra). All are suitable for brahmins, Yakṣī and Vajrī for Kṣatriyas and Vāpī and Ardhacandrā for Vaiśyas and Śūdras. A square piṇḍikā without a mekhalā is called a Vāpī; a square one with three mekhalās is a Yakṣī; square with two is a Vedī; 16-cornered and lotus-shaped with two mekhalās is a Vajriṇī; round with two mekhalās is a Maṇḍalā; Trikoṇā [is triangular and] has three mekhalās; Ardhacandrā has three mekhalās. 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.



2:126c–127b He should make it of the same material (svayonim) [as the liṅga) and he should not join it the wrong way round(?). 2:127c–128 Rudra is in the liṅga, Umā in the vedī: he should join them, venerate them and perform a homa and a rite of śānti. 2:128e–129 He should bestow (dakṣayet) the property of the liṅga upon the Brahmins and Sthāpakas: horses, elephants, gold, jewels, armlets, necklaces, villages, crops, wealth, cows, goats, sheep, vehicles, houses. He will be released from all debts and sins and will become a leader of Gaṇas (gāṇeśvaratvam). Now of course every detail of this summary cries out for explanation or commentary, 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): hiraṇyena tejasā cakṣur vimocayet “tejo’si” iti 13 liṅge cen nivartate, cakṣuṣor abhāvāt 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


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) literature, 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.



contrast, appears only to use the expressions pīṭha and vedī (with one uncertain 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 prescribed 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 impossible 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

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.



vāmadevena śvabhre tu yojayitvā gaṇāmbikām yonir viṣṇuḥ śivo liṅgaṃ sadya śvabhre niveśayet 2:116 imāṃ rudrāya24 vai japtvā pṛthivyāṃ vai śive punaḥ calite kampite vāpi saśabde bhramite tathā 2:117 patite ca tathā śāntiṃ divā saptaṃ yadākramet īśānyām āśṛte liṅge sarva[siddhi] . . . 2:118 trātāram indra25 japtvā tu piṇḍikāṃ bandhayet tataḥ adhobhāgaṃ nyase śvabhre piṇḍikasthan tu madhyataḥ 2:119 With the vāmadeva-mantra he should place the Mother of the Gaṇas [viz. the pedestal?] in the pit. The “womb” (yoniḥ) is Viṣṇu; Śiva is the liṅga. One should at once26 lower [the liṅga] into the pit. He should recite imā rudrāya. If Śiva [i.e. the liṅga should move or tremble in the earth, or should slip noisily or should fall, then pacification [will take place] when one has passed seven days (?).27 If the liṅga settles [pointing] in the North-East, then all siddhis . . . He should then fix the piṇḍikā28 while reciting [the mantra] trātāram indram. He should place the lower part [of the liṅga?] in the pit, resting on the piṇḍikā (?) in the middle. Certainly there is no unambiguous presentation in the Niśvāsa­tattva­saṃ­hitā of the sliding of a stone ring down over the liṅga, nor is there mention of an aśvattha-leaf-shaped representation of the vulva on the inside of the ring, nor is there any mention of the showing of the liṅgamudrā, a gesture that suggests sexual intercourse, immediately after this solemn joining of the liṅga with the piṇḍikā. I mention these absences particularly since Brunner’s article ‘The 24  imā rudrāya = Ṛgveda 1.114.1. 25  trātāram indraṃ = Ṛgveda 6.47.11. 26  Or could sadya be intended to mean ‘[while uttering] sadyojāta’? 27  The text and interpretation are uncertain here. One expects the officiant to be instructed to perform expiatory rites, as in this example that also involves a period of seven days (Uttara-Kāmika 31:51c–52b): liṅge bere ca pīṭhe vā svedo rāṣṭrabhayaṃ bhavet/ prakṣālyāstrāmbunābhyarcya śāntiṃ saptāham ācaret. ‘If sweat appears on the liṅga, on an image, on the pedestal, then there is danger to the realm. One should wash it with water empowered with the astra and one should perform a pacification-rite for seven days.’ 28  Or could these words be interpreted rather to mean ‘He should then bind [it to] the piṇḍikā’? Apart from this difficulty, it is not clear whether this piṇḍikā is the same object that was lowered into the pit at the beginning. If it is, then it might be being lowered again after the pacification-rite. If it is a different object, in other words, if it is the collar, then what is it that is about to be described as piṇḍikastham?

On Image-Installation Rites in the Early Mantramārga


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 central 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 29  An exception may occur in 2:112ab, a damaged half-line: śivasaṃhita(homa) . . . devaṃ samutkṣipet, which appears to refer to a mantra-group, the śivasaṃhitā. 30  Perhaps one could argue that the absence of a moment of instilling Śaiva tantric mantras is connected with the way tantric pūjā is constructed, on each occasion involving an invocation of the mantras into the locus worshipped, followed by a dismissal at the end, which might in turn be related to an early tendency to perform worship using perishable liṅgas of earth/clay and such; but then again, perhaps such a structure of pūjā developed in part because the sādhakas of the early Mantramārga typically appropriated readyinstalled liṅgas. 31   Diwakar Acharya has spoken of the parallels in the use of mantras in the Niśvāsatattvasaṃhitā and the Svāyambhuva-pañcarātra in unpublished lectures given at ‘Early Tantra’ Workshops in Pondicherry in 2009 and in Hamburg in 2010 and his edition of the hitherto unpublished Svāyambhuva-pañcarātra has since appeared from Pondicherry in 2015. But a full investigation of this question requires an edition of the account of pratiṣṭhā in the Guhyasūtra, particularly since the labels or incipits of Vedic mantras are especially often corrupted in transmission and so often difficult to identify.



permanent temple-building (prāsāda). This, she suggests, is in part because the ācārya has a larger rôle in getting ready the maṇḍapa and in part because every part of the maṇḍapa and each of the objects inside it is charged with symbolism and has some part to play in the ritual. But she cannot rid herself of the suspicion that the temporary yāgamaṇḍapa might simply be older than the temple as a place of worship. And indeed the construction of this yāgamaṇḍapa, with four wooden toraṇas of Palāśa (E), Udumbara (S), Aśvattha (W) and Vaṭa (N) seems to be a constant among old pratiṣṭhā rites, whether Śaiva (Guhyasūtra 2:67–68; Kiraṇatantra 56:26–27) or Vaiṣṇava (Svāyambhuva-pañcarātra 5), and it has antecedents in the Gṛhyapariśiṣṭas, particularly Atharvavedapariśiṣṭa 21.4:4–21.6:6 (for which see Einoo 2005a: 13–20). We may note that prescriptions about the personnel involved in an installation have changed between the Guhyasūtra and most known subsequent Śaiva literature. Brunner observes that the subject of the injunctions in the SP4 appears also to be the ācārya/deśika, who may be referred to as the sthāpaka. In numerous passages of the Niśvāsa­tattva­saṃ­hitā too, for instance the sthāpakalakṣaṇa passage (Guhyasūtra 2:2–10), it is plain that the sthāpaka is the ācārya. But in three places the word is used in the plural (1:77, 2:18, 2:128). Here it is possible that these extra sthāpakas are similarly qualified individuals who help. But in SP4, as well as in almost all other literature (beginning with the Sarvajñānottara, Kiraṇa, Devyāmata, Mataṅga [kriyāpāda 13:30]), such individuals are called mūrtipa or mūrtidhara, a category absent from the Niśvāsa­tattva­saṃ­hitā. Brunner (1998: x) plausibly assumes that their name is chosen because they correspond to the eight forms of the aṣṭamūrti, which may also be called mūrtīśa or mūrtipa and which are homologised with the ‘feet’ (pāda) of the temple (where there are nine of those rather than five). This would seem wholly convincing were it not for the fact that the Sarvajñānottara once mentions a group of eight mūrtidharas (19:88–89), but appears to include no rite of placing the ‘feet’ of the temple (padādhvara).32 This placing of ‘feet’, 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


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 involving the vāstupuruṣa, the spirit occupying the place selected for construction. The Niśvāsa­tattva­saṃ­hitā nowhere mentions the vāstumaṇḍala and the veneration 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 construction. 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): caturasram paścimaṃ bhāgaṅ kṛtvāṣṭāsran tu madhyamam 1:78 ṣoḍaśāsran tataḥ pūrvaṅ kṛtvā caiva tu varttulam kurkkuṭāṇḍasamaṅ kuryāc chatrākāra-śiro ’pi vā 1:79 dakṣiṇe unnataṅ kuryāt sarvakāmaphalapradam śāṇagomayaghṛṣṭan tu vālukāvālarajjubhiḥ 1:80 gomūtreṇodakenaiva prakṣālya caiva yatnataḥ

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!’



• 78c paścimaṃ] KpcW; paścimad N; paścime Kac • 79c kurkkuṭāṇḍa°] W; kukkuṭāṇḍa° K • 80d °rajjubhiḥ] conj.; °rajju(?)bhiḥ N; °rakṣabhiḥ K; °ranubhiḥ W

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. So-called bāṇaliṅgas and ‘self-born’ liṅgas do not follow prescriptions regarding their shape, but for man-made liṅgas, excluding the obviously exceptional mukhaliṅgas, almost all textual sources of the Mantramārga known to us describe only variants of the classical form,34 in which the lowest part of the liṅga is invariably square in section, the middle part invariably octagonal and the uppermost circular, and this is indeed the formula that the Guhyasūtra reflects a little later in the text, for instance in a question addressed by the goddess to Śiva (1:98c–100): caturasram adhaḥ kin tu aṣṭāsraṃ kin tu madhyataḥ 1:98 ūrdhvan tu varttulaṅ kin tu kiṃ vā lakṣaṇam ucyate kathaṃ līyej jagat sarvaṃ liṅge caiva carācaram 1:99 liṅgasya deham ākhyāhi pramāṇaṃ rūpam eva ca piṇḍikā ca bhavet kin tu kiṃ vā tasya tu mekhalā 1:100 The goddess spoke. What is the square part at the bottom? What is [ultimately] the octagon in the middle? Above that, what is the circular part? What is said to be their nature? How is the whole universe, composed of moving and

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


unmoving beings, dissolved in the liṅga? Tell me the body, dimensions and form of the liṅga. What is [ultimately] the pedestal? And what is its girdle? 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 several disparate and incomplete sources. Collated here are: a fragmentary ninthcentury 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 twelfthcentury 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.) tatra sarvasamaṃ śreṣṭhaṃ suvṛttonnatamastakam sarvadoṣavinirmuktaṃ sarvalakṣaṇasaṃyutam caturaśram adhobhāgam {aṣṭāśraṃ madhyataḥ sthitam 19:52 vṛttaṃ caiva tadūrdhvaṃ tu} prakurvīta ṣaḍānana

• 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


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

Among them, the best is one that is regular throughout, with a rounded high head, devoid of all faults and equipped with all [required] characteristics; he should make its lowest section square, {the one situated in the middle octagonal, and [the one] above that round,} o Skanda. suviśuddhasya liṅgasya kukkuṭāṇḍasamasya tu 19:53 yavamadhyaṃ tathā kuryāc caturaśram adhaḥ punaḥ paripūrṇataraṃ madhye adhordhvaṃ tu kṛśaṃ bhavet 19:54

• 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, underneath, rather fuller in the middle and thin at the top and bottom. aṣṭāśraṃ vā prakurvīta sacchatrākāramastakam samāśraṃ nirvraṇaṃ śuddhaṃ tejaskaṃ cārudarśanam 19:55

55b sacchatrākāramastakam] conj.; macchatrākāramastakam N; suvṛttaṃ kāramastakam T; suvṛttā­kāra­madhyakaṃ P; suvṛttākāramastakam M; supratākāramadhyakaṃ Te; supratārākāramadhyakam L (unmetrical) 55c nirvraṇaṃ] PM; nivraṇaṃ NTL; nipraṇaṃ Te 55d cārudarśanam] NPM; dāru­da­rśanam T; cārudarśanaḥ Te L

Or he may make it octagonal [throughout], with a head the shape of a parasol, with regular angles, without blemishes, pure, bright, beautiful to behold. mṛdaṅgaṃ kāmadaṃ jñeyaṃ mṛdaṅgākārasaṃsthitam caturaśram adhastāt tu taṃ liṅgaṃ chinnamastakam 19:56

• 56a mṛdaṅgaṃ kāmadaṃ jñeyaṃ] TPM; mṛdaṃgā kāmadāṃ jñeyaṃ N; mṛdaṃgakāmataṃ jeyaṃ Te L • 56b °saṃsthitam] N; °mastakam TPM; °mastakaḥ Te; ˽ stikaḥ L • 56cd] N; omit. TPM Te L • 56c adhastāt] conj.; adhastās N



The [liṅga known as the] mṛdaṅga-drum, which is fashioned in the shape of a mṛdaṅga, is known to bestow wishes: square [in section] underneath, that liṅga is one [the upper curve of] whose head is cut off.35 In the above passage, we seem to have four types of liṅga described: the classical 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 perhaps 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 prismatic liṅgas were installed in Pallava times, for their high polish and geometrical regularity 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 photograph, 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 templecompound 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

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.

Figure 4 and 5  Egg-shaped liṅgas, each with a square base, (cf. Sarvajñānottara 19:53c–54) of the pre-Angkorian period, kept respectively in the Ta Keo Museum and the Kampong Cham Museum, Cambodia. Photos: Dominic Goodall.




a tendency away from, full naturalistic curves towards geometricisation), but that is occasionally visible on early mukhaliṅgas: we reproduce an example from Badoh Pathari (see Figure 8), where the 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. We have not commented much on mukhaliṅgas until this point, which may surprise those who expect to see liṅgas with four faces considered as a reflection of the Mantramārgic notion that Śiva’s five brahmamantras are his five faces (the fifth and upper face being typically not represented). But it seems clear, firstly, that, even if they may come to reflect this notion (for an example, see 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 is plainly intended to be Aghora because of its exorbitant pupils and fangs), the earliest known examples do not (the heads of the one illustrated in Figure 8, for example, cannot easily be matched with the characteristics typically associated with the brahmamantras), and secondly that the earliest Saiddhāntika sources are not interested in mukhaliṅgas at all: the Niśvāsa­tattva­saṃ­hitā and Sarvajñānottara appear not to mention them, and the Svāyambhuvasūtrasaṅgraha, as we have seen above (21:15), mentions them only to specify that a sādhaka should not adopt one. It is only in later literature, beginning perhaps with the Kiraṇatantra (c. 8th century?), that they are included as possible options for someone installing from scratch. The following passage follows largely the testimony of the earliest surviving source for the Kiraṇatantra, a Nepalese manuscript dated to 924 AD (=N), since the Devakoṭṭai edition (=Ed) breaks off in the middle of a list and omits more than twenty verses from the chapter: layaṃ gacchanti bhūtāni saṃhāre nikhilāny ataḥ tena liṅgam iti proktaṃ sūkṣmatvāl liṅgam ucyate 51.2 trividhaṃ tat samākhyātam avyaktaṃ prathamaṃ bhavet dvitīyam uditaṃ vyaktaṃ vyaktāvyaktaṃ tṛtīyakam 51.3 procyate liṅgam avyaktaṃ pratimā vyaktam ucyate mukhaliṅgaṃ dvirūpaṃ syād avyaktaṃ pūrvam ucyate 51.4

• 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

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.



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

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.

‘both-manifest-and-not-manifest’. The ‘non-manifest’ is what is [commonly called] the liṅga; a [sculpted] image is said to be the ‘manifest’; the one that is both at once (dvirūpam) is the mukhaliṅga. First [in what follows], the ‘non-manifest’ will be taught. We should observe that the Sarvajñānottara’s account of liṅga-shapes includes one label that perdures for centuries, namely the ‘hen’s egg’ (kukkuṭāṇḍa), along with shapes that sound similarly curvaceous: ‘parasol’ (ātapatra), ‘halfmoon’ (khaṇḍendu), ‘cucumber’ (trapusa). But in later texts, beginning again perhaps with the Kiraṇatantra (51:53–58 [=Ed. 28–33]), such names are treated as labels for flat-sided and flat-headed shapes that differ only in the breadth of the short curve that links their vertical sides to their horizontal heads. This is well illustrated by Bhatt, Filliozat and Filliozat 2005, whose Figure 9 in volume 5



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 tripartite 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 consistently 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


edition is still quite unpolished. Aghoraśiva’s commentary is patchy and often makes no remarks about passages for which the interpretation and constitution of the text are doubtful. Parallel passages will no doubt help to elucidate many opaque lines, but several of these are in other unpublished sources and in the Kiraṇatantra, which as we have seen above, is often unreadable in the Devakoṭṭai edition, and so a great deal of further collation will be needed before the required parallels can be reconstructed. The pratiṣṭhā-rite that the Sarvajñānottara describes seems to be extremely simple, omitting, as we have mentioned above, numerous elements that we encounter in other sources, such as the vāstumaṇḍala and the construction of temporary maṇḍapas, but it must be acknowledged that it is impossible to reconstruct fully much of the ritual intended just from the instructions that it gives, partly because it does not give a linear narrative account, but rather hops back and forth, alluding, for instance, to the tracing of the marks three times, in 19:20, 19:51 and in 19:77–84, and mentioning optional alternatives. It may be one of the earliest accounts to include a few iconographical requirements (19:35–37) for the temple. 1.1 A Summary of Chapter 19 of the Sarvajñānottara 19:1–2 The definition of śivālaya and the installation (sthāpana) of the liṅga (in which everything is pratiṣṭhita) are announced. 19:3–8 Characteristics of the ācārya who may be a good sthāpaka and of those who should be avoided. 19:9–14 He should know the purifications (sthānaśuddhi, ātmaśuddhi, dravyaśuddhi, mantraśuddhi and liṅgaśuddhi) and perform them. 19:15–16b The limits of the sacred site (kṣetra) for liṅgas established by ācāryas (100 hastas) and ṛṣịs (500) and for a self-born liṅga (1000). 19:16c–17 Generating the kṣetra, appropriating the land (bhūparigraha), placing the [first] stone (śilānyāsa), and measuring out the lie of the temple (sūtrayitvā). 19:18 Ornamenting [the land] with gold, jewels and seeds, proclamation of an auspicious day (puṇyāhajayaśabdaiḥ) with musical instruments. 19:19 Worship of Vighnavināyaka(s?), diśābali,37 worship of Lokapālas. 19:20 Veneration of the ācārya and sthapati and rendering the liṅga endowed with the requisite marks (salakṣaṇam) by the ācārya, together with craftsmen (śilpin). 19:21 The sthapati may make the liṅga to fit the shrine proportionally, or vice versa. 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.



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 astraempowered 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


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 middle; 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!



19:59–64 Relative proportions of pedestal, door and temple. 19:65–66 Rendering the vedī smooth and constructing an uttaravedī (?). 19:67–73 More on relative proportions of liṅga, vedī, garbha, and other parts of different types of temple. 19:74–75 Fashioning of a classical tripartite liṅga. 19:76 The lowest part should be placed in the brahmavedī (?); the second in the pīṭhikā (?) and the third, the part that is worshipped (arcābhāga), should have the marks. 19:77 Leaving [bare] the upper part of the third part (ūrdhvaṃ tṛtīyakaṃ bhāgaṃ tyajya), he should engrave (utkiret) the shape of [the glans of] a penis (liṅgākṛtim) with [a stylus of] gold (suvarṇena). 19:78–79 Description of the lines traced. 19:80–84 Bad consequences of not tracing the lines or poor proportions in the liṅga and in the vedī emphasised. 19:85–86 When the liṅga, vedī and bull have been made, they should be brought mounted on a cattle-yoked cart (goyuktaratham) and made to circumambulate the town (pura). 19:87–88 Once brought to the temple, they should be laid on a soft bed for a night and the ācārya should perform adhivāsana. 19:89–90 At midnight, he instructs eight mūrtidharas to enter the shrine and prepare water-filled pots. 19:91 He wakes the liṅga (utthāpya) and bathes it. 19:92 He purifies with the liṅga, vedī and bull with dhāraṇās, bathes them, adorns them. 19:93–95 Placing of the brahmaśilā in the centre [of the site]. 19:96–112 One should place a tortoise (kūrma) on top [of the brahmaśilā?] and sets of eight minerals (dhātu), seeds and jewels, ranged around it in the cardinal and intermediate directions. 19:113–114 Substitutions in case not all the minerals, seeds and jewels should be available. 19:115 He should set up (sthāpayet) the liṅga. 19:116 He should not move the liṅga once it is set still. 19:117ab He should smear namaḥ sarvātmakāya with pāyasa on the bottom. 19:117c–118 He should [mentally] install a yellow Brahmā on its square part, Viṣṇu having the colour of flax blossom on the middle part, and on the round part he should install a white Rudra.42 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


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 afterwards 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. Drawing these various threads together, we appear to find a number of archaic features that are common in the accounts of pratiṣṭhā given in both the Sarvajñānottara and Niśvāsatattvasaṃhitā: the absence of an unambiguous reference to the vāstumaṇḍala; the absence of an account of the placing of the first stones; the absence of one standardised, geometricised form for the 43  It is unclear to me what such obstacles might be supposed to be.



liṅga; and the absence of a tendency to apply tripartite structure generally to all things that must be installed. The following further features of the account of the Niśvāsatattvasaṃhitā are worth recalling: not only is the construction of temporary maṇḍapas with archways in sacrificial woods maintained from an earlier ritual tradition, but it is Vedic mantras that appear to be used throughout as an integral and efficacious part of the pratiṣṭhā rites, whereas in the rest of the Niśvāsa-corpus no Vedic mantras are used (other than as brahma­ghoṣa); and no part of the ritual seems yet to have been singled out as a defining moment, as a core that would make the pratiṣṭhā logical in its new tantric Śaiva context. What all of this suggests is not just the (now perhaps unsurprising) conclusion that liṅgapratiṣṭhā largely consists of modified borrowings from earlier non-Mantramārga rituals of pratiṣṭhā, but also that the Guhyasūtra is at an earlier stage in the process of adapting the borrowings to Śaiva needs than is reflected by any other tantric account known to us. Aside from these awkwardnesses, the Guhyasūtra arguably reflects an odd uncertainty about the purpose of the pratiṣṭhā: is the liṅga to be the focus of a public temple or is it to be for private use? Such an uncertainty is not unparallelled in Śaiva literature: typically, accounts up to and including that of Somaśambhu give hints that reveal that they are sometimes speaking of the installation of images for sādhana or for private worship (see, for example, Brunner 1998: 68–9, n. 2, commenting on . . . muktyai, bhuktyai . . .) and sometimes of the installation of images for public temples. The Kiraṇatantra is unusual in that it maintains a similar but more circumstantial distinction: it occasionally labels options that are to be followed in a private house (mandira; cf. gṛhe in Sarvajñānottara 19:45c: prāsāde vā gṛhe) in contradistinction to options that are to be followed in a temple (prāsāda); and indeed it gives directions (Kiraṇatantra 19:14ff.) for the construction of a private temple (devatāgṛha) in the north-east corner of a private house. As for the Niśvāsa­ tattva­saṃ­hitā’s account, although it includes subjects that can surely only be relevant to large-scale (and therefore probably public) worship, such as a typology of vimāna-shapes (as we have seen above, Guhya­sūtra 2:20–36 cursorily describe sixteen named temple-types), and the positioning and orientation of a temple within a town (2:39–40; cf. Sarvajñānottara 19:30–32), it is accompanied by assertions to the effect that the pratiṣṭhā is taught to meet the sādhaka’s need of a liṅga in order to begin his pursuit, by means of mantras, of supernatural powers (sādhana). The topic is introduced, for example, as follows: siddhikṣetrāṇi puṇyāni yatra devo svayaṃ sthitaḥ ṛṣidevatajuṣṭāni siddhiliṅgāni yāni ca 1:34

On Image-Installation Rites in the Early Mantramārga


eṣām ekatame sthitvā liṅgaṃ vā sthāpya yatnataḥ taṃ vidhiṃ sampravakṣyāmi śṛṇuṣv’ ekamanā priye 1:35 Holy fields of power, wherever God has established himself [as a spontaneously 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. And chapter 3 of the Guhyasūtra, which immediately follows the account of pratiṣṭhā, begins as follows: gokulākulabhūdeśe coraḍāmaravarjite nissarppakīṭavalmīke ītibhiḥ parivarjjite 3:1 devadevam pratiṣṭhāpya susahā[yo] . . . In a country rich with herds of cattle, devoid of thieves and trouble-makers, without snakes, insects, ant-hills, and free of plagues, having installed the God of gods, with [the help of] a good [ritual] assistant . . . 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 subject 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 prescriptions 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 SP4 = Volume 4 of the Somaśambhupaddhati (Brunner 1998). TAK3 = Volume 3 of the Tāntrikābhidhānakośa.




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 śrībilvāraṇ­yasthalābhijātaśrī­macchai­vā­gamā­bdhi­pā­rī­ṇaśrīrāmasvāmiśivācāryavaryavarasūnunā 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 techniques 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 éditoriale 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


Pondichéry / École française d’Extrême Orient / Asien-Afrika-Institut, Universität Hamburg. And, for the still unpublished Guhyasūtra of the Niśvāsatattvasaṃhitā: NAK MS 1–227, NGMPP Reel No. A 41/14. Palm-leaf, early Nepalese ‘Licchavi’ script. The two apographs used here, both in Devanāgarī and on paper, are assigned the sigla K and W respectively: NAK MS~5–2401, NGMPP Reel No. A 159/18, and Wellcome Institute for the History of Medicine, London, Sanskrit MS I.33. The verse and chapter numeration used is that of our edition in progress. Diwakar Acharya, Peter Bisschop and Nirajan Kafle helped Goodall to produce the first complete transcription. Bṛhatsaṃhitā of Varāhamihira. The Brihat Samhita by Varâhamihira with the commentary of Bhaṭṭotpala, ed. Sudhâkara Dvivedî. Part 2. The Vizianagram Sanskṛit Series No. 12. Benares: E.J. Lazarus & Co., 1897. Bodhāyanagṛhyasūtra and Bodhāyanagṛhyaśeṣasūtra. The Bodhâyana Grihyasutra, ed. R. Shama Sastri. Oriental Library Publications Sanskrit Series 32 55. Mysore: University of Mysore, 1920. Mataṅgapārameśvarāgama, kriyāpāda, caryāpāda and yogapāda, with the commentary of Bhaṭṭa Rāmakaṇṭha up to kriyāpāda 11:12b. Mataṅgapārameśvarāgama (Kriyāpāda, Yogapāda et Caryāpāda) avec le commentaire de Bhaṭṭa Rāmakaṇṭha, ed. N.R. Bhatt. Publications de l’Institut Français d’Indologie 65. Pondicherry: Institut Français d’Indologie, 1982. Mohacūḍottara. NAK MS 5–1977, NGMPP Reel No. A 182/2. Paper manuscript in Devanāgarī. Rauravasūtrasaṅgraha. Printed at the beginning (pp. 1–16) and as Appendice I (pp. 173–194) of volume 1 of the Rauravāgama (see below). Rauravāgama. ed. N.R. Bhatt. 3 Vols. Publications de l’IFI No. 18. Pondicherry: IFI, 1961 [reprinted 1985], 1972 and 1988. Sarvajñānottara. NAK MS 1–1692. NGMPP Reel No. A 43/12 [=N]. Palm-leaf, early Nepalese ‘Licchavi’ script. Also GOML R 16829 [=M], GOML MS D 155595 [=Te] and IFP T. 334 [=L], paper transcripts in Devanāgarī, Telugu-script and Devanāgarī respectively. The testimony of the manuscripts transmitting Aghoraśiva’s commentary (see below) has also been taken into consideration. The verse and chapter numeration used in our annotation is that of Goodall’s edition in progress. Sarvajñānottaravṛtti of Aghoraśiva. Oriental Research Institute and Manuscripts Library, Trivandrum, MS 6578 [=T]. Grantha script, palm-leaf, probably originally from Madurai; 199 folios. Also IFP RE 47852 [=P], Grantha script, paper, 202 pages. Svāyambhuva-pañcarātra. See Acharya 2015 below. Svāyambhuvasūtrasaṅgraha. ed. Veṅkaṭasubrahmaṇyaśāstrī, Mysore, 1937. Readings from the following manuscript have also been quoted: Mysore, ORI, MS P 258, a palm-leaf manuscript in Nandināgarī script (= My).

82 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 palmleaf 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 sivaïte 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 österreichischen 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


Publications du département d’indologie 86.1. Pondicherry: Institut français de Pondichéry / École française d’Extrême-Orient. Goodall, Dominic. 2004. The Parākhyatantra, a scripture of the Śaiva Siddhānta. A critical edition and annotated translation. Collection Indologie 98. Pondicherry: Institut français de Pondichéry / École française d’Extrême-Orient. Goodall, Dominic, and Harunaga Isaacson. 2007. ‘Workshop on the Niśvāsatattva­ saṃhitā: The Earliest Surviving Śaiva Tantra?’ Newsletter of the Nepal-German Manuscript Cataloguing Project 3 (January-February 2007), 4–6. Goodall, Dominic. 2011. ‘The Throne of Worship: An “Archaeological Tell” of Religious Rivalries’, Studies in History 27(2) (2011 [appeared 2013]), 221–250. Goodall, Dominic, Alexis Sanderson, Harunaga Isaacson, et al. 2015. See Niśvāsatattvasaṃhitā above. 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. Le Thi Lien. 2005. ‘Gold Plaques and their Archaeological Context in the Oc Eo Culture’, Bulletin of the Indo-Pacific Prehistory Association 25 (2005), 145–154. Mills, Elizabeth. 2011. Dating and placing early Śaiva texts through prāsādalakṣaṇa. A study of prāsādalakṣaṇa material in six early Śaiva texts: the Bṛhatkālottara Mahātantra; the Niśvāse Mahātantre Pratiṣṭhātantra; the Mayasaṃgraha, with its commentary, the Bhāvacūḍāmaṇi; and the Brahmayāmala Jayadrathādhikāra Piṅgalāmata. Unpublished thesis submitted to the University of Oxford in 2010. von Mitterwallner, Gritli. 1984. ‘Evolution of the Liṅga’. In Discourses on Śiva. Proceedings of a Symposium on the Nature of Religious Imagery, edited by Michael W. Meister. Bombay: Vakils Feffer and Simons Ltd., 12–31. Nagaswamy, R. 1989. Śiva Bhakti. New Delhi: Navrang. Sathyanarayanan, R. 2015. Śaiva Rites of Expiation. A first edition and translation of Trilocanaśiva’s twelfth-century Prāyaścittasamuccaya (with a transcription of Hṛdayaśiva’s Prāyaścitta­ samuccaya) critically edited & translated by R. Sathyanarayanan with an introduction by Dominic Goodall. Collection Indologie 127. Pondicherry: Institut Français de Pondichéry / École française d’Extrême-Orient. Ślączka, Anna. 2007. Temple Consecration Rituals in Ancient India. Text and Archaeology. Brill’s Indological Library 26. Leiden: Brill. Ślączka, Anna. 2016. ‘The Two Iconographic Chapters from the Devyāmata, and the Art of Bengal’. In Tantric Studies: Fruits of a Franco-German Collaboration on Early Tantra, edited by Dominic Goodall and Harunaga Isaacson. Collection Indologie 131 / Early Tantra Series 4. Pondicherry: Institut Français de Pondichéry / École française d’Extrême-Orient / Asien-Afrika-Institut, Universität Hamburg, 181–246.



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.


The Ratnanyāsa (Placing of Gems) Ritual in the Devyāmata, an Early Śaiva Pratiṣṭhātantra 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 discovered in libraries of North India and Nepal and they are, for the major part, unedited and not translated. This situation is gradually improving, with several 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 treatises 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 century 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_005



and art from regions outside South India and to explain works of art and monuments 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 overview of Devyāmata’s chapters and, finally, I will discuss the ratnanyāsa ritual, which I will bring in relation to similar passages in other texts. 1

The Date of the Devyāmata

As mentioned above, the North Indian Śaiva pratiṣṭhātantras, including the Devyāmata, are earlier than the South Indian tantras and āgamas, i.e. earlier than the 12th century CE. But how much earlier and what place does the Devyāmata occupy within this group? A date of about the 6th or even the 5th century has recently been proposed on the basis of the chapters dealing with architecture,3 but the date might not be valid for the entire text, whose old core was presumably modified and extended over the centuries. The Devyāmata has survived in three manuscripts written in Newari script. The oldest one (manuscript A in the present article; fig. 1) was completed on the third day of the month of Māgha in the year Nepala Samvat 180, which corresponds to 1060 CE.4 The two remaining manuscripts (B and C) are slightly later. Manuscript B dates perhaps from the 12th century, to judge from palaeography, but its colophon is missing.5 Manuscript C is dated to the equivalent of 1136 CE. There exists also an apograph of this manuscript, in Devanāgarī script, dated 1912, at the Wellcome Library in London.6 Attributed citations from the Devyāmata are found in the early 11th-century Bhāvacūḍāmaṇi, a commentary on another early Saiddhāntika text, the Mayasaṃgraha.7 This demonstrates that the 3  See Elizabeth (Libbie) Mills (PhD thesis, Oxford 2011). 4  N AK MS 1–279, NGMPP A41/15. I am obliged to Prof. Alexis Sanderson for sharing this manuscript with me. 5  Manuscript B: NAK MS 5–446, NGMPP A41/13. I am obliged to Dr Dominic Goodall for sharing this manuscript with me. 6  Manuscript C: NAK 1.1003, NGMPP B 27/6. I want to thank Dr Elizabeth Mills for sharing this information, and the manuscript C, with me and to Dr Dominik Wujastyk who helped to locate the apograph in the Wellcome Library. 7  Sanderson 2003–2004: 441. Vidyākaṇṭha, the author of the commentary, was a pupil of the famous Śaiva scholar from Kashmir, Bhaṭṭa Rāmakaṇṭha II, who flourished between 950 and 1000 CE; see Goodall 1998: xvii–xviii.

The Ratnanyāsa Ritual in the Devyāmata

Figure 1


Manuscript A, Devyāmata. Courtesy Professor Alexis Sanderson, Oxford.

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 surviving 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–7thcentury 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.



building instructions given in the Devyāmata.13 This indicates that at least some sections of the Devyāmata might be quite early indeed. There are other indicators of an early date; one of them is the language. The Devyāmata lacks the so-called bhūtasaṃkhyās—words used instead of numbers, among the most common being veda used in the meaning of ‘four’ and ‘rudra’ in the meaning of ‘eleven’—which abound in treatises on art and architecture in a later period. In the 11th and 12th centuries the bhūtasaṃkhyās are already frequently found in works on art and ritual as can be seen, for example, in the South Indian āgamic and tantric literature, and the śilpaśāstras the Mayamata and the Mānasāra. They are also found in some North Indian texts that partly base themselves on the Devyāmata, such as the aforementioned Pratiṣṭhālakṣaṇasārasamuccaya and the Mohacūrottara.14 However, one of the strongest arguments for the antiquity of the Devyāmata is its claim that it belongs to the Niśvāsa-corpus, a set of very early Śaiva works also preserved in Nepalese palm-leaf manuscripts and thus far unedited.15 In fact, it styles itself the bījabheda, ‘the seed-recension’, of the Niśvāsākhyamahātantra, ‘the great tantra with the name Niśvāsa’.16 A plausible date for the Niśvāsa is the 5th or the 6th century CE and it might be the oldest surviving Śaiva tantra. The Niśvāsa was, in its time, widely known across the Indian subcontinent and beyond, for instance in Cambodia. Recent epigraphic research identifies the Niśvāsa-corpus with treatises mentioned in the 9th to 12th-century Sanskrit and Khmer inscriptions of Cambodia: those of Banteay 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 deities 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 passages 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


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 portion 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 iconography 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 correlation 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 heavily 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, possibly, 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.



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 characteristics are even listed in a separate short chapter. It appears therefore that the text, taken as a whole and in the form available 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. 2

The Structure of the Devyāmata

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 provided 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 20  See Goodall 2015. 21  Only manuscript C gives chapter numbers. Yet, it is just this manuscript that misses almost a half of the text. Moreover, the photographs of the palm-leaves of B and C do not follow the sequence of the text, and some folios in all three manuscripts are damaged with broken ends and lacunae, which makes comparing the chapters in the three manuscripts with each other rather time demanding. In any case, the total number of chapters should amount to 106 (the last numbered chapter in C is 104, followed by two chapters without a number). 22  The text seems to be missing already in the parent manuscript of C or it has been omitted consciously. Liṅgaparigraha, chapter 12 in A and B, is in C followed directly, on the same palm-leaf, by mūrtipā, chapter 21.

The Ratnanyāsa Ritual in the Devyāmata


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 concentrate 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). 2.1 The Installation of a Liṅga (2-[57])23 Following the introductory ‘chapter on questions’, the first chapter of this group (2) lists characteristics of a good architect, artisan and officiating priest. Next (3–5), we find a discussion on different types of liṅgas and their merit. The text speaks of seven types of liṅga depending on the material used: earth, wood, stone, metal, precious stones, flowers and fragrant substances (gandha). Each type is considered fit for worship by special categories of beings, for instance an earthen liṅga should be worshipped by the Aśvins, and so on. Further chapters (6–18) discuss the choice of the suitable material, such as different types of stone or wood, of which a liṅga can be made; the proportions; rounding the liṅga’s top (śirovartana); and incising special marks (lāñchana). It should be noted that the short chapter entitled liṅgaparigraha (12) does not speak— contrary to what one would expect—of appropriating a previously existing liṅga by a sādhaka, but warns against a liṅga being ‘out of proportion’: a liṅga too small, too large, with a too massive or too small top or bottom will cause all kinds of misfortune. This chapter, consisting only of 6.5 verses, of which the first two are missing in A, seems to be a remainder of a larger one, perhaps dealing with a genuine liṅgaparigraha. In the same way, the garbhalakṣaṇa (17) has nothing to do with a South Indian consecration ritual of garbhanyāsa (placing a consecration deposit), but simply lists all kinds of round blemishes to be found on stones. Next (19–20), the characteristics of the piṇḍikā and the brahmaśilā are explained. These are followed by the aforementioned very brief chapter on the mūrtipās (21), of which there should be four or eight. It is interesting that this information is given here, and not at the very beginning of the text, when other ritual performers are introduced (a later insertion?). The chapter commences, surprisingly, with the otherwise quite common statement that equals the liṅga to Parameśvara and the piṇḍikā to Umā:

23  Chapter numbers not given by C are my own addition and therefore written in square brackets.



piṇḍikā tu umā devī liṅgaṃ syāt parameśvaraḥ / tayor yaḥ kriyate yogaḥ sā pratiṣṭhāvarānane //3//24 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 installation 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 geographical 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]. 2.2 Iconography [58–63] This short section begins with listing the characteristics of a mukhaliṅga [58]. General proportions of images are given in [59], while [60–61] describe in detail the iconography of various gods and demons.26 The section ends with a warning against all kind of errors (doṣa) possible while making statues, their dire consequences [62], and the mantras that should accompany (presumably the placing of) the piṇḍikā [63].

24  Cf. for instance, Somaśambhupaddhati IV.I.1. 25  It is not certain whether we have here a separate chapter or just a section within a larger one. A and C seem not to include it (in A it might be because of manuscript damage), B reads: iti sruksruvalakṣaṇa, without the usual ‘paṭala’ at the chapter end. 26  For the edition of the last two, see Ślączka 2011a and 2016.

The Ratnanyāsa Ritual in the Devyāmata


2.3 Architecture (64-[106])27 The third and last group of chapters concerns the choice, preparation and ritual purification of the soil (which is suitable as a construction ground for a temple); determining the geographical directions by means of a gnomon and a number of cords (a common method found in various texts); digging the foundation; measurements for various parts of a temple; and the characteristics of different types of temples. The only chapter dealing with ritual is the śilānyāsa (84), which describes the ceremony of placing the first stones or bricks in the foundation pit. 3

The Ratnanyāsa Chapter [37] in the Devyāmata

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 3.1 Synopsis of the Ratnanyāsa Ritual as Described in the Devyāmata29 1–4. Offering of water (argha) from the water-vessel (arghapātra); purification of the ‘jewel-cavities’ (ratnadhārāṇi randhrāṇi)30 by sprinkling the pit (śvabhra) with the astramantra and ‘Śiva-water’ (śivavāri); covering the pit and the surface of the brahmaśilā with cloth; placing the darbha-grass on [or around] the pit;31 anointing the pit and the brahmaśilā with sandal-paste. 5. Having offered incense, the ācārya accompanied by the mūrtipās should begin the ratnanyāsa by depositing a handful of gold.

27  See Mills (PhD thesis). 28  Such cavities can be seen on several stones excavated from temples in South and Southeast Asia. For some examples of stones meant to hold consecration deposits (of a slightly different type), see Ślączka 2007. 29  Numbers in the synopsis refer to verse numbers; verse numbers are not given in the manuscripts, the numbering therefore is my own addition. The synopsis is based on mss A and B (as mentioned above, in C almost the entire section on pratiṣṭhā is missing). 30  Presumably the cavities on the surface of the brahmaśilā. Cf. Dīptāgama 20.244a: ratnadhārīṇi randhrāṇi. 31  Or: ‘scratching the pit with darbha’; both manuscripts are damaged on this point, but the verb used is probably ul-likh. Cf. also Pratiṣṭhālakṣaṇasārasamuccaya 16.29ab: darbhair ullikhya saṃprokṣya śivena śivavāriṇā.



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), however, 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 cognate 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


(sindūra), pyrites (mākṣika), and red chalk.37 When there is no collyrium available, 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, sandalwood, 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 everywhere 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 identical 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 (mentioned by the Devyāmata in the next verse):  kūrmākṣmāṃ merum ukṣāṇaṃ padmaṃ vā lakṣaṇānvitam /  eṣām ekatamaṃ madhye nyasen brahmaśilāvaṭe //



is made of eight metals or otherwise of silver, gold or copper and should be placed in the brahmasthāna.44 The bull should be made in gold or silver, the Meru in silver, the Earth in gold or silver. Or one should make a tortoise of gold or copper.45 31. The lotus and so on should be placed, together with the minerals, precious stones, grains and herbs in the ekabrahmasthānas (?).46 32–34. In the absence of the [prescribed] ingredients, one should place a ruby and gold with pādara. Or, in absence of all [these], one should place vajratāla,47 gold, sahadevā plants and barley grains. Or, together with the lotus, one should place gold and kṛsara.48 A liṅga established in this way will bring prosperity, welfare and success. 35–37. Together with the bull one should place wheat, gold, sarpi, dūrvāgrass, rocanā-orpiment and earth [touched by] cow’s horn. One should place the bull in the middle of the brahmā-stone, facing the door. One should place [it?] above the brahmaśilā, having made. . . .49 When the liṅga is established in this way the four varṇas rejoice, the quadruples thrive, everywhere there is plenty of grain. 38–40. Together with the Meru one should place madhūka-plants, unhusked grain and collyrium, for the sake of glory and prosperity of the land. Together with the Earth one should place gold and all kinds of grains. When the liṅga  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


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 happiness 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. 3.2 What is Ratnanyāsa? Before analyzing the ritual as given in the Devyāmata it is perhaps good to clarify what ratnanyāsa really is. The term commonly refers to the ceremony of placing a consecration deposit, which consists of gold and small objects of symbolic value, in the garbhagṛha, below the pedestal of an (iconic) image52 or a liṅga. Among items to be deposited, precious and semiprecious stones figure most frequently, and they give the name to the entire ceremony: ratna-nyāsa means ‘the placing of gems’ or ‘jewels’. The location—below the pedestal—is important as it distinguishes ratnanyāsa from numerous cognate rituals where similar items are placed in different parts of the temple under construction. To name only a few, during prathameṣṭakā- and mūrdheṣṭakānyāsa rituals, whose function is to ceremonially mark the commencement and completion of construction, a small deposit is laid among the foundation stones and ‘crowning’ stones respectively; garbhanyāsa concerns placing a deposit container called ‘embryo’ (garbha) in the base of the temple, usually near the door jamb; ṣaḍādhāra deals with depositing six objects, one upon the other, in the middle of the foundation pit. Further, small deposits, consisting only of gold or a few precious stones, may be placed beneath the threshold, inside or below the ‘jar’

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.



(kalaśa, a pot-shaped finial) installed on top of North Indian shrines, and in numerous other locations within the temple. 3.3 Sanskrit Texts that Mention Ratnanyāsa From among various consecration rituals, ratnanyāsa is the one most frequently covered by texts, which testifies to its popularity.53 Unlike other ceremonies, it is included in texts from all over India.54 It is also one of the earliest consecration rituals or, in any case, one of the earliest described.55 It is mentioned in the 6th-century Bṛhatsaṃhitā, the late-7th-century Guhyasūtra of the Niśvāsatattvasaṃhitā,56 the c. 10th-century Mohacūrottara and the Pratiṣṭhālakṣaṇasārasamuccaya, whose oldest manuscript dates 1160 CE. It is also found in a number of relatively early purāṇas, for instance the Garuḍa, Matsya, Viṣṇudharmottara, and the eclectic Agnipurāṇa.57 In later texts on ritual, it seems almost mandatory, and it is encountered in both Śaiva and Vaiṣṇava works.58 On the other hand, it is not described in the majority of the śilpaśāstras, including those from the South: the Mānasāra, the Mayamata

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


and the Śilparatna59 that base themselves on the South Indian āgamas, and in the āgama-śilpaśāstra Kāśyapaśilpa.60 3.4 The Term Ratnanyāsa As stated above, the term ratnanyāsa, although literally meaning ‘placing of gems’, is commonly used as reference to the entire ceremony, also when other objects are deposited. The term is not encountered in the earliest texts: the Bṛhatsaṃhitā (which seems logical as no precious stones are deposited here) and certain purāṇas,61 but it might have already been used, in the broader meaning, in the Guhyasūtra of the Niśvāsa.62 The term is usually not employed by Sanskrit treatises in reference to other ceremonies. This is different in translations and secondary literature in general where ratnanyāsa may refer 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 consecration 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 tortoise, 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 support 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 visible 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.



to any consecration ritual with terms ratna- and garbhanyāsa often used interchangeably.63 3.5 Development of the Ratnanyāsa as Reflected in the Textual Sources The earliest descriptions of what in later tradition became known as ratnanyāsa tend to be very concise. The Bṛhatsaṃhitā dedicates one verse to the ceremony and prescribes only a single piece of gold to be deposited in the cavity in the piṇḍikā (here probably just a synonym of pīṭha, the pedestal). The passages in the Garuḍa, Viṣṇudharmottara and Agnipurāṇa 60 are almost equally concise.64 Items other than gold, such as gems, seeds, and so on, are mentioned, but not specified. Contrary to this, texts dedicated to ritual, which have a lengthy section on pratiṣṭhā, often contain long lists of objects to be deposited, usually divided into ‘categories’, i.e. precious stones, grains, herbs, minerals, and so on. The objects within a single category amount to eight or nine, so that they can be distributed over the cardinal and intermediary directions and in the centre. They are placed clockwise, beginning with the east or the northeast (or sometimes with the centre), to the accompaniment of mantras. The longest descriptions name appropriate geographical directions for each and every object. The centre may receive additional items, which stand on their own, and may, or may not, allude to the sectarian affiliation of the temple. An image of a bull is, for instance, usually placed in temples dedicated to Śiva. A tortoise, on the other hand, does not necessarily allude to Viṣṇu (as seen in the Devyāmata itself), but should rather be interpreted as a support of the universe.65 Such elaborate lists are found already in the 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 rituals, 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 affiliations, 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


Pratiṣṭhālakṣaṇasārasamuccaya, the majority of the Vaiṣṇava Pāñcarātra and Vaikhānasa saṃhitās, the South Indian āgamas and tantras, and in the paddhatis of Īśānaśivagurudeva and Somaśambhu.66 The descriptions in the relatively early Guhyasūtra and the Matsyapurāṇa occupy a middle position between the earliest-and-shortest passages on the one hand, and the lengthy 11th-century-and-later passages on the other. Although not very elaborate, they do list groups of minerals, grains and precious stones, all of them specified, and the additional objects to be deposited in the centre.67 Correspondences between the Ratnanyāsa in the Devyāmata and in other Texts The chapter in the Devyāmata follows the usual pattern for the ratnanyāsa descriptions, resembling most closely those of the Pratiṣṭhālakṣaṇa­ sārasamuccaya, the Dīptāgama and, to a lesser extent, the Matsyapurāṇa. Similarities with the Pratiṣṭhālakṣaṇasārasamuccaya and the Matsya were already noticed in the chapter on the iconography of the deities and concerned both the contents and the language (Ślączka 2011a, 2016). Roughly one-fourth (about twelve verses) of the Devyāmata’s ratnanyāsa chapter is enclosed in Pratiṣṭhālakṣaṇasārasamuccaya 16, albeit in a different sequence and displaying a number of minor differences of the type usually found in various manuscripts of the same text, such as substituting caiva with atha, etc. Variances are also frequent in verses listing lesser known plants, minerals and other ‘ingredients’ to be used in the ritual, for instance: saugandhī/saurāṣṭrī, gorika/gairika, viṣṇukrāntā/viṣṇusaṃjña. The amount of simple mistakes in both texts appears similar, at least in this chapter, and judging on the bases of the known manuscripts. Further, some verses from the Devyāmata correspond to Dīptāgama 20, but their number is much smaller than in the case of the Pratiṣṭhālakṣaṇasārasamuccaya.68 It is interesting to note that some of them are only given by Dīpta manuscripts C2 and D.69 In general, it can be said that the ratnanyāsa in the Dīpta is slightly shorter and better structured than 3.6

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 mentioned, 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.



in the Devyāmata. Correspondences with Matsyapurāṇa 266 are restricted to verses listing the minerals, grains and herbs.70 Somaśambhupaddhati (Part IV.2 and IV.3) also prescribes similar objects to be deposited, especially those intended for the central cavity. As we have seen, in the Devyāmata these are: a lotus, a bull, the Meru (the cosmic mountain), the Earth and a tortoise. One should choose one of these (anyatama) and deposit it together with other, also specified, objects. The Somaśambhupaddhati gives a choice of the same five figures.71 Besides, in case of some of them, the accompanying objects are also identical, for example the Meru should be deposited together with the madhūka-plants, unhusked rice and collyrium; the lotus with gold and kṛsara. This is worth noting for, while images of tortoises are frequently prescribed for rituals of consecration,72 and the bull is something to be expected in a temple of Śiva, Meru and the Earth are mentioned in a much smaller pool of texts.73

The Date of the Ratnanyāsa Chapter

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 remember 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


tantric/āgamic literature.74 The ratnanyāsa in the Devyāmata is more elaborate than in the earliest available texts: the Bṛhatsaṃhitā, the Guhyasūtra and certain purāṇas. Of course, the purāṇas and the Bṛhatsaṃhitā are not ritual treatises sensu stricto, while the Devyāmata is, and therefore devotes a significant amount of text to all aspects of pratiṣṭhā. But the fact that the description here is more elaborate than that of the Guhya, a text on ritual, is certainly meaningful.75 As we could observe, Devyāmata’s ratnanyāsa section shows closest correspondences with the Pratiṣṭhālakṣaṇasārasamuccaya, the Dīptāgama, the Matsyapurāṇa and the Somaśambhupaddhati, which, in my opinion, would situate it somewhere between late 7th and early 11th centuries CE—a much later date than the 5th-6th century proposed for the chapters dealing with architecture. In fact, applying the latter date to the entire text would make the Devyāmata’s section on pratiṣṭhā earlier than its equivalent in the Guhyasūtra, which does not seem plausible. The Devyāmata is, of course, a compilation: a text that has constantly been revised and reworked, with certain passages added and other altered. It may therefore be possible that the section on architecture on the one hand, and the sections on liṅgapratiṣṭhā and iconography (which show certain mutual similarities in language)76 on the other hand, have their origin in two distinct sources: the section on architecture from a text cognate to the Bṛhatsaṃhitā, perhaps an early one, the two others from a (later) treatise on ritual.77 It should be stressed, again, that the (presumably early) chapters on architecture seem to describe temples of Gujarat, while the (presumably later) passages on 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) similarities 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 foundation] 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?



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 necessary 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 mentioning 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 consecration 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 mysterious ‘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


(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 earliest Saiddhāntika treatises and the later, 11th and post-11th century Śaiva texts, such as the Somaśambhupaddhati and the South Indian āgamas and tantras. Appendix

Chapters of the Devyāmata80

– praśnapaṭala [1] – sthāpakasthapatilakṣaṇapaṭalaḥ 2 – liṅgayonipaṭala 3 – ratnajaliṅgalakṣaṇapaṭalaḥ 4 – pārthivaliṅga 5 – vanapraveśa 6 – vṛkṣalakṣaṇa 7 – pāṣānalakṣaṇa 8 – vanādhivāsapaṭala [9]; present in C but no chapter number given – vṛkṣagrahaṇa 10 – liṅgasya bhāgamānalakṣaṇa 11 – liṅgaparigraha 12 – liṅga[yātrā] [13] – anyaddigliṅgaparigrahavidhāna [14] – liṅgavartana [15 in manuscript B] – śirovartana [16 in manuscript B] – garbhalakṣaṇa [17 in manuscript B] – lāñcchanalakṣaṇa [18] – piṇḍikālakṣaṇa [19] – brahmaśilālakṣaṇa [20] – mūrtipālakṣaṇa 21

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 – 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 – – – – –


pratimālakṣaṇe hitāhitalakṣaṇaparīkṣā [62?] pra[ṣth]āyāṃ samudāyanāmamantravidhānalakṣaṇa [63?]; for: pra[ti]ṣṭhāyām? jīrṇoddhāravidhāna 64 vṛṣabhalakṣaṇa 65 pratiṣṭhātantre digvibhāganirdeśaparikṣāyāṃ nagaragrāmakheṭakāyāṃ digvi[ve]śa 66 – bhūparikṣā 67 – hastalakṣaṇa 68 – bhūpramāṇa 69 – vanayātrā 70 – bhūmiviśuddhilakṣaṇa 71 – śaṅkulakṣaṇa 72 – dik[śo/sā]dhana 73 – kīlakasūtrakīlakaropaṇakhātapramāṇalakṣaṇa [74] – cayalakṣaṇa 75 – vāstuvibhāga 76 – vāstudehe devatāsthānanirdeśa 77 – vāstudevābhidhāna 78 – śir[ā]marmaparijñāna 79 – śalyoddhāra 80 – sūtrapāta 81 – vāstuyāga 82 – khātodiksūtrāvaraṇalakṣaṇa 83 – śilānyāsa 84 – mūlapādapīṭhalakṣaṇa 85 – pīṭhoparidiksūtrapāta [86] – prāsādāvayava 87 – prāsādabhūvibhāgavartana 88 – nāmaprāmāṇalakṣaṇa 89 – garbhabhittipramāṇaniyama 90 – prāsādaviśeṣalakṣaṇa 91 – dvāralakṣaṇa 92 – prāsādadravya 93 – prāsādānāṃ viśeṣeṇa sāmānyacayalakṣaṇa 94 – prāsādānāṃ samudāyena viśeṣalakṣaṇa 95 – prāsādodvasana 96 – guruśiṣyaviśeṣaṇa 97 – śāstraviśeṣaṇa 98 – cchandha/cchaddha/cchanda 99 – bhūmibandha 100

108 – – – – – –


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


Bṛhatsaṃhitā of Varāhamihira, ed. and trans. M. Ramakrishna Bhat. 2 vols. Delhi: Motilal Banarsidass, 1981. Dīptāgama, ed. Marie-Luce Barazer-Billoret, Bruno Dagens and Vincent Lefèvre with S. Sambandha Śivācārya. Vol. 1. Pondicherry: Institut Française de Pondichéry, 2004. Garuḍapurāṇa of Maharṣi Vedavyāsa, ed. Ramshankar Bhattacharya. Varanasi: Chowkhamba Sanskrit Series Office, 1964. Hayaśīrṣapañcarātra, ed. Bhuban Mohan Sānkhyatīrtha. 2 vols. Rajshahi: Varendra Research Society, 1952 and 1956. Īśānaśivagurudevapaddhati of Īśānaśiva Gurudeva, ed. T. Gaṇapati Sāstrī. 4 vols. Delhi: Bharatiya Vidya Prakashan, 1988. Jayākhyasaṃhitā, crit. ed. with an Introduction in Sanskrit by Embar Krishnamacharya. Baroda, 1931. Kasyapajñānakāṇḍa (Śrīmadvaikhānase Bhagavacchastre Kasyapajñānakāṇḍa; Kāśyapa saṃhitā), ed. R. Parthasarathi Bhattacharya. Tirupati: Tirumalai-Tirupati Devasthanams Press, 1948. Kāśyapaśilpa, ed. Kṛṣṇa Rāya Vajhe. Poona: Ānandāśra-mamudraṇālaya, 1926. Kāmikāgama, pūrvabhāga, ed. Svāmināthaśivācārya. Madras: South India Archaka Association, 1975. Kāśyapaśilpa, ed. Kṛṣṇa Rāya Vajhe. Poona: Ānandāśramamudraṇālaya, 1926. Kriyādhikāra (Bhṛgusaṃhitā), ed. Vaikhanasa Agama Pandit. Tirupati: TirumalaTirupati Devasthanams Press, 1953. Matsyapurāṇa, ed. Dvaipayanamuni. Poona: [s.n.], 1907. Mayamata, ed. and trans. Bruno Dagens. 2 vols. New Delhi: Indira Gandhi National Centre or the Arts, 1994. Marīcisaṃhitā (Le Temple Selon Marīci: Extraits de la Marīci-saṃhitā, étudiés, édités et traduits par Gérard Colas), ed. Gérard Colas. Pondicherry: Institut Français d’Indologie, 1986. Pādmasaṃhitā, ed. Seetha Padmanabhan, R.N. Sampath . . . [et al.]. Madras: Pancaratra Parisodhana Parisad, 1974–1982. Paramasaṃhitā (of the Pāñchatantra), ed. and trans. S. Krishnaswami Aiyangar. Baroda: Oriental Institute, 1940. Pratiṣṭhālakṣaṇasārasamuccaya, ed. Dāmodaraśarman and Bābukṛṣṇaśarman. Kathmandu: Nepāla Rājakīya Puastakālaya, 1996 and 1998. Rauravāgama, ed. N.R. Bhatt. 3 vols. Pondicherry: Institut Français d’Indologie, 1961, 1972 and 1988. Śatapathabrāhmaṇa with extracts from the commentaries of Sāyaṇa, Harisvāmin and Dvivedaganga, ed. Albrecht Weber. Varanasi: Chowkhamba Sanskrit Series, 1964. Śilparatna of Srī Kumāra, ed. T. Ganapatisāstrī. Vol. 1. Trivandrum: Government Press, 1922.



Samarāṅgaṇasūtradhāra of Bhoja, ed. T. Ganapati Sastri, revised and ed. Vasudeva Saran Agrawala. Baroda: Oriental Institute, 1966. Somaśambhupaddhati, ed. and trans. H. Brunner. 4 vols. Pondicherry: Institut Français d’Indologie, 1963, 1968, 1977 and 1998. Tantrasamuccaya of Nārāyaṇa with the commentaries Vimarśinī of Śaṅkara and Vivaraṇa of Nārāyaṇaśiṣya. Part 1, ed. V.A. Ramaswami Sastri. Trivandrum: Government Press, 1945. Tantrasamuccaya of Nārāyaṇa with the commentaries Vimarśinī of Śaṅkara and Vivaraṇa of Nārāyaṇaśiṣya. Part 3, ed. Dr K. Raghavan Pillai. Quilon: S.R.V. Press, 1962. Viṣṇudharmottarapurāṇa. Third khaṇḍa, ed. Priyabala Shah. Baroda: Oriental Institute, 1958, 1961. Viṣṇusamhitā, ed. M.M.T. Ganapati Sastri. Delhi: Sri Satguru Publications, 1990. Viṣvaksenasaṃhitā, ed. Lakshmi Narasimha Bhatta. Tirupati: Kendriya Sanskrit Vidya peetha, 1972.

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 manuscripts 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 iconography.’ 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


Goodall, Dominic. 2004. The Parākhyatantra: A scripture of the Śaiva Siddhānta. Pondichéry: Institut Français de Pondichéry; École française d’Extrême-Orient. Goodall, Dominic. 2005. The Pañcāvaraṇastava of Aghoraśiva: A twelfth-century South Indian prescription for the visualization of Sadāśiva and his retinue / an annotated critical ed. by Dominic Goodall [et al.]. Pondichéry: Institut Français de Pondichéry; École française d’Extrême-Orient. Goodall, Dominic. 2015. The Niśvāsatattvasaṃhitā: The earliest surviving Śaiva tantra. Volume 1. A critical edition & annotated translation of the mūlasūtra, uttarasūtra & nayasūtra, ed. Dominic Goodall in collaboration with Alexis Sanderson and Harunaga Isaacson, with contributions of Nirajan Kafle, Diwakar Acharya and others. École française d’Extrême Orient, Nepal Research Centre, French Institute of Pondicherry. Goudriaan, Teun. 1965. Kāśyapa’s book of wisdom (Kāśyapa-jñānakāṇḍaḥ): A ritual handbook of the Vaikhānasas. The Hague: Mouton and Co. Kane, Pandurang V. 1946. History of Dharmaśāstra. Vol. III. Poona: Bhandarkar Oriental Research Institute. Kramrisch, Stella 1946. The Hindu Temple. Calcutta: University of Calcutta. Matsubara, Mitsunori. 1994. Pāñcarātra saṃhitās and early Vaiṣṇava theology. Delhi: Motilal Banarsidass. Mills, Elizabeth. 2011. Dating and placing early Śaiva texts through Prasādalakṣaṇa, the characteristics of temples. PhD diss., Oxford. Unpublished. Mills, Elizabeth. 2016. ‘Bhūtasaṃkhyās as a dating tool in Pratiṣṭhā literature.’ In Tantric studies: Fruits of a Franco-German project on early tantra, ed. Dominic Goodall and Harunaga Isaacson. Pondicherry: Institut Français de Pondichéry / Ecole française d’Extrême-Orient / Asien-Afrika-Institut, Universität Hamburg, 247–260 Monier, Williams, M. 1899. A Sanskrit-English dictionary: New edition, greatly enlarged and improved. Delhi: Sri Satguru Publications. (Orig. pub. 1872). Rao, T.A. Gopinatha. 1914–1916. Elements of Hindu iconography. Vols. I–II. Madras: The Law Printing House. Rastelli, Marion. 1999. Philosophisch-theologische Grundanschauungen der Jayākhyasaṃhitā: mit einer Darstellung des täglichen Rituals. Wien: Verlag der österreichischen Akademie der Wissenschaften. Sanderson, Alexis. 2001. “History through textual criticism”. In Les sources et le temps: Sources and time. Colloquium, Pondicherry, 11–13 January 1997, ed. F. Grimal. Pondicherry: Institut Français de Pondichéry; École française d’Extrême-Orient, 1–47. Sanderson, Alexis. 2003–2004. ‘The Śaiva religion among the Khmers: Part I.’ Bulletin de l’École française d’Extrême-Orient 90–91, 349–463. Singh, Nag Sharan. 1983. The Matsyamahāpurāṇam: Text in Devanāgarī, translation and notes in English. Vol. II. Delhi: Nag Publishers.



Ślączka, Anna A. 2007. Temple consecration rituals in ancient India: Text and archaeology. 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 heritage: 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.


The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas Marie-Luce Barazer-Billoret My essay is based on the core of the āgamic corpus, that is to say the twentyeight main Śaivasiddhānta āgamas (mūlāgama)—as far as they have been preserved 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

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_006



section of āgamic literature concerns temple construction and installation ceremonies of divine images that are to ‘live’ in that temple. 1

The Various Stages of Building a Temple

At the beginning of our āgamic sources is often found an interesting ready-made expression. It announces that the treatise will deal with the ceremonies starting from ploughing and ending with installation (karṣaṇādi-pratiṣṭhāntam).1 The first stage of that set of ceremonies, the ploughing, is transposed from the construction ritual of Vedic altar, which started with the opening up of a furrow. It is clear that the whole process of building and consecrating the temple, as well as installing its divine inhabitants, comprises several Vedic reminiscences. It underlines the continuity of ritual processes, which have been more often updated than fully renewed. Before the ploughing ceremony, the choosing and examination of the building site of the temple have been performed. Then, there are the orientation and purification ceremonies ending by the foundation. This last step comprises the placing of the ‘first bricks’ (prathameṣṭakā) and a foundation deposit (garbha). Then the temple is built. At the end of the work, a new set of bricks (‘top bricks’, mūrdhneṣṭakā) is placed at the top together with a new deposit upon which finally rests the finial (stūpī, stūpikā). Thus the temple building work is carried out between the placing of two sets of bricks with corresponding deposits. The first of them marks the founding of the temple and the last its completion. Both ceremonies are very similar. The first one is performed for a temple as well as for each of the several buildings and annexes surrounding it (pavilions, galleries, chapels and sub-shrines, enclosure walls, entrance pavilions, etc.). The placing of the ‘top bricks’ seems to be reserved to buildings provided with a finial (such as a temple, most of the chapels and some types of pavilions and entrance pavilions). Between these two ceremonies, ritual activity at the temple building site seems to be scarce. However, it may be necessary, due to technical reasons, to sometimes place the liṅga before building the walls of the sanctum (when for instance the liṅga and its pedestal are too big to be introduced into the already built sanctum). In that case it is necessary to carry out the full installation ceremony of the liṅga and from then on to perform regularly its daily worship, the stone object having been turned into Śiva. From a practical point of view, archaeological restoration works carried out on in India as well as in South-East 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


Asia have shown that intermediary deposits were regularly placed at various levels of the temple elevation, between the foundation and crowning deposits, which suggests that corresponding ceremonies were performed. However, such deposits and ceremonies are not mentioned in the treatises.2 2

Images and their Installation

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 sanctum of a Śiva’s temple behind the liṅga. The pattern it describes roughly corresponds 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 similar 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 dismantled, transferred and rebuilt or various rebuilt or renovated temples at Angkor (Cambodia). 3  See our translation of the Rauravāgama, Introduction pp. xlvi–il.



manifestations: their number varies according to treatises. For instance, they are fourteen in the Rauravāgama, sixteen in the Dīptāgama (which includes Astra in the list) and four, five or eight according to the Rauravottarāgāma. Such variations are well in accord with what may be seen in temples: small temples have a limited iconographical program, while the important ones tend toward exhaustiveness. Some mūrtis of Śiva show the god surrounded by secondary deities who may be subjected to a separate—even if summary— installation. According to the Dīptāgama, the installation of Vaivāhyamūrti (Śiva’s wedding manifestation) includes special installation ceremonies for Brahmā, Viṣṇu, Śrī and Agni who each get a separate statue provided with an individual pedestal, and who are thus considered to be full-fledged divine entities. It is interesting to note that such a (brief) individual installation is prescribed by the Dīptāgama even for Apasmāra, in spite of his negative aspect, most probably due to his role of mount of dancing Śiva (vāhanākhyam apasmāram, Dīptāgama 39.24). The mythological story relating to a specific manifestation of Śiva may justify some details of his installation. As far as the Goddess as Śiva’s consort is concerned, she may be subject to a full installation ceremony performed at the same time than Śiva’s, such a double ritual ending with a marriage. However, it may happen that the installation of a pair such as Someśvara needs a double ceremony performed in two places and by two sets of officiating ritual specialists. Śiva’s images are not the only ones found in a Śiva’s sanctuary: the statues of his attendant deities must also be installed, the main ones being those of Vṛṣa, the bull, and of Mātṛkās, Gaṇeśa, Skanda, Caṇḍeśa, as well as Viṣṇu, Brahmā, Durgā, Jyeṣṭhā or Sūrya. 3

The Installation Ceremony

An installation ceremony comprises a long preparatory stage, which contrasts its shorter, sometimes very short apex. It has a double aspect in that it has both material and immaterial elements. Its aims are to turn a man-made object into a divine being as well expressed in the title of the work published by our Japanese colleagues Shingo Einoo and Jun Takashima (2005). I might add that we will not deal here with the installation ceremony of self-generated (svayambhu) images, the installation of which remains problematic. The material part of the installation ceremony is by no means insignificant and is often difficult to interpret. Its prescriptions begin with the making of the iconic or symbolic images and are especially concerned with iconography and iconometry. Later on during the ceremony itself sculptors are summoned to implement rules related to the drawing of lines on the liṅga (liṅgoddhara)

The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas


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. 3.1 The Installation of Main Images In a paper published in 1993–94 and also in other papers, I have dealt with the different stages of an installation ceremony and therefore I shall not describe them here. Instead, I intend to concentrate on the idea that, in each treatise, the installation ceremony follows more or less the same general pattern or model. In my opinion, the stage of the installation ceremony with the most obvious ritual pattern is the performance of the ‘sojourn’ (adhivāsa) in the installation pavilion. During that stage, the image and its pedestal (and sometimes the liners to be placed under the image, when it is a liṅga) is to sojourn in the sacrificial pavilion specially built for its installation ceremony. That pavilion will also shelter fire-pits (kuṇḍa) for the fire ritual as well as vases in which the mantras of the gods have to be kept before being transferred in the image. When the image is brought to sojourn in that pavilion it is not yet turned into the god and the question is to know if that change of nature is to occur during that sojourn ceremony or on the next day, when the installation itself takes place. The answer seems to vary from one text to another. A first pattern is found in some treatises, for instance, in the Rauravāgama (Chapters 28 and 30) and the Kiraṇāgama (Chapter 56). In these treatises, the turning into a god occurs clearly on the installation day at the time of the transfer of the mantras from the vase to the image; in that pattern, the sojourn ceremony performed the day before is nothing more than a lengthy sanctification of the image. In other texts, the presentation is confused and it is difficult to see when exactly the transformation occurs. Lastly, there are texts where, obviously, the turning of the image into a god happens during the sojourn ceremony, that is to say a day earlier than it is prescribed in the Rauravāgama and the Kiraṇāgama. In such texts, the sojourn ceremony receives a long and detailed description, but these works do not omit the transfer of the vase mantras on the installation day either—whose transfer they often call ‘soul imposition’ (jīvanyāsa). It is interesting to note that such a pattern is also found in a late chapter of the Rauravāgama dealing with the Nāgarāja installation.4

4  See Rauravāgama Kp.57.117 in N.R. Bhatt edition, Vol III, p. 114 and ibidem p. 115, fn 16, a quotation of Aghoraśivācārya’s Śivapratiṣṭhāvidhiḥ which follows the same pattern (see also Brunner 1998: 162).



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 corresponding 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 criterion: 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. 3.2 The Installation of Images of Various Other Categories In some works, there is a chapter dealing in a general way with the installation of various types of images which may be ‘mobile or immobile’ (cala-acala) with however some hints at the differences in the performing of the installation of each of the concerned categories. A good example of such a presentation is found in Ajitāgama where the general chapter ends by a list of these differences (see chapter 40: Pratimāsthāpanavidhiḥ, especially vv. 124–132 in N.R. Bhatt edition). Here we learn for instance that when an image is made of clay the ‘sojourn in water’ (jalādhivāsa) is to be omitted (interestingly, some other texts prescribe to perform it on a mirror or a similar object). In the same way Ajitāgama prescribes that the placing of jewels as an installation deposit (ratnanyāsa) is to be omitted when the image is painted while, in the case of a clay image, that ratnanyāsa is to be performed at the time of making its wooden frame-work (śūla). It has also been prescribed that when an image is a mobile (cala) that is to say when it is made of bronze, it must be taken in a procession at the end of the installation ceremony (Ajita 40.122–123). This is interesting in that it means the bronze image which is par excellence the form of the god aimed to be taken to a festival and other processions is “to come into office” as soon as it has been turned into that god. 3.3 Model and Function The bronze images are mainly used as necessary substitutes for the liṅga in some precise circumstances while the images placed on temple façades are more generally purely decorative. Such subordinate functions may explain why the rules regarding the installation of such images are most often simpler

The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas


than those to be applied to the installation of the liṅga. However such ‘simplification’ 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’ ceremony the liṅga is to lie upon a couch with the head to the East; there several 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, especially 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 transferred 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 worshipped 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 sanctification 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.



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 elements 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 pattern 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.

Conclusion: The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas

Returning to the title of our essay, the importance of pratiṣṭhā ceremonies in Śaivāgamas, we can say that it refers to two levels. The first concerns the huge number of installation ceremonies described in these texts, and the second expresses the relevance of these ceremonies for establishing the inner coherence of a specific treatise and its dating. In these works, which are mainly kriyāpādas, descriptions of installation ceremonies are numerous due to the fact that many installations are necessary for temple worship, as seen before. We may also add that some high-scale ceremonies may include one or even several subsidiary installation ceremonies. It is the case of the yearly temple festival—mahotsava or brahmotsava— during which the image of the bull is ritually installed on the flag raised at the 6  See Kāmikāgama, Pūrvabhāga, chapters 64 and, about images, 66.

The Importance of Pratiṣṭhā Ceremonies in the Śaivāgamas


beginning of the festival. Most interesting is the installation ceremony of a provisional liṅga (bālaliṅga), performed either before the building of a temple or before the beginning of an important renovation work, but in any case before the installation ceremonies regarding the temple and/or the liṅga, and most probably of other buildings, as well as of other images. The provisional installation becomes herewith an installation within the installation. In addition, we have put forward the question of ritualistic patterns, without however dwelling too much on their technical aspects, the matter being often difficult to apprehend. From a practical point of view, it needs a very careful reading of the concerned passages in order to identify often slightly sensible variations and grasp the treatise in all its completeness. Thus being the case, it is a fine tool to understand the evolving life of a treatise. A last point to be underlined regards installations as well as several other kinds of ceremonies described in kriyāpādas: it is the theological context in which they are to be performed. As already told the ritualistic sections of āgamas are very often kept apart and without the matching theological ones. Even when this matching exists, it is often not easy to draw clear correspondences between doctrine and rites. The common impression is often that there is a general pattern which is even independent from sectarian trends and which is seemingly more or less adapted to the concerned theological elements. A good example is provided by the hierarchical list of tattvas and its distribution: there are obviously discrepancies on these points between ceremonies without being possible to question the unity of the treatise or to suppose that it is a heterogeneous compilation of chapters borrowed from various sources. From my point of view—and it will be my conclusion—we are here in a field where theology is supplanted by rites. 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).



Other Editions and Studies

Barazer-Billoret, Marie-Luce. 1993–94. ‘L’installation des liṅga et images dans les temples selon les āgama śivaïtes’, Bulletin d’Etudes Indiennes, vol. 11–12 (1993–94), pp. 39–69. Barazer-Billoret, Marie-Luce, Bruno Dagens, Vincent Lefèvre and S. Sambandha Śivācārya. 2004. Dīptāgama, Edition critique: Vol. 1, Chapitres 1 à 21, Institut Français de Pondichéry (Collection Indologie n° 81.1). Barazer-Billoret, Marie-Luce, Bruno Dagens, Vincent Lefèvre, S. Sambandha Śivācārya and Christèle Barois. 2007. Dīptāgama, Edition critique: Vol. 2, Chapitres 22 à 62, Institut Français de Pondichéry (Collection Indologie n° 81.2). Barazer-Billoret, Marie-Luce, Bruno Dagens, Vincent Lefèvre, S. Sambandha Śivācārya and Christèle Barois. 2009. Dīptāgama, Edition critique: Vol. 3: Chapitres 63 à 111 (Appendice et index), Institut Français de Pondichéry (Collection Indologie n° 81.3). Brunner-Lachaux, Hélène. 1998. Somaśambhupaddati, volume IV, Institut Français d’Indologie, Pondichéry 1998 (Publication de l’Institut Français d’Indologie n°25). Dagens, Bruno and Marie-Luce Barazer-Billoret. 2000. Rauravāgama, un traité de rituel et de doctrine śivaïtes, introduction, traduction et notes par, 2 volumes, (li-640 pages), Institut Français de Pondichéry, Pondichéry, (Publications du Département d’Indologie n°89/I et 89/II). Einoo, Shingo and Jun Takashima (eds). 2005. From Material to Deity: Indian Rituals of Consecration, New Delhi: Manohar.


Studies in Dhāraṇī Literature IV: A Nāga Altar in 5th Century India Ronald M. Davidson Introduction In distinction to most of the papers in this volume, which examine the placement 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 understandably 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 spectrum 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 previous 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 scholars 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 iteration 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_007



themselves in a theological landscape, made room for the members of that horizon within their own ritual systems, and employed these systems to further their greater penetration into networks of production (Cohen 1998, Schopen 2002, DeCaroli 2004, Shaw 2004). In at least one professional discussion forum, the question has even been posed by William Bodiford as to whether it is possible for Buddhist communities to exist devoid of a similar system of interrelation with indigenous divinities of some variety, given that we do not see any successful establishment of Buddhist institutions without it (Cohen 1998: 360). In the Indian iterations, which established the paradigm, the ostensible relationship between the monastery and the cult site or cultic practices was to a great degree a cypher for the relationships within a community or between communities. Under the guise of engaging indigenous cults, Buddhist monks engaged individuals, families, landowners, tenured laborers, village priests of every stripe, indigenous sorcerers and any number of other persons in a complex network of interactions and support. As such, these interactions tended to precipitate questions of identity, religious boundaries, the lay-monastic relationship, to mention but a few of the many issues involved. Sometimes their interaction was cooperative, at others more competitive, but in most instances Buddhists seem to have understood their position as a minority tradition swimming in a brahmanical sea, especially notable during and after the period of the Gupta-Vākāṭaka dominion. This paper will examine the earliest complete presentation of a rain ritual dedicated to the nāga spirits, found in a text that went through remarkable transformations. The version of the text we will discuss may be tentatively reconstructed as the *Mūlamantra-sūtra or perhaps *Mūlamantradhāraṇīsūtra (T. 1007 牟梨曼陀羅呪經). It survives in three separate canonical Chinese translations, one Tibetan translation, and fragments found in the Gilgit manuscript corpus. As a ritual work, the earliest Chinese translation— which is attributed (with less than complete certainty) to the Liáng period (502–557 CE)—represents a late classical, pre-tantric form of Mahāyāna Buddhist ritual practice. The text is quite complex, and in the two later Chinese and one Tibetan translations, the earlier *Mūlamantra text was amalgamated with another text, the Mahāmaṇivipulavimāna-viśvasupratiṣṭhitaguhya-paramarahasya-kalparāja-dhāraṇī, whose lengthy name reflects a relatively common phenomenon of name inflation in post sixth century Indian Buddhism. The nāga rite articulates the fabrication of a nāga altar, and the ritual involvement with that altar, so that the nāga will be persuaded to bring rain. This paper will argue that the construction of the nāga altar and the form of the ceremony employed reflect three different tangents in classical Indian life:

A Nāga Altar in 5th century India


A. the method of worship of nāgas in farming communities in North India; B. the formation of a sthaṇḍila altar patterned after the optional (kāmya) practices found in the tradition of the gṛhyasūtras, particularly the later works and the supplements (śeṣa, pariśiṣṭa, vidhāna) appended to the gṛhya texts; and C. the mantra practices of sorcerers (vidyādhara), part of whose ritual syllabus dealt with snakes and snake lore. None of these source traditions individually can explain the ritual, but in aggregate they provided the raw materials for their amalgamation into the ritual system under discussion. It would appear that the Buddhists were the ones formulating the aggregate system reflected in our text, as rain rituals dedicated to nāgas endure in Buddhist communities to the present but are found almost nowhere else. In distinction, some remnants of the three processes continue in situ, an affirmation of cultural continuity in rural communities where all three have found their primary location. 1

The *Mūlamantra Text

Taishō Tripiṭaka no. 1007 is a medium-length text, but its length belies its difficulty, 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 interesting is, like other anonymously translated texts attributed to the Liáng period (502–557 CE), it incorporates into the work instructions not found in other versions 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



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 represents 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 background 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 complete 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 program 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 construction 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 controlling 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-viśvasupratiṣṭhitaguhyaparama-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 transcriptions 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


with the writing conforming to Sander’s Gilgit ms./Gupta Group B standard, thereby placing the manuscript in that era, if not slightly before (Sander 1968: 121–136). The Tibetan translation of the aggregate work is a Royal Dynastic production—attributed to the team of Vidyākaraprabha and dPal-gi lhunpo, revised by Vidyākaraprabha and dPal-brtsegs—and the aggregate title occurs in both the dKar-chag ldan-dkar-ma (or Lhan-dkar-ma, no. 335), and in the dKar-chag ’Phang-thang-ma (no. 303), the two surviving imperial catalogues. The Tibetan translation is closer to the Gilgit fragments than any of the Chinese translations, not an unusual circumstance. The Chinese title of T. 1007 is móulí màntuólúo (牟梨曼陀羅), the second part of which renders mantra, although these characters are sometimes used to represent maṇḍala. In our text, however, maṇḍala is represented as platform (壇), as is often the case in other translations, and in the current instance this can be verified by comparison to the other translations and the Gilgit manuscript.2 The other translations identify this part of the aggregated text as if it is to render something like *mūlamantra-paṭala, or *mūladhāraṇī-paṭala.3 The problem with the title of T. 1007 is not the móu, for this often renders mu/mū, as in the standard case of muni (牟尼), and so we would expect the Sanskrit to be mūla. The curiosity is the vowel in lí (梨), which does not easily render -la. Indeed, the móulí character combination is used to represent mūle in an earlier abbreviated translation of the Mahāmāyūrī-vidyārajñī (prāptamūle: 婆多牟梨), which would yield the less probable title of *Mūlemantra, a form that would raise other questions.4 One possible resolution of the phonological problem would be to acknowledge that Chinese translators sometimes transcribed -rya—a liquid with 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.



a semi-vowel—employing one or another form of the syllable li, as in the instance of ācārya (āshélí 阿闍梨) or ārya (ālì 阿唎). Thus the title of T. 1007 was possibly pronounced *mūlyamantra. This seemingly elegant solution is somewhat complicated by the presence of another use of the syllable li to render another word, súlìlúoshé (素唎囉闍 658a16), which may represent an original sūryarāja, Sun, the King. The li syllables are not the same, but are generally recognized as having similar phonological properties.5 However, it is also the case that we find Prakritic forms of sūrya, such as suriya or sūrī, both representing Ardhamāgadhi pronunciations; to complicate the phonology more, we find Sanskritic usages like saurī to refer to verses for the sun (Ṛgvidhāna 1.139) and mauli to denote a superior item. Elsewhere in the mantras, it would appear that the vowel epenthesis (svarabhakti) is employed to break up some of the consonant conjuncts, so that we find vajra represented as if it were pronounced vajira (bázhélúo 跋折囉). While some of these either represent received conventions from previous translators or simply illustrate the difficulty of Chinese to render adequately the richer phonological environment of Sanskrit, given the idiosyncratic nature of the translation found in T. 1007, there remains a possibility that the renderings attempt to record alternative, perhaps vernacular pronunciations. Consequently, there is some possibility that the móulí transcribed the Prakritic pronunciations, mūliya or mūliä for the written mūlya, and that the last vowel got lost in the transcription. Such loss of a final syllable is elsewhere seen in the transcription of Sanskrit into Chinese, most notably in the movement dhyāna > chánnà (禪那) > chán (禪), but also visible in other Chinese abbreviations of the aggressively polysyllabic Sanskrit vocabulary; in our text, for example, the name of the bodhisattva Vajrapāṇi is often abbreviated as simply Vajra. Despite this *mūlya possibility, the much greater probability lies with the supposition that the written title of the text was *Mūlamantra-sūtra or perhaps *Mūlamantradhāraṇī-sūtra. One reason is that, when *mūlya has been used in a title, as in the case of the Mūlyādhyāya-pariśiṣṭa, it has been employed to indicate funds (mūlya) for the ritualist, and this is not an issue in our text. Moreover, *mūlamantra has echoes of mūladeva, a divine enemy of Indra in the Ṛgveda (Ṛgveda 7.104.24, 10.87.2, 14), and suggesting the basis of power (mūla) of yātudhāna sorcerers in the same work (Ṛgveda 3.30.17ab, 10.87.10, 19; Sen 1968). Eventually, the character of Mūladeva will emerge as a trickster in the medieval period, the putative author of a thieves’ manual (steyaśāstra) and an expert in erotics (Bloomfield 1913). Consequently, *mūlamantra not only identifies the content of T. 1007—as the text begins with a discussion of a ‘root’ 5  Pulleyblank 1991, pp. 186, 188, provides 梨 and 利 approximately the same value, li.

A Nāga Altar in 5th century India


mantra (*mūlamantra)—it also connotes powerful ritual systems on the margins of respectability. Whatever the elusive historical processes involved, we should hew close to some iteration of *mūlamantra, a designation seen elsewhere in Indian ritual texts. In the case of the *Mūlamantra, not only is the title uncertain, but the translation is anonymous and presents something of a chronological problem. Our earliest surviving notice of the text is in Zhìshēng’s 730 CE Kāiyúan-shìjìaolù (開元釋教録: T. 2154), which lists it in several places, beginning with dhāraṇī scriptures having lost translators (539a19), which is generally where it is otherwise assigned.6 In one place, though, Zhìshēng (or whoever’s note he copied) suggests that it could be tidied up by inclusion in the body of Liáng Dynasty catalog of translations (603b7: 拾遺編入今附梁録). This is hardly the strongest of recommendations, since the reference to such a catalog (or catalog category?) is quite obscure.7 Nonetheless, T. 1007 is conformity to the other ritual texts assigned to this period and is not anachronistic with reference to the voluminous dhāraṇī translations done by Jñānagupta and Yaśogupta in the last quarter of the sixth century. Moreover, given the presence of a Gilgit manuscript and the retranslated, amalgamated text under Bodhiruci II, an early to mid-sixth century translation date for the *Mūlamantra is not out of the question. We will see that it is preceded by other rituals assigned to the early fifth century, and so—barring further information—I am tentatively inclined to accept an early to mid-sixth century date for this anonymous translation. We may suppose therefore that the text itself was composed in India over the last half of the fifth century, in the waning years of the imperial Gupta-Vākāṭaka classical era (ca. 320–550 CE). Overall, the text is primarily concerned with the employment of three mantras: the *mūlamantra, the hṛdayamantra and the upahṛdayamantra. These are identified towards the beginning of the *Mūlamantra section in all the surviving texts, with T. 1007 placing it after the nāga rain ritual, as we will see. The order of the various elements varies from recension to recension, but all of them contain a painting section, the nāga rite, a homa section, a section on mantras, one (or more) on mudrās, and a maṇḍala ritual. Interspersed are also found a number of smaller rituals, abbreviated mantra recitation rites, and numerous explanations of benefits. T. 1007 in addition has several other sections associated with it, especially those articulating various signs—wood,

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.



homa, etc.; these belong to the larger Indian literature on signs and must be considered in conjunction with those many other sources. 2

The Nāga Rain Rite

Every surviving complete recension of the *Mūlamantra contains some version of the following nāga rain ritual, with the fragmentary Gilgit recension standing as an exception. The ritual is found immediately associated with the *mūlamantra, and so is probably one of the oldest parts of the text. As already mentioned, the fullest expression of the ritual is found in T. 1007, which includes the mūlamantra itself, and may be translated as follows: 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. ii. [658b2] 先作方壇。若高臺平地隨時作之。又以青綠塗畫其壇。壇上 施龍。其壇四角各一水瓶各一香鑪。一燒薰陸香一燒栴檀香。一燒酥 合香一燒安悉香。

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 platform 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 officinalis), and in one bdellium incense (guggulu, Commiphora wightii). iii. [658b5] 又接壇外更作一壇。縱廣四肘牛糞塗上。四角安瓶。一盛 水和乳一盛水和酪。一盛水和乳粥。一盛水和脂。

Then on the outside of that platform but connected to it, make another platform [surrounding the first, the mekhalā], four cubits square, and 8  Compare uccasthāna in Amoghapāśamahākalparāja fols. 29a6–30a2, 33a5, 34a6–34b, 49b4–5.

A Nāga Altar in 5th century India


smear it over with cow dung. In the four corners [of this second platform], 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. iv. [658b8] 又於壇内燃燈八盞。又以脚極實𠼝(此云紅藍花未的)及麨莊 嚴其壇。又於四角各插一箭。又以五色線繫圍箭上。又作五色旛子懸 其箭頭。 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 corners, 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 platform, and employ five[-colored] variegated ornaments set out as offerings arranged on top of the platform. Then with the various kinds of flowers and fruit appropriate to the season, set them completely [covering] the inner platform. vi. [658b13] 於是呪師面向東。呪取白芥子滿八千粒。呪一遍打龍頭 上。如是呪打滿八千已。大雲四合即澍甘雨。從此已後一切那伽咸並 敬奉。 This done, the sorcerer (vidyādhara) faces east. Intoning mantras, he takes a full eight thousand white mustard seeds and saying mantras he throws the seeds one by one onto the head of the nāga, until he has completed the mantras and has thrown a full eight thousand times. Then a great cloud from the four quarters will saturate [the area] with sweet rain. From this time forth, all the nāgas will worship [the vidyādhara].

9  Uncertain identity. The interlinear gloss (紅藍花) indicates the safflower (Carthamus tinctorius), 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?



vii. [658b17] 若雨過多呪白芥子滿八百遍。投於有龍水中其雨即止。當 此呪時一不得語。若祈雨時疾雲暴風。苦霧卒雹乍聚乍散障其雨 者。作佉陀羅木橛打於龍淵岸側。則一切障難悉皆散去。

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 difficulties will entirely be scattered away. viii. [658b22] 若祈雨時被一切障難。令雨不得下者。當於壇内畫一毘 那藥劍。取白芥子八百粒。粒呪一遍用打毘那藥劍。如是滿足八百遍 已。能令一切作障難者。悉自被縛不復能障其雨即下。事既了已當用 乳汁洗去畫像。毘那藥劍甚大歡喜。從是已後一切所求無不如意。其 持呪者身及衣服。一切常須嚴肅淨潔。

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. ix. [658b29] 其根本曼陀羅功德略說如是。即說呪曰。 That is the summary of the qualities of this mūlamantra, and now to express the mantra itself:10

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 transcriptions 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


namaḥ sarvatathāgatānāṃ | oṃ vipulagarbhe | maṇiprabhe | tathāgatanidarśane | maṇi maṇi | suprabhe | vimale | sāgaragambhīre | hūṃ hūṃ | jvāla jvāla | buddhāvalokite | guhyādhiṣṭhitagarbhe | svāhā | 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 section 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 translation (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 earlier in the text in Chinese transcription and one siddhamātṛka text in a position 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. 3

Probable Sources for the Rite

In constructing this ritual text, the Buddhists were not spinning their performances 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 transcribes tathāgatanirdeśane; T. 1007.19.658c3–4 suggests tathatādarśane; siddhamātṛka T. 1005a.19.623b7 reads guhyādhiṣṭita. 11  T. 1005a.19.625a7–b1; T. 1006.19.641b15–c4; To. 506, fol. 301a2–b2. 12  To. 506, fol. 299a6–7; T. 1005a624a22–28.



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 evolving 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. 3.A Local or Folk Cultic Sources First, we find at the advent of the ritual a description of a specific kind of sculpture, an ephemeral one constructed from a combination of cow-dung and mud. The formula is not the same in the various recensions. Both the Tibetan and Bodhiruci II indicate it is to be made only of cow-dung, whereas Amoghavajra is quite specific on the mixture of mud with the cow-dung. The form of the figurine is interesting: a three-headed nāga. While three of the texts indicate that it is nothing more than the form of a snake, Amoghavajra specifies a combined zoomorphic-anthropomorphic form: 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 discussed 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


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: piśācaṃ mṛṇmayaṃ kṛtvā kākṣyaṃ ca dvijasattama | gandhair mālyas tathā vastrair alaṃkāraiś ca pūjayet | bhakṣyaiś ca lopikāpūpair māṃsaiḥ pānais tathaiva ca || 661 || Having fashioned an image of a piśāca of mud and leaves (kākṣya = ? pigeon pea), O best of Twice-born, it should be worshipped with incense, and garlands, clothing and ornaments, and food and sweets and cakes, meat and then water. 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 examination 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 person for destructive rites in Ṛgvidhāna 3.81–82. 15   Śrāvakabhūmi, p. 416.7: kāṣṭhāśmaśādakṛt.



fired mud images (terracotta) may be made by potters, brick manufacturers, or almost anyone. If they are fired, they could be baked in conjunction with any of the thousands of small temporary brick kilns in India, thus requiring virtually no special, dedicated resources (Huyler: 38–50). Terracotta images of nāga figurines do survive in a limited number of sites, despite being somewhat ephemeral and fragile themselves, especially if they are poorly fired. So Buxar, Ahichchhatrā, Jhūsī, Kauśāmbī, Sonepur, Basārh, Lauriyā Nandangarh, Mathura, Panna, Chandraketugarh, Kumrahar (Old Pāṭaliputra), Mangalkot (Mahasthangarh), Sonkh, Śṛiṅaverapura, and other locales throughout Bengal, Bihar and Uttar Pradesh have yielded whole or fragmentary examples of snake or nāga figurines.16 They may be in the forms of actual snakes, or an overtly anthropomorphized nāga/nāgī form, with a human upper body, as we see in Amoghavajra’s description.17 Among the most interesting are three from Kauśāmbī now in the Allahabad Museum, because they retain residue from the application of red paint, recalling the paint application in our ritual’s case. Indeed, we have seen that Amoghavajra’s translation calls for red paint on the body, and the Tibetan translation also affirms that the snake body is to be painted red.18 Unfired mud and cow dung ephemeral nāga images are additionally specified a little later, with the first examples of the nāga-pañcamī rituals mentioned in Sanskrit. Einoo (1994), who has studied these, has collected examples from the Bhāviṣyapurāṇa (1.36.62–64, 1.37.1–3,–18, 4.36.54–59), the Nāradapurāṇa (114.27), Garuḍapurāṇa (1.129.31cd-32) and the Bṛhaddharmapurāṇa (2.10.53–54).19 In the majority, the ritual admonition is to draw an image of a nāga on both sides of the door (e.g., Bhāviṣyapurāṇa 1.36.62: dvārasyobhayato lekhyā gomayena viṣolbaṇāḥ) and then make offerings to it. We note the purpose here, however, is generally alleviating fear of snake bite rather than control of rain, an issue to be revisited below. Moreover, the citations and descriptions are quite abbreviated within the Puranic literature overall, and speak of a rural tradition in which cow dung and snakes are among the most conspicuous items in the domestic environment.

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


The non-Sanskritic or oral-episteme nature of the practice is evident when we observe similar practices continuing in the 19th century, as recorded by Enthoven, based on the notes of Jackson. 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) Crooke records a variation on this theme, in which a ceremony is held at the end of the period of nāga worship, and a nāga figure is made. 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) Atkinson records a form of snake worship on nāga-pañcamī, remarkable similar to our own nāga altar. 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 brahman necessary, nor is a temple visit required, nor are any of the other brahmanical 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 temporary construction of an altar.20 In such instances, the procedures are accessible 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.



Even in the case of formal modern nāga worship in a temple, brahmans are not necessarily the ritualists. Oldham’s ethnography of 19th century nāga temples in what is now the states of Kashmir, Himachal Pradesh and Uttarakhand emphasized their non-orthodox priestly families: The priests of most of these temples, whether of the Nāgas or of the Devas, are Kshatriyas, or Khuttris as they are called in the vernacular . . . Orthodox Brahmans may sometimes be found officiating at these unorthodox temples, but this is very rarely the case. When it does occur, the position of the Brahman is but a subordinate one. (Oldham 1901: 467) 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. 3.B The Gṛhya-sūtra Contribution Almost lost in the discussions about the origin of tantrism—and the even more robust ritual invigoration of medieval India in purāṇic and other public performances—is the astonishing contribution of the gṛhya texts, the domestic works initially dedicated to the daily rituals and the rites of passage. The reason for the relative disinterest in these works has not been clear, but may perhaps be attributed to their lower status in comparison to the ‘solemn’ (śrauta) rites and to their strong connection to quotidian Indian family life. This was formally recognized with the superior status of the brahman who could perform the śrauta rituals because he kept three fires (āhitāgni), as opposed to the brahman who did not (anāhitāgni). In reality, most other twice-born males could perform many of the gṛhya rites. There are even indications, as in the Mānavagṛhyasūtra’s serpent offering (sarpabali), that the ritual could be performed by non-twice born, even non-twice born women!21 Clearly, the extension of authority outside of brahmanical status did not contribute to the stature of gṛhya rites. This, coupled with their modest costs, their relatively simple ritual program and their quotidian background all militated against their high status in the community. Yet the widespread distribution of the rites was a function of their penetration into every community with ritual 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


expectations of the twice born, so that, even if their status was low, nonetheless 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 (naimittika), 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 supplemental 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 consecration 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 rituals 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 uniformity 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 suffice, 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.



uttarato grāmasya vedyākṛtiṃ [sthaṇḍilaṃ] kṛtvā śākhābhiḥ parivāryāhataiś ca vāsobhiḥ sarvarasair ghaṭān pūrayitvā dikṣu nidadhyāt sarvabījaiś ca pātrāṇy avāntaradikṣu | Kāṭhakagṛhyasūtra 57.2 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. atha khalu yatra kva ca hoṣyant syād iṣumātrāvaraṃ sarvataḥ sthaṇḍilam upalipya ullikhya ṣallekhā udagāyatām paścāt prāg āyate nānāntayos tisro madhye tad abhyukṣya agnim pratiṣṭhāpya anvādhāya parisamūhya paristīrya purastād dakṣiṇataḥ paścād uttarata ity udaksaṃsthan tūṣṇīm paryukṣaṇam |Āśvalāyanagṛhyasūtra 1.3.1 Now wherever there may be offering, a low sthaṇḍila of about an arrow’s length on all sides [is constructed], having smeared it [with cow dung] and inscribed six lines, one [south to] north but west [of the fire location], then two on the [upper and lower] ends of that one [upper and lower facing east] and three in the middle [of those two]. Having anointed that [sthaṇḍila], he places the fire, adds fuel, wipes the ground around and then scatters [grass] around—east, south, west and north, so that he ends in the north. Silently he anoints that place. prāgudañcaṃ lakṣaṇam uddhṛtyāvokṣya sthaṇḍilaṃ gomayenopalipya maṇḍalaṃ caturasraṃ vā agniṃ nirmathyābhimukhaṃ praṇayet tatra brahmopaveśanam |Mānavagṛhyasūtra 1.10.1 (also MGS 2.2.1) Dresden’s translation: Having made a mark (by drawing lines) in northeastern direction and having sprinkled (this mark with water), having smeared the sthaṇḍila which should be circular or quadrangular with cow dung, and having kindled the fire by rubbing he carries the fire forward; this the devotion to brahman. In this latter instance, the older commentator to Mānavagṛhyasūtra allows that the sthaṇḍila is basically an equivalent designation to a maṇḍala: kim ākāraṃ limpet sthaṇḍilam ity āha—maṇḍalaṃ caturasraṃ catuḥkoṇaṃ tulyo vikalpaḥ | “What is the form of the sthaṇḍila that is to be smeared (with cow dung)? It is answered—it is considered the same as a maṇḍala with four

A Nāga Altar in 5th century India


corners, square.”24 Elsewhere, the Mānavagṛhyasūtra describes a more elaborate altar for the “long-lived horse” (dhruvāśvakalpa) ritual, probably motivated by the difficulty of maintaining equine health in South Asia: prāgastamayān niṣkramyottarato grāmasya purastād vā śucau deśe ‘śvatthasyādhastān nyagrodhasya vāpāṃ vā samīpe vedyākṛtiṃ kṛtvā tasyāṃ catuṣkoṇavanaspatiśākhāyām avasaktacīrāyāṃ gandhasragdāmavatyāṃ caturdiśaṃ vinyastodakummasahiraṇyabījapiṭikāyām apūpasrastaralājol­ lopikamaṅgala-phalākṣavatyāṃ sarvagandhasarvarasasarvauṣadhīḥ sar­ varatnāni copakalpya Mānavagṛhyasūtra 2.6.4 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 forest 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 barley-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 appropriation 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 Mahāmāyūrīvidyārājñī, 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.


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。

Now for the mantra platform of the Mahāmāyūrī-vidyārājñī: Use cow dung to smear the ground. Then employ and strew around seven colors of flowers, forty-nine flags, four knives, four mirrors, one hundred arrows, one bow and seven tall pots filled with thick broth (pāyasa? kṛsara?). Assemble string made from the wool of black goats, sixteen xin (尋 about 8 ft. in length = about 40 meters of string) long, and twenty-five crêpes. Then seven oil lamps; one vessel of curds, one vessel of parched grain and broth; one vessel of rice, one vessel of crêpes, and place one vessel of pomegranate and one vessel of flowers.25 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 contemporary 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 brahmanical ritual systems. The primary exception to this were the small images of the planets, made of various materials, found in many rituals on the pacification of the grahas, the seizers who possess and bring disease (Yano 2004). 3.C Sorcerers, Snakes and Local Spirits Perhaps the most interesting aspect of the nāga rain rite is the contribution of the local sorcerers, here identified as vidyādharas.26 Based on the anthropology of similar modern groups, it may be that this designation could be 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 cultivation 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


applied to one or another group of partially or wholly peripatetic castes. At least since the specter of the yātudhāna—sorcerers found throughout Vedic literature—analogous figures have been known in Indian literature, and continue to be identified to this day by a number of names in the modern period: ojha, sabari, baiga, jādūgar, etc. Similarly, there were several names by which these male—sometimes female—figures were addressed in the late classical period: īkṣaṇika, yātudhāna, aindrajālika, māyākāra, mantrin and, particularly, vidyādhara. They had a variety of skills exemplified in the literature, but were particularly adept at controlling spirits, casting spells of illusion, exorcism, raising the undead (vetāla), and a host of other valuable enterprises. Their influence on later tantric Buddhism was sufficiently profound for tantric Buddhists eventually to reference their discipline as vidyādhara-saṃvara and occasionally their scriptures as belonging to the vidyādhara-piṭaka (Davidson 2014b). Attempts to affiliate vidyādharas necessarily with one or another of the literary sectarian movements (Śaiva, Vaiṣṇava, Bauddha, Jaina, etc.) are misguided and misrepresent Indian religion, which has rewarded fragmentation more than amalgamation. This is not to say that they were not conscious of their own lineal authority, as some of the sorcerers certainly had their own independent traditions, as in the case of the aindrajālika of Harṣa’s Ratnāvalī. In his invocation of his gods, he pays homage to Indra and Saṃvara (Ratnāvalī 4.2), the two great magician gods of the Vedic literature; like other non-brahmanical figures he speaks in Prakrit, as do Buddhists, women and those not sanctified by Sanskritic status. The connections between these disparate figures and our ritual are several. First, we can have some confidence that the term for sorcerer or spell-master in Chinese (呪師, vi. above) is a translation of vidyādhara, since that is more accurately represented in the Tibetan translation of the same section (To. 506, fol. 301a6: rig pa ’dzin pa). Second, some of the activities explained are elsewhere associated with vidyādharas. The best example of this is the similarity between the ritual for controlling excessive rain (vii. above) and a ritual for controlling rain mentioned in the Bhaiṣajyavastu section of the Mūlasarvāstivādavinaya, part of the story of Sudhana that was also found in both the Divyāvadāna and the Bodhisattvāvadānakalpalatā.27 This part of the narrative involves a king of southern Pañcāla who wants to steal the young nāga Janmacitra from northern Pañcāla to bring favorable rain to his country. He needed the nāga’s power in order to make up for the cruelty of his own reign, because of which the rain god 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.



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. sa balyupahāravidhānārthaṃ gataḥ saptame divase āgamiṣyati | āgatyāsya hradasya catasṛṣu dikṣu khadirakīlakān nikhanya nānāraṃgaiḥ sūtrair veṣṭayitvā mantrān āvartayiṣyati | 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 division 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). atha mārayitukāmaḥ khadirakīlakaṃ kṛtvā strīpuruṣaṃ viciṃtya hṛdaye nikhanayet | kṣaṇād eva mriyate || 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

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 khadira 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: śaṅkuś ca khādiraḥ kāryo rajatena vibhūṣitaḥ | śaṅkuś caivopaveṣaś ca dvādaśāṅgula iṣyate |

Karmapradīpa 2.8.3

The stick is to be fashioned of Khadira wood, ornamented with silver. Both the stake and the fire-poke are to be 12 inches in length. 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 concerned 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 controlling 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.



iteration that had already begun to be eclipsed in brahmanical literature, which was moving to the synthetic aggregation of vināyakas with the god Hastimukha (‘elephant face’, also called Dantin ‘tusker’), and the epithet of gods like Bṛhaspati, Indra or Śiva as Gaṇeśa or Gaṇapati. Vināyakas in the Mahābhārata functioned, as a class, as in our own ritual: they, with the rākṣasas, piśācas and bhūtas caused problems at home (12.274.60d/423: vighnam kurur gṛhe) for which a prophylaxis is provided. As has been partially surveyed by previous scholars, rituals to deal with the groups of vināyakas are described in the Mānavagṛhyasūtra (2.14), the Baijavāpagṛhyasūtra, the Yājñavalkyasmṛti (I.270– 291), Bodhā­yanagṛhyaśeṣasūtra (3.10), Satyāṣāḍhagṛhyaśeṣasūtra (1.6.16),29 Sāmavidhānabrāhmaṇa (1.4.19), the Śāntikalpa of the Atharvavedapariśiṣṭa and Puranas or other texts related to the early Smārta schools in some way.30 Without a specific ritual, vināyaka is a source of obstacles mentioned in other Smārta literature, as in the Ṛgvidhāna (2.38) and the Atharvavedapariśiṣṭa (7.1.9, 20.3.1, 35.2.8). Together these diverse ritual treatments represent the Śuklayajurveda, the Maitrāyaṇīya-Kṛṣṇayajurveda, Taittirīya-Kṛṣnayajurveda, Kauthuma-Sāmaveda, Ṛgveda and the Atharvaveda branches among the Smārta schools, with the Yajurveda schools providing the most extensive sources. In view of this rich distribution, it would appear that the modern scholarly emphasis on the Atharvaveda at the expense of other Vedic branches might be reconsidered in light of the value of the domestic manuals from various branches to understand such previously unapproved traditions. Here, as in our instance, there is no iconographical description of the vināyakas in the earliest literature (e.g., Mānavagṛhyasūtra 2.14) so we do not know how the recommended drawing was to be done. There is certainly no reason to associate a vināyaka with an elephant-headed divinity in this Buddhist example, similar in that respect to the iconographical problem of the early vināyaka literature overall, which provides no iconographical directions for the demon’s representation. 29   Satyāṣāḍhagṛhyaśeṣasūtra 1.6.16 is essentially Bodhāyanagṛhyaśeṣasūtra 3.10, but there is also an earlier abbreviated form of the same ritual called the Gaṇapatipūjana in Satyāṣāḍhagṛhyaśeṣasūtra 1.3.1. 30  The Baijavāpagṛhyasūtra is quoted in Aparārka’s cy to the Yājñavalkya I.275. The nature of the relationship between the vināyaka materials and Gaṇeśa/Gaṇapati has been discussed by many scholars; important contributions include Winternitz (1898), Hazra (1948), Courtright (1985), Agrawala (1991–92), the essays by A.K. Narain (‘Gaṇeśa: A Protohistory of the Idea and the Icon’), M.K. Dhavalikar (‘Gaṇeśa: Myth and Reality’), and Ludo Rocher (‘Gaṇeśa’s Rise to Promience in Sanskrit Literature’) collected in Brown 1991: 19–83, Bolling 1913: 268–72.

A Nāga Altar in 5th century India


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 rapidly evolving Gaṇeśa persona, so that the Yajñavālkyasṃrti—while acknowledging 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 classical brahmanical strategy: the declaration of unity amidst bewildering plurality, facilitated by assigning multiple local divinities to one of the pan-Indian deities 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 traditions are incorporated into Sanskritic literature, we observe that they do not suddenly cease or evaporate from the religious landscape, as is so often presumed. The presumption is informed by the idea that high-status cults always displace, appropriate or redefine low-status ritual systems, whereas we actually 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 pursuit. 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.



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 problems, 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 indicative 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, unsurprising as some of the names eventually associated with Śiva or his retinue seem to have been understood by Buddhists as applicable to yakṣas, particularly 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. mahākālasya kaṇṭhe bandhayitavyaṃ dakṣiṇahaste saptavārā krodharājaṃ japya kruddhena capeṭan dātavyaḥ | mahākālaṃ rāvaṃ muñcati | vidyādhareṇa na bhetavyaṃ punaḥ kruddhena tāryayitavyan tato aśrūṇi pravarttayati vidyādhareṇa aśrū gṛhya tilakaṃ karttavyaḥ | adṛśyo bhaviṣyati sarvvanidhānāni paśyati | tato vidyādhareṇa puna capeṭaṃ dātavyaṃ tato mahākālaṃ rudhiraṃ vamati | vidyādhareṇa spṛsitavyam ākāśena gacchati mahāyakṣādhipatir bhaviṣyati | gugguludhūpaṃ datvā krodharājam anusmārayaṃ mahākā­ lasvarūpeṇa agratam upatiṣṭhati | yathākāmatā sarvvavarāṇi dadāti | sarvvamātṛgaṇaparivārā vaśagatā tiṣṭhanti | anyāni yathā kāmaka raṇīyāni sarvvakāryāṇi karttavya iti || tṛsūlapāśasādhanavidhisādhanam || ||33 Alternatively (the noose and trident combination already described) is to be bound to the throat of Mahākāla (statue). Having recited the 32  E.g., Divyāvadāna pp. 41–2; Suvarṇabhāsottama pp. 161.6. 33   Amoghapāśamahākalparāja ms. 30a7–b1; To. 686, ma: 54a1–55a1; T. 1092.20.258c13–259b4.

A Nāga Altar in 5th century India


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. 4

A Path not Taken: Brahmanical Rain and Nāga Rites

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 insignificant 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.



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 dimension 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 ritual 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 passing: 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


they could perform that service. In the Mārkaṇḍeyapurāṇa (16.3c–42ab) the gods express the relationship: vayam āpyāyitā martyair yajñabhāgair yathocitaiḥ | vṛṣṭyā tān anugṛhṇīmo martyān sasyādisiddhaye || 38 niṣpāditāsv auṣadhīṣu martyā yajñair yajanti naḥ | teṣāṃ vayaṃ prayacchāmaḥ kāmān yajñādipūjitāḥ || 39 ado hi varṣāma vayaṃ martyāś cordhvapravarṣiṇaḥ | toyavarṣeṇa hi vayaṃ havirvarṣeṇa mānavāḥ || 40 We are satiated by mortals, with the appropriate portions of sacrifice. We facilitate those mortals with rain, for the success of their crops, etc. And when the vegetation has ripened, the mortals sacrifice to us with sacrifices. We bestow on them their desires, propitiated by sacrifices and so on. Downwards we send rain, while mortals shower upwards— We with the rain of water, humans with the showers of ghee. While the orthodox theology of sacrificial acts has traditionally been dominated 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 formality 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 car39   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.



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 precipitation, 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 ophiolatry, it is for purposes other than rain.40 Probably because of the nāgas’ connection 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.g1), 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 underworld 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 sacrifice 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 miscopying from another rain ritual elsewhere in the chapter (Annāśāstri 27.5–6; Śarmā 2.45–46) yielded this misleading occurrence. The late medieval commentary, 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


in the Vājasaneyisaṃhitā mention rain in this context, but only seek protection through these powerful beings. In distinction to this probable error in copying, the most significant brahmanical 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 simply called the śravaṇā rite. As portrayed in the Mānavagṛhyasūtra, the motivation 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 bedding 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.



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 employing 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 placement 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 surprisingly 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


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 labhate 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 chapter 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 threeheaded) 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 specification 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 devavat). 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 kartavyam dhātujaṃ vā viśeṣataḥ.



associated with protection from snakes and the monsoon: the functions attributed 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 temples of the goddess by intrusion of the deities into a nāga’s site in the Nīlamatapurāṇa, which explains that Rāma set up a statue of the Sārṅgin (Viṣṇu) at a site of the nāgarāja Ananta: tasmād adūre puṇyodām anantasya mahātmanaḥ || 1189cd bhavanaṃ nāgarājasya tapas tepe sudāruṇam | patiṣṭhānaṃ tathā cakre tasya devasya sārṅgiṇaḥ || 1190 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 previously 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



The Buddhist Formulation

In distinction, Buddhists monks have invoked nāgas for the purposes of bringing 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 proximity 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 proximity 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 surviving 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 example 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 excavation 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 necessarily mean monastic ritual involvement or to acknowledge that the ritual agendas of the communities might have diverged.



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 perhaps 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


the name of the nāgarāja. O son of good family! There will be a transformation 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 liturgical 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 others. 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 contemporary 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 recognized 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



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 conspicuous 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 predisposition 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 inscriptions 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 supposition 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


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 observation. 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 placement 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 continuing 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 If he recites the [hṛdaya-]mantra over white mustard seeds fully 8000 times and scatters them in space before him, then he will be able to cause the rain to fall. All the nāgas together must pay homage and obey him. 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 ritual 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 behaviors 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 serpents for the purpose of bringing rain to communities of their patrons. The 54   Hevajra-tantra I.ii.20, Amoghapāśamahākalparāja fol 29b5 are later examples.



cultivation of nāgas for such weather manipulation goals proved sufficiently important that it became an essential tool for Buddhist missionary activity. But we should also remain cognizant that the nāga ritual systems exceeded rain rituals per se, and nāga supplication rituals extended to include trips to the underworld (pātāla), the search for treasure, securing supernormal powers, to name but the more important; many of these were shared with the larger Indic ritual universe. All such goals secured for rites of nāga worship/coercion a central place in Buddhist missionary activity. There are nāga shrines scattered throughout almost every culture in which Buddhism prospered worldwide. In formulating these rituals, they knit together aspects of ritual systems from the immediate environment: folk offerings to the nāgas, domestic ritual systems incumbent on every twice-born, and the vidyādhara sorcery of the families of sorcerers inhabiting north and central India since well before the time of Buddhism. Such an institutionalization had many benefits. It proposed the placement of monastic communities in the center of an alternative divine cosmos, one that claimed to bring some of the same benefits as the brahmanical cosmos, but with different tools and with different expectations. It spoke to the needs and identities of local groups, often underclass or outcaste, who would not or could not be patrons of the brahmanical worship of the gods. It elicited an emotional response of site-specific validation, for the Buddhists affirmed local spirits and Buddhist mantras rather than the pan-Indian divinities employing the Vedic mantras of the brahmanical texts. Finally, it was a vehicle for Buddhists to provide an exchange of services that did not depend on caste, undercutting the primary raison d’etre of brahmanical authority. Integrating pre-existent folk divinities into a new ritual basis, the Buddhists engaged Indian agrarian communities in a manner comprehensible, for it was based on a cosmos those communities themselves had devised. 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 encouragement 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


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-VedaSaṁhitā: 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.



Bodhāyanagṛhyaśeṣasūtra. Sastri, R. Shama. 1920. The Bodhāyana Gṛihyasutra, pp. 178– 376. Mysore: Government Branch Press. Brahmapurāṇa. Jhā, Ācārya Tāriṇīśa, ed. 1993. Brahmapurāṇam (Hindī anuvād sahit). Ilāhābād: Hindī Sāhitya Sammelan. Bṛhatsamhitā. Dvivedi, Krisnacandra. 1996. Bṛhatsaṁhitā by Śrī Varāhamihirācārya, with the Commentary of Bhaṭṭotpala. 2 vols. Varanasi: Sampuranand Sanskrit University. *Dhāraṇīsaṃgraha, 陀羅尼雜集, T. 1336. Divyāvadāna. Cowell, E.B. and R.A. Neil. 1886. The Divyâvadâna, A Collection of Early Buddhist Legends. Cambridge: Cambridge University Press. dKar-chag lDan-dkar-ma. Herrmann-Pfandt, Adelheid. 2008. Die lHan kar ma: Ein Früher Katalog der ins Tibetische übersetzten buddhistischen Texte. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Lalou, Marcelle. “Les Textes Bouddhiques au Temps du Roi Khri-Sroṅ-Lde-Bcan.” Journal Asiatique 241/3 (1953): 313–353. dKar-chag ’Phang-thang-ma. Kawagoe, Eishin. 2005. dKar chag ’Phang thang ma. Sendai: Tohoku Society for Indo-Tibetan Studies. Fānfànyǔ. 翻梵語. T. 2130. Garuḍapurāṇa. Bhattacharya, Ramshankar. 1964. Garuḍapurāṇam of Maharṣi Vedavyāsa. Varanasi: Chaukhambha Sanskrit Sansthan. Gautamadharmaśāstra. Stenzler, A.F. 1876. The Institutes of Gautama. London: Trübner & Co. Gautamagṛhyapariśiṣṭa. Sathe, Jayashree. 2004. Gṛhya Rituals of Sāmavedins (As Propounded by Gautama). Delhi: Pratibha Prakashan, 196–370. Gobhilagṛhyasūtra. Bhattacharya, Chintamani. 1936. Gobhilagṛhyasūtram with Bhaṭṭanārāyana’s Commentary. Calcutta: Metropolitan Printing and Publishing House, Ltd. Harivaṃśa. Vaidya, Parashuram Lakshman. 1969. The Harivaṃśa: Being the Khila or Supplement to the Mahābhārata. Poona: Bhandarkar Oriental Research Institute. Hevajratantra. Snellgrove, David L. 1959. The Hevajra Tantra: A Critical Study. 2 vols. London: Oxford University Press. Hiraṇyakeśigṛhyasūtra. Kirste, J. 1889. The Gṛihyasūtra of Hiraṇyakeśin, With Extracts from the Commentary of Mātṛidatta. Vienna: Alfred Hölder. Jaiminigṛhyasūtra. Caland, Willem. 1922. The Jaiminigṛhyasūtra Belonging to the Sāmaveda. Lahore: Punjab Sanskrit Book Depot. Kāiyúan-shìjìaolù. 開元釋教録. T. 2154. Karmapradīpa. Schrader, Friedrich. 1889–1900. Der Karmapradīpa. 2 vols. Halle, a.S: Buchdruckerei des Waisenhauses. Kāṭhakagṛhyasūtra. Caland, Willem. 1925. The Kāṭhakagṛhyasūtra, with Extracts from Three Commentaries, and Appendix and Indexes. Lahore: Research Department, D.A.V. College.

A Nāga Altar in 5th century India


Kātyāyanagṛhyasūtra. Mīmaṃsaka, Yudhiṣṭhiro. 1983. Kātyāyanagṛhyasūtram. Bahalagarha, Sonipata-Harayana: Ramalala Kapura Trasta. Kauśikasūtra. Limaye, V.P., et al. 1982. Keśava’s Kauśikapaddhati on the Kauśikasūtra of the Atharvaveda. Pune: Tilak Maharashtra Vidyapeeth. Mahāmaṇivipulavimāna-viśvasupratiṣṭhita-guhyaparamarahasya-kalparāja-dhāraṇī. Bodhiruci II trans. T. 1006. Amoghavajra trans. T. 1005a. To. 506. sTog 468. Mahāmāyūrī-vidyārājñī. Takubo, Shūyo. 1972. Ārya-Mahā-Māyūrī Vidyā-Rājñī. Tokyo: Sankibo. T. 982–988. Mahāmegha-sūtra. Dharmakṣema trans. T. 387. Mahāvyutpatti. Sakaki, Ryōzaburō, ed. 1916–25. Bonzō Kanwa Shiyaku taikō honyaku meigi daishū. 2 vols. Kyoto: Shingonshū Kyoto Daigaku; rpt. Tokyo: Suzuki Gakujutsu Zaidan, 1962. Mānavagṛhyasūtra. Sastri, Ramakrishna Harshaji. 1926. Mānavagṛhyasūtra of the Maitrāyaṇīya Śākhā, With the Commentary of Aṣṭāvakra. Baroda: Central Library. Dresden, M.J. 1941. Mānavagṛhyasūtra: A Vedic Manual of Domestic Rites— Translation, Commentary and Preface. Groningen: J.B. Wolters. Mañjuśriyamūlakalpa. Śāstri, T. Gaṇapati. 1920–25. Āryamanjusrîmûlakalpa, 3 vols. Trivandrum Sanskrit Series No. LXX. Trivandrum: Government Press. Mantrapāṭha. Winternitz, Maurice. 1897. The Mantrapāṭha: Or the Prayer Book of the Āpastambins. Oxford: Clarendon Press. Mārkaṇḍeyapurāṇa. Banerjea, K.M., ed. 1855. The Márcandeya Purána. Calcutta: Asiatic Society. Matsyapurāṇa. Ānandāśrama Paṇḍitas, ed. 1903. Śrimaddvaipāyanamunipraṇītaṃ Matsyapurāṇam. Ānandāśramasaṃskrtagrahthāvaliḥ no. 54. Pune: Ānandāśrama. *Mūlamantra-sūtra. T. 1007. See also the Mahāmaṇi-vipulavimāna-viśvasupratiṣṭhitaguhyaparamarahasya-kalparāja-dhāraṇī. Nāradapurāṇa. Khemarāja Śrīkṛṣṇadāsa et al. 1984. Śrīnāradīyamahāpurāṇam. Delhi: Nag Publishers. Nīlamatapurāṇa. de Vreese, K. 1936. Nīlamata or Teachings of Nīla: Sanskrit Text with Critical Notes. Leiden: Brill. Qīfǒ bāpúsà sǔoshūo dàtúolúoní shénzhòujīng, 七佛八菩薩所說大陀羅尼神呪經, T. 1332. Pāraskaragṛhyasūtra. Bākre, Mahādevaśarmā. 1917. Pāraskaragṛhyasūtram śrīkarkopādhyāya-jayarāma-harihara-gadādhara-viśvanātha-praṇīta-bhāṣyapañcakasamalaṃkṛtam. Bombay: Gujarati Printing Press; rpt. New Delhi; Munshiram Manoharlal, 1982. Ratnāvalī. Kale, M.R. 1921. The Ratnāvalī of Śrī Harṣa-Deva. Rpt. Delhi: Motilal Banarsidass, 1995. Rauravāgama. Bhatt, N.R. and Jean Filliozat. 1961–1988. Rauravāgama: edition critique. 3 vols. Pondichéry: Institut Français D’indologie. Trans. B. Dagens and M.-L. BarazerBilloret. 2000. Le Rauravāgama: Un Traité de Rituel et de Doctrine Śaivaïtes. 2 vols. Pondichéry: Institut Français D’indologie.



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 Vāgharakula­­bhūṣaṇamahārājaśrīkālanāthapraṇītayajurmañjaryākhyavyākhāsamalaṅkṛtam, 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 TherîGâthâ: (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-YajurvedaSaṁhitā. 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


Vāyupurāṇa. Āpṭe, Hari Nārāyaṇa. 1905. Mahāmuniśrīmadvyāsapraṇītaṃ Vāyupurāṇam. Ānandāśramasaṃskrtagrahthāvaliḥ no. 49. Pune: Ānandāśrama. Yājñavalkyasmṛti. Apte, H.N. 1902–03. Yājñavalkyasmṛti. Ānandāśramagranthāvali no. 46, 2 parts. Pune: Anandasrama. Yajurmañjarī. In Śuklayajurvidhāna, Annāśāstri, ed., pp. 51–199.

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.



Davidson, Ronald M. 2012. ‘Observations on an Uṣṇīṣa Abhiṣeka Rite in Atikūṭa’s Dhāraṇīsaṁgraha.’ In István Keul, ed. Transformation and Transfer of Tantra in Asia and Beyond, Berlin: Walter de Gruyter, 77–98. Davidson, Ronald M. 2014a. ‘Studies in dhāraṇī literature II: Pragmatics of dhāraṇīs.’ Bulletin of the School of Oriental and African Studies 77/1, 5–61. Davidson, Ronald M. 2014b. ‘Studies in Dhāraṇī Literature III: Seeking the Parameters of a Dhāraṇī-piṭaka, the Formation of the Dhāraṇīsaṃgrahas and the Place of the Seven Buddhas.’ In Richard Payne, ed. Scripture:Canon::Text:Context—Essays Honoring Lewis R. Lancaster. Berkeley: Institute of Buddhist Studies and BDK America, Inc., 119–180. DeCaroli, Robert. 2004. Haunting the Buddha: Indian Popular Religions and the Formation of Buddhism. Oxford: Oxford University Press. Deeg, Max. 2005. Das Gaoseng-Faxian-Zhuan als religionsgeschichtliche Quelle. Wiebaden: Harrassowitz. Duquenne, Robert. 1988. ‘Gaṇapati Rituals in Chinese.’ Bulletin de l’École française d’Extrême-Orient 77/1, 321–354. Einoo, Shingo. 1994. ‘The Nāgapañcamī as Described in the Purāṇas and its Treatment in the Dharmanibandhas.’ Journal of the Japanese Association for South Asian Studies 6, 1–29. Einoo, Shingo. 1996. ‘The Formation of the Pūjā Ceremony.’ Studien zur Indologie und Iranistik 20 [Festschrift Paul Thieme], 73–87. Einoo, Shingo. 2005. ‘The Formation of Hindu Ritual.’ In From Material to Deity: Indian Rituals of Consecration, eds Shingo Einoo and Jun Takashima. Japanese Studies on South Asia No. 4. New Delhi: Manohar Publishers, 7–49. Enthoven R.E. 1914. Folk Lore Notes: Vol. I—Gujarat. Bombay: British India Press. Fergusson, James. 1868. Tree and Serpent Worship: or Illustrations of Mythology and Art in India in the First and Fourth Centuries after Christ. London: India Museum. Gonda, Jan. 1980. Vedic Ritual: The Non-Solemn Rites. Leiden: Brill. Grafe, Jörg. 2001. Vidyādharas: Früheste Zeit bis zur Kaschmirischen Bṛhatkathā. Frankfurt am Main: Peter Lang. Gupta, Parmeshwari Lal. 1965. Patna Museum Catalogue of Antiquities. Patna: Patna Museum. Härtel, Herbert. 1976. ‘Aspects of Early Nāga Cult in India.’ Journal of the Royal Society of Arts 124, 663–683. Härtel, Herbert. 1993. Excavations at Sonkh: 2500 Years of a Town in Mathura District. Berlin: Dietrich Reimer. Hazra, R.C. 1948. ‘Gaṇapati-Worship, and the Upapurāṇas Dealing with it.’ Journal of the Ganganatha Jha Research Institute 5/4, 263–276. Huyler, Stephen P. 1996. Gifts of Earth: Terracottas & Clay Sculptures of India. New Delhi: Indira Gandhi National Centre for the Arts.

A Nāga Altar in 5th century India


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 PhilologischHistorische 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.



Sen, Sukumar. ‘On Mūradeva, Mūladeva and Śiṣnādeva.’ In Mélanges d’Indianisme a la Mémoire de Louis Renou. Paris: Éditions E. de Boccard, 677–681. Shaw, Julia. 2004. ‘Nāga Sculptures in Sanchi’s Archaeological Landscape: Buddhism, Vaiṣṇavism, and Local Agricultural Cults in Central India, First Century BCE to Fifth Century CE.’ Artibus Asiae 64/1, 5–59. Shaw, Julia. 2013. ‘Archaeologies of Buddhist propagation in ancient India: ‘ritual’ and ‘practical’ models of religious change.’ World Archaeology 45/1, 83–108. Storch, Tanya. 2016. ‘Fei Changfang’s Records of the Three Treasures Throughout the Successive Dynasties (Lidai sanbao ji 歷代三寶紀) and its Role in the Formation of the Chinese Buddhist Canon.’ In Spreading Buddha’s Word in East Asia, eds Jiang Wu and Lucille Chia. New York: Columbia University Press, 109–142. Straube, Martin. 2006. Prinz Sudhana und die Kinnarī: Eine buddhistische Liebesgeschichte von Kṣemendra, Texte, Übersetzung, Studie. Marburg: Indica et Tibetica Verlag. Strauch, Ingo. 2014. ‘The Evolution of the Buddhist rakṣā genre in the light of new evidence from Gandhāra: The *Manasvi-nāgarāja-sūtra from the Bajaur Collection of Kharoṣṭhī Manuscripts.’ Bulletin of the School of Oriental and African Studies 77/1, 63–84. Tokuno, Kyoko. 1990. ‘The Evaluation of Indigenous Scriptures in Chinese Buddhist Bibliographical Catalogues.’ In Chinese Buddhist Apocrypha, ed. Robert Buswell. Honolulu: University of Hawai’i Press, 31–74. Varma, K.M. 1983. Stucco in India: From Pre-Mohenjodaro Times to the Beginning of the Christian Era. Santiniketan, West Bengal: Proddu. Varma, K.M. 1987. Technique of Gandhāran and Indo-Afghan Stucco Images. Santiniketan, West Bengal: Proddu. Vira, Raghu and Lokesh Chandra. 1995. Gilgit Buddhist Manuscripts: Revised and Enlarged Compact Facsimile Edition. 3 vols. Delhi: Sri Satguru Publications. Vogel, J. Ph. 1926. Indian Serpent-Lore, or the Nāgas in Hindu Legend and Art. London: Arthur Probsthain. Winternitz, Moriz. 1888. ‘Der Sarpabali, ein altindischer Schlangencult.’ Mitteilungen der Anthropologischen Gesellschaft in Wien 18, 25–52, 250–264. Winternitz, Moriz. 1998. ‘Gaṇeśa in the Mahābhārata.’ Journal of the Royal Asiatic Society 3, 380–84. Wright, Arthur F. 1948. ‘Fo-T’u-Têng 佛圖澄: A Biography.’ Harvard Journal of Asiatic Studies 11/3–4, 321–371. Wright, Arthur F. 1990. Studies in Chinese Buddhism. Edited by Robert M. Somers. New Haven, CT: Yale University Press. Yano, Michio. 2004. ‘Planet Worship in Ancient India.’ In Studies in the History of the Exact Sciences in Honour of David Pingree, eds David Edwin and Charles Burnett. Leiden: Brill, 331–348.


At the Crossroads of Art and Religion: Image Consecration in the Pāñcarātrika Sources Marzenna Czerniak-Drożdżowicz The cult of a concrete deity is one of the characteristic features of the Pāñcarātra, an early Vaiṣṇava Tantric tradition of India. According to it, this deity is perceived as the only object of worship, but present in his many forms, among them also material representations. Since images of the god are at the same time treated as the god himself, the procedures concerning the process of transformation of a material, art object into a powerful deity’s representation requires a particular practice and a particular theory behind it. This object of art and, at the same time, the object of devotion should also be beautiful and pleasant to be looked at (darśanīya) because, being like that, it stimulates devotion.1 In these installation and consecration procedures art and religion are combined. The ideas concerning the idols and the god’s presence in them can be found in the Pāñcarātrika literature as well as in the sources belonging to the Śrīvaiṣṇavas, a tradition closely related to the Pāñcarātra that elaborates ideas on the presence of god in his representations.2 One of the crucial skills of the Pāñcarātrikas, of course those entitled to it, was invoking, inviting (āvāhana) the god into his idols in the temple. The creation of these representations of the deity was subject to particular regulations since the idols had to be fashioned and created in a strictly regulated way and the rules for these procedures are explained in the Pāñcarātrika literature. Thus Pāñcarātra created an elaborate concept of different forms of god, among them so-called modes of his existence and then manifold forms accompanied with mythical stories explaining their appearance and role. * 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 (CzerniakDrożdżowicz 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_008



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 manifestation. 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 devotees 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 characteristics (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 qualities, 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


Emanations known as vyūhas, namely Vāsudeva, Saṅkarṣaṇa, Pradyumna and Aniruddha are each characterized by two particular guṇas out of the group of six. In the case of Vāsudeva all of these six appear in the highest degree, but in the other vyūha forms two qualities are prevailing. Accordingly, Saṅkarṣaṇa is characterized by jñāna and bala, Pradyumna by aiśvarya and vīrya, while Aniruddha by tejas and śakti.8 Each of them plays a particular role in the creation theory of the Pāñcarātra: Vāsudeva provides opportunities to acquire divine favour, while Saṅkarṣaṇa, governing souls (jīva), destroys cyclically the world and hands down the knowledge in the form of śāstra; Pradyumna creates, implements and strengthens dharma, and Aniruddha protects and hands down the knowledge. These four forms possess particular iconography, among many representations, the one which is called viśākhayūpa—a pillar out of which spread four branches—vyūhas. This concept refers to the unity of the vyūha forms and the highest Vāsudeva. Vibhavas are known also as avatāras, and the number of them, according to different sources, varies from 10 to even 36. These forms refer to the god’s manifestations in the moment of distress, when they play a specific role. Sākṣādavatāra or svarūpāvatāra are those in which god’s nature manifests directly, and the ones known as āveśāvatāra are those in which the deity reveals himself only temporarily. Less important are incarnations of some aspects, companions, or weapons of god known as ‘partial incarnations’—aṃśāvatāra. The fourth mode is known as antaryāmin—‘inner controller’ which is constantly present inside each living creature.9 It is a subtle form of god and such an idea appeared already in the Upaniṣadic texts. For example, the Subalopaniṣad, belonging to the White Yajurveda, in chapter 7 identifies the insider ātman, present in every creature, with Nārāyaṇa (eṣa sarvabhūtāntarātmā apahatapāpmā divyo deva eko nārāyaṇaḥ /—‘This inner soul of all creatures, the one who overpowered evil, he is divine, the only God Nārāyaṇa.’10 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.



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 representations 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 CzerniakDrożdżowicz 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


Kautukabera, sometimes synonymous with utsavabera, can be a movable idol (cala) or, if made of stone, permanently residing in the temple. It is used for daily offerings (nityārcana). The kautuka idol usually remains in its place when the temple can afford all five types of idols and in such a case the role of processional idol is taken up by utsavabera, which is a movable idol usually made of metal.15 Balibera is used for bali offerings (of food) while snapanabera16 is used for bathing the deity, especially in daily rites.17 Pāñcarātrika texts (saṃhitās), especially those created later, when the temple cult became prevalent, included a lot of information referring to the idol, namely to the description of different forms of god, of their construction, installation and consecration (sthāpana or pratiṣṭhā).18 The worship in the temple constitutes one of the four main topics of the saṃhitās,19 called kriyāpāda.20

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 rituals 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).



The descriptions of the pratiṣṭhā21 can be found in several texts, among them22 the Sātvatasaṃhitā, Jayākhyasaṃhitā, Pauṣkarasaṃhitā, Pādma­ saṃhitā, Paramasaṃhitā, Īśvarasaṃhitā, Pārameśvarasaṃhitā, Viṣṇusaṃhitā, Viṣvaksenasaṃhitā, as well as Devāmṛtapāñcarātra.23 Some of these texts, for example, the Paramasaṃhitā and the Viṣṇusaṃhitā very clearly state that one should worship god only in a form (ākāra) that is human or human-like.24 When the proper form has been chosen and created, installation and consecration can begin—the physical object is provided with 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 concerning installation’—pratiṣṭhāvidhi, Jayākhyasaṃhitā ch. XX, ‘rule concerning installation’—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 published, 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 appendix 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


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 previously 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 transitional 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 installation, 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).



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 invocation 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 substances needed for offerings as well as jewels, minerals and seeds; then come also musicians. Bad omens are pacified, which is done by offerings of clarified 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 pedestal 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: diamond (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


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 sacrificial 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.



An additional issue is the conservation and reconstruction of damaged idols and temple buildings. This is not a secular domain, because it is strictly regulated with the rules of religion and is described in religious texts, and not only in art manuals. The fact that the god is really present in the temple and in the idol means that one cannot manipulate it and operate in the temple as if it were a secular, profane domain. In the Pāñcarātrika texts, therefore, there are portions entitled jīrṇoddhāra—‘recreation of the destroyed/removal of the destroyed,’ or sometimes called navīkaraṇa—‘renovation.’ Usually the texts order to keep the idols and use the temples as long as they are not very much damaged.37 During renovation of the idols the potencies of god present in it have to be temporarily removed, which is done by using the kumbha vessel into which the god’s power is moved from the idol. One of the important elements of the ceremony is the pouring of water from this vessel over the head of the renovated idol or, in the case of the temple renovation, from the top of the temple by the priests who climb the superstructure. This way the potencies come back to the idol and to the temple. Descriptions of these ceremonies one can find, for example, in the Viṣvaksenasaṃhitā (ch. 36 jīrṇoddhāravidhi), Pādmasaṃhitā (ch. 17 jīrṇoddhāra) and in the Viṣṇusaṃhitā (ch. 24), which (in verse 24.3) compares the renovation and change of the place of the god’s residence with the abandonment of the old body by the soul.38 The old idol, which cannot be used any more, should be taken in the procession to a river and thrown into the water. The descriptions of getting rid of such idols are also detailed, since they are treated as a kind of nirmalya, the remnants from offerings and, due to the presence of god’s potencies, could be very dangerous if left unattended.39 The material presented above shows some mutual relations between religious ideas, verbalized as concrete prescriptions for the procedures, and the artistic and technical needs of the installation process. But this technical process has its value only as far as it concerns the real presence of the god. In the case of the body of literature we are concerned with, it is the idea of 37  Damages can be caused by different factors: atmospheric circumstances, but also by breaking 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


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 offerings 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 additional 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 40   Īśvarasaṃhitā 20.16cd–17 (see appendix 2.3). 41  See also Granoff 2006: 401–42. Also Viṣṇudharmottarapurāṇa 46.1–5 (see appendix 8) speaks about a necessary transformation of god’s body into one which is accesible to devotees.



satisfaction an invocation of Viṣṇu, the Supreme God, O Vajra is indicated/ ordained by the wise men.’ The position of the Jayākhyasaṃhitā seems to be similar, even if it reads that the consecration assures the permanent presence of god (yena sannihitaḥ sadā)42 and through it the image becomes the god himself (yena sarveśatā vipra bimbasyāsya prajāyate // 20.221). At the conclusion of the consecration, the sādhaka addresses the god in his image saying that now he will reside in it forever (vastavyam iha sarvadā).43 Nevertheless, in the JayS every daily rite begins with āvāhana and the reason for the fact could be that these daily rites refer to a personal worship of temporary objects of cult and not to the temple images, in which god permanently resides.44 The Viśvamitrasaṃhitā reads (10.108) that āvāhana is omitted only in the case of the idols consecrated by gods, siddhas and munis,45 so it is needed for those consecrated by men, thus for most of the temple images. Vaikhānasas for example direct āvāhana mostly to temporary images, invoking to them the god from the main idols.46 A kind of incompatibility of the texts and the actual temple ritual was also observed by Hélène Brunner47 who mentions a practice to interpret āvāhana not as an invocation of the god already present in his image, but as a means of attracting his attention during a particular act of worship (sammukhīkaraṇa), and a similar opinion is expressed by a Pāñcarātrika text, the Viśvamitrasaṃhitā—āvāhanam bhaved asya sammukhīkaraṇam dvija / 10.109ab. The Pādmasaṃhitā, while elaborating on the concept of rūpa and mūrti of god, understands mūrtis as different divine figures arising in the process of emanation and being concretizations of the Highest Vāsudeva, who in the

42   Jayākhyasaṃhitā 20.207cd-208 (see appendix 3.1). 43   Jayākhyasaṃhitā 20. 331 (see appendix 3.2). 44  For the supposition that these two sets of rites were meant for different groups—temple priests and Brahmins worshipping in private see Granoff 2006. 45   Viśvamitrasaṃhitā: suraiḥ siddhaiḥ sattvaniṣṭhair munibhiś caiva kārite / āvāhanaṃ na kurvīta sthāpite niścale harau // 10.108. 46  Colas 1986: 96, Marīcisaṃhitā 4.2.2.: ‘En dehors de ce (ritual d’installation) on ne fait ni invocation (āvāhana) ni congédiement (visarjana); on vénère journellement et correctememnt (le Dieu) selon la règle (prescrite) pour le culte quotidian (nityārcana).’— anyatrāvāhanavisarjanaṃ ca hitvā nityārcanavidhānena nityaṃ samyag arcayet //; see also Colas 1996: 297–298. He also mentions that Sanatkumārasaṃhitā of Pāñcarātra (BrR IX, 1–4) distinguishes the cult (arcanā) for the fixed image (sthāvara) and for the movable one (asthāvara) (Colas 1996: 297, n. 5). 47  Brunner 1993: 12.

Image Consecration In The Pāñcarātrika Sources


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 installation 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 highest 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 manifestation 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.



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 heavenly 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 transcendental 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 considers 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 hidden 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, irregular, 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


Śrīvaiṣṇavas, understanding god’s manifestations in the temples to be superior to all other forms, believe that Viṣṇu in his images is present in every temple and in every image, and that the procedure of āvāhana and visarjana, invoking and dismissal of the god, undertaken before and after the worship, are performed even in the case of the consecrated images. Their role is to draw attention and full involvement of the god present in countless representations in this particular ritual for this particular image. Different personalities of the local manifestations are described in the sthalapurāṇas, the stories praising particular places of worship. Narayanan (1985: 57) says: ‘In theological terms, the same Lord is said to reside in the many holy places with no difference in status among local manifestations, just as the manifestation of Viṣṇu as arcā, as present in the heart (hārda), and as incarnate in the world (avatāra) are ontologically equivalent.’ Since men cannot approach and understand god in his transcendent state, he, showing his favour, reveals himself in forms that are accessible to men, among them mūrtis. André Padoux (1990: 1), referring to the Sanskrit root murch meaning ‘to become solid/thicken’ and to its derivative mūrti meaning ‘a condensation’, wrote that it is this invisible absolute who evolves and condensates and by means of the fact provides the forms which are then worshipped by men. These forms have to be fashioned in a particular way and created according to the rules. To enable the real, full presence of the god in such an image, the installation and consecration process, which joins the elements of art/technique with ritualistic/religious ones, is necessary. Only images that follow the traditional canons prescribing these procedures could be effective, giving the real presence of god. 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).



Jayākhyasaṃhitā: Jayākhyasaṃhitā. 1931. Crit. edited with an Intro­duc­tion in Sanskrit, Indices etc. by Embar Krish­nam­­acha­rya. (Gaekwad’s Oriental Series 54). Baroda. Devāmṛtapāñcarātra. Unpublished critical edition by Diwakar Acharya (e-text courtesy to the author of the edition). Paramasaṃhitā [of the Pāñcharātra]. 1940. Edited and Translated into English with an Introduction by S. Kri­sh­na­swami Aiyangar. (Gaekwad’s Oriental Series 86). Baroda. Paramasaṃhitā. Ms 62.045 (10.G.27). Theo­­so­phical Socie­ty, Adyar. Pādmasaṃhitā (PādS). (Part I). 1974. Crit. Ed. by See­tha Padmanabhan and R.N. Sampath. (Part II). 1982. Crit. Ed. by Seetha Padmanabhan and V. Varadachari. Madras: Pancaratra Parisodhana Parisad. Pārameśvarasaṃhitā (PārS). 1953. Śrīpāñcarātrāntargatā Śrīpārameśvarasaṃhitā. Śrī Govindācāryaiḥ saṃskṛtā. Śrīraṅgam. Pauṣkarasaṃhitā (PauṣS): Pauṣkara-Saṃhitā. 1991. Vol. 1 Critically edited by Prabhakar Pandurang Apte. Tirupati: Rashtriya Sanskrit Vidyapeetha; 2006. Vol. 2 Critically ed. and translated into English by Prabhakar Pandurang Apte. Tirupati: Rashtriya Sanskrit Vidyapeetha. Mahābhārata (MhBh). 1933–59. Critical Edition, edited by Vishnu S. Sukthankar and S.K. Belvalkar. 19 vols. Poona: The Bhandarkar Oriental Research Institute. Viśvamitrasaṃhitā. 1970. Edited by Undemane Shankara Bhatta. Tirupati: Kendriya Sanskrit Vidyapitha. Viṣṇudharmottarapurāṇa (VDhP). 1958. Edited by Priyabala Shah. Third Khaṇḍa, vol. 1, Gaekwad’s Oriental Series No. 130. Baroda: Oriental Institute. Viṣṇusaṃhitā. (ViṣṇuS). 1991. Edited by M.M. Gaṇapati Sāstrī with an Elabo­rate Introduction by N.P. Unni. (Tri­van­drum Sanskrit Series 85). Revised & Enlarged Edition. Delhi: Nag Publishers. Śrīpraśna Samhitā. 1969. Edited by Seetha Padma­­nabhan with the Foreword of V. Raghavan. Tirupati: Rashtriya Sanskrit Vidyapeetha. Sātvatasaṃhitā (SS). 1982. With Com­ment[a]ry by Ala­śiṅga Bhaṭṭa. Fore­word by Gaurinath Sastri, edited by Vra­ja Vallabha Dwivedi. (Library Rare Texts Publi­ca­tion Series 6). Varanasi.

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


Brunner-Lachaux, Hélène. 1963, 1968, 1977, 1998. Somaśam­bhu­pa­d­dha­ti. Texte, Traduction et Notes. 4 vols. Pondicherry: Publications de l’Institut Francais d’Indologie. Brunner, Hélène. 1993. ‘Jñāna and Kriyā : Relation between theory and practice in the Śaivāgamas.’ In Ritual and Speculation in Early Tantrism: Studies in Honor of André Padoux, ed. Teun Goudriaan. Delhi: Sri Satguru Publications. Colas, Gerard. 1986. Le Temple selon Marīci. Pondicherry: Publications de l’Institut Francais d’Indologie. Colas, Gerard. 1996. Viṣṇu, ses images et ses feux. Les métamorphoses du dieu chez les vaikhānasa. Paris: Presses de l’Ecole francaise d’Extreme-Orient. Czerniak-Drożdżowicz, Marzenna. 1998. Sṛṣṭikrama—order of creation in the Paramasaṃhitā. In Studies in Hinduism II. Miscellanea to the Phenomenon of Tantras, ed. Gerhard Oberhammer. Wien: Österreichische Akademie der Wissenschaften, 43–54. Czerniak-Drożdżowicz, Marzenna. 2003. Pāñcarātra Scripture in the Process of Change. A Study of the Paramasaṃhitā. Vienna: Publications of The De Nobili Research Library. Czerniak-Drożdżowicz, Marzenna. 2008. Studia nad pańćaratrą. Tradycja i współczesność. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego. Czerniak-Drożdżowicz, Marzenna. 2011. Studia nad pańćaratrą. Część II. W poszukiwaniu tożsamości. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego. 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, H. Richard. 1999. Lives of Indian Images. Princeton, NJ: Princeton University Press 1999. Eck, Diana. 1998. Darśan. Seeing the Divine Image in India (3rd ed.). New York: Columbia University Press. Einoo, Shingo and Takashima, Jun (eds). 2005. From Material to Deity. Indian Rituals of Consecration. New Delhi: Manohar. Goodall, Dominic. 2004. The Parākhyatantra. A Scripture of the Śaiva Siddhānta. A Critical edition and annotated translation. Pondicherry: Institut francais de Pondichéry. Goudriaan, Teun. 1965. Kāśyapa’s book of wisdom (Kāśyapa-jñānakāṇḍaḥ): A ritual handbook of the Vaikhānasas. The Hague: Mouton and Company. Granoff, Phyllis. 2006. ‘Reading between the lines. Colliding attitudes towards image worship in Indian religious texts.’ In Rites Hindous, Transfers et Transformations, eds Gerard Colas and Gilles Tarabout. Paris: Collection Puruṣārtha, Editions de l’Ecole des Hautes Etudes en Sciences Sociales, 389–421.



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 ĀgamaKosha). 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


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.



Appendix: Quoted Verses from Selected Texts


Devāmṛtapāñcarātra 1.7–17

pratiṣṭhā tasya devasya śrotum icchāmi tattvataḥ | kasmin dravye tu kartavyā tasya vai pratimā śubhā || 7 kiṃ pramāṇaṃ tu vijñeyam aṅgapratyaṅgayos tathā | śarīraṃ kiṃpramāṇaṃ tu kartavyaṃ lakṣaṇānvitam || 8 makuṭasya pramāṇaṃ tu śirasaś ca kathaṃ bhavet | lalāṭaṃ tu punas tasya kartavyaṃ kiṃpramāṇataḥ | 9 nāsikā tu kathaṃ kāryā netre caiva pitāmaha | gaṇḍayoḥ kiṃ pramāṇaṃ tu karṇayoś ca kathaṃ bhavet || 10 oṣṭhau kiṃ tasya kartavyau cibuke kiṃ nu lakṣaṇam | bāhoś caiva tadaṅgulyā grīvāyāḥ kiṃpramāṇataḥ || 11 vakṣaś caiva kathaṃ kāryaṃ stanake nābhimaṇḍalam | jaṭharasya pramāṇaṃ tu kaṭyā vai liṅgam eva ca || 12 ārukau jānunī caiva nalakau gulphapādayoḥ | aṅgulyā nakhaṃ aṅguṣthau sarvalakṣaṇam eva ca || 13 vāhanaṃ nu punas tasya kartavyaṃ tu janārdane | āyudhāni ca devasya kīdṛśāni tu kārayet || 14 sarvalakṣaṇasaṃpannaṃ kathaṃ viṣṇuṃ vinirdiśet | prāsādaṃ tu kathaṃ kāryaṃ śobhanaṃ yāgamaṇḍapam || 15 piṇḍikā tu kathaṃ kāryā garbhasūtraṃ kathaṃ bhavet | grahanakṣatralagnaṃ ca tithivāraṃ tathaiva ca || 16 maṇḍapasya vibhāgaṃ tu vedyāś caiva tu lakṣaṇam | sthāpakasya tu cihnāni ye ca mūrtidharāḥ smṛtāḥ | etat sarvaṃ samāsena bhagavan vaktum arhasi || 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


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 |





svayaṃvyakte tathā divye saiddhe cārṣādyamānuṣe | jīrṇoddhāravidhiṃ vakṣye śṛṇudhvaṃ munisattamāḥ ||

kṣetreṣu puṣkarādyeṣu svayam eva jagatpatiḥ || 20.16 arcārūpeṇāvatīrya vyaktaiś cakrādilāṃchanaiḥ | samanvitaḥ sannidhatte 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 (. . .).’



2.4 20.263cd–264

namas te pararūpāya namas te vyūharūpiṇe || namo vibhavarūpāya namas te tv antarātmane | namas te ‘rcāvatārāya nānākārāya te namaḥ || ‘Salutation to [your] highest form, salutation to [you] having a form of emanation (vyūha), salutation to [your] form as manifestation (vibhava), [and] to [your form as] inside ruler/inner controller (antarātma = antaryāmin), salutation to [you] as descending into the idol (ārcāvatāra), salutation to you in many forms.’

3 Jayākhyasaṃhitā 3.1 20.207cd–208

athādivāsanaṃ kuryād vidhidṛṣṭena karmaṇā || dhyānākhyaṃ niṣkalaṃ śuddhaṃ yena sannihitaḥ 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


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 | ‘sthāpanā, asthāpanā, as well as saṃsthāpanā, prasthāpanā and pratiṣṭhā—these are the 5 kinds of installation.

Image Consecration In The Pāñcarātrika Sources


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



tejonidānamadhyeyaṃ nārāyaṇam ajaṃ vibhum | nivṛttimantram uccārya pratimāyāṃ niveśayet || 28.60 brahmarandhreṇa mārgeṇa praviṣṭaṃ pratimākṛtau | vyāpya sthitaṃ smaret sarvaṃ mantreṇa kamalāsana || 28.61. Translation by Silvia Schwarz: ‘In front of the chief icon, remaining in the bound lotus posture, having held the breath, the pure-minded one should meditate on the eternal brahman [as]: [52] Vāsudeva, unborn, tranquil, luminous, eternally manifest, without beginning, middle and end, one, immovable and permanent, [53] nothing but consciousness, supreme bliss, beyond the darkness, imperishable, fully endowed with knowledge (jnāna), power (śakti), strength (bala), sovereignty (aiśvarya), valour (vīrya), and splendor (tejas); [54] not tangible as far as feet and hands [are concerned and] as far as the faculties of seeing, hearing and so on [are concerned, yet] having hands, voice and feet everywhere, having eyes, heads and mouths everywhere; [55] free from growth and decline, blazing like millions of suns, consciousness, all-pervading, eternal, the one whose wonderful world beyond heaven [is] whole consciousness. [56–57a] After having fixed the thought on Viṣṇu in such a form, as the Supreme Viṣṇu, consisting of the mūlamantra, [who is] eternal joy, consisting of the prāṇava, the Great Viṣṇu, he should make Viṣṇu enter into the lotus heart [and], having brought [Him] near, he should worship him according to the regular order. [57b–58] Afterwards, having brought [Him] together with Śrī to the throne within the añjali by means of the puruṣātmamantra, then the guru should meditate with [the help of] the viśvātmamantra; [59] after having uttered the nivṛttimantra he should make Nārāyaṇa enter into the cult-image, unborn, allpervading, to be understood as essentially splendour; [60] he should think of [Him] entered into the cult-image through the way of the brahmarandhra by means of the mantra, abiding there by pervading it [i.e. the image] completely, o lotus born. [61]’ I am grateful to my colleague Dr. Silvia Schwarz for allowing me to quote her translation from her Ph.D. dissertation before it was published.

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


ākāre tu kṛtā pūjā stutir vā dhyānam eva vā | vidhinā śāstradṛṣṭena deva eva kṛtā bhavet || 3.8 bhaktyā paramayā ‘kṛṣṭo devadevas sa yogibhiḥ | teṣām anugrahārthāya rūpaṃ bheje caturbhujam || 3.9 tasmāt tenaiva rūpeṇa devadevaṃ samarcayet | phalabhedāt dvidhā tasya pūjanaṃ śāstracoditam || 3.10 phalam abhyudayaḥ pūrvaṃ nirvāṇaṃ tu paraṃ phalam. ‘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.’


Pārameśvarasaṃhitā 11.236–237

vāsudevābhidhānas tu devaḥ ṣāḍguṇyavigrahaḥ | karmiṇām upakārārthaṃ prāsādaṃ sthūlavigraham || 236 sarvajñānakriyāḍhyaṃ ca sarvatatvasamāśrayam | samāsādyānugṛhṇāti sadārcāntargataḥ prabhuḥ ||.237

196 7


Śrīpraśnasaṃhitā 2.54cd-55ab

manmūrtayaḥ pañcadhā vadanty upaniṣatsu ca|| paravyūho hārda iti vibhavo ‘rceti bhedataḥ | ‘They say in Upaniṣads about my forms as fivefold: para, vyūho, hārda, vibhavo and ārca separately’


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 ‘Without a form, smell and taste, without speech and touch is Puruṣa described by you, therefore how He could have this form? Mārkaṇḍeya said Prakṛti ([inborn] nature/natural/fundamental) and vikṛti (change/changeable/modification) are the two forms of the Highest Soul. His form without characteristics is known/proclaimed as prakṛti. The one with a shape known as vikṛti is declared the whole world. It is possible to perform worship and meditation for the one having a form. Thus due to [his] nature God with a form should be worshipped according to the rules. [If] a non-manifested way [/form is chosen], embodied ones obtain difficulty/ unhappiness. Therefore which [form] is shown by the God according to [His] wish, this form of deity, in [His] manifestations, they worship.’

Image Consecration In The Pāñcarātrika Sources


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 ‘No one is able to meditate upon God without a form. If one’s mind is averted from all forms, where does it fix? If the mind is averted, it wanders or it is overtaken by sleep. Therefore the wise one should worship with his mind [only] the one having the form. But he should remember that it is not really Him.’ (Eck 1998 p. 45: ‘Without a form, how can God be meditated upon? If (He is) without any form, where will the mind fix itself? When there is nothing for the mind to attach itself to, it will slip away from meditation or will glide into a state of slumber. Therefore the wise will meditate on some form, remembering, however, that the form is a superimposition and not a reality’.)


Choosing an ācārya for Temple Construction and Image Installation Elisabeth Raddock 1 Introduction How do those desiring liberation make a temple for you, O God? 25 And what is the rule for the protectors of the mūrti of the ācārya? And [what is] the regulation for a sacrifice to the vāstu? And [what is] the rule for the giving of arghya? 26 What is the regulation for the placing of the stones? As well as [what is] the [regulation for] the preparation of the sacred ground etc.? And [what is] the rule for the temple? And [what is] the rule for the image? 27 [What is the rule with regards to] the entire five-fold temple? Thus also, what is the method of erecting the flag-staff? And whatever else that would be additional to the temples that [too I] asked about, O Sureśvara. 28 (Hayaśīrṣa Pañcarātra, 1.25cd–28) 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 characteristics 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_009

Choosing an ācārya for temple construction



The Text

The Hayaśīrṣa Pañcarātra (c. 8th–9th century CE2) is a Vaiṣṇava Pāñcarātra text that focuses on both the construction of temples and sculpting of images, which are to be placed inside. Pāñcarātra literature discusses four steps or phases in the construction of the temple: the planning and construction of the prāsāda (temple), the design and sculpting of the pratimā (image), pratiṣṭhā (installation) and pūjā (institution of worship performed daily and occasionally within the temple after its completion). The first three steps are discussed in the Hayaśīrṣa Pañcarātra. 3

Historical Context

The Hayaśīrṣa Pañcarātra was composed in the early stages of what scholars, 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.



with distinct philosophical systems, such as Pāñcarātra, to which the Hayaśīrṣa Pañcarātra belongs. These schools shared both ideas, articulating how the cosmos and the world are organized, as well as practices/rituals that enabled followers to act most efficaciously in the cosmos. Though these practices had as their primary, explicit goal the personal relationship between God and devotee, and the transforming action of his or her condition, these practices also collectively brought about social relations among people, structuring community, authority and hierarchy within the society. During this period (ca. 700–1200) ‘religious communities devoted to the gods Viṣṇu and Śiva had largely supplanted those loyal to older religious formations, such as that of the Vedas and the heterodox Buddhists and Jains, in elite support and institutional resources’ (Davis 1997: 26). According to Romila Thapar (2002: 387), ‘[t]he temple was closely associated with the belief and practice of Purāṇic religions.’ Purāṇic religions promulgate bhakti and the worship of Viṣṇu, Śiva or the goddess. In the Purāṇic religion, the temple emerges as the center of public worship (and, as Thapar and others have noticed, a center of economic and political power). Earlier, relatively simple codes for temple construction, still in the process of developing, were elaborated at the time of Temple Hinduism into more complex regulations. These regulations are laid out, in similar fashion, in multiple texts from various places across the subcontinent. There are differences of course, determined by the particular religious affiliation of the texts’ authors, regarding location as well as purpose (whether ritual or architectural). Even so, the similarities are striking, particularly in terms of structure, stress on the rituals concerning the vāstupuruṣamaṇḍala (with exceptions such as the Pādma Saṃhitā) and the significance of proportion and cardinal directions. During the early medieval period, the temple became the dominant religious institution in South Asia and Hindu-influenced Southeast Asia. Myriad temples of every size were constructed and dedicated, not only to Viṣṇu and Śiva but also to the Goddess, as well as other divinities regarded as lesser according to the dominant Vaiṣṇava and Śaiva points of view. As the permanent residence of a god in the community, the temple became much more than the location for pūjā offerings. The temple was, and is, the locus of a wide range of personal and social activities, both ritual and other, providing, for example, a stage for performing and literary arts.4 Many temples were major economic institutions acting as feudal lords, employers, landholders and moneylenders. The temple was the social center of the community and was an 4  Arts may have ritual elements incorporated as well (see Thielemann 2002: 9).

Choosing an ācārya for temple construction


arena in which relationships of authority and rank were constituted, contested and displayed.5 Since a temple is a place for contact with a god, as well as a social, economic and political institution, construction of large structures becomes a major event. A king, a chieftain, or an assembly of important and wealthy locals, having decided that they would like to have a temple constructed, search out the people to do the job for them, and then step back, taking on the role of yajamāna (patron). Sanskrit texts lay out the steps for constructing a temple. The steps are described in similar ways by the different schools and in the different parts of the subcontinent. In spite of differences, there is a general pattern that each text follows. First, one needs to locate an appropriate plot of land, which has the right type of soil, as well as surroundings. Next, the area is examined to make sure that there are no malignant spirits or objects hiding there. If there are, certain rites must be performed to remove them. Then, the space is ritually and spatially marked, and the appropriate deities invited. Only after these place-taking ceremonies are carried out can construction start. While complicated rituals, calculations and rules may not have been followed in the case of small roadside shrines, there are many indications that they were important for the construction of large, permanent temples, built as homes for the gods.6 As the construction of a temple is an important event, it is not surprising that the individuals who carry out the work are carefully selected and/ or characterized by the text. One might read particular attention to the individual workers qualifications as a synecdoche: the employee, particularly the ācārya, represents the whole temple construction, or a metonymy: the holiness of the temple (or the principle of selectivity and exclusivity) spreads from the temple to encompass its workers. In both cases the importance of the employees’ character is great. If one reads it as a synecdoche the employees’ character colors the temple, while if one reads it as a metonymy the holiness of the temple spreads to the workers (which would be disastrous if they did not behave properly).

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 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 architectural form.’ At first glance, this formulation sounds similar to Coomaraswamy 1934/94, who attempts to articulate the spiritual truths symbolized by architectural forms. Michell’s formulation, however, isolates three moments: theological 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 intermediary of ritual and rule, which is central to the genre of śilpa śāstra and better 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 representation.’ 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 written 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


part of a received body of sacred knowledge as give guidance to masons or sculptors.’ Michell’s thesis that rules are an intermediate step facilitating the translation of theology into architectural form is rather different than Meister’s statement that śāstra justifies temple construction. These two narratives articulating the role of śilpa śāstra, though contradictory, need not be mutually exclusive. Jones has noticed that there are three sorts of claims to the authority of standardized architectural stipulations: 1. ‘[T]here are certain universally applicable rhythms and proportions,’ seen in nature and mathematics and replicated in architecture; 2. God decreed certain ritual-architectural prescriptions, which should be followed in architectural design; 3. Prestigious ancestors have established patterns that ought to be replicated in architecture. As Jones (2000: 35, 48) notes, these categories are not mutually exclusive. It is clear that they all apply to the construction of the Hindu temple. Proportions, rhythm and alignment with the directions8 all fall within the first stipulation, while many śilpa śāstras fall within the second, with their strong soteriological focus on a great god. Again other śāstras emphasize the third type depending on who supposedly composed them. While the Hayaśīrṣa Pañcarātra narrator is a god, Viṣṇu in his form as Hayaśīrṣa, other texts, such as the Samaraṅgana Sūtradhāra attributed to Bhubhojan Deva, have a human as the composer. It may well be the case that texts such as the Hayaśīrṣa Pañcarātra and the Agni Purāṇa mainly focus on the ritual confirmation of the structure rather than specific architectural details beyond proportioning because they are intended more as ritual manuals than architectural śīlpa śāstras. The texts deal with actual architectural details in a vague fashion. Their focus is on ritually significant moments during construction. In the Hayaśīrṣa Pañcarātra and the Agni Purāṇa more text is actually devoted to consecration and other rituals than actual descriptions of construction. The quantity of rituals described gives the Hayaśīrṣa Pañcarātra and the Agni Purāṇa a very different appearance than śilpa śāstra texts, such as the Mānasāra and the Mayamatam, which contain very detailed measurements, proportions, etc. but are virtually silent about rituals. The one ritual śilpa śāstras do commonly discuss in any length is the vāstupuruṣamaṇḍala. This fact has caused scholars to propose that the maṇḍala has a vernacular origin, i.e. that it is part of the architectural tradition rather than the brāhminical.9 There are thus two categories of texts serving different purposes: śilpa śāstra and ritual manual.

8  The directions, which in the Hindu tradition are divine (see Wessels-Mevissen 2001). 9  See, for example, Meister 2007.



Alternatively it is possible to argue that the Hayaśīrṣa Pañcarātra falls into the ‘makara’ category to which Anna Ślączka (2007: 11) assigns the Kaśyapa Śilpa.10 Ślączka argues that the Kaśyapa Śilpa meets the criteria of both ritual manual and śilpa śāstra. While the Hayaśīrṣa Pañcarātra, on the one hand, provides information mainly on ritual, it also provides details with regards to the layout of a temple, thickness of walls, height, number of pillars in maṇḍapas and their relative location to the garbhagṛha. On the other hand, it provides little detail with regards to embellishments besides the doorframe. While the text does discuss the features of gods and goddesses to be carved (or painted) it generally says little about their placement, and occasionally the text is satisfied by merely mentioning their names, such as in the chapter on yoginīs.11 The Hayaśīrṣa Pañcarātra, is, therefore, a ritual manual with some characteristics of śilpa śāstra. Texts such as the Hayaśīrṣa Pañcarātra and the Agni Purāṇa were composed to emphasize the importance of the temple consecration rituals. These texts were not written for someone whose main interest is the actual making of temples, sculptures and the like. It was composed to make sure people know basic theological-ritual facts, including the fruits of actions like making temples. For example, the author/ compiler of the Agni Purāṇa says (in chapter 38) that just thinking of building a temple frees one from the sins of 100 births. Perhaps, rather than being written by/for śilpins, it is written for priests, who need to know about rituals; priests who need the basic knowledge of temples, sculptures, and their placement. This, I would argue, is the point of the rest of the Agni Purāṇa, in which, ritually important points of all possible aspects of life are included such as grammar, meter, bathing rituals, rituals for kings, etc. It is also possible that the Hayaśīrṣa Pañcarātra is written for the yajamāna since several chapters in the beginning discuss the qualifications of the ācārya; providing a checklist for someone who wants to hire personnel to build a temple. Many portions of the text resemble a checklist rather than actual descriptions. For example, the descriptions of deities often are mere enumerations of what they should hold in their hands. Lists are an interesting generic feature of many varieties of Sanskrit literature. 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 (concentration), 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 consecration (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


Within the complex system of proportioning based on the vāstu­ puruṣamaṇḍala there is still a great deal of license in which way the artist/ architect may use his creativity in order to plan the temple. Probably, only a small elite set of specialists had even a rudimentary command of these operative, abstract principles. Meister particularly emphasizes the secrecy, which he said increased over time, associated with the knowledge of ritual grids. Meister (1985) documents the extreme care with which a Hindu architect would select only a few disciples who would be taught the principles behind the architecture they had created. Meister relates this phenomenon to contemporary times. Indeed, ‘few people are cognizant of the logic behind the architectural forms in which they live their daily lives; fewer still explicitly understand the subtleties of the structures in which they worship’ (Jones 2000: 59). Meister’s point is, no doubt, correct. On the one hand, before the advent of printed literature and general education, pre-modern systems of education, including ancient Hindu ones, were based on family, caste and guild-like groups. One learned by apprenticeship. On the other hand, and not unrelated to the phenomenon of guild secrecy, the ideology of mokṣa-oriented jñāna, from Upaniṣadic times to tantric sects, valorize and reward esoteric, specialist knowledge. By way of analogy, one may note that modern professional societies share some characteristics, of specialist knowledge and charisma, with pre-modern guilds and castes. If these standards of ‘correct’ proportioning and orientation are, generally, secret (or at least a highly privileged) knowledge, how do the majority of people, uneducated, and not participating in the rituals before and during construction, experience these architectural conventions? Is the adherence to the tradition of importance to them? And how do, if they do, these conventions affect their religious experience? Historically verifiable answers to these questions are, obviously, impossible. However, seeking to generalize, Alberti has noted that perfectly proportioned architecture (or art) engenders a nearly magical, certainly transrational sense of realization and fulfillment. No preparatory training, cultural cultivation or taste is required. According to Alberti, to stand in front of rightly proportioned (beautiful) works of art effects a profound transformation in that spectator and ‘contributes to the honest pleasure of the mind.’12 This is a statement that seems to resonate with traditional Hindu artistic ideology. Right proportions are of central importance and are emphasized in all manuals of art. Indeed, creating the right proportions may be the main purpose of the vāstupuruṣamaṇḍala. This is a critical aspect setting śilpa śāstra apart from other brāhminical texts. The spatial facts of architecture—particularly the phenomenon of proportion—interact with 12  Alberti cited in Freedberg 1989: 44–47. Compare Eco 1988: 85.



mythological, theological, sociological, pictoral and narrative motives familiar from other genres of Sanskrit literature. The śilpa texts constantly remind the reader (hearer) that adherence to rules are of critical importance to the welfare of the community. There is the inducement (phala śruti) that the person who builds a temple will gain heaven, commonly not only for himself but for all his family in many generations. The benefit may also, more generally, encompass the entire cosmos. In addition the poor person who builds a small shrine will gain as much as a rich person who constructs a grand temple (Agni Purāṇa chapter 38.1–2113). The gods will also be happy by this perfectly proportioned temple. Only through construction of this perfectly proportioned temple will the gods be enticed to take up residence and, thus, make their presence manifest—thus affording the devotees direct access to the divine powers. The usage of grids is also supposed to help guard the temple (Meister 1985: 249, 256). Thus, texts such as the Hayaśīrṣa Pañcarātra simultaneously provide ritual validation for the temple construction and preserve a sacred and (more or less) secret knowledge regarding proportion and alignment, which can be appreciated after the fact by devotees lacking esoteric knowledge. 5

Employees of the Temple Project

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


The text states that the one who desires to build a temple should search for an ācārya: One who desires to gain merit by means of [constructing] a temple for my image ought to search for an ācārya who is endowed with [the following] characteristics (Hayaśīrṣa Pañcarātra, 2.11).15 This verse is important because it indicates that the text, or at least this portion 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 chapters 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 hiring, 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.



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 selection 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. He who is a pure descendent of Brahmā and he who knows the destruction, origin and maintenance of the universe, who is wellborn and wise, he is worthy of the state of an ācārya. 2.1220 Free from consumption and loss of memory, without leprosy, whose limbs are not too many or few, young and marked by the auspicious signs. 2.1421 He should take as a guru one who is a gṛhastha and a Brahmācarya22 and who has avoided the eight [things] that begin with ‘k’23 and who is intent on a vow of fasting, always. 2.1724 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


He should have an honest nature, he should be a sweet and gentle speaker. When such a guru, the best of men, is established then the ruler of the country will sacrifice25 and the artists, beginning with the sthāpakas, are honored for a long time. Here there is no doubt. 2.18–1926 It is important that the ācārya be of an appropriate caste: 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 Some of the qualities the ācārya should not have include:

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.



He should not be red-eyed, have honey-colored eyes or cats’ eyes. He should not be greedy, have inflammation of the neck glands,31 or be inclined to hypocritical behavior. 3.12 He should not be one who is lacking means or location, nor harsh or focused on meanness, nor pitiless or wanting in power, nor be one who is completely lacking skills. 3.13 Even if he knows the methods in the tantras for [constructing] temples he should not be one who has not reached the far shore of the Veda and aṅga texts.32 He should avoid with great care the non-believer,33 even if omniscient34 as well. 3.14 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 logical 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. A god enshrined by any of these [persons named above], is in no manner a giver of fruit. 3.16

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 paśusāstraparāṇmukhaḥ (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


If a building for Viṣṇu is made anywhere by these excluded [types] then [that temple] will not give rise to enjoyment and liberation and will yield no reward, of this there is no doubt. 3.1738 However the next chapter (chapter 4) discusses the fact that a twice-born liberated 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 characteristics 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

38  HBV 19.118. 39  HBV 19.100. 40  Fourfold saṃhitā. The commentary of the 1952 edition tells us that the Hayaśīrṣa Pañcarātra, which is fourfold, belongs to Ādi Saṃkarṣāna and teaches the pratiṣṭhā (base), taḍāga (pool or tank) (commentary reads tadaṅga), prāsāda (temple) and pratimā (image). That is the four parts of śilpa śāstra one needs to complete a temple building project. 41   H BV 19.102. This verse makes it clear that the text puts the knowledge of the Pāñcarātra tradition foremost. 42  The term iṣyate (root √iṣ) has several meanings, such as the passive meaning acknowledge or accept used here, it can also mean laid down or prescribed which is the meaning used in verse 5.9 below. 43  HBV 19.103. 44  HBV 19.119.



The achievement and success of each step of the construction and consecration of the temple depend on the ācārya.45 Perhaps the Hayaśīrṣa Pañcarātra and other similar Pāñcarātra texts, such as the Pādma Saṃhitā, were composed 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. 45   Pādma tantra, kriyā. 24.2b. (See Smith 1960). 46  See Smith 1960. 47  Kane tells us that an anulomaka is a śūdra who is born from a high-caste father and a lower-caste mother (Kane vol. 2, 1968: 449).

Choosing an ācārya for temple construction


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 building 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 principal 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).



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: He [with whom one constructs a temple] should not be a Śaiva, or a Saura, nor a Naiṣṭhika,53 nor a naked one, nor born of mixed marriage, nor unclean, old, or one who is of a despicable form or marked by great sin.54 3.255 Nor should he have leprosy,56 deformed nails, white leprosy,57 brown teeth, be a consumptive,58 one born in Kacchadeśa,59 or from Kāverī or Koṅkana.60 3.361 Nor originating in Kāmarūpa62 or Kaliṅga, or Kāñcī, Kāśmīra or Kośala, nor one having bad behavior, bad company or come from Mahārāṣṭra. 3.463 He should not be stupid, have a fat lip, be one who spits, or have an indistinct voice, nor have a tumor, nor be a charmer64 nor be deformed, proud or deaf. 3.565

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 ( Alphito). 58  Consumption—kṣayī. On consumption see Zimmermann (1987) where he mentions consumption 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


He should not be a punarbhū,66 a svayambhū,67 a widow’s bastard, or a non-believer,68 nor irrational, pale, bald or crippled69 or fat. 3.7 He should not have forsaken his vows or fasting nor be the husband of a śūdra, nor living on trade or theater. He should not be an adulterer with a bought woman. 3.9 He should not hate the ācārya, putraka and others, be a servant of others, a glutton, attendant, prone to disasters,70 wicked or afflicted with disease. 3.10 He should not be very sickly, a physician,71 violent, neglecting the right time, or one who abuses the twice-born, respectable elder or the god. A sthāpaka72 [who has these qualities] should be avoided. 3.1173 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 brahminical 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 mentioned 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 development 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 disqualifications for both groups of people—who would want a hypocritical employee or someone who is completely lacking skills?



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 mountains (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 aforementioned mountain ranges as the ‘Land of the Āryans’ (āryavārta). This land, the natural range of the black antelope, whose skin is used in various 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 ‘twiceborn’ 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 approximately 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) prescribes 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 preparatory 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.

74   Manusmṛti 2.22–4 (transl. by Goldman 2007: 2). 75  See Kane 1930/1962, vol. 2: 11ff, especially 18–19.

Choosing an ācārya for temple construction


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 auspicious 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, perhaps 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.



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 buildings. 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 construction 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 category 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 necessarily 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 acknowledge 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 discussion on the various titles, their interdependence, organization, etc., as described both in inscriptions and texts.

Choosing an ācārya for temple construction


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

Primary Sources (including translations of Sanskrit Texts) Agni Purāṇa

Dutt, K.N. (transl.), K.L. Joshi Shastri (ed.) 2005. Agnimahāpurāṇam, Sanskrit Text and English Translation. Delhi: Parimal Publications. Mitra, Rājendrālala (ed.) 1873–1879. Agni Purāṇa: a collection of Hindu mythology and traditions. Calcutta: Asiatic Society of Bengal.

Hari Bhakti Vilāsa

Hayaśirṣa Pañcarātra



Pādma Saṃhitā

Gosvāmī Sanātana (com.), Bhumipati Dāsa (transl.), Purṇaprajña Dāsa (ed.), 2005, Hari Bhakti Vilāsa, Rasbiharilal & Sons, Vrindavan.

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: Critical Edition, 24 vols. 1933–1970. With Harivamsa. Critically edited by V.S. Sukthankar et al. Poona: Bhandarkar Oriental Research Institute.

Dagens, Bruno. 1994. Mayamatam. New Delhi: Indira Gandhi National Center for the Arts.

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.




Viṣṇudharmottara Purāṇa

1953. Śrīpāñcarātrāntargatā Śrīpārameśvarasaṃhitā. Śrīraṅgam: Śrī Vilasam Press.

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


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.



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.


‘Re-Installation’ of Idols Replacing Damaged Ones, with Special Reference to the Ritual Literature of Kerala S.A.S. Sarma 1

Re-Installation and Early Ritual Manuals

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 rituals 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 performed 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 12thcentury Śaiva manual on expiations says: If one’s own liṅga is dropped, destroyed, burnt, stolen or taken away by a rat, kite (ātāyin), crow, dog or monkey, one will be purified after reciting one lakh of aghora, [and] after installing, according to the rules, another liṅga. And the same [rule] applies for the piṇḍikā. But if the liṅga falls from one’s hand into flowing or stagnant (sthite) water, in that case too one should recite [aghora] one lakh times and [the liṅga] requires re-inauguration (punaḥ saṃskāram arhati). If the liṅga is licked by fire and its beauty is not diminished, [expiation is achieved] by [reciting aghora] thirty thousand times. If [the liṅga] is mutilated, burnt, loses its lustre, is used by out-castes, falls in a terrible hole [or] is stolen by kings or others, then one should abandon the liṅga, recite one lakh repetitions of aghora, install another liṅga and [thereby] one attains purification.1 1  Satyanarayanan 2015: 237–8:  sveṣṭaliṅge paribhraṣṭe naṣṭe dagdhe hṛte ‘pi vā // 119cd //  nīte vā mūṣikātāyikākakauleyavānaraiḥ /  japtvā lakṣam aghorasya vidhinā liṅgakāntaram // 120 //

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_010



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 deinstallation 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 reinstalling 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 chapter also indicates the difficulties one may face in worshipping a damaged liṅga and proposes the installation of a new one in its place: duṣṭaliṅge sthite rāṣṭre prajārājñor na śobhanam / tasmād uddharaṇaṃ kāryaṃ vidhinā mantrapūrvakam // Somaśambhupaddhati in its chapter on jīrṇoddhāra (Brunner 1998: 357–377) discusses in detail the procedure for replacing liṅga that are old, weakened, burned, partly damaged or ones made with wrong measurements. According to this, when the liṅga is moved, its accouterment such as its base (piṇḍī) and the bull, all that need to be removed and thrown in water with the  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


necessary rituals. Jñānaratnāvalī proposes re-installing a liṅga that is installed or worshipped by non-Śaivites such as the Pāśupatas, Somasiddhāntins, Kālamukhas etc. as well as damaged ones. It gives a detailed description of the rituals that are to be followed for the replacement of the damaged ones with new ones (Madras R. 14898, pp. 564–571). 2

Jīrṇoddhāra in Kerala Ritual Texts

Most of the ritual texts of Kerala, such as the unpublished Śaivāgamanibandhana, the Prayogamañjarī, the Īśānagurudevapaddhati, the Viṣṇusaṃhitā and the Tantrasamuccaya, concern temples. Very often they also contain a chapter, usually called Jīrṇoddhāra, ‘Renovation,’ devoted to the renovation of temples, which also discusses the re-installation of a new liṅga or idol in the place of a damaged one. There are also some short manuals which deal solely with the jīrṇoddhāra.2 The following table shows the ritual texts of Kerala that deal with jīrṇoddhāra: Texts

Chapter Details



Chapter 22 (213 verses)

Unpublished (Manuscripts of this text are available at the IFP, Pondicherry; Oriental Research Institute, Trivandrum; University of Calicut Manuscript Library; Government Oriental Manuscripts Library, Chennai.).

2  Manuscripts of these short texts are available in public as well as private manuscript libraries: 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).



(cont.) Texts

Chapter Details



Chapter 21 (96 verses)

ed. Tripunithura, Kerala.


Vol. IV (pp. 612–615), (52 verses)

ed. Trivandrum Sanskrit Series, Trivandrum, Kerala.


Chapter 24 (93 verses)

ed. Trivandrum Sanskrit Series, Trivandrum., Kerala.


Section II, Chapter 1 (90 verses)

Unpublished (Manuscripts of this text are available at the IFP, Pondicherry; Oriental Research Institute, Trivandrum; University of Calicut Manuscript Library; Government Oriental Manuscripts Library, Chennai.).


Chapter 11 (232 verses)

ed. Panchangam Books, Kunnankulam, Kerala.



Chapter 11 (100 verses)

ed. Trivandrum Sanskrit Series, Trivandrum, Kerala.

Under What Circumstances is jīrṇoddhāra Performed?

As the Prayogamañjarī of Ravi, an eleventh century Śaiva ritual manual of Kerala, observes, jīrṇoddhāra is necessary, when a temple is weak, decayed, burned or when the liṅga installed in a temple breaks or shows visible changes.3 Thus it is essential to perform the jīrṇoddhāra not only when a temple is in a state of decay but also when the liṅga or idol installed in the temple is 3  jīrṇoddhāravidhiṃ vadāmi śithile jīrṇe ‘tha dagddhe ‘thavā /  prāsāde sphuṭite viparyayagate liṅge ‘thavā cobhaye // (Prayogamañjarī 21:1ab).

‘ Re-Installation ’ of Idols Replacing Damaged Ones


damaged. Īśānagurudevapaddhati, another ritual manual of Kerala, indicates various situations which necessitate the renovation of a temple. Compared to earlier writers, the author of this text adds new contingencies such as the theft of the idol, wrong measurements (i.e. either shorter or larger than prescribed), hollowness, or when the idol is washed away by a river or the waves of the ocean etc.4 According to other Kerala ritual texts (for example, Kriyāpaddhati 392), when the idol is removed from its pedestal by thieves or in some other way, when the hand or legs, eyes or nose of the idol are damaged beyond repair, if the idol is broken at throat, stomach or hip then the damaged one should be de-installed and a new one installed. However it is interesting to note that, in certain specific cases, such as when only part of the hand of the idol is broken, the idol need not be replaced, but can be repaired. According to the Viṣṇusaṃhitā (24:61–66)5 the broken portion can be covered or replaced by golden parts. In practice, in certain cases, idols which are known to have been installed by great sages, or whose sanctity is affirmed through astrological predictions need not be replaced even though they are broken or badly damaged but may be protected by a golaka, or a protective covering made of gold or the five-metals (pañcaloha).

4  jīrṇoddhāravidhiṃ vakṣye liṅgaprāsādayaḥ kramāt /  khaṇḍite sphuṭite bhinne calite patite hṛte //  dagdhe jīrṇe ‘ṅgahīne ca sagarbhe vā kṣate cyute /  ajñānād adhike vonamāne liṅge tu mānuṣe //  pratiṣṭhite vā kṣudrārthaṃ liṅge nadyabdhipīḍite /  tyajanti mantrās tal liṅgaṃ devāś cāpi yataḥ sphuṭam //  (Īśānagurudevapaddhati, 4.64:1–3). 5  bāhucchede parityāgaḥ pratimāyāḥ kare tathā /  yasminn avayave bhagne vairūpyaṃ tatra tāṃ tyajet //  yady ekakaraśākhā tu bhagnā dve vātra na tyajet /  ataḥ paraṃ parityāgas tricchede kaiś cid işyate //  pādaśākhāparicchede ‘py evam anyatra kalpayet /  sphuṭite ca parityāgo bhinne ca parikīrtitaḥ //  sauvarṇīṃ sāṅguliḥ kāryā yā bhagnā dve ca te tathā /  lohādau cec chilābimbe tāmreṇa rajatena vā //  heticchede tu sarvatra sauvarṇaṃ tat prakalpayet / makuṭe kuṇḍale caiva vastrādiṣu ca śasyate //  pratimāyāḥ parityāge corāhṛtivad iṣyate /  sthāpanaṃ pūjanaṃ cātra na grāhyaṃ pīṭham akṣatam // (Viṣṇusaṃhitā 24:61–66).

228 4


he Present Situation as Regards the Replacing of Damaged Idols T in Kerala Temples

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 management 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. 5

Two Types of jīrṇoddhāra

According to the Kerala ritual texts there are two types of jīrṇoddhāra, namely niṣkramaṇa and saṅkoca: niṣkrāmaṇena vidhinātra tadantaraṅge vaikalyabhāji vitanotu yathāpuraṃ tat / saṅkocanena bahiraṅgavipady adaś ca niṣkrāmaṇena cirakālacikīrṣitaṃ cet // (Tantrasamuccaya 11:2) 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

6  For a detailed study on this subject see Tarabout 2007. 7  Even though the theme bālālayapratiṣṭhā or constructing a small temporary shrine where the deity will be worshipped while the temple is being renovated, appears in the Kerala ritual

‘ Re-Installation ’ of Idols Replacing Damaged Ones



Rituals Related to Removing of the Damaged Idol

When it is only a damaged liṅga or idol that needs to be deconsecrated and removed and a new one re-installed, a three-day ritual is followed. Among the different rituals performed during such a re-installation, the rite of extracting the life (jīva) from the idol or jīvodvāsana and the uninstalling the damaged idol from its pedestal or bimboddhāraṇa are the most important, and they can be performed only by an experienced priest. While the installation process (pratiṣṭhā) is performed in the ‘order of creation’ (sṛṣṭi) in which the priest makes the presence of the god permeate the entire temple complex, the deconsecration, or extraction of life from the idol (jīvodvāsana), is performed in the ‘order of destruction’ (saṃhāra) in which the priest makes the presence of the god be reabsorbed back into the idol before it is extracted. Now I will mention certain preliminary rituals performed before the previously stated jīvodvāsana ritual. Once the new idol is ready for installation and when the auspicious time has been fixed for the installation, offering of sprouts (aṅkurārpaṇa), cleaning of the prāsāda (prāsādaśuddhi), cleaning of the idol (bimbaśuddhi), the oblations in fire for expiations (prāyaścitta-homa), oblation with the tattva mantras (tattvahoma), worshipping the pot (kalaśapūja) and the ablution with the pot that is worshipped as anujñākalaśa etc., are to be performed. Once these rituals are performed and the oblation (bali) to Kṣetrapāla is finished, the priest should pray in the assembly of the temple administrators to the gods present in the entire temple complex about the necessity of the re-installation process thus: śṛnvantu devatāḥ sarvāḥ prāsādam imam āśritāḥ / yad ucyamānam asmābhir anugṛhṇantu tena naḥ // pūrvaiḥ pūrvaṃ kṛtam idaṃ devālayam asāravat / prāptaṃ kālavaśāt bhūyo vayaṃ tat kartum udyatāḥ // kālenaitāvtā bhūyaḥ prāsāde ‘smin punar nave / prāpayāmo vayaṃ satyaṃ tad anuñjātum arhatha //8

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.


Figure 1


Obtaining the idol for installation from the craftsman (bimbaparigraha). Photo Ajithan.

‘ Re-Installation ’ of Idols Replacing Damaged Ones


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 musical 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: pūjābimbam idaṃ viṣṇoḥ sthāpitaṃ pūrvasūribhiḥ / anena dūṣaṇenādya dūṣitaṃ ceha varttate // duṣṭabimbasya coddhāraḥ kartavya iti śāsanam / āgamasya vayaṃ kartuṃ vyavasāyam upāsmahe // atredaṃ yadi kartavyam āgamārthatayā bhavet / bhavanto no ‘nujānantu bhavemājñākarā vayam //

Figure 2

Opening of the eyes (netronmīlana). Photo Ajithan.



This is followed by a prayer to the devotees assembled there to make them aware that the liṅga or idol which is damaged is equal to the ‘remains of an offering’ (nirmālya): yathaiva devapūjāyāṃ viniyuktam aninditam / dravyaṃ puṣpādikaṃ paścān nirmālyam iti nindyate // evaṃ biṃbam aduṣṭaṃ yat pūjāyāṃ viniyujyate / tad eva dūṣitaṃ paścān nirmālyam iti nindyate // tasmād duṣṭam idaṃ biṃbaṃ bhavadbhir muktasaṃśayaiḥ / nirmālyabuddhyā tyaktavyam iti śāstrasya śāsanam // 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.

Figure 3

Placing the idol in water for a stipulated time (jalādhivāsa). Photo Ajithan.

‘ Re-Installation ’ of Idols Replacing Damaged Ones


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 imagines the God as `cinmātra’ and does the removal of the connection of channels (nāḍīviccheda), and then opens the sanctum. 7


Then in front of the idol a pot meant for the jīvakalaśa is installed and worshipped with necessary rituals. 8


Next the priest, holding flowers in both hands and with the prescribed mantras, 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 placing 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 9



Then the priest, accompanied by four brahmins, perambulates the temple and enters the sanctum, all of them wearing a thread on the wrist (kautuka). Puṇyāha is performed and the puṇyāha water sprinkled on the four brahmins who then change the position of their sacred thread (yajñopavīta) looping it as if it were a garland; then they all do the prāṇāyama with the mantras of Nirmālyadhāri. Next, the priest, holding flowers in both hands, absorbs the vāsanāśakti from the heart of the God while reciting oṃ udvāsayāmi (udvāsana) and takes it from the idol to the sūryamaṇḍala. After the recitation of prescribed sūktas, the priest and the four brahmins remove the medicinal paste placed between the seat and the idol (aṣṭabandha) installed in the pīṭha with the help of a golden axe. A rope made of grass and the hair from a cow’s tail is tied to the idol and the other end to the neck of a bull standing in front of the sanctum. Then all five pull the idol from its pīṭha and replace it supine on the pīṭha, with the head on the southern side, and wash it with water. The priest once more moves the vāsanāśakti from the idol to the sūryabimba, and covers it with cloth. The idol is brought out of the sanctum with the head foremost. It is then placed with its head in the south, on the southern side of the saptamātṛs which are installed in the prākāra, on the darbha grass, with its tips facing south. All five untie their tufts, and holding sesame and rice in their hands they spread these around the idol while doing three pradakṣiṇa (circumambulation) anti-clockwise [fig. 4]. Then they remove the cloth and wash the idol again. The priest takes the remaining vāsanāśakti from the idol to the Sun, covers the idol with cloth and carries it on a ladder to a body of water (jalāśaya), a pond, river etc. Standing beside the water the priest purifies himself with the mantras of Nirmālyadevatā and performs the pīṭhapūjā in the deep part of the water; he then invokes the Nirmālyadevatā into the idol, worships it and drops it into the water. If the idol is made of wood it is taken out of the temple and offered in the fire. Then the brahmins, and everyone assembled there, take a bath and return to the temple. The pīṭha and napuṃsakaśilā are also to be dropped into the water if they are going to be replaced with new ones. If they are to be reused then they should be kept in pits. That evening, necessary rituals are to be performed for the jīvakalaśa and also for the idol to be installed, both of which are kept on adhivāsa. The priest remains there all night without sleeping. The next day after the daily routines, and other necessary rituals, the new idol, the jīvakalaśa etc. are taken in procession from the adhivāsamaṇḍapa. The new idol is installed in the pīṭha with necessary rituals, such as, parāvāhana, avasthāvāhana,

‘ Re-Installation ’ of Idols Replacing Damaged Ones

Figure 4

The damaged idols are kept on the southern side of the temple. Photo Ajithan.




mūlāvāhana, ṣaḍadhvanyāsa etc. Then, while performing the jīvāvāhana, the priest slowly pours the water from the jīvakalaśa three times while reciting the mantras prescribed for the jīvāvāhana. Sthūlāvāhana, stotrāvāhana and mantrāvāhana are also to be recited. On the fourth day, special ablutions are to be performed such as one with 1000 pots or more [fig. 5]. 10

Expiatory Rituals Related to the ‘Installation’

Like the late South Indian Śaiva ritual manuals, the ritual texts of Kerala describe expiations related to temples, some of which are particularly associated 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 occasions 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.

Figure 5

1000 pots after the rituals. Photo Ajithan.

‘ Re-Installation ’ of Idols Replacing Damaged Ones



Case of De-Installation and Re-Installation Performed to Protect A an Age-Old Idol

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 especially, its precious main-idol (mūlavigraha) installed in the temple. The priests thought of a way to protect the idol: the caitanya from the mainidol 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 renovation 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



extensively during rituals, especially during the re-installation process during which more than twenty-five hymns are chanted. Even though the involvement of astrologers in matters such as the fixing of the mode of renovation required for a temple, is not specifically mentioned in the texts, nowadays in practice, no renovation or installation processes are carried out in Kerala without an astrological consultation (devapraśna). It is fascinating to note that, even today, the rituals related to temples follow the ritual manuals strictly in most temples in Kerala, and many vernacular ritual manuals for the use of priests are regularly published. An example is the recently published Kriyāpaddhati, a prose ritual manual written in Malayalam based on the well-known Tantrasamuccaya. References

Manuscripts and Printed Sources

Īśānagurudevapaddhati of Īśānagurudeva, ed. T. Gaṇapati Śāstri, 4 Vols., Delhi: Bharatiya Vidya Prakashan, 1990. (Reprinted with a substantial new introduction dated to 1987 by N.P. Unni, from Trivandrum Sanskrit Series Nos. 69, 72, 77 and 83, Trivandrum, 1920, 1921, 1922, 1925.) Kalaśacandrikā, IFP MS T. 879. Paper. Kiraṇatantra see s.v. Dominic Goodall Kriyādīpikā or Puṭayūrbhāṣā of Vāsudeva, ed. Raman, Kunnaṃkulam (Kerala): Panchangam Books, 1988. Kriyāpaddhati, ed. Kāraṇatt Śrīdharan Nampūtiri, Trichur (Kerala): Sripuram Publications, 2010. Jñānaratnāvalī of Jñānaśiva. Oriental Research Institute, Mysore, MS P 3801. Palm-leaf, Nandināgari. Also GOML MS R 14898 and its appograph IFP MS T. 231, as well as pp. 13–60 of IFP MS T. 106, all paper transcripts in Devanāgarī. Tantrasamuccaya of Nārāyaṇa, with the commetnary Vimarśinī of Śaṅkara, ed. T. Ganapati Śāstri, Delhi: Nag Publishers, 1990. (Reprinted with a new introduction by N.P. Unni, from Trivandrum Sanskrit Series Nos., 67 and 71, Trivandrum, 1919, 1921.) Tantrasamuccaya of Nārāyaṇa, with the commentary Vimarśinī of Śaṅkara and Vivaraṇa of Nārāyaṇaśiṣya, in three parts. Part I (Paṭalas 1 to 4) ed. V.A. Ramaswami Sastri, Part II (Paṭalas 5 to 8) ed. K.S. Mahadeva Sastri, Part III (Paṭalas 9 to 12) ed. K. Raghavan Pillai. Trivandrum Sanskrit Series Nos., 151, 169, 200, Trivandrum, 1945, 1953, 1962. Tantrasamuccaya of Nārāyaṇa with the Malayalam translation in three volumes, by K.P.C. Anujan Bhattatirippad. Kunnamkulam (Kerala): Panchangam Books, 1989. 1991, 1992.

‘ Re-Installation ’ of Idols Replacing Damaged Ones


Niśvāsatattvasaṃhitā, see s.v. Goodall, Dominic 2015. Puṭayūrbhāṣā, see s.v. Kriyādīpikā. Prayogamañjarī of Ravi, ed. Si. Ke. Rāman Nampiyar with Ke. Acyutappotuvāl. Tripunithura: Sanskrit College, 1953–4. Pratiṣṭhālakṣaṇasārasamuccaya, ed. Dāmodaraśarman and Bābukṛṣṇaśarman. Kathmandu: Nepāla Rājakīya Pustakālaya, 1966 and 1968 (Vikram Saṃvat 2023 and 2025). Prāyaścittasamuccaya of Trilocalaśiva, see s.v. Satyanarayanan, R., 2015. Viṣṇusaṃhitā of Sumati, ed. T. Gaṇapati Śāstri, Delhi: Nag Publihseres, 1991. (Reprinted with a new introduction by N.P. Unni, from Trivandrum Sanskrit Series No. 85, Trivandrum, 1925.) Śeṣasamuccaya with the Malayalam commentary by Maheśvaran Bhaṭṭatirippāṭ. ed. Divakaran Nambutirippad, Aluva (Kerala): Tantravidyapitham, nd. Śaivāgamanibandhana, IFP MS T. 379. Paper. Sarvajñānottara. NAK MS 1–1692. NGMPP Reel No. A 43/12. Palm-leaf, early Nepalese ‘Licchavi’ script. Also GOML R 16829, GOML MS D 155595 and IFP T. 334, paper transcripts in Devanāgarī, Telugu-script and Devanāgarī respectively. Sarvajñānottaravṛtti of Aghoraśiva. Oriental Research Institute and Manuscripts Library, Trivandrum, MS 6578. Grantha script, palm-leaf, probably originally from Madurai; 199 folios. Also IFP RE 47852, Grantha script, paper, 202 pages. Somaśambhupaddhati see s.v. Brunner, Hélène.

Translations and Studies

Anujan Bhattatirippad, K.P.C. 1989, 1991, 1992. see s.v. Tantrasamuccaya. Brunner, Hélène (ed. and transl.) 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. Rituels 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 sivaïte selon Somaśambhu. Quatrième partie: rituels optionnels: pratiṣṭhā. Publications de l’IFI No. 25. Pondicherry: IFI. Czerniak-Drożdżowicz, Marzenna. 2006. ‘Viṣṇusaṃhitā’s five-fold classifications and the explanation of the name Pāñcarātra.’ Cracow Indological Studies 8, 131–148. Goodall, Dominic (ed. and transl.) 1998. Bhaṭṭa Rāmakaṇṭha’s Commentary on the Kiraṇatantra. Volume I: chapters 1–6. Critical edition and annotated translation. Publications du département d’indologie 86.1. Pondicherry: Institut français de Pondichéry / École française d’Extrême-Orient. Goodall, Dominic. 2015. Niśvāsatattvasaṃhitā, The earliest surviving Śaiva Tantra, Vol. I, A critical edition and annotated translation of the Mūlasūtra, Uttarasūtra &



Nayasūtra, eds Dominic Goodall et al. Pondicherry: Institut français de Pondichéry / École française d’Extrême-Orient. Satyanarayanan, R. 2015. Śaiva rites of expiation, A first edition and translation of Trilocanaśiva’s twelfth-century Prāyaścittasamuccaya (with a transcrip­ tion of Hṛdayaśiva’s Prāyaścittasamuccaya), critically edited and translated by R. Satyanarayanan with an introduction by Dominic Goodall. Pondicherry: Institut français de Pondichéry / École française d’Extrême-Orient. Sarma, S.A.S. 2009. ‘The eclectic paddhatis of Kerala.’ In: Indologica Taurinensia 3, 319–340. Sarma, S.A.S. forthc. ‘The synthesis of Śaiva and Vaiṣṇava systems of worship in ritual manuals—a unique feature of the ritual literature of Kerala. Paper presented in the ‘Agamas and Tantras’ section of the 14th World Sanskrit Conference, Kyoto, Japan, 1–5 September, 2009. Tarabout, Gilles. 2007. ‘Authoritative Statements in Kerala Temple Astrology.’ Rivista di Studi Sudasiatici 2, 2007, 85–120. Unni, N.P. 1991. see s.v. Viṣṇusaṃhitā.


The Planting of Trees and the Dedication of a Garden: A Comparison with the Image Pratiṣṭhā 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 garden are included into the utsarga, and in this article I first analyse the descriptions 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: HirGŚS 1.7.3 [96,22–98,25].3 Mbh 13.99.22cd–33 (see Kane 1974, pp. 894–895).4 Bhaviṣyapurāṇa–90.5

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 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.’

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_011



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–29.13 Bhaviṣyapurāṇa–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 accomplishing 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 consecration 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 corresponds 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 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


Bhaviṣyapurāṇa–36ab.15 Bhaviṣyapurāṇa–74.16 Bhaviṣyapurāṇa–7 =/ Bhaviṣyapurāṇa–7.17 Bhaviṣyapurāṇa–10.18 Bhaviṣyapurāṇa–33.19 Bhaviṣyapurāṇa–5.20 Bhaviṣyapurāṇa–13.21 Bhaviṣyapurāṇa–4.22 Bhaviṣyapurāṇa–11.23 Bhaviṣyapurāṇa–6.24 Bhaviṣyapurāṇa–18.25 Bhaviṣyapurāṇa 4.128.1–45.26 15  It concerns the dedication of a garden: 1a ārāmādau viśeṣo yo vakṣyate, ‘Special rules concerning a garden and others are explained.’ 16   A garden is suggested in the dedication mantra given in . .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 planting 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–9ac corresponds to Bhaviṣyapurāṇa–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.’



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 mentioned in the previous notes 3 to 26, some Purāṇas such as Agnipurāṇa 70.1a pratiṣṭhāṃ pādapānāṃ, Bhaviṣyapurāṇa =/ Bhaviṣyapurāṇa kṣudrārāmapratiṣṭhāṃ, Bhaviṣyapurāṇa pratiṣṭhāvidhivistaram, Bhaviṣyapurāṇa athāśvatthapratiṣṭhāyāṃ, Bhaviṣyapurāṇa vakṣye bilvapratiṣṭhāṃ, Bhaviṣyapurāṇa puṣpārāmapratiṣṭhāṃ and Bhaviṣyapurāṇa 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 established 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 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


Among the ritual texts that treat the rites of tree planting and the dedication of a garden, the oldest one may be ŚāṅkhGS 5.3.1–5. There are two examples 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.–36ab consecration of a garden (ārāmādipratiṣṭhā),–50 the construction of a dam (setubandhana),–15ab consecration of pastureland (gopracārapratiṣṭhā),–18 directions of the boundaries of various constructions, bad results when one digs or ploughs pastureland,–26 merits of a donation of gocarman,28– 48ab consecration of a temple (maṇḍapapratiṣṭhā),–48ab consecration of a place for supplying water (prapāpratiṣṭhā),–69 consecration of a well (kūpapratiṣṭhā),–74 consecration of a garden,–7 consecration of a small garden (kṣudrārāmapratiṣṭhā),–10 the rite of planting one tree or many trees (ekādivṛkṣāṇāṃ vidhi),–36 consecration of a reservoir of water,–6 consecration of a reservoir of water,–33 consecration of a garden,–7 consecration of a small garden,–5 the rite of planting one tree or many trees,–13 planting of aśvattha trees,–4 planting of vaṭa trees,–11 planting of bilva trees,–20 rātripratiṣṭhā or seemingly a rite of planting a certain tree the main actions of which are performed in the night (?),–4 consecration of a Viṣṇu temple(?),–12ab consecration of a temple,–14ab consecration of a tṛṇaveśman or a house made of grasses (?),–15 consecration of a place for supplying water,–17 consecration of a well,–6 consecration of a flower gardens (puṣpārāmapratiṣṭhā),–18 planting of tulasī plants,–20ab the construction of a dam,–23 the construction of a small dam (kṣudrasetubandhana),–18 consecration of pastureland,–10 a consecration ritual performed only one day 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 gavāṃ śataṃ vṛṣaś caiko yatra tiṣṭhaty ayantritaḥ, ‘that land which a hundred 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.



for a poor man,–29 consecrations of different deities (1–8 of Kālī/ Durgā, 9–14ab of the Śivaliṅga, 14cd–18 of a śālagrāma stone, 19–22ab of Sūrya, 22cd–24 of Vārāhī and Tripurā, 25–29 of Bhuvaneśvarī, Mahāmāyā, Ambikā, Kāmākṣī, Indrākṣī and Aparājitā),–179 appeasement of evil omens. 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–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 madhyamaparvan, 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 useful, 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


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 beneficiaries. 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 shadows.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 living 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 30  Mbh 13.99.30ab puṣpitāḥ phalavantaś ca tarpayantīha mānavān. 31  Viṣṇudharmottarapurāṇa 3.297.20 patrapuṣpaphalair nityaṃ chāyayā caiva śākhinaḥ / pareṣām upayujyante tad dhi teṣāṃ mahat tapaḥ /20/. 32  Mbh 13.99.28 puṣpaiḥ suragaṇān vṛkṣāḥ phalaiś cāpi tathā pitṝn / chāyayā cātithīṃs tāta pūjayanti mahīruhāḥ /28/ Padmapurāna 6.27.15cd–16ab puṣpaiḥ suragaṇān patraiś cāpi tathā pitṛṛn /15/ chāyayā cātithīn sarvān pūjayanti mahīruhāḥ /. 33  Viṣṇudharmottarapurāṇa 3.297.13cd–14ab devān prasūnaiḥ prīṇāti chāyayā cātithīṃs tathā /13/ phalair manuṣyān prīṇāti. 34  Viṣṇusmṛti 91.5–8 vṛkṣaprado vṛkṣaprasūnair devān prīṇayati /5/ phalaiś cātithīn /6/ chāyayā cābhyāgatān /7/ deve varṣaty udakena pitṝn /8/. 35  Bhaviṣyapurāṇa 4.128.4–5ab prāṇinaḥ prīṇayanti sma cchāyāvalkalapallavaiḥ / ghanacchadāḥ sutaravaḥ puṣpair devān phalaiḥ pitṝn /4/ puṣpapatraphalacchāyāmūlav alkaladārubhiḥ /. 36  Padmapurāṇa 1.58.6 patanti yāni patrāṇi jale parvaṇi parvaṇi / tāni piṇḍasamānīha pitṝṇām akṣayaṃ yayuḥ /6/. 37  Viṣṇudharmottarapurāṇa 3.297.15cd–16ab deve varṣati yad vṛkṣāt patrebhyaḥ sravate jalam /15/ tena tṛptim avāpnoti paralokagato dhruvam /. 38  Padmapurāṇa 1.58.7 khādanti patagās tatra phalāni kāmato dhruvam / brahmabhakṣyasamaṃ tasya puṇyaṃ bhavati cākṣayam /7/ Padmapurāṇa 1.58.34cd– 35ab phalāni yasya khādanti jantavo ‘tha sahasraśaḥ /34/ āśritā vihagāḥ kīṭāḥ patagāḥ śalabhādayaḥ/. 39   See Padmapurāṇa 1.58.9 uṣṇe cchāyāṃ pragṛhṇanti gāvo devadvijātayaḥ / kartuḥ pitṛgaṇānāṃ ca svargo bhavati cākṣayaḥ /9/ Padmapurāṇa 1.58.35cd–36ab chāyāśritāś ca



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 mentioned. 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 importance 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 different 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 committed in the past three births perish and the heavenly world will be attained.47 ye satvās tatsaṃkhyātāḥ pṛthagjanāḥ /35/ tasya kiṃkaratām yānti śataśo devatārcitāḥ / See also Bhaviṣyapurāṇa 4.128.3cd–4ab, Viṣṇudharmottarapurāṇa 3.297.14cd–15ab, Bhaviṣyapurāṇa 4.128.7 and 15. 40  HirGŚS 1.7.3 [98,16] āropya vidhivad dhīmān anantaphalam aśnute //. 41  HirGŚS 1.7.3 [97,24–25] pippalaḥ śaṅkaradvāri vaṭo mārge catuṣpathe / jalāśaye gavāṃ goṣṭhe ropitaḥ sarvakāmadaḥ //. 42  Matsyapurāṇa 59.17 anena vidhinā yas tu kuryād vṛkṣotsavaṃ budhaḥ / sarvān kāmān avāpnoti phalaṃ cānantyam aśnute /17/ See also Padmapurāṇa 1.28.18cd–19ab. Padmapurāṇa 1.58.33cd–34ab mentions various wishes such as heaven, enjoyment, kingdom and so on. 43   HirGŚS 1.7.3 [97,16–17] dhātrīkapitthabilvānāṃ ropaṇaṃ kīrtivardhanam. See also Bhaviṣyapurāṇa 44  Mbh 13.99.24cd–25 kīrtiś ca mānuṣe loke pretya caiva phalaṃ śubham /24/ labhate nāma loke ca pitṛbhiś ca mahīyate / devalokagatasyāpi nāma tasya na naśyati /25/. 45  Bhaviṣyapurāṇa 4.128.16 ataḥ paraṃ pravakṣyāmi vṛkṣasyodyāpane vidhim / sarvapāpapraśamanaṃ sarvakīrtivivardhanam /16/. 46  Agnipurāṇa 70.8cd pāpanāśaḥ parā siddhir vṛkṣārāmapratiṣṭhayā /8/. 47  Bhaviṣyapurāṇa janmatrayādikaṃ pāpaṃ vināśya svargam ādiśet /.

The Planting Of Trees And The Dedication Of A Garden


Many passages mention heaven as a reward of planting trees or dedicating a garden. One who plants various trees in various places goes to heaven (Padmapurāṇa 1.58.12–13), stays there pleasantly (Mbh 13.99.32, Padmapurāṇa 1.58.40–42ab, Viṣṇudharmottarapurāṇa 3.297.4cd–5 and 11cd–12) and for a very long time.48 Such a person is not driven away from heaven49 and the yonder world one attains becomes imperishable.50 One who plants an aśvattha tree goes to Viṣṇu’s world,51 or when one plants trees, one goes to Yama’s world (Viṣṇudharmottarapurāṇa 3.297.24–27). When a person plants various trees in various places in various seasons, he goes to the identification with Brahmā, ‘brahmasāyujya’ (HirGŚS 1.7.3 [98,17–25]). According to Bhaviṣyapurāṇa one who plants a hundred trees becomes Brahmā and one who plants a thousand trees becomes Viṣṇu. Padmapurāṇa 1.58.38 says that one who serves mango trees becomes Gaṇeśa. According to HirGŚS 1.7.3 [97,26– 29] by planting a nimba tree together with an aśvattha tree on a crossway Śiva will be worshipped.52 HirGŚS 1.7.3 [97,17] simply says that Śaṃkara/Śiva will be pleased by planting trees. In the post-Vedic period a certain ritual is evaluated by equating it with a certain śrauta ritual (Olivelle 1993: 54, Bodewitz 1999: 114). Viṣṇudharmottarapurāṇa 3.297.1cd–4ab hands down this kind of evaluation. One who plants bushes (gulma) or creepers (vallī and latā) obtains merit of the godāna, by planting trees bearing blossoms (puṣpavṛkṣa) and fruit trees (phalavṛkṣa) rewards 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 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 desiring 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 the planting of nimba brings damages to domestic animals (paśuvināśana) and Bhaviṣyapurāṇa counts it among trees the consecration of which is not to be performed.



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. kapittha, 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 enjoyment 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 person 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–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 vaṭarūpī śivo yataḥ.

The Planting Of Trees And The Dedication Of A Garden


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 one liberates his ancestors from pāpa, namely evil or the bad situation in the yonder world. A similar idea was expressed concerning 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 various 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 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 number 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 (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 (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–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).



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 religious 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

63  Mbh 13.99.30–31 puṣpitāḥ phalavantaz ca tarpayantīha mānavān / vṛkṣadaṃ putravad vṛkṣās tārayanti paratra ca /30/ tasmāt taḍāge vṛkṣā vai ropyāḥ śreyorthinā sadā / putravat paripālyāś ca putrās te dharmataḥ smṛtāḥ /31/ Padmapurāṇa 6.27.17cd–18ab is a combination of Mbh 13.99.30ab and 31cd. See Kane 1973: 895. 64  Viṣṇusmṛti 91.4 vṛkṣāropayitur vṛkṣāḥ paraloke putrā bhavanti /4/. 65  Mbh 13.99.27ab tasya putrā bhavanty ete pādapā nātra saṃśayaḥ, ‘these trees become his sons, there is no doubt concerning this,’ and Padmapurāṇa 6.27.14cd putrapautrā bhavanty ete pādapā nātra saṃśayaḥ. 66  Viṣṇudharmottarapurāṇa 3.297.13ab eko ‘pi ropito vṛkṣaḥ putrakāryakaro bhavet / ‘Even one tree planted may do a duty of the son.’ 67  Padmapurāṇa 1.28.22cd–23ab aputrasya ca putritvaṃ pādapā eva kurvate /22/ tīrtheṣu piṇḍadānādīn ropakāṇāṃ dadanti te / ‘Trees take the place of the son for one who has no sons. On the banks of rivers they offer rice balls for the sake of those who planted them.’ Viṣṇudharmottarapurāṇa 3.297.31 putrāḥ pradiṣṭāḥ puruṣasya vṛkṣāḥ svayaṃ kṛtās teṣu nareṇa bhāvyam / snehena nityaṃ puruṣaṃ mṛtaṃ te kāmais tu divyaiḥ paritarpayanti /31/ ‘It is said that trees are sons of a man who planted them by himself. He may have future in them. They affectionately satisfy the dead person with divine wishes forever.’ 68  Bhaviṣyapurāṇa aputrasya hi putratvaṃ pādapā iha kurvate / yatnenāpi ca viprendrā aśvatthāropaṇaṃ kuru /37/ śataiḥ putrasahasrāṇām eka eva viśiṣyate / See also Padmapurāṇa 1.28.24ab. 69  Bhaviṣyapurāṇa 4.128.45 putrair vinā śubhagatir na bhaven narāṇām duṣputrakair iti tathobhayalokanāśaḥ / etad vicārya sudhiyā paripālya vṛkṣān putrāḥ purāṇavidhinā parikalpanīyāḥ /45/ ‘Without sons there may be no happy future for the people. With bad sons there will be loss of both worlds. Thus pondering, a wise man should keep trees and endeavour to have sons according to the rule of the Purāṇas.’

The Planting Of Trees And The Dedication Of A Garden


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 description of the tree planting very often exalted the usefulness of the trees themselves 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 sonlessness 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/.



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 society to invent some religious motivations for inspiring rich people such as kings, ministers and landowners to plant trees continually, to release a bull repeatedly 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 planting 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 protects 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


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 explanation: 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 mentioned 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.



[174,4–24], Viṣṇusmṛti 86.1–20, Bhaviṣyapurāṇa 4.131.1–22, Devīpurāṇa 60.1– 13, Garuḍapurāṇa 2.5.39cd-46ab (in the pretakalpa), Garuḍapurāṇa 2.6.1–27, Garuḍapurāṇa 2.41.1–13, Matsyapurāṇa 207.1–41, Viṣṇudharmottarapurāṇa 1.146.40cd-59 and 1.147.1–19. In these texts the relationship with the ancestor worship is expressed in three ways. First, the effects that concern the ancestors are introduced.82 The vṛṣotsarga liberates the ancestors or the dead person: When the bull drinks, eats, raises up the tail fully wet with water, it pleases the gods, the seers and the ancestors and it raises the past and future kinsmen up to the seven generation on both sides.83 What is set free for the sake of the dead person liberates the dead person from great hell.84 The fathers are satisfied: The bull and the four young heifers drink water and shake off drops of water, then a mantra is recited: May the ancestors go, being satisfied.85 The Garuḍapurāṇa emphasizes that the dead person attains mokṣa.86 And the pitṛs are provided with water and food sufficiently: When the bull which has been set free and drinks water in a reservoir, the whole reservoir serves the ancestors of the performer; when the proud bull makes lines anywhere upon the ground with its horn, then food and drink serves the ancestors abundantly.87 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 prosperity 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 reading 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


Second, the vṛṣotsarga is gradually connected with the rite for the dead person. 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: 88  Viṣṇudharmottarapurāṇa 1.147.3ab mṛtāho yasya tasyārthe tasminn ahani śaṃkaram. 89  ĀśvGPŚ 3.18 [174,18] tasmād ekādaśe ‘hni pretāya vṛṣam utsṛjet. Garuḍapurāṇa 2.5.39–46 repeatedly insists that the vṛṣotsarga is to be performed on the eleventh day after death. 90  Bhaviṣyapurāṇa 4.36.47a nārāyaṇabaliḥ kāryaḥ sarpadaṣṭasya dehinaḥ / dāne piṇḍapradāne ca brāhmaṇānāṃ ca tarpayet /46/ vṛṣotsargas tu kartavyo gate saṃvatsare nṛpa (Einoo 1994: 6). 91  Mbh 3.82.71–72 tato gayāṃ samāsādya brahmacārī jitendriyaḥ / aśvamedham avāpnoti gamanād eva bhārata /71/ tatrākṣayavaṭo nāma triṣu lokeṣu viśrutaḥ / pitṝṇāṃ tatra vai dattam akṣayaṃ bhavati prabho /72/ ‘From there he arrives at Gayā, observing chastity and subjugating the senses. He attains the reward of the aśvamedha only by going there, o descendant of Bharata. There is an undecaying banyan tree famous in the three worlds; what is given to the ancestors there becomes undecaying, o King.’ 92  Mbh 13.88.14 eṣṭavyā bahavaḥ putrā yady eko ‘pi gayāṃ vrajet / yatrāsau prathito lokeṣv akṣayyakaraṇo vaṭaḥ /14/. 93  Mbh 3.85.5–8 yatra sā gomatī puṇyā ramyā devarṣisevitā / yajnabhūmiś ca devānāṃ śāmitraṃ ca vivasvataḥ /5/ tasyāṃ girivaraḥ puṇyo gayo rājarṣisatkṛtaḥ / śivaṃ brahmasaro yatra sevitaṃ tridaśarṣibhiḥ /6/ yadarthaṃ puruṣavyāghra kīrtayanti purātanāḥ / eṣṭavyā bahavaḥ putrā yady eko ‘pi gayāṃ vrajet /7/ mahānadī ca tatraiva tathā gayāziro ‘nagha / yatrāsau kīrtyate viprair akṣayyakaraṇo vaṭaḥ / yatra dattaṃ pitṛbhyo ‘nnam akṣayyaṃ bhavati prabho /8/ ‘There flows the lovely and holy River Gomatī, which is visited by Gods and seers, the sacrificial terrain of the Gods, and the slaughter site of Vivasvat. The great sacred mountain of Gaya is there, which is honored by royal seers, the



Kṛṣṇaśuklāv ubhau pakṣau gayāyāṃ yo vasen naraḥ / punāty āsaptamaṃ rājan kulaṃ nāsty atra saṃśayaḥ /84/ eṣṭavyā bahavaḥ putrā yady eko ‘pi gayāṃ vrajet / yajeta vāśvamedhena nīlaṃ vā vṛṣam utsṛjet // ‘When one stays in Gayā for two dark and bright fortnights, he purifies his family up to seven generations, there is no doubt in this point. It would be desirable to have many sons; if even one son goes to Gayā, or if one performs the aśvamedha or if one sets free a bull of dark colour.’ Here the greatness of Gayā is not due to its akṣayavaṭa, but because that by staying there for a month seven generations of one’s ancestors are purified or liberated 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


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 generations 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 mountain 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 concern for the dead person (Einoo 2002: 713–712). The idea that the dead father or fathers are cared and worshipped properly 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

96  See Einoo 2002: 711–710. 97  Mbh 13.99.16 sa kulaṃ tārayet sarvaṃ yasya khāte jalāśaye / gāvaḥ pibanti pānīyaṃ sādhavaś ca narāḥ sadā /16/ Padmapurāṇa 6.27.3 is very similar to it. 98  BodhGŚS 4.4.18 evaṃ prayuñjāno daśa pūrvān daśāparān ātmānaṃ caikaviṃśatiṃ paṅktiṃ ca punāti /18/ See also Bhaviṣyapurāṇa 4.127.59–60ac, 60d-65ab, 67cd-69ab, 69cd–70ab, 76, 88 and Padmapurāṇa 6.32.53cd–54ab, 56cd–57ab. 99  See AVPŚ 39.1.7, PārGSPŚ [404,8], ĀśvGPA 29 [262,12–263,3], ĀśvGPŚ 4.9 [179,22–24], Bhaviṣyapurāṇa–223ab, Bhaviṣyapurāṇa 4.127.32cd-34cd, Matsyapurāṇa 58.42– 43, Padmapurāṇa 1.27.44–45 The similar action is prescribed also in ĀśvGPŚ 4.9 [179,22– 24], Bhaviṣyapurāṇa 4.127.32cd-34ab and Somaśambhupaddhati 4.12.16–17ab. 100  ĀśvGPA 29 [263,1] svayaṃ pucchāgre lagnaḥ anvārabdha uttīrya.



treated repeatedly by a number of the Purāṇas101 and the Dharmanibandhas102 and it is performed even today regularly. It may be the most popular religious rite among the Hindu people. Therefore the idea underlying the śrāddha was employed to motivate the people to perform the group of rites called utsarga. While the image of a deity itself is religiously meritorious, the objects of the utsarga such as trees, reservoirs of water and the bulls are only regarded as meritorious when they help others. The other difference between the pratiṣṭhā and the utsarga is that the temple construction and the image installation are performed as religious ceremonies often even today in India, but the rites belonging to the category of the utsarga have seemingly stopped being religious performances. The pratimāpratiṣṭhā may be the climax of a series of ceremonies performed in the course of the construction of a temple. When a temple is once established and a beautiful statue of a meritorious deity is installed, this temple may become a center of the pilgrimage and it produces merits for the donor of the temple and the statue. The temple requires a number of priests to worship the deity and to maintain the temple. If the temple becomes a well-known centre of pilgrimage, it creates further job opportunities for many Brahmins. On the other hand, a bull works for a certain time. A water reservoir may continue to be useful for agriculture as a source of irrigation and other facilities. Trees once planted begin to grow and begin to bring useful things to the people. But these three do not produce further occasions of employment of Brahmin priests. The very fact that the utsarga stops to be performed widely is another important difference of the utsarga from the pratiṣṭhā of a deity image. 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


AVPŚ: Atharvavedapariśiṣṭa, The Pariśiṣṭas of the Atharvaveda, Edited by G.M. Bolling and J. von Negelein, Vol. I: Text and Critical Apparatus, Leipzig: Otto Harrassowitz, 1909, 1910. ĀśvGPA, Āśvalāyanagṛhyapariśiṣṭa, Edited by K.P. Aithal, The Adyar Library Bulletin, Vol. 27, pp. 217–287, 1963. ĀśvGPŚ, Āśvalāyanīyaṃ gṛhyapariśiṣṭam, Edited by V.G. Apte, Ananda Ashrama Sanskrit Series 105, pp. 141–183, Varanasi, 1936. ĀśvGS, Āśvalāyanagṛhyasūtra-Bhāṣyam of Devasvāmin, Critically Edited with an Introduction by Pandit K.P. Aithal, Madras: The Adyar Library and Research Centre, 1980. Bṛhatsaṃhitā, The Brihat Saṃhitā by Varāhamihira with the Commentary of Bhaṭṭotpala Edited by Mahāmahopādhyāya Sudhākara Dvivedī, The Vizianagram Sanskrit Series, vol. X, part I, 1895, part II, 1897, Reprint, Varanasi: Benares Sanskrit University, 1968. Bhaviṣyapurāṇa, The Bhaviṣyamahāpurāṇam, 3 Vols, Delhi: Nag Publishers, 1984, 1985. BhārPŚS, Bhāradvājapariśeṣasūtra, Edited by C.G. Kashikar, the Śrauta, Paitṛmedhika and Pariśeṣa Sūtras of Bharadvāja, Poona: Vaidika Saṃśodhana Maṇḍala, 1964. Devīpurāṇa, the Devī Purāṇa, Edited by Acarya Pancanan Tarkaratna, Calcutta: Nava Bharat Publishers, 1975. GB, Das Gopatha Brāhmaṇa, herausgegeben von Dieuke Gaastra, Leiden: E.J. Brill, 1919. GautDhŚ, Śrīgautamadharmaśāstram: The Institutes of Gautama, Edited with an Index of Words by Adolf Friedrich Stenzler, London: Trübner & Co., 1876. Harivaṃśa, The Harivaṃśa, Being the Khila or Supplement to the Mahābhārata, for the First Time Critically Edited by Parashuram Lakshman Vaidya, 2 Vols., Poona: Bhandarkar Oriental Research Institute, 1969, 1971. Harivaṃśapurāṇa, The Harivaṃśapurāṇam, Delhi: Nag Publishers, 1985. HirGŚS: Hiraṇyakeśigṛhyaśeṣasūtra, Satyāṣāḍhaviracitaṃ Śrautasūtram, Vol. 8, after Praśnas 19 and 20 that are the Hiraṇyakeśigṛhyasūtra, with new pages 1–126, Ananda Ashrama Sanskrit Series 53, Varanasi, 1929. JaimGS, The Jaiminigṛhyasūtra Belonging to the Sāmaveda, Edited with an Introduction by W. Caland, 1922, Lahore: The Punjab Sanskrit Book Depot. KāṭhGS, The Kāṭhakagṛhyasūtra with Extracts from Three Commentaries, an Appendix and Indexes, Edited by Willem Caland, Lahore: The Research Department, D. A. V. College, 1925. Kālikāpurāṇa, The Kālikāpurāṇa (Text, Introduction & Translation in English), B.N. Shastri, Part III, Delhi: Nag Publishers. KauṣGS, The Kauṣītakagṛhyasūtra, with the Commentary of Bhavatrāta, Edited by T.R. Chintamani, New Delhi: Panini, 1982. Liṅgapurāṇa, The Liṅgamahāpurāṇam, Delhi: Nag Publishers, 1989. Matsyapurāṇa, Śrīmaddvaipāyanamunipraṇītaṃ Matsyapurāṇam, Ananda Ashrama Sanskrit Series, No. 54, Poona: Ananda Ashrama Press, 1981.



Manusmṛti, Manusmṛti with the Sanskrit Commentary Manvartha-Muktāvalī of Kullūkabhaṭṭa, Edited by J.L. Shastri, Delhi, Varanasi, Patna: Motilal Banarsidass, 1983. Mahābhāgavatapurāṇa: The Mahābhāgavata Purāṇa (An Ancient Treatise on Śakti Cult), Critically Edited with Introduction and Index by Pushpendra Kumar, Delhi: Eastern Book Linkers. Mbh, The Mahābhārata for the First Time Critically Edited by Vishnu S. Sukthankar and S.K. Belvalkar et al., 19 Vols., Poona: Bhandarkar Oriental Research Institute, 1933–1959. MānGS, Mānavagṛhyasūtra of the Maitrāyaṇīya Śākhā with the Commentary of Aṣṭāvakra Sastri, Edited by Ramakrishna Harshaji, New Delhi: Panini, 1982. MānŚS, The Mānava Śrautasūtra: Belonging to the Maitrāyaṇī Saṃhitā, Edited by Jeanette M. van Gelder, New Delhi, International Academy of Indian Culture, 1963. Nāradapurāṇa, The Nāradīyamahāpurāṇam, Delhi: Nag Publishers, 1984. Padmapurāṇa, The Padmamahāpurāṇam, 4 Vols., Delhi: Nag Publishers, 1984, 1985. PārGSPŚ: Pāraskaragṛhyasūtrapariśiṣṭa, Grihya-Sūtra by Paraskar with Five Commentaries of Karka Upādhyāya, Jayarāma, Harihara, Gadādhara and Viśvanātha, Edited by Mahadev Gangadhar Bakre, Bombay: Gujarat Printing Press, 1917, Reprint, New Delhi: Munshiram Manoharlal, 1982, pp. 404–548. Ṛgvidhāna, see Bhat 1987. ŚāṅkhGS, Śāṅkhāyana Gṛhya Sūtra (Belonging to the Rgveda), by S.R. Sehgal, Second Revised Edition, Delhi: Sri Satguru Publications, 1987. Somaśambhupaddhati, rituel dans la tradition śivaïte selon Somaśambhu/ texte, traduction et notes, Hélène Brunner-Lachaux, 4e pt., Pondichéry: Institut français d’Pondichéry, École française d’Extreme-Orient, 1998. Skandapurāṇa, The Skandamahāpurāṇam, 7 Vols., Delhi: Nag Publishers, 1986–1987. Vāmanapurāṇa, The Vāmana Purāṇa, Critically Edited by Anand Swarup Gupta, Varanasi: All-India Kashiraj Trust, 1967. Vāyupurāṇa, The Vāyumahāpurāṇam, 1983, Delhi: Nag Publishers. Viṣṇudharmottarapurāṇa, The Viṣṇudharmottarapurāṇam, 1985, Delhi: Nag Publishers. Viṣṇupurāṇa, The Viṣṇumahāpurāṇam, Delhi: Nag Publishers, 1985 Viṣṇu Smṛti, Viṣṇusmṛti with the Commentary Keśavavaijayantī of Nandapaṇḍita, Edited by Pandit V. Krishnamacharya, Madras: The Adyar Library and Research Centre, 1964. VaikhGS, Vaikhānasasmārtasūtram, the Domestic Rules of the Vaikhānasa School Belonging to the Black Yajurveda, Critically Edited by W. Caland, Calcutta: The Asiatic Society of Bengal, 1927.

The Planting Of Trees And The Dedication Of A Garden

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.



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.


The Sūrimantra and the Tantricization of Jain Image Consecration Ellen Gough1 For eight days in December of 2008, hundreds of Jains from both imageworshiping 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 conversation 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 symbols, the mantras used in the pratiṣṭhā ceremony explicitly call the Jina into the icon, and laypeople 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.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_012



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 themselves the full image consecration ceremony. To undertake this ceremony, the Digambaras invited a ritual specialist (pratiṣṭhācārya) from India to perform the pañcakalyāṇaka-pratiṣṭhā in which the icons to be consecrated are made to reenact the five (pañca) auspicious 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, examining 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 4  On the Kānjī Svāmī Panth, see Jain 1999: 101–117. 5  See Paul Dundas’s discussion of this issue in his article on Jain image consecration (Dundas 2009). The present study is indebted to this work as well as Dundas’s earlier study of the sūrimantra (Dundas 1998).

sūrimantra and tantricization of Jain image consecration


mendicant leaders—ācāryas—could perform pratiṣṭhā rites using the ritual formula, the sūrimantra, imparted to them upon their promotions to the highest rank of mendicancy (ācāryapada-sthāpanā). As we will see, the notion that only ācāryas have the authority to consecrate images seems to have developed amongst medieval tantric traditions, which I understand as cults of deities led by ritual adepts initiated (dīkṣita) into the cult by means of non-Vedic, esoteric mantras.6 Digambaras likely adopted this practice sometime in the medieval period, allowing only naked monks who had been imparted the sūrimantra to consecrate temple images. Therefore, because of this history, Kānjī Panthīs, despite rejecting mendicancy, unknowingly honor this older practice by requiring the ritual specialist to temporarily become, if only in appearance, a naked monk. To gain some insight into these pre-modern Digambara uses of the sūrimantra, much of this essay will rely on Śvetāmbara sources, since Śvetāmbaras have published dozens of medieval texts on the sūrimantra, whereas Digambaras have published limited materials, mostly dating from the last century. Contemporary Digambara ācāryas are not imparted a sūrimantra upon their ordinations,7 and the antiquity of the different Digambara sūrimantras used in image consecration today cannot be established. Śvetāmbara texts, on the other hand, confirm that Śvetāmbara ācāryas, for upwards of 900 years, have used the sūrimantra conferred to them at their ordinations to consecrate temple images. In these texts, the transfer of the sūrimantra in the promotion of an ācārya, in which the guru pronounces the mantra into the ritually decorated ear of the disciple, is replicated in image consecration. As we will see, these sources, some of which explicitly make reference to Digambara practices, could reveal some of the history of the Digambara sūrimantra. Digambaras and Śvetāmbaras seem to have influenced one another’s image consecration ceremonies, and members of both sects have debated, in a similar manner, the role of the ācārya in pratiṣṭhā ceremonies. Therefore, 6  I follow the concise explanation found in Goodall 2004: xxi: ‘The central fact that characterizes 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.



after focusing on Śvetāmbara practices, determining which texts were the most influential, this essay will return to the few available Digambara sources on the sūrimantra to suggest that in the medieval period, Digambaras and Śvetāmbaras both used an invocation termed ‘sūrimantra’ to promote ācāryas, and the contents of this mantra may have even been the same. 1

Early Śvetāmbara Mentions of the Sūrimantra

Examining the extant medieval Śvetāmbara literature on the promotion of a mendicant leader and the consecration of a temple image can give us some understanding of when the sūrimantra emerged as an important component of Jain ritual culture. According to Paul Dundas, the earliest known mention of a formula by the name of ‘sūrimantra’ may be found in Ācārya Haribhadrasūri’s Brahmasiddhāntasamuccaya (eighth century?), which describes the promotion of a sūri, another term for ācārya. In this ceremony, ‘a pupil, on the appropriate day and in a suitable state of purity, should be given the sūrimantra by his teacher along with sprinkling of the head with sandalwood powder (adhivāsana)’ (Dundas 1998: 36). To my knowledge, an explicit reference to the sūrimantra in the context of the pratiṣṭhā ceremony is not found until much later, perhaps not until Śāntisūri’s Prakrit text Ceiyavaṃdaṇamahābhāsa, which has been dated to the eleventh century.8 However, Śāntisūri only mentions that the pratiṣṭhā of jina icons is performed by the sūri with the sūrimantra; he provides no instructions as to how, exactly, the mantra should be used. The earliest complete account of the use of the sūrimantra in the pratiṣṭhā ceremony may be found in another Śvetāmbara text, the Nirvāṇakalikā. Among other topics, this Sanskrit treatise outlines mendicants’ daily rituals (nityakarman) and repentance rites (prāyaścittavidhi), the ceremony of renunciation (dīkṣā), the promotion of a mendicant head (ācāryaabhiṣeka), the iconographical features of the twenty-four tīrthaṅkaras and other gods, and the consecration rituals (pratiṣṭhāvidhi) for various objects such as a door (dvāra) and a temple icon (bimba). The terms ‘sūrimantra’ and ‘ācāryamantra,’ which designate the same formula, are mentioned several times in both the section on the promotion of a mendicant leader (ācārya-abhiṣeka) and in the section on temple image consecration (bimbapratiṣṭhā). In the description of the abhiṣeka of the ācārya, the key moment of transmission of the ācāryamantra from the guru to disciple 8  Ceiyavaṃdaṇamahābhāsa, v. 412, cited in Dundas 1998: 48n22.

sūrimantra and tantricization of Jain image consecration


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 assembly (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 (performing nyāsa) and reciting the ācāryamantra three, five, or seven times (Nirvāṇakalikā, pp. 23a–23b). The text also gives the option of reciting a different 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 enlightened 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 discussion 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 uttering 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 formula imparted to mendicants between the ranks of muni and ācārya. On this mantra, see Shah 1941. See also footnote 47 below.



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): oṃ namo arihaṃtāṇaṃ | oṃ namo siddhāṇaṃ | oṃ namo āyariyāṇaṃ | oṃ uvajjhāyāṇaṃ | oṃ namo loesavvasāhūṇaṃ | oṃ namo ohijiṇāṇaṃ | oṃ namo paramohijiṇāṇaṃ | oṃ namo savvohijiṇāṇaṃ | oṃ namo aṇantohijiṇāṇaṃ | oṃ namo kevalijiṇāṇaṃ | oṃ namo bhavatthakevalijiṇ āṇaṃ . . . (Nirvāṇakalikā, pp. 23a–23b) oṃ praise to the enlightened ones, oṃ praise to the liberated ones, oṃ praise to the mendicant leaders, oṃ praise to the mendicant teachers, oṃ praise to all the mendicants in the world, oṃ praise to praise the jinas who have clairvoyant knowledge (avadhi), oṃ praise to the jinas who have supreme clairvoyant knowledge, oṃ praise to the jinas who have complete clairvoyant knowledge, oṃ praise to the jinas who have infinite clairvoyant knowledge,14 oṃ praise to the omniscient ones, oṃ praise to the omniscient ones who are still living . . . This set of praises to Jain ascetics with certain special powers such as clairvoyance comprises the first section (pīṭha) of the different sūrimantras of competing 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 promotion 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 containing 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 14  Part of this translation is taken from Dundas 1998: 40.

sūrimantra and tantricization of Jain image consecration


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 commentary on the Pravacanasāroddhāra of Nemicandra’ (Sanderson 2011a: 3). 18  On the āgamas, a problematic category that now generally refers to a ‘canon’ of 45 scriptures, but has designated different groupings of texts throughout history, see Folkert 1993: 53–112.



Haribhadrasūri’s eighth-century(?)19 Pañcāśakaprakaraṇa.20 Crucially for this study, one of these quotes from a so-called ‘āgama’ mentions the sūrimantra, instructing the ācārya to recite it three times before covering the icon with a white cloth as part of the adhivāsana rite.21 While this Prakrit verse cannot be found in the extant āgamas or in works of Haribhadrasūri, it probably does come from an earlier source, as there does not seem to be any reason for the composer of the Nirvāṇakalikā to have invented it entirely. From this survey of the earliest known mentions of the sūrimantra, we can deduce that by the time of the eighth-century Haribhadrasūri, Śvetāmbara ācāryas transmitted to their disciples the sūrimantra to promote them to the ranks of ācārya, but we cannot be sure of the contents of this mantra or whether or not it was used in image consecration. Neither the Pañcāśakaprakaraṇa nor Haribhadrasūri’s other text on image consecration, the eighth chapter of Ṣoḍaśakaprakaraṇa, mention the sūrimantra. However, this does not necessarily mean that the sūrimantra was not used in image consecration at that time. It is reasonable to hypothesize that the requirement that an ācārya impart his sūrimantra to the image emerged between the eight and eleventh centuries.22 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




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: All these texts agree on three main prerogatives: dīkṣā, pratiṣṭhā, vyākhyāna: the ācārya (and he alone) will give initiations to those deserving, perform the installation of Śiva’s images for those who ask for it, and comment on the Scriptures (the Āgamas) (Brunner 1990: 15).

when he mentions in the Ceiyavaṃdaṇamahābhāsa that ‘there are many icons consecrated 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.



In the Jain context, at the end of the ācārya-abhiṣeka in the Nirvāṇakalikā, after promoting the disciple, the guru names these same three prerogatives: He announces that the new ācārya now may initiate new disciples, preach on the Jain scriptures, and perform pratiṣṭhā (Nirvāṇakalikā, p. 9a). This announcement not only aligns with contemporaneous Śaiva Saiddhāntika sources Brunner examined, it may well be a direct copy from the Śaiva text, the Siddhāntasārapaddhati.26 This is not to say that the Siddhāntasārapaddhati, per se, had any clear influence on Jain image consecration rites. As noted above, ācāryas were already consecrating images before the composition of the Nirvāṇakalikā, and even after its composition, the Nirvāṇakalikā may have been far less influential than scholarship has suggested. Since its first publication by M.B. Jhaverī in 1926, multiple editions of the Nirvāṇakalikā have been published,27 and it is often cited in secondary scholarship on Jain art and ritual,28 including the only two articles in English that have been devoted to Jain image consecration.29 However, when I asked the Śvetāmbara scholar-nun of the Kharatara Gaccha, Sādhvī Saumyaguṇā, to explain the rites in the Nirvāṇakalikā, she showed little interest in discussing the details of the text, claiming it is ‘influenced by Hindus’ and is ‘not in use today.’30 While the Nirvāṇakalikā remains a fascinating text that documents Śaiva influence on at least some Jains, it should not be seen as evidence that Jains ‘tantricized’ their image consecration ceremonies and other rites by copying Śaiva texts. Rather, Jains tantricized their ritual practices in uniquely Jain ways to correspond to the ritual norms of all medieval Indian traditions. 3

The Texts and Practices of Present-day Śvetāmbara Consecrations

A close examination of the practices of present-day Śvetāmbaras and the texts that influence them supports Sādhvī Saumyaguṇā’s claim. Manuscript collections, the contents of influential medieval texts, and present-day rituals 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 consecration. 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


suggest that the Nirvāṇakalikā played a marginal role in the development of Jain mendicant promotion and image consecration. Instead, texts by the twelfthcentury Ācārya Candrasūri, the fourteenth-century Ācārya Jinaprabhasūri, and the fifteenth-century Ācārya Vardhamānasūri should be seen as more foundational 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 ācāryapadasthāpanā 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 mendicant 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, composed in 1411–12 by the ācārya Vardhamānasūri. Manuscript collections suggest that these texts were more widely used than the Nirvāṇakalikā in the premodern 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.


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.

included in any other known Jain manuals on the promotion of an ācārya. Indeed, Jainism may have been the only lasting tantricized tradition to not adopt abhiṣeka as the means of ritual promotion of an ācārya, as Pāñcarātra, Buddhist, and Śaiva ācārya-abhiṣeka ceremonies are well documented.34 Instead, the Vidhimārgaprapā describes the transfer of the sūrimantra as follows: After the disciple circumambulates a model of the Jina’s preaching assembly (samavasaraṇa) and the guru three times, he sits on a newly made seat to the right of the guru (Vidhimārgaprapā, p. 67, lines 7–8).35 Then, at an auspicious time, the guru recites the ‘mantra of the lineage of the guru,’ here referring to the sūrimantra, three times into the disciple’s right ear that has been rubbed with sandalwood powder (Vidhimārgaprapā, p. 67, line 8.). Contemporary manuals agree on this key rite,36 and interviews with sitting ācāryas of multiple samudāyas confirmed that they were promoted in this way (Figure 1).37 Ācāradinakara also mentions the decoration of the ear with 34  For a recent study of abhiṣeka, see Shinohara 2014. 35  Citations from the Vidhimārgaprapā come from the edition edited by Vinayasāgara 2000. 36  Devendrasāgara 1943: 138–139; Khāntivijayagaṇi 1926: 141. 37  In Surat, Gujarat, on 23 October, 2013, I witnessed the ācāryapada-sthāpanā (there termed sūri/ācāryapad pradān) of two monks who would become Nayacandrasāgarasūri and Pūrṇacandrasāgarasūri. The ceremony closely followed the instructions given in

sūrimantra and tantricization of Jain image consecration


sandalwood powder and triple recitation of the sūrimantra (Ācāradinakara, p. 115a) as does a twelfth-century outline of ācāryapada-sthāpanā by Ācārya Candrasūri in his manual on ascetic conduct, Subodhasamācārī (Subodhasamācārī, p. 17b). Candrasūri’s text, rather than the Nirvāṇakalikā, which makes no mention of this rite, should therefore be seen as a predecessor to texts followed today like the Vidhimārgaprapā. For image consecration, as well, this twelfth-century Candrasūri seems to have been more influential than the credited author of the Nirvāṇakalikā, Pādaliptasūri. Today, Śvetāmbaras of all image-worshiping mendicant lineages follow an early seventeenth-century image consecration manual, the Pratiṣthākalpa, by Upādhyāya Sakalacandragaṇi.38 When crediting his sources for the text, Sakalacandragaṇi says nothing of Pādaliptasūri, but instead recognizes Candrasūri’s text on image consecration, the Pratiṣṭhāvidhi.39 Here, Sakalacandragaṇi follows in the footsteps of Jinaprabhasūri40 and Vardhamānasūri,41 who also name Candrasūri as one of the sources for their pratiṣṭhāvidhis and do not give credit to Pādaliptasūri.42 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 preserved 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.



Sakalacandragaṇi’s manual differs considerably from these earlier Śvetāmbara sources on pratiṣṭhā, however. Unlike the earlier sources mentioned above, the text organizes the consecration around the reenactment of the five auspicious events in the Jina’s life (pañcakalyāṇaka). As we will see, Digambara manuals have structured the ceremony around these events since at least the twelfth century, so Śvetāmbaras may have adopted this practice from Digambaras in order to make image consecration not about obscure rites common with other traditions, but about the life of the Jina. Contemporary consecration rites also include a variety of rituals that originally developed separately from pratiṣṭhā, such as the worship of particular ritual diagrams (maṇḍala). Today, the most popular edition of Sakalacadragaṇi’s text has been edited, with Gujarati commentary, by Ācārya Somacandrasūri. This edition prescribes the ten days of worship that precede the day when the icon is established in the temple as follows: Day One: Ritual procession with pots of water ( jala-yātra), establishment of the pots (kumbha-sthāpanā) and a lamp (dīpaka), raising of the banner ( jvārāropaṇa), establishment of the kṣetrapālas, erection of the pillar (sthambha) and arch (toraṇa) Day Two: Worship of the nandyāvarta-maṇḍala Day Three: Worship of the guardians of the ten directions (daśa-dikpāla), the protector deity bhairava, the sixteen “spell goddesses” (vidyādevī),43 the nine planets (navagraha), and the eight auspicious symbols (aṣṭamaṅgala)44 Day Four: Worship of the siddhacakra maṇḍala Day Five: Worship of the bīsasthānaka maṇḍala Day Six: Reenactment of the descent of the Jina into the mother’s womb (cyavana-kalyāṇaka) Day Seven: Reenactment of the birth of the Jina ( janma-kalyāṇaka) Day Eight: Eighteen different ritual ablutions (abhiṣeka) and the celebration of the birth (janma-kalyāṇaka) and ritual naming of the Jina Day Nine: Reenactment of the education, wedding, royal consecration, and renunciation (dīkṣā-kalyāṇaka) of the Jina Day Ten: Adhivāsana, eye-opening (añjanaśalākā), and the reenactment of the Jina’s enlightenment (kevalajñāna-kalyāṇaka) and liberation (nirvāṇa-kalyāṇaka)

43  On the Jain vidyādevīs, see Shah 1947: 114–77. 44  On the Jain aṣṭamaṅgala, see Wiley 2009: 45.

sūrimantra and tantricization of Jain image consecration


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: The monk first purifies himself through meditation. He ritually ensures that the image is in a spiritually stable condition, so that the consecration will be successful. He uses a golden needle to brush a paste of powdered gems onto the outline of the eyes of the image. The monk looks at the image in a mirror while reciting a mantra (spell) calling on the Jina’s omniscience to illuminate the cosmos. He says a special mantra into the

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.



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 consecration 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 ceremonies, we can understand this text as a largely insignificant, if interesting, medieval document. 4

Śvetāmbara Controversies

Jain ceremonies of image consecration and the promotion of an ācārya also differ from their Śaiva equivalents because Jains, unlike members of other tantric traditions, did not develop separate ordinations for mendicants and 46   The Gujarati commentary to Pratiṣṭhāpāṭha, p. 276 prescribes sandalwood power (sukhaḍ) instead of sugar. 47   Pratiṣṭhāpāṭha, pp. 272–273 prescribes the recitation of the vardhamānavidyā into the right ear. However, this instruction relates to monks of a lower rank than ācārya. Conversations with ācāryas confirm that they recite the sūrimantra. See Vidhimārgaprapā, p. 102, lines 17–18, which instructs a regular monk (sāmānyayati) to recite the vardhamānavidyā, here termed the ‘pratiṣṭhāmantra.’ 48  See the Gujarati commentary to Pratiṣṭhāpāṭha, p. 276. 49   Ācāradinkara, p. 184a, prescribes the guru to pronounce the ācāryamantra not thrice, but seven times after decorating the ear with sandalwood powder and unhusked rice (akṣata). The guru should then use his right hand to touch the icon while performing the cakramudrā and reciting the pratiṣṭhāmantra, i.e. the vardhamānavidyā.

sūrimantra and tantricization of Jain image consecration


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 mendicant; 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 person who had received a tantric ordination by worshiping a maṇḍala and receiving 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 composed 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 Rejection of monastic participation in image consecration was also passed to members of the A(ñ)cala Gaccha, a lineage that emerged from the Paurṇamīyaka Gaccha in the mid-twelfth century.51 These divergent rites between members of competing mendicant lineages, along with slightly different sūrimantras used by each gaccha, may have been a means by which Śvetāmbaras of different groups could lay claim to certain temples. In the twelfth century, for example, an ācārya of the Kharatara Gaccha, Jinadattasūri, emphasized that temple images consecrated by laypeople were not worthy of worship.52 Paul Dundas (2007: 238n103) has also documented the concerns recorded in the seventeenth-century Senapraśna over whether or 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.



not members of the Tapā Gaccha should worship icons consecrated by members 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 participate 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. 5

Digambara Sūrimantras and Image Consecration

In the last few decades, the Digambara community has contested the role of the ācārya in pratiṣṭhā ceremonies in a manner remarkably similar to medieval 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 consecration 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 mendi53  Ācārya Kalāpaprabhasūri, interview by author, Mumbai, August 15, 2013. 54  According to a ritual specialist who has performed image consecration ceremonies for members of all gacchas, ācāryas of the other extant mendicant lineage that emerged from the Paurṇamīyaka Gaccha, the Āgamika (Tristuti) Gaccha, today participate in image consecration. Yaśvant Golecchā, interview by author, Jaipur, 16 June 2013.

sūrimantra and tantricization of Jain image consecration


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 distinctions 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 different 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 published, 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 manuals 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.’



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 recitation of the sūrimantra, do not necessarily follow a single text, and instead combine 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 medieval 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


Day One: Hoisting of the banner (dhvajārohaṇa) Worship in the temple, ācārya’s command to perform the ceremony, congregation’s invitation of the ācārya to participate in the ceremony, procession of auspicious pots and icons to be consecrated from the temple to the maṇḍapa representing Ṛṣabha’s birthplace, Ayodhya, purification (śudhi) of the maṇḍapa, ablution (abhiṣeka) of the mūrti established on the platform (vedikā) in the maṇḍapa, establishment of the auspicious pots on the platform, hoisting of the banner at the entrance to the maṇḍapa, lighting of a lamp (dīpa-prajjvalana), evening lamp offering (āratī), lecture by the monks Day Two: The Jina’s Conception, Part I (garbha-kalyāṇaka-pūrvārddha) Daily abhiṣeka, temple worship, procession of auspicious pots (ghaṭayātra), cleansing of the temple with herbs (dhāmapraukṣaṇa), ritual purification of the ritual actors (sakalīkaraṇa), worship of the yāgamaṇḍala,62 evening lamp offering (āratī), lecture by monk (guru-pravacana), reenactment of the events during the pregnancy of Ṛṣabha’s mother such as her sixteen dreams63 Day Three: The Jina’s Conception, Part II Daily abhiṣeka, fire offering (havanaśānti), celebration of the anniversary of the renunciation (dīkṣā) of the two monks (muni) in Ācārya Vibhavasāgara’s saṃgha, recitation of the Bhaktāmara Stotra, lecture on the scripture Iṣṭopadeśa,64 worship of the yāgamaṇḍala, reenactment of the descent of the Jina into the mother’s womb, evening āratī, lecture from the monks, reenactment of the events of the final stages of Ṛṣabha’s mother’s pregnancy such as the interpretation of the dreams Day Four: The Jina’s Birth (janma-kalyāṇaka) Daily abhiṣeka, garbha-kalyāṇaka-pūjā, fire offering, birth of the Jina, the gods’ ablution (abhiṣeka) of the Jina on Mount Meru, entry of Ṛṣabha into the palace decorated with all the auspicious adornments (śriṅgāra), decoration of all the icons 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 tradition, see Wiley 2009: 47. 64  For an English translation of this seventh-century philosophical treatise, see Modi 2010.



to be consecrated with beautiful clothes, lamp offering (āratī), worship of the monks, monks’ lecture on the scriptures, dance of Indra, reenactment of the childhood games of Ṛṣabha Day Five: The Jina’s Renunciation (dīkṣā-kalyāṇaka) Daily abhiṣeka, janma-kalyāṇaka-pūjā, recitation, vāstuśāntipūjā and fire offering in the four directions, reenactment of Ṛṣabha’s marriage and his establishment of human civilization (introduction of modern warfare, writing, agriculture, etc.), Ṛṣabha’s life in the palace, his renunciation (dīkṣā), evening lamp offering, worship of the monks, lecture on the scriptures and evening drama reenacting Ṛṣabha’s sons’ war over control of the kingdom65 Day Six: The Jina’s Enlightenment (jñāna-kalyāṇaka) Morning abhiṣeka, dīkṣā-kalyāṇaka-pūjā, lecture by the monk and ritual feeding (āhāra) of Ṛṣabha, adhivāsanā, performance of nyāsa on the icons, mouth-opening rite (mukhoddhāṭaṇa), eye-opening rite (netronmīlana), *recitation of the sūrimantra, establishment of the icon in the model of the Jina’s preaching assembly (samavasaraṇa), evening lamp offering, worship of the monks, lecture on the scriptures, cultural performance Day Seven: The Jina’s Liberation (mokṣa-kalyāṇaka) Morning abhiṣeka, reenactment of Ṛṣabha’s death and attainment of liberation (mokṣa) on Mt. Kailāsa, the gods’ transportation of the hair and nails of Ṛṣabha to heaven, viśvaśānti fire offering, ritual procession of the icons from the maṇḍapa to the temple, establishment of the icons in the temple, establishment of the auspicious pot (kalaśa) and the banner atop the temple, dismissal (visarjana) of the deities invoked in the ceremony66 During these seven days, ritual specialists (pratiṣṭhācārya) and laypeople performed 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 description of a Digambara image consecration ceremony in Gujarat in 1974, see Jain and Fischer 1978: 4–14.

sūrimantra and tantricization of Jain image consecration


ceremony that Digambara monks undertake.67 The second time occurred on the sixth day, when the icon gained enlightenment (jñāna-kalyāṇaka). Like contemporary Śvetāmbaras, Digambaras perform the adhivāsana, eye-opening, and transfer of the sūrimantra back-to-back as part of the jñānakalyāṇaka ceremony. In the ceremony I witnessed, the monks did not perform these rites in the middle of the night like the Śvetāmbaras, but in the afternoon, out of the view of the lay congregation, on the platform behind a curtain.68 After the adhivāsana rite, ritual specialists and clothed, lower-level members of Ācārya Vibhavasāgara’s saṅgha (kṣullaka, brahamcāriṇī) were allowed behind the curtain as the mendicants performed the mouth-opening (mukhoddhāṭaṇa) rite, in which they touched the seeds of seven different grains to the mouths of the pratimās. The monks then recited ‘Svastyayana,’ a blessing containing Sanskrit verses invoking each of the twenty-four tīrthaṅkaras,69 and performed the eye-opening rite, in which they recited a particular mantra 27 times and used golden sticks to apply sugar, milk, and camphor to the eyes of the icons. After these performances, the naked monks were left alone behind the curtain to pronounce the sūrimantra 108 times while showing the icons a mirror.70 After this secret recitation, the main icon used to reenact Ṛṣabha’s biography was revealed from behind the curtain with great pomp, musicians blaring hymns over the loudspeakers, hundreds of audience members cheering, and five lamps glowing in front of the main image to symbolize the attainment of omniscience, or all five types of knowledge.71

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 sacraments, 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 possessions 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 consecration 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.



This performance of jñāna-kalyāṇaka was clearly modeled on Jayasena’s Pratiṣṭhāpāṭha, as the earlier texts make no mention of the sūrimantra, only agreeing that the eye-opening should be performed after the adhivāsana.72 Jayasena, on the other hand, not only prescribes the recitation of the sūrimantra after the eye-opening,73 but he also emphasizes its importance. He requires an ācārya to pronounce it74 and stresses that while the other ritual acts of the pañcakalyāṇaka-pratiṣṭhā are simply for increasing the devotion (bhakti) of practitioners, the recitation of the sūrimantra, along with the dīkṣā-kalyāṇaka, the adhivāsana, prāṇapratiṣṭhā, and eye-opening rites, are the main (mūla) ritual acts that transform the icon into the essence of the Jina (jinatva).75 Indeed, for both Śvetāmbaras and Digambaras today, the recitation of the sūrimantra represents the point at which the inert stone, metal, etc. of the image becomes, in some way, a physical manifestation of the Jina.76 However, unlike Śvetāmbaras today, who all require monks to recite the sūrimantra and essentially agree on the mantra’s general structure and contents,77 Digambaras 72  After the adhivāsana, Pratiṣṭhāpāṭha, vv. 120–121, only outlines the mouth and eyeopening 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 outlined in this paper.

sūrimantra and tantricization of Jain image consecration


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 wellknown 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 ceremony 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 different 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.



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 transformed 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 appendix 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 liberal about the recitation of the mantra. They are the only saṅgha, to my knowledge, 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


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.

pañcakalyāṇaka-pratiṣṭhā in Hastinapur in January 2013, a celibate lay disciple of the Jñānamatī, Ravindrakīrti Svāmī, dressed in orange robes and in plain sight, recited the mantra into the ears of the icons not at the time of the Jina’s enlightenment, but on the final day of the ceremony, during the reenactment of the Jina’s liberation (Figure 2). Here, Ravindrakīrti Svāmī essentially acted as a bhaṭṭāraka, though bhaṭṭārakas in South India do not officially recognize him as one. Bīsapanthīs do not necessarily use the sūrimantra provided in Jñānamatī’s edition, and instead employ the mantra that has been passed down from guru to guru. Notebooks of ritual specialists and Digambara monks are filled with sūrimantras they have collected from their teachers and gurus. According to a few laypeople and monks I interviewed, bhaṭṭārakas today use a sūrimantra that is basically a combination of the Vedic gāyatrī-mantra,83 the Prakrit 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).



pañcanamaskāra-mantra, and the second half of the formula given by Śāstrī and Jñānamatī.84 I cannot confirm these claims, however, as in my experience, these clerics keep their sūrimantras a well-guarded secret. One bhaṭṭāraka did confirm to me that he uses four different types of sūrimantras, and the one used to consecrate metal ritual diagrams (yantra) does resemble the gāyatrīmantra, but he would not give me the specific contents of the formula. This secrecy over the sūrimantra is common. According to one prominent ritual specialist in Mumbai, the exposure of the sūrimantra to non-experts has only occurred within the last couple of decades. Thirty years ago, he only received the mantra from his teacher, another ritual specialist, after years of training and prodding. Amongst some practitioners, transmission of the mantra became a bit more liberal at the end of the twentieth century, after the stillactive Ācārya Kunthusāgara published his versions of the formula in a popular compendium of tantric texts, the Laghuvidyānuvād. In this text, Kunthusāgara enjoins a ritual specialist, not a monk, to recite the mantra, and instead of ‘sūrimantra,’ or ‘mantra of the mendicant head,’ Kunthusāgara names the invocation ‘sūryamantra,’ or ‘mantra of the sun (sūrya).’ This name is not unique to this text, as the Kānjī Svāmī ritual specialist I interviewed also claimed that the mantra is sometimes called the ‘sūryamantra.’ Unlike Śvetāmbaras, Digambaras today do not often use the term ‘sūri’ for ācāryas, so this name may have evolved from mispronunciation due to unfamiliarity with the term ‘sūri.’ Because Kunthusāgara, like Kānjī Panthīs, does not consider this mantra to be the domain of the ācārya, or sūri, alone, the importance of the term ‘sūri’ in the name of the mantra has been lost. Kunthusāgara provides four different sūryamantras: One for an icon of a jina, one for all the protector deities of the jinas (śāsanadeva), one for the Padmāvatī, the protector goddess (yakṣī) of the twenty-third tīrthaṅkara Pārśvanātha, and another for Pārśvanātha’s protector god (yakṣa), Dharaṇendra (Kunthusāgara and Vijayamatī n.d.: 236–237). For an image of a jina, he outlines a sūryamantra of three parts. First, the pratiṣṭhācārya should repeat, 21 times into the ear of the icon, one short Sanskrit mantra that consists of a series of seed syllables and a praise of the arhats.85 Then, placing a mirror in front of the icon, he 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


should read another short Sanskrit mantra five times. This mantra, which contains 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 different interpretations of how it should be imparted. Some Digambaras do not require the recitation of the sūrimantra, others instruct a clothed ritual specialist 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 vardhayantu dharmācāryā huṃ ruṃ kuru kuru svāhā svādhā (Kunthusāgara and Vijayamatī n.d.: 237).



In response to such concerns, in 2000–2001 in the Hindi monthly Jain journal Sanmati Sandeś, Prakāś Hitaiṣī Śāstrī (n.d.), a ritual specialist associated with the Kānjī Svāmī Panth, published an article, ‘What is the Sūrimantra (sūrimantra kyā hai?),’ which argues that texts and historical practices dictate that a pratiṣṭhācārya, not an ācārya, should give the sūrimantra to temple images.89 When ritual texts use the term ‘ācārya’ in their prescriptions of pratiṣṭhā, Śāstrī contends, they refer not to a renunciant, but to a ritual specialist (Śāstrī n.d.: 13). The most persuasive of his many arguments in favor of this claim may be his interpretation of a line of ungrammatical Sanskrit in the only pre-modern text to mention the sūrimantra, Jayasena’s Pratiṣṭhāpāṭha. While the grammar of the sentence is not correct, Śāstrī insists that it instructs the ācārya to ‘become naked’ (digambaratvagata) after the adhivāsana rite and before the recitation of the ‘Svastyayana’ and the other rites required for the jñāṇa-kalyāṇaka.90 If the term ‘ācārya’ referred to a naked muni, Śāstri reasons, then there would be no need to require him to become naked (Śāstrī n.d.: 27). The term ‘ācārya’ here thus must designate the ritual specialist, a ‘pratiṣṭhācārya,’ literally, a ‘master’ (ācārya) of consecration (pratiṣṭhā). Aware that his manual on pratiṣṭhā was being interpreted in this way by Kānjī Panthīs and other Digambaras, the composer of the pratiṣṭhā text followed by Kānjī Panthīs, Nāthūlāl Jain Śāstrī, addressed this issue of nudity in an article entitled, ‘The Importance of the Digambara Sūrimantra in the Consecration of a Jina Image (jinbimb pratiṣṭhā meṃ sūrimantra kā mahatva).’ In this piece, Śāstrī is adamant that only a faultless Digambara muni who has undertaken dīkṣā can give the sūrimantra; a layperson who simply becomes naked does not become a muni (Śāstrī 2012: 192).91 The key to his argument lies in his translation of the term ‘avagata’ in the portion of Jayasena’s text addressed above. He claims that the Sanskrit instruction in Pratiṣṭhāpāṭha that Prakāś Śastrī claims describes the ācārya as ‘having gone (avagata) to the state of nakedness (digambaratva)’ really should be translated as describing the ācārya as ‘having the appearance (avagata) of nudity’ (Śāstrī 2012: 192). This adjective merely describes the permanent state of the ācārya. According

89  I thank Pawan Jain for directing me to this article. 90  The line, with my suggested corrections in brackets, reads: adhivāsanā[ṃ] niṣṭhāpya sarvān janān apasṛtya digambaratvāvagata[ḥ] ācāryaḥ . . . svastyayanaṃ paṭhet | (Pratiṣṭāpāṭha, p. 282). ‘Having completed the adhivāsanā and dismissed all the people, the ācārya, having become naked, should . . . recite the svastyayana benediction.’ 91  I thank John Cort for directing me to this article.

sūrimantra and tantricization of Jain image consecration


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 specialists. 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? 7

Digambara Sūrimantras and the Promotion of an Ācārya

While, to my knowledge, present-day bhaṭṭārakas do not receive a sūrimantra upon their promotions, texts and inscriptions suggest that the imparting of the sūrimantra, the promotion to the rank of a bhaṭṭāraka, and the ability to consecrate images were intimately related in the early modern period, if not earlier. Nāthūrām Premī (1912) has outlined the contents of a manuscript in Idar, Gujarat, that prescribes the ordination of a bhaṭṭāraka. This ritual manual describes how the bhaṭṭāraka-to-be should remove his clothes, pull out his hair (perform keśa-luñcana), and receive the sūrimantra. A similar document was published in 2002, in a manual entitled ‘Various Rituals for the Initiation Rites of Passage,’ Vividh Dīkṣā Saṃskār Vidhi. This manual contains, in Sanskrit, the prescriptions for the rites of initiation for a regular Digambara monk (muni), a mendicant teacher (upādhyāya), a mendicant leader (ācārya) and a pontiff (bhaṭṭāraka) that were said to have been copied from an ‘ancient’ (prācīna) manuscript a monk from the first half of the twentieth century, Muni Sanmatisāgara, found in a manuscript house in the pilgrimage site of Śrī Atiśay



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 (bhaṭṭārakapadasthāpanā-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, receiving 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 example, in the early seventeenth century, a disciple of Bhaṭṭāraka Kumudacandra, Gaṇeśa, composed a hymn praising his teacher (gurustuti) that reads: saṃvat sol chapanne vaiśākhe prakaṭ padovar āpyā re | ratnakīrtti gor bāraḍolī var sūr maṃtra śubh āpyā re | bhāī re man mohan munivar sarasvatī gacch sohaṃt | kumudacandra bhaṭṭārak udayo bhaviyaṇ man mohaṃt re ||95 In Saṃvat 1656 (1599 C.E.), in the month Vaiśākha (April-May), [he] ascended the throne (pada) of the bhaṭṭāraka. He received the supreme, auspicious sur[i]mantra [from] the monk (gor) Ratnakīrtti in Bāraḍolī (Gujarat). 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


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 regular 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 illustrates 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 further 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 manuscript 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 practices, 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.



However, Jayasena’s prescription that the ācārya must ‘become digambara’ before the jñāna-kalyāṇaka ceremony does suggest that these clothed clerics were honoring older practices of naked ācāryas. Inscriptions from at least as early as the eleventh century describe Digambara mendicant heads consecrating icons in South India (Vijaymūrti 1952: 281–300, inscription # 213),97 and in his twelfth-century Pratiṣṭhāsāroddhāra, Āśādhara enjoins mendicant heads to perform the rites of the dikṣā-kalyāṇaka in the image consecration ceremony.98 He does not, however, make any mention of the ācārya’s role in the jñāna-kalyāṇaka, when the sūrimantra today is recited. Simply because ācāryas participated in image consecrations does not mean they used the sūrimantra. But there are some hints that medieval naked Digambara monks used a sūrimantra identical to that of Śvetāmbaras. The Śvetāmbara Jinaprabhasūri’s fourteenth-century manual on the sūrimantra documents a portion of what he terms the sūrimantra of the Digambara tradition (āmnāya) (Dundas 1998: 36). This ‘Digambara sūrimantra,’ resembling the first section (pīṭha) of its Śvetāmbara equivalent (translated above), includes eight different praises to spiritually advanced ascetics who have achieved certain special powers (Sūrimantrabṛhatkalpavivaraṇa, pp. 81–82). While I have not found any Digambara text to name this set of praises the ‘sūrimantra,’ Digambaras from as early as the first half of the first millennium did employ a similar, if longer, list of praises to ascetics with superhuman powers as a ritual invocation99 and by at least the beginning of the eleventh century, forty-eight such praises began to be inscribed on ritual diagrams (valaya).100 Interestingly, this diagram, termed the ‘gaṇadharavalaya,’ or the ‘ring of the disciples of the tīrthaṅkaras (gaṇadhara),’ made its way into medieval Digambara pratiṣṭhā manuals. These manuals describe how different ritual 97  On bhaṭṭārakas’ participation in pratiṣṭhā, see also Parmār 2010. 98   atheṃdrāḥ . . . jinaniṣḳramaṇakalyāṇakriyāṃ kuryuḥ sasūrayaḥ || (Āśādhara, Pratiṣṭhāsāroddhāra, 4.112). ‘Then, the indras (ritual specialists) should perform the dīkṣā kalyāṇaka along with the mendicant heads (sūri).’ 99  These praises make up the auspicious opening (maṅgala) to the fourth chapter of the Ṣaṭkhaṇḍāgama. See Wiley 2012 and Gough 2015. 100  Amitagati’s text on lay ritual conduct, Śrāvākācāra, vv. 46–48, composed in the first quarter of the eleventh century (see Williams 1963: 23), records the contents of a circular diagram (valaya) with two intersecting triangles at its center surrounded by the praises to special powers that make up the Digambara gaṇadharavalaya and the first pīṭha of the Śvetāmbara sūrimantra.

sūrimantra and tantricization of Jain image consecration


diagrams are to be worshiped in the days leading up to the establishment of the temple image. Each type of image to be consecrated (a goddess, a siddha, etc.) is associated with a different ritual diagram, and as seen above in the outline of the Digambara consecration ceremony, for the consecration of an image of a jina, the worship of the yāgamaṇḍala is required. For the consecration of an image of an ācārya, however, Āśādhara in the early thirteenth century requires the worship of the gaṇadharavalaya (Pratiṣṭhāsāroddhāra, pp. 230b–232b), clearly associating this mantra with mendicants of the highest rank.101 The exact same diagram is then described as a visual representation (yantra) of the sūrimantra in the earliest available Śvetāmbara manual devoted to the use of the sūrimantra, Siṃhatilakasūri’s thirteenth-century Mantrarājarahasya (Mantrarājarahasya, v. 69).102 It seems that medieval Digambaras and Śvetāmbaras both associated the same diagram with the consecration of an ācārya, either in stone or in flesh. Another interesting document—a manuscript of a single folio—sits in the manuscript collection of the Śrī Digambar Jain Candraprabhujī Mandir in Jaipur. This folio clearly belonged to a larger document at some point. One side of the folio contains a discussion of savaiyā, a type of meter of pre-modern Hindi poetry (McGregor 1984: 89n203). The other side begins in the middle of a discussion in Sanskrit of what it calls the sūrimantra, outlined as irimeru girimeru pirimeru sirimeru hirimeru āirimeru. Meru, the manuscript declares, means “arhatship,” or embodying enlightenment. Iri, it continues, provides the pleasures of this world (saṃsāra); it is the wish-fulfilling tree (kalpadruma), guaranteeing that one will not be blind or deaf and bringing about the success of worldly and otherworldly desires. Giri means kingship (rājya) and the ability to make others prosper, piri stands for renunciation (dīkṣā), siri means enlightenment (kaivalya) and omniscience (sarvajñatā), hiri means one commands great respect (mahāpūjā) and is never obstructed throughout the three worlds, and āiri stands for the mendicant community (saṅgha) and liberation (mokṣa). The manuscript summarizes how the six syllables of this mantra represent the path to liberation: The first two syllables (iri and giri) stand for pleasures in this world, piri represents abandoning these pleasures and renouncing, and

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.



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 manuscript 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

By tracing the historical development of one component of the 2008 pratiṣṭhā ceremony in Arizona—the recitation of the sūrimantra—this essay has shed some light on the tantricization of Jain image consecration practices and the interactions between Śvetāmbaras and Digambaras as they developed their image consecration and mendicant promotion rites. My examination of Śvetāmbara sources revealed that Ācārya Candrasūri’s twelfth-century text on pratiṣṭhā seems to have had a more significant influence on Śvetāmbara practices than a text of roughly the same period, the Nirvāṇakalikā, which displays idiosyncratic components likely copied from a Śaiva text, the Siddhāntasārapaddhati. Unlike the Nirvāṇakalikā, Candrasūri’s text on the promotion of a monk to the rank of ācārya prescribes that the guru decorate his disciple’s right ear with sandalwood powder in order to impart the sūrimantra. This exact act is performed in image consecration ceremonies to this day, making Jain pratiṣṭhā practices, as I see it, ‘tantric’ in that they require initiates of 103  Śrī Digambar Jain Candraprabhujī Mandir, Aṅkroṃ kā Rāstā, Kiśanpol Bāzār, Jaipur, manuscript 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


the highest rank—ācāryas—to use a special mantra in a ritual that replicates their own ordination ceremonies.105 As the secondary sources surveyed above have shown, the requirement that ācāryas perform key components of pratiṣṭhā seems to have developed in tantric traditions in the medieval period. However, the Jain practice of whispering the mantra into the ear of the disciple to be promoted and into the icon to be consecrated does not align directly with the practices of other tantric traditions. Firstly, Jains, unlike members of other tantric traditions, combined tantric and monastic ordinations and required that the ācārya be a mendicant. This has caused controversies amongst both Śvetāmbaras and Digambaras over role of the monk in image consecration ceremonies, since Jain monks are not supposed to perform image worship. In addition, while Śaivas and other tantric sects stressed the importance of abhiṣeka for the promotion of an ācārya, Jains have not adopted abhiṣeka as a means of promotion. Indeed, since the Nirvāṇakalikā prescribes that the ācārya be promoted not by decorating his ear with sandalwood powder and reciting the sūrimantra, but instead through abhiṣeka, if this text were our only available source on medieval Śvetāmbara image consecration and the promotion of an ācārya, then there would be little reason to suggest that the present-day Digambara whispering of the sūrimantra into the ear of a mūrti is modeled on the medieval promotion of an ācārya. Thankfully, we have other Śvetāmbara medieval sources, and it seems that Digambaras and Śvetāmbaras not only both modeled pratiṣṭhā on ācāryapada-sthāpanā, but that they developed their presentday image consecration ceremonies in conversation with one another. Sources available today suggest that the key moment of image consecration emerged as a combination of Śvetāmbara and Digambara practices. Today, both Śvetāmbaras and Digambaras recite the sūrimantra into the ear of the temple image on the day or night preceding the establishment of the image in the temple. They do so in conjunction with the eye-opening ceremony performed during the reenactment of the fourth auspicious event in the life of the Jina, enlightenment (jñāna-kalyāṇaka). It is possible that this contemporary jñāna-kalyāṇaka ceremony is a composite of the originally Śvetāmbara practice of reciting the sūrimantra and the originally Digambara practice of understanding the image consecration ceremony as the replication of the five auspicious events in the life of a Jina. 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.



Digambara sources do not mention a sūrimantra until perhaps the fourteenthfifteenth centuries, the time of the composition of Jayasena’s Pratiṣṭhāpāṭha, while Śvetāmbara sources as early as the Nirvāṇakalikā and Ācārya Śāntisūri’s eleventh-century Ceiyavaṃdaṇamahābhāsa require the use of the sūrimantra in pratiṣṭhā. On the other hand, all available Digambara sources, from Vasunandi (twelfth century) onwards, organize the events of the ceremony around the five events in the life of a Jina, while Śvetāmbaras do not seem to have adopted this practice until the composition of the text used today, Sakalacandragaṇi’s early seventeenth-century Pratiṣṭhākalpa. This hypothesis likely simplifies the historical development of the ceremony, however. With the hope that scholars will unearth other sources related this topic, including medieval Digambara manuals on ācāryapada-sthāpanā, I look forward to further research to solve some of the mysteries presented here. References Ācāradinakaraḥ of Vardhamānasūri. Prathamavibhāgaḥ & Dvitīyavibhāgaḥ. Bombay: Pandit Kesarisinha Oswal Khamgamwala, 1922–1923. Balbir, Nalini. 1994. ‘An Investigation of Textual Sources on the samavasaraṇa (The Holy Assembly of the Jina).’ In Festschrift Klaus Bruhn, ed. Nalini Balbir and J.K. Bautze. Reinbek: Verlag für Orientalische Fachpublikationen, 67–104. Bentor, Yael. 1996. Consecration of Images and Stūpas in Indo-Tibetan Tantric Buddhism. Leiden, New York, Köln: E.J. Brill. Bhattacharyya, A.K. 2010. Historical Development of Jaina Iconography. Delhi: Bharatiya Kala Prakashan. Brunner, Hélène. 1990. ‘Ātmārthapūjā versus Parārthapūjā in the Śaiva tradition.’ In The Sanskrit Tradition and Tantrism, ed. Teun Goudriaan. Leiden, New York, København, and Köln: E.J. Brill, 4–23 Carcarī of Jinadattasūri. 1927. In Apabhraṃśakāvyatrayī: Three Apabhraṁśa Works of Jinadattasūri with Commentaries. Ed. Lalchandra Bhagawandas Gandhi. Baroda: Oriental Institute, 1–27 Carrithers, Michael. 1990. ‘Jainism and Buddhism as Enduring Historical Streams.’ Journal of Anthropological Society of Oxford 21/2, 141–63. Ceiyavaṃdaṇamahābhāsaṃ of Śāntisūri. Mumbai: Śrī Jinśāsan Ārādhanā Ṭrasṭ, 1986. Cort, John E. 2005. ‘Devotional Culture in Jainism: Manatunga and His Bhaktamara Stotra.’ In Incompatible Visions: South Asian Religions in History and Culture. Essays in Honor of David M. Knipe, ed. James Blumenthal. Madison: University of Wisconsin Center for South Asia, 93–115.

sūrimantra and tantricization of Jain image consecration


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 AntiPaurṇamīyaka 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.



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


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.’ (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.



Pratiṣṭhākalpa (Añjanaśalākā) of Sakalacandragaṇi. 1985. Edited by Muni Somacandravijaya Gaṇin. Surat: Śrī Nemacaṃd Melāpcaṃd Jhaverī Jain Vāḍī Upāśraya Ṭrasṭ. Pratiṣṭhāpāṭha Saṭīka of Jayasena. 1925. Edited with Hindi commentary by Hīrācand Nemacand Dośī. Sholapur: Śeṭh Hīrācand Nemacand Dośī. Pratiṣṭhāsāroddhāra of Pandit Āśādhara. 1917. Translated into Hindi and edited by Pandit Manoharlāl Śāstrī. Bombay: Jaingranth Uddhārak Kāryālay. Pratiṣṭhātilaka of Nemicandra. Translated into Hindi and edited by Jñānamatī Mātā. Hastinapur: Digambar Jain Trilok Śodh Saṃsthān, 2006. Premī, Nāthūrām. 1912. ‘Bhaṭṭāraka.’ Jaina Hitaiṣī 7 [cited from a reprint of unspecified selections, no date]. Summarized in Peter Flügel, ‘Demographic Trends in Jaina Monasticism.’ In Studies in Jain History and Culture, ed. Peter Flügel. London: Routledge, 2006, 312–398 Rastelli, Marion. 2003. ‘Maṇḍalas and Yantras in the Pāñcarātra Tradition.’ In Maṇḍalas and Yantras in the Hindu Traditions, ed. Gudrun Bühnemann. Leiden: E.J. Brill, 119–52. Śāstrī, Prakāś Hitaiṣī. n.d. Sūrimantra Kyā hai? Edited by Rākeś Jain. Nāgpur: n.p. Compiled from Dec. 2000, Jan. 2001, and Feb. 2001 editions of Sanmati Sandeś. Śāstrī, Nāthūlāl Jain. 2012. ‘Jinbimb Pratiṣṭhā meṃ Sūrimantra kā Mahatva.’ In Paṇḍit Nāthūlāl Jain Śāstrī Smṛti Granth, eds Anupamā Vikās Jain, Devkumār Jain, Arvind Kumār Jain and Jinendra Kumār Jain. Indore: Dr. Ajitakumārsingh Kāslīvāl Amar Granthālay, 191–194. Śāstrī, Nāthūlāl Jain. 1998. Pratiṣṭhā Pradīp: Śrī Digambar Jain Pratiṣṭhā Sambandhī Vidhi Vidhān. 2nd ed. Indore: Vīr Nirvāṇ Granth Prakāśan Samiti Kāryalay. Sanderson, Alexis. 2011. ‘The Appropriation of Śaiva Sources and Models in the Production of Jain Ritual Paddhatis from the 10th to the 15th Century.’ Lecture Handout, Abteilung für Kultur und Geschichte Indiens und Tibets, Asien-AfrikaInstitut, University of Hamburg, 4 July. Sanderson, Alexis. 2011a. ‘The Śaiva Age: Further Evidence of the Buddhist and Jain Adaptation of Śaiva Models and Sources in Early Medieval India.’ Lecture Handout, Columbia University, New York, 19 September. Sanderson, Alexis. 2015. ‘The Jaina Appropriation and Adaptation of Śaiva Ritual: The Case of Pādliptasūri’s Nirvāṇakalikā.’ Lecture Handout, 15th Annual Jaina Lecture, SOAS, 19 March. Saumyaguṇā, Sādhvī, ed. and trans. into Hindi. 2005. Vidhimārgaprapā. Mumbai: Śrī Mahāvīr Svāmī Jain Derāsar Ṭrasṭ. Saumyaguṇā, Sādhvī. 2006. Jain Vidhi-Vidhān Sambandhī Sāhitya kā Bṛhad Itihās. Bhāg 1. Edited by Sāgarmal Jain. Shajapur, Madhya Pradesh: Prācya Vidyāpīṭh.

sūrimantra and tantricization of Jain image consecration


Shah, Umakant P. 1941. ‘Varddhamāna-Vidyā-Paṭa.’ Journal of the Indian Society of Oriental Arts IX, 42–51. Shah, Umakant P. 1947. ‘Iconography of the Sixteen Jaina Mahavidyas,’ Journal of the Indian Society of Oriental Art 15, 114–77. Shah, Umakant P. 1987. Jaina-Rūpa-Maṇḍana ( Jaina Iconography). New Delhi: Abhinav Publications. Shinohara, Koichi. 2014. Spells, Images, and Maṇḍalas: Tracing the Evolution of Esoteric Buddhist Rituals. New York: Columbia University Press. Shinohara, Koichi. 2014b. ‘The Esoteric Buddhist Ritual of Image Installation.’ In Buddhist and Jaina Studies: Proceedings of the Conference in Lumbini, eds J. Soni, M. Pahlke and C. Cüppers. Bhairahawa, Nepal: Lumbini International Research Institute, 265–298. Smith, H. Daniel. 1984. ‘Pratiṣṭhā.’ In Agama and Silpa: Proceedings of the Seminar held in December, 1981, ed. K.K.A. Venkatachari, 50–68. Bombay: Anantacharya Indological Research Institute. Somasundaravijaya, ed. 2007–2008. Śrī Sūrimantra Mahāpūjan evaṃ Vardhamān Vidyā Mahāpūjan. n.p. Śrāvākācāra of Ācārya Amitagati. In Śrāvākācāra-Saṃgraha, Prathama Bhāga, edited and translated into Hindi by Hīrālāl Śāstrī, 263–421. Solapur: Jain Saṃskṛti Saṃrakṣak Saṃgh, 2001. Śrīsubodhāsamācārī of Candrācārya. Edited by Shaha Jivanchand Sakerchand Javeri. Bombay: Sheth Devchand Lalbhai Jain Pustakoddhar Fund, 1924. Syādvādamatī Mātā, Āryikā, comp. n.d. Vimal Bhakti Saṃgrah. Bhāratavarṣaya Anekānt Vidvat Pariṣad. 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, 115–42. Tiwari, Maruti Nandan, Kamal Giri, and Harhar Singh (eds.) 2010. Encyclopedia of Jaina Studies: Volume I, Jaina Art & Architecture. Varanasi: Parshwanath Vidyapeeth. Velankar, Hari Damodar. 1944. Jinaratnakośa: An Alphabetical Register of Jain Works and Authors, Volume I, Works. Poona: Bhandarkar Oriental Research Institute. Vidhimargaprapā of Jinaprabhasūri. Edited by Mahopādhyāya Vinayasāgara. Jaipur: Prākṛt Bhāratī Akādamī, 2000. Vijaymūrti, Pandit, comp. 1952. Jainaśilālekhasaṃgraha. Dvitīyo Bhāgaḥ. Bombay: Māṇikacandra Digambara Jainagranthamālāsamitiḥ. Wiley, Kristi L. 2009. The A to Z of Jainism. Lanham, Toronto and Plymouth, UK: Scarecrow Press, 2009.



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.


Āgamic Temple Consecration Transnationalized Annette Wilke 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 context, 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 fieldwork 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 HammUentrop 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 developments, 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 1  This is presently changing with the second generation. 2  From now on I will mostly use the short-form ‘Kamadchi Temple.’ I refrain from using diacritic marks for the temple names, but instead use the transcription by which the temples represent themselves in Germany. 3  The first book on Sri Lankan Tamils in German speaking and Scandinavian countries (Baumann, Luchesi and Wilke 2003) appeared shortly after consecration and contained already a description of the memorable event (Luchesi 2003: 223–271, in particular 237–244), which I want to take up and extend. I will add some new perspectives and material, and also discuss more recent developments in the extremely dynamic and fluid field of Tamil Hindu diaspora temple culture.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_013



what exchanges took place with the new homeland culture? What local, national and transnational dynamics did the consecration trigger off—regarding 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 intentional. The consecration of the Kamadchi Temple marked the new visibility of a largely invisible migrant religion. The advent of ‘ethnic Hinduism’ is a novelty 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 profound 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 construction 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 transliteration ‘Maha Kumbhabhisekam’ was used to announce the event.

Āgamic Temple Consecration Transnationalized


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 erection 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 transnational 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. 1

ho was Able to Achieve this Little Wonder?—Notes on the W Wider Context

The presence of ethnic Hinduism in Germany (and continental Europe) is a very recent development. Until the mid-1990s we find new religious movements like the ISKCON, that erected some temples since the 1970s, and about 36.000 Indian Hindus, mostly professionals, who live in various parts of Germany and who founded cultural associations, but have no temple culture. This situation changed due to forced migration. The late 1980s and 1990s brought about 60.000 refugees from Sri Lanka to Germany, about 45.000 of them Hindus, and some 7000 Hindu migrants from Afghanistan.5 Both groups have been very active in the reconstruction and institutionalisation of their religious life. Within two decades the Afghani Hindus erected 7 temples and the Sri Lankan Hindus around 30. At present at least 40 can be counted, with a concentration 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.



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—nowadays 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 popular 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 previous ‘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 happily 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 imprisonment or being sent back to war-torn Sri Lanka.

Āgamic Temple Consecration Transnationalized


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 married 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 undoubtedly 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 privately sponsored. Sri Paskaran was the first who started public processions at the yearly temple 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) outside 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 personnel, 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



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 humbly in 1989 in a basement in the city area of Hamm. Already in 1993–94, however, 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 participants 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 transfigurations 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, supporting 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 possible 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


quite a bit of leisure time and honorary work, including active help and coordination in the yearly temple festivals and guided tours for German groups visiting 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 advisory 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 cooperation of priest and architect. The architect was found by the priest in an oracle-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 structure, 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. 2

he Consecration Rites (2002) of the Sri Kamadchi Ampal Temple T e.V. Europe

On July 7, 2002, the eight days of consecration reached their peak and public 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



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 symbolically 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 industrial 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 infusion 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, notwithstanding the importance given to mahākumbhābhiṣeka on the eighth day. This day was advertised as major one. It witnessed the biggest crowd and largest presence of press and television—who in fact had a major share in creating 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 mantrasādhana, daily pūjā and japa, and selflessness (Takashima 2005: 132, 134). 9  Historically, prāṇapratiṣṭhā does not seem to be as central. Takashima (2005: 116–125) mentions 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


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 LalitāSahasranāma, 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 foundation 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. 2.1 Preliminary Rites Underlying the ceremony in May 1999 was the ancient Indian rite for temple and house construction on the construction ground, in which a sacred geometrical diagram is foundational (vāstumaṇḍala).10 The measurements of the vāstumaṇḍala (27×27 metres, corresponding to the planned temple size) had been calculated by the South Indian Sthāpati M.G. Nagaraj, who also fashioned the new temple images. The rite was performed by Sri Paskaran, assisted by his brother-in-law and several priests from abroad. There was a crowd of some 200 persons present. After purifying the construction ground by the veneration of the earth (bhūmipūjā), vāstupūjā was performed aiming at sanctifying the ground and connecting the world of men with the world of the gods and the divine guardians of the quarters. The square-shaped vāstumaṇḍala with horizontal and 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.



vertical lines (harbouring the deities to be worshipped on the construction ground) was drawn on the ground on which the temple was going to be built, and the deities of the vāstumaṇḍala were worshipped, starting with Īśana in the north-east corner. Thereafter, the rite of the laying of the foundation stone was performed.11 For purification two cows were led to the spot where today Kamadchi’s central sanctum (garbhagṛha, ‘womb chamber’) is located. Right at the place where the goddess’ image was going to be installed, a pit was dug and stabilized. Into this vacuum were inserted a number of auspicious and symbolically loaded items while mantras were being recited: a copper-plate with the date and name of the temple founder (priest Sri Paskaran) and his assistants and sponsors, earth from seven eminent Indian temples, sand from three oceans, shells filled with milk and holy Ganga water, nine precious stones (navaratna) and five metals (pañcaloha), and finally also 3000 small copper plates engraved with the (ritual diagram) śrīcakra. The yantras had been purchased by lay devotees for 200 German Mark (or some smaller or larger optional amount) some time before May 1999, in order to have their names engraved on the backside of the yantra. These yantras, notably their large (symbolically charged) number, and the esoteric śrīcakrapūjā (āvaraṇapūjā) performed by Sri Paskaran, were obviously an addition to traditional elements. It filled priest Paskaran with special pride to have the only temple in the world with 3000 magically powerful yantras as its base. An even more obvious ritual ‘invention of tradition’ (Hobsbawm and Ranger 1984) was the topping-out ceremony in following year, which took up a German local custom, probably to please the German craftsmen. This foreign and profane custom was ‘Hinduised,’ and again the śrīcakra played a pivotal role. The construction work had started in spring 2000 and quickly developed, so that priest Paskaran and the German architect invited for the topping-out ceremony already in August 2000. The invitation mentioned explicitly the German name ‘Richtfest’ and attracted besides Hindu-Tamil guests local dignitaries from the churches and city administration who spoke some greeting words. All German and Indian workmen were present. The latter had recently arrived to build the interior shrines, and the richly sculptured surface of the gopura and vimāna towers. The assembled crowd was smaller 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 foundation 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


at this occasion (around 50 persons). They witnessed a garland being erected to crown the top of the roof (‘Richtkranz’), i.e. the gopura, and elements of a newly invented ritual. Sri Paskaran performed on the gopura a fire sacrifice (homa), and had a new set of donated copper śrīcakras fixed with mortar on the temple tower (where in 2002 the kalaśas would be placed). In a speech the priest explained the śrīcakra as major symbol of the goddess and its auspicious power. So, the consecration rites in 2002 were preceded by a pair of foundational sanctifying rituals in which both cultures were merged performatively by ritual actions. 2.2 Āgamic Consecration Globalized: Tradition and Change Dynamics of change and merger like the ones described above are typical of rituals when transferred to other cultures. Although Sri Lanka Tamils try to reconstruct their home-tradition as faithfully as possible, transformations can occur more easily in diaspora situations. Rituals may not only get lost, but also witness emendations and extensions, and even deliberate inventions. The rituals described above seem typical to me in terms of the situational weaving together traditional ritual forms and newly created additions and adaptions to the new environment. Individuals like the zealous chief priest and Śrīvidyā adept Sri Paskaran Kurukkal may take on new roles as ritual inventors. It is at the same time traditional imagination which is adhered to and finds new globalized outlets. In the background—personally present during the consecration rites in 2002 and 2014—was Swami Brahmananda, a Śrīvidyā guru from the Bhuvanesvara-Pitha, Tamilnadu, who now lives in the US. A further novelty brought about by diaspora contexts and the modern ‘deterritorialization’ of cultures are new transnational networks of ritual specialists and professionals in temple building touring around the world. Three of the Indian artisans (śilpins), who were present at the topping-out ceremony, had already been in Houston, Texas, for building a South Indian style temple there (Luchesi 2003: 233). Their ‘master’ Sthāpati Nagaraj, who had calculated the vāstumaṇḍala, came in 2002 personally to Hamm-Uentrop for the eye-opening ceremony. He was also responsible for the first Swiss temple in the style of South Indian architecture, which was inaugurated 2013 in Trimbach (Baumann & Tunger-Zanetti 2014: 29). Similarly, Dr. N. Somashkandak Kurukukkal, an eminent head priest and professional in temple building from Canada who is as well member of the council of Hindu priests in Australia, was present in Hamm-Uentrop and Trimbach to guide and supervise the consecration ceremonies. He was present again in Hamm-Uentrop during re-consecration in 2014.



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 challenge 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 remarkable 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: They unanimously prescribe the determination of the correct time, selection of the personnel for the ceremony, erection of a special shed 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 rituals of post-twelfth century. It is well known that early texts do not yet include temple consecration. 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 ‘synopsis’ 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


(maṇḍapa [also called yāgaśāla]) and the necessary rituals such as ablution of the image, the fire-offering to be performed therein . . . [After ablution] 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 sanctorum 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 certain 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 ceremony (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 goddess 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 ceremonial 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.



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 diaspora-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. 2.3 The Eight Days of Consecration ( June 30–July 7, 2002) My reconstruction only aims at describing major outlines, not a full and detailed description—which would go beyond both my capacities and the limits of this article.14 The intention is to give a glimpse of the big-style ceremony and draw attention to the fact that the performance of consecration was at the same time a ritualization of the social bonds with the new society. It continued the endeavour to integrate the German surroundings and people, apparent already during the phase of construction (sign board ‘Hier entsteht ein Hindutempel’) and the preliminary rites of consecration (foundation stone and topping-out ceremonies). The following summary illustrates that part of the worship during the eight days of consecration constituted at the same time rituals of interaction and exchange. Sometimes, different rituals would take place simultaneously (e.g. during the first day). All the eleven priests from Sri Lanka, respectively India, and also Sthāpati Nagaraj from Chennai, were present from the first day onwards, each of them having his own specific ritual function, e.g. at one of the fire-altars in the ritual tent. The first and the succeeding days were devoted to extensive rituals of protection and purification, from the fifth and sixth day onwards the rituals became more dense and focused on divinization. 2.3.1 First Day, June 30, 2002 (a Sunday) During the first day the temple grounds were purified and sanctified in a complex ceremony that lasted several hours: Sri Prabhudevan drew the outlines of the vāstupuruṣa on the ground near the Western entrance of the temple. An altar was established on the ‘body’ of the vāstupuruṣa on which eight metal 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


pots sealed with coconuts and mango leaves, indicating the presiding deities of the directions, were placed. A long ceremony followed, including extensive 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. 2.3.2 Second to Fourth Day, July 1–3 (Monday to Wednesday) Ceremonies of protection and purification continued inside the new empty temple. On Monday, the divinities of the directions were invoked and worshiped, on Tuesday the nine planets (navagraha) were offered worship, and on Wednesday a sequence of purification rites was performed. This included bringing a cow (the epitome of purity) into the empty sanctum sanctorum which was going to be the permanent abode of Goddess Kamadchi. The beautiful brown cow had been selected by priest Paskaran. The local farmer, however, led it through the dense crowd into the empty ‘womb chamber.’ He stayed on in the following days, as the cow was to be constantly present in yajñaśāla to assure the purity and auspiciousness of this place of fire sacrifices. 2.3.3 Fifth Day, July 4 (Thursday) This was the day of the eye-opening ceremony, the goddess’ awakening to the world and her bestowal with sight.15 An eminent role was played by the Śilpin Nagaraj who had fashioned the (over human-size) Kamadchi image of black granite. He slit the contours of the pupil of the eye, the eye-lids 15  Cf. Hikita 2005: 191–194. The ceremony is known by several names. The following description 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.



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 accessible 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 programmebrochure 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 neighbourhood 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 preceptor 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. 2.3.3 Sixth Day, July 5 (Friday) From this day onwards, the programme got more and more dense. In the morning, after a solemn procession around the temple, the sacred copper vessels (kalaśas) were fixed on the temple roof, i.e. gopura tower, by the priests and the Sthāpati’s younger brother). Each kalaśa was worshipped with a pūjā. Thereafter, each of the new shrines in the temple-interior was crowned with a kalaśa. Meanwhile, the preparations for establishing the special shed, i.e. sacrificial tent, next to the temple for fire sacrifices—the yajñaśāla—had come to an end. Ten altars had been erected on which ten water pots (kumbhas)

Āgamic Temple Consecration Transnationalized

Figure 1


Fire rituals in the yajñaśāla performed in 2002 and 2014. photo May 1, 2014, copyright A. Wilke.

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 corresponded 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 temple 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 foundation stone at each of the pīṭhas by depositing the deity’s sanctified yantra and



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. 2.3.3 Seventh Day, July 6 (Saturday) This day was a very special event for lay devotees. They got the rare chance to enter the shrines and anoint the divine images with sesame oil [Fig. 2]. Thereafter, it would be the sole privilege of priests to enter the shrines and get into bodily contact with the deity. Accordingly, there was a huge crowd throughout the day and long queues in front of the shrines to get the opportunity to give the deities an oil massage und have thus a very personal and embodied contact with them. Meanwhile, again extensive fire sacrifices with mantra utterance took place in the neighbouring tent, i.e. the yajñaśāla, to invoke divine power and presence in the water jars so that the deity’s power would be drawn into them. Later in the evening a long cord made of holy Kuśa grass was spanned from Kamadchi’s central kalaśa (pot, jar) and fire altar to her anthropomorphic image in the sanctum of the temple. Side strands of the cord led to the water jars, fire pits and temple images of the other divinities. The priests’ task was now to transfer the divine energy which had accumulated in the yajñaśāla to the anthropomorphic images in the shrines to make them divine icons. They did so by solemnly walking thrice to and fro from the sacrificial tent to the temple, following the course of the cord and constantly touching it with their ritual sticks and sacrificial spoons [Fig. 3]. When they arrived at the images they emptied the spoon with ghee on the head of the icon and touched its body parts in nyāsa-like fashion with the ritual stick. Only very few lay worshipers were left—mostly sponsors of the shrines—to witness this sacramental transubstantiation after midnight until around 1.30 a.m.

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 addition 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

Figure 2

Oil bath by lay devotees performed in 2002 and 2014. photo May 3, 2014, copyright A. Wilke.



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


It was already two or three o’clock in the morning, when the magic evening 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 worshipers 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 supposed to perform it only once in his lifetime. 2.3.4 Eighth Day, July 7, 2002 (Sunday) This was the great final day of the official consecration rites, the mahākumbhābhiṣeka, or ‘great ablution with the jars,’ when the empowered water in the jars was poured on the temple (the copper kalaśas on the roof), the shrines and the images. Several thousand people, some of them from abroad, many local dignitaries and much press had come to Hamm-Uentrop for this magnificent event. So, a considerably larger crowd than the previous days (a certain exception was only the oil-bath the previous day) heard the loud and happy sounds of the temple music and observed how the jars were taken out from the yajñaśāla by the priests, sponsors, and guests of honour. The procession of pot-bearers circumambulated the temple clockwise, led by the musicians, before climbing up the scaffold to the temple roof, the vimāna and gopura towers where the kalaśas were poured out and pūjās performed. A special highlight and reason for common rejoice and loud jubilation was when at the end, flower petals and drops of holy water were sprinkled in all directions and on the crowd below. The same procedure was repeated inside the densely packed temple. First, the various priests climbed each a different shrine and showered sacred water from additional holy jars on the kalaśas on the top shrines, before performing the ablution of the image inside the shrine. Like all other rites during the previous days, everything was perfectly concerted among the priests, so that all the ten shrines and deities received almost simultaneously the sacred bath. Thereafter, the shrines were closed while the crowd was waiting for their


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


re-opening and the ‘vision’ (darśan) of the beautifully decorated and richly dressed deities, most splendid of all Kamadchi Devi. A long solemn pūjā ceremony with 64 offerings (upacāra) and the waving of many lamps was following. Besides these traditional religious actions, performed by the priestly specialists and in part also the Sri Lankan Tamil sponsors, some of the festive incidents 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 consecration (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 previous 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 relation 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 leaflet 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 ŚaivaSiddhānta 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



of success—and although a deeper second glance and thick description will reveal also some draw backs and difficulties which make the development less splendid than it superficially looks, one cannot doubt, that the rapid rise of the Kamadchi temple to a new pilgrimage place of transnational importance and international fame in the past decade, was a great success and is an amazing achievement. The temple became also an important contact zone on the national and local level, in which positive encounters between migrants and members of the host society developed. My hypothesis is that the constitution of a ‘real’ temple, i.e. a building which looked like a temple, and a traditional ceremony which confirmed it, were immensely important for this development. Today, the Kamadchi temple is a good example of religious ‘glocalization.’ Local, national and transnational levels intersect. 3.1 Developments at the Local Level Already in February 19, 2002, the FAZ, one of the biggest German newspapers, reported that Hamm was something like the capital of exile Hinduism in Germany. The reports about the inauguration of the Kamadchi temple in the industrial area of Hamm-Uentrop were enhancing this claim. An interesting development was that one year after the consecration in July 2002, the city of Hamm lanced a flyer, inviting for a pilgrimage to Hamm. The red and white flyer showing the temple tower, illustrates the big changes triggered off by the consecration: good contacts with the German locals and the rapid rise of a new Tamil Hindu pilgrimage place of transnational impact. A visitor to Hamm stepping into the tourist office in 2003, would have found the new colourful flyer inviting in English language for pilgrimage to the nearby Kamadchi Ampal temple. Whereas the temple had been set up by Tamil Hindus, the advertising flyer was issued by the township shortly thereafter. Hamm is a middle size German city on the border of the Ruhr area, Westphalia, and unremarkable except for a relatively high percentage of different migrant populations who constitute 22% percent of the citizens. The native German citizens have not always been as well-disposed towards their local Hindu community as the flyer suggests. But a lot has changed since the erection of a building in the style of classical South Indian temple architecture in the industrial area of Hamm-Uentrop in July 2002. For Tamil Hindus a new pilgrimage place was born and for Hamm a new tourist attraction. The more or less pale city at the border of the heavily industrial Ruhr area gained an exotic and international flair. Road signs help foreigners to find the ‘Hindu temple’ and recently, the nearby bus stop was re-named to include the reference ‘Hindu temple.’ Nowadays, the township values the Kamadchi temple as an important ‘upholder of Hindu culture’ across the nation. Interreligious dialogue meanwhile is one of the most cherished projects of the city council of Hamm, and

Āgamic Temple Consecration Transnationalized


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 gestures 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 deliberately 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 important 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



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. 3.2 National Level With the imposing sacral structure a ‘real temple’ originated which was of distinctly different quality than other Hindu places of worship in Germany, a temple which looked exactly like one at home in Sri Lanka and South India. This is how many Tamil migrants felt, but equally important was the radiation into the majority society whose members became aware that there was a place like this by the media attention around the consecration. It was only then that they realised that Tamil refuges had brought along their traditional Hinduism to Germany. One should remember the specialness of the temple building in 2002. The Kamadchi temple itself had initially been a small basement shrine in the backroom of a bowling-alley. When it grew larger—still in the city of Hamm—it had its own conflict history with German citizens, which started when the ‘alien’ Hindu religion went public by processions. Despite being forced to ‘withdraw’ to an industrial no-man’s land, it achieved public recognition. Not only locally, but nationwide it became something like the official representative of Hinduism. The sacral structure created a new contact zone profile whose own dynamism and social effervescence called processes into being that were not yet imaginable in 2002. To my assessment the sacral architecture was eminently important for both Tamil Hindus and German natives as it created a shared space of imagination and a material reference point for both the socio-cultural groups. While for the migrants, it was an authentic “home away from home” (Ballard 1994), it was ‘India text door’ for members of the majority society. A young Tamil woman, who had already been raised in Germany and was living in Frankfurt, stated: ‘If we go there [to the Kamadchi temple], it is for us like a pilgrimage of India,’ and a German native said he had never been to India, and did not feel necessity anymore, as: ‘India has come to me.’17 For many migrants only a traditional architecture and Āgama consecration 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 statement (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


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 traditional 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 reputation 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 attribution 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 religious 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.’



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 certain 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 approximation 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; moreover, 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 architecture, 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 temple 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, witnessed 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


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 fundraising 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. 3.3 Transnational Level It is noteworthy that the original temple name already contained a proud ‘Europe e.V.’—which was initially nothing more than an imagination and wish of the founding priest Sri Paskaran, but in fact gained reality after 2002, as indeed Sri Lankan Tamil Hindus from all over Europe participate in the yearly temple festival of Hamm-Uentrop. Even more proud is the label ‘international’ of the aspired cultural centre. Let us see what the future will bring, it may be less brilliant and great as hoped for, but there can be no doubt that at present the temple stands out within the German temple-scape for its internationalization. The temple acquired the repute of a new European Hindu pilgrimage place. Its fame spread far beyond the borders of Germany—even to India, the US and Canada. The great chariot procession (tēr) at the temple festival attracts not only Sri Lankan Tamils from all over Germany, but also from Denmark, Holland, France, Italy, Switzerland, and Great Britain. It also increasingly attracted German locals, some of whom travel for hours to get there. From 2001 to 2015 the temple festival, i.e. the chariot procession, attracted (with notable exceptions in 2009, 2012–13) an increasing crowd of 10,000 to 20,000 participants from Germany and neighbouring countries. The largest number of participants was so far witnessed in 2008: Local newspapers recorded 25.000 participants and an article appeared even in one of the two major Swiss newspapers, the Tagesanzeiger of Zurich. Since 2007, the sponsor of the Tēr day was a Sri Lankan Tamil Hindu from England, respectively, a British Tamil Hindu family who sponsored a major part of the day (the religious part and the movable toilets). This transnationalizing tendency largely augmented since 2002, when the newly built ‘real temple’ was inaugurated. Around a month after consecration, there was a particularly spectacular Tēr festival with a large number of vow fulfillers, including hook-swinging. This first hook-swinger in Germany was British who sought to attain the right of residence in Britain with his vow.



He indeed succeeded, which is why he repeated his vow of self-affliction in 2003 to thank the goddess. In this year there was even a second hook-swinger from Germany. However, his swinging beam broke, and the violent practice was no more allowed in the years to come. Vows that continued were kāvaṭi dance (including smaller hooks) performed by men and the bearing of pots with burning camphor or milk by (mostly young) women. Rolling (men) or constantly prostrating (women) on the processional route was simplified, i.e. restricted to circumambulating the temple. For many years, the ritual specialists who apply the hooks and rent the kāvaṭis came from France. Priests assisting and conducting the procession arrive from Sri Lanka and India. At the consecration, the Giri Trading agency from Mylapore (Chennai) was present for the first time with a large stall of pūjā items and audio material. Not only the eleven priests, but also this Brahmin trader from South India stayed on for another 45 days, and continued to come to the temple festivals every year. Moreover, a festival market selling goods from home grew ever larger each year. It is a ‘little India’ with jewellery, garments, and foods (particularly delicious mangos coming directly from Sri Lanka) and heavily frequented by Sri Lankans and Germans alike. Special mention deserves the fact that the festival attracts a great number of Tamil youth who enjoy meeting each other every year across the local and national borders. In interviews they report that they look forward to this day the whole year through. The festival is an absolute ‘must go’ for the youngsters—mostly not for religious reasons, but for socializing, shopping (at the festival market), looking for potential marriage partners, searching political exchange, strengthening cultural ties and identity formation, but also showing German friends what their culture is about. Many of the young generation perceive the big event as something very special, they feel that their culture is unique, something to be proud of and the girls love wearing beautiful and expensive traditional dresses and jewellery. The traditional temple architecture, the big-style yearly festivals, the good press (even of violent rites), the newly created networks through the big events, and a zealous priest have been favourable for transnationalizing processes. The temple and priest gained international fame, spreading to the US and far-off Canada, where many Sri Lankan Tamils migrated. Sri Paskaran was invited to a temple consecration in Memphis, Tennessee, and restored an old Vaiṣṇava temple in South India, which was out of use, because it stood under the curse that anybody attempting the re-establishment would die. This did of course not happen to Sri Paskaran—one of his strategies to contribute to his fame as special and saintly person.

Āgamic Temple Consecration Transnationalized


3.4 Some Shadows The story of the Kamadchi temple is not only one of success. There have been impediments and difficulties all along, starting with financial straits. The temple festivals have been the most important resources, so that decline of participants is a serious threat. This happened in 2009 where only 7000 visitors were counted (Westfälischer Anzeiger Hamm 8.6.2009)—probably due to the heavy new outburst of violence in Sri Lanka: Many migrant families lost relatives and did not visit temples in 2009 and 2010 because of ritual pollution. The number augmented again to 18.000 in 2011 (Westfälischer Anzeiger Hamm 14.6.2011), but in 2012 and 2013 there was a new substantial decrease because of unfortunate weather conditions: cold and heavy rains. Although the temple festival in July 2014 brought again hot weather and increasing numbers of vow-fulfillers,19 the crowd on the whole did not reach the figures of previous successful years,20 and the peak-number of 2008 (25.000 participants according to local press) remained a singular event. Although big festival events belong to the major financial resources, expenditure is also very high. I was informed by a board member that the temple festival in 2014 barely covered all the costs. Presently, the Kamadchi temple seems to be in a serious financial crisis—after it had been nearly without debts a few years after consecration. Unfortunately, the security police superimposed more and more costly injunctions, which the temple is hardly able to fulfil. A financial and personal traumatic disaster for the whole priestly family happened in November 2013. The priest’s house and temple office fell a victim to a violent robbery. The raid (performed by masked men) was never elucidated. It left not only fear and terror in the priest’s children, but also a deep hole in the temple’s donation box (containing large sums donated for the cultural centre). Moreover, the goddess was robbed of her golden jewellery and the house and temple doors were destroyed with an axe, etc. It is said that the damage amounted to 40.000–50.000 Euro. One can easily imagine that the re-consecration in May 2014 in particular included also very high costs (starting with the air-fair of fourteen priests from South Asia) and seriously added to emptying the treasure boxes. Another 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 ceremony 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 probably therefore abstained from visiting again the temple festival.



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 success 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 subjects. 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 additional 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 sponsoring, 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



Re-Consecration 2014

Priest Paskaran (who worked as an electrician before acting as full-time priest) informed me that according to Indian custom, it is necessary to re-charge the temple’s energy and power every 12 years, as the mantra-power is gone or has faded after this period. As all sorts of people—‘good’ and ‘bad’ ones—can come to the temple, ‘bad people’ (like the robbers in 2013) have taken away the holy energy of prayers. The energy charge of the battery therefore needed recharge and new prāyaścitta. An Internet article (28.03.2014) and video around the re-consecration mentions more practical problems: The harsh German winters had heavily destroyed the sculptures on the vimāna and gopura, so that they badly needed repair and better protection. The idea to paint the towers with variegated bright colours was explained by the German architect not with the South Asian custom (at least in modern times) or the wish for beautification (which at the first consecration could not be fulfilled due to shortage of money), but with the necessity of better protection. It was communicated to the public that repair was only allowed each 12 years. In short: The traditional time-span of re-consecration after 12 years was kept intact because the temple needed both material and spiritual renewal. For the former, again five Indian artisans (śilpins) from Mr. Nagaraj’s team were present to renovate and paint the temple towers—unfortunately, however, the painting was not completed due to visa problems. Due to denial of a longer visa, the artisans arrived much later than scheduled and had only a good month of time instead of half a year, so the painting could only partially be completed. The major concentration was laid on the vimāna (above the garbhagṛha) where energy is supposed to be the strongest. It was a strenuous race with time, as the ritual part was going to start on 27th April 2014. Sri Paskaran made clear that only Indian artisans were able to do the job, due to their knowledge about the looks of the gods and their upholding purity rules (strict vegetarian diet, no eggs, no alcohol, etc.) which the priest considered as absolutely necessary. Moreover, he considered it necessary to hold a special pūjā when the artisans arrived, to ask permission of the goddess before they started their work. It is like the immigration office, he said, we need labour permission from her. For the ritual part, again a great number of priests arrived from South Asia, so that altogether 16 priests actively performed re-consecration—the mental and ritual purification and re-charge of sacred power. This included 12 days of śrīcakrapūjā and a full ‘maṇḍala’ of 48 days of rituals, prayers, and mantra-repetition to purify the pīṭha and spread ‘good vibrations’ in the temple, as Sri Paskaran explained. Whereas these extra prāyaścitta and so-called maṇḍala-abhiṣeka rituals extended until July 6, 2014, the Tēr day of the temple



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 sacredness 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-consecration 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 temple 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 thunderstorm that caused the backside of the temple stand in water, so that circumambulation 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


days will be given preference. The announcement on the board marked them also as major days. It mentions only: 27.04.2014 start of the consecration/Kumbhabhisekam Ceremony 02.05.2014 Yaga-Pooja (consecration of the copper peaks [kalaśa vessels]) 03.05.2014 bath of all deity images with sesame oil by the (lay) devotees25 04.05.2014 major ceremony of consecration: Maha Kumbhabhisekam Ceremony 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 worshiping 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 pedestal 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) sponsors 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.



the installation of the foundation stone of the pīṭha was being renewed with nine kinds of jewels, medicinal substances, and seeds.—The major fire sacrifice 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. 4.1 Sixth Day, Friday, May 2, 2014 This day (which I did not personally observe) mirrored the first consecration in its focus on the kalāśas (on the top of the roof and towers). They were renewed in the morning. Before and after it long pūjās and mantra recitation took place in the yajñaśāla. I was informed that in the afternoon of the very same day began already the rite which is most popular and cherished among the laypeople: the intimate ‘washing’ or anointing the icons with sesame oil. It gives lay-persons the opportunity to get active themselves and be closer than ever to the divinities, enter the shrines otherwise reserved to the priests, and have very direct body contact with the deities. 4.2 Seventh Day, Saturday, May 3. 2014 This was the major day of the ‘oil-bath.’ It witnessed a large crowd of devotees touching, anointing, rubbing in, massaging and caressing the images with large amounts of oil—from early morning to around 6 p.m. when the major activities changed to the yajñaśāla. While I noticed particularly long and tender rubbing in the oil on Kamadchi’s black granite body [Fig. 2], I also noticed equally long queues at the shrine of Murugan, the old Tamil god who is much favoured by many Sri Lankan Hindu devotees. In the ritual hut with the fire-sacrifices, ritual actions took place as described above (nyāsa on the central pot, dhyāna, mārtṛkā homa, etc., and the sponsors taking active roles at the altars were the pots of their shrine were placed). Late in the evening, from 10 p.m. (on May 3) to 1 a.m. (of May 4), the transference of energy and prāṇapratiṣṭhā took place. First, empowered water from the pots (which had been brought from the sacrificial shed to the sanctum) was poured by all the priests on Kamadchi’s image in the womb chamber. A priest told me the ritual action on this evening was called bimba-śukla (bimba referring to mūrti, and śukla to abhiṣeka). I suppose ‘ablution’ (abhiṣeka) in this case means the sprinkling of waters from the pots and of ghee from the spoons on the image before, respectively after, the ritual transfer of energy. For this to happen, again a cord of holy Kuśa grass was strung from the yajñaśāla

Āgamic Temple Consecration Transnationalized


to Kamadchi’s image in the garbhagṛha, and three times the priests would symbolically lead energy from the sacrifical shed to the image, keeping constant 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.



4.3 Eighth Day, Sunday, May 4, 2014 The priests did not get much sleep in the night, as prāṇapratiṣṭhā was over around 1 a.m., and the mahā-kumbha-abhiṣeka ceremony would already start at 5 a.m. At 9 a.m. the ablutions of the temple (roof and towers) were already over, and the parallel actions on shrines and images inside the temple were finished around 12 a.m. These rites involved the bringing of a cow into the garbhagṛha, which turned out to be absolutely impossible. The farmer was not present this time, and the poor young and beautiful cow seemed to be afraid to enter the crowded temple in which loud life music was being played. So, after a long and unsuccessful attempt to persuade the cow, a substitute was brought into the ‘womb-chamber.’ After the festive pūjā, the singing of bhajans by Swami Vishvananda and his group of (German) disciples took place. 5

gamic Temple Consecration Transnationalized: Conclusions and Ā Uncertain Trajectories

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 MahāKumbha-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 ethnicization 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 managed to integrate into the social life of Hamm and gain enormous public recognition. 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 generated 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


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. 5.1 Transnationalization of the Ethnic The Kamadchi temple is a good example of religious glocalization. Āgamic consecration initiated and enforced the glocalization processes—the local becoming global and the global local. This pertains to Tamil Āgamic Hinduism implanted into German space, as well as to a culturally insignificant industrial park in the Ruhr area which attained significance by becoming a Hindu pilgrimage place. The Kamadchi temple was the first traditional style temple in continental Europe and is so far still the largest Hindu temple on the continent. Within a few years it developed into an eminent Hindu pilgrimage centre not only for Tamil Hindus, but even for (native) Germans. Particularly those of the second generation, who have already been born in Germany, say they feel like on a pilgrimage to India, when coming to the temple. Quite similarly, the temple is welcomed by German India lovers who rejoice finding India in Hamm-Uentrop. Nowadays, the temple is also internationally known and becomes a transnational meeting-point at the yearly temple festivals. The bigstyle chariot processions (tēr) attracted in the past 14 years up to 20.000 visitors and pilgrims from Germany and all over Europe—in 2008 even 25.000. The transnationalizing tendencies and networks pertain to the Sri Lankan Tamil diaspora community, as well as to the temple—visibly in big style already at the consecration in 2002, when artisans, priests, swamis, and traders of South Asia were present, some of whom had been the first time to Gemany, whereas others had occasional or permanent work and residence in the US, Canada, and Australia, and toured as professionals in temple building around the world. Transnationalizing also concerns religious ideas and practices. The Kamadchi temple became a contact zone wherein nowadays non-Hindus have arcanas performed and seek spiritual advice from a Hindu priest. 5.2 Aspects of Change Because of its architecture and the consecration rites the Kamadchi is perceived as ‘real’ and ‘authentic,’ unlike other Tamil Hindu places of worship which are conceived more like house-shrines by Sri Lankan migrants. While I do not deny the intention to reconstruct tradition as faithfully as possible, change is almost unavoidable if the cultural contexts change. There is need of adaption and, sometimes, old traditions get remoulded completely. The de-territorializing of culture which is so typical for modern times affords liberalisation of traditional customs (starting with the prohibition to leave the holy Āryabhata and travel overseas) and possibly attracts specialists who are



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 building, 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 violations 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. 5.3 Role of Aesthetics Both architecture and rituals can be understood as public representation systems, imagined homeland, and attracting mediators (whether received as irritating or attractive). The sensory aesthetics of the Kamadchi temple and its ceremonies have been amplifiers and ‘agencies’ of the public display of Tamil home culture. At the same time they functioned as mediators and connectors of previous and present home culture. To have a temple, which also looks like one in South India and Sri Lanka (and was consecrated accordingly), proved exceedingly crucial for both the migrants and the mainstream society. An interesting issue, which I have discussed elsewhere in more detail, is that contrary to the Muslim mosques, the temple was received exceedingly well and had very good press—even regarding violent rituals at the temple festival (Wilke 2006). I have called the Kamadchi temple a ‘contact zone’ and at another place an ‘unproblematic parallel society’ (Wilke 2013b). Both can go together, parallel society and contact zone. The merger that made this possible was not least a visible sacred building, a ‘real temple.’ It created a shared space of imagination among the minority culture of the migrants and the majority culture of native Germans, and a place were hosts und guests are situationally inverted. A beautified exterior and interior of the temple approximating it to great temples at home, colourful festivities and ritual ‘authenticity,’ such as the splendid consecration rites and the yearly temple festivals, remained great attractors. There was a tendency to top them every year. For 2014 was scheduled to

Āgamic Temple Consecration Transnationalized


present the temple towers in colourful paint. That only the vimāna could ‘make it,’ reveals some of the difficulties of Āgama culture going abroad. 5.4 Contingencies and Uncertain Future Transformations happen in times of cultural exchange particularly fast. There was a high amount of movement in the past decade, starting with the move of a charismatic non-Brahmin priest from the periphery to the centre and the effervescence this centre developed from local to global. But there is also a high amount of contingency regarding the near or distant future. We can only acknowledge the striking story of success in recent years, for which both the traditional architecture and the temple festivals were the driving motors. It is clear, however, that the big crowd on the Tēr day does not reflect belonging, but just an interest in a great annual event. A particular challenge regarding the future poses the young generation. At present the youngsters are no more as enthusiastic as their parents to invest into the temple life, and many are critical of priests, rituals, and Sanskrit language (Marla 2015). Although the temple festivals are an absolute must go, their role is less religious than providing big events and a meeting place. Did the festival move from a religious event to a spectacle in which the staging and consumerist character predominate? What does the declining number of participants at the re-consecration indicate? Will Sri Paskaran be able to build the International Cultural Centre? Will it be successful? There are many open questions left. The temple’s success remains a contested and fragile field. The article shows not only the successful sides, but also problematic ones. It is not hostile neighbourhoods, fear of foreigners, or blunt discrimination which the Kamadchi temple (and not only this temple) is presently confronted and struggling with, but most of all the finances and an all-to-strict German bureaucracy imposing additional burdens. There are also challenges from within the community. What the next twelve years will bring is difficult to say. Certain is only one trajectory: The temple remains a fascinating field of research and so does Āgamic temple consecration in the diaspora. 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.



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.


Ritual Observed: Notes on the Structure of an Image Installation 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 conversation 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 smilingly 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 afternoon in 1995, a small but constant stream of devotees had been visiting the shrine regularly over the years. The performance of the three-day installation ritual had obviously generated the desired effects, transforming an artful sandstone sculpture of a popular deity into a ‘living’ image worthy of worship. The consecration of temple images belongs—together with life-cycle rituals, initiations and ordinations—to a category of complex South Asian

© koninklijke brill nv, leiden, ���7 | doi ��.��63/9789004337183_014



rituals where efficacy is not only highly relevant but also immediately palpable: 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 communities. In a similarly straightforward manner, an image consecration is performed 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 concern 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 constitutive 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 communicative 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, rituals 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


‘[t]here must exist an accepted conventional procedure having a certain conventional effect,’ a procedure that includes ‘the uttering of certain words by certain persons in certain circumstances’ (1962: 14). And, according to Searle, who in many respects continued and elaborated upon Austin’s work, there are ‘sets of constitutive rules’ in accordance with which sentences must be uttered in order to perform speech acts (Searle 1969: 37). In an analogous way, the force—or the performative efficacy—of rituals such as the ones mentioned above depends on their being part of socially recognized, conventional procedures, where qualified persons (religious specialists) utter selected syllables, words and sentences (passages, verses, mantras from ritual texts) and perform various practices, and on their invocation of certain well-established and socially accepted, constitutive rules. The consecrations of temple images in South Asian contexts have a long tradition, reaching back to at least the late Gṛhya texts from the 4th–5th century CE (Einoo 2005: 95), and are part of a vast and diverse network of ritual and cultural practices.3 Consecrations of man-made images are deemed to be necessary preconditions for the legitimacy and functioning of places of worship. Such rituals are usually performed according to rules laid down in minute detail in handbooks, and combine on various levels elements that connect it to related systems and practices (e.g. Vedic recitation, Vedic sacrifice, elements of Puranic and tantric ritual systems, yogic practices, and others). The social recognition of these complex rituals is therefore beyond dispute. In the concluding part of this essay, I intend to focus briefly on the inner structure of the installation ritual, on some of the rules according to which its elements and different subsystems engage with each other, as well as on a number of questions related directly to its performance. In the last fifty years, the study of religions has seen the rise of a cluster of approaches that Thomas Tweed labeled in a recent article the ‘Quotidian Turn.’ Addressing the advantages and limitations, gains and losses of these developments, Tweed points out that with the obvious gains came along less obvious losses, such as ‘a tendency to attend most fully to ordinary people and everyday life, which minimizes the significance of clergy, beliefs, ecclesiastical institutions, prescibed rituals and consecrated spaces’ (Tweed 2015: 379). Even if image installations belong to the repertoire of organized, ‘prescribed’ religion, the present essay can be nevertheless located in an interpretive tradition that

3  For example, over the centuries, public installation rituals were often part of larger socioeconomic and political schemes that included land donation and administration, as well as religious expansion (Colas 2010: 319).



analyzes the everyday.4 It looks at practiced religion, the actual performance of a ritual that is, to be sure, based on prescriptive texts and is highly choreographed, but that nevertheless also has distinctive individual, particular features: its environment, the actors and the artefacts involved, the liberties taken (adaption, improvisation, etc.). The main methods applied in the research for this essay were participant observation and intensive conversations with participants in the consecration, religious specialists and devotees alike, along with a close reading of relevant texts from the ritual manual that served as script for the installation. The following description and analysis of a complex ritual attempts thus to hold together—at least to some extent—the ‘interpretive binaries’ pointed out by Tweed and others, with an emphasis placed on religious practice. 1


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 vulnerable 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 publisher 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

Figure 1

The sculptor begins.

Figure 2


Taking shape.

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 produce drafts of the new statue, giving them a number of specifications regarding 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 figure 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 conventional 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 intention 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 mischief, playfulness and power needed to be captured adequately in the new image.



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, aiming 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 followed 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 mixture 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 damaged 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 conversations with the purohit and the temple owner.

notes on the structure of an image installation

Figure 3

The clay model.

Figure 4


The sandstone image.

On the basis of astrological calculations, the family priest had already determined 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 owner’s relatives, to friends and acquaintances,10 and to some important personalities 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 yantras) 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.



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 positioning 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 prescriptions. 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 northwestern corner of the ritual enclosure. On the square-shaped navagrahamaṇḍala (the diagram 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


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 sacrificial space and sat down at their designated places. The team of five ritual experts comprised the family priest (purohit), who acted during the consecration 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 definition 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 contributions in Padoux 1986 and Bühnemann et al 2003. 14  The sarvatobhadramaṇḍala (the all-auspicious diagram) was square/shaped and symmetrical, with the basic pattern of 19 × 19 square fields that were grouped in larger geometrical 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 goddesses, along with the main goddess of the diagram, Śrī (Keul 2002: 15–16).


Figure 5


The protagonists.

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 vessels, 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


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 during 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 drinking pañcagavya in the same way as fire consumes firewood.’ This essay focuses in its theoretical considerations primarily on the consecration’s performance 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 imagined to be present on the sacrificial site. In other cases the abbreviated version with



placed on a heap of yellow-colored rice in the immediate vicinity of Gaṇeśa and worshiped as a pūrṇakalaśa, a pot of plenty.19 At the end of this sequence, one of the ṛtviks touched the yajamāna’s head and shoulders three times with the kalaśa. Ritual entreaties for an auspicious day (punyāhavācana)20 were followed by a brāhmaṇapūjā with a special focus on the hands of the ritual specialists. The figure of the brahmin ācārya is described in many texts as utmost venerable and as containing all the deities, being therefore an eminent object for worship. Their hands are seen in Vedic ritual contexts as equivalent to the sacrificial fire. Offering sacrificial materials into the fire or the ācārya’s hand establishes a connection with the world of the ancestors and gods.21 This was followed by a benedictory invocation of auspiciousness (svastivācana) and the scattering of boiled rice, and by another saṃkalpa and punyāhavācana accompanied by the sprinkling of water from the previously established kalaśa vessel on the sacrificer and in the direction of the diagrams, under the constant recitation of verses. The first maṇḍala the ritual group turned to was the one of the sixteen mothers. They were invoked one by one, and a ṛtvik touched every areca nut (pugīphala) with his right ring finger and applied kumkum paste on it. The yajamāna poured water (for rinsing the mouth and for washing the feet), threw a few grains of unbroken rice (akṣata) and laid a piece of cotton string symbolizing the offering of vastra on each of the sixteen squares. The ritual procedures at the śrīyantra were similar, except that the deities there were red dots on a canvas, and, while the ṛtvik still applied kumkum paste on each of these (this time with the middle finger of his right hand), the yajamāna threw the rice in the direction of the yantra and also performed the libations in its direction. The coordination between the members of the team of ritual specialists was smooth and efficient. During the worship of the mātṛkā-goddesses and of the śrīyantra-deities performed by the yajamāna, jointly guided and assisted by the main priest and one of the ṛtviks, the remaining priest and the two assistants prepared the implements for the nāndīśrāddha, the ritual for the ancestors: on a large plate made of ḍhāk leaves they laid out several bundles of kuśa grass, into which the ancestors (more precisely, the three forefathers 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


of the great-grandfather of the yajamāna’s father and mother, respectively22) were ritually invoked and offered the worship with sixteen services. The ancestors’ presence at the ritual was a necessary requirement, but it also was a temporary one, their brief visit reflecting a clear difference in the imagined divine capacities of the superhuman beings ritually gathered at the sacrificial site: the goddesses and gods were expected to dwell for the duration of the entire day; the ancestors, on the other hand, were formally dismissed immediately after receiving the ṣodaṣopacārapūjā. Lacking the attribute of omnipresence, their attendance may have been required by another descendant in some other ritual context.23 The leaf plate containing the grass bundles was then carefully folded, tied with a string, and placed on the ground near the sarvatobhadramaṇḍala for the whole of the remaining part of the installation ritual. In another nāndīśrāddha-related sequence, emotionally loaded but not sanctioned in the ritual texts, the yajamāna’s late father was remembered and his photograph garlanded. The picture remained in the vicinity of the sarvatobhadra during the entire three-day installation procedure. Well into the preliminary phase of the installation, the yajamāna then formally appointed the priests in a sequence called ācāryavaraṇa.24 Each ritual specialist was honored with a pañcopacārapūjā and given a checkered red and white scarf (that some of the priests wore on the next day of the ritual), and a yellow lower garment, as well as a copper vessel. The priests also received on their foreheads a tīkā administered by the yajamāna. All these sequences were accompanied by the constant recitation of various texts. A rather amusing episode during the honoring of the priests reminded the observers of the brahmins’ traditionally god-like ritual status. The garland offered to the purohit proved to be somewhat short and remained suspended in limbo, with one side dangling from the priest’s ear and spectacles frame instead of gliding down smoothly to his neck. Motionless like a statue, the priest did not budge. He continued his recitation unblinkingly until the yajamāna adjusted the garland, freeing it—with a mildly impish smile—carefully from behind the elderly man’s ear. The following sequences were related to each other in that they were addressed to deities imagined as directly connected to the location of the event. First, iron pegs were symbolically inserted into the ground (lohaśaṅku) 22  The nāndīśrāddha or nāndīmukhaśrāddha is often addressed only to the forefathers on the father’s side. Here, however, it explicitly included the mother’s side, too. 23  I am grateful to Pandit Induśekhar Pāṇḍey, the officiating head priest, for patiently answering my persistent consecration-related questions. 24  The importance of this rite is discussed, for example, in Welbon 1984. On the conditions that a potential ācārya is supposed to fulfill, see Joshi 1959: 72f.



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 supervising ṛtvik manifested his approval. This rite was aimed at the nāgas, the serpent 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 drawing 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 teamwork required for the smooth and efficient performance of most of the ritual activities in the consecration: the purohit was reciting the relevant text passages, including the names of the individual deities, one of the rtviks indicated 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, positioned around the yajñamaṇḍapa. The yajamāna followed the priests, bending 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 worship 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 center of attention. The god was invited to take his seat in the middle of the diagram, where a large kalaśa was placed, topped by a coconut. Before that, the

notes on the structure of an image installation


priests performed a quick instilment of life-breath (prāṇapratiṣthā) into a flat, small,25 golden Hanumān image, wrapped it in a betel leaf and placed it under the coconut onto the tāmrapātra on the kalaśa. The main element of the procedure was nyāsa (the yajamāna held the image in his hand, while repeating the seed syllables, bījamantras, recited by the purohit), to which we will return in more detail in connection with the installation of the sandstone image. The small golden image, ritually invested with divine power, served several purposes in the consecration: it was meant to doubly ensure the god’s presence on the sarvatobhadramaṇḍala and in the ritual area; it was to play the same supporting role in the circumambulation of the city on the event’s second day; and it was part of the ritual payment received by the main priest. After having established and secured Hanumān’s presence in the yajñamaṇḍapa and the performance of a detailed pūjā for the deity (including the offering of vastra, yajñopavīta, and the recitation of popular devotional prayers), the ritual activities were transferred to the neighboring bathing site (snānamaṇḍapa) where the sandstone image stood. The first sequence there was a short invocation of the protectors of the world (lokapālas), followed by the purifying touching of the image with kuśa grass. In addition to purification, the latter was intended to correct and atone for possible mistakes and flaws that may have occurred during the statue’s production. The yajamāna and the young assistants successively applied oil, ghee, and pañcagavya mixture on the sculpture in a lengthy procedure, taking utmost care that no crevice was left out. In the process, the sandstone changed color from a pale ochre to a dark red-brown. The assistants then poured—again successively—cow urine (gomūtra), the five nectars (pañcāmṛta), and a mixture of water from every pitcher that had been established on the maṇḍalas and had been collected in a huge brass vessel placed in front of the image. These were succeeded by affusions with several kinds of water from small brass vessels over the image: rain water (meghajala), water from the confluence of two rivers (saṅgamjala), water from a crossing or pilgrimage place (tīrthajala), and, finally, Ganges water (gaṅgājala). Immediately thereafter, the yajamāna garlanded the statue and applied red powder and yellow rice onto its forehead, and covered it with a damp, plain white cloth, which was meant to represent the sequence of jalādhivāsana, ‘the purification and consecration with water,’ during which an image in the process of installation is usually left overnight in a water tank. 25  The approximate size of this image was 2.5 × 3.5 cm.


Figure 6


Changing color.

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 usually 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 consecration 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


square fields with colored rice, on the diagram there were also three larger triangular 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 varying 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, creating 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 quantities 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—



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 diagrams. 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 recitation of devotional hymns continued at the sarvatobhadramaṇḍala, culminating in exclamations such as ‘bajrāṅgabalī kī jai!’ ‘Hail to the mighty Diamondlimbed 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 animated 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 (relatives 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

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.


Figure 8


Araṇipūjā. In the background the image after puṣpādhivāsana.

renovated shrine.27 Interestingly, although the image could not yet count as fully ritually enlivened at this stage, it nevertheless received the highly respectful, at times tender and playful, personalized treatment that often characterizes 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 rubbing block (the lower rubbing element, or adharāraṇi, aśvattha-wood), pulling vigorously the ends of the rope wrapped around the pole and swivelling the 27  The painting is presently part of the temple decoration.

notes on the structure of an image installation


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 glowing 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 blowing 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 implements 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.


Figure 9


Purification with butter.

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 coming 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 excessive heat was a vestimentary one: the elderly, bald purohit tied a handkerchief around his head, thus probably deviating somewhat from the dress code prescribed 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


Figure 10 Applying sugar and essential oil.

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 ‘hairline.’ 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



Figure 11 On the way out.

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 playing 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 necklace, 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


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 plaster 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 yellow 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 gathered behind them, forming a procession. Meanwhile it was late afternoon and still very hot outside. The festive procession 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, occasionally rushing ahead and stopping to watch from the margins. Despite the heat, many of his relatives and friends present during the days of the consecration 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 neighboring 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 29  See below for a discussion of this substitution. 30  On nyāsa in the pratiṣṭhā, see Einoo 2005: 109–113.



Figure 12 Aṅganyāsa.

Puruṣasūkta-verses and holding a mango in his left hand, he touched the designated 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 including 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


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 spontaneous 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 enthusiastic group of men (uncles, cousins, and friends of the yajamāna) managed 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 acknowledge 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.



Figure 13 To the temple.

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 flowers, (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 dh