Goddess Traditions in Tantric Hinduism: History, Practice and Doctrine [1 ed.] 1138823341, 9781138823341

Citation preview

Goddess Traditions in Tantric Hinduism

Hinduism cannot be understood without the Great Goddess and the goddessorientated S´a-kta traditions. The Goddess pervades Hinduism at all levels, from aniconic village deities to high-caste pan-Hindu goddesses to esoteric, tantric goddesses. Nevertheless, the highly influential tantric forms of South Asian goddess worship have only recently begun to draw scholarly attention. This book addresses the increasing interest in the Great Goddess and the tantric traditions of India by exploring the history, doctrine and practices of the S´a-kta tantric traditions. Tantric forms of South Asian goddess worship form a major part of what is known as ‘S´a-ktism’, and is often considered one of the major branches of Hinduism next to S´aivism, Vais.n.avism and Sma-rtism. S´a-ktism is, however, less clearly defined than the other major branches, and the book looks at the texts of the S´a-kta traditions that constitute the primary sources for gaining insights into the S´a-kta religious imaginaire, ritual practices and history. It provides an historical exploration of distinctive Indian ways of imagining God as Goddess, and surveys the important origins and developments within S´a-kta history, practice and doctrine in its diversity. Bringing together contributions from some of the foremost scholars in the field of tantric studies, the book provides a platform for the continued research into Hindu goddesses, yoga, and tantra for those interested in understanding the religion and culture of South Asia. Bjarne Wernicke Olesen is a Fellow at the Oxford Centre for Hindu Studies, UK, and a teacher of classical South Asian languages and religions at Aarhus University, Denmark. He is the director of the S´a-kta Traditions research project together with Gavin Flood, University of Oxford.

Routledge Studies in Tantric Traditions Series Editor: Professor Gavin Flood, University of Oxford The Routledge Studies in Tantric Traditions series is a major new monograph series which has been established to publish scholarship on South, East, and Southeast Asian tantric traditions. The series aims to promote the serious study of both Hindu and Buddhist tantric traditions through the publication of anthropological and textual studies and will not be limited to any one method. Indeed, the series would hope to promote the view that anthropological studies can be informed by texts and textual studies informed by anthropology. The series will therefore publish contemporary ethnographies from different regions, philological studies, philosophical studies, and historical studies of different periods which contribute to the academic endeavor to understand the role of tantric texts and their meaning in particular cultural contexts. In this way, the series will hope to establish what the continuities and divergencies are between Buddhist and Hindu tantric traditions, and between different regions. The series will be a major contribution to the fields of Indology, Sinology, History of Religions, and Anthropology. Identity, Ritual and State in Tibetan Buddhism Martin A. Mills The Khecarı-vidya- of Adina-tha A critical edition and annotated translation of an early text of hat.hayoga James Mallinson An Introduction to Tantric Philosophy The Parama-rthasa-ra of Abhinavagupta and its Commentary by Yogara-ja Lyne Bansat-Boudon and Kamalesha Datta Tripathi Tantric Mantras Studies on mantrasastra Andre Padoux Tantric Visual Culture A Cognitive Approach Sthaneshwar Timalsina Goddess Traditions in Tantric Hinduism History, practice and doctrine Edited by Bjarne Wernicke Olesen

Goddess Traditions in Tantric Hinduism History, practice and doctrine

Add Add

Edited by Bjarne Wernicke Olesen

AddAdd AddAddAdd Add AddAdd Add

First published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2016 selection and editorial material, Bjarne Wernicke Olesen; individual chapters, the contributors The right of Bjarne Wernicke Olesen to be identified as author of the editorial material, and of the individual authors as authors of their contributions, has been asserted by them in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Olesen, Bjarne Wernicke, author. Goddess traditions in Tantric Hinduism: history, practice and doctrine/ Bjarne Wernicke Olesen. pages cm. – (Routledge studies in tantric traditions) Includes bibliographical references. 1. Tantrism–History. 2. Hindu goddesses–History. I. Title. BL1283.83.I54 2015 294.5'2114–dc23 2015004295 ISBN: 978-1-138-82334-1 (hbk) ISBN: 978-1-315-74212-0 (ebk) Typeset in Times New Roman by Taylor & Francis Books

Contents

List of illustrations List of contributors Foreword Acknowledgement Introduction

vi vii x xii 1

BJARNE WERNICKE OLESEN

1

S´akti in early tantric S´aivism: historical observations on goddesses, cosmology and ritual in the Nis´va-satattvasam . hita

16

SHAMAN HATLEY

2

The (Un)Dreadful Goddess: Aghorı- in early S´a-kta tantras

33

JUDIT TÖRZSÖK

3

Varieties of melaka in the Jayadrathaya-mala: some reflections on the terms hat.ha and priya

51

OLGA SERBAEVA

4

Snakebite goddesses in the S´a-kta traditions: roots and incorporations of Tvarita-, Kurukulla- and Bherun.d.a-

74

MICHAEL SLOUBER

5

The Ca-n.d.a-lı- as S´akti: untouchable women in some tantric texts

96

MIKAEL AKTOR

6

S´a-ktism and hat.hayoga

109

JAMES MALLINSON

7

S´a-ktism, polity and society in medieval Malabar

141

RICH FREEMAN

Index

174

List of illustrations

Figures 4.1 A twelfth-century folio of the Tvarita-mu-lasu-tra in the Kaiser Library, Kathmandu 4.2 Kurukulla- yantra 7.1 Nampu-tiri Brahman novitiate priests offering bali as part of regular worship to the standardized, modular representation of the Seven Mothers, with accompanying Vı-rabhadra and Gan.apati on the southern side of the main shrine 7.2 Rough sketch plan giving major shrines and their relative orientation in the Ko-lattiri’s royal temple of Ma-t.a-yi Ka-vu 7.3 The Teyyam of Bhairavan, performed by Malayan dancers, modeled on the costumed ritual-begging of the Co-yi (Yogi) caste who settled in Malabar 7.4 Nampu-tiri of a prominent Va-dhyar family offering gurusitarpan.am from a large cauldron of artificial blood at his feet over a woven man.d.alam to their family goddess of Rakte-s´vari in a domestic rite

80 86

145 150

159

165

Tables 3.1 The ritual dimensions of the two varieties of melaka 4.1 Parallel passages in the Agnipura-n.a that are dependent on the Tvarita-mu-lasu-tra 4.2 Prakrit spell teachings for Bherun.d.a-

61 81 89

List of contributors

Mikael Aktor is Associate Professor of History of Religions at the Institute of Philosophy, Education and the Study of Religions, University of Southern Denmark. He holds a PhD from the University of Copenhagen, a part of which was carried out at the School of Oriental and African Studies, London University. His field of expertise is within the study of Dharmas´a-stra, in particular with a focus on caste and untouchability. He is presently engaged in a research of North Indian S´aiva temple ritual and temple sculpture as part of a general interest in ritual studies and religious aesthetics. Gavin Flood is a renowned Professor of Hindu Studies and Comparative Religion at the University of Oxford with a specialization in S´aivism and phenomenology and with research interests that span South Asian traditions. Since 2005 he has been the Academic Director of the Oxford Centre for Hindu Studies, which is a Recognised Independent Centre of the University of Oxford. He is the author of the acclaimed An Introduction to Hinduism. His other publications include: Body and Cosmology in Kashmir Saivism and Beyond Phenomenology: Rethinking the Study of Religion. He is also Editor of the Blackwell Companion to Hinduism. [John] Rich[ardson] Freeman is a cultural anthropologist with a background and research interest in Indology (Sanskrit and Dravidian languages and literatures). He currently teaches courses on Indian civilization, Hinduism, and the anthropology of religion for the faculties of History and Religion at Duke University, USA. His special area of research is Kerala, where he has carried out many years of ethnographic and textual work on the Teyyam traditions of spirit possession in Malabar, the wider region’s Brahmanical system of temple worship, and the literary history of Malayalam in its cultural relations with Sanskrit and Tamil. Shaman Hatley (Concordia University, Montréal), researches the literature, ritual, and social history of Tantric S´aivism in medieval India, and religion in premodern Bengal. Hatley’s dissertation, The Brahmaya-malatantra and Early Saiva Cult of Yoginı-s, analyses the history of the S´aiva cult of yoginı-s

viii

List of contributors

and provides a partial critical edition of one of its earliest scriptural sources, the Brahmaya-mala. Hatley is a contributor to the Ta-ntrika-bhidha-nakos´a (‘A Dictionary of Technical Terms from Hindu tantric Literature’), and has authored several articles and book chapters concerning tantric practices and goddess cults. His current research focuses on the ritual roles and divinization of women in early tantric S´aivism and Buddhism. He completed his PhD in Religious Studies at the University of Pennsylvania in 2007, under the direction of Harunaga Isaacson. James Mallinson is Lecturer in Sanskrit and Classical and Indian Studies at SOAS, University of London. He took his BA in Sanskrit and Old Iranian at the University of Oxford, followed by an MA in Area Studies (South Asia), with Ethnography as his main subject, at SOAS. His doctoral thesis, submitted to the University of Oxford, was a critical edition and annotated translation of the Khecarı-vidya-, an early text of hat.hayoga. Dr Mallinson has published eight books, all of which are editions and translations of Sanskrit yoga texts, epic tales and poetry. His recent work has used philological study of Sanskrit texts, ethnography and art history to explore the history of yoga and yogis. He is currently working on a monograph entitled Yoga and Yogis: The Texts, Techniques and Practitioners of Early Hat.hayoga. Bjarne Wernicke Olesen is a Fellow at the Oxford Centre for Hindu Studies, UK, and a teacher of classical South Asian languages and religions at Aarhus University, Denmark. He is founder and co-leader of the South Asian Religion research unit at Aarhus University and director of the S´a-kta Traditions research project together with Gavin Flood, Oxford. His research interests include Hindu and Buddhist traditions in South Asia (especially medieval S´a-kta traditions), Hindu Studies, and the history of research on religion. He is co-author of a new Danish translation of the Hat.hayogapradı-pika- (forthcoming) and has written one of the first Sanskrit grammars with a reader in Danish. Olga Serbaeva is a post-doctoral researcher at the Department of Indology, Asia-Orient-Institute, University of Zurich. Her primary interests include the representation of d.a-kinı-s in tantric Buddhism and S´aivism, the conceptualization of altered states of consciousness in the texts of the Vidya-pı-t.ha, such as provoked possession (a-ves´a), breaking of the codes of the mantras, and the application of the statistics in defining the closeness of the Sanskrit texts in general. Michael Slouber is Assistant Professor of South Asian Studies at Western Washington University, USA. His research focus is on the history of religion and traditional medicine of South Asia. Judit Törzsök studied English and Indian Studies at ELTE University, Budapest (MA, 1992), and completed her DPhil under Professor Alexis Sanderson’s

List of contributors

ix

supervision at Merton College, Oxford (2000). After a Junior Research Fellowship at Emmanuel College, Cambridge and a post-doctoral year at the University of Groningen, she was appointed Assistant Professor (maître de conférences), at the University of Lille III, where she still teaches Sanskrit and classical Indian religions and literature. In the past four years she has also lectured at the Ecole Pratique des Hautes Etudes, Paris, where she obtained her habilitation in 2010, supervised by Lyne Bansat-Boudon. Her main research areas are pre-twelfth-century S´aiva Tantrism (especially the early phase of goddess cults), classical Sanskrit literature (particularly drama), and pura-n.ic S´aivism, with occasional incursions into classical Tamil devotional literature. She has published articles on various aspects of S´aivism and tantrism, contributed to the Clay Sanskrit Library with two volumes, and participates in the Skandapura-n.a Project (Groningen) and the Ta-ntrika-bhidha-nakos´a (Hindu Tantric Dictionary Project, Vienna).

Foreword

This book brings together some of the most important scholars working on S´a-kta traditions today. Bjarne Wernicke Olesen’s book is groundbreaking in the sense that it is the first serious attempt to publish a broad range of work concerned at understanding what might be called S´a-ktism. Although there have been important general works on tantrism in an earlier age (one thinks, for example, of Hindu Tantric and S´a-kta Literature by T. Goudriaan and Sanjukta Gupta; and Jan Gonda’s Medieval Religious Literature in Sanskrit), it has only been in the last twenty-five years or so that our knowledge of the tantric and S´a-kta traditions has significantly improved. This has been largely due to the work of Alexis Sanderson, who has changed our understanding of Indian religions. Many of his students have developed important work on these traditions too, some of whom are represented in this volume. One of the features of the present volume, and the S´a-kta project of which this is a product in general, is the combination of textual scholarship with anthropology. Although one must be wary of reading present practices into the past, it seems to me undoubtedly true that present practices shed light on the meanings of past texts more than they obscure those meanings. Conversely, in the Indian context, we cannot understand present practices without reference to the texts of tradition. These approaches are well represented here. In his introduction Bjarne Wernicke Olesen points out the importance of Goddess traditions in India and the need for a systematic programme of research to understand these forms of religion. One of the difficulties is in drawing the boundaries around what we mean by S´a-kta traditions. A simple definition might be those religions whose focus is a goddess or range of related female deities. But then we have the non-Saiddha-ntika traditions, which are nominally S´aiva, focused on S´iva, and yet which have the Goddess at their esoteric heart (as Sanderson shows). These S´aiva-S´a-kta religions clearly cannot be excluded from our inquiry, but the boundaries between S´a-kta and S´aiva are not always clear. Part of the project is therefore to disentangle the historical complexity of these religions and the ways in which they interacted with each other, with contemporary politics, and with social systems. While there are interesting philosophical and theological questions generated by S´a-kta religion, the emphasis in this book is textual and historical, partly because it is necessary

Foreword

xi

to generate accurate forms of textual and historical knowledge before significant philosophical and theological work can be achieved, which is a second order reflection once the philological and anthropological work has been done (although never completed). But the study of these traditions does raise important philosophical and theological questions that we need to think about. Why is it that worship of a goddess is so often connected with blood sacrifice and, in the S´a-kta traditions, with antinomian ritual behaviour? What are the implications of reflecting upon a transcendent theistic reality in terms of the Goddess? Why is Sanskrit philosophical discourse gendered in such a way that S´iva is the male, transcendent deity while S´akti is the female power associated with chtonic forces? What kind of cultural psychology is at work here? These and other questions will need to wait for other spaces of reflection. Further historical questions also need to be asked. In particular, there is clear need for detailed textual and anthropological study at the regional level, as Rich Freeman has done for Kerala in the present volume and as Sanderson has done for Kashmir. We need detailed regional studies along with broader-sweep historical accounts that show the interrelationship between regions and the historical impact of S´a-kta traditions in different parts of the sub-continent. This book is an important contribution to this field of inquiry and lays the foundations for further historical, textual and anthropological accounts of the goddesses and their worship. Gavin Flood

Acknowledgement

We gratefully acknowledge the help and support of the Nehru Centre, Mr Shivdasani, Gisela and Børge Olesen, and the Oxford Centre Centre for Hindu Studies (OCHS), who generously granted funds for the organisation of the S´a-kta Traditions conference held in Oxford. Staff and students of the OCHS as well as staff of St Antony’s College and Somerville College were of great support in making this conference such a success. Gavin Flood wrote the summary of Rich Freeman’s paper. Marianne Fibiger, Mikael Aktor and Gavin Flood read an early draft version of the introduction, criticising and improving it. The OCHS standard for Sanskrit terms developed by Bjarne Wernicke Olesen and Gavin Flood has been used throughout this book: Sanskrit words in general are italicized and have lowercase. Uppercase is used for proper names and nouns designating religions and traditions, but not for modes of religious activity or textual genres. Lowercase is used for words adopted into the English and for adjectives, but not in the case of S´a-kta, S´aiva, Sma-rta, and Vais.n.ava and certain groups of texts for conventional reasons. Full mantras are written with capital letters and italicized. English endings are non-italicized as are Sanskrit words adopted into the English.

Introduction Bjarne Wernicke Olesen

General remarks As an increasing number of scholars have pointed out (e.g. Coburn 1984; Brooks 1990; Pintchman 1994; Flood 1996), Hinduism cannot be understood without the Indian Great Goddess and the goddess-oriented S´a-kta traditions. The Great Goddess in her many forms pervades Hinduism at all levels, from aniconic village deities to high-caste pan-Hindu goddesses to esoteric, tantric goddesses. Indeed, as David Kinsley has stated, ‘the great variety of goddesses allows one to find in their mythology and worship expressions of almost every important Hindu theme. In short, a study of Hindu goddesses is not so much a study of one aspect of the Hindu tradition as it is a study of the Hindu tradition itself ’ (Kinsley 1987: 4). Furthermore, as Alexis Sanderson has shown in his detailed philological studies (e.g. Sanderson 1988, 1995, 2009), S´a-kta tantric traditions have also had a profound influence on tantric Buddhism or Vajraya-na – in some cases (e.g. the Yoga-nuttara Tantras) to an extent that would allow for the term ‘S´a-kta Buddhism’. Jainism also has been influenced by tantric goddess worship, although to a lesser extent (Sanderson 2009). Nevertheless, the very influential S´a-kta tantric forms of South Asian goddess worship have only recently begun to draw a more broad scholarly attention. These traditions form the esoteric part of what has become known as ‘S´a-ktism’, i.e. traditions focusing on a goddess or a group of goddesses with a tendency to identify the Goddess as S´akti (‘power, force’) with ultimate reality, ens realissimum. By some scholars, S´a-ktism is seen as a network of both tantric and non-tantric traditions and considered one of the major branches of Hinduism next to S´aivism, Vais.n.avism and Sma-rtism. To others, S´a-ktism is first and foremost an esoteric development within tantric S´aivism, a subsect, or a form of ‘feminised S´aivism’ – sometimes also emphasised as ‘tantra par excellence’. Thus, ‘S´a-ktism’ is less clearly defined than the other major branches of Hinduism and it can be surprisingly difficult to discern S´a-ktism from S´aivism in its tantric forms. In the broader perspective of the study of religion, these often very complex and challenging forms of religion provide a test case for our understanding of Hinduism and raise important theoretical and methodological questions with regard to the study of religious traditions in South Asia as well

2

Bjarne Wernicke Olesen

as to the more general and comparative study of religion. However, these meta-theoretical questions will be dealt with elsewhere, as they form one of the major themes at the second international S´a-kta Traditions conference and subsequent volume.

This volume The main approach of the present volume is textual Indology as the texts of the S´a-kta traditions constitute our primary sources for gaining insights into the S´a-kta religious imaginaire, ritual practices, and history. However, this approach is supplemented by an art-historical perspective, an anthropological perspective and the overall perspective of the history of religions. The volume answers the increasing interest in the Great Goddess and the S´a-kta tantric traditions of South Asia, indicated by the growth of popular and secondary literature in recent years. Unfortunately, most of this new literature is written without the proper knowledge of the original texts and many misconceptions persist. The goal of this volume is therefore to shed new light on what has been called ‘some of the most complex, challenging and influential forms of South Asian religion’,1 based on solid research that is grounded in the original texts (most of which have yet to be translated into European languages). The book brings together contributions from some of the foremost scholars in the field of tantric textual studies with the particular aim of tracing and understanding developments in the history and rise of tantric worship of the Indian Great Goddess, covering a long period of Indian religious history from the early medieval period up to the present. The volume acts as a historical exploration of distinctive Indian ways of imagining God as Goddess (and goddesses), a survey of important origins and developments within S´a-kta history, practice and doctrine in its diversity, as well as an insight into the ‘weird and wonderful’ S´a-kta religious imaginaire and ritual practice that may be considered distinctive and thus sets ‘S´a-ktism’ apart from other South Asian religious traditions. Finally, the volume intends to fill a gap between dedicated monographs that focus on a single goddess, a single tradition, or a group of goddesses, and the few general surveys of ‘S´a-ktism’ or ‘S´a-kta religion’ at hand. For those interested in goddess worship and esoteric traditions in general, this book will establish important points of departure for further comparative work in the area of comparative religion. For those concerned with the specific understanding of religion and culture in South Asia, especially in its esoteric tantric forms, it will provide a platform for the continued research into Hindu goddesses, yoga, and tantra.

The S´a-kta Traditions project The S´a-kta Traditions project was started by Gavin Flood, Academic Director of the Oxford Centre for Hindu Studies (OCHS) and Professor of Hindu Studies and Comparative Religion at the University of Oxford, and Bjarne

Introduction

3

Wernicke Olesen, Research Fellow at OCHS and PhD scholar at Aarhus University. The project was inspired by Wernicke Olesen’s doctoral work and a mutual interest in a somewhat neglected area of research, as well as the recognition that any headway in this field would indeed be of great value for the future study of religions in South Asia. A research desideratum and aim for the project was thus formulated by Gavin Flood: Research on the S´aiva traditions has been quite extensively developed in recent years. Important work has been done on the Skanda Pura-n.a,2 the Pa-s´upatas,3 the S´aiva Siddha-nta,4 the non-dualistic S´aiva traditions, and their philosophical articulation in the Pratyabhijña-.5 But less work has been done on what might be called S´a-kta traditions, those traditions, tantric and non-tantric, focused on an independent Goddess (Devı-) or on S´iva’s power (s´akti). Research has been done on the Kubjika- tradition6 and on S´a-kta oriented S´aiva traditions7 but a sustained research programme that inquires into the history, doctrine and practices of what might be called ‘S´a-ktism’ is a desideratum. The aim of this research project is therefore to address fundamental questions such as the clarification of the distinction between S´aiva and S´a-kta traditions, questions about S´a-kta textual lineages and their interrelationship, the clarification of doctrines and practices of the different schools, questions about the relationship between the tantric and the pura-n.ic Goddess traditions, questions about the relationship between local Goddess traditions (such as the Teyyams in Kerala) and the pan-South Asian traditions, raising questions about the relationship between esoteric practices and the exoteric temple cults, asking what the delimitation of S´a-kta doctrine is, and what developments there are in contemporary S´a-kta worship. The research project is ambitious and intends to address these questions from a number of perspectives, i.e. a text-historical or philological perspective, an anthropological perspective on contemporary practice, a doctrinal or conceptual focus on theological reflection based on the textual material that has been established to date, an art-historical angle, as well as the meta-theoretical perspective of the study of religion. The project is being carried out as a collaboration between the Oxford Centre for Hindu Studies and the Department for the Study of Religion at Aarhus University. Project outputs are a number of international conferences held at Oxford and Aarhus, with papers and reports on ongoing research published in dedicated conference volumes as well as the establishing of an international S´a-kta research network with webpages at Aarhus and OCHS.8

The conference The first international S´a-kta Traditions conference was held in Oxford on 10–11 September 2011 at Somerville College. The conference was one of

4

Bjarne Wernicke Olesen

participation by invitation, and priority was given to the Indological perspective, as the texts of the S´a-kta traditions and their text-historical boundaries are hardly established. The conference was highly successful, with over fifty participants and twelve international speakers. The keynote address was given by Alexis Sanderson, Spalding Professor of Eastern Religions and Ethics at All Souls College, Oxford, considered by many to be the world’s foremost scholar on Sanskrit, Indology, and the tantric traditions. The conference was hosted by Professor Gavin Flood, and manager of the conference was Research Fellow Bjarne Wernicke Olesen. The first day of the conference concentrated on Indological and philological papers. Each session lasted one hour and each speaker chaired the following session. Alexis Sanderson gave the keynote address on the rise of goddess worship in early medieval India and was chaired by Gavin Flood. The keynote was followed by papers based on particular S´a-kta-related tantric texts, especially the Nis´va-satattvasam . hita (Shaman Hatley, Concordia), Brahmayamala and the Siddhayoges´varı-mata (Judit Törzsök, Lille–Paris), Datta-treyayogas´a-stra and the Hat.hapradı-pika- (James Mallinson, Oxford), Jayadrathaya-mala (Olga Serbaeva, Zürich) and the Yonitantra (Mikael Aktor, Odense). The day was concluded with a conference dinner for the speakers in the library at the Oxford Centre for Hindu Studies. The second day was of a more interdisciplinary nature, including papers from the perspectives of Indology, art history, the comparative study of religion, and anthropology (or ‘all the weird stuff’, as Gudrun Bühnemann humorously put it). Harunaga Isaacson (Hamburg) presented the first paper on the sa-dhana of female deities in Indian tantric Buddhism of the tenth to twelfth century and was chaired by Gavin Flood. The session was followed by . papers on s´aktilingas and similar sculptures from the perspective of Indology and art history (Gudrun Bühnemann, Wisconsin–Madison), Devı- worship as point of departure for a comparative project (Marianne Qvortrup Fibiger, Aarhus), snakebite goddesses in the S´a-kta traditions with particular reference to the Ga-rud.a Tantras (Michael Slouber, Berkeley), and the S´a-kta influence in the goddess cults of Malabar based on textual, ethnographic, and historical sources in Malayalam, Tamil and Sanskrit (Rich Freeman, Duke). The day was concluded with a final discussion led by Gavin Flood focusing on the use of the terms ‘S´a-kta’ and ‘S´a-ktism’.

Particulars In addition to the author’s abstracts initiating every chapter, some of the main features of the papers will be summarised below in an approximate chronological order, revealing some of the S´a-kta motifs and themes that recur throughout the volume.9 However, it should be emphasised that the points of view expressed here do not necessarily reflect those of the authors, although an attempt has been made to keep formulations and points close to the original. First and foremost, the following synoptic overview may prove useful in providing the

Introduction

5

reader with a general impression of how the heterogeneous sources and material show a remarkable congruence and continuity in the longue durée of Indian religious history covered by this volume – hopefully stimulating the reader as well as facilitating the communication between scholars working in different S´a-kta-related areas of research. In his keynote, Alexis Sanderson provided a revised and original overview of the rise of goddess worship in early medieval India based on new evidence, which in his own words suggests a major elaboration or extension of his previous work.10 In his lecture, Sanderson attempted to identify an independent S´a-kta tradition, which first appears inside Ka-pa-lika atimargic sources (fifth century CE) and then exists refined within the Mantrama-rga or tantric S´aivism, being assimilated to various degrees. In his article ‘The S´aiva Exegesis of Kashmir’ (2007), Sanderson made the point that the Mantrama-rga (fifth century CE +) in all its versions, both saiddha-ntika and non-saiddha-ntika, operates on a single liturgical system.11 In his lecture, this point was qualified by stating that there is indeed strong evidence for an independent Ka-pa-lika S´a-kta tradition coming in from the side and being to a greater or lesser extent assimilated into the matrix of the S´aiva common ritual system. One of the distinctive features of this S´a-kta tradition is the collective, orgiastic worship in which the individual identity is merged in a group. According to Sanderson, this tradition was not a creation of or a reformation within the Mantrama-rga, but an ancient tradition preserved at the margins among esoteric groups. What has been hitherto represented as a more elevated, mystical tradition within S´aivism is indeed so from the perspective of later S´aiva practitioners, but it is also – in Sanderson’s view – a very archaic and independent S´a-kta tradition from the outside, which has been preserved in spite of various attempts at reformation and assimilation. Furthermore, this was not just a ‘weird and wonderful’ phenomenon at the margins, but it flowed into the high cultural world, which became very evident in the work of the great S´a-kta-S´aiva exegete Abhinavagupta (tenth century CE), who came from this very Ka-lı-kula-Krama S´a-kta tradition of the non-saiddha-ntika division. Abhinavagupta intellectualised it and reinterpreted it as part of a new kind of cultural self-definition for learned Brahmans in Kashmir. Thus, Abhinavagupta stands out as the final culmination of this process of intellectualisation of the S´a-kta tradition or ‘high cultural S´a-ktism’, that begins in Kashmir in the ninth century and culminates in the late tenth century. It left its stamp on the culture of Kashmir but also had a huge influence on India as such. First and foremost, it was taken up in South India by the very influential cult of Tripura-sundarı-, a kind of purified ‘sweet S´a-ktism’ as opposed to the ‘shocking S´a-ktism’ of North India. It was assimilated by the orthodox Sma-rta Brahmanical system as well as by Pa-ñcara-tra Vais.n.avism and even by Dighambara Jainism and later S´vetambara Jainism, thus indicating the pervasive influence and popularity of S´a-kta traditions and goddess worship at the time. Finally, Sanderson emphasised, Buddhism from the eighth century onwards was completely transformed along the same lines of the S´a-kta developments which we see in S´aivism.

6

Bjarne Wernicke Olesen

In his chapter, Shaman Hatley examines the nature and historical development of goddesses and the category of S´iva’s ‘power(s)’, s´akti, in what may be the oldest surviving S´aiva scripture, namely the Nis´va-satattvasam . hita, dating from the fifth to seventh century CE. As pointed out by Hatley, the Nis´va-sa contains much of interest for the history of S´a-kta traditions, although the text places little cultic emphasis upon goddesses. Based on a diachronic analysis of the five books of the Nis´va-sa corpus – following the hypothetical stratification suggested by its editors (Goodall et al., forthcoming) – Hatley suggests that early tantric goddesses were frequently apotheoses of feminine-gender cosmological categories. This was followed by a movement from feminine-gendered principles to embodied goddesses that may be seen as characteristic of the early Mantrama-rga. The equation of goddesses with S´iva’s s´akti, and the ‘femaleness’ of s´akti, thus appears to be innovations of the early Mantrama-rga, aspects of which can be studied in the Nis´va-sa. Furthermore, we find in the Nis´va-sa early and very interesting complementary articulations of the idea of the singular Great Goddess (Maha-devı-) in the intersecting spheres of lay religion (laukikadharma) and tantric S´avism (Mantrama-rga). Noteworthy in the Nis´va-sa’s description of lay religion (in the Mukha-gama book) is the identification of Maha-devı- with prakr.ti and the absence of identification with s´akti and ma-ya- – the two other cosmogonic principles identified by Pintchman (1994) in the ‘symbolic complex’ of the Great Goddess found in the Pura-n.as. Thus, the Nis´va-sa may reflect a relatively early stage in the formation of the Great Goddess, preceeding the classic formulation of pura-n.ic S´a-kta theology in the Devı-ma-ha-tmya of the Ma-rkan.d.eyapura-n.a. However, this is complemented by a Mantrama-rga-oriented formulation of the Great Goddess found in a late stratum of the Nis´va-sa, namely the Guhyasu-tra book. Here we find the conflation of prakr.ti and s´akti that is critical to the pura-n.ic construction of the Maha-devı-. This becomes evident in the deification of prakr.ti and in the subsumption of all such cosmological goddesses within s´akti in the cosmology of the Nis´va-sa. It may be added that Hatley’s larger objective is to reconstruct the processes of the S´a-kta transformation and reconfiguration of the S´aiva cosmology in the Brahmaya-mala as well as to edit chapters from this text. The Brahmaya-mala is among the earliest extant tantras with a S´a-kta cultic orientation, and thus offers interesting insights into how the cultic emphasis on goddesses was extended into doctrine. In her paper, Judit Törzsök investigates the place of the goddess Aghorı- in early tantric scriptures (ca. 7th–8th century CE) and the possible factors contributing to the shaping of her cult and identity as a supreme goddess. Aghorı(‘undreadful’) is a frequently occurring name for the highest goddess in these early tantric sciptures as well as in non-tantric contexts at the time. In particular, Törzsök looks at the Siddhayoges´varı-mata of the Trika and the Brahmaya-mala, and points out that it is in the Brahmaya-mala that the goddess Aghorı- (or Aghores´-ı) is first described in detail as the supreme goddess of the cult. Furthermore, as Aghorı-’s cult is strongly associated with cremation ground rituals, it is suggested by Törzsök that Aghorı- might even have been

Introduction

7

the principal deity of the skull-bearing Ka-pa-lika ascetics, their tradition being an important precursor of later S´a-kta tantric currents. A number of remarkable ka-pa-lika elements thus figure in the Brahmaya-mala, including the ‘great vow’ (maha-vrata), the ‘great churning’ (maha-mantha-na), provoked posession (a-ves´a), and other cremation ground rituals. Among other continuities with pretantric currents, Törzsök emphasises the continuities with the La-kulas. She shows how the idea of three kinds of female powers (s´akti) headed by Aghorais already present in their scriptures predating the S´a-kta tantras. However, it is only in the S´a-kta tantras that the role of male and female deities and mantras are reversed and goddesses take the leading role. In their mantric identities, it can thus be observed how the tantric Aghora- or Aghorı- often seem to ‘s´a-ktise’ existing male S´aiva entities by replacing them with feminine ones. The Siddhayoges´varı-mata creates a fully S´a-kta doctrine and pantheon inspired by what was originally a Vedic mantra, namely the Aghora-mantra, and in the Brahmaya-mala, Aghorı-’s nine-syllable mantra is used as a substitute for the male nava-tman-mantra. According to Törzsök, it is also quite evident how Aghorı- was used to draw the cult of the seven Mothers into the S´a-kta tantric doctrine and ritual. In this case, Aghorı- would become the eighth Mother and supersede them, just as the tantric cult of the Mothers is supposed to supersede the nontantric cult of the Mothers. It therefore seems, that the cult of the Mothers was not a direct precursor of S´a-ktism, but instead was adopted and transformed in the same way the S´a-kta tantras adopted and transformed pretantric S´aiva doctrines and practices of the Atima-rga. However, some doubt still remains as to what extent Aghorı- functions merely as an epithet or does indeed refer to an identifiable supreme goddess with ka-pa-lika antecedents. In either case, Aghorı- clearly played a significant role and provides us with interesting insights into the development of S´a-kta tantric doctrine and practice. Olga Serbaeva explores particular instances of the tantric practitioner or sa-dhaka’s encounters with yoginı-s as they are described in the Jayadrathaya-mala (pre-tenth century?). These encounters are termed melaka or mela-pa and usually come at the end of a mantra recitation or homa. They invariably take place in desolate and dangerous places (cremation grounds, mountain tops, etc.) where the tantric sa-dhaka finds himself surrounded by yoginı-s in a visionary (and often intoxicated) state. He satifies the yoginı-s by offering them a bloody argha from his own limbs whereafter they transform him into a superhuman, often compared to Bhairava. These encounters constitute the core of the post-initiation tantric practice in the Vidya-pı-t.ha texts to which the Jayadrathaya-mala belongs.12 The text is influenced by ideas of the Krama and later Trika and constitutes a gathering of multiple traditions and material – ranging from descriptions of Ka-pa-lika-style human sacrifice to internalised yogic dhya-na. The text consists of four parts or s.at.kas (supposed to contain 6,000 verses each) of which the first s.at.ka is the most archaic and least S´a-kta in language and orientation. The first s.at.ka may be separated by a century or more from the remaining s.at.kas, which – taken together – can be said to form one of the most ‘S´a-kta’ tantric texts in existence, as suggested by

8

Bjarne Wernicke Olesen

Serbaeva. The deities invoked are almost exclusively female while Bhairava is reduced to a speaker and rarely appears in the man.d.alas. In some instances, the sa-dhaka is said to achieve the state of yoginı-s, to become a yoginı- or even to become like Ka-lı- herself. Serbaeva’s point of departure is the distinction made by Abhinavagupta in his Tantra-loka between two types of melaka, the hat.ha- and priyamelaka. The hat.hamelaka is also mentioned in the Brahmaya-mala as a potentially dangerous practice in which the sa-dhaka may risk his life and be killed by the yoginı-s/d.a-kinı-s. Thus, Serbaeva explores the nature of the hat.hamelaka in the Jayadrathaya-mala, including its relation to the practice of provoked possession (a-ves´a) and whether the same clear-cut hat.ha/priya distinction is found in this pre-Abhinavagupta text. However, a preliminary linguistic analysis of the occurences of the term melaka and its variants in the Jayadrathaya-mala does not reveal a clear-cut distinction, just as the potential violent character of the hat.hamelaka and the potential sexual character of the priyamelaka seems to be neither fully confirmed nor contradicted. On the other hand, a ‘melaka-manual’ is found in the Jayadrathaya-mala (3.38–39) which deals exclusively with technical aspects of melaka. The last verses of this manual do in fact allow for an alternative interpretation of some melakas as being ‘violent’ (hat.ha) in nature, as it is the yoginı-s who are ‘forced’ by the sa-dhaka to appear at a time and place chosen by him. This can be contrasted with the priya variety in which the sa-dhaka has to humbly await the yoginı-s in a place and time where they ‘love’ (priya) to manifest themselves. Thus, in the hat.ha variety in the Jayadrathaya-mala the sa-dhaka becomes the main orchestrator of the encounters with yoginı-s. This practice contrasts with earlier tantras and is based on a set of powerful mudra-s provoking altered states peculiar to the Jayadrathaya-mala. These mudra-s are complex ritual procedures including body postures, mantras, shouting, breath control, and so on, related to almost all instances of melaka in the text through which the sa-dhaka gains access and control over the yoginı-s. As the Jayadrathaya-mala links melaka with mudra-s and the yogic mastery of the subtle body, the text constitutes an important step in the proces of the interiorisation and intellectualisation of the melaka as it is known from the works of Abhinavagupta. After being internalised, the melaka was often encoded by kun.d.alinı--related procedures as a practice for the expansion of consciousness (in which the priya/hat.ha distinction became irrelevant). Furthermore, Serbaeva suggests that one consider the preparatory phase of the melaka as a visible ritual praxis in which the priyamelaka is distinguished by secret signs (chomma) as well as by a low-risk collective practice ending in a collective altered state, i.e. a regular ‘tantric feast’ with human yoginı-s. This is opposed to the exceptional and individual high-risk hat.hamelaka involving non-human yoginı-s, aiming at the personal transformation of the sa-dhaka in a solitary visionary state. Serbaeva concludes with some interesting reflections on the nature of this hat.hamelaka-related transformation of the sa-dhaka. Michael Slouber takes a closer look at early identities and transformations of snakebite goddesses found in the S´aiva Ga-rud.a Tantras, who were later

Introduction

9

incorporated into the influential S´a-kta traditions of the 9th–12th century. The prevalence of snakebite in South Asia was from an early time coupled with popular goddess worship, resulting in a number of influential S´a-kta traditions dedicated to goddesses associated with snakes and healing snakebite. Although the literature and worship of these goddesses became widespread, little work has been done on their textual traditions, as pointed out by Slouber. He suggests, that some of these goddesses were popularised by the Ga-rud.a Tantras. In his paper, he focuses on the three goddesses Tvarita-, Kurukulla- and Bherun.d.a-. The goddess Tvarita- (‘The Swift One’) was in time identified with high goddesses such as Kubjika-, Durga-, and Ka-lı-, as well as with Padma-vatı- in the Jaina Tantras (probably via the influential cult of Tripura- Sundarı-). Kurukulla- is best known as a tantric Buddhist goddess often identified with Ta-ra-, but she also figures as an independent goddess in the S´rı-vidya- and Ka-lı-kula sources from the ninth to the sixteenth century apart from the Ga-rud.a Tantras. As Slouber shows, it remains a complex question whether her origin is to be found in a S´aiva or Buddhist context. In the case of Bherun.d.a-, Slouber shows the usefulness of distinguishing between the goddess’ identity as an independent snakebite goddess of the Ga-rud.a Tantras and her identity as an ancillary goddess in other S´a-kta sources. It becomes clear, that Tvarita-, Kurukulla- and Bherun.d.a- each have multifaceted identities as well as separate literature, visual forms, spells, and identities in the early S´aiva Ga-rud.a Tantras and the early S´a-kta Tripura-sundarı- literature. In relation to the important question of the relationship between the tantric and the pura-n.ic goddess traditions, Slouber presents new evidence that Pura-n.ic chapters on snakebite goddesses were borrowed directly from tantric sources. Thus we find parallel passages in the Agni Pura-n.a that are obviously dependent on the Tvarita-mu-lasu-tra, the latter being ascribed to the Trottalatantra, i.e. one of the most cited canonical Ga-rud.a Tantras. Mikael Aktor takes a closer look at the Ca-n.d.a-lı-’s role in tantric texts. She is the lowest of low-caste women, belonging to what Dumont has labelled ‘the old prototype of the Untouchable’. At the same time she is the ‘Chief Goddess, Lady of hosts’ in tantric texts. Thus, the Yonitantra (eleventh century CE) evokes a yantra representing the yoni with the Ca-n.d.a-lı- as the central divinity. In her worship (yonipu-ja-) the worshipper becomes equal to S´iva. As Aktor observes, low-caste women such as the Ca-n.d.a-lı- or the D . ombı (washer woman) figure prominently in many of the tantric texts – especially in the Buddhist and Vais.n.ava tantric Sahajiya- literature – in which she personifies the kun.d.alinıs´akti. This raises the question of the reasons behind such antinomian tantric tributes to the untouchable woman par excellence. Through a study and analysis of the literary images of the lowest castes in other genres, Aktor shows how the use and ritual significance of these caste labels in the tantric texts becomes intelligible. A multifaceted image of the untouchable woman as a reservoir of contrasting forces emerges from this study. In general, female sexuality is an expression of s´akti, and as seen in the law books (e.g. Ma-navadharmas´a-stra), a more free flowing sexuality is attributed to women, especially from the lowest

10

Bjarne Wernicke Olesen

castes, who are not restricted and ‘tamed’ by marriage into one of the four varn.as. Thus, unmarried, childless and sexually active low-caste women are highly potent and preferred as agents of ritual transformation in an antinomian tantric context. But even more so is the Can.d.a-la woman, who is often associated with savage life beyond civilised norms. In classical texts (e.g. Cha-ndogyopanis.ad), the Can.d.a-la represents a category of untouchable castes rather than a single caste. Two different images of the Can.d.a-la emerge in the early centuries of composition of the Maha-bha-rata and early law books, namely the domesticated caste of Can.d.a-la and the savage Can.d.a-la – the latter representing a tribe rather than a caste. The tribal stereotype as an antithesis to Brahmin values is preserved in later narrative literature such as Ba-n.abhat.t.a’s novel Ka-dambarı(sixth century CE). As can be observed in the novel, ambiguous attitudes toward untouchability are at stake – the wild and unspoiled Ca-n.d.a-lı- is also seen as disturbingly attractive. Aktor suggests that the rationale is turned on its head in the novel, that is, beside the level of social norm and public censure dictated by kinship and class, there is another secret level at which the untouchability protects the unspoiled beauty and pure sexuality of the Untouchable against the impure touch of the Touchable. However, the Can.d.a-la’s impurity is not only the lowest in a continuum. Patañjali’s commentary on one of Pa-n.ini’s rules (2.4.10), as well as Buddhist Ja-taka stories, reveals the idea that the Can.d.a-las carry with them an inborn ‘magic’ quality and pollution that cannot be removed by ordinary means. A ritual transformation of this inauspicious quality can therefore be highly powerful in a tantric context. Furthermore, as Aktor points out, the association between Ca-n.d.a-lı-s and menstruation can be traced back to the earliest texts dealing with Untouchables. In the law books, the term ca-n.d.a-lı- is frequently used to designate a woman on her first (and most polluting) day of menstruation. Thus, whereas the Yonitantra places the Ca-n.d.a-lı- at the centre of the yantra, the Br.hadyonitantra places the ‘flowering’ or menstruating woman (pus.paru-pin.-ı) at the centre. As can be seen in mythical aetiologies, death, grievious sin, female sexuality and menstruation are often related. The - inversion of norms is also at stake in relation to the D . ombı in the Caryagıti, an eleventh-century CE Buddhist Sahajiya tantric text, as Aktor shows. Here we again find the secret level serving as the playground for the Siddhas’ experiments with sexual energy and mystical experience – expressed in the tantric language of implicit meanings (sandhya-bha-s.a-). As Aktor points out (with a reference to Dumont’s yoga–bhoga dichotomy), instead of rejecting the enjoyments of the senses (yoga), the ritual enjoyment (bhoga) of experiences otherwise considered impure became possible through this inversion of norms at a secret level. For some of the tantric traditions of the Mantrama-rga, the attribution of ritual power to bhoga became one of the primary means for attaining yogic states of consciousness – and thus to have your cake and eat it, as it were. This became the case with the disturbingly attractive Ca-n.d.a-lı-. James [Jim] Mallinson looks at hat.hayoga’s relationship with S´aivism and its S´a-kta manifestations and examines one of the ways S´a-ktism endured after its heyday in what Alexis Sanderson (2009) has termed the ‘S´aiva Age’ (5th–13th

Introduction

11

century CE). The thirteenth century CE thus marks the end of the heyday but also the beginning of hat.hayoga’s rise to a position of dominance among India’s soteriological methods. In his paper, Mallinson shows how this very succesful form of yoga – first and foremost popularised by the Hat.ha[yoga] pradı-pika- (ca. 1450 CE) – was a result of the ‘S´a-ktisation’ of hat.hayoga and its appropriation by the Na-ths, a S´a-kta cult of the Pas´cima-mna-ya or western stream of Kaula S´aivism. Mallinson starts off by summarising the distinguishing features of early hat.hayoga with the Vais.n.ava Datta-treyayogas´a-stra as point of departure and then goes on to summarise the techniques of hat.hayoga as taught in a number of pre-Hat.hapradı-pika- texts identified as the corpus of early hat.hayoga (neither S´a-kta nor Na-th). Thus, in its early formulations, hat.hayoga was closely associated with traditional ascetic practices of boosting the beneficial effects of celibacy, i.e. the preservation of semen (bindudha-ran.a). This early hat.hayoga was characterised by a variety of yogic techniques later known as mudra-s (e.g. viparı-takaran.-ı and khecarı-mudra-) with the primary aim of keeping the semen (bindu, amr.ta) in the head of the hat.hayogin. According to a pervasive yogic physiology (first found in the Amr.tasiddhi), bindu/amr.ta rains down from the moon in the head and is burned up in the sun at the base of the spine. As the bindu was considered the vital principle, its loss was supposed to lead to weakness and death, that is, if preventive yogic techniques were not employed. Should the hat.hayogin be as unfortunate to ejaculate, he might even use the technique of vajrolimudra- as a damage control to create a vacuum in his abdomen and resorb his bindu. Mallinson then goes on to show how S´a-kta techniques of yoga were grouped separately from hat.hayoga as layayoga in a typology that became commonplace, but was first found in the Datta-treyayogas´a-stra. Thus, in the same period as early hat.hayoga we find the influential kun.d.alinı- system of yoga developed by S´a-kta traditions. In this system, the yogin visualised the rising of the serpent goddess kun.d.alinı- from her home at the base of the spine (a-dha-ra) up through a series of cakras (usually six) to union with S´iva or, as is the case in the S´a-kta-oriented Khecarı-vidya-, the great store of amr.ta in the head, with which she floods the body on her journey back down. This stands out as a remarkable contrast to the cavity-sealing khecarı-mudra- of early hat.hayoga. Furthermore, vajrolimudra- as described in the S´a-kta hat.hayogic text S´ivasam . hita is a technique not for damage control but for absorbing the commingled sexual fluids (bindu and rajas) of both male and female partners in order to combine them with one’s own body. The Datta-treyayogas´a-stra also mentions a number of sam . ketas or ‘secret techniques’ in relation to layayoga that are found in other texts as well. These secret techniques focus on visualisations of, and meditations on, places and energies in the body – again in contrast with the more physical yogic techniques of early hat.hayoga. To the techniques of layayoga is further added the concentration on na-da or ‘internal sound’ (said to be the best form of layayoga in the S´ivasam . hita). As Mallinson points out, the yoga of early texts associated with the Nath gurus (e.g. Goraks.as´ataka and Yogabı-ja) seem to correspond more with layayoga than with early hat.hayoga. However,

12

Bjarne Wernicke Olesen the kun.d.alinı- system and techniques of layayoga developed in a S´a-kta milieu were then overlaid onto the techniques of hat.hayoga in a number of texts. An early attempt is the Vivekama-rtan.d.a while the later S´ivasam . hita presents a more coherent synthesis. However, the Hat.hapradı-pika-’s fusion became the locus classicus and dominant form of hat.hayoga as well as of yoga more broadly conceived. In this text, S´a-kta techniques of kun.d.alinı--oriented layayoga was included under the rubric of the otherwise bindudha-ran.a-oriented hat.hayoga. At the same time, the purpose of hat.hayoga was to a great extent realigned to that of layayoga (e.g. the raising of kun.d.alinı-). Mallinson identifies early practitioners of hat.hayoga as munis such as Kapila and Datta-treya in contrast to layayogapractising, S´a-kta-oriented siddhas such as Goraks.a – a distinction that seems to persist today between S´a-kta Na-ths and the more orthodox Das´ana-mı-s and Ra-ma- nandı-s. Finally, Mallinson locates these developments in a wider perspective, and suggests that the co-option of hat.hayoga by a S´a-kta tradition can be seen as representative of the development within S´a-ktism of a less sectarian and more universal yoga (i.e. a democratisation of yoga) and of the formation of the Na-th sam . pradaya. This originally non-celibate S´akta tradition developed into the highly influential celibate order of Nath ascetics, who are nowadays more in keeping with the expurgated ‘sweet S´a-ktism’ of the south Indian S´rı-vidya-, having the benevolent goddess Ba-la- or Tripurasundarıas their focus of worship. In the last paper Rich Freeman presents a case study of S´a-kta traditions in Kerala during their formation in the later medieval period and traces their continuities into the present. In this important paper Freeman integrates knowledge of Sanskrit and Malayalam texts with knowledge gained through years of fieldwork. In particular, Freeman shows that many of his findings in the anthropology of Kerala are borne out in the textual findings of Alexis Sanderson’s critical Indology. Freeman begins with the contemporary situation in Kerala, that in the temples and in performative cults there has been a synthesis of puranic and tantric ideas and practices. But beneath the general and vague invocations of a ‘Mother Goddess’ or ‘Bhagavatı-’ worship, there are more complex textual and historical instantiations of S´a-kta traditions in Kerala. One of the main findings Freeman presents is the way in which the Nampu-tiri Brahman caste has accommodated and appropriated S´a-kta traditions, stripping them of any association with blood sacrifice, non-vegetarian and alcoholic offerings, and ritualised sex and sexual imagery. This Kerala tantric tradition is an amalgam of vedic and tantric rites, focused on the temple, in which vedic as well as tantric mantras are used in temple ritual that contains no elements of impurity characteristic of lower-caste S´a-kta rites. The dominant Brahmans are known as Tantris, those who install images in temples, perform annual festivals, and have the hereditary knowledge and authority in these matters, in contrast to those Brahmans of subordinate ritual status called S´a-ntis, who are engaged in daily rites of worship. This Nampu-tiri tradition has been in place since the fifteenth century and the main text of the tradition, the Tantrasamuccaya, is still used in temples today. The importance

Introduction

13

and centrality of this text is witnessed by layers of Sanskrit and Malayalam commentaries. There is also the earlier Is´a-nas´ivagurudevapaddhati, a composite text that is generally framed in a S´aiva Siddha-nta tradition that goes back to the twelfth century, but that incorporates Vais.n.ava and S´a-kta cults of worship. A third important text, the S´es.asamuccaya, composed within the same Nampu-tiri family as the Tantrasamuccaya, presents the remainder of the deities not dealt with in that text and includes S´a-kta pantheons of the seven Mothers, Sarasvatı- and related deities (S´rı-, Parvatı-, Jyes.t.ha-), and the demon slaying Rurujit. Here we have an absorption of S´a-kta elements into the tradition that reflects the Brahmanical control of lower social groups and their deities. In Kerala there existed a complex social structure of sambandham relations in which the younger siblings of the eldest Brahman male had liaisons with upper-caste, non-Brahman women. This meant, as Dumont had noted, that most Brahman males’ vedic ritual status was compromised and so, Freeman argues, an alternate tantric order offered a ‘socio-religious compensatory validation’. In the context of these textual and social traditions, Freeman presents an example from Nı-les´varam and the teyyam festival at which alcohol and blood are offered to, often, female deities embodied in possessed male dancers. Here at Nı-les´varam, a S´a-kta priestly caste, the Pit.a-rar, had royal patronage and Freeman shows that the royal shrine at Nı-les´varam and others along the Malabar coast represent the worship of Rurujit from the S´es.asamuccaya. In a way parallel to the Orissan Paippala-din Atharvavedins that Sanderson has shown adapted Trika and other S´a-kta cults, so the Kerala Brahmans under royal patronage adapted S´a-kta traditions. The shrine at Ma-t.a-yi Ka-vu presents an architectural reflection of a process of assimilation of a S´a-kta cult. Rurujit is the goddess who conquers the demon Ruru, a demon who is assimilated into an enemy of the former kingdom. After killing the demon the goddess is pacified and installed in the temple as the royal protector. The goddess is incarnated in the spirit-possessed teyyam in the temple’s outer courtyard on an annual basis. This pattern of the localisation of a goddess who upon defeating a demon has to be controlled is often encountered in Kerala. Indeed, we see a general pattern in Kerala of Nampu-tiri Brahmans responding to S´a-kta and tantric traditions by attempting to gain control of them and, as Freeman observes, ‘bring them within the orbit of their social authority’. One of the main features of these S´a-kta traditions in Kerala is the blood rites. While the teyyams are offered real blood from decapitated chickens, the Nampu-tiris offer substitute blood to appease their deities. But the teyyams are also offered substitute blood tainted by real, chicken’s blood. Freeman shows two contrasting logics behind this. On the one hand the Brahmans wish to avoid pollution, on the other the castes who offer to the teyyams wish to increase the amount of blood offered. Here we have in stark contrast a distinction and tension between the S´a-kta and the vedic traditions still present on the Malabar coast. Ørslev Kloster, 2014

14

Bjarne Wernicke Olesen

Notes 1 2 3 4 5 6 7 8 9

10

11 12

Centre for Tantric Studies, Hamburg (www.tantric-studies.uni-hamburg.de/about-us/). Adriaensen et al. 1998; Bakker and Isaacson 2004. Bisschop 2006. Goodall 1998, 2004. E.g. Sanderson 2007: 231–442. Goudriaan and Schoterman 1988, 1994; Heilijgers-Seelen 1994. Sanderson 2007, 2009; Padoux 1994; Vasudeva 2004; White 2003. www.ochs.org.uk/research/sakta-traditions (Oxford); http://projects.au.dk/sakta-tradi tions/ (Aarhus). Alexis Sanderson, Harunaga Isaacson, Marianne Fibiger and Gudrun Bühnemann did not contribute with papers to this volume. However, a summary is given of some of the main points of Alexis Sanderson’s keynote address, as his lecture provides an important framework of understandning for the papers collected in this volume. It is our hope that their contributions may form part of subsequent conference volumes in the S´a-kta Traditions project. In his influential article on S´aivism and the Tantric traditions (1988), Sanderson presented a broad overview of the traditions, counting two branches of the Atima-rga (second century CE +), namely Pa-s´upata and La-kula, and two branches of the Mantrama-rga (fifth century CE +), namely Siddha-nta and Ka-pa-lika-style cults. However, in the light of new evidence, Sanderson suggested the addition of a third branch of the Atima-rga (3rd–5th century CE), encompassing Somasiddha-ntins and Ka-pa-likas. In the context of S´a-kta traditions, this would imply that an early S´a-kta dimension can be traced within the Atima-rga as well. Sanderson stated that this point was made partly to undermine the tendency of scholars in the field ‘to hermetically seal themselves off inside one or other of these branches and not look over the wall to see what’s on the other side’. Cf. Sanderson’s mapping of the Tantric traditions (Sanderson 1988).

References Adriaensen, Rob, Bakker, T. Hans and Isaacson, Harunaga. 1998. The Skandapura-n.a, Vol. I. Adhya-yas 1–25. Critically Edited with Prolegomena and English Synopsis. Supplement to Groningen Oriental Studies. Groningen: Egbert Forsten. Bakker, T. Hans and Isaacson, Harunaga. 2004. The Skandapura-n.a, Volume II A, Adhya-yas 26–31.14: The Va-ra-n.ası- Cycle. Critical Edition with an Introduction, English Synopsis and Philological and Historical Commentary. Supplement to Groningen Oriental Studies. Groningen: Egbert Forsten. Bisschop, Peter. 2006. Early S´aivism and the Skanda Pura-n.a. Sects and Centres. Groningen Oriental Studies, 21. Groningen: Egbert Forsten. Brooks, Douglas Renfrew. 1990. The Secret of the Three Cities. An Introduction to Hindu S´a-kta Tantrism. Chicago, IL and London: University of Chicago Press. Coburn, Thomas. 1984. Devı--Ma-ha-tmya. The Crystallization of the Goddess Tradition. Delhi: Motilal Banarsidass Publishers. Flood, Gavin. 1996. An Introduction to Hinduism. Cambridge: Cambridge University Press. Goodall, Dominic. 1998. Bhat.t.a Ra-makan.t.ha’s Commentary on the Kiran.aTantra. Volume I: chapters 1–6. Critical Edition and Annotated Translation. Publications du départment d’Indologie 86.1. Pondicherry: Institut Français de Pondichéry/École française d’Extrême-Orient.

Introduction

15

Goodall, Dominic. 2004. The Para-khya Tantra. A Scripture of the S´aiva Siddha-nta: A Critical Edition and Annotated Translation. Collection Indologie 98. Pondicherry: Institut Français de Pondichéry/École française d’Extrême-Orient. Goodall, Dominic, et al. Forthcoming. The Nis´va-satattvasam . hita: The Earliest Surviving S´aiva Tantra. Volume 1. A critical edition of the Mu-lasu-tra, Uttarasu-tra, and Nayasu-tra. Franco-German Early Tantra Series. Goudriaan, Teun and Schoterman, Jan A. 1988. The Kubjika-matatantra. Kula-lika-mna-ya Version. Critical edition. Orientalia Rheno-Traiectina, XXX. Leiden: Brill. Goudriaan, Teun and Schoterman, Jan A. 1994. The Kubjika- Upanis.ad. Edited with a translation, introduction, notes and appendices. Groningen Oriental Series, volume X. Groningen: Egbert Forsten. Heilijgers-Seelen, Dory. 1994. The System of Five Cakras in Kubjika-matatantra, 14–16. Groningen Oriental Series 9. Groningen: Egbert Forsten. Kinsley, David. 1987. Hindu Goddesses. Visions of the Divine Feminine in the Hindu Religious Tradition. Delhi: Motilal Banarsidass Publishers. Padoux, André. 1994. Le Coeur de la Yoginı-: Yoginı-hr.daya avec le commentaire Dipikad’Amr.ta-nanda. Paris: de Boccard. Pintchman, Tracy. 1994. The Rise of the Goddess in the Hindu Tradition. Albany, NY: State University of New York Press. Sanderson, Alexis. 1988. “S´aivism and the Tantric Traditions”, pp. 660–704 in The World’s Religions, ed. S. Sutherland, L. Houlden, P. Clarke and F. Hardy. London: Routledge and Kegan Paul. Sanderson, Alexis. 1995. “Vajraya-na: Origin and Function”, pp. 89–102 in Buddhism into the Year 2000. International Conference Proceedings. Bangkok and Los Angeles, CA: Dhammaka-ya Foundation. Sanderson, Alexis. 2007. “The S´aiva Exegesis of Kashmir”, pp. 231–442 in Mélanges tantriques à la mémoire d’Hélène Brunner/Tantric Studies in Memory of Hélène Brunner, ed. Dominic Goodall and André Padoux. Pondicherry: Institut français d’Indologie/École française d’Extrême-Orient. Sanderson, Alexis. 2009. “The S´aiva Age: The Rise and Dominance of S´aivism during the Early Medieval Period”, pp. 41–349 in Genesis and Development of Tantrism, ed. Shingo Einoo. Tokyo: Institute of Oriental Culture, University of Tokyo. Vasudeva, Somadeva. 2004. The Yoga of the Ma-linı-vijayottaratantra. Pondicherry: Publications de l’Institut français d’Indologie, no. 97. White, David Gordon. 2003. Kiss of the Yoginı-: ‘Tantric Sex’ in Its South Asian Contexts. Chicago, IL: University of Chicago Press.

1

S´akti in early tantric S´aivism Historical observations on goddesses, cosmology and ritual -1 in the Nis´va-satattvasam . hita Shaman Hatley

Introduction This essay seeks to elucidate conceptions of s´akti and the roles of goddesses in the early form of tantric S´aivism represented by the Nis´va-satattvasam . hita. ´ Probably the oldest surviving tantric Saiva scripture, portions of the text could date to as early as the fifth century C.E.2 The essay emerges from preliminary observations I made concerning continuities between the Nis´va-satattvasam . hita and the Brahmayamala, the latter being one of the oldest surviving tantric S´aiva texts with a S´a-kta cultic orientation (Hatley 2007). The S´a-kta cults of vidya-pı-t.ha works such as the Brahmaya-mala evidence profound transformations of S´aiva cosmology,3 with myriad s´aktis displacing the male deities who presided over the hierarchy of tattvas or ontic levels as delineated in earlier Siddha-ntatantras. While the aims of the present essay are limited, my larger objective is to reconstruct the processes of transformation underpinning the Brahmaya-mala’s S´a-kta cosmology, in tandem with the project of editing chapters of this text concerned with the intersecting subjects of cosmology and initiation (pat.alas 32–38). In studying the Nis´va-satattvasam . hita – hereafter, ‘the Nis´vasa’ – I rely heavily upon the foundational studies of Alexis Sanderson (2006) and Dominic Goodall (forthcoming), particularly in the area of cosmology. My reading of the Nis´va-sa is based on the provisional editions circulated among participants in the Nis´va-sa workshop (Pondicherry, 2007) and Early Tantra project (2008–2010);4 and the more recent draft editions and translations of Goodall et al. (forthcoming), and, for the Mukha-gama, Nirajan Kafle (forthcoming). I approach the text diachronically, following the hypothetical stratification being proposed by its editors (Goodall et al., forthcoming) – namely, that the scripture’s five books were composed in the following chronological order: the Mu-lasu-tra, Uttarasu-tra, and then Nayasu-tra, followed by the Guhyasu-tra and Mukha-gama, the whole presumably being complete before the end of the seventh century. In the present essay I for the most part omit discussion of the Nis´va-saka-rika-, a large and poorly transmitted supplement to the Nis´va-sa which might belong to a somewhat later period.

S´akti in early Tantric S´aivism

17

Maha-devıI begin from the observation that the Nis´va-sa places little cultic emphasis upon goddesses; nonetheless, it contains much of interest for the history of S´a-kta traditions. In describing laukikadharma, the non-initiatory religion of the laity (and not specifically S´aiva laity), the Nis´va-sa’s Mukha-gama (3.107c–11) briefly describes worship of the ‘Great Goddess’ (maha-devya-s tu pu-janam, 107d). Present here is an unmistakable and perhaps comparatively early articulation of the idea of the singular Maha-devı-, with the spouse of S´iva heading the following list of divine names and epithets: Uma-, Ka-tya-yanı-, Durga-, Rudra-, Subhadrika-, Ka-lara-tri, Maha-gaurı-, Revatı-, Bhu-tana-yika-, Arya-, Prakr.tiru-pa- (‘She who - takes the form of Prakr.ti’), and Gan.ana-yika- (gan.a-na-m . nayika, ‘Leader of S´iva’s Troops’) (108c–9). The goddesses identified with Parvatı- or Uma- span sectarian allegiances and might suggest a Gupta-era or late-Gupta milieu, potentially comparable to that of the old Skandapura-n.a (circa sixth to seventh - centuries), or possibly even the Harivam . s´a. Arya and Revatı, for instance, would seem unlikely inclusions in a medieval list of the Goddess’s principal identities.5 Noteworthy is the identification of the Maha-devı- with prakr.ti, the cosmogonic principle of Sa-m . khya, as well as the absence of explicit identification with s´akti and ma-ya-. These are the three cosmogonic, feminine-gender principles Pintchman (1994, 3–5, etc.) identifies as coming together in the Pura-n.as to form the ‘symbolic complex’ of the Great Goddess. While it is difficult to draw conclusions from so short a passage, the absence of ma-ya- and s´akti may suggest that the Mukha-gama reflects a relatively early stage in the formation of the Great Goddess, perhaps similar to that of the Skandapura-n.a. The latter attests the idea of the Maha-devı- emanating other goddesses, as well as her identity with prakr.ti (13.24, 42cd– 44ab);6 her identification with the cosmogonic principles of s´akti and ma-ya- is also intimated, but only in passing, receiving little emphasis or elaboration.7 The three cosmogonic principles come together emphatically in the subsequent Devı-ma-ha-tmya of the Ma-rkan.d.eyapura-n.a,8 which is explicit in advancing what we might call a pura-n.ic S´a-kta theology on this basis. The Mukha-gama’s eulogy of the Maha-devı-, in the context of lay religion, is complemented by a Mantrama-rga-oriented formulation of the Great Goddess found in a dialogue between the divine couple in the Guhyasu-tra (8.128–40), a late stratum of the text. There, rather than being eulogized as the unitary Goddess encompassing the feminine divinities of popular religion, S´iva praises her as mother of the universe, and as the apotheosis of all manner of female cosmological principles: the ontic levels (tattva) designated by feminine-gender names, such as ma-ya-. Similarly, S´iva is identified as father of the worlds and as all male-gendered facets of divinity. The categories invoked are primarily those of Mantrama-rga cosmology, rather than the deities of popular religion. However, the technical terminology of cosmology also gives way to Pura-n.alike devotional praise of the divine couple as universal progenitors: all that exists originates in S´iva’s seed, emerging from the womb of the Goddess;9 all

18

Shaman Hatley

that comes into being represents various transformations (vika-ra) of her.10 Viewing these formulations in tandem, we find in the Nis´va-sa relatively early and complementary articulations of the idea of the Great Goddess in the intersecting spheres of laukikadharma and the Mantrama-rga, i.e. tantric S´aivism.

S´akti, cosmology and theology Viewing the Nis´va-sa corpus according to the stratification proposed by its editors, it appears evident that the nature and role of S´iva’s feminized power or powers evolves significantly with the text’s historical development. The contrast is marked between the earliest stratum (the Mu-lasu-tra) and latest strata (Guhyasu-tra, then Nis´a-vaka-rika-); and it seems that some distinctions are discernible between these and the middle strata as well (Uttarasu-tra and Nayasu-tra). Present in the Uttarasu-tra and Nayasu-tra, but not clearly evident in the earlier Mu-lasu-tra, are several key elements of the theology of s´akti familiar from later sources, such as the cosmogonic function of the supreme S´akti and the role of the ‘descent of S´iva’s power’ (s´aktipa-ta) in grace and initiation. The most fascinating material is present in the Nis´va-sa’s treatment of cosmology. By ‘cosmology’ I refer to both cosmogony and cosmography – subjects expounded in a variety of contexts, especially those of initiation – and to the origins of mantra, scripture, and language. The cosmology of the Nis´va-sa has been the subject of groundbreaking studies by Sanderson (2006) and Goodall (forthcoming), who demonstrate that the universe of the Nis´va-sa’s Mu-lasu-tra consists of an ascending hierarchy of worlds (bhuvanas), rather than the ontic levels or tattvas which come to characterize more developed systems.11 In the Mu-lasu-tra, s´akti is conspicuously absent, both as a level of the cosmos and as a cosmogonic power. The closest one comes to a feminine creative principle is Va-gı-s´varı-, Goddess of Speech, discussed below. The conception of s´akti as S´iva’s cosmogonic power is, however, evidenced in the middle and later strata of the Nis´va-sa. The Uttarasu-tra (1.5–6) articulates the notion that S´iva’s s´akti, conjoined with his ‘energy’ (tejas12), gives rise to bindu, the primordial creative ‘drop’ or ‘point’.13 This appears to function as material cause for the emanation of a descending series of ontic levels (tattva): Is´vara, Vidya-, Ma-ya-, Ka-la, Niyati,14 Purus.a, Prakr.ti, and the material creation. Nayasu-tra, chapters 3 and 4, describes meditation upon the tattva-series as deities, a subject receiving further elaboration in the Guhyasu-tra (Chapter 8); this might represent an historical development vis-à-vis the Uttarasu-tra, for in the latter’s exposition of meditation upon the tattvas, their deification is not explicit.15 In contrast, the yoga of the Nayasu-tra involves visualization of the tattvas as deities with anthropomorphic qualities who embody the powers associated with their ontic levels (3.20ff). Tattvas with feminine-gender names are represented as goddesses, comprising Prakr.ti, Niyati, Ma-ya-, Vidya-, and the transcendent S´akti. The visualization of Ma-ya- (3.34c–35) provides an early example of a

S´akti in early Tantric S´aivism

19

‘fierce’ tantric goddess – a deity ‘black in colour, red-eyed, with long teeth, very hairy, with tawny hair … coarse-bodied, big-bellied, [she is the one] who causes [all] creatures, from Brahma- downwards, to fall again and again’.16 Goodall (forthcoming) observes that Ma-ya- is not here identified as having cosmogonic functions, as she is in later, developed systems of S´aiva cosmology. The Nayasu-tra’s meditation upon s´akti (4.7–39) portrays S´iva’s power as an ontic level possessing cosmogonic and soteriological functions. She is salvific, for through mastery over the s´aktitattva one ascends to realization of the highest S´iva (4.34). The ‘Great Power’ (maha-s´akti) transcends Is´vara and Sada-s´iva. She is fourfold, her divisions being the ‘subtle, extremely subtle, immortal, and immortal of immortals’ (su-ks.ma-, susu-ks.ma-, amr.ta-, and amr.ta-mr.ta-) (4.7cd–9ab). This tetrad mirrors but transcends that of the four kala-s of the Lord (Is´vara), who is identified with the primordial, cosmogonic ‘drop’ (bindu). Goodall draws attention to the fact that the Nis´va-sa’s presentation of the kala-s or fourfold powers of the Lord (and/or of bindu) is archaic, lacking the fifth member, s´a-ntyatı-ta, almost ubiquitous in the later tradition, including the Nis´va-saka-rika-.17 Nonetheless, the s´aktitattva’s tetrad of aspects has continuity with the later S´a-kta cosmology of the Brahmaya-mala, which develops this theme by situating various sets of four kala-s in its hierarchy of ontic levels.18 In the middle stratum of the Nis´va-sa, represented by the Nayasu-tra, are present key aspects of the Mantrama-rga vision of s´akti’s role in grace and initiation. Though conceived of variously in the tradition, the basic premise is that S´iva’s grace (anugraha) operates through his s´akti, and manifests in or as tantric initiation (dı-ks.a-). Chapter 1 of the Nayasu-tra introduces the notion of the ‘descent of S´iva’s power’ (s´aktipa-ta): s´ivas´aktinipa-tena dı-ks.a-[jña-]nam . prayacchati ||88|| - 19]ta- caiva sada-s´ivah || so ’nugrahah. smr.to [hy evam da . . ‘Through the descent of S´iva’s power, he bestows initiation and the scriptural wisdom. It is this which is known as “grace”, and its giver is Sada-s´iva.’ The nexus of s´akti, grace, and initiation extends to the officiant as well; it is yogic mastery of the s´aktitattva that allows a guru to give salvific initiation (4.27cd–29, 38–40). The guru (des´ika) whose initiation is efficacious ‘brims with S´iva’s s´akti’ (s´ivas´aktyopabr.m . hitah.), and is capable of transmitting grace through his mere gaze, speech, or touch.20 While aspects of this nexus of ideas may have pre-tantric, Atima-rga roots (Sanderson 2006: 191–92), its formulation in the Nayasu-tra clarifies that it belongs to an early, though probably not initial, phase of the Mantrama-rga. Beyond deified tattvas, several other divinities the Nis´va-sa treats as goddesses also represent apotheoses of feminine principles. The Guhyasu-tra (7.293–98) provides an intriguing prescription for meditation upon Sus.umnaas a cosmological goddess, potentially, but not necessarily a deification of the central channel of the yogic body.21 Of much interest is the goddess of speech,

20 Shaman Hatley Va-gı-s´varı-, who, as mentioned, appears in the archaic bhuvana-based cosmology of the Mu-lasu-tra. There (5.13) she represents the level of the cosmos just prior to the pran.ava or om . kara. The sequence of ascending levels here mirrors and is presumably based on an Atima-rga cosmological system presented in the Mukha-gama – that of the La-kula sect (Sanderson 2006: 163ff). In this, the speech-goddess appears as the first level of the ‘pure universe’ (s´uddhama-rga), 22 and the ‘source’ (yoni) of the om . kara above it. This suggests a parallel with more developed Mantrama-rga cosmological thought, insofar as the tattva known as s´uddhavidya- (‘pure knowledge’), the first level of the ‘pure universe’ (s´uddha-dhvan), is sometimes viewed as the source of the ontic levels arrayed above it. On this subject Goodall (forthcoming) remarks, ‘What is odd about that conception is that it [s´uddhavidya-] is placed at the bottom of the pure universe, and with its supposed evolutes therefore ranged “above” it, quite unlike the evolutes of prakr.ti, which are ranged “beneath” it.’ The source of this oddity may lie in the fact that s´uddhavidya- inherits the cosmogonic function of the Atima-rga’s Va-gı-s´varı-, a doctrine complexified by the Mantrama-rga addition of bindu, which is also viewed as the material cause for tattvas of the pure universe – but from ‘above’ rather than below. Concerning speech, it is also notable that s´akti has no generative role in the Uttarasu-tra’s description of the emergence of scripture (s´ivatantrotpatti, 1.22ff). Scripture, in the form of the ‘primordial sonic resonance’ (na-da), is said to emerge from the quiescent S´iva (Uttarasu-tra 1.23), without further elaboration; or is said to emerge from the alphabetical matrix, the ma-tr.ka- (1.41).23 Like a number of other cosmogonic categories, the ma-tr.ka- is deified – not, however, as a goddess, despite the gender of the term, but rather as Ma-tr.ka-s´iva (5.27).24

S´akti sets The Nis´va-sa attests a number of sets of multiple feminine principles or powers familiar from the later tradition; however, there are archaic aspects to some of these. Early and late strata of the Nis´va-sa both attest a triad of s´aktis almost ubiquitous in later S´aivism: that of Va-ma-, Jyes.t.ha-, and Raudrı-. This is present in the Mu-lasu-tra (5.3–4), perhaps the oldest textual stratum, where the initiation of the ‘liberation seeker’ (muktidı-ks.a-) involves linking the initiate’s soul to the ‘base of the cosmos’ (ka-la-gnirudra) using the Va-ma-s´akti (va-maya-), purifying it with Jyes.t.ha-, and raising it upwards to the ‘hell-worlds’ (naraka) via the Raudrı- s´akti. Here mantras of the three are not explicated, but probably implied.25 The Guhyasu-tra (7.259cd–60ab) situates the three s´aktis at the level of Is´vara in its presentation of the ascending hierarchy of ontic levels; this apparently elaborates on a passage in the Mu-lasu-tra (5.16) where their presence is implied. The triad becomes ubiquitous in the Nis´va-saka-rika-, for instance homologized with the three central channels of the yogic body and other threefold categories.26 Also present in all layers of the Nis´va-sa is the extended, ninefold form of the s´akti triad, beginning with Va-ma- and ending with Manonmanı-. This

S´akti in early Tantric S´aivism

21

series comprises female counterparts of the vidyes´as, deities whose names are abstracted from the archaic mantra of Va-madeva (Goodall 2004: 181). As presented in the Mu-lasu-tra, the nine s´aktis seem to function in a manner analogous to later traditions; installed on a lotus of eight petals upon Sada-s´iva’s throne (a-sana), they form the ‘inner layer’ (a-varan.a) of his man.d.ala (cf. Goodall et al. 2005, Fig. 4).27 In Guhyasu-tra 7.252, they appear to be divisions of the tattva-goddess Maha-vidya-.28 The s´akti triad and its ninefold form hence are present at the earliest levels of the Mantrama-rga, to which it appears distinctive; I am unaware of any links to the pre-tantric Atima-rga. In contrast, the triad of S´iva’s powers of knowledge, action, and volition – jña-nas´akti, kriya-s´akti, and iccha-s´akti – seems absent as such from the Nis´va-sa’s early and middle textual strata. This is also true of other early Siddha-nta Tantras.29 More striking is the fact that the very idea of S´iva’s power of volition or will, iccha-s´akti, seems present only in the Guhyasu-tra (8.79a), where there is no indication of it being grouped with jña-nas´akti and kriya-s´akti. Rather, jña-nas´akti and kriya-s´akti find mention as a pair, both in the Mu-lasu-tra (5.15) and Guhyasu-tra (7.260cd, and probably 8.65b); in both cases they are referred to as vidya-s (dve vidye) and are enumerated immediately after the triad of va-ma-, jyes.t.ha-, and raudrı-, as though forming a pentad. In the cosmology of the Guhyasu-tra, however, S´iva’s volition emerges as an additional s´akti. A lacunose passage appears to refer to its veneration in a description of an ascending series of cosmological deities (8.79ab); and at the close of Guhyasu-tra, chapter 8, in the Pura-n.a-style dialogue referred to earlier, S´iva describes the Goddess as mother of the worlds, and identifies her with a series of feminine-gender cosmological principles that includes his volition.30 Only in the Nis´va-saka-rika- are kriya-, jña-na, and iccha- spoken of as a triad of S´iva’s powers,31 distinct from and perhaps mirroring the triad of va-ma-, jyes.t.ha-, and raudrı-. The position of iccha-s´akti as a late entrant into the cosmology of the Nis´vasa seems to confirm Hélène Brunner’s suggestion (1992: 1–7) that the pair jña-nas´akti and kriya-s´akti underlies later groupings of three or more. Indeed, S´iva’s dual powers of knowledge and action are a topic emphasized in Pa-s´upata literature,32 where, it may be noted, the ‘femaleness’ of S´iva’s s´akti seems confined wholly to grammatical gender. Identification of S´iva’s volition as a s´akti may have developed to explain how the quiescent, desireless S´iva should stir into action at all, echoing the Upanis.adic idea of a primeval divine aspiration which sets creation in motion. This is conceptually distinct from the cosmogony intimated in passages from the Uttara- and Nayasu-tra, which emphasize the desireless (nis.ka-ma) nature of S´iva and S´akti, whose conjunction nonetheless results in the origin of bindu. The metaphor invoked is that of the solar S´iva’s rays shining upon the sunstone of S´akti, giving rise to the appearance of fire.33 S´akti produces bindu when conjoined with or possessing S´iva’s tejas – his radiance, or perhaps generative power34 – a potentially problematic third cosmogonic factor. In contract, and in a less philosophical register, the Guhyasu-tra invokes the image of divine progenitors when it speaks of the Goddess as the ever-pregnant creatrix and S´iva as her lover (ka-mayitr.).35

22

Shaman Hatley

Mother-goddesses I turn now to the Mother-goddesses (ma-tr.), who appear to have been central to the early development of S´a-kta-oriented cults. This is evident from early vidya-pı-t.ha scriptures, which categorize goddesses and sometimes practitioners into clans headed by the Mothers, incorporate them as cult deities, or in other ways evidence their historical primacy (Hatley 2012: 107–19). It is also the cult of Mothers alone among goddesses that finds mention in Vara-hamihira’s list of major sectarian groups of the sixth century.36 The Nis´va-sa describes ma-tr.s as deities primarily of public, lay religion (laukikadharma) – not as tantric deities possessing mantric identities. The Mukha-gama (2.28cd–30ab) lists the ma-tr.s among cult deities of temples,37 and refers to devotees of and places sacred to the Mother-goddesses, among many other divine and semi-divine beings.38 There is nothing to confirm whether the Nis´va-sa knew of the brahmanical Seven Mothers (sapta ma-tarah.), important cult deities from at least the fifth century; the Nis´va-sa’s nameless ma-tr.s could instead resemble the multitudinous and diverse Mothers attested in the Maha-bha-rata, Kus.a-n.a-era statuary, and medical literature. (These are attested alongside the Seven Mothers in the old Skandapura-n.a.) As in other works of tantric literature, temples of the Mothergoddesses are listed among the places appropriate for performing solitary tantric ritual, along with crossroads, jungles, mountain peaks, S´iva temples, and so forth.39 It appears to be the liminality of the ma-tr. temple – presumably a secluded shrine rather than public monument – that makes it suitable for the rituals envisioned, for these do not directly involve their worship. Some passages may refer to individual goddesses associated with ma-tr.s in ways potentially suggesting their incorporation as tantric deities. In a fascinating passage of the Uttarasu-tra (1.31–38), Ka-lı-, and possibly the goddess Vijaya-, are listed among the interlocutors involved in the revelation of tantric literature,40 alongside a variety of gods, gan.as, sages, bhairavas, unnamed ma-tr.s, yaks.as (guhyaka), and other divinities. This suggestive list, and vague reference to scripture’s manifoldness, may point towards the existence of diverse tantric literatures in the period of the Uttarasu-tra’s composition, beyond the canon of twenty-eight texts it delineates – texts that would come to be viewed as the canon of Siddha-nta Tantras.41 The presence of bhairavas,42 ma-tr.s, and so forth among the interlocutors may raise the possibility that there existed texts belonging to or prefiguring scriptural genres such as the Bhairava Tantras, and those - dedicated to goddess cults (Va-ma- or Bhaginı- Tantras, Ma-tr. Tantras, D . akinı Tantras); Sanderson (2001, 2009) adduces evidence pointing toward the existence of such literature in or prior to the seventh century. In the Guhyasu-tra alone do we find allusion to the fierce Can.d.ika-, whose identity frequently intersects with that of Ca-mun.d.a-, leader of the Seven Mothers: in passing, the text provides an otherwise anomalous ‘mantra of Can.d.-ı’ (can.d.imantra) and its application, alluding to a tradition of magical ritual centered upon this deity.43 The Guhyasu-tra also attests the incorporation of ma-tr.s in its exposition of cosmology, providing what may be a significant link to later S´a-kta-oriented

S´akti in early Tantric S´aivism

23

traditions. Chapter 5 lists ma-tr.s among the presiding deities (patayah.) of a series of seven pa-ta-las, ‘netherworlds’, together with a variety of other divinities, including rudras, gan.as, na-gas, ra-ks.asas, bhaginı-s, and yogakanya-s.44 None of these are presented as tantric cult deities with mantras. By and large, the Guhyasu-tra’s higher cosmological spheres are populated with male mantra-lords (mantres´vara, etc.) and manifold rudras. While in later S´a-kta accounts of the cosmos, goddesses would largely eclipse male deities, in the Guhyasu-tra the goddesses of the netherworlds are exceptional. In the lowest pa-ta-la are present ‘groups of Mothers’ (ma-tr.gan.a-h.) and ‘Sisters’ (bhaginyah.).45 These could potentially refer to the most famous representatives of the classes, namely the Seven Mothers and the Four Sisters of Tumburu; however, most probably are intended to be amorphous bands of minor female deities. Two other sets of . Mother goddess find mention: ‘tawny’ (pingala) Mothers who bear blue lotuses are present in the third netherworld, while in the fourth preside kapa-lama-tr.s, ‘Skull Mothers’. The latter, who have parallel in the kapa-larudras of the next higher cosmological sphere, appear to represent a transformation of the Mothers into deities whose ka-pa-lika iconography presages that of the S´a-kta Vidya-pı-t.ha’s cult goddesses.46 Positioned higher in the series are yogakanya-s, ‘yoga maidens’, deities of the sixth and seventh netherworlds. Here described merely as ‘possessing great power’ (maha-vı-rya-h.),47 goddesses of this particular collocation are not wellattested in S´aiva textual sources familiar to me. Yet as powerful, youthful goddesses connected with yoga who transcend the Skull Mothers, these ‘yoga maidens’ appear to prefigure the deities later referred to as yoges´varı-s, ‘female masters of yoga’, or yoginı-s. This connection is drawn much later by a Kashmirian, Ks.emara-ja, in commenting upon a parallel passage in the Svacchandatantra; he understands yogakanya-s as yoginı-s who possess their yogic powers from the very moment of birth.48 This seems consistent with the term’s use in the Harivam . s´a, where it describes the goddess Ekanam . s´a as the 49 infant girl substituted for Kr.s.n.a. It hence appears that the Guhyasu-tra describes a hierarchy of divinities encompassing goddesses with typological similarities to those later brought within the rubric of the yoginı-: multiple categories of Mother-goddess, the Sisters, and maiden goddesses possessed of yogic powers.50

Ritual Although the cosmology of the Guhyasu-tra only faintly intimates developments suggestive of the formation of the S´a-kta Vidya-pı-t.ha, its ritual offers substantial material for comparison. This late stratum of the Nis´va-sa is rich in siddhi-oriented practices that foreshadow varieties of ritual elaborated upon in the Bhairavatantras. Particularly noteworthy are its mortuary (ka-pa-lika) practices, the use of magical substances (siddhadravyas), and coital ritual. The ka-pa-lika rites of the Guhyasu-tra appear largely magical in orientation, which places them in much closer relation to those of the Mantrama-rga’s

24

Shaman Hatley

Bhairavatantras than the liberationist ka-pa-lika ritual of the Atima-rga. This is particularly evident in the Guhyasu-tra’s prescriptions for preparing magical substances in a skull; in one case the end-product is ash that would turn the practitioner into a vidya-dhara (celestial wizard), and in another an eye-ointment that induces invisibility. Similar procedures are common in vidya-pı-t.ha sources, where, however, more marked use of the ‘impure’ is made.51 Perhaps the most extreme of the ka-pa-lika magical practices taught in the Guhyasu-tra is chapter three’s rite of fire sacrifice, performed in the mouth of a corpse, which in structure and aims parallels veta-lasa-dhana as described in the Brahmaya-mala and Hars.acarita.52 As in the Brahmaya-mala, the rite culminates with the corpse’s tongue emerging, which when severed becomes a magical sword.53 Elsewhere, the Guhyasu-tra describes rituals for magically enlivening a corpse that does one’s bidding.54 Such magic was undoubtedly ancient, described for instance in the Jaina Prakrit Vasudevahin.d.i (Dezso˝ 2010: 399). As for sexual ritual, which is characteristic of the early S´a-kta Vidya-pı-t.ha, this finds some precedent in the Guhyasu-tra, which teaches a tantric version of the asidha-ra-vrata (‘the razor’s edge observance’; 3.38cd–43ab). This involves the participation of a beautiful young woman skilled in the erotic arts, although one who succumbs to lust in her embrace falls into hell. What distinguishes the Guhyasu-tra’s inflection of this brahmanical ritual is a shift from ascetic to magical ends, which places it in continuity with the coital rituals of vidya-pı-t.ha practice systems. I have elsewhere sought to trace the development of the asidha-ra-vrata, and suggested that it may have been adopted from an Atima-rga source (Hatley, forthcoming). Absent from the coital rite of the Guhyasu-tra is any indication of interest in manipulating the dangerous potency of the ‘secret nectars’ (guhya-mr.ta), i.e. sexual fluids, a subject of considerable fascination in vidya-pı-t.ha sources. However, besides metal ores, neem oil, and the like, the Guhyasu-tra’s ritual occasionally taps the power of conventionally impure substances, such as blood and beef, and in one instance menstrual blood.55 In the Guhyasu-tra, some glimpses may be had of the cult of spirits prominent in Tantras of the Bhu-ta- and Bhairava- scriptural streams. In an example from Chapter 11, one who fasts, smears the body with crematory ash, and performs twelve-lakh repetitions of the mantra OM.ANA-THA-YANAMAH. has the dars´an of ‘spirits’ (bhu-ta), who offer him ‘magical substances’ (siddhadravya) that induce invisibility.56 Such magical, transactional experiences are greatly elaborated upon in the vidya-pı-t.ha yoginı- cults, wherein ‘visionary encounters’ (mela-pa) with the goddesses become central to the aims of ritual. Also noteworthy is the Guhyasu-tra’s prescription for gaining the aid of a yaks.in.-ı (‘dryad’). Further reference is made to achieving power over such female spirits as the bhu-tı-, pis´a-cinı-, and na-ginı-, although generally, feminine-gender spirits are little emphasized.57 Erotic magic is present, such as a rite in the Guhyasu-tra wherein one magically transforms a female ‘goat or sheep’ (aja-) into a woman who fulfills ‘all of one’s desires’.58 Also noteworthy are the numerous references to joining the ranks of the vidya-dharas, suggesting that even at this level of the tradition, magical perfection and the attainment of embodied divinity

S´akti in early Tantric S´aivism

25

had emerged as well-defined aims of ritual. In general character, the Guhyasu-tra thus suggests the extent to which the ritual of the Bhairava Tantras and early S´a-kta cults had deep roots in earlier tradition, representing a shift in emphasis rather than something altogether novel.

Conclusion A number of historical observations emerge from the preceding analysis of the Nis´va-sa corpus. First, it is evident that a theology emphasizing S´iva’s cosmogonic and grace-bestowing ‘power’ (s´akti) comes into view in the middle stratum of the Nis´va-sa corpus. This nexus of ideas receives further elaboration in the Guhyasu-tra, which, for instance, incorporates the idea of S´iva’s ‘power of volition’ (iccha-s´akti), accompanied by a shift in cosmogonic imagery. In general terms, a disproportionate amount of the material concerning s´aktis and goddesses belongs to the Guhyasu-tra, a comparatively late stratum of the text. Diachronic analysis of the Nis´va-sa corpus suggests that early tantric goddesses were frequently apotheoses of feminine-gender cosmological categories; a movement from feminine-gendered principles to embodied goddesses may be characteristic of the early Mantrama-rga. The equation of goddesses with s´akti, and the ‘femaleness’ of s´akti, appear to be innovations of the early Mantrama-rga, aspects of the early development of which are visible in the Nis´va-sa itself. Furthermore, the conflation of prakr.ti and s´akti, which appears critical to the Pura-n.ic construction of the Maha-devı-, becomes evident in the cosmology of the Nis´va-sa in the deification of prakr.ti and in the subsumption of all such cosmological goddesses within s´akti. It is also evident that many of the ritual forms and concerns of the early S´a-kta Vidya-pı-t.ha are presaged in the Guhyasu-tra of the Nis´va-sa.

Notes 1 I would like to thank Dominic Goodall, Harunaga Isaacson, and Alexis Sanderson for their comments on the version of this essay presented at the conference on S´a-kta traditions held at the Oxford Centre for Hindu Studies, in September 2011. 2 This is argued in some detail by Goodall et al. (forthcoming, Introduction). 3 Concerning texts of the Vidya-pı-t.ha, see Sanderson (1988: 668–79, 2009: 19–20, 45–49). 4 ‘Early Tantra: Discovering the Interrelationships and Common Ritual Syntax of the S´aiva, Buddhist, Vais.n.ava and Saura Traditions’, directed by Harunaga Isaacson and Dominic Goodall, and funded by the Agence Nationale pour la Recherche (France) and the Deutsche Forschungsgemeinschaft. 5 For several early references to Arya-, see White (2003: 39–40). She is also mentioned in Skandapuran.a 31.106a. 6 This model is evident in Pa-rvatı-’s emanation of Kaus´ikı-/Vindhyava-sinı-, and Kaus´ikı-’s of the Mother-goddesses; see the discussion of Yokochi (2004: 49–55, 99–113). On Pa-rvatı-’s identification with prakr.ti, note especially 13.24, and 13.43–44ab (iyam . ca - - prakr.tir devı- sada- te sr.s.t.ika-ran.am | patnı-ru-pam . samasthaya jagatkaran.am agata || namas tubhyam . sades´ana devyas´ caiva sada namah. |). 7 The Mantramarga theology of the cosmogonic s´akti seems only faintly alluded to in the Skandapura-n.a; the Maha-devı-’s identification with S´iva’s ‘power’ (s´akti) is

26

8 9 10 11

12 13 14 15 16 17 18 19 20 21

22 23 24

25 26 27

28

Shaman Hatley mentioned in Skandapura-n.a, Chapter 32 (para-m . ca pravr.ttim . mahes´asya s´aktim . ). Cf. the weaker case of Skandapura-n.a 3.14ab, which (in passing) refers to S´iva as ‘singular, [though] conjoined with the s´akti, and not devoid of s´akti’ (ekasmai s´aktiyukta-ya as´aktirahita-ya ca). As for the identification of Pa-rvatı- with the creative, magical power known as ma-ya-, note especially Skandapura-n.a 60.41c. This eulogy of Pa-rvatı- (vv. 38–41), spoken by Vis.n.u, is remarkably s´a-kta in character, speaking of the Goddess as creator and destroyer, as prakr.ti, bearer of the world, and maha-ma-ya-, among other divine epithets. On the dating of the Devı-ma-ha-tmya, see Yokochi (2004: 21–23). Guhyasu-tra 8.129: a- brahmastambaparyantañ jagat sarvañ cara-caram | mama bıjodbhavam . sarvam . tava yonivinirgatam || Guhyasu-tra 8.133cd: vika-ra-s tava ja-yante yat sarvam . sampratis.t.hitam. Goodall (forthcoming) remarks, ‘It is only in the textual layers later than the Mu-lasu-tra that we find the tattvas presented, sometimes to homologise them with mantra-elements, sometimes to serve as the focal point for yogic meditation and, . finally, in the case of the higher (non-Sa-nkhya) tattvas, as layers in the upper reaches of the cosmos.’ See the discussion below. Uttarasu-tra 1.5cd–6ab: tasya s´aktih. s´iva- nitya- s´ivatejopabr.m . hita | s´ivatejena - s´akter ja-yati bindukah. sam yukta . . The tattva-goddesses Ma-ya-, Niyati, and Vidya- are discussed by Goodall (forthcoming). Such deification might, however, be implicit, as suggested by the instruction to make an effigy of ma-ya- (Uttarasu-tra 5.25). Translation of Goodall et al. (forthcoming). See kala- (definition 6) in Ta-ntrika-bhidha-nakos´a, vol. 2; and Goodall et al. (forthcoming). While the Brahmaya-mala attests the fivefold form of the kala-s, its cosmology places importance upon sets of four, as presented in pat.ala 32, on the subject of ‘cosmography’ (prakriya-). The editors note that the old palm-leaf manuscript (siglum N) is not fully legible here; the bracketed syllables are transmitted by its two apographs. Nayasu-tra 4.40: caks.us.a- va-caya- spars´a- manasa- -d- udakena va- | dı-ks.ayet sarvajantu-ni s´ivas´aktyopabr.m . hitah. ||. Sus.umna- is said to be white, with the shape of a lotus stalk, and to ‘emerge from the body of S´iva’ (padmasu-tra-kr.tih. s´ukla-, 294c; s´ivaka-ya-d vinih.sr.ta-, 297b). What precisely sus.umna- represents in the Guhyasu-tra is unclear, for as Goodall et al. (forthcoming) point out, it is not yet the ‘central channel’ (na-d.-ı) in the yoga of the Uttarasu-tra. Mukha-gama 4.95ab and 4.127ab; cf. Guhyasu-tra 8.50cd. Uttarasu-tra 1.41: s´a-strakot.isahasra-n.i mantrakot.ya- hy anekas´ah. | ma-tr.ka-sambhava-h. sarve na-sti mantram atah. param ||. In contrast, Guhyasu-tra 8.128 might identify the alphabet with both S´iva and the God- - dess: aham . tvañ ca vis´alaks.i mulam . sarvajagasya tu | matr.kas´ivarupen.a siddhisambhavapadapam || – perhaps ‘You and I, O woman of long eyes, are the root of the entire world. As the ma-tr.ka-s´iva, [we are] the tree giving rise to [the fruit of] the perfections’. See discussion of Goodall et al. (forthcoming) in the annotation thereon. See e.g. Nis´va-saka-rika-, IFP Transcript no. 17, 40.6: id.a-ya-m samsthita- va-ma. .. - . va-ma-nge sarvadehina-m | sus.umna-ya-m . punar jyes.t.ha raudrı tu pingalas´rita || . Describing the basic man.d.ala for ‘worship of S´iva’ (s´iva-rcanavidhi, 2.1b), Mu-lasu-tra, ch. 2 enjoins installing the Nine s´aktis upon a white lotus (tasyopari sitam . padmam . navas´aktisamanvitam, 2.2cd). Cf. Guhyasutra 1.108, and 8.65a, which refers to ‘the ninefold s´akti pantheon/worship’ (navadha- s´aktiya-gam . ). See the discussion ad loc. by Goodall et al. (forthcoming).

S´akti in early Tantric S´aivism 29 30 31 32 33 34

35 36 37 38 39 40

41 42 43 44 45 46

47

27

Goodall, personal communication. A relatively early example of kriya--, jña-na-, and iccha- as a triad of powers is provided by the Svacchandatantra (1.65–68, etc.). Guhyasu-tra 8.140ab: mama iccha- na hanya- tvam . tvam . hi s´aktibalodaya. See, for instance, IFP Transcript no. 127, p. 357: s´ivajña- tu maya-khya-ta- om . nameti triyaks.aram | jña-nas´akti[h.] kriya-s´aktir iccha-s´akti[s] tr.tı-yaka- | te yada- jña-payed eva[m . ] tada siddhyanti nanyatha |. See for instance the Pañca-rthabha-s.ya of Kaun.d.inya ad Pa-supatasu-tra 4.23. Uttarasu-tra 1.5–6, Nayasu-tra 2.6–9; note particularly 2.7–8: a-dityasya man.ir yadvat ta-pito raviras´mibhih. | vahnis sañja-yate tatra na rave tatra ka-mata- || man.er api na ka-mitvan tadvad devasya ces.t.itam | a-dityavac chivo jñeyah. s´aktir mman.ir iva sthita- ||. Uttarasu-tra 1.6ab: s´ivatejena sam . yukta s´akter jayati bindukah.. For tejas as a creative power, note Cha-ndogya Upanis.ad 6, 2.3, and cf. Couture’s (2001, 319) remarks on the term in the Harivam . s´a: ‘Tejas is, as it were, a divine seed (bıja), or a spark, a medium which allows the gods to descend upon the earth. The gods need only a ‘portion’ or a ‘fragment’ (am . s´a or bhaga) of their unlimited tejas in order to generate (sr.j, visr.j, utsr.j) duplicates of themselves and appear in the human world.’ . Guhyasu-tra 8.133ab: tvam . sada gurvin.ı devi ahan kamayita sada. - 59.19. Br.hatsam hita . - - ganes´am ma-taram ravim Mukha-gama 2.28cd–30ab: brahma-n.am . skam . dam . rudran.ım . . . - yum dharmmañ jales´varam | yo yasya stha-panan. ´ || vahnim s atakratum yaks am va . . . . kurya-t pra-sa-de tu sus´obhane || pu-jaye paraya- bhaktya- so ’mr.to hy asya lokata-m | Mukha-gama 3.33–34ab. Uttarasu-tra 2.4c and Guhyasu-tra 6.32. Cf., e.g., Siddhayoges´varı-mata 6.2–4, and Brahmaya-mala 14.11cd–13ab and 18.18–19. Uttarasu-tra 1.34d. The old MS. reads vijayo ka-lir eva ca; vijaya- seems more likely, however, for Ka-lı- (or a variant thereof) and Vijaya- (or Jaya-) occur together in a number of deity lists: e.g. Ma-linı-vijayottara 20.44a; Siddhayoges´varı-mata 20.42cd; and Bha-gavatapura-n.a X, 2.11cd. Concerning Ka-lı- or Maha-ka-lı-’s relationship to the ma-tr.s, note their association in the Kuma-rasambhava (VII.30–39); her presence (alongside Maha-ka-la) in the Ra-van.-ka--ka--ı shrine of the ma-tr.s at Ellora-; and e.g. Skandapura-n.a 68.9. For Vijaya- as a ma-tr., see e.g. Skandapura-n.a 64.26. She is better known as one the ‘Four Sisters’ (caturbhaginı-), cult deities of the ‘leftward stream’ (va-masrotas) of S´aiva revelation. . Note also reference to a derivative ‘compilation’ (sangraha) genre of scripture in . Uttarasutra 1.32a–c: yo ’nyah. parataro bhedo s´ivatantres.u pat.hyate | sangrahah. sa tu vijñeyo. The verse quarter 1.34c appears to read bhairava-s´ ca gan.amukhya-[h.], where, the editors point out, bhairava-h. could instead be an adjective describing gan.amukhya-h.. The passage giving the mantra is unfortunately lacunose; the old MS. (siglum N) reads, om . can.d.ike krama … can.d.imantro ’yam (3.25 ab). Guhyasu-tra 5.1–21. 5.6cd: anye ma-tr.gan.a- rudra- bhaginyas´ ca gan.a-s tatha-. . See Guhyasu-tra 5.9ab (ma-tara-[h.] pingala- yatra vasanty utpalahastika-h.), 5.11cd (kapalamataro yatra va … t.ayah. [lacuna; perhaps read vasanti s´atakot.ayah.]), and - s are 5.14ab (tatha- kapa-larudra-s´ ca asam . khyeyah. puna[h.] sthitah.). Kapalamatr .. also mentioned in Skandapura-n.a 7.23, and apparently the old S´ivadharmasangraha (Peter Bisschop, personal communication). yogakanyaka-s are mentioned in Guhyasu-tra 5.15d, in a description of the city Ratnavatı- of the sixth pa-ta-la; other denizens include atharvarudras (atharvarudra-s tatraiva vasante yogakanyaka-h.). ‘Yoga maidens possessed of great power’ (yogakanya- maha-vı-rya-s, 5.19a) are also met with in a description of the seventh pa-ta-la, along with ‘sons’ of Agni and Va-yu (agnikuma-raka-h., 18d; va-yukuma-raka-h., 19b). The fifth pa-ta-la contains agnikuma-rika-s, female counterparts of the ‘sons of Agni’. Given the similarity of the collocations, and their occurrence together, it is possible, if

28

Shaman Hatley

unlikely, that yogakanya- should be understood along the lines of agnikuma-rika-, as ‘maiden [born] of Yoga’. 48 Svacchandatantra (KSTS edition) 10.116cd–19ab, describing the abode of S´iva–Ha-t.akes´vara: - yadu-rdhve caiva sauvarn.am . patalam . parikırtitam | tatra vasaty asau devo hat.akah. parames´varah. ||116|| purakot.isahasrais tu samanta-t pariva-ritah. | siddhai rudragan.air divyair bhaginı-ma-tr.bhir vr.tah. ||117|| yoginı-yogakanya-bhı- rudrais´ caiva sakanyakaih. | siddhadravyasamair mantrais´ cinta-man.irasa-yanaih. ||118|| siddhavidya-samr.ddham . vai hat.akes´asya mandiram | Here yoginı-s are mentioned alongside siddhas, rudras, bhaginı-s, ma-tr.s, yogakanya-s, and perhaps rudrakanya-s, in the entourage of S´iva-Ha-t.akes´vara. Ks.emara-ja interprets the ‘yoga maidens’ (yogakanya-s) as a high grade of yoginı- (yoginyo yogena siddha-h., yogakanya-s tu ja-tama-tra- eva sam . smaritayogah., ‘yoginıs are perfected through [practice of] yoga; but yogakanya-s are caused to recall their yoga upon merely being born’). This passage in the Svacchandatantra is an elaboration upon Guhyasu-tra 5.16cd–17ab, where the deities mentioned are rudras, vidya-s, and vidyes´varas: - sauvarn.am . saptamam . jñeyam . patalam . nagasevitam | yatra citravatı- na-ma purı- rudrasama-kula- ||16|| tatra-sau ha-t.hako devo vidya-vidyes´varair vr.tah. |. 49 Harivam . s´a 96.14. See also the passages occurring in some manuscripts between 48.36–37 in the critical edition. Also, cf. the category of the ‘natural born’ (pra-kr.ta-) yoginı- in Brahmaya-mala, pat.alas 56 and 96, and similar notions in a variety of other sources. I am grateful to Harunaga Isaacson for his suggestions on this subject (personal communication). 50 Cf. Brahmaya-mala pat.alas 56 and 74, which contain detailed taxonomies of yoginı-s. Compare also the list of potentially malevolent divinities in Netratantra, ch. 2, the female ones among these being ‘hordes/groups of s´a-kinı-s and yoginı-s; the bhaginı-s, rudrama-tr.s, etc.; d.a-vı-s, d.a-marika-s, and ru-pika-s’ (°s´a-kinı-yoginı-gan.aih. | bhaginırudrama-tra-did.a-vı-d.a-marika-dibhih. || ru-pika-bhir … , 13b–14a). This list hence includes both Mother-goddesses and ‘Sisters’, the latter, according to Ks.emara-ja’s commentary, ‘arising from partial incarnations of [the Seven Mothers,] Bra-hmı-, etc.’ (brahmya-dyam . s´akottha bhaginyah.). 51 A process for producing invisibility-inducing kohl in a human skull is given in Guhyasu-tra 3.81–82. The recipe involves nothing more offensive than ghee. The process in 11.110 utilizes ash: om . yogadhipa namah. | anena mantren.a bhasma gr.hya kapa-lasamput.e stha-pya ta-vaj japed ya-vad †a-vartitamatih.† tatas tu tenoddhu-lane vidya-dharo bhavati | … (‘OM.YOGA-DHIPANAMAH. – taking hold of ashes using this mantra and placing them in the hollow of a skull, one should repeat the mantra until … (?). Then, when one is dusted with these ashes, he becomes a vidya-dhara … ’). Compare with Brahmaya-mala, pat.ala 50, which taps the powers of rather less innocuous substances: kros.t.hukasya tu pis´itam . haritalamanacchila | -krtva- tu pı-sayet ||8|| rocana- ca maha-ma-m sam ekı . . . . - kapa-lasam . put.am . kr.tva atmaraktena mis´ritam | sahasra-s.t.a-dhikam . japtva tr.s.kr.tva tilakam . kuru ||9|| bhavate bhu-tale siddho adres´yah. ka-lava-s´inah. | - 9c sahasra-s.t.a-dhikam . ] em. Isaacson; sahasrantadhikam . MS. 10b adres´yah.] em.; adres´yoh. MS.

S´akti in early Tantric S´aivism

52 53 54

55

56

57

58

29

‘One should mix together and make a paste of the flesh of a jackal, the harita-la and manah.s´ila- minerals, yellow pigment, and human flesh. After placing this in the hollow of a skull mixed with one’s own blood, and reciting the mantra one-thousand and eight times, make a bindi [with this] thrice. He becomes perfected on this [very] earth, invisible, having power over death.’ This brief Chapter, the kros.t.ukakalpa, has as its theme magic utilizing jackal (kros.t.uka) flesh. For a detailed and insightful study of veta-lasa-dhana, see Dezso˝ (2010). Guhyasu-tra 3.60cd–64ab, quoted by Goodall et al. (forthcoming, Introduction); this passage is also discussed by Dezso˝ (2010). Guhyasu-tra 11.86 describes magically enlivening an unmutilated corpse, which if male becomes a slave whom one may ride as a mount and go anywhere; if female, the corpse becomes like a celestial maiden, with whom one may live ten thousand years, invisible. The more elaborate rites described in 14.127–29 are performed in a cremation ground, and the enlivened corpse (veta-la) may be dispatched to do a particular task or fetch magical substances or treasure: tatas´ cottis.t.hati bruvate ca | bho vı-rapurus.a kim . karomıti | sa vaktavyah. | ıpsitam . kamam . dadasveti | tatah. - dayati | atha va-ñjanarocanamanahs´ila- hy ausadhiratnanidha-nam vasarvam sampa . . . . a-nayasveti | tatas tat karma kr.tva- tatraiva gatva- nipatati || (‘And then [the veta-la] arises and says, ‘O heroic man, what shall I do?’ He is to be told, ‘provide the [following] desired wish’. Then he accomplishes everything. Or [one should say,] ‘fetch ointment, yellow pigment, the manah.s´ila- mineral, or herbs, gems, or hidden treasure’. Then, after doing that work, [the corpse] goes there and falls [back] down’) (excerpt from 14.127). Note for instance that Guhyasu-tra 10.87 mentions homa using beef (goma-m . sa), while 14.66 describes smearing an ‘effigy’ (pratikr.ti) with blood as part of a ‘rite of subjugation’ (vas´-ıkaran.a). A magical recipe using ‘menstrual blood’ (na-rı-rajas) is attested in 3.54c–56b, quoted in Goodall et al. (forthcoming, Introduction). Guhyasu-tra 11.64: anena mantren.a s´mas´a-nabhasmana- sna-tva- nira-ha-ro dva-das´alaks.am . japet bhutagan.ani pas´yati [em.; pas´yanti Cod.] | siddhadravyan.i prayacchanti | taih. siddhadravyair antarhito bhavati || (‘Having bathed in ashes using this mantra, while fasting, one should repeat the mantra twelve hundred thousand times. He sees groups of spirits; they bestow magical substances. Through those magical substances, he becomes invisible’). Guhyasu-tra 10.81–84. This procedure, called yaks.in.-ıvidhi (yaks.in.ya- es.a vidhih.), involves worship of an image that comes to life when the rite is complete: siddha- sa- kim . karomıti bharya me bhavasveti | taya saha ramate yavad acandratarakam (‘when mastered, she [says] “what shall I do?” “Be my wife.” He enjoys himself with her for the duration of the moon and stars’). The subsequent verse (10.84) provides means for making a wife of a ‘snake goddess’ (na-ginı-). Cf. Guhyasu-tra 14.83, which describes rites for subjugating female spirits – the yaks.in.-ı, pis´a-cinı-, and bhu-tı-. ru-pavatı- strı- bhavati sa- sarvaka-ma-[n] dada-ti. The rite is described in Guhyasu-tra 14.153.

References Primary sources Cha-ndogya Upanis.ad. V. P. Limaye and R. D. Vadekar, eds. Eighteen Principal Upanisads, vol. 1. Poona: Vaidika Sam . s´odhana Man.d.ala, 1958. Nis´va-saka-rika-. Institut Français de Pondichéry Transcript nos. 17, 127, and 150. Electronic transcriptions by S. A. S. Sarma, Nibedita Rao, and R. Sathyanarayanan.

30

Shaman Hatley

Nis´va-satattvasam . hita. Electronic transcription and draft edition by Dominic Goodall,Peter Bisschop, Diwakar Acharya, and Nirajan Kafle, based on National Archives of Kathmandu ms. no. 1–227 (Nepal-German Manuscript Preservation Project reel A41/14), and two apographs: NAK 5–2406 (NGMPP reel A159/18), and Wellcome Institute Sanskrit ms. I.33. Nis´va-satattvasam . hita. See Goodall et al. (forthcoming), The Nis´vasatattvasam . hita: The ´ Earliest Surviving Saiva Tantra. Volume 1. A critical edition of the Mulasu-tra, Uttarasu-tra, and Nayasu-tra. Franco-German Early Tantra Series. Nis´va-satattvasam . hita. Nirajan Kafle, ed. Netratantra. Madhusu-dan Kaul Sha-strı-, ed. The Netra Tantram, with Commentary by Kshemara-ja. 2 vols. Kashmir Series of Texts and Studies, nos. 46, 61. Bombay: Research Department of Jammu and Kashmir State, 1926, 1939. Pa-s´upatasu-tra. R. Ananthakrishna Sastri, ed. Pa-s´upatasu-tras with Pañca-rthabha-s.ya of Kaun.d.inya. Trivandrum Sanskrit Series, no. 143. Trivandrum: Oriental Manuscripts Library of the University of Travancore, 1940. Br.hatsam . hita. Hendrik Kern, ed. The Br.hatsam . hita of Varaha-Mihara. 1864–1865. Reprint (7 fascs. in 1), Bibliotheca Indica, vol. 48. Osnabrück: Biblio Verlag, 1982. Brahmaya-mala. Palm-leaf manuscript: NAK 3–370 (ngmpp reel A42/6). Paper manuscripts: nak 5–1929 (ngmpp reel A165/14), nak 1–143 (ngmpp reel A166/1), nak 1–286 (ngmpp reel A165/13), and nak 6–2608 (ngmpp reel A1319/4). Brahmaya-mala. See Hatley (2007). Bha-gavatapura-n.a. J. L. Shastri, ed. Bha-gavata Pura-n.a of Kr.s.n.a Dvaipa-yana Vya-sa, with Sanskrit Commentary Bha-va-rthabodhinı- of S´rı-dhara Sva-min. Delhi: Motilal Banardidass, 1983. Ma-rkan.d.eyapura-n.a. K. M. Banerjea, ed. 1855–1862. Reprint, Bibliotecha Indica, vol. 29. Osnabrück: Biblio Verlag, 1980. Ma-linı-vijayottaratantra. Madhusu-dan Kaul S´a-strı-, ed. Kashmir Series of Texts and Studies, no. 37. Bombay: the Research Department of Jammu and Kashmir State, 1922. Siddhayoges´varı-mata. Judit Törzsök, ed. ‘‘The Doctrine of Magic Female Spirits’. A Critical Edition of Selected Chapters of the Siddhayoges´varı-mata(tantra) with Annotated Translation and Analysis’. PhD thesis, University of Oxford. Skandapura-n.a. Rob Adriaensen, Hans T. Bakker and Harunaga Isaacson. The Skandapura-n.a, Vol. I. Adhya-yas 1–25. Critically Edited with Prolegomena and English Synopsis. Supplement to Groningen Oriental Studies. Groningen: Egbert Forsten, 1998. Skandapura-n.a. Kr.s.n.aprasa-da Bhat.t.ara--ı, ed. Skandapura-n.asya Ambika-ka-n.d.ah.. Mahendraratnagranthama-la-, no. 2. Kathmandu, Nepal: 1988. Svacchandatantra. Madhusu-dan Kaul Sha-strı-, ed. The Svacchanda Tantram, with Commentary by Kshemara-ja. 6 vols. Kashmir Series of Texts and Studies, nos. 31, 38, 44, 48, 51 (vol. 5a), 51 (vol. 5b), 56. Bombay: the Research Department of Jammu and Kashmir State, 1921–1935. Harivam . s´a. Parashuram Lakshman Vaidya, ed. The Harivam . s´a, being the khila or supplement to the Maha-bha-rata: for the first time critically edited. 2 vols. Poona: Bhandarkar Oriental Research Institute, 1969–1971.

Secondary literature Brunner-Lachaux, Hélène. 1992. “Jña-na and Kriya-: Relation between Theory and Practice in the S´aiva-gamas”. In Ritual and Speculation in Early Tantrism. Studies in

S´akti in early Tantric S´aivism

31

Honor of André Padoux, ed. Teun Goudriaan, pp. 1–59. Albany, NY: State University of New York Press. Couture, André. 2001. ‘From Vis.n.u’s Deeds to Vis.n.u’s Play, or Observations on the Word Avata-ra as a Designation for the Manifestations of Vis.n.u’. Journal of Indian Philosophy, 29: 313–326. Dezso˝, Csaba. 2010. ‘Encounters with Veta-las: Studies on Fabulous Creatures I’. Acta Orientalia Academiae Scientiarum Hung, 63(4): 391–426. Goodall, Dominic. Forthcoming. ‘How the Tattvas of Tantric S´aivism Came to Be 36: the evidence of the Nis´va-satattvasam . hita’. Goodall, Dominic. 2004. The Parakhyatantra. A Scripture of the S´aiva Siddha-nta. Collection Indologie, no. 98. Pondicherry: Institut Français de Pondichéry/École française d’Extrême-Orient. Goodall, Dominic, et al. Forthcoming. The Nis´va-satattvasam . hita: The Earliest Sur´ viving Saiva Tantra. Volume 1. A critical edition of the Mulasutra, Uttarasu-tra, and Nayasu-tra. Franco-German Early Tantra Series. Goodall, Dominic, Nibedita, Rout, R. Sathyanarayanan, S.A.S. Sarma, T. Ganesan and S. Sambandhas´iva-ca-rya eds. 2005. The Pañca-varan.astava of Aghoras´iva-ca-rya: A Twelfth-century South Indian Prescription for the Visualization of Sada-s´iva and his Retinue. Collection Indologie, no. 102. Pondicherry: Institut français de Pondichéry/ École française d’Extrême-Orient. Hatley, Shaman. Forthcoming. ‘Erotic Asceticism: the Knife’s Edge Observance of Brahmaya-mala xl (the asidha-ravratapat.ala), with a critical edition and translation’. Tantric Studies, 3. Hatley, Shaman. 2012. “From Ma-tr. to Yoginı-: Continuity and Transformation in the South Asian Cults of the Mother Goddesses”. In Transformations and Transfer of Tantra in Asia and Beyond, ed. István Keul, pp. 99–129. Berlin: Walter de Gruyter. Hatley, Shaman. 2007. ‘The Brahmaya-malatantra and Early S´aiva Cult of Yoginı-s’. Ph.D. thesis, University of Pennsylvania. Pintchman, Tracy. 1995. The Rise of the Great Goddess in the Hindu Tradition. Albany, NY: State University of New York Press. Sanderson, Alexis. 2009. “The S´aiva Age”. In Genesis and Development of Tantrism, ed. Shingo Einoo, pp. 41–350. Institute of Oriental Culture Special Series, no. 23. Tokyo: Institute of Oriental Culture, University of Tokyo. Sanderson, Alexis. 2006. ‘The La-kulas: New Evidence of a System Intermediate between Pa-ñca-rthika Pa-s´upatism and Agamic S´aivism’. Ramalinga Reddy Memorial Lectures, 1997. In The Indian Philosophical Annual, 24: 143–217. Sanderson, Alexis. 2001. “History through Textual Criticism in the Study of S´aivism, the Pañcara-tra and the Buddhist Yoginı-tantras”. In Les Sources et le Temps. Sources and Time. A Colloquium. Pondicherry 11–13 January 1997, ed. François Grimal, pp. 1–47. Pondicherry: Institut Français de Pondicherry/École française d’Extrême-Orient. Sanderson, Alexis. 1988. “S´aivism and the Tantric Traditions”. In The World’s Religions, ed. S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp. 660–704. London: Routledge and Kegan Paul. Reprinted in The World’s Religions: The Religions of Asia, ed. F. Hardy. London: Routledge and Kegan Paul (1990), pp. 128–172. Ta-ntrika-bhidha-nakos´a, vol. 2. Hélène Brunner, Gerhard Oberhammer, and André Padoux, eds. Ta-ntrika-bhidha-nakos´a. Dictionnaire des terms techniques de la

32

Shaman Hatley

littérature hindoue tantrique. 2 vols. Beiträge zur Kultur- und Geistesgeschichte Asiens, no. 44. Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2004. Yokochi, Yuko. 2004. ‘The Rise of the Warrior Goddess in Ancient India. A Study of the Myth Cycle of Kaus´ikı--Vindhyava-sinı- in the Skandapura-n.a’. PhD dissertation, University of Groningen. White, David Gordon. 2003. Kiss of the Yoginı-: ‘Tantric Sex’ in Its South Asian Contexts. Chicago, IL: University of Chicago Press.

2

The (Un)Dreadful Goddess Aghorı- in early S´a-kta tantras1 Judit Törzsök

Introduction Aghorı-, literally ‘undreadful’, is a frequently used name for the highest goddess in tantric as well as nontantric contexts, from the earliest period of tantrism. It could therefore be considered just another appellation or epithet of the supreme goddess, employed in a way similar to the word S´iva: in both cases, a potentially dangerous and frightening deity is called benign or ‘undreadful’ by antiphrasis, to avert the deity’s wrath or to transform it into benevolence. Aghorı-, however, seems to be more than just another epithet, at least in some early S´a-kta tantras, whose relevant passages I propose to present and analyse in this chapter. First, one could argue that if Aghorı- were an epithet, it would have the form of the adjective aghora-, and not aghorı-. Such an argument could be used if our texts followed all the rules of classical Sanskrit grammar; but it cannot apply to early tantras, in which word formation is often irregular and which thus use the two forms aghora- and aghorı- sometimes interchangeably. To give one example: the Siddhayoges´varı-mata (of the seventh century approximately) has a set of eight goddesses headed by Aghora-. This Aghoraappears consistently as Aghorı- in a reccurring compound that denotes this group: aghorya-dyas.t.aka, the ogdoad starting with Aghorı-. That the compound does not refer to a different set of eight starting with Aghorı- instead of Aghora- is confirmed by Jayaratha, the commentator of Abhinavagupta’s Tantra-loka (ad 30.27). He explicitly says that the mantra of Aghorı- is the one that includes the epithet or name Aghora- in the vocative. (OM. AGHORE HRI-H. ITI AGHORYAH . ). This and other examples show that just as in the case of other deities, the difference between an epithet and a name is not always clear. In spite of the confusion between Aghora-s and Aghorı-s, epithets and proper names, one could retain as a rule of thumb that when a single goddess is meant, she is generally called Aghorı-. How is tantric Aghorı- related to the vedic mantra and its personification, Aghora? In what company and in what ritual context can she usually be found? And how, where and when did she become the supreme goddess? These are

34

Judit Törzsök

the questions I shall attempt to answer, using the evidence of early S´a-kta tantras, mostly dating from about the seventh and eighth centuries CE.

1 The triad of Aghora, ghora and ghoraghoratara The male deity or epithet aghora plays an important role already in pre-tantric S´aivism. The vedic Aghora-mantra, which evokes three forms of Rudra, is one of the five mantras of the Pa-s´upatas.2 The S´aiva Siddha-nta adopts these five mantras (commonly known as the five Brahma-mantras), and, at a later date, associates them with five heads of S´iva.3 The Aghora mantra thus becomes identified with the frightening Southern face and Bhairava.4 This is a rather important detail in the context of S´a-kta scriptures, for early S´a-kta tantras categorise themselves as Bhairavatantras and claim to go back to the teaching of Bhairava or Aghora himself.5 The scriptures are written in the form of dialogues in which the goddess is taught by Bhairava, the originator of the doctrine. This is the case of the Siddhayoges´varı-mata, for example, which in fact derives itself from the Svacchandabhairava, and also claims to be part of the S´a-kta tantras of the Vidya-pı-t.ha. The Ma-linı-vijayottara, which in turn mentions the Siddhayoges´varı-mata as its source, agrees with this in principle, but not in all the details. The supreme god of the dialogue is called Parames´vara, but it is claimed that the teaching was originally propounded by Aghora.6 The Aghora-mantra itself mentions three forms of the god: the undreadful or benign (aghora), the dreadful (ghora) and the more dreadful than the dreadful (ghoraghoratara). Although some S´a-kta tantras do not adopt this and other vedic mantras in their core rituals,7 the three forms of Aghora figure prominently in several of them. Instead of being forms of Rudra, they represent three groups of female powers, s´aktis, who govern the world. The earliest S´a-kta tantra that mentions them is probably the Siddhayoges´varı-mata: [The Benign ones:] The Female Powers who bestow grace by clearing up the darkness of ignorance which resides in the body of the bound soul are called ‘the Benign Ones’ (aghora-h.), they give people [help to attain] S´iva[hood]. There are Rudras governed by these Benign Powers. Focussing their minds on Sada-s´iva, they release the limited individual soul from its bondage.8 [Dreadful ones:]9 [Other Female Powers] who obstruct the way to liberation are known as ‘the Very Dreadful Ones.’ (?) The Rudras who are possessed by them perform creation and dissolution, they play in the body like children with clay bulls.10 [Extremely dreadful ones:] Those [Female Powers] who cause a downward flow [of the souls into lower levels of the universe] and gratify the soul, who make it indulge in objects of enjoyment and in the condition of being bound, obstruct the

The (Un)Dreadful Goddess

35

way to liberation. They are called ‘the Ones Surpassing the Dreadful’ (ghoratarya-h.) The Rudras whose minds are empowered by them are governed by them. They also throw down and down those who are given to the pleasures of the bound soul.11 [Summary:] Since possession by the three [kinds of] Powers is always beneficient/ Since S´iva is infused with the three kinds of Powers, they are called12 the ‘Ones More Dreadful Than the Dreadful,’ the ‘Dreadful Ones’ and the ‘Benign Ones.’ Thus, established in the Power of Rudra, they govern this world.13 In the preceding and subsequent passages, the Siddhayoges´varı-mata also clarifies that these are three manifestations of the supreme s´akti called the Power of Rudra (rudras´akti). The text in fact draws a parallel between the three s´akti groups and the main triad of goddesses of the Trika, Para-, Para-para- and Apara-. It thus creates a fully S´a-kta doctrine and pantheon, inspired by elements of a Vedic mantra.14 The teachings about these three s´akti groups, who appear to be of pure, mixed and dark (sa-ttvika, ra-jasa and ta-masa) nature, subsequently become very widespread. The same kind of description can be found not only in a closely related Trika tantra, the Ma-linı-vijayottara (3.31–33), but also in other . s´a-kta schools and scriptures: in the Bhairavamangala- (282–7), in the Jayadrathayamala (3.24.73–83; 4.61.43), in the Devıpañcas´atika- (4.38–40), as well as in the Netratantra (29.159cd–160). Further triads came to be homologised with them: such as various goddess triads,15 or the three functions of Volition, Knowledge and Action (iccha-, jña-na, kriya-).16 The later parallels and the occurrence of the Aghora-triad in the Siddhayoges´varı-mata, the earliest scripture of the Trika, suggest that the s´a-ktisation of members of the Aghora mantra happened in the Trika. This, however, does not seem to be the case. Before the Trika, the La-kulas or Prama-n.apa-s´upatas already expounded a doctrine to this effect. The only surviving passage of their scriptures,17 a citation given by Ks.emara-ja (11th cent.),18 attests that they posited the existence of three groups of Rudras and s´aktis in very similar terms. [Benign Rudras:] ‘Dreadful’ (ghora) is the name of the net of bonds, terrible and associated with crimes. Those who lack this net of bonds are known to be the Benign Ones (aghora), Rudras starting with Va-mes´vara, established above the source of this net. They are called the benign ones, listen to the Dreadful Ones told in brief.19 [Dreadful Rudras:] Rudras headed by Gopati and ending with Gahana are the Dreadful Ones (ghora), inhabiting various worlds.20 [The More Dreadful than the Dreadful Ones]

36

Judit Törzsök And those who start with Vidyes´vara and end with Ananta, the Great Ma-hes´varas are yet different, known as the More Dreadful Than the Dreadful Ones (ghoraghoratara), residing below.21 [Their s´aktis] Thus, there are Benign Aghoras, Dreadful Ghoras and More Dreadful Than the Dreadful Ghoraghorataras, as well as s´aktis of the highest lord, Parames´vara, who are always established in them. The s´aktis of these forms effect everything concerning creation, destruction and maintenance, as well as the acts of binding and release. The individual pays homage to all these forms.22 The homage is an act of donation [of one’s self], which is characterised by cause23 and effect/ having the characteristics of the cause of a [particular] effect.24

This passage is very close in wording and content to the description of the s´akti groups found in the Siddhayoges´varı-mata, and the structure of the two descriptions is also similar, although there are a few important differences. The Siddhayoges´varı-mata lacks the strong allusion to the wording of the Aghoramantra, which the La-kula passage has at the end, and does not mention the names of Rudras, unlike the La-kula text. In this way, it is further removed from the Pa-s´upata background. At the same time, although the s´aktis are mentioned in the Pañca-rthaprama-n.a, too, it is the male triad that the passage focuses on. The Siddhayoges´varı-mata s´a-ktises the passage in that it reverses the order of presentation, and clearly treats the s´aktis as supreme powers, to whom the Rudras are subject. Now does the Siddhayoges´varı-mata borrow this passage and doctrine directly from the La-kulas? If it did borrow much of this passage from somewhere, its source may have been a text of the Siddha-nta rather than this La-kula passage directly. For there is a very odd detail in the Siddhayoges´varı-mata’s version that cannot come from the La-kulas, and it does not fit in the doctrine of the Siddhayoges´varı-mata either: it is the fact that the benign Rudras in the Siddhayoges´varı-mata concentrate on Sada-s´iva (sada-s´iva-rpitadhiyo), the godhead of the Siddha-nta, while releasing the individuals25 from their bonds. Whether the direct source of the Siddhayoges´varı-mata was a La-kula or a S´aiva Siddha-nta passage here, the parallels show us some details of the way in which the S´aiva doctrine of the Aghora triad became s´a-ktised and provide us with yet another example for continuities between pre-tantric and tantric currents, including the presence of s´aktis in the pre-tantric scriptures of the La-kulas.

2 Aghorı- in groups of eight goddesses 2a The octad starting with Aghorı-/AghoraAnother way in which the Siddhayoges´varı-mata and the system of the Trika s´a-ktised the Aghora mantra is that they introduced a S´a-kta mantra starting with the feminine vocative of Aghora-, which thus functioned as some kind of

The (Un)Dreadful Goddess

37 female version of the vedic Aghora mantra. This female mantra (vidya-), which includes a number of meaningless seed syllables and a series of proper names in the vocative, incorporates a set of eight goddesses in its word elements (padas), in the following order: Aghora- (‘undreadful’), Paramaghora- (‘extremely dreadful’), Ghoraru-pa- (‘she who has a dreadful form’), Ghoramukhı(‘she who has a dreadful face’), Bhı-ma- (‘terrifying’), Bhı-s.an.a- (‘frightening’), Vama[nı-] (‘vomiting/creating’), and Piba[nı-] (‘drinking/withdrawing’). The full sequence forms one of the main female mantras of the system, the mantra of the middle goddess of the Trika, Para-para-.26 Om . Aghore hrıh. Paramaghore hum . Ghorarupe hah. Ghoramukhi Bhıma-Bhıs.an.e Vama Piba he ruru rara phat. hum . hah. phat.. The eight goddesses incorporated in it, about whose iconography the text remains silent (although it describes the three main mantra goddesses in detail), have an important function: they are projected onto eight body parts of the practitioner from head to foot (Siddhayoges´varı-mata 7.27 ff.) during his ritual transformation into Bhairava, after he has placed the three goddesses on his head, in his heart and on his feet. The eight goddesses are also present on the man.d.ala of the cult: before placing the goddess triad on the tips of a trident in the centre of the man.d.ala, the practitioner must project this ogdoad on the eight petals of the lotus from which the trident rises.27 The eight goddesses in fact replace eight male deities who would occupy the petals in the S´aiva Siddha-nta and related systems. For the Siddha-nta usually has the eight Vidyes´varas in this arrangement, while the Svacchandatantra puts eight Bhairavas there.28 Thus, the Siddhayoges´varı-mata and related Trika texts, which took over these mantras, produced a S´a-kta version of two ritual elements. First, they created a S´a-kta version of the vedic Aghora mantra, or, more precisely, a S´a-kta mantra somewhat reminiscent of the Aghoramantra. Second, they replaced the Siddha-nta’s and the Svacchandabhairava’s eight male deities on the lotus petals of the man.d.ala with eight goddesses.29 The thirty-eight-syllable mantra of Para-para- and its octad of mantra goddesses then found their way from Trika scriptures (the Siddhayoges´varımata, the Tantrasadbha-va and the Ma-linı-vijayottara) into related s´a-kta tantras: they figure in an inflected form in the Kubjika-mata (18.34),30 and they are prescribed in a different form but still as parts of Para-para- and as the ‘Octad Starting with Aghorı-’ (aghorya-dyas.t.aka) in the Kularatnoddyota (5.135, 6.258–266).31 2b Octads ending with AghorıIn addition to the so-called aghorya-dyas.t.aka or the octad starting with Aghorı-, one also finds a very different set of eight goddesses in the Trika, which,

38

Judit Törzsök

however, functions as an alternative to the Aghorı- octad. In this different set, the name Aghorı- does not always figure, and when it does, it is the last one, rather than the first. The Tantra-loka (33.3) mentions Ma-hes´-ı, Brahman.-ı, Ska-ndı-, Vais.n.avı-, Aindrı-, Yama-tmika- (i.e. Ya-mya-), Ca-mun.d.a- and Yogı-s´-ı (i.e. Yoges´-ı). The first six names derive from the well-known Mother goddesses, who epitomise the s´aktis of their respective male deities, while the last two, Ca-mun.d.a- and Yoges´-ı or Yoges´varı-, are independent goddesses. It is the last one, Yoges´-ı, a synonym of yoginı-, who is alternatively called Aghorı- (Tantrasadbha-va 3.216). The octad is commonly called Brahman.ya-di (Tantrasadbha-va 3.216, 8.22) or Ma-hes´ya-di32 (Ma-linı-vijayottara 20.37). An almost identical set of deities ending with Aghorı- is described in the Siddhayoges´varı-mata (ch. 16) as eight goddesses presiding over groups of consonants of the Sanskrit alphabet. They are listed following the order of consonants they are to be worshipped with. Bra-hmı- is identified with the velars, Vais.n.avı- with the palatals, Ma-hes´-ı with the retroflex consonants, Ya-mya- with the dentals, Kauma-rı- with the labials, Va-savı- (a synonym of Aindrı-) with the semi-vowels, Karn.amot.-ı with the sibillants and Aghores´-ı, the supreme one (para-), with the ligature of Anger, KS.A. The order of these Mother-goddesses is different from the above-mentioned sets and Karn.amot.-ı stands for Ca-mun.d.a-. It would require a separate study to examine the order of these goddesses and the possible reasons of how and why such lists differ. Here, I would only like to point out that in spite of differences in order, differences that concern both the beginning and the middle of the list, Aghorı- or Aghores´-ı always remains at the end. This is also the case in the Brahmaya-mala, whose composition perhaps slightly predates that of the Siddhayoges´varı-mata. Its standard set of eight33 consists of Ma-hes´-ı, Brahma-n.-ı, Vais.n.avı-, Kauma-rı-, Ya-mya- (called Vaivasvatıand sometimes replaced by Va-ra-hı- in the latter half of the text), Aindrı-/ Va-savı-/Ma-hendrı-, Ca-mun.d.a- (also called Carcika- as well as Can.d.ika-) and the last one alternatively called She Who Completes the Mothers (4.257) and Aghores´-ı (89.92), who is also said to be Yoges´-ı or Yoges´vares´varı- (58.1).34 The various names of the last goddess, her position in the ogdoads and the fact that she is generally absent from early representations of the standard set of Mother goddesses (ma-tr.) suggest that she was added later to the already existing pantheon of Mothers. She is also absent from the Brahmaya-mala’s yoginı- clans, which are based on the Mother goddesses and end with Ca-mun.d.aas the Leader of Mothers.35 As is common for newcomers in an already existing pantheon, one of her main functions appears to supersede all other Mother goddesses. She is therefore called the Supreme one (para-). Aghorı-’s role and position among the Mothers also suggests that through her, the nontantric cult of mothers was assimilated and at the same time allegedly superseded in the Bhairavatantric systems. For, as the case of the original Svacchandatantra shows, the mother goddesses were not fully part of the original Bhairavatantric pantheon. In the Svacchanda, they do not appear as alphabet goddesses in the earlier, shorter recension of the text, and were

The (Un)Dreadful Goddess

39

clearly added there at a later date, including some inconsistencies in their enumeration.36 This also implies that by the time the earliest S´a-kta tantras were redacted, the cult of Mothers as manifestations of s´aktis must have been developed. This does not need any demonstration, but certain details in the development of the cult of Mothers as s´aktis and their iconography may shed more light on the dating of s´a-kta tantras. One rather interesting detail concerns Bra-hmı-, who is associated with the antelope hide in Siddhayoges´varı-mata 29.24.37 This iconographic feature appears in the sixth and seventh centuries in various regions of India, but disappears by the eighth century, when a larger scale standardisation seems to take place.38 In addition to confirming the tentative dating of the Siddhayoges´varı-mata to a period probably around the seventh century, the borrowing of this iconographic detail also suggests that the cult of mothers was perhaps integrated into S´a-kta tantrism around this time, or, in any case, not much earlier or not in a more archaic form. No matter how exactly this assimilation or adaptation happened, the addition of Aghorı-/Yoges´-ı allowed the smooth integration of the seven Mothers into the tantric system, which commonly arranged deities in circles of eight. Aghorıthus indeed completed the Mothers conveniently, who needed a last member, so to speak, in order to be worshipped on the eight petals of the generic lotus design of the Tantras.

3 Aghorı- or Aghores´-ı as the supreme goddess Aghorı- played a much more important role than just completing and superseding the Mothers in early S´a-kta tantrism. Aghorı- is the name of Svacchandabhairava’s consort in the Svacchandatantra, and Aghores´-ı or Aghores´varı- is also one of the names of the supreme mantra goddess in the Brahmaya-mala.39 Surprisingly, she also appears once as the highest goddess in the Siddhayoges´varı-mata, whose main pantheon is supposedly different. At the end of its third chapter, a few lines celebrate the infinite power of the mantra goddesses and the power of the system or doctrine of yoginı-s, the Siddhayoges´varı-mata itself. In the last verse of this passage, the three mantra goddesses or s´aktis are said to bestow all kinds of success. They are called the Heart of All S´aktis, and are declared to be the means to attain the goddess Aghores´-ı.40 It is in the Brahmaya-mala that Aghorı- or Aghores´-ı is described in detail as the supreme goddess of the cult. This goddess is also called Bhairavı-, and, more specifically Can.d.a- Ka-pa-linı-, for she appears in, or rather, as the nine-syllable root mantra of this system: HU-M. CAN.D.E KA-PA-LINI SVA-HA-.41 Her name also figures in the form of Can.d.ika- or Can.d.as´akti (46.95). The mantra, or rather, vidya-, that is Aghores´-ı or Can.d.a- Ka-pa-linı- is the source of the six yoginı-s of the system, who are said to emerge from the limbs of Aghores´varı-. This Aghores´varı- is clearly distinguished from the one who is the last member of the series of Mother goddesses.42

40

Judit Törzsök

In this context, it is tempting to speculate on whether this nine-part mantra is not a S´a-kta version of the very widespread Nava-tman mantra. The Nava-tman is an important S´aiva mantra already in the earliest surviving tantra, the Nis´va-sa,43 and it is also described and used in the Siddhayoges´varı-mata and related texts.44 Later, the Kubjika-mata makes him the male consort of its supreme goddess.45 However, it rarely figures in the Brahmaya-mala. It is in fact absent in the first and earlier half of the text, which uses the compound nava-tman or nava-tmaka in a different meaning. For in the system of the Brahmaya-mala, the practitioner does not normally transform his body with the help of the Nava-tman-mantra. One of the most common expressions that describe the body of the practitioner as transformed into the deity is avadhu-tatanu, which does not mean that he has discarded his (worldly) body (as a straightforward interpretation of the compound would suggest), but refers to the fact that he projects the nine aks.aras of the main goddess of the cult, also called Avadhu-ta-, onto his body.46 It seems therefore possible that the ninesyllable Can.d.a- Ka-pa-linı- usurps the place of the otherwise ubiquitous nine-letter Nava-tman-mantra, at least in the earlier section of the Brahmaya-mala.47 Whether this Aghorı-/Can.d.a- Ka-pa-linı- represents a S´a-kta version of the Nava-tman or not, she was adopted in an altered form in the Tantrasadbha-va, which calls her Aghorı-, and gives her mantra in the form OM. CA-MUN.D.E KA-PA-- - 48 The fusion between various fierce goddesses, Can.d.ika-, AghorıLINI SVAHA. and Camun.d.a, here becomes complete. Interestingly, the Brahmaya-mala tends to call this main mantra goddess Can.d.a- in chapters concerning mantra descriptions, while in narrative passages, such as in the story of the transmission of the tantra or in the narration of ritual events, she is more commonly called Aghorı-. This may or may not be due to exigencies of the metre. In any case, the very first chapter of the Brahmaya-mala puts this name in the foreground. The first chapter relates, as is customary, the mythical transmission of the text: the tantra-vata-ra. Again, as is customary in S´a-kta Tantras, the goddess, here in the form of Aghorı-, asks Bhairava to teach her the secret doctrine.49 Bhairava relates that he himself obtained the teaching from S´rı-kan.t.ha and transmitted it to the goddess. She, however, started teaching it to her attendants without any authorisation. Seeing this, Bhairava was filled with anger and destroyed the goddess’s knowledge of the scriptures. At the supplication of the goddess, he finally conceded that she could regain her knowledge if she was to be reborn and to worship him. The goddess then took up a new body . in the house of a brahmin in Praya-ga and worshipped the linga devotedly. When she obtained perfection (siddhi), Bhairava bestowed knowledge upon her with the following words (tr. Hatley):50 You are that s´akti, O highly fortunate one, whom of old I had cursed when a transgression was made. Remember yourself! Your name is now Aghorı-. You are my Great S´akti, the agent of grace for all. I shall teach you that knowledge which was lost when you made a transgression, the

The (Un)Dreadful Goddess

41

great [tantra] spoken by S´rı-kan.t.ha, after it came into being from the sada-s´iva tattva (?),52 together with its secrets, O Maha-devı-; listen with single-minded attention.53 51

Here, the name Aghorı- is presented as a reward. The same name, in the form of Aghores´-ı, also appears in titles of scriptures listed in the Brahmaya-mala: three s´a-kta tantras of the Vidya-pı-t.ha (within the Bhairava current) bear the titles Aghores´-ı, Aghores´varı- and Krı-d.a-ghores´-ı.54 Given the prominence of Aghores´-ı/Can.d.a- Ka-pa-linı-, it is surprising that one finds hardly any mention of her in the man.d.ala-description of the Brahmaya-mala. It is observed by Sanderson (1988: 672) that the text presents Bhairava as the supreme deity in the iconic representation of the pantheon, even if he is transcended by Can.d.a- Ka-pa-l.inı- on the mantric plane. In addition, as Kiss (2014: 20) remarks, the man.d.ala-description of chapter 3 shows almost no awareness of a female central deity, even in her mantric identity. The description of the man.d.ala and the rituals performed around it lack any explicit mention of an all-transcending central goddess. Csaba Kiss concludes that the man.d.ala description itself may belong to an early phase of the cult of Kapa-lıs´a, which was not yet dominated by a s´akti. Thus, it seems that Aghores´-ı/Can.d.aka-pa-linı-, both in her all-transcending form and in her lower manifestation as the eighth Mother-goddess, represents an addition to already existing pantheons and cults, at least in the system of the Brahmaya-mala. Although Can.d.a- Ka-pa-linı-/Aghorı- is absent from the man.d.ala and the rituals connected with it, she is very much present in a series of chapters that prescribe and describe rituals whose aim is to obtain supernatural powers, carried out in the cremation ground. These chapters form a cluster near the end of the first and earlier part of the text, from chapters 46 to 49. She is mentioned and appears in each of them, often together with groups of female spirits or with Bhairava. She is most often called Aghorı-, but occasionally also Can.d.as´akti (e.g. 46.95) and Can.d.ika- (e.g. 46.107). The first rite is called the Great Churning (maha-mantha-na), the second prescribes the construction of a Magic Pavilion (siddhaman.d.apika-, wrongly entitled in the manuscripts as magic man.d.ala), the third one explains the Worship of Deities in a Hole (garta-ya-ga) and the last one concerns the use of yantra diagrams (yantra-rn.ava). A number of elements link these chapters together, which are of strongly ka-pa-lika character. They prescribe rites in the cremation ground, and in almost all of them the practitioner must transform his body according to the maha-vrata as a preliminary (maha-vratatanusthitah.), i.e. he must wear the five insignia, which include bone ornaments, the skull bowl, etc. During most of these rites, he will be possessed primarily by Bhairava, but also by his retinue of goddesses; in this way, his transformation into Bhairava is completed through possession, just as it was said to happen for ka-pa-likas.55 Aghorı- plays various important roles in these rites: she is a mantra to be recited in one’s heart (49.32). The practitioner can see her form (cha-ya-) as a reward, and she bestows all kinds of siddhis on him as a boon (48.20ff).

42

Judit Törzsök

Alternatively, she can send thousands of spirits (bhu-ta) to give him a boon, or the seven Mothers may appear and elect the practitioner as their eighth member (as.t.amaka 48.27ff). Aghorı- commands groups of s´a-kinı-s, yoginı-s (maha-vı-ra-) and ca-mun.d.a-s (47.18, 26); and she appears in front of the sa-dhaka to accept him as her son at the end of the so-called Great Churning. Because of Aghorı-’s significant role in the last rite, it may be useful to summarise it briefly. The churning is performed in the cremation ground and preceded by a worship of deities in nine skulls arranged on a man.d.ala and filled with blood, alcohol (madira-) and the mingled sexual fluids (picu). For the churning, which is carried out on top of a corpse, the practitioner must use materials obtained on the spot: the ad hoc pavilion he constructs is made of human bones, the vessel used is fashioned of clay he finds in the cremation ground, the churning stick is a large bone such as the tibia, and the rope is made of human hair and intestines taken from the dead. The rite mainly consists of imitating the cosmic churning of gods and demons in its cremationground version. Accordingly, it produces various miraculous objects, such as the Kaustubha gem or the moon, and makes groups of mythical beings and powers present, whose help may be refused if one aspires for greater powers or results. The practitioner in fact prepares his own soma of impure substances in a vessel called stha-lı-, which he identifies with Aghorı-, while the churning stick is to be identified with Bhairava and himself.56 Near the end of this rite, Bhairava appears with his goddesses, and they enter the practitioner. Then, at the very end, Aghorı- herself appears before him and starts a pura-n.ic-type conversation: Well done, noble being, you who are honoured by gods and demons alike. You have performed a true miracle here, concerning this sa-dhana, my son. You have become the ornament of the world, an eminent man. Sit on my lap, my son, and choose yourself the right boon: immortality, bhairavahood, through which you can fulfill [even] the wishes of Brahma-, Vis.n.u or Indra, or whatever you wish, my dear son, together with your fellow practitioners.57 The practitioner then replies: If you are satisfied with me, o goddess, then give me your breast, o Mother. Hearing these pleasant words of the sa-dhaka, the goddess [says]: Come, come, noble being, drink from my breast, my adopted son. Which other noble being than you would deserve to be my son? Therefore, I shall embrace you, heroic practitioner, and give you my breast. … The goddess then does so and the eminent practitioner, after drinking the nectar of immortality left-and-right [i.e. from both breasts] will become omniscient in a second. He will become Bhairava in person, the Lord of Guhyaka-s. All the seventy-million mantras will enter his heart.58 Contrarily to the mantric Can.d.a- Ka-pa-linı-, this Aghorı- seems to be a visually perceivable being. When she appears, her features are described in a few lines:

The (Un)Dreadful Goddess

43

While he [the practitioner] churns for one more second, making obeisance to Can.d.ika-, the goddess Aghorı-, who bestows supernatural powers, will arise there [in the churning vessel]. Her form is like the rainbow, she looks like ten million suns. She is adorned with all kinds of ornaments and is seated on a throne [made of/decorated with] skulls. She is surrounded by young maidens like herself and her body extends from the underworld up to the sky, filling the whole space.59 This chapter of the Brahmaya-mala combines various aspects of Aghorı-: the mantra-goddess, who must be recited mentally during the rite, the ruler of yoginı-s, s´a-kinı-s and other spirits, the all-transcending cosmic s´akti, and the Mother goddess, who adopts the practitioner as her son. It is therefore all the more interesting to see once again that she does not figure, at least not with the name Aghorı-, on the man.d.ala described at the beginning of the chapter.60 The order of the Mothers in the eight directions, surrounding a male god in the middle and each accompanied by their own male consort, is as follows (starting from the East): Ma-hes´-ı, Brahma-n.-ı, Vais.n.avı-, Kauma-rı-, Vivasvatı- (= Ya-mya-), Va-savı- (= Aindrı-), Bhairavı-/ Yoges´-ı and Guhyas´akti. Another surprising detail is that the passage describes the seventh Mother, the emaciated Bhairavı-/Yoges´-ı, in the most detailed manner and therefore suggests that she is the central one in the system. Her two names, Bhairavı- and Yoges´-ı, are alternative names of Aghorı-/Can.d.a- Ka-pa-linı- in the Brahmaya-mala, which also points to their identity. However, it is the last Mother, Guhyas´akti, whose radiant appearance agrees with Aghorı-’s. While no detail is given on the previous six Mothers at all, the last two goddesses are described as follows: Bhairavı- is flat-nosed, red-eyed and wrathful. She is emaciated, seated on a corpse and has a thousand creeper-like arms. Facing Bhairava, she looks like ten million suns. She is [also called] yoges´-ı [i.e. a yoginı-] she wears all kinds of ornaments and is placed on the northern petal. One should place [the eighth goddess] Guhyas´akti in the north-eastern direction, o fortunate one. She is like the rising sun and is known as the Mother of All.61 Whether one identifies Aghorı-’s lower manifestation with the emaciated seventh or with the radiant eighth mother in this case, with neither or with both of them, it is certain that this passage shows some hesitation as to who is the supreme Mother and her potential equivalent on the man.d.ala. I take this to imply, once again, that Aghorı-’s cult was not directly connected to the Mother goddess cult, in spite of her having motherly characteristics when she appears at the end of the so-called Great Churning.

Conclusion To summarise these investigations concerning Aghorı- in early S´a-kta Tantras, the following factors may have contributed to the shaping of her identity and cult:

44

Judit Törzsök

First, continuities with pretantric currents, in particular the La-kulas, should be emphasised. The idea of the three kinds of s´aktis headed by the benign Aghora-s predates the S´a-kta tantras, although it is only in the S´a-kta sources that the role of male and female deities becomes reversed and goddesses dominate their Rudras. Second, it is possible that some elements of Aghorı-’s cult came from a pretantric Ka-pa-lika cult. As I pointed out in this chapter (and elsewhere), a number of particularly remarkable Ka-pa-lika elements figure in the Brahmaya-mala, in the cluster of chapters (46–49) in which Aghorı- also appears as a main goddess: they include the practice of the Maha-vrata, possession by Bhairava (and possession, a-ves´a is often attributed to the Ka-pa-likas), cremation ground rituals, etc. Aghorı- is also named a skull-bearer in her mantra (ka-pa-linı-), and in this context one cannot but recall the name of the Ka-pa-lika practitioner from the Ma-latı-ma-dhava: Aghora-ghan.t.a. Therefore, it is not impossible that Aghorı- as a supreme goddess has some Ka-pa-lika antecedents. Third, Aghora- or Aghorı- in their mantric identities often seem to s´a-ktise, so to speak, existing male S´aiva entities. In the Trika system, Aghora- is the first member of the Para-para- mantra, whose eight goddesses replace the male Vidyes´varas/Bhairavas on the man.d.ala; and in the Brahmaya-mala, Aghorı-’s nine-syllable mantra functions as a substitute of the male Nava-tman. Fourth, Aghorı- is also used, quite evidently, to draw the cult of the seven Mothers in the s´a-kta tantric doctrine and ritual. She is the eighth Mother, but at the same time she supersedes them, just as the tantric cult of the Mothers is meant to supersede the non-tantric one. In this sense, it seems that the cult of the Mothers is not a direct precursor of S´a-ktism, but is rather one of the cults that S´a-kta tantras adopted and transformed, just as they adopted and transformed pretantric S´aiva doctrines and practices of the Atima-rga. Finally, in the midst of these multiple Aghora-s and Aghorı-s, one should not forget that Aghorı- may also function as an epithet: one with strong vedic associations, recalling the universal strands of sattva, rajas and tamas, conveniently usable for an all-transcending as well as all-encompassing principle and goddess. Whether Aghorı- is just another name or an identifiable, possibly Ka-pa-lika, deity, she played no small role in the process by which S´a-kta Tantric doctrine and practice evolved.

Notes 1 I am grateful to the participants of the conference ‘The Goddess: Understanding the S´a-kta Traditions’ (Somerville College, Oxford, 10–11 September 2011, held by the Oxford Centre for Hindu Studies, the University of Oxford and Aarhus University) for their comments and suggestions, in particular to Prof. Alexis Sanderson, Prof. Harunaga Isaacson, Dr Shaman Hatley and Dr James Mallinson. I would also like to thank Prof Gavin Flood for inviting me to give a paper and Dr Bjarne Wernicke Olesen for his kind help. In this paper I have not commented on particularities of tantric or Ais´a Sanskrit, unless the interpretation is problematic, for the main point of the paper concerns religious history rather than philology.

The (Un)Dreadful Goddess 2 3 4

5

6 7 8 9

10 11

12 13 14

15 16 17 18 19 20 21 22

45

See Pa-s´upatasu-tras 3.21. This, however, does not happen in the core part of the earliest tantra, the Nis´va-sa, as Goodall 2009 points out. The Southern face is said to be frightening (raudra) already in the later layers of the Maha-bha-rata (13.128.6cd), as Bakker 2002: 398 shows. For representations of the faces, which systematically include a ferocious one as early as the beginning of the Christian era, see Bakker 2002: 398 and Kreisel 1986: 54ff. See Hatley 2007: 185: ‘The emergence of Bhairava in the tantric S´aiva pantheon, whatever his roots may be, appears to have involved some degree of identification with Aghora, the southern, fierce face of Sada-s´iva who is said to reveal the bhairavatantras.’ maya-py etat pura- pra-ptam aghora-t parama-tmanah.. Ma-linı-vijayottara 1.14ab. The Siddhayoges´varı-mata does not appear to use them, but the Brahmaya-mala (ch. 12ff.) prescribes the sadya-di to create five mantric faces (pañcavaktra-n.i vı-rasya sadya-dı-naiva da-payet). - pramr.jya-jña-natimiram . pas´udehe vyavasthitam / yah. s´aktayo ’nugr.hn.anti aghoras tah. s´ivaprada-h. // rudra-s ta-bhir aghora-bhih. s´aktibhih. samadhis.t.hita-h. / sada-s´iva-rpitadhiyo bandhana-n mocayanty an.um // Siddhayoges´varı-mata 2.24–25. One line or more may be missing here, for the characterisation of these powers is unusually brief compared to the others. Moreover, it is also possible that a line is misplaced or became corrupt, for it is also odd to describe these s´aktis as very dreadful: ghoratarya-s. muktima-rganirodhinyo †ghoratarya-s† tu ta-h. smr.ta-h. // a-vis.t.a-h. s´aktibhis ta-bhih. sargapralayaka-rin.ah. / krı-d.ante vai tanau rudra- ba-la- mr.dvr.s.abhair iva // Siddhayoges´varı-mata 2.26cd-27. adhah.srotavidha-yinya-h. pudgalam . rañjayanty api / bhoges.v eva pas´utve ca pudgalam . rañjayanti ya-h. // muktima-rganirodhinyo ghoratarya-s tu ta-h. smr.ta-h. / upodbalitacaitanya rudras tabhir adhis.t.hitah. // pas´ubhoges.u sam . saktan adho ’dhah. patayanty api / Siddhayoges´varı-mata 2.28-30ab (N.B.: pas´utve is a conjecture, the manuscripts read paradve.) There is a reference to the names of S´iva or Rudra here. Since the three kinds of . powers are beneficient (s´am . kara), or rather, since S´ankara is infused with these three kinds of powers, they are known by his three names. . s´aktitrayasama-ves´o yasma-t sarvatra s´ankarah. // ghoraghoratara-ghora-ghora-s ta-h. parikırtitah. / evam . bhuvana paleyu rudras´aktivyavasthitah. // Siddhayoges´varımata 2.30cd-31. This vedic triad seems central in the system of the text, for its very first verse (1.1, which has been transmitted in a corrupt and incomplete from) appears to say that they form the essence of the tantra: aghoraghoraru-pa-n.i ghoraghoratara-n.i ca / †sthita-ni yasya su-tram . tu – – tantratmam eva ca†. See e.g. Jayadrathayamala 3.24.73ff. In Netratantra 29.159cd-160. For this fact and an analysis, see Sanderson 2006: 175. Citation introduced by s´rı-pañca-rthapram . an.e tu, in the commentary (-uddyota) to Svacchandatantra 1.43. - tu na vidyeta hy ghoreti pa-s´aja-la-khyam . papayuktam . bhayanakam / tad yes.am . aghorah. parikırtitas // vames´varadayo rudra jalamuloparisthita-h. / te hy aghora-h. samakhyatah. s´r.n.u ghoran samasatah. // prokta- gopatipu-rva- ye rudra-s tu gahana-ntaga-h. / te tu ghora-h. sama-khya-tana-na-bhuvanava-sinah. // vidyes´vara-dyananta-nta- maha-ma-hes´vara-s´ ca ye / ghoraghoratara-s tv anye vijñeya-s tv adha a-s´rita-h. // The wording echoes part of the Aghoramantra itself.

46

Judit Törzsök

23 The exact understanding of this expression is not straightforward. The effect (ka-rya) could be creation and the cause (ka-ran.a) the deity; or, rather, the cause may be the homage and the effect liberation. The compound is reminiscent of the compound ka-rya-karan.a (denoting the twenty-three tattvas or the body, see e.g. Pa-s´upatasu-trabha-s.ya ad 5.13), and therefore one may be tempted to emend ka-ran.a to karan.a, but that would result in a major violation of the metre. 24 ete aghora- ghora-s´ ca ghoraghoratara-s tatha- / etes.v avasthita- nityam . s´aktayah. pa-rames´vara-h. // sthitipralayasarges.u bandhamoks.akriya-su ca / sarva-rthaprerakatvena ru-pes.v etes.u s´aktayah. // ru-pebhya ebhyah. sarvebhyo namaska-ram . karoty an.uh. / namaska-rah. paritya-gah. ka-ryaka-ran.alaks.an.ah. // 25 Note that an.u is used in both texts. 26 For the encoded description of the mantra, see Siddhayoges´varı-mata 3.23ff., Ma-linı-vijayottara 3.42ff., and Tantra-loka 30.20ff. 27 For the man.d.ala itself, see e.g. Siddhayoges´varı-mata 6.12ff.; for this arrangement of the eight goddesses, see e.g. Siddhayoges´varı-mata 6.28ab: pu-rvapatra-dike nyasya tato ’ghorya-dikam . gan.am. For an illustration, reconstuction and analysis of the man.d.ala, see Sanderson 1986: 171ff. . 28 For an example of the Siddha-nta, see Matangapa-rames´vara, kriya-pa-da 1.33ff., for the Svacchandatantra see 5.57ff. and Törzsök 2003: 200. The design of the man.d.ala of nine lotuses is the same in both cases, in which the arrangement of the central deities on the nine lotuses reproduces their arrangement on the middle lotus. 29 Para-para-’s octad may also be added to an equivalent set of male deities, as a passage of the Ma-linı-vijayottara suggests. See 8.77cd: aghora-dyas.t.akam . dhyayed aghorya-dyas.t.aka-nvitam. 30 This octad is said to be situated in Aghori (sic!): 18.33 purya-s.t.akam aghoristham . - ru aghore hrı-m shaum aghorya-yai prathamam yatha-vastha-m . vadami te / 18.34 hrım . . . . s´irah. //1// hrım . ru paramaghore hum . shaum . paramaghorayai mukham //2// hrım . ru - ru shaum ghoramukhi shaum - shaum . ghorarupe shaum . ghorarupayai hr.di //3// hrım . . . - ghoramukhyai guhye //4// hrı-m . phat. bhımaname shaum . bhımayai daks.in.abhuje //5// - bhı-sane shaum bhı-sana-yai va-mato bhuje //6// shaum hrı-m hah vama shaum hrı-m . hum . . . . . . . . . . - phat piba he shaum pibanyai va-matoru //8// vamanyai daks.in.oru //7// hrı-m . hum . . . -. 18.35 etat purya-s.t.akam . devya aghoryas.t.akasam . yutam / nyased as.t.avidhangam . tu as.t.apattres.u sa-dhakah.. - - 31 aghorya-dyas.t.akam . devi padas.t.akasamanvitam / paraparaya vidyayah. nyasamargo yatha- bhavet (5.135). 32 This appellation may be due to an attempt to s´aivise the Mothers by making S´iva’s s´akti, rather than Brahma-’s, the first Mother. 33 For the set, see e.g. Brahmaya-mala 4.255–7 and 46.27–32. For a detailed discussion of the early history of the Mothers and their role in tantric currents, especially in the Brahmaya-mala, see Hatley 2007: 31ff. 34 For this synonym, see Hatley 2007: 67, referring to e.g. Brahmaya-mala 35.32, Tantrasadbha-va 14.155b, and Tantra-loka 29.52d. 35 See Hatley 2007: 415ff. 36 For the two recensions of the Svacchandatantra and the question of Mothers in them, see Sanderson 2001: 21, Törzsök 1999: 198ff. and Hatley 2007: 149. 37 In fact, the text prescribes that the practitioner should show the mudra-s of the ascetic bowl and the antelope skin to yoginı-s belonging to the Bra-hmı- clan. Just as in the other cases, the mudra-s represent attributes, which the yoginı-s should recognise. 38 Panikkar 1997: 149. 39 See Hatley 2007: 34ff. - 40 uktam . s´aktitrayam . devi divyasiddhipradayakam / hr.dayam . sarvas´aktın.am aghores´ı prasa-dhakam // 3.54.

The (Un)Dreadful Goddess 41 42 43 44 45 46 47

48 49 50 51

52 53

54 55 56 57

58

47

See Hatley 2007: 251. See Hatley 2007: 254ff. See Nis´va-sa Mu-lasu-tra 2.3ff. Siddhayoges´varı-mata 17.8, 17.15, and especially 20.34–38. See Kubjika-mata chapter 2, Sanderson 1988: 686–7, and Kiss in Ta-ntrika-bhidha-nakos´a vol. III at nava-tman. For Avadhu-ta- as a mantra goddess, projected on the parts of the body, see e.g. Brahmaya-mala 4.503, 12.96, 49.34. The Nava-tman is known only in the latter, chronologically later, half of the text; but even there the passage describing it is clearly borrowed from the Nis´va-sa, as Hatley 2007: 219–20 points out. It is noteworthy in this context that the Siddhayoges´varı-mata shows a similar distribution of the Nava-tman: its first, probably earlier, half does not mention the Nava-tman at all, while the second, later, half describes and uses it. See Tantrasadbha-va 21.154cd-156ab (Hatley 2007: 71 and 226). The tantra-vata-ra of the Brahmaya-mala is rather complicated; here I summarise only the section that is relevant in the context of this paper. For the edition of this passage and its translation, which I cite here without alteration, see Hatley 2007: 296 and 347. Hatley signals in his translation that the interpretation of maha-n (‘great’), which is difficult to account for grammatically, is tentative. It indeed appears to refer to the doctrine, the tantra itself. The purport may be that initially the text had its full, original length, i.e. it was longer than the actual ‘human recension’; but it is also possible that maha-n is just a common verse-filling adjective. It is odd that the doctrine of the Brahmaya-mala should be said to originate in the Sada-s´iva realm, which belongs to the Siddha-nta; perhaps the line or the expression was borrowed from a text of the Siddha-nta. - - - - aghorısa- s´aktis tvam . mahabhage ya s´apta viplave kr.te / pura maya smaratmanam . nama te ’dhuna // madıya tvam . mahasaktih. sarvanugrahakarika / tat pravaks.yami te - s´iva-t ´ jñanam yad bhras t am viplave kr te // s rı kan t hena maha n proktam bhu tva sada . . .. . .. . padat / sarahasyam . mahadevi s.r.n.us.vekagramanasa // Brahmayamala 1.31–33. See Hatley 2007: 296. See Hatley 2007: 218. For additional analyses and details, see Törzsök 2011 and Törzsök forthcoming. - - - / s´aktistha-m s´aktiru-pa-ñ aghorya-[m . ] sthalirupan tu dhyayen mantrı sus´obhanam . . - nabhairavam devam s´uddhasphatikanirmalam ca dhya-ye soma-tmake sthita-m // mantha . . . . . / sahasrabhujaparyantam . cinten manthanarupin.am . // 46.48–9. sadhu sadhu mahasatva surasuranamaskr.tah. // mahas´caryam idam vatsa kr.tan te . sa-dhanam . prati / pr.thivyan tilako jatam tvam es.a purus.ottamah. // utsa[nge] tis.t.a me putra varam . vr.n.ıs.va sobhanam . / amaratvam . bhairavatvañ ca brahmavis.n.endraka-madam . // yañ ca yacayase vatsa sakhayaih. saha putrakah. / 46.111cd-4ab. My understanding of ka-madam . is tentative. It may also be intended in the sense that the sa-dhaka obtains whatever he wants, i.e. brahmatvam . vis.n.utvam . indratvam . va svaka-matah.. Alternatively, it could also mean that bhairavahood is something all these deities wish to obtain (i.e. ka-madam . for ıpsitam . ?). Finally, one could understand that bhairavahood bestows the qualities of Brahma, Vis.n.u, Indra and Ka-ma (the -tva suffix being ommitted). The construction of the compound would normally allow none of these interpretations. - sa-dhakova-ca: yadi tus.t.a-si ma-n devi stanam . me dada ambike // srutva vakyam . tato devya-m . sadhakasya susobhanam . / ehi ehi mahasatva stana me piba putrakah. // tvam . muktva- tu maha-satvah. ko nyo putratvam arhati / paris.va[jya] tato vı-ram . stanan dada-mi sa-dhakah // 46.114cd-116. evam . kr.tvapi vai devi m amr.tam . sadhakottamah. / - savyasavyan tato pıtva sarvvajño bhavate ks.an.am . // bhairavo tha svayam . saks.a - - vimsati tasya vai hrdi // guhyaka-na-m . . prabhu bhavet / saptakot.yas tu mantran.am . .

48

Judit Törzsök

46.120–1. vim . sati is to be emended to vis´anti, following Shaman Hatley’s suggestion. - / tatrottisthati vai devi aghorı59 ks.an.ama-tram . mathed yava namaskr.tva tu can.d.ikam . .. siddhida-yika- // s´akraca-panibha-ka-ra- su-ryakot.isamaprabha- / sarvva-bharan.asam . yukta - - - kapa-la-sanasam . sthita // atmanah. sadr.s´aih. kanyaih. samanta parivaritah. / patalañ ca nabhañ caiva sarvvam a-pu-rya sam . sthita // 46.107–9. 60 It is to be noted that this man.d.ala is different from the elaborate one described in chapter 3 of the text. - samraktalocana-m // nirma-nsa-[m] pretam a-ru-d61 bhairavı-m . cipit.anasan tu kruddham . . . . . -m / bhairava-bhimukha-m kruddha-m su-ryakotisamaprabha-m ha-m sahasrabhujabalları . . . . . . - yoges´-ım uttare dale / -ıs´a-ne tu maha-bha-ge guhyas´aktin // sarvva-bharan.asam . yuktam . - jananı- smrta-m / 46.30cd-33ab. tu vinyaset // udaya-rkkanibha-ka-ram . . . sarvves.am .

References Primary sources Kubjika-matatantra, ed. Goudriaan, T. and Schoterman, J. Leiden: 1988. E-text by Somdev Vasudeva, GRETIL. Kularatnoddyota, MS K: Bodlian Library, Oxford: CSS MS no. C 348 ; MS . Kh : NAK MS no: 1–1653 ; MS G NAK 1–16 ; MS Gh NAK 4–2454 ; MS N NAK 5–5151 , MS C NAK 5–4807 ; MS Ch NAK 5–487 ; MS J NAK 5–5142. E-text edited by Mark Dyczkowski, MIRI. Jayadrathaya-mala, NAK 5–4650 (s.at.ka 1 & 2) ; 5–722 (s.at.ka 3) ; 1–1468 (s.at.ka 4 A 151–116) E-text by Olga Serbaeva, draft transcription prepared in 2004. I am grateful to Olga Serbaeva for making her transcription available to me. Jayadrathaya-mala, Yoginı-sam . caraprakaran.a (an originally independent work transmitted as part of the Jayadrathaya-mala) Prof. Alexis Sanderson’s draft edition prepared in 2004. I am grateful to Prof. Sanderson for making his edition available to me. Tantrasadbha-va, NAK 5–1985 and NAK 5–445, unpublished edition of chapter 4 by Somdev Vasudeva, unpublished edition of Chapters 16 and 25 by Judit Törzsök. Complete e-text established under the supervision of Mark Dyczkowski. MIRI. Tantra-loka of Abhinavagupta, with a commentary -viveka by Jayaratha. 8 vol. ed. with an introduction R.C. Dwivedi and N. Rastogi. Delhi: Motilal Banarsidass. 1987. (Reprint of KSTS 1918–1938) E-text MIRI. Devı-pañcas´atika-: Ka-lı-kulapañcas´atika- also known as Devı-pañcas´atika-, NAK MS 5–5183, 5–358, and 1–252. Electronic edition by Mark Dyczkowski, MIRI. Nis´va-sa(tattvasam . hita), NAK 1–277; London, Wellcome Institute for the History of Medicine, Sanskrit MS I 33. E-text by Dominic Goodall, Peter Bisschop, Diwakar Acharya, and Nirajan Kafle. Mu-la(su-tra), Naya(su-tra) and Uttara(su-tra) ed. D. Goodall and A. Sanderson. Forthcoming. I am grateful to the editors for making their work available to me. Netratantra, with a commentary (-uddyota) by Ks.emara-ja. Ed. V.V. Dvivedi. Delhi: Parimal Publications. 1985. E-text MIRI. Pa-s´upatasu-tras: Pa-s´upatasu-tra with the Pañca-rthabha-s.ya of Kaun.d.inya. Ed. R.A. Sastri. Trivandrum: The Oriental Manuscripts Library of the University of Travancore. (Trivandrum Sanskrit Series vol. 143) 1940. E-text by Oliver Hellwig, GRETIL. Brahmaya-mala, NAK Ms. No. 3–370. E-text by Shaman Hatley, revised by Csaba Kiss. Draft edition of the first 49 Chapters by Csaba Kiss.

The (Un)Dreadful Goddess

49

. Bhairavamangala-, NAK pam 687 S´aivatantra 144. E-text by Somdev Vasudeva 1997. . . Matangaparames´vara: Matangapa-rames´vara-gama (Kriya-pa-da, Carya-pa-da et Yogapada) avec le commentaire -vr.tti de Bhat.t.a Ra-makan.t.ha. Ed. N.R. Bhat.t.. Publications de l’IFI no. 65. Pondichéry, 1982. Maha-bha-rata, ed. V.S. Sukthankar (1927–1943) and S.K. Belvalkar (from 1943) with Shrimant Balasaheb Pant Pratinidhi, R.N. Dandekar, S.K. De, F. Edgerton, A.B. Gajendragadkar, P.V. Kane, R.D. Karmakar, V.G. Paranjpe, Raghu Vira, V.K. Rajavade, N.B. Utgikar, P.L. Vaidya, V.P. Vaidya, H.D. Velankar, M. Winternitz, R. Zimmerman et al. 19 vols. Poona: Bhandarkar Oriental Research Institute, 1927–1959. Ma-linı-vijayottaratantra, Ed. Acharya Krishnanand Sagar. Varanasi: Krishna-nand Sa-gar. 1985. (1st ed. Madhusu-dan Kaul, Bombay 1922 KSTS 37) E-text by Somdev Vasudeva on GRETIL, and MIRI. Siddhayoges´varı-mata, edition based on NAK Ms. No.5–2403; Asiatic Society of Bengal, Calcutta. 5465 (G). V. Törzsök 1999 and Törzsök forthcoming. Svacchandatantra, with the commentary -uddyota by Ks.emara-ja. 2 vol. ed. Dvivedi, V.V. Delhi: Parimal Publications. 1985. E-text based on the original KSTS edition, MIRI.

Secondary literature Bakker, H.T. 2002. “Sources for reconstructing ancient forms of S´iva worship”. In Les Sources et le Temps, ed. F. Grimal, pp. 397–419. Pondicherry: IFP. Goodall, D. 2009. Lectures on the Nis´va-satattvasam . hita at the École Pratique des Hautes Études, Paris (Unpublished. Forthcoming as part of Nis´va-sa(tattvasam . hita) above). Hatley, S. 2007. ‘The Brahmaya-mala and Early S´aiva Cult of Yoginı-s’. Unpublished thesis, University of Pennsylvania, Philadelphia (Forthcoming from Routledge). Kiss, Cs. 2014. The Brahmaya-malatantra or Picumata II. The Religious Observances and Sexual Rituals of the Tantric Practitioner, Chapters 3, 21 and 45 (Forthcoming in the Early Tantra Series of the IFP/EFEO). Kreisel, G. 1986. Die S´iva-Bildwerke der Mathura--Kunst. Stuttgart: Franz Steiner Verlag. Panikkar, Shivaji K. 1997. Saptama-tr.ka-. Worship and Sculptures. New Delhi: D.K. Printworld. Sanderson, A. 1986. “Man.d.ala and Agamic Identity in the Trika of Kashmir”. In Mantras et Diagrammes Rituels Dans l’Hindouisme, ed. A. Padoux, pp. 169–214. Paris: CNRS Editions. Sanderson, A. 1988. “S´aivism and the Tantric Traditions”. In The World’s Religions, ed. S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp. 660–704. London: Routledge. Sanderson, A. 2001. “History through Textual Criticism in the Study of S´aivism, the Pañcara-tra and the Buddhist Yoginı-tantras”. In Les Sources et le Temps, ed. F. Grimal, pp. 1–47. Pondicherry: IFP. Sanderson, A. 2006. ‘The La-kulas: New Evidence of a System Intermediate between Pa-ñca-rthika Pa-s´upatism and Agamic S´aivism’. Indian Philosphical Annual, 24: 143–217. Tantrikabhidhanakos´a (Dictionary of Hindu Tantric Terms) vol. III. Ed. D. Goodall and M. Rastelli, forthcoming. Vienna: Österreichische Akademie der Wissenschaften.

50

Judit Törzsök

Törzsök, J. 1999. ‘The Doctrine of Magic Female Spirits. A critical edition of selected chapters of the Siddhayoges´varı-mata (tantra) with annotated translation and analysis’. Unpublished thesis, University of Oxford. See Törzsök, J. forthcoming. Törzsök, J. 2003. “Icons of Inclusivism: Man.d.alas in some early S´aiva tantras”. In Man.d.alas and Yantras in the Hindu Tradition, ed. Gudrun Bühnemann, pp. 179–224. Leiden: Brill. Törzsök, J. 2011. “Ka-pa-likas”. In Brill’s Encyclopedia of Hinduism, ed. Knut A. Jacobsen, Helene Basu, Angelika Malinar and Vasudha Narayanan, vol. III, pp. 355–361. Leiden: Brill. Törzsök, J. Forthcoming. The Teaching of Powerful Yoginı-s. A critical edition of the Siddhayoges´varı-mata with an introduction and annotated translation.

3

Varieties of melaka in the Jayadrathaya-mala Some reflections on the terms hat.ha and priya Olga Serbaeva

Introduction There exists in the Vidya-pı-t.ha Tantras a special term denoting all kinds of encounters with yoginı-s.1 The term in question is melaka or mela-pa. In the overwhelming majority of cases, such encounters do not pertain to normal, everyday practices, but rather come at the end of a recitation of a mantra and homa. They also invariably take place in desolate and dangerous places, such . as a mountaintop, or by a solitary linga or tree, or on a cremation ground.2 The sa-dhaka, usually in a state of inebriation, suddenly hears indistinct sounds, enters into a visionary state and then finds himself surrounded by yoginı-s. These may be deformed or appear with the faces of animals. On being addressed by these yoginı-s in a particular manner, he offers them a bloody argha by piercing one his limbs on the left side, usually his arm. Satisfied by the offering, the yoginı-s transform him into a superhuman being possessed of omniscience and the ability to fly, and who is frequently then compared to Bhairava. This would describe the most frequent pattern of melaka. In the present paper, I intend explore some of the particular instances of it as described in the Jayadrathaya-mala (JY). The Jayadrathaya-mala assembles material drawn from all sorts of earlier texts, some of which are now lost in their original forms. It consists of four s.at.kas, or parts, each of which is supposed to contain 6,000 verses. If s.at.kas 1, 3, and 4 still conform, more or less, to that number, a significant part of s.at.ka 2 has been lost. The first s.at.ka is the most archaic and the least S´a-kta in its language and orientation. It is possible that at least one century separates it from the remaining three. These, on the other hand, are very similar in style and all presuppose the existence of the others, leaving a strong impression of their having been compiled by one and the same person.3 The last three S.at.kas together constitute probably one of the most ‘S´a-kta’ texts in existence: the deities invoked are almost exclusively female, while Bhairava remains present only as a speaker and rarely appears in the man.d.alastructures. Yoginı-s of various kinds constitute a practitioner’s only means of access to a chosen goddess. It is not rare to find expressions such as ‘the sa-dhaka will become a yoginı-’, or ‘will achieve the state/domain of yoginı-s’.

52

Olga Serbaeva

The composition as a whole, influenced by the ideas of the Krama and the later Trika, is clearly pre-Abhinavagupta, for he cites the Ma-dhavakula – itself a semi-independent part of s.at.ka 4.4 But what is the purpose and usefulness of such a study of melaka in the JY? Two texts influenced my approach. The first of these is the Tantra-loka, which divides the varieties of melaka into two clear-cut types, respectively termed priya and hat.ha.5 Abhinavagupta associates the violent (hat.ha) variety with the fact that one should protect the ‘holes’ or chidras. Chidra is a technical term in the Vidya-pı-t.ha Tantras for vulnerable points through which it is believed that a person could be liable to attack by yoginı-s and similar sorts of beings/spirits. The other, priya, is characterised by ka-maca-ritva, meaning that one might accomplish it in accordance with one’s desire.6 In a key passage in chapter 28, Abhinavagupta links this procedure to different mantroddha-ras, or mantra-raising codes. He proposes, moreover, an interpretation of the process that is strongly non-dualistic: representing melaka as a union with the total consciousness or samvit, whose oneness, although it appears to be split through the separation of bodies, can nevertheless be grasped. The melaka in the TA constitutes a union with one’s own nature: precisely that total consciousness. The other text pertinent to my approach is the translation of Brahmaya-mala 99.10–13 by Shaman Hatley in his PhD thesis.7 In this, hat.hamelaka is presented as a practice fraught with great danger for the sa-dhaka, who, it would appear, might even potentially be killed by the d.a-kinı-s during the transformative process. Thus it became my aim to discover whether it would be possible to find the same clear-cut separation into priya and hat.ha types in the JY. I also hoped to explore all the meanings of the various types of melaka present and to discover if the JY is ultimately any more explicit about whether the hat.ha variety does or does not pose a mortal danger to the sa-dhaka.

Preliminary linguistic analysis of the collocation of the terms The hat.hamelaka Let us begin with a definition of hat.hamelaka. Hat.hamelaka is a practice in which the practitioner risks his life: for, according to BY, should he make the slightest mistake, or should his mental state at a given moment waver, or should some other such disturbance occur, then he will undoubtedly be devoured by the yoginı-s, and Rudra himself would not be able to save him.8 From a selection of early S´aiva tantric texts mostly belonging to the Vidya-pı-t.ha or reflecting its influence (UK, JY, BY, TST, TA, and NT), it would appear that hat.ha, in the sense of ‘force’ or ‘violence’, might refer to the following aspects of practice: (1) particularly violent yoginı-s, (2) the risk of death for the sa-dhaka,9 or even (3) to the procedure of melaka itself, which in this context has nothing at all to do with sexual union,10 but rather with the union with another through the

The Jayadrathaya-mala: varieties of melaka

53

consumption of his/her subtle and perhaps in some cases gross body. In relation to passages concerning killing of the pas´u and the extraction of his blood, all the texts in question share a particular feature: the practices are related not from the point of view of the sa-dhaka, who in such a context has a pretty good chance of ending up as the sacrificed pas´u, but from the point of view of the ‘aggressor’ – the yoginı-s. The procedures for draining blood or pra-n.a, extracting and consuming the subtle body, or simply killing the victim all have a similar basis: the manipulation of the channels of the pas´u. As I have argued elsewhere, these procedures, often encoded by the same set of mudra-s, are used by the S´aiva tantric guru in a successful possession-provoking form of dı-ks.a- as described in the texts of the Vidya-pı-t.ha.11 Let us now address the occurrences of melaka – of the hat.ha type as well as of and other types – in the JY in more detail, so as to discover which of these three presuppositions is confirmed by the text. In JY the word melaka and its variants occurs more than 250 times. Melaka or mela-pa does not cover all encounters with yoginı-s and there are numerous passages describing it without using the term – the linguistic marker in this case will be a-ya-nti yoginyo or a-ya-nti devata- – but the described procedure is the same. The term hat.hamelaka occurs ten times in total, in s.at.kas 2 and 4. In s.at.ka 2, chapters 10, 11, 15, and 25 are concerned. In s.at.ka 4, chapter 2 and chapter 30 are the relevant ones.12 In the last s.at.ka the term hat.hamelaka also occurs in chapters 52, 69 and 80. JY 2.10 provides a long list of sa-dhanas bearing the names of animals -(sim . ha, marjara, etc., all of which are animal-faced dutıs of the main goddess, who is Mantramantres´varı). The deities appear before the sa-dhaka after the performance of japa and homa accompanied with the most transgressive offerings. Among them, the ma-rjara sa-dhana is related to the hat.hamelaka.13 According to this procedure, having repeated the vidya- of the du-tı- called Ma-rjaravaktra- (Cat-faced) for 3 lakhs of times, and having offered a quasiastronomical number of mice into the fire, the sa-dhaka obtains siddhis of Ma-rjarı- (what exactly these consist of is not mentioned). Further, he is to proceed to a cremation ground to obtain hat.hamelaka, where he performs the ya-ga that calls Ma-rjarı- and her companions. The yoginı-s appear and ask him for blood offering. Having heard their request, the sa-dhaka pierces his left limb, whereby the yoginı-s grasp the argha and give him the caru, which is meat. In partaking of that substance, the sa-dhaka’s nature is transformed into that of vı-racakres´vara, the heroic leader of the cakra. The account of this particular procedure makes no reference whatsoever to the idea that the sa-dhaka might be in danger of suffering violence of any form on the part of the yoginı-s; and it must be said that the description of melaka here is, as a result, somewhat classical. - Chapter 11 is related to Maha-hum . karin.ıcakra in its colophon. The final passage of the chapter explains the hat.hamelaka procedure.14 The sa-dhaka, having attracted a human victim to secluded house, drinks his blood and eats

54

Olga Serbaeva

his flesh; by which means he worships Maha-hum . karin.ı, who is characterised in the dhyana passage as holding a human skeleton. He offers the victim’s flesh in the fire on the fourteenth day of the dark fortnight, and after his having made a thousand offerings, the devata-s appear and apparently grasp the flesh with terrifying sounds. They surround the practitioner and say again and again: ‘O hero! Lord of the choisest beautiful cakra!’ Having heard that, the sa-dhaka pierces his left arm and having given the argha, he flies up into the sky. He becomes Bhairava and is worshipped by the multitude of yoginı-s. The passage suggests that hat.hamelaka is a particular kind of melaka requiring a human sacrifice if it is to be obtained. In chapter 15 of the second s.at.ka we find a long list of those practices that are termed in the form ‘X plus dhvaja-’. In this case, dhvaja- appears to be a code word for dead body.15 Sim . hadhvaja is one of these, in which Bhairava explicitly states that he will now explain the horrid hat.hamelaka.16 Having gone to the terrible forest of the pitr.s, the sa-dhaka draws the six-angled man.d.ala on the ground, in the middle of which he worships S´iva- by means of dhu-pas made of eyes and teeth and which are placed upon a human head[s], heaps of flesh, and vessels of blood. He himself is inebriated, and, having finished that part of the ritual, he attracts the ‘great pas´u’, i.e. a human being, whom he grasps in the manner of Narasim . ha – tearing him apart with his nails, drinking his blood, etc. He laughs a terrible laugh and begins the uccara – the repetition of mantra in a particular way – over the first of the blood-filled vessels. Should that vessel manifest a particular sign, then he has obtained a yoginı- from the family which, apparently, corresponds to the direction in which the pot is situated. She appears in front of him, becoming his servant and saying, ‘O vı-ra, do that ya-ga,17 and you will obtain siddhis.’ The sa-dhaka prostrates himself in front of her, his head on the ground, and further offers her the argha from his own veins. He proceeds, moreover, with the performance of the ta-n.d.avamudra-, which attracts the siddhayoginı-s, who initiate him to the state of cakres´vara, in which he attains various pleasures. Here once more, hat.ha refers ‘only’ to the fact that the sa-dhaka performs the most repugnant and transgressive acts. The violence involved is perpetrated on the pas´u, not the sa-dhaka. The last occurence of the term in chapter 25 of s.at.ka 2, just mentions hat.hamelaka among other siddhis.18 In the fourth s.at.ka, chapter 2 deals with more than one hundred mudra-s, the majority of which provoke possession-like states and promise either visions of deities or melaka with yoginı-s. Only two of these mudra-s are related to the hat.hamelaka.19 The first of these is a hand-lock called kacchapa, which allows the practitioner to control cet.akas and kin.karas, and which is pleasing to all devı-s and Bhairava. It fills up all and bestows hat.hamelaka. The second mudra- is called mukula.20 Also a hand-lock, it summons deities in the very instant, because it is an excellent ‘vehicle for the mantras’, and it bestows hat.hamelaka. Chapter 30 of the fourth s.at.ka contains an important reference, because it plainly states, in the introductory part of the chapter – where the goddess

The Jayadrathaya-mala: varieties of melaka

55

provides a summary of the Tantras already heard by her – that the hat.hamelaka is hat.havedha, meaning the violent form of provoked possession, literary a ‘violent piercing’.21 The term vedha is more often used in the context of s´aiva initiation than in that of melaka. The chapter itself does not clarify the issue. Chapter 52 places the hat.hamelaka in a list of siddhis.22 While explaining how to obtain melaka with the whole multitude of the ma-tr.s by the use of a violent method (hat.ha-t), Chapter 69 proposes a more esoteric, internal interpretation of the process: the mastery of pra-n.a gives rise to a series of psycho-physical signs, such as visions.23 It is the mastery of pra-n.a – blocking it in a particular way by means of what appears to be a mudra- – that opens the way to the higher state, a state free from both nis.t.ha and anis.t.ha (stability and instabilty, or perfection and imperfection), and it is this which enables the sa-dhaka to experience melaka. Hat.ha-t here clearly constitutes something quite different from the usual types of melaka. For whereas melaka are commonly orchestrated by the yoginı-s, here it is rather the sa-dhaka’s mastery of the pra-n.ic flow which allows him to do what he wants with the ma-t.rs; and do it, moreover, in apparent safety. In chapter 80, the last clear occurrence of hat.ha, it is again a mudra- that bestows hat.hamelaka.24 All these references raise questions about the exact nature of the hat.hamelaka. The main characteristic of the hat.ha in the JY lies not in what happens to the sa-dhaka, nor in the type of yoginı-s which appear before him, but rather in the offering of human flesh and blood required for its practice. The occurrences of melaka in the last three s.at.kas taken together, however, rather argue against such a presupposition. Violence and references to human sacrifice appear also in the description of melakas not termed hat.ha or ghora, for example in JY 2.8.61–65. By the same token, maha-melaka is a term for a procedure related to the offering of human flesh and to the invocation of various kinds of yoginı-s.25 To muddle things even further, however, while there are numerous instances where a simple melaka, unqualified by ghora/ maha-/hat.ha, is described as involving offerings of the most transgressive substances, while in s.at.ka 3, maha-melaka is just a usual melaka.26 Thus, before we can begin to form any conclusions about hat.hamelaka, we should briefly look at the occurrences of the priya- variety. The priyamelaka There are eight occurrences in s.at.kas 2–4. In s.at.ka 2, Chapters 13, 17 and 26 are the ones concerned; in the first two instances, however, priyamelaka are merely included in lists of siddhis.27 Priyamelaka is one of many siddhis to be obtained by the worship of Is´a-nakalı. This form of worship requires an offering of human flesh; the only difference from the types previously described consisting in the fact that the goddess herself appears from the fire.28 Ka-mamelaka appears to be synonymous with priya-, but again the procedure as it is described is no different

56

Olga Serbaeva

from hat.ha.29 Only once, at the very end of the second s.at.ka, in a passage that is in every likelihood an interpolation added to complete a part previously lost, is priyamelaka explicitly sexual and occurs in pa-ta-la with divine women maddened by love – and apparently takes place, moreover, with some regularity (i.e. vı-raparvan.i, the special days of heroes, are mentioned).30 In s.at.ka 3, chapters 17 and 38 are those concerned. In Chapter 17, a single pada of a particularly long mantra is said to bestow priyamelaka.31 Chapter 38 states that knowledge of the secret signs is obligatory for obtaining it.32 In chapter 2 of the last s.at.ka, two mudra-s bestow priyamelaka.33 Priyasam . gama in this context appears to be a synonym of the priyamelaka. It is also mentioned in Chapters 69 and 75. If the first of these constitutes more a yogic procedure, the second is a collective sexual practice attracting deities.34 Thus in conclusion of our survey, of about 250 occurences of melaka only a small minority of them are qualified as priya- or hat.ha-. Nowhere in the JY is the hat.ha opposed to priya, or compared to it. The potential sexual character of priya-, moreover, is neither confirmed nor contradicted by JY’s descriptions. Melaka-manual in JY 3.38–39 But drawing conclusions about the relative meanings of the terms priya- and hat.ha- is in fact only impossible when taking JY a whole, and when the practices recounted in it are considered without any account being taken of their geographical or chronological origins. There do exist, on the other hand, two chapters in the JY which deal exclusively with technical aspects of melaka. These chapters conclude the third s.at.ka, and follow directly after a semi-independent part of the text called the Yoginı-sam . caraprakaran.a, which itself promises to explain melaka, but in fact fails to do so. In chapter 38, Bhairava provides a summary of the general rules of melaka; he lists appropriate meeting places; and emphasises the need for secrecy and respect for the rules: warning that the yoginı-s protect those who are respectful, but strip the fat and even the life from those who are not. Taking the life of an abuser of the rules is not presented here as a hat.hamelaka.35 Bhairava first, as if following the Tantrasadbha-va 16, provides a classification of the yoginı-s in accordance with the ma-tr.s. He also lists their subclasses, again in close accordance with TST 16, and further remarks that they are uncountable.36 Having, moreover, briefly listed the supernatural capacities of yoginı-s, and their practices,37 he then goes on to describe the physical features of the yoginı-s of the clans of the ma-tr.s and the external signs by which they can be identified – including the stories they respond to, their particular methods of extracting blood or killing, and the specific signs they paint on their houses. The entire passage follows the logic of TST 16, and the majority of these yoginı-s are dangerous; some indeed are said to extract blood, or even kill, by mere touch.38 Further, he explains the chommas: verbal signs that are dependent on the phases of the moon and related to particular families – the whole being encoded as a ka-lacakra.39 This is followed by the non-verbal signs: the

The Jayadrathaya-mala: varieties of melaka 57 vertical nya-sa that these comprise linking the families of yoginı-s to the tattvas, knowledge of which would allow priyamelaka to be obtained, v. 102ab.40 Fifty verses list the qualities of sa-dhakas deserving of melaka.41 Among other things, these require the ability to swallow anything at all as proof that a nondual state has been achieved. Verses 151–167 exlain the cilla-vrata, a part of which requires that the sa-dhaka should put on, or visualise (it is unclear), ka-pa-lika attire. He should behave in an antisocial, unpredictable manner and call himself a ‘ka-pa-lika eager to melt with the rays’.42 Finally, verses 167cd190 explain the ‘lunatic observance’, or unmatta vrata, which is obligatory for the obtainment of melaka, but which is extremely difficult to accomplish even for Bhairava.43 Once obtained, however, the sa-dhaka is able to see the deities within seven days, and gradually, within a month of conversing with yoginı-s, he becomes like Ka-lı- herself.44 This particular claim, however, is corrected in the very last verses of the chapter, where the final state is said to be that of Bhairava. The subject continues in the last chapter of the s.at.ka, called maha-mela-pasiddhi – or the chapter on the supernatural effects arising from the great mela-pa. Here, the goddess, worried that the vratas explained in the previous chapter are too difficult for the sa-dhakas, asks Bhairava to explain a sukhopa-ya method.45 Bhairava agrees and explains the priyamelaka that consists in attraction by means of offerings of transgressive substances to the various animal-faced deities. These deities are likely to share with the practitioner secret knowledge during the resulting intercourse, which knowledge allows the incomparable melaka to be obtained.46 Verses 23cd-28 describe the melakas that follow this basic priya, all of which involve a state of trance, termed in this case yoganidra-. It is precisely in this state that the sa-dhaka is to encounter the yoginı-s.47 Another variety appears in v. 29–54ab. This includes a ritual arrangement that is smeared with all sorts of transgressive substances obtained from one’s own body and which is considered to constitute particularly potent ‘vehicles for the mantras’; it is decorated with offerings and lamps which are, by the same token, made of a mixture of fish and human fat. The resulting construction is to be worshipped for seven days, after which the lamps will spontaneously light themselves, indicating the success of the ritual. If the lamps start to smoke, by such token the sa-dhaka will obtain the ability to travel instaneously to any place he wishes close by. He uses this ability to worship the yoginı-s in the places where they are most likely to appear, and obtain siddhi from them, including melaka. The second level of achievement is marked the apparition of heat, v. 44, which signifies that the sa-dhaka will gain the capacity to move through the air a distance of ten yojanas and thus obtain melaka with different classes of yoginı-s. By this same melaka he becomes omniscient and obtains the knowledge of the great samvara in the body, and also of invisibility. Verses 48ff. explain what happens when the lamps finally burst into flame. This signifies that the sa-dhaka’s present life is his last and that he has acquired the ability to move through the entire brahma-n.d.a; becoming like

58

Olga Serbaeva

Bhairava himself, able to curse and to bless, to create and to destroy the universe. He is served by thousands of yoginı-s, and has a constant supply of caru.48 The last verses of the chapter and of the s.at.ka explain the melaka proper, v. 54cd-80 – its uttama variety beginning at v. 58. The ultimate character of the melaka consists in the fact that it occurs wherever the sa-dhaka displays the mudra-s and repeats the mantra (rather than in the places where the yoginı-s fix it, as would usually be the case). The time necessary for each practice is reduced from days to a matter of mere minutes.49 The mudra-s, taken as complex ritual procedures including a number of elements, such as appropriate body-positions and movements, mantras, shouting, pra-n.a control, etc., are a particularity of the JY, and the overwhelming majority of melakas and visions of deities (which are virtually the same thing) are achieved through them. The passage, moreover, allows yet another possible interpretation of hat.ha variety of melaka as ‘violent’, this time to the yoginı-s, who are ‘forced’ to appear in the places and at the moments chosen by the sa-dhaka displaying those powerful mudra-s, whereas, in the priya- variety, he would be humbly waiting for them in such a place and at such times as they ‘love’ to manifest themselves. The unusual varieties of mela-pa Besides the aforementioned hat.ha and priya, other varieties of melaka exist. These appear only in the JY s.at.ka 4. Chapter 46 mentions rahasyamelaka.50 Although the du-tı- is needed for the preliminary stages, the melaka itself is more a yogic procedure, in which the yoginı-s (rays here) manifest themselves simultaneously inside and outside. Cı-llasam . yogamelaka as a particular kind of melaka is described in chapter 49; it is related to the vidya- of Mela-paka-lı-.51 Her dhya-na suggests that cilla- is a particular kind of bird, whose meat is also to be offered into the fire. The yoginı-s also take the form of birds. In addition to all of this, chapters 75 and 76 present melaka in sexual terms. Collective practice is called vı-ramelaka.52 However, these three unusual terms in no way constitute a viable alternative to the priya- and hat.ha- classification that is so limpid in the TA.

Conclusion to the analysis of collocation of selected terms If we are seeking a clear-cut linguistic separation of the priya- and hat.havarieties, then it must be concluded that the JY, taken as whole, does not provide it. Only a minoritity of occurences are qualified in the text by the terms hat.ha or priya. The general procedure, moreover, is essentially similar to both. As for evidence of the violence allegedly done by yoginı-s to the sa-dhaka, the JY does not confirm with any clarity that such constitutes the norm in hat.hamelaka. While it is true that the yoginı-s are generally bloodthirsty and

The Jayadrathaya-mala: varieties of melaka

59

they kill pas´us in order to obtain their own siddhis, the prescriptions for the sa-dhaka, such as respect for the rules, the maintenance of control over the mind, the injunction to cease japa and mudra- following certain particular signs, would appear to amount to no more than advice of a general sort, rather than instructions especially related to hat.hamelaka cases. Contary to the earlier tantras (SYM, BY, TST), in JY – as its compiler skilfully emphasises – it is rather the sa-dhaka himself who becomes the main orchestrator of melaka; the new, mudra--based practice allowing him to obtain melaka in any location (and not in given geographical spots) and at any time (rather than on particular moon-days), and within a very short period of time: days or even minutes – as opposed to a minimum of six months in the earlier texts. Although TST suggests some practices that may allow yoginı-s to be placed under one’s control, the approach is different from that encountered in the JY, where it is the sa-dhaka himself who can choose to force the yoginı-s to appear. More specifically it is through a set of particularly powerful, alteredstate-provoking mudra-s that the sa-dhaka is able to first gain access and then to obtain mastery over the yoginı-s.53 These mudra-s, which enable the combination of the tantric vidya-s with the yogic mastery of the subtle body (one’s own and that of others), constitute a feature peculiar to JY. Inside the JY, however, they coincide with almost every occurrence of melaka.54 It is precisely these mudra-s that make possible for both the sa-dhakas and the yoginı-s in all texts ranging from TST to JY to extract essences and even jı-vas from others. The fact that the JY links melaka with the mudra-s makes it an important step in the process of interiorisation of the melaka: a process already embryonic in the TST, and which goes on to become a self-evident reality in the works of Abhinavagupta. And, once the melaka has been so internalised – most often encoded by kun.d.alinı--related procedures – even the question of the difference between priya and hat.ha becomes irrelevant. In becoming a practice for the expansion of consciousness, the melaka loses its visionary external tantric character and becomes rather a non-dualistic philosophical expression of a quasi-absolute state.

Second reading: melaka as a visible ritual described in the Vidya-pı-t.ha texts Thus, finding myself ultimately unable to clearly distinguish linguistically the difference between the hat.ha- and the priya- varieties of melaka, I would like to try to approach the problem from a slightly different perspective. I propose here to consider the preparatory phase of the melaka as a visible ritual praxis. The clue that led me to the possibility of this new interpretation was an association of the priyamelaka with the knowledge of the secret signs.55 Let us then for the moment allow the suggestion that the difference between priya and hat.ha might lie in the role of these secret signs, which are proper to the former.

60

Olga Serbaeva

Bringing together all the passages expounding these signs (chomma) and the context of their usage,56 it becomes apparent that they fall into the following categories: (1) secret signs for the recognition of yoginı-s, i.e. when and how one can understand that a woman wearing this or that colour is in fact the one; (2) the means, verbal (secret language) and nonverbal (gestures), by which they may be conversed with; and (3) between the above two is the set of signs allowing knowledge of the ‘family’ of the yoginı-. The essential issues for our study of melaka, are the fact that the specific purpose of the secret language is precisely to discuss the time, place and conditions of participation in the melaka, and, even more importantly, the fact that all these yoginı-s are human women. If we assemble a glossary of all the words used in the many varieties of the ‘secret language’, it transpires that the melaka itself appears to consist in a collective partaking of meat and alcohol, and probably in a collective pleasure-giving ritual. There exists within that glossary, however, no special term that would designate any special state of consciousness that such a ritual would provoke for its participants. The ritual is, nevertheless, more than likely to culminate in some a-ves´a-like collective state of ‘bliss’, the vı-ramelaka.57 The description of the collective ritual in the Kula-rn.avatantra, moreover, appears to infer the very same idea.58 This question of chommas, and of their application to what essentially are ordinary human women, allows us to further advance the definition of the frontier between priya and hat.ha types of encounter. Indeed, the sa-dhaka invoking yoginı-s alone in a desolate place is placed in an altogether different ritual context than that of a collective tantric ‘party’. His ability to recognise the yoginı-s suddenly manifesting themselves before him is not predicated on the colour they are wearing; but rather on the sign they are unable to cross.59 The kind of beings he will encounter, moreover, is essentially different: these are non-human entities, grotesquely deformed, or having animal features, or even wearing the heads or feet of animals. A further essential difference should be noted: the bloody argha never plays any part in the context of the tantric feast,60 whereas it is a sine qua non element of the solitary yoginı--related practice. The feast lacks any tension between the sa-dhaka and the yoginı-s, while the descriptions of the solitary practice are full of it. Such solitary practice is risky, aiming as it does at attaining something which is virtually impossible – his acceptance by the yoginı-s as a dominating male figure, the leader of their cakra. This contrasts starkly with what would ‘usually’ transpire should a person find himself surrounded by yoginı-s: he would become their victim and be devoured by them. One of the best examples of such an orchestrated shift occurs in the Kubjika-matatantra 23, where the practitioner declares that he is a samaya-breaker (i.e. the one to be devoured by the d.a-kinı-s or sa-kinı-s in this sort of literature), he offers the dha-tus from his body to d.a-kinı-s, and expects a quick death.61 However, as soon as the yoginı-s appear, the sa-dhaka presents himself as Bhairava, their master, not their victim. Such cheating of d.a-kinı-s is, however, rather exceptional, and the idea generally conveyed in the earlier and more wild and visionary texts remains that

The Jayadrathaya-mala: varieties of melaka

61

found in melaka and especially in its hat.ha variety: the practitioner comes in contact with dangerous and savage beings, whom he would be well advised to avoid angering. Of course, one might well imagine that, having become the master of the cakra, the ex-sa-dhaka would thereafter feast daily in the company of subdued yoginı-s, i.e. the priyamelaka. The texts of the Vidya-pı-t.ha do not, however, elaborate to any extent on the subject. Table 3.1 below presents in summary form the ritual dimensions of the two varieties of melaka: Thus, if this proposed distinction of the priya- and the hat.ha- varieties is Table 3.1 The ritual dimensions of the two varieties of melaka PriyaKnowledge of secret signs Collective practice, i.e. sa-dhakas and yoginı-s (human) Eating, drinking and enjoyment ending in a collective altered state Tentatively, tib. tshogs kyi ‘khor lo, i.e. gan.acakra, i.e. a party with human yoginı-s Rather human yoginı-s, at least externally. They are to be found and recognised among human-looking women. Limited risk: mostly related to one’s status in society Does not presupposed a shift in nature The state of ‘bliss’ or trance does not appear to result in an essential, deep and drastic change of the nature of the practitioner, i.e. men and women stay human beings, can continue their life and participate in other ‘parties’ Regular practice

Hat.haProtection of one’s chidras/marmas Sa-dhaka versus yoginı-s Personal transformation in a solitary visionary state Tentatively, tib. gcod, i.e. ritual calling forth the demons and especially the d.a-kinı-s 62 Non-human, i.e. animal headed or deformed beings/spirits by definition High risk: the position of sa-dhaka itself means that he will either master the circle or be eaten by the yoginı-s Based upon the shift from a potential victim to Bhairava-state Yoginı-s transform the sa-dhaka into their equal, i.e. super-human being, possessing various supernatural capacities, or even becoming [like] yoginı- himself. Classical expression is, however, to be like a Bhairava Exceptional and often final practice

correct, the overwhelming majority of melakas described in the JY (and also in other Vidya-pı-t.ha texts, except specifically the chomma-passages) are in fact of the hat.ha type. The vı-ramelaka – the only clear description of ‘melaka with secret signs’ in the JY – belongs to a set of at least two chapters, of which the preceding one describes a sexual ritual with a female partner called a-diya-ga, while the subsequent one is, to all intents and purposes, the collective extension of the ya-ga, allowing the inference that it constitutes a regular practice.

62

Olga Serbaeva

The most interesting question, however – what exactly in the sa-dhaka in the context of wild and solitary practice is transformed by the yoginis? – remains unanswered. What is the essential difference between the human being and Bhairava? And by what means can that difference be erased? Why and how can yoginı-s effect such a fundamental shift in no time? Such questions divert us from the subject of melaka proper into the domain of the modification of the su-ks.ma by various yogic and tantric methods, and into the realm of vampiric practices and other strange and violent rituals. All these subjects are broad and complex enough that they require articles to be committed to each of them individually.

Concluding remarks consisting in some reflections on hat.hamelaka-related transformation I wish here to concentrate on the final stage of the melaka, i.e. the transformation of the sa-dhaka. Let us suggest that the terms priya- and hat.ha- might also refer to the manner of the change provoked in the sa-dhaka by melaka. In case of priya-, one can conjecture that the yoginı-s’ transformation of the sadhaka takes place in a relatively gentle, even ‘loving’, way – whether with or without any sexual connotation and corresponding practice. I would like, however, to concentrate here on the predominant hat.ha variety of melaka. I would tentatively suggest that hat.ha variety of melaka can be defined as that mode of unity of consciousness between two beings in which one is sucked/ transported/extracted and absorbed into the other; precisely the varieties of practices referred to in the UK 2 and NT 20.63 Although they are not specifically termed hat.hamelaka, the following instances from the JY support this idea. The end of chapter 9 of s.at.ka 2 – which is concerned with the worship of Kulacakres´varı- – presents melaka as an internal process: the expansion of consciousness provoked by powerful mantras combined with the extraction of the subtle elements from the bodies of others by means of the hela-dola- procedure and of the same mudra-s as those used in the initiation.64 The hat.hamelaka, taken to mean what we may call a ‘violent’ or even ‘vampiric’ mode of union, potentially contains at least three different aspects specifically related to that ‘violence’. The first is the imminent danger to the sa-dhaka, who must scrupulously observe particular precautions and proper reactions when the visual and audible signs begin to be manifested, for he risks his life in these encounters. The principal of these precautions includes perfect control of one’s mental state and respect of the code of behaviour with the yoginı-s, the offerings of his own blood, etc.65 The second, which, as we have seen previously, links the hat.hamelaka with human sacrifice, would suggest that the yoginı-s use the pas´u offered by the sa-dhaka with extreme violence, consuming the victim’s flesh or his subtle essences.66 But it is not the physical violence done to the victim that appears to be qualified by hat.ha. Rather it is the procedure of melaka itself: i.e. the

The Jayadrathaya-mala: varieties of melaka 63 subtle and gross body of the victim is consumed/absorbed by the yoginı-s. This is, of course, exactly the same as what in the Tantras the yoginı-s are believed to do all the time; and they can very well do it to the sa-dhaka too, should he make any mistakes or show disrespect for the rules. The JY permits us to suppose that melaka is not limited to sa-dhaka–yoginı-s relations alone, but might include a victim–yoginı-s modality too. This triangle – consisting of yoginı-s/victim/sa-dhaka – potentially casts BY 99 in a different light. Although the material contained in the JY does not provide any other clear statements that the yoginı-s may kill the sa-dhaka in the process of melaka, the idea in itself remains interesting and provocative. The sa-dhaka experiencing the assault of the yoginı-s would certainly be called a victim (pas´u), and thus we can reasonably apply those instances of pas´uyoga described in the Netra tantra 20 and in JY itself to such cases. Both texts suggest that the pas´u – killed/sacrificed and eaten/consumed during such practices – changes his ontological nature and either becomes one with S´iva, or is reincarnated with greater capabilities. Seen in this light, to be killed and devoured by the yoginı-s does indeed constitute a melaka, defined in the Vidya-pı-t.ha Tantras as being a powerful transformative practice. In addition to this and conversely, however, there is also a whole set of passages demonstrating that it is the sa-dhaka who extracts the subtle essences or kills in the same manner as yoginı-s, and it is he who then inherits those same capabilities precisely after melaka.67 While suggesting that it is the sa-dhaka who extracts the blood of other beings by yogic means, just as yoginı-s do, Chapter 23 of the third s.at.ka does not use the terms priya or hat.ha to categorise his multiple melakas.68 The chapters at the very end of JY again concern the sa-dhaka who has the capacities of yoginı-s and is able to extract blood and kill pas´us, especially the ‘seven-times-born’.69 But how should this be understood? Besides the obvious literal interpretation – that the melaka is – in fact – a carefully orchestrated transformation of a human sa-dhaka into a non-human blood-drinking monster, another option is also possible and latent in the text. The most interesting and ambiguous occurrences of melaka appear in the last s.at.ka to Chapter 33, which retranslates melaka as a yogic pervasion and out-of-body experience.70 Chapter 34 presents the external places for melaka as being located within the practitioner, thus the body itself becomes a sacred ground, where kun.d.alinı- awakens, and it is, moreover, precisely on the contact points of the channels that the sacred places are situated. In chapter 69, Bhairava is clear on the point that priyamelaka occurs when the thousand-petalled lotus is pierced.71 Thus and in conclusion we have discovered a gathering of multiple traditions in this one text: from the extremely physical description of human sacrifice performed by a sa-dhaka who is drunk and probably possessed – suggesting that a ka-pa-lika-like praxis was one source for the JY – to the yogic dhya-na with the yoginı-s in which the sacralised places of meeting with them are internalised – which would suggest that some more recent Tantras, clearly

64

Olga Serbaeva

non-dualistic and yoga-oriented in their nature, were also a source. The link, within this single text, between the yogic and the ‘tantric’ – in the sense of the most extreme and violent practices – remains unclear and would require an independent research project to explicate. Until such time as this research has been carried out, conclusions about the precise meaning of the terms related to melaka can only remain preliminary and partial.

Notes 1 For the classification of the S´aiva tantric traditions see Sanderson 1988/1990. 2 Serbaeva Saraogi forthcoming a, b. 3 This conclusion is based on the preliminary analysis of the frequency of syllables with the program Sanskrit_Parallels, created by Dmitry Serbaev and the author. Besides this, what holds the last three s.at.kas together is a similar structure of the Chapters and Subchapters: the explanation of the exceptional character of the knowledge to be revealed is followed by a coded vidya- and a dhya-na of the goddess; further, the technicalities of the ritual invocation are given: these include the construction of the man.d.ala or offerings; this part ends up with the appearence of the yoginı-s, transformation of the sa-dhaka and the list of siddhis he is supposed to obtain. Within the procedure of invocation, the JY emphasises the importance of the mudra-s. 4 TA, in chapters 15, 24 and 29; more on this in Sanderson 2007. 5 TA.28.371–374ab: yoginı-melako dvedha- hat.hatah. priyatas tatha- / pra-cye cchidra-n.i . sam . raks.et kamacaritvam uttare // sa ca dvayo’pi mantroddhr.t prasange dars´ayis.yate - nam prapadyate // tena tatparva tadvac ca sva/ yoginı-melaka-c cais.o’vas´yam jña . . - . santa-na-dimelanam / sam . vitsarvatmika dehabhedad ya sankucet tu sa // melake’nyo. nyasanghat.t.apratibimbadvikasvara /. The two variants of melaka are provoked by different mantras in TA.30.96–98ab. Abhinavagupta names TST as the source of it. 6 The word kama, or desire, may or may not have sexual connotations. Moreover, the subject and object of ‘desire’ remain unclear in the passage. In TST 16, the priyamelaka is called forth by a vidya- that sexually exites the yoginı-s (after v. 359), and it is that state, that apparently keeps them under control, and prevents their desire to harm the practitioner. However, this also appears a common way to deal with them, regardless of the kind of the melaka, see for example TST 16.239, and 329. There is also a similar expression in JY.2.17.410. 7 I am indebted to Shaman Hatley for his comments and suggestions concerning the preliminary version of this article. BY.99.10–13, Hatley 2007: 335 : mela-paka-s - - niyojayet tatha-nye ye hat.ha-f prokta- vara-nane / te tu d.a-kinivr.nda-na-m . na s´uddhanam . - pakam pra-pta-h sa-dhakendra- vara-nane / pra-pnuvantı-psita-n ka-ma-n vilo// evam mela . . . - - ma-t tu viparyayam // yatha- devı-padam . prapta vilomad. d.akinı bhavet / sadhako ’pi - rgena s´́ aktivijña-nagha-tana-t tatha- devi tanmadhye pas´uta-m vrajet // kim tu tenaivama . . . - - punah / muktva- deham ava-pnoti ja-tijñas / prabha-vena kula-na-m . tu so ’pi samanyatam . . ́ ca prajayate // 8 BY.14.215cd-220ab, Hatley 2007: 15–16: japet mantro maha-satvo digva-so daks.in.a-- mukham . // saptaratren.a yoginyo agacchanti mahabhayah. / raudrarupa tathas´uddha sakrodha maran.atmika // tat dr.s.tva tu na bhetavyam . vırasatvena mantrin.a / arghan - ta-sa-m . pradatavyah. pran.ipate kr.te sati // … pramada yadi ks.ubhyeta satvahınas tu sadhakah. / tatks.an.ad devi khadanti yoginyo yogadarppitah. // na ta raks.ayitum . s´akto rudro’pi svayam a-gatah. / 9 That is precisely the function of the yoginı-s as described in the NT.20, TST.16 (translated in Serbaeva Saraogi 2006) and throughout the text. In TST.16 the yoginı-s are said to obtain their powers by the list of four means: tapas, yoga,

The Jayadrathaya-mala: varieties of melaka

10 11 12 13

14

15 16

17 18 19

20

65

sacrifice of the janmapas´u, and by the vision of theirmantra (i.e. mantrasiddhi, when the deity of the mantra appears in front of the practitioner), v. 170 (ed. A. Sanderson): yogena tapasa- va-pi svamantrasya ca dars´ana-t / ya-nti yoges´varı-sam . stham . tatha- janmapas´or bala-t //. The same text, especially in chapter 7, teaches how to find and recognise such a pas´u, while Chapter 27 is a manual of blood-extraction. This is rather a predominant interpretation of melaka that is propounded in the book by David White (2003: 215 and 217), who presents even the hat.hamelaka in sexual terms. Serbaeva Saraogi 2010. Serbaeva Saraogi forthcoming a, b. The JY still awaits its critical edition. I reproduce the manuscript here ‘as is’, with minimal conjectures and corrections. JY.2.10.149–156ab [f. 29v6–30r1, P59rv]: - - ma-rja-ravaktra- can.d.ogra- jagatsam . harakarika / asya laks.atrayam . japyam . mahavı ranives´ane // mu-s.akotthasya sa-sasya juhed as.t.[-h]a sahasrakam / tada-sau siddhyate tasya ma-rja-rı- parames´varı- // pas´ca-d gacched vı-ravaro hat.hamela-pasiddhaye / … // - ya-vat prakrı-d.ate vı-ra- ta-van melapam a-pnuya-t / a-gatya yoginı- sarva- bhu-tam . bhutagratah. priye // bravanti sa-dhakendrasya dehya pas´ ca-ru[P: dehya pam /yam caru] . . - - - . bha-gakam / evam . s´rutva tatas tasam bhitva vamangam atmanah. // hatvargham . ca carum . pas´cad deyam . mam . sam . surarcite / tas´ carum . gr.hya vırendram . suddhasye['] bhyartha yanti hi // tada-sau sa-dhako vı-racakres´vara vibhur bhavet / [ma-rja-rısa-dhanam] JY.2.11.82cd-88, [f. 34r9–34v4, P68rv]: hat.hamela-paka-mna-yam . s´r.n.u para. mpara-gatam // prapu-jya s´u-nyaves´ma-dau maha-kanka-radha-rin.-ım / maha-pas´um . tadrajatva-X[cu-?/mu-?]llaya-gavidha-natah. // pı-tva- tadrohitam . [P: tallohitam . ] vıro - - bhaks.ayeX[ttu?] maha-mis.am / s´-ıs.tam . gr.hıtva bhutahni tis.ya yogotha va girih. // -ru-pam h[/k?]ara-n.kagam / sahasram juhuya-d ya-van huta-s´anam pratarpyeta devı . . . - tavad a-ya-nti devata-h. // gr.hı-tva- kalakala-ra-vam . bhıtabhıtas´ ca vihvalah. / avr.tya vı rarajendram bruvanti ca punah punah // ehi vı ra saha sm[/y?]a kam ramacakravar. . . . - - bhitva- va-man.gam a-tmanah // argham datvaes´vara / evam . s´rutva tatah.s tasam . . . mahorcasam utpataty avanıtalat / vicared bhairavo bhutva devatagan.apu-jitah. // I thank Alexis Sanderson who clarified this point in 2004, pers. comm. JY.2.15.210–219ab [47v5–48r1, P99v-100r]: hat.hamela-pam atyugram . s´r.n.u vı rendravandite / gatva pitr.vanam . bhımam . vidhanam idam acaret // s.at.kon.am . bhuvanam . tatra kr.tva madhye yajec chivam / naramun.d.opari sada dantanetrarghadhupakaih. // … // mahapas´um . s´amanıya pragr.hya narasim . havat / vidarya - - nakharaih. pı-tva- lohitam . phuks.iret punah. // haharavam . naded ghoram . prathamam . . pa-tram uccaret / yada- vijr.m . bhat sphurjeta tada tadyoginıkulam // agatya kinkaram . bhutva sadhakasya vaded idam / bho bho vırendra dehy ekam . bhagam asmac cat [/r?]u-ttamam // stanaya-gavidha-nottha- tatah. siddhim ava-psyasi / evam . s´rutva s´irah. - rgham a-dadet // ta-sa-m puras tu tacchı-rsam vı-rara-Xi nivedayet / pa-tya tad -ıyam ca . . . . - suta-ndava-n // astadha-pranaded ghoram datva- ta-sa-m . tada bhuyo badhva mudram . .. .. . pratidigbhımamañjasa / tatah. stah. siddhayoginyau cakres´atve narottamah. // niyojayenti deves´i maha-bhogabharolban.am / This ya-ga is apparently sexual because it is related to breasts (stana-), etc. JY.2.25.529cd, [f. 121v7–9]: hat.hamela-pasam . siddhir anyanokta maya hi sa // - s´eyam kin.kara-na-m tatha- priye / priya- ca JY.4.2.98–100ab: cet.aka-na-m prana . . . - sarvvadevı-na-m . devadevah. priya sada // yagadike ca yaddravyam . naram . bhavati parvvati / sarvvaprapuraty es.a devı nasyah. sama kvacit // vidyate varamudres´-ı hat.hamela-pasiddhida- / JY.4.2.130–134ab: mukula-khya- maha-mudra- s´r.n.u vı-rendravatsale / yayaprabandhaya- devi mantrakot.is´ata-ny api // sa-dhakobhimukha-h. sarvva bhavantı-ha surarcite / [hasta explanation] / mukula-khya- bhavaty es.a- mantra-va-hanasiddhida- // … pu-rvvavı-ryabalenaiva hat.hamela-pasiddhida- /

66

Olga Serbaeva

21 JY.4.30.32–33 [119r1–3]: anya- ca vı-ram atula- yogamela-pakarata- / anya vai s´am .vara-ja-la- ru-pabheda-ny anekas´ah. // hat.hamela-paka-s´ ca-nye hat.havedhap[/y?]asmr.ta- / avya-va sr.s.t.iyogena vedha-yantı- bala-t pas´u-n // 22 JY.4.52.11cd-16ab [f.152v], esp. v. 15ab:pres.aka-bhu-tasiddhis´ ca hat.hamela-paka-dikam . / . 23 JY.4.69.173cd-177ab [f. 194v]: tatrasthan kum . bhayet pran.am . yavad urddhapadam . labhet // [mudra-] // bha-vayed bha-vabodhena svabha-vam / tatrastho . ıs´varantagam . . laks.ayed dhı-ma- nis.t.a-nis.ta vica-ran.a-t // as´es.ama-tr.vr.m . dasya melakan kurute hat.hat / 24 JY.4.80.92cd-93 [f. 218r]: leliha-na- smr.ta- mudra- sarvvavı-rapraka-s´ita- // anaya- band- hama-tra-ya- hat.hamela-pata- bhavet / plavate gagano bhogam . bhairavakaravigrahah. // 25 JY.2.15.241–254ab [f. 48v3–49r2, P101v-102v]: s´r.n.u [']nyat parames´a-ni ghoram . . - s´akraks.ayankaram / vidha-nam adbhutam . bhımam mahamelapasiddhidam // gatva - jya- maha-ma-msa-sava-dibhih / sarvavı-ropaca-rena madira-nandananditah hara-syadam pu . . . . . // [243–245, dhya-na of the terrible goddess, most probably S´iva-du-tı-, 246–251 describe the horrid preparations, 252–254ab a mudra-] tada- gacchati deves´i sarva-kravya-dama-tarah. // svam . catam . [/rum . ?] gr.hya pratyeka kr.tva kolahalam . mahat / c[P: - - khr]et.ika-sa-kinı-bhis´ ca ru-pika- gocarı-kulam // dikcarı- d.a-kinı-na-m . ca khecarı bhucarı kulam / a-gatya svam . carum . tasya prayacchanti na sam . s´ayah. /. The same can be seen in JY.2.17.209–215 [f. 69r4–8, P148r], where after dhya-na, japa and homa Yoges´varı- appears in the fire, v. 209–210, while melaka with yoginı-s happened after that, and it requires the most transgressive things, such as a human heart, and a display of a mudra-. 26 JY.3.9.62–65ab [f. 71v3–5]: maha-mela-pam anvicchet tada- kurya-d ayam . vidhih. / - . - gatva- pitr.vanam . bhımam . pragr.hya s´avam uttamam // svalankr.tam . prasannatma - rayet tada- / tatprstha-ru-dha eva-sau japed vidya-m abhı-tavat // s´atocca-ra-d vedistham ka ... . . - - - eva devyo sama-gaccham . ti suvrate / tasam argham . pradatavyam . bhitva vamamgam a-tmanah. // tatas ta-h. siddhida-h. sarva-h. sadhaka-ya maha-tmane /. In JY.3.17.131–132 [f. 126r6–7], a single pada of a long mantra bestows the ability to receive knowledge in dreams from gods and goddesses, puts the yoginı-s moving in the space, earth and in between under control, and secures maha-melaka: ma-sama-tram . yada devam . hr.tpadmantar jayed yada / tada jñanam . svapnapade devadevya dadanti ca // vasagas tasya tis.t.hanti yoginya bhurbhu-vah. khaga-h. / maha-mela-pika-siddhi devasaukana ja-yate // 27 JY.2.13.29ab: maha-veta-lasam . siddhi priyamelapasadhanam / 28 JY.2.17.457–475, the offerings at the preliminary state do include human flesh, v. 460ab, 465cd. Priyamelaka belongs to the list of siddhis that is given by the goddess: JY.2.17.471cd-473 [f.76v7–9], esp. v. 471cd: pra-sena- vividha-h. sarva-h. priyamela-pam uttamam // 29 JY.2.17.486cd-491 [f. 77r7–9, P166rv]: mela-paka-ma dvau laks.au japtva- sa-dhakasattamah. // pas´ca-d gacched vara-va-sam . sadhakah. kr.tanis´cayah. /… tadyogam . sarva- dhakes´vare / … // yadi ghora-m das´a-[m]ru-dha-m smared mela-pe smarttavyam sa . . . . . . vidya-m ananyadhı-h. / ta-vad a-ya-nti yoginyo na-na-ru-pasahasras´ah. // va-ma-ngam a-tmanas ta-sa-m . bhitva[']rgham . sam . prakalpayet / tatas tus.t.ah. pranr.ttyanti sadhakam . varayanti hi // 30 JY.2.26.164cd-168ab [131r9-v]: unmadam . jayate tıvram . strın.am madanadıpanam // tatah. pravis.t.ah. pa-ta-le ta-sa-m . svamı bhavis.yati / mahabhogabharas´akto bhaved ´ brahma-yur eva sah. // atha va- s´rı-mukha-gre[']sau japam tis t hati nis calah . .. . / vıraparvan.i vırendras tadagacchati yos.itah. // tasa saha vis´ed devi patalam . bhogasagaram / - pu-rvavad bhogabha-gı-sya- sa-dhakah. sa-nugobalı- // evam . patalasam . siddhih. priyamelapam ucyate /. 31 JY.3.17.188cd-189 [128v8–129r1]: priyamela-pam anena sa-dhakem . dra labhet sada // yada kum . d.alinıcakre mantram a[/e?]tad udaharet / tada sam . sadhayet sarvvam . trailokye yatpratis.t.hitam // 32 To be discussed further.

The Jayadrathaya-mala: varieties of melaka 33 34

35

36 37 38

39 40 41 42 43

44

67

Maha-yoges´varı-mudra-, JY.4.2.150ab: baddhaya- tatks.an.a-d devi priyamela-pam . bhavet /, and vicitramudra-: JY.4.2.372cd-373ab: paradehapraves´a-rtham bandhaye va-ra-yogatah. . . // hela-dola-viha-ren.a priyasangamamelake / JY.4.75.60cd-63 [205v-206r]: trikon.abhagakum . d.a tal lim . gas´rucya juhed alam . // a-dyam . tacchukram atulam . tarpan.artham . tato juhet / evam . homam . samatis.thed ya-vad vai ks.an.ama-trakam . // tavad avis´ate dutim . kr.tva kam . t.haravam . mahat / snehat kaulikam evais.a kathayen na-tra sam . s´ayah. // tatkala eva sam . cara kramapurn.n.aih. praja-yate / priyamela-pam a-ya-nti devya-devyas tu sa-dhake // - JY.3.38.25–30ab [227r2–6]: sa-navaivaya-ga-ti niyatam . marmavam . ti ca / tatpuja- na-m tepi sures´a-ni samaya-m pa-layamti vam danam nityam kurvati mr galocane // pra . . . . . . . . - ta-h. / pa-dadispars´anam . des´e yagakhyenaiva tah. kvacit // des´ajñabhesamutpanne - - rudamti ca hasamti ca // … // devata-hasam . ti ca lalanti ca / mahamarakavelayam . . . nimdake mu-d.ho gurva-ca-ravikalpini / praharam . ti sada rasmyo haram . ti ca vasamis´am -ram ta-jı-vitam ca haramti hi / ´ // s´os.ayam ti s arı . . . . JY.3.38.34cd-35ab. JY.3.38.42cd-55. These verses imply that the yoginı-s in question are rather human, and they have to consume various trangressive substances as a part of their samaya. JY.3.38.56–86 [228r8–229r6], thus Brahmı- takes away life while singing, v. 60ab: - - ha-sa-rati sa- jı-vam . X geyapta natra sam . s´ayah. /; Mahes´varı is just called ‘greedy for anything’ (sarvalampat.a-) v. 64ab: vasa-mis.a-di haran.-ı sa- s´akta- nityam eva ca /; Kaumarı- takes away life, v. 66ab: lampat.a- kusuma-modeva- ca- jı-vam . haraty asau /; Vais.n.avı- is constantly pleased by drinking blood and she takes away jı-va by - - touching: 68ab: rakra-pa-napriya- nityam . spars´aj jıvam . haraty asau /; Varahı, v. 70cd, extracts some essence from the ‘holes’ and gives numerous diseases: cchidrair a-haret sa- tu vya-dhya-dibhir anekadha- //; Aindrı- takes away jı-va by sniffing, v. 74ab: - - - a-ghra-n.a-d harate jı-vam . kruradya nirghr.n.ahi sa /; terrible Yamı extract soul by - na-m sa- bhaya-nika- //; fat Ca-munda- takes ´ looking, 77cd: dr.s.t.ipa-ta-d harej jı-vam pas u . .. - away soul by kissing, v. 79d: cum . banaj jıvam aharet //; while Yoges´i likes dravyas and blood, v. 82cd; sadamr.takathasveva ramate sakr.t advyathe //. JY.3.38.87–95. JY.3.38.101cd-102ab [230r1–2]: cchommaka-di kramam . jñeyam . tridha trigun.am eva ca // sadhayet priyamelapa siddhaye sadhakas sada / JY.3.38.103cd-150. -JY.3.38.163–164 [f.232r6–232v2]: ka-pa-likosmi kam . kalı ras´mimelapalolupah. / sar- s´-ı vı-racakres´varo hy aham // evam va-dı- bhaven nityam vicared dhıvabhaks.o pipam ca . . . rara-d sada- / maha-vı-ryakala-ru-d.ho bhairava-ca-rapa-lakah. // JY.3.38.169–178ab [f. 232v2–233r1]: ks.an.am . haset ks.an.am . gayet ks.an.am . rodam . ks.an.am . radet / ks.an.am . plavet ks.an.am . narttet ks.an.am . dhavet ks.an.am . lalet // ks.an.am . s´antam . ks.an.am . vıram . ks.an.am . bhıtatsavad bhavet / ks.an.am . raudrarasavastho ks.an.am eva bhaya-nakam // ks.an.am s´r.m ga-rin.am devi ks.an.am ha-syaikatatparah. / . . . . . - dho ksanam ka-runyam a-sthitah // na-na-rasasama-visto na-na-bks.an.am angutasam ru . . . . . . . . .. - - ha-vasama-sthitah. / na-na-vila-sasam . yukto nanagıtaravakulah. // … // ks.an.an kros.t.ukaravısya ks.an.am . cillaravakulah. / hayehes.aravam api ks.an.am . gajaravam . vadet // . - - - - ks.an.am . marjararavısyat ks.an.am . sarasavam . nadet / gardabhakhyan ks.an.am . ravam . - tyas´vavam narah / naden na-dam ca kesaram // ks an am mam gacatus ken a valvaja . . . . . . . . . . ka-pa-liko[']ham . ca vades´ cam . d.alo[']ham . vadet punah. // ras´minatho[']smi matto['] smi ks.udho[']smi pralapa-my aham / - - JY.3.38.180–185: tada- sapta-ha ma-tren.a devı-na-m . dars´anam . bhavet / nanarupaih. kurupais´ ca yukta-yuktam . vadam . ti tah. // tatah. param . prayaccham . ti siddhadravyan.i yani ca / … tada sa plavate vyomni s´aX[mbhu?]ntam . yavad eva hi // mahabhairavana-thena pu-jyate sarvaka-likam / yatha- sa- parames´a-nı- ka-lika- bhairaves´vari // tadvad eva bhaved devi sa-dhako bahuna-tra kim / evam . vratottamam . devi dus´caram . bhairaves.v api //

68

Olga Serbaeva

45 JY.3.39.3ab. - 46 JY.3.39.13–22ab [233v9–234r6]: sam . tarpyatmanam evatra sadaghurn.n.italocanah . / . - - vrajed vı-ragr.ham . mantrı s´unyaves´mam athapi va // tatra puryamukhan khe[sve?] - vam . padam ekam . samuddharet / sphukirad yavad evasau tavad ayanti devatah. // s´vanama-rja-raru-pa-dyair na-na-ru-pair madotkat.a- / na-na-bha-s.a- vadam . ty etah. sadha- ganam / ka-gre n.a [/na]sam . s´ayah. // tabhis sarddham . tattvates´ ca plavate gaganam . . sa-dhako devadeves´i bhairaveva bhaved atah. // athava- tena madyena pı-tenaiva - hi ma-sa ta-h. / yoginı-melakam . jayet priyakhyam . sarvakamadam // athava pıtamadyas tu paryat.eda-pana-m . bahum / yatra rocam . ti bhasam . ti rudam . tis tu kadam . vakah. // narttanti va- maha-bha-ge tatra tadvat sama-caret / paravı-rye sama-yukto paramarmabalot.kakah. // evam . vai kurvatas tasya tatra ya yoginıpriye / sopasarpatitasyas´u prıyama- / tasyajña-naina-muda-nvita- // yad rahasyam ca tat tasya kathayen mithunena sa . - - valam . ghya syat sadhakena vijñanata // bhaven melapam atulam . yat surem . dres.u durlabham / - 47 JY.3.39.26–28 [234r9-v3]: yoganidra-m . tato yati sadhako[']sau mahabalah. / nidraga- na-ru-pam anekas´ah tas tada- so vai vı-ro pas´yan maha-matih. // devata-vr.m dam akhilam na . . . - - taram // tatra ta-bhih pu-jyate[']sau / ta-bhis sa-rddham . masamatrat plavate gaganam . . yatha- bhairavana-yakah. / sarvajñah. sarvakartta- ca maha-yoges´vares´varah. // 48 JY.3.39.39–53 [234v9–235v3]: s´r.n.u kramen.a vaks.ya-mi siddhasya-syaiva laks.an.am / dhu-ma-yitena subha-ge s´mas´a-nasthasya tasya ca // gudades´avyavasthasya khagatis trin.a yojana-t / yam . yam . sthanam . vim . cim . tya tam . tam . tam . yati ks.an.antarat // tatrastham . dravyam a-da-ya ks.an.a-d a-ya-ti vegatah. / tatrastha- yoginı- ya- tu bhu-carı- khecarı- thava- - // d.a-marika- siddha- milate sa- na sam . s´ayah. / kramajñanaprada sa tu sadhakasya na - su[/ma?]bha-ge s´mas´ tena s´rnuyatsivyate priye / maha-vı-ragrha-t tasthe ´ sam s ayah // u .. . . . khagatir das´ayojana-t // t[?bhi]intiram . dhraih. prayaty es.a sadhako natra sam . s´ayah. / - - - sahamela-pabha-g bhavet // yena mela-pakenaiva sa-dhakas bhu-carın.a-m . ca lamanam . sarvavid bhavet / hartta-kartta yatha- rudro bhavaty eva hi sa-dhakah. // antardha-nam . praya-ty es.a ma-nus.a-n.a-m . sada priye / mahas´ambaravijñanam . dehe casya prajayate // adhuna- jvalitasya-sya s´r.n.u vaks.ya-mi vai bhavet / etad vidulabhataram . vahanedrasya - pya janmana-pas´cimena tu / ya-ti brahma-ndam sa-dhanam // tatra-pi jvalitam pra . .. akhilam . gatis´ casya nirudhyate // tam . vahanıkr.tam . yena sadhakena mahatmana / sa bhaved bhairavah. saks.ıt sarvayoges´vararcitah. // s´apanugrahakr.d vıro sarvamela-pako dhepah. / sarvana-tha- sarvagatir maha-teja- maha-balah. // ru-pa-d ru-pasahasra-n.i ks.an.a-d eva karoty asau / laks.an.a yoginı-na-m . ca sevyate[']sau mahabalah. // mahacaru bhujo nityam . nityam . siddhiphalanvitah. / s´ivas´aktiyutas´vasau bhavaty eva va [na, serb] sam . s´ayah. // 49 JY.3.39, especially v. 57ab: tada- sapta-hama-tren.a sa-dhako melakam . labhet /; v. 63ab: tada- sapta-hama-tren.a vı-ro melapabhag bhavet /; 66cd: tada- sapta-hama-tren.a - sa-dhako melakam . labhet //; 67cd: tatres.t.hika yajed agre saptahan melakam . labhet //; 70cd: tada- mela-pabhyag -ısya-t sa-dhako ghat.ika-dvaya-t //. 50 The term rahasyamelaka appeares in JY.4.46.1cd in the question of the goddess. Bhairava explains the sexual character of preparation in JY.4.46.4cd-22ab [145v146r]: rahasyam amalam . guhyadevyahr.dayam uttamam . // rahasyam . sthanam a-sa-dya du-tı-yukto guru-ttamah. / madira-nam dacaitanyah kr to balividhikramah . . . . // - samuddharet / [v. 6cd-9ab, code of the pra-s´ya pañca-mr.tam . tatstham . tato vidyam . vidya-, followed by the mantratarpan.a and dhya-na of the goddess, v. 9cd-13ab] - sarvvavı-ropaca-ren.a pas´ca-d vidya-m . japen Xvai // tallayas tanmano bhutva mudram . - - ka-ma-tura-m . japet / evam . yavad atikranta muhurttas traya eva hi // vacadrasmikulam . caiva viva-se sa-dhakes´vare / saba-hya-ty antaram . sarvvam . vyaktim eti na sam . s´ayah. // mela-pam atra ja-yeta rahasya-khyam . mahadbhutam . / evam . sarvvasvavasthasu melapam . na vinas´yati // militasyasya melapam . melapamilitas tu sah. / s´ivavad ras´mimela-pa-n mela-pa iha kı-rttitam . //, etc. 51 Cilla-sam . yogamelaka appears in the question of the goddess in JY.4.49.1ab. Bhairava calls it maha-melaka and explains its nature in v. 21–37 [f.149v-150r]: s´r.n.us.va

The Jayadrathaya-mala: varieties of melaka

52 53 54 55

56

57 58 59 60 61

69

sam . pravaks.yami mahamelapam uttamam . / yena dhıtena subhage bhaved vıres´var´ esvaram . . // gatva haraspadam . ratrau tatra vidya[m . , serb] samuddharet / kakaladyars.ita-n kr.tva vadhuviti madhyagah. // tatoddharen maha-vidya-m . sarvvayogini[ı, - na, v. 23cd-29] evam dhya-tva- maha-raudra-m serb]kars.an.-ım / [code and dhya . . . - - msam pu-jayen na[lacuna]dale / pratyagre parames´a-nı-m . sarvvavıropacaratah. // cillama . . . - - juhet tatra[lacuna] sneha-ktam . nares´varam . / pas´cac caguhikakarya cillasankucyalu-khale // ba-lakesara samopeta-m dhu-pa-rtham vı-rana-yakah. / sahasra-mantrita-m kr.tva. . . . - namdace[/ai]tanyo darpitah parames´vari / tatra pas´ca-n gacchet pitr.gr.ham // madira . . .. vedi sama-s´ritya mudra-m . daha tatam . ratrau yavat sad. badhva karankin.ı // dhupam - vad a-ya-nti yoginyah cilla-ru-pa- maha-bala-h // kotis´o devadeves´i hakapum gavah / ta . . . . - -. kr.tva- ka-lakala-ravam . / tasam argham . pradatavyam . bhitva vamangam atmanah. // - naiva tam bhı-sayamti hi / abhı-stan tu varam pra-rthya ta-sa-m tatastra varayam ti ta . . . . .. . . gacchet sadaiva ca // yoges´varavapur bhu-tva- krı-d.ate yatra rocate / evam . tava sama-khya-tam . cillasam . yogamelakam . // The most important parts of Chapter 75 were edited and translated by Alexis Sanderson in 2007: 284–287, notes 176–177. JY.2.19.37–46 [f.91v8–92r3, P199v-200rv], especially v. 39: chinna-granthivika-s´a-tmapañcadha-mavika-s´akah. /, further, hasta is explained ending in a melaka with multiple sorts of yoginı-s. An article on mudra-s in the JY is now being prepared for the journal Tantric Studies. JY.3.38.101cd-102ab: cchommaka-di kramam jñeyam tridha- trigun.a-m eva ca / sa-dhayet priyamela-pa[m . ] siddhaye sadhakas sada //, most probably a reflection of TST.16.355cd-358ab (ed. A. Sanderson): cchommakais´ ca vicitrais´ ca bhairavair . angajaih. s´ubhaih. // sa-dhakena tato jña-tva- hr.dayena visarjayet / anyatha- kupyate - devyas´ cchidrayanti dura-sada-h. // evam . devi mayakhyata hat.han melapakam . priye / - - na kartavyam . vararohe pramadat sadhakena tu // tasmat kim . bahunoktena manasapi na cintayet /. This passage suggests that one makes priya out of hat.ha by satisfying the devı-s with the proper protocol (i.e. secret language, proper timing, proper composition of mantras). The subject of secret signs is very well presented in the Vidya-pı-t.ha texts, as for the BY; the Chapters 53 and 71 (Hatley 56, 74 and 93) are concerned, JY 3.38, and 4.71, SYM 27 provides their classification and insists that these must be transmitted orally and never written down; finally TST, chapter 16 to 19 also provide a lot of information on the subject. The term chomma is not used, but the same is decsribed in the concluding chapter of the SVT (chapter 15). Already discussed, Sanderson 2007. KT.11 throughout. The sign of their ‘family’, i.e. s´akti for Kauma-rı-, etc.; such practices can only be found only in TST 16. Although various transgressive and impure substances might be consumed during such feast. The practice is not called hat.hamelaka, but utkra-nti, consisting in a construction of a man.d.ala, in which the a-tman of the practitioner is placed in the midst of six terrible yoginı-s, who also receive the offerings of his own flesh, blood and other products of the body, and are attracted by this, KMT.23.132–138cd. The practitioner further by means of a particular vidya- invites the d.a-kinı-s to grasp the offering [v. 140ab-141]. He actually says that he wants to quit, and calls himself a - - samaya-breaker, v. 141cd-143: gr.hn.antv idam . maya dattam atrajña parames´varı // duh.khito ‘ham . virakto ‘ham . bhras.t.o ‘ham . samayojjhitah. / gr.hn.an tu devatah. ks.i- svaka-m tanum // a-jña- yadi prama-no ‘sti prama-nam yadi ca-npram . maya dattam . . . . . vayam / tena satyena gr.hn.antu matpradattam . marıcayah. //. Within seven days the yoginı-s appear and offer the siddhis of knowledge, v. 144–146: evam . viraktadehas tu ya-vat kurya-d dine dine / ta-vad a-ya-nti yoginyah. saptame ‘hani bha-sura-h. //

70

62 63 64

65 66 67

68

Olga Serbaeva jña-nasiddhiprasiddhasya saptara-tra-ntaka-vadhim / ks.apayanty anyatha- naiva s´-ı - - nigrahedam praka-rayet / tad ghram .. sam . harayanti tah. // atha ced dus.t.akarman.am . . a-tma-ngasamudbhu-tam . kiñcid dravyam . na gr.hn.ayet //. They apparently can kill him, and that is what they do with the being whose parts have been offered; that is why the sa-dhaka, if he uses this procedure for destructive magic, should not use his own dravyas. However, it is not a suicide proper, because what we read next is a surprising transformation (the translation of this passage by Goudriaan 1983: 101–106 is misleading). The sa-dhaka now presents himself not as a victim but as Bhairava, by means of dhya-na (both are ‘surrounded by yoginı-s’ in the texts), v. 147–148: brahman.a-lepyama-tma-nam . pas´cad dhyanam . niyojayet / paramatmasvarupo ‘ham . bhairavo - ‘ham . mahaprabhuh. // iti matva prayuñjıta s.od.hanyasam . svake tanau / kr.tva s.od.as´a va-ra-n.i tato vajratanur bhavet // Chöd Practice Manual and Commentary, 51–69. Detailed discussion of those passages can be found in Serbaeva Saraogi 2010. JY.2.9.49cd-59ab [f. 24v9–25r5, P49v, a folio preceeding that is missing]: sapta-ha-t pu-ja- bhru-ra s´aktis´iva-tmakah / na-c cakram . hr.dbhavam . parames´vari // vikasamaty asandeham . . tad eva melakam . subhrur bahir viharan.anta hi // vikasadasya cakrasya sarvamelapaka-dhipah. / bhavaty a vas´yam . vırendro sarvamantraran.ih. svayam //[P starts] evam . - - - vidhim . mahacakram . heladolaviharavat / saure tv avahayogena hr.nmandara[/man.d.ala] vika-s´anam // a-pivet pu-rvavidhina- para-dehotthitam . priye / s.ad.grasantena santarpya - - - bhairavı- rasalampat.a- // tada- khecarata-m . yati saptaharan mahabala / pıt.haram . sam . pra- tis.t.ha-pya jyes.t.ha-gram . pr.cchac chidritam // mahamatr.gr.he karyam . vidhim evam . mahes´vari / sam tatra maha-bhairavabhairavı- // vidhimet ka-rukam sarva-m . pujya purvavat . . . hela-dola-viha-ravat / pa-s´a-nkus´aprayogena ca-turdantabalena ca // vivasya-kars.ayed [d] ravyam . yattasya manasepsitam / vivaram . puritam . tena bhavaty asya na sam . s´ayah. // - tras´asyate / evam esa- sama-khya-ta- pindavyayamantraprakarttavyam sam cayam na . . . . .. dravyarasa-tmika- // a-kr.s.t.ir devadeves´i kim anyat kathaya-my aham /, also JY.2.17.606– . 610: [f. 80v2–9, P174rv]: vamangottha mahalaks.mı tatah. khacaratam iyat / nocen maha-caruh. sa-dhyo hela-m . dola viharavat // amr.takr.s.t.iyogena purvoktena tu sadhakah. / - tram prapu-rayet // tam pragrhya vrajed gauri s´matena-pu-rya svakam vaktram punah pa . . . . . . - rin-ım badhva- vidya-m samyag ihoccaret // saptava-ra-n s´a-nam . siddhasevitam / tatra phetka . . - . . tato devi svayam . devı karankin.ı / samabhyeti pradatavyam . carun tasyan nivedayet // tam . pras´ya devadeves´ı sadhaka svapuram . nayet / atha sam . varam akhilam . prabravımi tava-khilam // These aspects are described in TST 16, Serbaeva Saraogi 2006. This would explain a good part of the described melakas requiring the offerings of human flesh and blood, or even human sacrifices. JY.2.17.599cd-610, [f. 80v2–9, P174rv]: the first part of the passage describes a usual melaka, ending in the fact that the sa-dhaka obtains the ability to fly. From 606cd, it is a transformed sa-dhaka who extracts blood in order to go higher in the hierarchy of the melakas: nocen maha-caruh. sa-dhyo hela-m . dola viharavat // amr.ta-kr.s.t.iyogena purvoktena tu sa-dhakah. / tena-pu-rya svakam vaktram . . punah. pa-tram . prapurayet // tam . pragr.hya vrajed gauri s´mas´anam . siddhasevitam / tatra - phetka-rin.-ım badhva- vidya-m . samyag ihoccaret // saptavaran tato devi svayam . devı . - . karankin.ı / samabhyeti pradatavyam . carun tasyan nivedayet // tam . pras´ya devadeves´-ı sa-dhaka svapu-ram . nayet / atha sam . varam akhilam . prabravımi tavakhilam //. JY.3.23.143cd-166ab [f. 167r8–169r6]: melaka gives rise to the vampiric abilities, from v. - . - 148: pas´ca-t pidha-payec chidram u-rddhasam . put.agam . narah. / mahasthinas´ankuna va tato dhyanaratah. priye // tatra tis.t.hed atidignotac ca dhyanam ihocyate / ahr.tya pas´ayogen. allasvayec cankus´a-hatam // sravan.tam . s´ulasam . vidvam . cimtayet pas´uvigraham / heladolaviharen.a pran.aprekhanapas´anaih. // a-ha-radojasarvvakh[/sv?]am . tasyavaX khagasya ca / evam . dhyanaratasyasya hattakalalam . bhavet //. The extraction of the subtle essences of the victim allows the sa-dhaka to find finally the seven-times- born pas´u, who, being consumed, bestows the ability to fly: evam . pas´unam .

The Jayadrathaya-mala: varieties of melaka

71

- yatra devı- krsodarı- //. sapta-na-m .. . bhaks.an.ad amr.tam . priye / plavate gaganabhogam . Being himself transformed into a flying being, the sadhaka can now get the highest yogic siddhis from the khecarı-s: an.ima-dyas.t.akam . vatha khecaram . va padottamam / pa-ta-lam athava- devi yatha-va-tasya rocate /, while the caru, which, I believe, is still that seven-times-born being, makes happy all sorts of yoginı-s: evam . nam . dam . ti ta devyo yoginyo d.a-mara-s tatha- // d.a-ginyas´ caiva la-ma-s´ ca mahocchus.ma- cca ma-tarah. / u-rddhanis´va-sika-s´ ca-nya- hy adhonis´va-sika- parah. // nakracu-s.yas tatha-ca-nya- yoginyo ya-h. prakı-rtita-h. / ya-vam . tyah. s´am . vare tam . tre puraprokta suvistare //. The text refers to a certain S´am . varatantra as the full exposition of such practice. 69 JY.4.81.67cd-72ab [220v7–221v3]: maha-pitr.vanam . bhımam . mudrabandhasamacaret // tata a-ya-nti ta-h. sarvva-h. s.ad.sa-ha-[s?]ralam . pat.ah. / yoginya divyavapus.o bhogamoks.aphalaprada-h. // ta-sa-m argham . pradeyam . tadbhagam . vapi caruttamam . / evam . mela-pam a-khya-tam . pas´uyogan maya tava // melapartham . bhaved devi pas´usaptakabhaks.an.a-t / sarvvajña- ja-yate martyah. sarvvakartta- mahes´varah // tan na-sti yan na - badhva- puroktavidhina- tatah kurute tan na-sti yan na sa-dhayet / mudra-m . va kevalam . . - vı-rah khecarı-cakrana-yakah / // plavaty evam maha . . . 70 JY.4.33.21cd-26 [126v]: ta-m . pravis´yaprayogena akule s´as´vate pade // cittacaitanyayogena mantre sapta-das´a-ks.are / rodhayitva- caddhau spam . dau laya[/o?]day- - avivarjitau // tyajate pra-kr.tam . deham . aham ity evananyatha / purvvapas´cimakarman.i - t priye / na tesuramate prakr.ta-ks.ata-kr.ta- // vighnaru-pa-n.i te sarvva mohayam ti bala . . cittam . cittam . madhye nives´ayet // naham asti na canyosti kevalah. s´antarupake / - vayet // tada- sa khecaro bhu-tva- yoginıks.an.am apy atra vis´ra-mam sahajam yadi bha . . - krtva- pravis´ed dehaja-n bahun // melakam . bhavet / tanmadhye bhavanam . . 71 JY.4.69.128cd-132 [192v-193r]: sahasradalasam . vedhı priyamelapakarakam . // … / svejasam . cid api cintayet // matras´atanirodhena pas´yate ks.ir. tajalam . kr.tva na kim - asa-garam . / tadurmimadhyagam . padmam . dr.s.t.va urmikalaplutam . // tatrastham . bam .- ya-vat ksobhah praja-yate / ksudhvasya bhavate drstim divya-m dayet sam ... . . dhyam . . . . . jña-navibha-karı-m . //.

References Sources UK: Urmikaula-rn.ava tantra. Ed. M.S.G. Dyczkowski. MS 5–5207 (sic. 5–5202); NGMPP: B 115/9; folios 83; Newari script. E-text: Muktabodha, 04.2008. KMT: The Kubjika-matatantra. Kula-lika-mna-ya version. Ed. T. Goudriaan and J.A. Schoterman. Leiden: E.J. Brill (1988). E-text: GRETIL, 2005. KT: Kula-rn.ava Tantra. Text with English tr. by Ram Kumar Rai. Varanasi: Prachya Prakashan (1983). Kula-rn.ava Tantra. Ed. Taranatha Vidyaratna, M.P. Pandit and Arthur Avalon. Delhi, Patna: Motilal Banarsidass (1975). E-text: Muktabodha, accessed 2006. JY: Jayadrathaya-mala tantra. All four parts are transliterated by O. Serbaeva Saraogi, 2007–2009. The other and older (palmleaf) manuscripts of this text were kindly provided by Dr Diwakar Acharya; these are NGMPP A 995–996 and A 996–991 for s.at.ka 1, A 996–992 for s.at.ka 2, B 26–29 for s.at.ka 3, A 996–993 and A 997–991 for s.at.ka 4 (all palmleaf mss are marked P). JY.1 S.at.ka 1, NAK 5–4650, NGMPP B122/7, paper, Devana-garı- script. JY.2 S.at.ka 2, NAK 5/4650, NGMPP A153/2, paper, Devana-garı- script. JY.3 S.at.ka 3, NAK 5/1975, NGMPP A152/9, paper, Newari script. JY.4 S.at.ka 4, NAK 1/1468, NGMPP B122/4, paper, Newari script.

72

Olga Serbaeva

Tantrasadbha-va. Pat.ala 16. Critically edited from the codex unicus by A.G.J.S. Sanderson, 27.08.2004. MS NAK 5–445, NGMPP A44/2, folios 186, palmleaf, Kut.ila script. Pat.ala 16 occupies ff.108r3–118r1. Tantrasadbha-va, partially and provisionally edited by M.S.G. Dyczkowski, MS K 1–1985 S´aivatantra 1533, NGMPP A 188/22; folios 132; Devana-garı- script. MS Kh: 1–363 S´aivatantra, NGMPP A 44/1; folios 140, Newari. MS G: 5–445 S´aivatantra 185, NGMPP A 44/2, folios 210; Newari. E-text: Muktabodha, in 2005. The Tantra-loka of Abhinavagupta. With Commentary by Ra-janaka Jayaratha. Ed. Madhusudan Kaul S´a-strı-. KSTS 23, 28, 29, 30, 35, 41, 47, 52, 57, 58, 59. Bombay and Srinagar (1918–1938). E-text: Muktabodha, 04.2008. The Netra Tantram, with Commentary by Ks.emara-ja. Ed. Madhusudan Kaul S´a-strı-. KSTS 46 and 61. Bombay (1926 and 1939). E-text: Muktabodha, in 2005. Brahmaya-mala (Picumata). Ed. A.G.J.S. Sanderson, Oxford, 7 October 2004. Chapters 53–54 and 56. (MS NAK 3–370, NGMPP A42/6, ff.344, palmleaf.) Hatley, Shaman. The Brahmaya-malatantra and Early S´aiva Cult of Yoginı-s. Pennsylvania: University of Pennsylvania (2007). Unpublished Ph.D. thesis with edn of the selected chapters. Törzsök, J. Siddhayoges´varı-mata. The Doctrine of Magic Female Spirits. A Critical Edition of Selected Chapters of the Siddhayoges´varı-mata (tantra) with Annotated Translation and Analysis. Unpublished D.Phil. thesis and abstract. Submitted 1999, Merton College, Oxford.

Literature Dorje, Thekchok and Jamgon Kongtrul Lodo Taye. 2007. Chöd Practice Manual and Commentary. Ithaca, NY and Boulder, CO: Snow Lion Publications. Goudriaan, T. 1983. “Some Beliefs and Rituals Concerning Time and Death in the Kubjika-mata”. In. Selected Studies on Ritual in the Indian Religions, ed. M. Heerma Van Voss, E.J. Sharpe and R.J.Z. Werblowsky. Leiden: Brill, 92–117. Sanderson, A.G.J.S. 1990. “S´aivism and the Tantric Traditions”. In The World’s Religions: The Religions of Asia, ed. F. Hardy. London: Routledge, pp. 128–172. First published in 1988 in The World’s Religions, ed. S. Sutherland, L. Houlden, P. Clarke and F. Hardy. London: Routledge and Kegan Paul, 660–704. Sanderson, A.G.J.S. 2007. “The S´aiva Exegesis of Kashmir”. In Mélanges Tantriques à la Mémoire d’Hélène Brunner/Tantric Studies in Memory of Hélène Brunner, ed. Dominic Goodall and André Padoux. Pondicherry: Institut français d’Indologie/ École française d’Extrême-Orient. Collection Indologie 106, pp. 231–442 and (bibliography) pp. 551–582. Sanderson, A.G.J.S. www.alexissanderson.com. Serbaeva Saraogi, O. 2006. ‘Yoginı-s in S´aiva Pura-n.as and Tantras. Their Role in Transformative Experiences in a Historical and Comparative Perspective’. Unpublished PhD thesis, University of Lausanne. Serbaeva Saraogi, O. 2010. “When to Kill Means to Liberate: Two Types of Rituals in Vidya-pı-t.ha Texts”. In Grammars and Morphologies of Ritual Practices in Asia, Section I, ed. Axel Michaels and Anand Mishra. Wiesbaden: Harrassowitz, pp. 65–84. Serbaeva Saraogi, O. Forthcoming a. Yoga from Yoginı-s’ Point of View. (Proceedings of the 13th World Sanskrit Conference, Edinburgh, Scotland, July 2006).

The Jayadrathaya-mala: varieties of melaka

73

Serbaeva Saraogi, O. Forthcoming b. Can Encounters with Yoginı-s in the Jayadrathaya-mala Be Described as Possession? Forthcoming in the proceedings of the International conference ‘Yogini – History, Polysemy, Ritual’, held at NTNU, Trondheim, 1–2 October 2010. White, D.G. 2003. Kiss of the Yoginı-: ‘Tantric sex’ in Its South Asian Contexts. Illustrated edn. Chicago, IL and London: University of Chicago Press.

4

Snakebite goddesses in the S´a-kta traditions Roots and incorporations of Tvarita-, Kurukulla- and Bherun.d.aMichael Slouber

Introduction The prevalence of snakebite in South Asia coupled with the deeply rooted traditions of goddess-worship there has given rise to a variety of S´a-kta traditions of snakebite goddesses, some of which are popular down to the present day. Although the literature of and references to some of these goddesses is widespread, little scholarship has been done on their textual traditions. This lacuna is due in part to the fact that most of the early primary sources on goddess traditions remain unedited and unpublished, and therefore difficult to access. Over the last several years I have begun editing some of the surviving scriptures of the Ga-rud.a Tantras, the most fundamental and influential sources for the topic of snakebite medicine. The Ga-rud.a Tantras present us with a variety of traditions used by Ga-rud.ikas, practitioners who treat snakebite and other envenomings. The Ga-rud.ikas’ repertoire consisted of herbal medicines, mantras to the bird-king Garud.a, protective diagrams, and other spell systems such as vidya-1 goddesses. I have been keeping track of references to these goddesses and with this paper I would like to acquaint you with some of my initial findings on three of them – Tvarita-, Kurukulla- and Bherun.d.a-. Tvaritawas the subject of a vast literature, and though much of it has been lost, a substantial amount of material reaches us in manuscripts from Nepal and elsewhere. The literature on Kurukulla- is also large, although the early Ga-rud.a Tantra passages are consistently thin on detail. Bherun.d.a-’s references were widespread, but like Kurukulla-, they lack depth. I might have also told - -2 -. -3 you about S´a-kta snakebite goddesses such as Jham . karin.ı, Jangulı, or Suvarn.arekha-,4 for example, but for conciseness I limit myself to the abovenamed three, who became known to later tradition primarily through derivative traditions.

Tvarita‘The Swift One’ (tvarita-/tu-rn.a-/s´-ıghra-) is a fitting title for a goddess whose most celebrated function was saving the lives of those bitten by venomous

Snakebite goddesses and S´a-kta traditions

75

snakes such as a cobras, vipers, or kraits. Her ultimate origin may be lost in antiquity, but the earliest surviving source is one of the most widely cited canonical Ga-rud.a Tantras called the Trottala.5 In time she was identified with various goddesses of extensive renown: Kubjika-, Durga-, and Ka-lı-6 in the wider S´a-kta traditions, as well as Padma-vatı- in the Jaina Tantras.7 In the introduction to his massive fourteen-volume edition and translation of the Mantha-nabhairava’s Kuma-rika-khan.d.a, Mark Dyczkowski has an eight-page section on Tvarita-. His discussion is very informative, and he points to a long and important section of the unpublished Kulakaulinı-mata that discusses Tvaritaat length.8 I also discovered a Tvarita-vidha-nasu-tra ascribing itself to the 9 Caturvim . s´atisahasrasam . hita (i.e. the Manthanabhairavatantra), but it is incomplete and damaged and I have been unable to trace it in other manuscripts of this massive work. We have many texts that know Tvarita- as an autonomous snakebite goddess, but it seems probable that most or all of these derive from the aforementioned Trottala, so I will only briefly mention them in the section below on ‘Borrowings’. Since we have several manuscripts of texts which ascribe themselves to the Trottalatantra, let us turn to them now and explore what Tvarita- was like in these early sources. Tvarita- in the Trottalatantra I am aware of two surviving works that ascribe themselves to the Trottalatantra: the 700-verse Tvarita-mu-lasu-tra and the 200-verse Tvarita-jña-nakalpa. Both use the titles ‘Trottala’ and ‘Trottalottara’ interchangeably, although the latter is listed as a separate text in the canonical lists. The Tvarita-mu-lasu-tra positions itself as an extraction from the (presumably mythical) version of over 100,000 verses. I have introduced the text and edited and translated most of the first chapter for Somadeva Vasudeva’s forthcoming Tantrik reader S´ivasudha-prapa-pa-lika-, so here I will dispense with introductions and describe the most salient features of each of its nine chapters. The origin story of Tvarita- in the first chapter is a rather grand tale in ninety-four verses. Some stanzas are directly parallel to the similarly named Nis´va-samu-lasu-tra, and presumably the Tvarita-mu-lasu-tra redactor has that text in mind. The frame story is that Garud.a approaches S´iva after many aeons of penance and requests teachings on a variety of topics. As an aside, I should remind you that Garud.a is an independent deity in the S´aiva Tantras and is almost never associated with Vis.n.u, despite the present-day popularity of his Vais.n.ava identity. In response, S´iva begins to narrate Tvarita-’s origination story to Garud.a: a band of yoginı-s, themselves originated from S´iva’s Bhairava form, approached S´iva and requested that he preside over their caru ritual, which, the context implies, involved sexual rites. When Gaurı- learned that S´iva was engaged in this ritual, she became dreadfully angry and generated herself as Tvarita- by uttering the nine-syllable Tvarita- vidya-. Here she has eighteen arms, which the text reminds us correspond in pairs to the nine syllables of the vidya-, and she is seated on a lion. This mount brings to mind

76

Michael Slouber Durga-, but the similarity stops there and whether Durga-, Gaurı-, and Tvaritaare three goddesses or one is a matter for another paper – here I treat them as separate personages.10 Several verses are devoted to describing her fierce lion, and then we return to Tvarita-’s own appearance:11 Tvarita- is mounted on [the lion] with her left shank hanging down and the right folded in so the sole of her foot touches her leg. The nails on her feet sparkle like twinkling stars, are lotus red and marked with lotuses. The soles of her feet are beautiful with long and symmetrical toes. She has well-rounded thighs and fleshy calves, and her buttocks and hips are broad. She has a deep naval, a belly with three folds, and is beautified by a streak of abdominal hair, broad hips and breasts like golden pitchers. She has a shell-like neck, lips like the Bimba fruit, eyes like blue lotus petals, a beautiful nose, curved bow-like brows, attractive ears, and she glows with a crown. She shines with snake earrings and a necklace consisting of the king of snakes, brilliant with a thousand hoods and radiant with glittering jewels. The goddess’s arms shine, beautified with snake bracelets, and likewise the girdle on her hips, frightening because of being made of a great serpent. On the feet of the goddess are two terrible snake anklets, and her left foot is placed in the middle of a thousand-petalled lotus. One garment, adorned with flowers, looks like forked lightning in the sky. Another garment is like a rainbow draped across her breasts. It is studded with numerous flowers, and tied tightly with a great serpent. The goddess’s hands carry weapons and the tips of her shell-like nails are sharp. On the right she holds a vajra, staff, sword and discus, a mace, a shining spear, arrow, and javelin, and she displays the gesture of granting boons. On the left she holds a bow, noose, sword, bell, a threatening finger, a conch and a goad, and she also displays a gesture of goodwill and holds a lotus. This elaborate description of Tvarita- contrasts with the simple two, four, or eight-armed forms which Dyczkowski references in the Kulakaulinı-mata and Tantrara-ja,12 a point to which I will return in the section below on ‘Incorporations’. The Tvarita- spell itself is an interesting topic, and here we will briefly describe some features of it. The mu-lavidya- usually consists of three times three syllables: HU-M. KHE CA CCHE KS.AH. STRI-M. HU-M. KS.E PHAT., and often it is prefixed with om . . This is the form of the vidya that we get spelled out plainly in 13 the Tvaritajñanakalpa. It agrees with a versified enumeration in the same text’s twenty-fourth verse.14 In the Tvarita-mu-lasu-tra’s second chapter we are given the syllables of the mu-lavidya- in a simple code based on the standard layout of the Sanskrit alphabet:15 The last of the heated group with a dot and mounted by the sixth vowel - ). The first seed syllable of the basic spell has been told, O (h+u-+m . =hum .

Snakebite goddesses and S´a-kta traditions

77

Lord of Birds. Now, the second of the [soft-]palatal class joined with the eleventh vowel (kh+e=khe). The first [vowel] in conjunction with [that of] the tongue-palate [class] should be alone (c+a=ca). Beneath that same one, one must use the second of that [class] with the eleventh vowel (c+ch +e=cche). Now one must join the second of the heated beneath the first of the [soft-]palatal class together with the sixteenth vowel (k+s.+ah.=ks.ah.). One must use the first of the tongue-teeth (t) below the third of the heated (s) and the second of the mixed class again combined beneath with the fourth vowel (s+t+r+ı-=strı-). One must use the second of the heated joined beneath to the first of the [soft-]palatal class combined with the eleventh vowel (k+s.+e=ks.e). The last of the heated together with a dot and mounted by the fifth vowel (h+u+m . =hum . ). And the second syllable of the labials is another to be joined to that which is first when the tip of the tongue touches the palate; this is to be extracted with a half vowel (pha+t.+a/2=phat.). This is the Exalted Spell-Goddess Tvarita- who grants all success. She should be prefixed with om . and should always have ‘homage’ (namah.) at the end. For fire rites it ends in sva-ha-.16 Thus the vidya- given here is: (OM.) HU-M. KHE CA CCHE KS.AH. STRI- KS.E HUM. PHAT. - - (TVARITAYAI NAMAH . /SVAHA). This differs from the Tvaritajñanakalpa version in several respects: the syllable strı lacks anusvara, the syllables ks.e and hum . are reversed, and the syllable hum in the eighth position has a short vowel. The . significance of these differences is not currently apparent to me, but we may - are elsewhere generally interchangeable, with the note that hum . and hum . latter occurring only moderately more frequently than the former. The fact that the author went to the trouble to spell out that the first is with the sixth vowel whereas the second is with the fifth vowel suggests that the difference was significant in his tradition. Next the Tvarita-mu-lasu-tra teaches the ancillary mantras: a three-syllable Heart mantra, a Head mantra with an unclear number of syllables, a five-syllable Crest mantra, a five-syllable Armour mantra, a mantra to the three Eyes whose number of syllables is unclear, and a four-plus-one-syllable Weapon mantra, whose first four syllables correspond to weapons placed in the four cardinal directions. The specific syllables of these ancillary mantras, where they can be clearly determined, differ from those plainly enumerated in the Agnipura-n.a parallel, so for the time being I will leave this puzzle unsolved. The text emphasises that without these ‘secret ancillary mantras’, one cannot have success with the Trottalatantra. Thereupon the Tvarita-mu-lasu-tra, still in the second chapter and starting with verse 38, teaches the simpler ‘vidya--ancillaries’:17 The first and second are the Heart. The third and fourth are proclaimed to be the Head. The fifth and sixth are taught as the Crest. The seventh

78

Michael Slouber and eighth are the Armour. The star syllable (phat.)18 is the Eye qualified with its half-syllable as being the ninth.

So we have the following scheme for the vidya-’s ancillaries: Heart hu-m . khe Head ca cche Crest ks.ah. stri Armour ks.e hum Eye phat. Next the Tvarita-mu-lasu-tra teaches the ten-syllable spells for each of ten female attendants (du-tı-): S´akra’s Vajratun.d.a-, Agni’s Jva-linı-, Yama’s S´abarı-, . Nirr.ti’s Kara-lı-, Varun.a’s Plavangı-, Va-yu’s Dhu-nanı-, Kubera’s Kapila-, Rudra’s Raudrı, Vis.n.u’s Cakravega, and Brahma’s Brahmaveta-linı-. Their vidya-s are characterised by beginning and ending with the respective syllables of Tvarita-’s mu-lavidya-, thus each syllable represents one of the female attendants. The way this works out to fit nine syllables to ten attendants is that phat. is taken to be two-in-one, so the pha element is Cakravega- and the t. element is Brahmaveta-linı-. The other syllables of the attendants usually include their name in the vocative, sometimes alias names, and either individual syllables or imperatives appropriate to each. Thus Jva-linı-, the attendant of Fire, is told ‘blaze!’ and Manovega-, the attendant of Wind, is told ‘go!’ The first eight attendants clearly correspond to the eight compass points starting in the east and they are placed in this configuration around Tvarita- in many of the rituals taught in the Tvarita-mu-lasu-tra. For example, in the eighth chapter there are instructions for making a ‘Vajra-bolt’ (vajra-rgala) diagram, here for the purpose of killing an enemy, using a circuit of the first eight female attendants surrounding Tvarita- on the petals of a lotus.19 TheVajra-bolt is Tvarita-’s signature man.d.ala. It is the same one that S´iva was presiding over with the yoginı-s in the first chapter, and that which Tvarita- was persuaded to enter by all of the terrified gods. My summary has covered most of the first two chapters of the Tvarita-mu-lasu-tra. I will now very briefly look at the contents of the remaining seven chapters. The third teaches the installation of Tvarita-’s weapon mantras on the hands and body of the mantra practitioner, which affords him invulnerability from gods, demons, or any evil influences. The fourth Chapter is on mudra-s – hand gestures used in the worship of Tvarita- and rituals involving

Snakebite goddesses and S´a-kta traditions

79

her. Twenty-eight mudra-s are described, many corresponding to the eighteen weapons/gestures in Tvarita-’s hands. The fifth chapter teaches initiation. The ritual begins with an elaborate worship of Tvarita- in the Vajra-bolt man.d.ala, with her mounted on a five-faced S´iva acting as her throne. This pose is probably meant to demonstrate her superiority to Sada-s´iva, the prototypical five-faced S´iva of the Siddha-nta Tantras. The initiation also involves ritual generation of fire in a vulva-shaped pit and offering grains and ghee into it while reciting the basic vidya- along with the ancillary spells. At one point the text says ‘And he becomes initiated by just one oblation, O Bird; in this way he would be authorised. Now listen further concerning liberation.’20 If my understanding is correct, this ritual departs significantly from the S´aiva norm where the most basic initiation grants liberation and further initiation is required for those seeking powers. Against this interpretation is the fact that the opening of the chapter calls the initiation both power-granting and liberation-granting. Many benefits of initiation are listed, such as obtaining a kingdom, success with mantras, destruction of poverty, and obtaining sons. At one point there is a choice to either dismiss the goddess and dismantle the man.d.ala, or for those who are authorised to continue with offerings that include animal and human blood. The goddess is praised as present in a long list of deities, in fact, as all-pervasive. The Chapter closes with a fascinating discussion of who may be initiated and who is unqualified. That chapter will be a fruitful source for future research. The briefer sixth chapter gives instructions for locating a site on which to practice, ranging from a dreadful cremation ground to cities, towns, or villages where people are predominantly S´aiva. Instructions are given for several basic rites that I will not discuss here. Chapter 7 begins and ends on the topic of different extractions of the syllables of the vidya- for various purposes, but most of the chapter is rather a detailed description of the creation of various deities culminating in Tvarita-’s appearance. The details of this creation story would certainly be of comparative interest to other scholars of S´aivism and S´a-kta traditions, but space permits me to note only a few features. The basic image is one of chaos in the universe with various exceedingly powerful forces coming into existence and clashing. A battle between Garud.a and Vis.n.u ends with Vis.n.u being vanquished and leaving the egg of Brahma. The chaos does not come to an end until Trotala-, Tvarita-’s nom de guerre, is established as the protectress (tra-yaka-) and terrifier (tra-saka-) of the world. This serves as a folk etymology of her name. The creation story in the first chapter has little in common with the one here, so we must assume that the text is preserving two separate accounts of her creation. The long eighth chapter is a collection of various practical applications (prayoga) of the vidya-. We see instructions for making magical diagrams (yantra) on funerary cloths, skulls, or less grim, walls and leaves. Goals include the standard black magic actions like killing an enemy, sowing dissension, driving a rival out of town, or controlling women; white magic actions such as creating peace and well-being; royal work like defeating an enemy army; and more

80

Michael Slouber

Figure 4.1 A twelfth-century folio of the Tvarita-mu-lasu-tra in the Kaiser Library, Kathmandu

specialised actions like destroying possessing demons and fevers. We also see, of course, several rituals for destroying poison and healing snakebite victims. We will mostly pass over the ninth and final chapter on yoga, as I am not experienced in this subject. Suffice it to say that here it involves meditation, breath control, and visualisation culminating in a vision of Tvarita-. The Tvarita-jña-nakalpa, which I have mentioned several times already, is very parallel to the Tvarita-mu-lasu-tra. It is only 200 verses in extent,21 and the colophon places it as the thirty-fifth chapter of the 11,000-verse Trot.alottara. We have three Nepalese manuscripts of it and the earliest is paleographically similar to manuscripts from the eleventh or twelfth centuries. It is not obvious from the parallel passages whether the kalpa depends directly on the mu-lasu-tra, but it is certainly abbreviated and shares many verses. It is notable in giving various applications (prayoga) of the Tvarita-’s basic syllables, sometimes using only a few of them, and sometimes more. These applications feature snakebite cures much more centrally than the mu-lasu-tra, although other topics are also given. Borrowings We see Tvarita- material in many other texts, but we have also determined that little of her grandeur in the Tvarita-mu-lasu-tra carries over in subsequent literature. I will discuss her other identities shortly, but here I would like to point out some parallel passages in the Agnipura-n.a that are clearly dependent on the Tvarita-mu-lasu-tra: The column on the left is extracted from forty-two verses in the Tvarita-mu-lasu-tra that are redacted as only five and a half verses in the Agnipura-n.a. This continues, with the next line in the Agnipura-n.a (310cd) picking up with Tvarita-mu-lasu-tra 2.1. I have given the Tvarita-mu-lasu-tra’s Sanskrit for verses 2.11–20ab in note 15 on Tvarita-’s mu-lavidya-, which one can compare to Agnipura-n.a 310.10–18. It is mostly word-for-word except in lines like Tvarita-mu-lasu-tra 2.11 where the vocative ‘O Lord of Birds’ (khages´vara) did not fit the context of the Pura-n.a and the line was simply dropped. In several cases the redactor tried to clean up the non-standard Sanskrit forms like u-s.ma-n.asya by changing it to u-s.man.as´ ca, but in a few cases we see the redactor corrupting the sense even further; for example: ‘mounted by the fifth vowel’

Snakebite goddesses and S´a-kta traditions

81

Table 4.1 Parallel passages in the Agnipura-n.a that are dependent on the Tvarita-mu-lasu-tra Tvarita-mu-lasu-tra 122 Agnipura-n.a 31023 as.t.a-das´abhuja-m as.t.a-das´abhuja- devı. sim . he . - sanasthita- // 41 // va-majan;gha- pratis.t.hita- / daks.in.adharmasim ha . dvigun.a- tasya-h. … . pa-dapı-t.he samarpita- // 3 // tvarita- tatra ca-ru-d.ha- va-majan;ghana-gabhu-s.a-m pralambita / . vajradan.d.e khad.gam . - krama-t / cakram daks.in.a- dvigun.a- tasya-h. pa-dapr.s.t.he . gadam . s´u-lam samarpita- // 45 // . s´aram . tatha s´aktim . varadam . … daks.in.aih. karaih. // 4 // - - tarjanıvajradan.d.a-sicakram dhanuh. pa-s´am s´aram . ca gada s´ulam . . ghan.t.am . . .. ´ ´ mahojjvalam / m s an ;kham an ;kus am / . s´aram abhayam . s´aktis´ ca varadam . daks.in.ena . ca tatha vajram . vamapars´ve kr.ta-yudha- // 55 // dhr.ta-yudham // 5 // - . dhanus.pa-s´adharam pu-jana-c chatruna-s´ah. sya-d ra-s.t.ram . ghan.t.a tarjanı s´an; . . kham an;kus´am / jayati lı-laya- / abhayam dı-rgha-yu- ra-s.t.rabhu-tih. sya-d . ca tatha padmam . vamapars´ve kr.ta-yudhadivya-divya-disiddhibha-k // 6 // taleti saptapa-ta-la-h. … - la-gnibhuvana-ntaka-h / ka yas tu pu-jayate bhaktya- gr.he nityam . . - om svas´aktitah. / . karadisvararabhya yavad . brahman.d.avacakam // 7 // s´atravo vilayam . yanti is.t.asattvaparan; taka-ra-d bhra-mayet toyam mukha-h. // 66 . totala jayate parara-s.t.ra-n.i lı-laya- pu-jita- tu sah. / tvarita- tatah. / vibhu-tayas´ ca vipula-m a-rogyam . dırgham a-yus.am // 67 // sidhyanti sarvaka-rya-n.i divya-divya-ny anekadha- / … taleti sapta pa-ta-la-n kala-gnibhuvana-ntika-n - om . karadi svar arabhya yavad brahma-n.d.a–va-cakam // 82 // traka-ra-t tra-yate sarva-m . s trasate caiva sarvatah. / trotala- tena a-khya-ta- tantra-rtho ’yam . pratis.t.hitah. // 83 //

82

Michael Slouber - (pañcamasvara–m–a-ru-d.ham . ) ! ‘mounted by five vowels’ (pañcasvarasamarud.ham . ). Needless to say, M. N. Dutt’s ‘translation’ of the Agnipuran.a passage in Joshi’s edition is complete nonsense and notably altogether skips this and several other verses in the chapter. The parallels may be summarised as follows: Tvarita-mu-lasu-tra ! Agnipura-n.a 1–4 310 5–6 7–8

311 312

Chapter 313 in the Agnipura-n.a has no material on Tvarita-, but it resumes with chapter 314, ‘Tvarita-jña-nam’. Such a title makes us suspect that it may be drawing from the Tvarita-jña-nakalpa, but I found no parallels. Agnipura-n.a 309 also opens with ‘Now I shall tell the tvarita-jña-nam’, but it too appears unrelated to the Tvarita-jña-nakalpa that reaches us. I do, however, see that most of Agnipura-n.a 309 is parallel with Na-ra-yan.a’s Tantrasa-rasam . graha 22, starting with verse 47 and going to almost the end of the chapter. Is´a-nas´ivagurudevapaddhati 22 also has some parallels with this chapter. It is not clear to me now whether one of these texts is copying from another or if they independently copy from the same source text or texts, but I would not rule out the possibility that the Agnipura-n.a is drawing on post-canonical digests in some cases. It appears to use Na-ra-yan.a’s work in chapter 294, unrelated to Tvarita- but of interest to my broader work on the Ga-rud.a Tantras. These identifications of parallels just scratch the surface of what remains to be discovered by careful textual work taking into account the rich treasure of unpublished sources. We have many Tvarita- texts that I have not yet mentioned and have no time or space to explore in detail now, such as the aforementioned Tvarita-vidha-nasu-tra,24 Pa-rames´varı-mata 39 which describes itself as drawn from the Trotalottara, and S´a-rada-tilaka 10 (up to around verse 50) with Ra-ghavabhat.t.a’s useful citations of many other Tvarita- texts. In her book The Iconography of Hindu Tantric Deities, vol. II, Gudrun Bühnemann points to the tenth-century Prapañcasa-ra, verses 13.26–31, as the source of her description in S´a-rada-tilaka 10.25 Incorporations of TvaritaMuch of the Tvarita- material in other traditions cannot at present be attributed to direct borrowing from the Trottala corpus. In this regard, let us return to the topic of her appearance. Dyczkowski lists three important features common to most of her visualised forms: that she is a tribal woman (s´abarı-),

Snakebite goddesses and S´a-kta traditions

83

that she is adorned with snakes, and that she is associated with peacocks.26 These criteria agree with her appearance in Agnipura-n.a 309; however, the Tvarita-mu-lasu-tra visualisation mentions nothing of her being a tribal woman or associated with peacocks, and so these features are absent in the chapters drawn from that work (310–312). Agnipura-n.a 314 has been taken to be a third unique visualisation of Tvarita- in that text27 – as two- or eight-armed – but details of her visualisation are not given and the eleven attendants match those in chapter 309, and so I think it is safe to assume it is coming from the same tradition. What tradition might that be? In Kulakaulinı-mata 3, the main form in which Tvarita- is visualised agrees with Dyczkowski’s attributes, but it also mentions an alternative eighteen-armed form for use in magical rites.28 This would seem to be a reference to our Tvarita-mu-lasu-tra version. On the provenance of this chapter, Dyczkowski points out that it does not mention Kubjika- at all, but identifies her as Tripura-, and that he suspects this entire chapter was drawn from a tantra of another school.29 I assume this other school would be some early form of the cult of Tripurasundarı-. We might then classify the early Tvarita- literature into two camps: the Trottala corpus and its borrowers on the one hand, and the texts apparently derived from an early or proto-Traipura tradition. Examples of the latter generally share the following features not present in the Trottala corpus:    

Tvarita- is visualised as a tribal woman with leaf-clothing, peacock feathers, and parasol; she is accompanied by eleven attendants plus two door guardians; the presence of the Tvarita- ga-yatrı- mantra; the core vidya- is enhanced by the addition, twice, of the syllable hrı-m ..

It appears likely, then, that the origin of Kulakaulinı-mata 3, Agnipura-n.a - 309 and 314, Na-ra-yan.a’s Tantrasa-rasam . graha 22, and Is´anas´ivagurudevapaddhati 22 all may lie in the early Tripura- cult. This furthermore appears to be the form of Tvarita- in which she was adopted into wider pantheons in the form of a nitya-, yoginı-, ma-tr.ka-, or s´akti in the retinue of another goddess. For example, in later Traipura scriptures like the Jña-na-rn.ava and S´rı-vidya-rn.ava, Tvarita- features as an attendant (here nitya-) in the retinue of Ka-mes´varı-, a synonym of Tripurasundarı-, and her vidya- is the twelve-syllable version enhanced with two hrı-m . s. The key question is whether Tvarita- ultimately emerges from the Trottala corpus or that of Tripurasundarı-, and the evidence points to the former. Although it may be tempting to suggest that a simpler visualisation of Tvaritaas tribal snakebite goddess was the source of the more complex and encompassing eighteen-armed Tvarita-, it may not be the case. The cult of Tripurasundarı- was devoid of the ferocious hordes of prior Kaula pantheons and was set to be incorporated into mainstream religion, and so had a need for powerful yet non-threatening deities. But perhaps there is a third model.

84

Michael Slouber

Perhaps there was an original Tvarita- cult attached to an early Trottala scripture wherein Tvarita- had a more humble appearance that was adapted to be more Durga--like as the cult grew in prominence as we see in the Tvarita-mu-lasu-tra. We cannot be sure, but perhaps these speculations will be useful avenues for future research. We may also note that this Traipura Tvarita- was closely associated with the Jain goddess Padma-vatı-. The third verse of the Bhairavapadma-vatı-kalpa – an important Jain tantric work from the eleventh century with significant dependencies on S´aiva tradition – runs as follows:30 Totala-, Tvarita-, Nitya-, Tripura-, Ka-masa-dhanı-: these are names of the goddess Padma-, and so is Tripurabhairavı-. In his Hamburg lecture entitled ‘The Appropriation of S´aiva Sources and Models in the Production of Jain Ritual Paddhatis from the 10th to the 15th Century’, Alexis Sanderson pointed to this verse as one among many pieces of evidence that the Bhairavapadma-vatı-kalpa drew on S´aiva sources, in this case, he asserted, the S´a-kta cult of Tripurasundarı-. The evidence which I have offered on the widespread presence of Tvarita- in Traipura sources both supports and is supported by Sanderson’s cogent thesis. The connection between Padma-vatı- and Totala-/Tvarita- must have been an easy one, since Padma-vatı- was closely associated with snakes through her previous life’s role as snake-protectress of Pa-rs´vana-tha, the twenty-third tı-rtham .kara in Jainism, himself associated with curing snakebite. She remains an important goddess in Jainism even today, often pictured with a series of cobras with flared hoods shielding her from above. Tvarita- is also worshipped today. We may point to the popular Tulja Bhavani temple in the Tuljapur district of Maharashtra, whose website informs us that Tulja- is a Marathi form of Sanskrit Tvarita-.31 The description of the idol and the descriptions on the website, however, make it clear that she is regarded as Durga-, slayer of the buffalo demon, so we can only wonder about the roots of this particular temple. Looking for Tvarita- on the internet, one predominantly finds references to her as a nitya- goddess, which is the identity that I believe was popularised in the early Tripurasundarı- literature. At Celextel.com,32 under the category ‘Yantras’ Tithi-Nitya’, one can buy a copper Tvarita- yantra that they promise protects the owner from ‘poverty and poisonous attacks’. At Shiva-Shakti. com,33 we get a description of Tvarita- as a Nitya- goddess, evidently drawn from Traipura sources.

KurukullaThe goddess Kurukulla- is best known as a tantric Buddhist goddess and often identified with Ta-ra-. Her Buddhist identity is so popular, that even so eminent a scholar as Jan Meulenbeld remarked that a reference to her by the ninth

Snakebite goddesses and S´a-kta traditions

85

century S´aiva physician Ma-huka was to a tantric Buddhist goddess. He backed this statement up with nearly twenty references to her in the secondary literature.35 What he didn’t know, and what the scholars he cites didn’t know, is that Kurukulla- actually has a complex S´aiva identity that cannot 34

easily be reduced to borrowing from the Buddhist traditions. Whether Kurukulla- originally sprung from Buddhist or S´aiva roots cannot be easily determined, but we can at least here briefly show that it is not a simple question. S´aiva references In S´aiva/S´a-kta literature, it is useful to distinguish between two Kurukullaidentities: the first as a goddess who heals snakebite and keeps a home safe from snakes and harmful influences, and the second as a subsidiary figure in various other goddess traditions, usually not associated with snakes or poison. The oldest references, so far as I can determine, are to the first identity. The earliest S´aiva references come from the ninth century. We have Ma-huka’s citation of the power of Kurukulla- and Bherun.d.a- in the opening verses of the first chapter of his Haramekhala-, but these are extremely brief – only one line for Kurukulla-: ‘Kurukulla- drives away snakes [when] inscribed at the threshold of the house’.36 The anonymous commentator fleshes this out somewhat by telling us that ‘threshold’ means a certain part of the door – I would assume it is the lintel in conformity with the practice one sees in modern Nepal for the Na-g Pañcamı- festival – and that one is to post a yantra there on birch bark which has been inscribed with the syllables of Kurukulla-’s vidya- on the six corners of two interlocking triangles, as in the following figure: Judging from the widespread references to it, the apotropaic practice of hanging this yantra in one’s house may be the core of Kurukulla-’s fame in the . - early S´aiva tradition. S´ankuka’s Sam . hitasara also mentions this practice in his section on Kurukulla as a Garud.a goddess. This text is roughly contemporaneous to the Haramekhala- (both c. ninth century), and at six verses, this is sadly the longest passage that I have seen on the early S´aiva Kurukulla-. In these verses we also learn that the Ga-rud.ika mantra practitioner would install the syllables on his body and be able to carry out various magical acts just like Garud.a, incant a string with the vidya- and ritually place it on a patron to ward off snakes, and incant gravel to be thrown in a house to drive out Na-gas. For details on this text and the practices mentioned, see Slouber 2011: 51–56. In the Kriya-ka-lagun.ottara, a scripture from around the tenth century drawing on older Ga-rud.a and Bhu-ta Tantras, we get a few more specifics:37 - om . kurukulle svaha. This vidya is to be written on a sheet of birch bark on the door of a house, facing out. She drives off a snake. By facing inward, it would enter again.

86

Michael Slouber

f fT

^

&

wr

f ?%

Figure 4.2 Kurukulla- yantra

D . alhan.a gives us several more references to this Kurukulla in his twelfth-century commentary on the Sus´rutasam . hita. Regarding 1,46.447 on kings avoiding poisoned food, he says that the mantras used to purify food refers to infallible mantras which render the poisoned food harmless such as those of Kurukulla- and Bherun.d.a-.38 Commenting on 5,5.9, he again mentions these two as exemplary of antivenom mantras that he thinks the root text is referring to, but notes that he will not give them since they are taught in other works. He mentions Kurukulla- a third time in his commentary on 5,5.51. All of this points to her prominence as an antivenom and anti-snake goddess in medieval Hindu India. The nineteenth Chapter of the Ga-rud.a Pura-n.a is a sort of hyper-condensed Garud.a Tantra in thirty-five verses. Verses 14–17 are on Kurukulla-, and they - are so similar to the six verses on Kurukulla- in the Sam . hitasara, that it makes one suspect that to be the Puran.a’s source, although it is also possible that each draws on a third source text. The Bhairavapadma-vatı-kalpa, which I mentioned before regarding Tvarita-, also includes a single verse along the same lines:39 A na-ga would not linger in a house where a Ga-rud.ika (na-ga-ribandha) has written the Kurukulla- vidya- in the middle of a six-cornered diagram.

Snakebite goddesses and S´a-kta traditions

87

I think we need not make much of the fact that this is a Jain text – we need not posit a separate Jain Kurukulla- – because so much of this text is drawn from S´aiva sources. Nor do we need to detail all of the other texts which refer to such a yantra, but they include Ud.d.a-mares´varatantra 15.1, verse 121 in the unpublished Yogaratna-valı- of S´rı-kan.t.hapan.d.ita, and S´a-rada-tilaka 24.8. All of these references from texts from the ninth century up to the sixteenth century know Kurukulla- as an independent S´aiva goddess whose vidya- may be used against snakes and poison. But this is not her only S´aiva identity. Like Tvarita-, she is also found in many texts as an attendant goddess to another deity. That the situation was complex is evident when we read S´aktisam . gamatantra 3,14 where the text emphasises that the Kurukulla it teaches is different than the one in the S´rı-vidya- tradition. There she is only an ancillary . (anga) of Ka-lı-, but here she is a maha-vidya- and a nitya- of Ka-lı-.40 Skimming the chapter, it is clear that this Kurukulla- has no associations with snakes or poison, but is rather used predominantly for love magic. Going back to the older S´rı-vidya- text Tantrara-jatantra, we get a 101-verse Chapter (22) on Kurukulla- that also focuses predominantly on love magic. It does, however, also mention the yantra to drive off snakes and a few other antivenom-type usages. Buddhist KurukullaThe Kurukulla- of Tantrara-jatantra 22 has a suspicious number of Buddhist - features. Her encoded ten-syllable vidya- (om . tare tuttare ture svaha) in this chapter is identical to that of Kurukulla in a number of early Buddhist Tantras such as the Kurukulla-kalpa and the Guhyasama-ja. I have not seen this vidya- in any other S´aiva sources. So does this make a Buddhist origin likely? Perhaps, but not necessarily. Wiesiek Mical, a doctoral candidate at the University of Hamburg, is writing his dissertation on Kurukulla-, primarily from Buddhist sources. His deep research has also led him to the aforementioned S´rı-vidyaand Ka-lı-kula sources, and he has explored the origins of Kurukulla- at length. Nevertheless, he remains uncertain as to the tradition in which Kurukulla- first arose, and he was only partially aware of her Ga-rud.a tantric identity. Chronologically, the Buddhist sources appear to be earlier, but issues in dating most of these texts raise significant doubts. We eagerly await Mical’s forthcoming dissertation. The early Buddhist Sam . mitıya sect had a subdivision known as Kurukullaka as early as the second century AD. This Therava-da sect bore no resemblance to Maha-ya-na or Vajraya-na Buddhism and may have even burned Tantric Buddhist scriptures in the medieval period.41 According to Mical, we have sources linking the name of this sect to a mountain as well as other sources linking Kurukulla- to a mountain in Gujarat. Thus, although Mical finds the evidence tenuous, we could theoretically have a tradition of the goddess originating from a mountain independently from the earlier orthodox sect of the same name.42 The main source for the Buddhist Kurukulla- cult was the Kurukulla-kalpa. This seems to be the source of several Kurukulla- meditations in the

88 Michael Slouber Sa-dhanama-la-, and could perhaps even be older than the Hevajratantra. While I have not read this text in detail, I can say that it focuses on using Kurukullafor both healing poison and for love magic. We also see the Kurukulla- yantra to be placed on a door to drive out snakes in verses 4.23–24. On the other hand, Mical sees Kaula influence in some of its chapters, although the chronological feasibility keeps him rightly doubtful.43 To conclude, the question of Kurukulla-’s roots are far from solved and depend heavily on relative dating of various under-studied texts and traditions on both the Buddhist and S´aiva sides. It is also possible that the influences went both ways, in which case the relevant question is how the traditions influenced each other and not which came first.

Bherun.d.aThe name of the goddess Bherun.d.a- immediately conjures avian imagery. According to K. N. Dave, the bherun.d.a bird was either a bearded vulture, adjutant stork, or dodo (1985: 397–399). The first two seem plausible because of their enormous size and strikingly fierce appearance. The bearded vulture can have a wingspan of up to ten feet, and the adjutant stork over eight feet. Hemacandra’s lexical Aneka-rthasam . graha suggests that it may have referred to both birds: ‘The word bherun.d.a refers to two fierce birds, [while] bherun.d.ais a specific deity.’44 We may note in passing the two-headed bird named Gan.d.abherun.d.a, associated with Vis.n.u’s Narasim . ha incarnation and part of the official seal of the state of Karnataka, although I see no connection between this mythical bird and the goddess Bherun.d.a- aside from the name. Features in the early references As with Kurukulla-, it is useful to distinguish the stand-alone snakebite-goddess of the Ga-rud.a Tantras and dependent literature from her identity as an ancillary goddess in other S´a-kta sources. I have already mentioned several texts in connection with Tvarita- and Kurukulla- that also feature Bherun.d.a-: the - Sam . hitasara, Haramekhala, and D . alhan.a’s commentary to several Sus´rutasam hita passages all seem to refer to the independent snakebite goddess. We . also have references to her in the Rasaratna-kara’s toxicology (vis.acikitsa-) section, Yogaratna-valı- 122, and Bhairavapadma-vatı-kalpa 10.12. Without exception the passages are brief, sometimes only half a verse. They say that the Bherun.d.a- spell should be chanted in the ear of a bite victim to free him of the venom. With the exception of D . alhan.a, who declines to give the spell directly, all of these sources or their commentaries teach a Prakrit spell for - Bherun.d.a-. The longest version is that following Sam . hitasara 151, which runs for five lines in my edition. The others are briefer; the Rasaratna-kara version is only one line, but clearly based on the same Prakrit spell. In Table 4.2 that follows, I give the spell as found in each of the five sources, though I abbreviate - that of the Sam . hitasara:

Snakebite goddesses and S´a-kta traditions

89

Table 4.2 Prakrit spell teachings for Bherun.d.a- Sam sam . hitasara . joe mpae bherum . d.ae vi_bhariakaram . d.ae tam . ta mam . ta visa -ahosaï jambhaï mohaï tha-vara jamgama kimtima jaja ja-hi re jaja ja-hi . . . - /… re maha-pasa-u bhara-d.-ıe haru visa karu n.ivvisu hu-m . hum . [continues] Haramekhalaom . joe mae bherun.d.ae vijjabhariakaran.d.ae mantam . sun.uha jaha ghosaï hikka-raï taha visu n.a-saï tha-varajam . gamao tham . bhaï jam . bhaï mohaï jahi re jahi re Yogaratna-valıom . joyasasan.e bherun.d.e vijjahabharis´ca karan.d.e tam . tu mam . tu a-ghosaï phekarat vis´u na-s´ayi tha-varajam . gama mare stambhan.e mohan.a ja-hire gara ja-m . jah. Bhairavaom . ekahi ekamate bherun.d.a vijjabhavikajakaran.d.e tam . tu mam . tu padma-vatı-a-mosaï hum . kara vis.a nasaï thavara jam . gama kittima am . gaja om . kalpa phat. - bı-jam bhaviakarande tantra Rasaratna-kara om . eha matra bherun.d.e aïum . . .. - re visa na-s´aï stha-varajan.gameti mantra agdos.a -ın hu-m ka . . manhukaï

Clearly the vidya- is in need of editing, but I do not feel confident enough to fix it. The Haramekhala- version seems more coherent in some respects. I partially translate: om . joe Mother Bherun.d.a whose basket is filled with spells (?), listen to the mantra! As you cry out, screech, so must you destroy the poison, be it from a plant or animal – terminate it! Destroy it! Make it fail! Go! re!, Go! re! A Ga-rud.apura-n.a 19 reference sounds like our stand-alone Bherun.d.a-, but the spell is not in Prakrit: OM. HRI- HRAU HRI-M. BHIRUN.D.A-YAI SVA-HA-. Other S´a-kta identities We find many other texts using Bherun.d.a- as an attendant goddess. The Tantrara-jatantra has Bherun.d.a- as a nitya- alongside Tvarita- and Kurukulla-. In verses 3.35–37, the text gives her nine-seed-syllable spell in code working out to: OM. KROM. BHROM. KROM. JHROM. CHROM. JROM. SVA-HA-. It is completely different from the Prakrit vidya- of her independent identity, but here too she is said to be able to destroy the three types of poison, but rather than chanting in the ear of the victim, it need only be recalled by the initiate. The Tattvacinta-man.i has a similar series of syllables for its Bherun.d.a- nitya-. In the Brahma-n.d.apura-n.a, the Vis´vasa-ratantra, and the Matsyendrasam . hita, she is listed as one of fifteen or sixteen nitya- goddesses, but no details on her form or mantra are given. In the Matasa-ra, Bherun.d.a- is one of eight goddesses of speech (va-ges´varı-).

90

Michael Slouber

Conclusion Tvarita-, Bherun.d.a- and Kurukulla- each have multifaceted identities in the S´aiva and S´a-kta literature. We have separate literature, visual forms, spells, and identities in the early Ga-rud.a Tantra material and in the early Tripurasundarı- literature. Which came first is difficult to prove, but the fact that the Ga-rud.a Tantras give no hint that the goddesses are borrowed from another system is suggestive. On the other hand, the Tripurasundarı- literature, and later S´a-kta systems that drew on it, frequently mention the ability of these goddesses to heal poison and drive away snakes. I close with one final passage from the S´rı-vidya-rn.ava:45 May the vidya- who is called Suvarn.arekha-, the one said to be an eradicator of snakes, give ease to me. May the mighty-looking vidya- called Kurukulla-, arisen from the mouth of the Lord of Birds, always be present - on the tip of my tongue. May the one called Jham . karin.ı always be present in my body. [May the one] named Remover of Poison be a cleaver to the form of the Kali age. May Bherun.d.a- always be present in my throat. May Totala- be present in my head. And likewise may Suvarn.arekha- also . always be present at my base. Let Ja-ngulı- make my speech perfect for the destruction of poison.

Notes 1 Vidya-s are both spells, the female equivalent of mantras, and goddesses. The sonic spell was understood as the embodiment of the goddess. - 2 See Slouber 2011 for an edition and translation of the ninth-century Sam . hitasara passage on Jham . karin.ı. 3 Predominantly known as a Buddhist goddess from the sa-dhana literature, she was also known as a S´aiva goddess. Asha Archives manuscript number 3152 contains a . S´aiva Ja-ngulı-vidya- on folios 5 recto to 8 verso. We also see a S´aiva passage on . Jangulı in the Tantrasadbha-va passage starting with 23.296. - 4 See the aforementioned Sam . hitasara (Slouber 2011) for a passage on Suvarn.arekha. 5 The spelling of this title varies. We also see it as Trotala, Trotula, Totula, Totala, and Trot.ala. 6 Mark Dyczkowski 2009: 83–85 (vol. 2). 7 See Bhairavapadma-vatı-kalpa 1.3 in Jhavery 1944. 8 The colophons of the Kulakaulinı-mata ascribe it to the Caturvim . s´atisahasra (i.e. the Mantha-nabhairavatantra). Thanks to Mark Dyczkowski for sharing his draft edition of the Tvarita- chapter of this work. 9 Microfilmed in NGMPP A59/13. Folios 16–34 carry the Tvarita-vidha-nasu-tra passages. 10 The crucial verse on the topic of Gaurı-/Tvarita-’s identity is: Tvarita-mu-lasu-tra 1.37: -tatah. kruddha- tu sa- gaurı- tad.itkot.isamaprabha- / tvaritam . tvarita vidya tribıja trigun.ı kr.ta- // ‘And then Gaurı- was enraged, with a fiery splendour equal to millions of lightning bolts. Instantly (tvaritam . ) she [uttered] the Tvarita vidya, consisting of three times three seed syllables.’ The verse has no verb, so inserting ‘uttered’ is an

Snakebite goddesses and S´a-kta traditions

11

12 13 14

15

16

91

interpretation. We could alternatively supply ‘became’, but that is also a theologically loaded assumption. Tvarita-mu-lasu-tra 1.45–56, ff.5v–6v (Here and henceforth I cite folio numbers for manuscript H170/3). I offer the following provisionally edited Sanskrit text, but the grammar is often highly non-standard, sometimes to preserve the meter: tvarita. tatra ca-ru-d.ha- va-majangha- *pralambita- (conj., pralambita-m . Cod.) / daks.in.a dvigun.a tasyah. padapr.s.t.he samarpita- // sphuranta-rakavad devya- bhra-jante pa-dayor nakha-h. / ambhoja-run.a*varn.a-bha-h. (conj., -varn.a-bha- Cod.) padmala-ñchana*laks.an.a-h. (conj., . -laks.an.a- Cod.) // ra-jate caran.a-dhasta-t susama-nguli*–m–a-yata- (conj., -ma-pata- Cod.) . -noru *vistı-rna (conj., vistı-rne Cod.) jaghanoru sa- // gambhı/ suvr.tam jan ghapı . . . ra*na-bhis (conj., -na-bhi Cod.) trivalı- romara-jı-sus´obhita- / vistı-rn.a- kat.ides´e tu hemakumbhapayodhara- // kambugrı-va- tu bimbos.t.hı- nı-lotpaladaleks.an.a- / suna-sa- ca-pabh. ru-bhanga- sukarn.a- *mukut.ojjvala- (conj., makut.ojvala- Cod.) // *visphuranman.ibhir (conj., visphuretman.ebhir Cod.) *dı-ptah. (conj., dı-ptaih. Cod.) sahasraphan.idı-ptima-n / na-gara-jakr.toha-rakarn.akun.d.alabha-sinı- // na-gabandhakr.ta- s´obha- *bhujau (conj., bhujo Cod.) devya- vira-jate / mekhala- kat.ides´e tu maha-hikr.tabhı-s.an.a- // pa-dayor . *nu-purau (conj., nupurau Cod.) ghorau devya-h. kr.tabhujangamau / sahasradale ca-mbhoje va-mapa-dakr.todara- // viyattad.illata-*bha-sam (conj., -bha-sa-m . . Cod.) vasah. kusuma*s´obhitam (conj., -s´obhita-m . Cod.) / indrayudhanibham . canyam . *vasah. (conj., ra-sam . Cod.) kr.tapayodharau // anekapus.paracitam . dr.d.ha*bandha (conj., -vadha- Cod.) mahoragam / kr.ta-yudhakara- devı- tı-ks.n.a-granakha*s´uktayah. (conj., -muktayah. - Cod.) // vajradan.d.a-sicakram . ca gada s´ulam . *mahojjvalam (conj., mahojvalam . Cod.) / s´aram . s´aktis´ ca varadam . daks.in.ena kr.tayudha .// *dhanus . (conj., dhanu- Cod.) . pa-s´a*dharam . (conj., -haram . Cod.) ghan.t.a tarjanı s´ankham ankus´am / abhayam . ca tatha- padmam . vamapars´ve kr.tayudha // Dyczkowski 2009: 88–89. Following verse 46, ff.4v–5r in manuscript A59/15. - ka-radvayasamyuktam khe ca cche Tvarita-jña-nakalpa A59/15, f.2v: (om . ) hum . . . - kse phat ca vai smrta-h // 24 // padabhus.itam / vargatıtam . . . visargas´ ca strım . hum . . . Taken literally, we might assume that hum . karadvayasam . yuktam . means that the - , but I think it is rather just indicating that the entire vidyavidya- begins hu-m . hum . will have two hu-m . syllables. The om . at the beginning does not fit the meter and may have been added later. The text divides the Sanskrit syllables into eight groups, as is typical, but the names of several vargas appear to be peculiar to this text and the Agnipura-n.a passage that draws on it. The verses preceding the vidya- clearly list the groups in order: svaravarga, ta-luvarga, jihvata-luka, ta-lujihva-gra, jihvadanta, os.t.haput.a, mis´ravarga, and u-s.ma-n.a. In manuscript H170/3 it is ff.5v–6v (AP=Agnipura-n.a 310 parallel, which I only report when I accept its reading over ours): s.as.t.hasvarasama-ru-d.ha-m *u-s.ma-- - n.a-ntam . (AP, us.man.anta- Cod.) sabindukam / mulavidyadikam . bıjam . kathitam . tu khages´vara // taluvarga*dvitıyam . (AP., -dvitıyas Cod.) tu svaraikadas´ayojitam / jihva-ta-lu*sama-yoge (AP, sama-yogah. Cod.) prathamam . kevalam . bhavet // tad eva taddvitı-yam . tu adhastad viniyojayet / ekadas´asvarair yuktam . prathamam . taluvargatah. - t tasya yojayet / sodas´asvarasamyuktam u-sma-nasya // u-s.ma-n.asya dvitı-yam tu adhasta . . . . . . tr.tı-yakam // jihva-dantasama-yoge prathamam . yojayed adhah. / mis´ravargadvitıyam . tu adhastat punar eva ca // caturthasvarasambhinnam ta luvarga dimam punah / u s ma . . . . n.asya - - dvitı-yam . tu adhastad viniyojayet // svaraikadas´asambhinnam us.man.antam . sabindukam . / pañcamasvara–m–a-ru-d.ham os.t.hasamput.ayogatah. // dvitı-yam aks.aram . canyam . jihvagre ta-luyogatah. / prathamam . *yac ca sam . yojyam . (conj., pañcame yojyam . Cod.) svar- a-rdhenoddhr.ta- ima- // tvariteyam . mahavidya sarvasiddhipradayika / om . kar-adisama-yukta-–m–ante namo japet sada- // sva-ha-ntamagnika-ryesu… . I assume we should also supply tvarita-yai before namah./sva-ha-.

92

Michael Slouber

17 Folio 13v: a-didvihr.dayam . proktam . tricatuh. s´iramis.yate // pañcas.as.t.has´ikha prokta kavacam . saptamas.t.akam / *taraka (conj., tarakam . Cod.) tu bhavennetram . nava-rdha-ks.ara*laks.an.am (AP, raks.an.am Cod.) // 18 The context makes it clear that ta-raka- refers to phat., but we may also point to the similar sounding code word for phat. found in Dakshinamurti’s Uddha-rakos´a: turaga. 19 See ff. 62v–63v. 20 ff. 26r–v: ekaya- caiva a-hutya- dı-ks.ito *bhavate (conj., bhavete Cod.) khaga / adhika-ro bhaved *evam . (conj., devam . Cod.) moks.am . s´r.n.u atah.param // 21 The text gives this figure itself. It is actually around 150 plus prose, but the traditional way of measuring the length of a verse-text is by counting the average syllables per line, which would include the prose too. 22 For the following words I deviate from H170/3: pralambita-, B126/9, pralambita-m . H170/3; dhanus.pa-s´adharam B126/9, dhanupa-s´aharam H170/3; is.t.asattvapar. . . . - a-nmukha-h. B126/9, dus.t.asattva- para-nmukha-h. H170/3; traka-ra-t conj., om . karat - ra-t H170/3; sarva-ms tra-sate B126/9, sarva-n tra-s´ate H170/3. B126/9, aim ka . . 23 Here are the words that I take from the Agniopura-n.a etext (‘APe’) rather than Joshi’s - printed edition: samarpita- APe, samı-psita- Joshi; om . karadisvararabhya APe, - - - om . karad ıs´varad arabhya Joshi. It is evident from citations of these chapters by other scholars that the transmission of the Agnipura-n.a includes a great deal of variation. 24 Although the Tvarita-vidha-nasu-tra is put in the mouth of S´rı-vakra- (Kubjika-), she and her interlocutor S´rı-kan.t.ha reference the ‘Trottala-mata’ as a source and the material appears more closely aligned to the Tvarita-mu-lasu-tra and Tvarita-jña-nakalpa than other Kaubjika material I have seen. 25 Bühnemann 2000: 207 (vol. 2). 26 Dyczkowski 2009: 89 (vol. 2). 27 Dyczkowski 2009: 88, citing Mallmann 1963: 160. 28 Ibid. 88. 29 Dyczkowski, personal communication. 30 Jhavery 1944: 373 (p. 1 of the edition’s pagination): totala- tvarita- nitya- tripuraka-masa-dhanı- / devya- na-ma-ni padma-ya-s tatha- tripurabhairavı- // 31 http://ncbinfotech.appspot.com/tulja/tulja.php?pageid=TV20. 32 www.celextel.com/tvarita-yantra-320-mm-p-916.html. 33 www.shivashakti.com/nitya.htm. 34 Meulenbeld 1999; vol. IIA: 134. 35 Ibid. vol. IIB: 151. 36 Edition of Bhattarai 1972: 5–6. - 37 Kriya-ka-lagun.ottara 35, mantra seventeen: om . kurukulle svaha / bhurjapatre iyam . . vidya lekhya gr.hadvare paran-mukham // sarpam uccat.ayati / sam . mukhena punah. pravis´ati / 38 siddhair avyabhica-ribhih. kurukulla-bherun.d.a-prabhr.tibhir hatavis.am annam iti sambandhah. / 39 Verse 10.41 in Jhavery 1944: s.at.kon.abhavanamadhye kurukulla-m . yo likhed gr.he vidya-m / tatra na tis.t.hati na-go likhite na-ga-ribandhena // With Bandhus.en.a’s commentary: ‘s.at.kon.abhavanamadhye’ s.at.kon.acakramadhye / ‘kurukulla-m . ' kurukulla-na-madevya- mantrah. / ‘yo likhed’ yah. ko ’pi mantrava-dı- likhet / kva? ‘gr.he’ . gr.hadehalya-m, svava-sottara-nge / ka-m? ‘vidya-m’ kurukulla-devya- vidya-m / ‘tatra’ tasmin gr.he / ‘na tis.t.hati’ na stha-ti / kah.? ‘na-gah.' sarpah. / kasmin kr.te sati? ‘likhite’ - phat sati / kena? ‘na-ga-ribandhena’ garud.abandhena // mantrah.––om . kurukulle hum . . // 40 S´aktisam . gamatantra, Sundarıkhan.d.a, Chapter 14, verses 8–9. Thanks to Wiesiek Mical for pointing out this chapter to me and sharing his draft translation of it. 41 According to Ta-ra-na-tha’s History of Buddhism in India, translated by Chattopadhyaya 1970: 279.

Snakebite goddesses and S´a-kta traditions 42 43 44 45

93

Personal communication. Personal communication. 3.173cd: bherun.d.au bhı-s.an.akhagau bherun.d.a- devata-bhidi // The numbering in the electronic edition is odd, but it cites it as p. 342 of the 1947 Shrinagar edition: suvarn.arekhin.-ı prokta- vidya- ya- procyate kila // nirmu-linı- bhu. - janga-na-m . sa karotu sukham . mama / kurukulleti vikhyata paks.irajamukhodbhava // - - ya vidya- sa- maha-ru-pa- jihva-gre stha-tu me sada- / *jham . karin.ıti (conj., om . karin.ıti Cod.) vikhya-ta- dehe stha-tu sada- mama // *vis.a-paha-rin.-ı (conj., vidya-paha-rin.-ı Cod.) na-ma kaliru-pavida-rin.-ı / bherun.d.a- stha-tu me kan.t.he *totala- (conj., torala- Cod.) stha-tu mastake // tatha- *suvarn.arekha-pi (conj., s´avalarekha-pi Cod.) mu-le stha-tu . . sada- mama / *ja-ngulı- (conj., ja-ngalı- Cod.) vis.ana-s´a-ya va-ca-m . siddhim . karotu me //

References Primary sources Agnipura-n.a, see Joshi 2001. Aneka-rthasam . graha, see Zachariae 1893. Bhairavapadma-vatı-kalpa, see Jhavery 1944. Brahma-n.d.a Pura-n.a, GRETIL Etext Archive. URL: http://fiindolo.sub.uni-goettingen. de/gret_utf.htm#BrahmandaP, accessed 20/08/2011. Caturvim . s´atisahasra, see Manthanabhairava. Ga-rud.a Pura-n.a, see Sresthi 1906. Guhyasama-ja, see Matsunaga 1978. Haramekhala-, see Bhattarai 1972. Hevajratantra, see Snellgrove 1959. Is´a-nas´ivagurudevapaddhati, see Sastri 1920. . Ja-ngulı-vidya-, unedited Asha Archives manuscript #3152. Jña-na-rn.avatantra, see Malaviya 2001. Kriya-ka-lagun.ottara, unedited. See Slouber 2007. Kulakaulinı-mata, unedited. See Dyczkowski 2009. Kurukulla-kalpa, see Pandeya 2001. Mantha-nabhairava, see Dyczkowski 2009. Matasa-ra, unedited, see NGMPP reel B 28/16. Matsyendrasam . hita, see Sen Sharma 1994. Narayan.a’s Tantrasa-rasam . graha, see Aiyangar 1950 and Slouber 2009. Nis´va-samu-lasu-tra, see Goodall (forthcoming). Pa-rames´varı-mata, Muktabodha Digital Library. URL: http://muktalib5.org/digital_ library.htm, accessed 20/08/2011. Prapañcasa-ra, see Bühnemann 2000. Rasaratna-kara, see Srikrsnadas 1909. Sa-dhanama-la-, see Bhattacharya 1968. S´aktisam . gamatantra, Muktabodha Digital Library. URL: http://muktalib5.org/digital_ library.htm, accessed 20/08/2011. - Sam . hitasara, see Slouber 2011. S´rı-vidya-rn.ava, see Sharma 1966. Sus´rutasam . hita, see Bhattacaryya 1891. Tantrarajatantra, see Sastri and Woodroffe 1918.

94

Michael Slouber

Tantrasadbha-va, unpublished edition of Dyczkowski based on NGMPP reels A 188/ 22, A 44/1, and A 44/2. Tattvacinta-man.i, see Bhatt 2005. Trottala, see Slouber (forthcoming). Trottalottara, see Slouber (forthcoming). Tvarita-jña-nakalpa, unedited, see Slouber (forthcoming) and NGMPP reels A 59/15, B 26/14, and B 126/15. Tvarita-mu-lasu-tra, unedited, see Slouber (forthcoming) and NGMPP reels B 126/9, H 170/3, and C 6/7. Tvarita-vidha-nasu-tra, unedited, see NGMPP reel A 59/13. Ud.d.a-mares´varatantra, see Zadoo 1947. Uddha-rakos´a, see Taki 1978. Vis´vasa-ratantra, Muktabodha Digital Library. URL: http://muktalib5.org/digital_libra ry.htm, accessed 20/08/2011. Yogaratna-valı-, unedited, see NGMPP reel A 210/10 and elsewhere. Manuscripts widespread.

Secondary literature Aiyangar, M.D. 1950. Tantrasarasangraha. Madras: Government Oriental Manuscripts Library. Bhatt, V.P. 2005. Tattvacintamani. Delhi: Eastern Book Linkers. Bhattacaryya, J.V. 1891. Nibandhasamgrahkhya susrutasamhitaya vyakhya. Calcutta: Sarasvati Press. Bhattacharya, B. 1968. Sadhanamala. Baroda: Oriental Institute. Bhattarai, K. 1972. Haramekhala- sat.-ıka-. Kathmandu: National Archives. Bühnemann, Gudrun. 2000. The Iconography of Hindu Tantric Deities, Volume II. Groningen: Egbert Forsten. Chattopadhyaya, A., Chattopadhyaya, D. and Chimpa, L. 1970. Taranatha’s History of Buddhism in India. Shimla: Indian Institute of Advanced Study. Dave, K.N. 1985. Birds in Sanskrit Literature. Delhi: Motilal Banarsidass. Dyczkowski, M. 2009. Mantha-nabhairavatantram. New Delhi: Indira Gandhi National Centre for the Arts. Goodall, D. Forthcoming. The Nis´va-satattvasam . hita: The Earliest Surviving S´aiva Tantra. École française d’Extrême-Orient and Nepal Research Centre. Jhavery, M.B. 1944. “Bhairavapadma-vatı-kalpa”. In Comparative and Critical Study of Mantrasastra with Special Treatment of Jain Mantravada, being the Introduction to Sri Bhairava Padmavati Kalpa, pp. 367–450. Ahmedabad: Sarabhai Manilal Nawab. Joshi, K.L. 2001. Agnimaha-pura-n.am. Delhi: Parimal Publications. Malaviya, S. and Malaviya, R. 2001. Isvaraproktam Jñanarnavatantram. Varanasi: Krsnadasa Akadami. Mallmann, M.T. 1963. Les Enseignements Iconographiques de l’Agni Purana. Paris: Presses Universitaires de France. Matsunaga, Y. 1978. The Guhyasamaja Tantra: A New Critical Edition. Osaka: Toho Shuppan. Meulenbeld, G.J. 1999. A History of Indian Medical Literature. Groningen: E. Forsten.

Snakebite goddesses and S´a-kta traditions

95

Pandeya, J.S. 2001. Kurukullakalpah = phags-ma sgrol-ma kurukullei rtog pa. Varanasi: Durlabha Bauddha Grantha Sodha Anubhaga, Kendriya Ucca Tibbati Siksa Samsthana. Sastri, L. and Woodroffe, J.G. 1918. Tantrara-jatantra. London: Luzac & Co. S´a-stri, M.M.T.G., ed. 1920–25. Is´a-nas´ivagurudevapaddhati. 4 vols. Trivandrum Sanskrit series 69, 72, 77, 83. Trivandrum: Superintendant, Government Press. Reprint, Delhi and Varanasi: Bharatiya Vidya Prakashan, 1988. Sen Sharma, D. 1994. Matsyendra Samhita. Calcutta: Asiatic Society. Sharma, B. 1966. Sri Srividyarnava Tantram. Prayaga: Kalyana Mandira. Slouber, M. Forthcoming. “Introduction, Edition, and Translation of the First Chapter of the Tvarita-mu-lasu-tra”. In Vasudeva, Somadeva, S´ivasudha-pra-pa-lika-. . - Slouber, M. 2011. ‘S´ankuka’s Sam . hitasara: Edition and Translation of Verses 1–5, 75–85, and 129–154 with an Anonymous Commentary’. Master’s thesis, University of Hamburg. Slouber, M. 2009. E-text of Na-ra-yan.a’s Tantrasa-rasam . graha. Hamburg: Centrefor Tantric Studies. URL: http://tantric-studies.org/e-texts/shaiva/NaTaSaSa.txt, accessed 20/08/2011. Slouber, M. 2007. ‘The Manuscripts of the Kriya-ka-lagun.ottara’. Newsletter of the NGMCP, 5, 9–16. URL: www.uni-hamburg.de/ngmcp/nl5light_e_e.pdf, accessed 20/ 08/2011. Snellgrove, D.L. 1959. The Hevajratantra: A Critical Study. Oxford: Oxford University Press. Sresthi, K. 1906. Garuda Mahapurana. Mumbai: Sri Venkatesvara Steam Press. URL: http://reader.digitale-sammlungen.de/resolve/display/bsb10211126.html, accessed 20/08/ 2011. Srikrsnadas, K. 1909. Rasaratna-kara. Mumbai: Sri Venkatesvar Steam Press. Taki, S.1978. Sri-Daksinamurti-viracita Uddharakosah. New Delhi: Munsiram Manoharalal. Zachariae, T. 1893. Der Anekarthasamgraha. Vienna: A. Holder. Zadoo, J. 1947. The Uddamaresvara Tantram. Srinagar: Normal Press.

5

The Ca-n.d.a-lı- as S´akti Untouchable women in some tantric texts Mikael Aktor

Can.d.a-lı- – Chief Goddess, Lady of Hosts The Hindu tantric text, the Yonitantra, evokes a yantra representing the yoni (vagina, womb), in which the central divinity is presented as ‘Ca-n.d.a-lı-’, a woman belonging to an untouchable caste (YT 3.14–20). This low-caste woman is nothing less than the ‘Chief Goddess, Lady of Hosts’, whose worship makes the worshipper equal to S´iva.1 Other tantric texts, most prominently the Buddhist and Vais.n.ava tantric Sahajiya- literature, make use of a similar rhetoric in texts where the Ca-n.d.a-lı- or the D . ombı washerwoman becomes a personification of the kun.d.alinıs´akti of the Hindu Tantra systems (Schoterman 1980: 24–25; Bhattacharyya 2005: 216). These identifications raise a number of questions about the social reality and the ritual significance behind such caste labels. The general antinomian character of the tantric movements as well as the names of tantric masters can be read as a critique of Brahminical social exclusiveness by groups that were either non-Brahmin by birth or had left their Brahmin communities. In so far as the names of these masters indicate their caste background, many came from low castes like leather-workers (‘Camari’), washermen (‘D . ombı’), hun´ ters (‘Sabarı’), oil-grinders (‘Tailo’), weavers (‘Tanti’) etc. Among those who were born Brahmins, Saraha chose to become an Untouchable and married a low-caste woman (Bhattacharyya 2005: 263). This impression of an anti-Brahmin attitude is obscured, however, by the idea that tantric practice was more an esoteric layer of personal religion than an identifiable feature of a particular recognisable group. Among what was probably a common Kaula stock of verses found in various tantric texts, Yonitantra 4.20 states: Privately a S´a-kta, outwardly a S´aiva, among people a Vais.n.ava – bearing various outward appearances, the followers of the Kula-system spread over the earth.2 According to this principle, tantric sa-dhakas might have belonged to various layers of society, and it would not always be easy to identify them.

The Ca-n.d.a-lı- as S´akti

97

The yonipu-ja-, the ritual that is the subject of the Yonitantra, involves sexual intercourse in order to extract the sexual fluids of the partner. These fluids, the yonitattva, consist both of the fluid produced in the vagina through the stimulation by the penis of the candidate during sexual intercourse, and the menstrual blood, which in these contexts – in contrast to the public Brahminical view – is regarded as an extremely powerful ritual remedy. In the Brahminical code menstruation and everything related to it is segregated from the ritual field, while in tantric texts it can occupy a significant aspect of the ritual practice. From the point of view of the Yonitantra, intercourse is ideally with a female partner during her menstruation, a practice abhorred in the texts of Brahminical law. The yonitattva is used in two ways: As a daily tilak (forehead mark) and in the worship of the female partner (S´akti) as Devı-, the latter also involving the consumption of the yonitattva mixed with wine (Schoterman 1980: 28–29; White 2003: 109–110). The social status of the female partner is flexible like that of the sa-dhaka, although somewhat ambiguous. Low-caste women are preferred, but not exclusively: An actress, a kapa-lika woman, a prostitute, a washer woman, a barber’s daughter, a Brahmin woman, a daughter of either a S´u-dra, a cowherd or a garland maker, these nine young women [kanya-] are recommended; or a woman from any caste who is experienced and wanton. Except for a mother, he may have intercourse with any woman between twelve and sixty years. … Only a yoni that has already been wounded is suitable for worship, never that of a virgin; by the worship of a virgin there will be destruction of powers every time, O Goddess. (YT 2.3–5, 3.25–26)3 Other texts are more restrictive both in terms of social status and age. Some exclude any woman belonging to the three upper varn.as and limit the upper age at 25 (Schoterman 1980: 20). However, that they are referred to in the text as ‘kanya-h.’ is, according to Schoterman, an indication that, although not virgins, they should still be unmarried (ibid.: 18). He also regards it as ‘a wellknown fact that especially women living on the fringe of society were chosen as S´akti in tantric rituals’ (ibid.: 24). Whatever the social facts are behind such caste labels in tantric texts, many other texts make it possible to understand the ritual significance of these labels. What emerges is a multifaceted image of ‘the Untouchable woman’ as a reservoir of contrasting forces that arouse contempt, horror and attraction at the same time. Untouchable women are powerful, first because female sexuality in general is an expression of s´akti; in the Law Books, a much more freely flowing sexuality is attributed to women than to men; according to these texts women are driven by a constant itch of sexuality (MDhS´ 9.14–15; Leslie 1989: 260). Second, there are reasons to believe that this sexuality is manifested even

98

Mikael Aktor

more strongly in low-caste women, because it remains untamed by marriage to men from the four varn.as. In general, marriage is the remedy for unrestrained female sexuality, and, in addition, Can.d.a-las are often associated with a savage life beyond the norms of civilised culture. These unmarried but sexually active low-caste women possess an untamed sexuality and as such they are highly potent agents of ritual transformation in an antinomian tantric context. Their position in society without any role in the kinship system of the varn.as and as an underprivileged group with poor access to protection from outside also makes them easily available. Apart from the general secrecy it involved, the use of low-caste women could therefore have been a way of extricating the tantric ritual from the constricting public norms of family and marriage.

The Can.d.a-la woman In classical South Asian texts belonging to a wide range of genres, The Can.d.a-la represents a category of untouchable castes more than it denotes one single caste (for an overview, see Aktor 2010). It is, as Dumont has expressed it, ‘the old prototype of the Untouchable’ (Dumont 1980: 52). In the later layers of the vedic literature we already find clear expressions of contempt for Can.d.a-las, and it is interesting in the context of the Yonitantra that one of the earliest formulations (in Cha-ndogyaupanis.ad) is explicitly with regard to the Can.d.a-la-yoni, which is far from being a source of liberation or uplift. In conjunction with the new doctrine of karman and rebirth we can read: But people of foul behaviour can expect to enter a foul womb [yoni], like that of a dog, a pig, or an outcaste [can.d.a-la] woman. (ChU 5.10.7)4 We do not know exactly what was understood by ‘can.d.a-la’ in this early text. But during the centuries of the composition of the Maha-bha-rata and of the early Law Books two different images had emerged. One was the domesticated caste of Can.d.a-la, that is, low caste groups living at the borders of villages and cities but still performing duties within them, typically related to the impurities of death, excrement or criminality. This image is dominant in the Law Books, whose very purpose is to record the customs of ‘good people’ and on that basis prescribe social norms. The very nature of these Can.d.a-las is a transgression of kinship rules, in that, as is explained in this literature, they are regarded as the outcome of illegitimate sexual relations between Brahmin women and S´u-dra men (e.g. GDhS 4.18; MDhS´ 10.12). The other type is the savage Can.d.a-la representing a tribe rather than a caste. These are mostly depicted in the epic and other narrative literature as ferocious hunters living in the jungle. An example is the well-known episode from Maha-bha-rata about Vis´va-mitra, who steals dog’s meat from a Can.d.a-la hamlet during a period of starvation (MBh 12.139.27–30). In contrast to the Law Books, the notion of mixed classes (varn.asam . kara) is absent in such stories, probably because in

The Ca-n.d.a-lı- as S´akti 99 this narrative literature these tribal Can.d.a-las represent a contrast to Aryan civilisation and as such are depicted as being unrelated to its ideal kinship system, the four varn.as (Aktor 2008: 87–104). Later narrative literature preserves this tribal stereotype. In the novel Ka-dambarı- by Ba-n.abhat.t.a and his son dating from about 600 CE, the hero, magically transformed into a parrot, is abducted by ferocious, savage Can.d.a-las. Shocked by the scene of the Can.d.a-la hamlet in the jungle, Vais´ampa-yana, the parrot, gives us a long description of this terrible place. The description is clearly an expression of all the standard stereotypes of the Can.d.a-la as representing the antithesis to Brahmin values: Here and there the village was hidden in a thick forest of bamboo and could be inferred only by smoke, which had the stench of meat. Everywhere there were enclosures of fences made mostly of skulls, heaps of feces alongside roads that were made mostly of bones, and courtyards sloshing with the slop of chopped meat, fat, marrow, and blood. The people’s livelihood consisted mainly of the hunt; their food was mainly meat; their oil was mainly bone marrow; their clothes were mainly of raw silk; their beds were mainly of hides; their retinue consisted mainly of dogs; their mounts were mainly cows, their objects of human pursuit [purus.a-rtha] were mainly women and wine; their offerings to gods were mainly blood; and their deeds of duty [dharmakriya-] were mainly sacrifices of beasts. It was the shape of the Hell of Torment. It was like the cause of all evils. It was like the place of all cremation grounds. It was like a city of all sins. It was like a temple of all hellish punishments. … The men were more shameless in all their business dealings than were the hearts of all men. The behaviour was no different in children, young people, or the aged. Intercourse was had with one’s own and others’ women. It was a market for only foul deeds. (Layne 1991: 341–342; Ka-d pp. 505–506)5 But, while this place may be terrifying in every decent respect, the wild and unspoiled Can.d.a-la women are also disturbingly attractive. At the start of the novel a young woman from this Can.d.a-la tribe brings the wonderful parrot to S´u-draka, the noble king of the country. The text describes vividly the sight of the woman and the king’s reaction: She, like a swoon, rendered men unconscious. … As if insubstantial, she avoided touch. She, like a painting, was enjoyment only for the eye. … She was in her early adolescence, and her figure was supremely lovely. S´u-draka (gazed at her with unblinking eyes), was struck with amazement, and turned these thoughts over in his mind: ‘Aho! What a wasted effort by the Creator, pouring such beauty into such an unsuitable vessel. For if she has been so created that her beauty mocks all other beauty, then why give her a caste with which the pleasures of sexual intimacy and touch are forbidden? It seems to me that the Lord of Creatures may have

100

Mikael Aktor produced her without even handling her, out of fear of the stain of a . Ma-tanga’s [i.e. a Can.d.a-la’s] touch, else where comes this beauty of such perfection; for (such beauty as this does not come about for limbs spoiled by the touch of hands). Fie! fie! upon the Creator who brings about (such an improper combination)! For this maiden whose shape is so very enchanting vexes one, since love-making with her will always be condemned because of her low caste.’ (Layne 1991: 14, with additions/alternatives in brackets from Smith 2009: 47; Ka-d pp. 24–25)6

S´u-draka’s thoughts reveal ambiguous attitudes to untouchability. He seems to turn its rationale on its head. That the Can.d.a-la girl is untouchable protects her from the touch of others more than it protects others from her touch. The fear of her touch aroused in others keeps her in a state of perfect beauty, unsullied by the touch of lustful hands. Untouchability is a social fact that the king has to obey, however much he may want to touch her. But below or beside this level of social norm and public censure, there is another level at which the Untouchable embodies a pure beauty that is protected from the impure touch of the Touchables.

The level of society The rules preventing king S´u-draka from having sexual relations with the beautiful Can.d.a-la girl were formulated in increasing detail in the medieval dharmas´a-stra texts. While untouchables were considered ritually polluting, they also made up an essential unskilled labour force. Their upward mobility was, therefore, impeded by rules restricting many different types of contact, especially sexual contact and economic transactions. The Para-s´arama-dhavı-ya, a huge mid-fourteenth-century commentary on the Para-s´arasmr.ti, contains long chapters on Untouchability rules. Judging from the amount of text that is expended, for a man from any of the four varn.as to have sex with a Can.d.a-la woman is a very serious transgression. A vast number of rules are cited, and since they prescribe different penances to repair the same transgression, each penance has to be explained in relation to a specific circumstance. The penances range from a Praja-pati Penance, which is a fast lasting twelve days, to penances that involve the suicide of the sinner. The criteria for distinguishing different degrees of seriousness are many, typical ones being intention or knowledge of the character of the act, frequency with which the sin was repeated, completion or interruption of the sexual intercourse, and fertilisation of the woman (Aktor 2008: 119–129). But what is the nature of the Can.d.a-la’s devastating touch. Are Can.d.a-las impure? Yes, they are more impure than any other caste, but this is not simply a matter of being the lowest in a continuum of impurity. The nature of the Can.d.a-la’s impurity is different. We get a very precise definition of this special quality in one of the earliest sources to the Untouchability complex. This is Patañjali’s commentary on Pa-n.ini’s rule (2.4.10) regarding the use of names of

The Ca-n.d.a-lı- as S´akti

101

certain groups of S´u-dras in the Sanskrit copulative compound (dvandva). Pa-n.ini prescribes that names of S´u-dras ‘who are not excluded’ (aniravasita) can be used in the group of dvandva compounds that have the value of neuter singular collectives (sama-ha-radvandva). In his commentary Patañjali, who like other grammarians and like many lexicographers classified Can.d.a-las and similar groups as low categories of S´u-dras, explains who the ‘excluded’ are and what the criteria are for their exclusion. Following common practice, Patañjali casts his argument as a debate, first presenting the prima facie view that segregation is a matter of exclusion from a geographic territory, the country known as Arya-va-rta. But this has to be rejected because some dvandva compounds formed by names of foreign people and known from common usage do not conform to this demarcation (that is, these names can be members of a sama-ha-radvandva even though the people that they designate live outside Arya-va-rta). Hence, it is suggested that it is a matter of being excluded not from Arya-va-rta but from Aryan settlements within this country: In that case, what is meant is rather compounds of ‘S´u-dras who are not excluded from an Aryan dwelling-place’. – But what is an Aryan dwellingplace? – It is a village, a cattle farm, a city, or a market-place. – But that means that when Can.d.a-las and Mr.tapas live inside these large habitations, there the compound ‘can.d.a-lamr.tapa-h.’ [plural] would also be wrong.7 In other words, in common parlance it is correct to say ‘can.d.a-lamr.tapa-h.’, that is ‘Can.d.a-las and Mr.tapas’ (the latter probably being guardians of the cremation ground) joined in one dvandva compound inflected in the plural, but not ‘can.d.a-lamr.tapam’ inflected as neuter singular. That being the case, these people must, in fact, be excluded from something, since only names of S´u-dras who were not excluded could figure as members in this latter type of dvandva compound. So the job is to define the criterion of exclusion. The next suggestion is exclusion from sacrificial rituals. Only Twice-born men are allowed to perform vedic sacrifices. But this does not work either, since it is also correct to speak of ‘Carpenters and Blacksmiths’ (taks.a-yaska-ram) and ‘Washermen and Weavers’ (rajakatantuva-yam) in the neuter singular despite the fact that these artisans, like any other S´u-dra, are definitely excluded from performing the sacrifice. Finally we get the solution: In that case, what is meant is compounds of ‘S´u-dras who are not excluded from the exchange of food vessels’. When food vessels are regarded as being purified after they have been used and cleaned correctly, then the people that have used them are the ‘not excluded’. When food vessels are not regarded as purified after they have been used, even though they have been cleaned correctly, then the people that have used them are the ‘excluded’.8 That is, if food can be eaten from a food vessel that has been used by another and cleaned properly afterwards, then this other person belongs to those who

102

Mikael Aktor

are not excluded. This is why it is correct to construct compounds out of names of S´u-dra artisans inflected in the neuter singular. Although they cannot perform sacrifices, they are not excluded from the exchange of food vessels that have been properly cleaned. This also explains why it is wrong to construct sama-dha-radvandva compounds with names such as ‘Can.d.a-la’ and ‘Mr.tapa’. These people carry a kind of pollution that cannot be removed by ordinary cleaning. Food vessels that have been used by them cannot under any circumstances be used by others, and the only way their names can be combined is in dvandva compounds inflected in the plural. In the Buddhist Ja-taka stories we find other early texts that confirm the same idea. The Satadhammaja-taka (Ja-t 2.82–85) is narrated as a warning against monks who make their living unlawfully. Such illicit ways are compared to eating the food of Can.d.a-las. The story tells us about a Brahmin, who wanders along the road but who has not brought any food. He meets a kind and polite Can.d.a-la who offers him food, but the arrogant Brahmin rejects this offer. However, when he starts getting hungry, he secretly takes some of the Can.d.a-la’s food thinking that if he only removes the surface layer of food, polluted by the Can.d.a-la when he had his meal, it would be acceptable. But no. Even the inner part of the food, which has not been in direct contact with the Can.d.a-la while he partook of his meal, is severely polluting. The Brahmin immediately gets sick and dies alone in the jungle. Another story, the Setaketuja-taka (Ja-t 3.232–237), describes a young Brahmin who commands a nearby Can.d.a-la to move away, because he feared that, after touching the Can.d.a-la’s body, the wind might touch his own. The idea is made explicit in Cittasambhu-taja-taka (Ja-t 4.390–401) where young women who happen to see two Can.d.a-las in the street exclaim: ‘Seeing this is an evil omen!’ (4.390). All these examples express the idea that Can.d.a-las carry with them an inborn quality that affects their surroundings. This quality is ‘magic’ in the sense that only ritual means can remove its influence. This also implies that a transformation of this quality can be potentially powerful, as are other inauspicious elements used in tantric ritual.

Menstruation Where the Yonitantra placed the Ca-n.d.a-lı- at the centre of the yoni, another text, the Br.hadyonitantra, identifies the same as pus.paru-pin.-ı, that is ‘flowering’, a general euphemism for a menstruating woman (Schoterman 1980: 24). And indeed the term ‘ca-n.d.a-lı-’ is frequently used in law texts and elsewhere to designate a woman on her first day of menstruation, that is, the most polluting day. The relation between Can.d.a-la women and menstruation in general is therefore well established. But she is not alone: She is declared to be a Can.d.a-la woman on the first day [of menstruation], a Brahmin-killer on the second day and a Washerwoman on the third day, while she is purified on the fourth day.9

The Ca-n.d.a-lı- as S´akti 103 Ma-dhava-ca-rya, the commentator of this verse from the Para-s´arasmr.ti, admits that this is metaphorical language but maintains that it should be understood in practical terms: The consequences of having sexual intercourse with a menstruating woman on these days decrease to the same extent as those of have sexual intercourse with a Can.d.a-la woman and the other women mentioned. It is with regard to this that these names are used.10 The relation between Can.d.a-las and menstruation can be traced back to the earliest texts that point out those considered to be untouchable. Apart from Can.d.a-las these include menstruating women and women who have recently given birth as well as grievious sinners and people who have touched a corpse (GDhS 14.30). Indeed, death, grievious sin, female sexuality and menstruation are all connected in a mythical aetiology: When Indra kills Vis´varu-pa, he is condemned by the three worlds as a brahmin-killer. To escape the effects of this most heinous crime, he persuades the earth, trees and women to assume one-third of his guilt each. In return, he grants each a boon: the earth, when dug, will heal within one year; trees, when cut, will grow again; and women, unlike other animals, may enjoy sexual intercourse at any time (Taittirı-yasam . hita II.5.1ff), even in advanced pregnancy …. Indra’s guilt took the form of fissures in the earth, sap in trees – hence one should avoid ‘red secretions’ and resin from cut trees (Manu V.6; cf. Tait.Sam . . II.5.4) – and menstrual blood in women. (Leslie 1989: 251, see also Leslie 1994)

- The level of mystical experience: the D . ombı of the Caryagıti This perception of the Untouchable as a polluting force both in the home (as menstruating women) and in society (as Can.d.a-las, D . ombas and other low castes) was rooted in the social norms of family and class, according to which the purity of lineage and caste was the precondition of social respect and success. At the same time these were the very same norms that centuries of world-renouncing ideologies had portrayed as a lethal trap. The attraction of women, sexual pleasure, home and family form only a thin covering over a pit of suffering, old age and death. But unlike other such ideologies the tantric movements insisted on both yoga and bhoga. For this reason, rather than simply rejecting the enjoyments of the senses (yoga), they inverted social norms in such a way that experiences considered impure according to the public social norms were attributed ritual power, allowing their enjoyment (bhoga) to become a means of attaining yogic states of consciousness.11 This inversion of norms is also articulated in relation to untouchability. One of the early expressions of this is seen in the Buddhist Sahajiya- tantric

104

Mikael Aktor

texts known as the Carya-gı-ti (Kværne 2010).12 This is a collection of songs in a seemingly simple lyrical style, in which a complex tantric cosmology and practice is encoded. Composed in the eleventh century or earlier (ibid.: 7) in a language characterised as Old Bengali or Old Maithili (ibid.: 3), these songs are attributed to a group of nineteen tantric siddhas (ibid.: 4–5), whose names also appear in other medieval lists (White 1996: 78–86). The text of the songs is embedded in a thirteenth-century exegetical commentary by Munidatta, a scholar well-trained in the Buddhist Tantras as well as in a range of older Maha-ya-na works (Kværne 2010: 2, 19). In relation to the present subject the first two verses of the tenth song are especially interesting. Per Kværne (2010) has edited the song together with Munidatta’s commentary and a Tibetan translation. His English translation of the songs reproduces the original verses with the essence of Munidatta’s exegetical gloss in brackets. Omitting these brackets, the first two verses of song 10 read: Outside the town, o D . ombı, is your hut; the shaven-headed brahmin (boy) goes constantly touching (you). - Ho D . ombı! I shall associate with you, (I,) Kan.ha, a kapali-yogin, shameless and naked. (Ibid.: 113)13 According to Munidatta, these verses apply a language of implicit meanings (sandhya-bha-s.a-, Kværne 2010: 116) in which the social level of the narrative refers to a different level of mystical experience. The two levels are opposite, and everything is turned around: In society the D . ombı occupies the lowest status as a polluting Untouchable, but here she represents the purified avadhu-ti, the central vein or psychic channel, through which the bodhicitta moves upwards towards the Cakra of Great Bliss in the head to be transformed in the supreme bliss of non-duality. The Brahmin, whose vedic knowledge and observance of purity rules secures him the highest rank in the social-religious hierarchy, is here ridiculed as a youth constantly touching the Untouchable without ever attaining the embrace of the illuminating Wisdom embodied by the D . ombı. Only the socially despised Skull-Bearer, Kan.ha, enjoys her fully. The narrative transcends the level of social hierarchy to indicate a process of transcending the sphere of the sense objects. Regardless of the explicit antinomian character of tantric texts we may here recognise an old theme in Indian literature, the inversion of social hierarchy as an expression of imparting esoteric knowledge. The fickle Brahmin boy14 of these verses is one other S´vetaketu/ Setaketu as known from the Upanis.ads and the Pali Canon (ChU 5.3–10; Ja-taka 3.232–33) whose knowledge is proven inadequate when confronted with true wisdom, always represented by a person of lower status. Both of these levels are contained in the notion of touch. At the social level of the Law Books the D . ombı is ‘untouchable’, expressed by the

The Ca-n.d.a-lı- as S´akti

105

prescriptive participle aspr.s´ya, but here, at the level of wisdom, she is instead ‘not associated with touch’, expressed by the descriptive adjective aspr.s´a/aspars´a,15 that is, she embodies a state beyond any contact with sense objects and is therefore ‘outside the town’.16 Therefore, ‘touch’ in the Carya-gı-ti songs is overlaid with several layers of psychological and esoteric meanings and Munidatta’s commentary even more so. Still, the logic is similar to that of King S´u-draka in Ka-dambarı-: What at the social level is considered a protection for the Touchables against the polluting touch of the Untouchables becomes at the level of wisdom an expression of the purity of something that in the narrative assumes the form of an Untouchable. Thus, as in Ba-n.abhat.t.a’s drama, Untouchability is turned on its head. It preserves the purity of the ‘Untouchable’ against the touch of the Touchables.

A tentative conclusion Untouchability has here been illustrated by texts in which the meaning of touch and physical contact is understood in different degrees. In Ka-dambarıtouch is physical and is connected with intimacy and love-making; the untouchability of the young Can.d.a-la woman both explains and preserves her beauty. In the Yonitantra touch is understood ritually; sexual contact is a means of producing the ritually potent yonitattva, but at the same time the Ca-n.d.a-lı- is reinterpreted as a goddess and is thereby clearly referring to a different level of meaning than that of social hierarchy. In the Carya-gı-ti song, and especially in Munidatta’s interpretation of it, the psychological and esoteric meanings of touch and the D . ombı seem to dominate. ‘Purity’ is ambivalent; the untouchable D . ombı is an adequate expression of a notion of purity beyond any purity/impurity dualism. All three texts assume a two-levelled world. On the surface there is the world of the respectable life according to the expectations of family and class. Here kinship is the leading principle, and sexuality is seen in the context of preserving and protecting the family line. Accordingly, there are good wombs and bad wombs, and Untouchability preserves this distinction. The foul yoni of the Ca-n.d.a-lı- in the Cha-ndogyaupanis.ad is the womb of a mother. But there is another secret level where sexuality is free from associations with mothers, children and family obligations. This is the level of King S´u-draka’s dreams of lovemaking with the beautiful, unspoiled Can.d.a-la woman, and this is the level of the Siddhas’ experiments with sexual energy and transcendent states of consciousness. Rejecting the values of family and social hierarchy while insisting on enjoyment means that relations are turned upside down. Seen from this level, observance of Untouchability rules represents the unripe, unenlightened state of the public social world. At the same time it preserves the unspoiled s´akti, which is embodied by young, untouchable and ever childless women, whose yoni becomes at once the source of sexual joy and, allegedly, the opening to blissful states of consciousness beyond any contact with the world of sense experience.

106

Mikael Aktor

Notes 1 yonimadhye pradha-na- ca ca-n.d.a-lı- gan.ana-yika- / tasya-h. pu-janama-tren.a mama tulyo na sam . cayah. // (YT 3.20). - - 2 antah. s´a-kta- bahih. s´aiva-h. sabha-ya-m . vais.n.ava matah. / nanarupadharah. kaula vicaranti mahı-tale // (Schoterman 1980: 16; YT 4.20). . 3 nat.-ı ka-pa-lika- ves´ya- rajakı- na-pita-ngana- / bra-hman.-ı s´u-drakanya- ca tatha- gopa-lakanyaka // malakarasya kanya ca nava kanya-h. prakı-rtita-h. // athava- sarvaja-tı-ya- vidagdha- lolalocana- // ma-tr.yonim . parityajya sarvayonim . ca tad.ayet // dvadas´abdadhikayonim . - naiva pu-jayet // aksata-ya-vat s.as.t.im . samapayet // … ks.atayonih. pujitavya aks.atam . . pu-jana-d devi siddhiha-nih. pade pade // (YT 2.3–5, 3.25cd-26). 4 atha ya iha kapu-yacaran.a- abhya-s´o ha yat te kapu-ya-m . yonim apadyerañ s´vayonim . - canda-layonim va-. (ChU 5.10.7). va- su-karayonim va . .. . 5 aham . [Vais´ampayana] tu … itastato visragandhidhu.modgamanumıyamanasan- ntaritaves´masamnives´am, sarvatah karankapra-yavrtiva-tam, asthi´ dravam s avana . . . . . . - pra-yarathya-vakaraku-t.am, utkr.ttama-m . samedovasasr.kkardamaprayakut.ırajiram, - ya-s´anam, vasa-pra-yasneham, kaus´eyapra-yapar´ a-kh- et.akapra-ya-jı-vam , pis itapra . . . - - idha-nam . , carmaprayastaran.am . , sarameyaprayaparivaram . , dhavalıprayavahanam . , strı - paha-r´ madyapra-yapurus.a-rtham, asr.kpra-yadevata-balipu-jam , pas u . - - , apra-yadharmakriyam, a-karam iva sarvanaraka-n.a-m . , karan.am iva sarvakus´alanam . - na-na-m, pattanam iva sarvapa-pa-na-m, a-yatanam iva sar´ ´ ´ sam nives am iva sarvas mas a . . vaya-tana-na-m . … lokahr.dayebhyo ’pi nirghr.n.atarasarvasam ..vyavaharasamastapurus.m, avis´es.a-ca-raba-layuv-asthaviram, avyavasthitagamya-gamya-nganopabhogam, apun.yakarmaika-pan.am . pakkan.am apas´yam / (Kad, pp. 504–506). 6 mu-rcham iva manoha-rin.-ım … amu-rtam iva spars´avarjita-m, a-lekhyagata-m iva dars´anama-traphala-m … aciroparu-d.hayauvana-m, atis´ayaru-pa-kr.tim, animes.alocano dadars´a / samupaja-tavismayasya ca-bhu-n manasi mahı-pateh. / aho vidha-tur astha-ne ru-panis.pa-danaprayatnah. / tatha- hi / yadi na-meyam a-tmaru-popahasita-s´es.ar- u-pasam . pad utpa. dita kim artham apagataspars´asam . bhogasukhe kr.tam . kule janma / manye ca matangaja-tispars´ados.abhaya-d aspr.s´ateyam utpa-dita- praja-patina- / anyathakatham iyam aklis.t.ata- la-vanyasya / na hi karatalaspars´akles.ita-na-m avayava-na-m -ı dr.s´-ı bhavati ka-ntih. / sarvatha- dhig vidha-ta-ram asadr.s´asam . yogakarin.am atimanohara-kr.tir api kru-raja-titaya- yeneyam asuras´rı-r iva satataninditasurata- raman.-ıya-py udvejayati / (Ka-d, pp. 24–25). - - 7 evam . tarhy aryanivasad aniravasitanam / kah. punar aryanivasah. / gramo ghos.o nagaram . sam . vaha iti / evam api ya ete mahantah. sam . styayas tes.v abhyantaras´ can.d.a-la- mr.tapa-s´ ca vasanti tatra can.d.a-lamr.tapa- iti na siddhyati // (MBha-s. 1.475.4–7). - - 8 evam . tarhi patrad aniravasitanam / yair bhukte patram . sam . skaren.a s´uddhyati te - rena-pi na s´uddhyati te niravasita-h // (MBa-s ‘niravasita-h. / yair bhukte pa-tram sam ska . . . . . 1.475.8–10). 9 prathame ’hani can.d.alı dvitıye brahmaghatinı / tr.tıye rajakı prokta caturthe ’hani s´udhyati // (PS 2.7.18c-19b). 10 can.d.a-lya-digamane ya-va-n pratyava-yas ta-va-n udakya-gamana ity abhipretya tair na-mabhir vyavaha-rah. / (PM 2.7.18c-19b, p. 168). 11 Admittedly, I use the yoga-bhoga dichotomy here as a broadly accepted trope of ‘Tantra’ rather than with precise historical reference. This general use goes back to, among others, Louis Dumont in his famous essay on world renunciation (Dumont 1980: 279) where he quotes Kula-rn.avatantra 2.23: yogı- cen naiva bhogı- sya-d bhogıcen naiva yogavit / bhogayoga-tmakam . kaulam . tasmat sarvadhikam priye // Yonitantra 3.18c-19b is a variant of the same verse: yogı cen naiva bhogı- sya-d bhogı- ca na tu yogava-n / yogabhoga-tmakam . kaulam . yadi yoniprapujakah. // Both variants express the idea that the essence of Kaula doctrine is the union of what on the surface seems a contradiction, i.e. yoga and bhoga. The YT variant emphasises the connection to the yonipu-ja-.

The Ca-n.d.a-lı- as S´akti

107

12 I want to thank Dr Hartmut Buescher for his keen and critical comments to a previous draft of this section on the Carya-gı-ti. 13 nagara ba-hiri re d.ombi tohori kur.ia- / choi choi ya-i ba-hman.a na-r.iya- // a-lo d.ombı- toe . sama karibe ma sa-nga / nighin.a Ka-n.ha ka-pa-li joi la-ga // (Kværne 2010: 113). 14 According to Munidatta, the Brahmin with his childish and fickle mind represents the Bodhicitta of the uninstructed yogin: bra-hman.a iti … capalayogatva-t cittaba- bodhicittam (Kværne 2010: 116). t.ukam . / asampradayayoginam . . 15 aspr.s´ayogatva-t d.ombı- iti (Munidatta in Kværne 2010: 116); aspars´a- bhagavatıyasma-t tasma-d d.ombı- prakathyate (HT 1.5.18; Snellgrove 2010, part 2: 18). 16 The full passage about the D . ombı and the Brahman.a reads: aspr.s´a-yogatvat d.ombı iti paris´uddha-avadhu-tı- naira-tma- boddhavya- / bra-hman.a iti brahma-hu-m . -kara-bıja- bodhicittam samvrtija-tam . capala-yogatvat citta-bat.ukam . / asampradaya-yoginam . . . . - s´ukra-ru-pam . man.imulad viramanandam . spr.s.t.va spr.s.t.va gacchati / (Kværne 2010: 116, his orthography).

References Aktor, Mikael. 2008. Ritualisation and Segregation: The Untouchability Complex in Indian Dharma Literature with Special Reference to Para-s´arasmr.ti and Para-s´arama-dhavı-ya, Corpus Iuris Sanscriticum et fontes iuris Asiae Meridianae et Centralis, ed. Oscar Botto, vol. 9. Turin: Comito ‘Corpus Iuris Sanscriticum’/CESMEO. Aktor, Mikael. 2010. “Untouchability”. In Brill’s Encyclopedia of Hinduism, vol. 2, ed. Knut A. Jacobsen, Helene Basu, Angelika Malinar and Vasudha Narayanan, pp. 876–881. Leiden: Brill. Bhattacharyya, N.N. 2005 [1999, 1982]. History of the Tantric Religion: An Historical, Ritualistic and Philosophical Study. 2nd rev. edn. New Delhi: Manohar. BORI. 1971–1976. The Maha-bha-rata: Text as Constituted in Its Critical Edition. 5 vols. Pune: Bhandarkar Oriental Research Institute. Dumont, Louis. 1980. Homo Hierarchicus: The Caste System and Its Implications. Rev. edn. Chicago, IL: University of Chicago Press. Ka-le, Mores´vara Ra-macandra. 1895–1896. Ka-dambarı- by Ba-n.abhat.t.a and His Son. 2 vols. Mumbai: Sha-rada-krı-dan Press. Kielhorn, F. 1962–1972 [1880–1885]. The Vya-karan.a-Maha-bha-s.ya of Patañjali. 3rd edn, 3 vols. Pune: Bhandarkar Oriental Research Institute. Kværne, Per. 2010 [1977]. An Anthology of Buddhist Tantric Songs: A Study of the Carya-gı-ti. 3rd edn. Bangkok: Orchid Press. Layne, Gwendolyn (transl.). 1991. Ka-dambarı-: A Classic Sanskrit Story of Magical Transformations. New York: Garland Publishing. Leslie, I. Julia. 1989. The Perfect Wife (Strı-dharmapaddhati). Oxford: Oxford University Press. Leslie, I. Julia. 1994. “Some Traditional Indian Views on Menstruation and Female Sexuality”. In Sexual Knowledge, Sexual Science: The History of Attitudes to Sexuality, ed. Roy Porter and Mikuláš Teich, pp. 63–81. Cambridge: Cambridge University Press. Olivelle, Patrick. 1998. The Early Upanis.ads: Annotated Text and Translation. New York: Oxford University Press. Olivelle, Patrick. 2000. Dharmasu-tras: The Law Codes of Apastamba, Gautama, Baudha-yana, and Vasis.t.ha. Delhi: Motilal Banarsidass. Olivelle, Patrick. 2005. Manu’s Code of Law: A Critical Edition and Translation of the Ma-nava-Dharmas´a-stra. New York: Oxford University Press.

108

Mikael Aktor

Schoterman, J.A. 1980. The Yonitantra: Critically Edited with an Introduction by J.A. Schoterman. New Delhi: Manohar. Smith, David. 2009. Princess Ka-dambarı-. vol. 1. The Clay Sanskrit Library. New York: New York University Press. Snellgrove, David. 2010 [1959]. The Hevajra Tantra: A Critical Study. 2nd edn. Bangkok: Orchid Press. White, David Gordon. 1996. The Alchemical Body: Siddha Traditions in Medieval India. Chicago, IL: University of Chicago Press. White, David Gordon. 2003. Kiss of the Yoginı-: ‘Tantric Sex’ in Its South Asian Contexts. Chicago, IL: University of Chicago Press.

6

S´a-ktism and hat.hayoga1 James Mallinson

Introduction In this paper I shall examine one of the ways that S´a-ktism endured after its heyday, the ‘S´aiva Age’ magisterially documented by Alexis Sanderson in a recent book-length article.2 The thirteenth century, the end of S´aivism’s period of dominance, coincided with the rise of yoga, specifically yoga which used the techniques known as hat.ha, to a position of dominance among the soteriological methods employed in India.3 It is hat.hayoga’s relationship with S´aivism and its S´a-kta manifestations that is the subject of this paper. Scholarship on hat.hayoga, my own included, unanimously declares it to be a reformation of tantric yoga introduced by the gurus of the Na-th sam . prada-ya, in particular their supposed founder, Goraks.a. In much secondary literature the phrases ‘Na-th yoga’ and ‘hat.hayoga’ are used interchangeably. When other traditions are seen to employ the practices or terminology of hat.hayoga, they are said to be borrowing from the Na-ths.4 Goraks.a, who probably flourished in the twelfth century, and Matsyendra, who according to tradition was Goraks.a’s guru but is likely to have lived three centuries before him, were exponents of the S´a-kta cult of the Pas´cima-mna-ya or western stream of Kaula S´aivism.5 To this day the Na-th Yogı-s are renowned as ta-ntrika adepts and their monasteries are often situated near important goddess temples, such as Jwalamukhi in the Kangra district of Himachal Pradesh or Devi Pattan near the India–Nepal border. The Pas´cima-mna-ya tradition is no longer extant and Na-th liturgy is now more in keeping with the tradition of the foremost surviving S´a-kta cult, that of the Daks.in.a-mna-ya or southern stream of Kaula S´aivism, with the goddess Ba-la- or Tripura--sundarı- being the focus of their worship.6 The Na-ths’ S´a-kta credentials are pukka.7 The ascription to them of both the invention of hat.hayoga and the composition of the corpus of Sanskrit texts which teach its practice stands on shakier ground: it derives from the claims of the Na-ths themselves and from the fact that of the small number of texts on hat.hayoga that have been edited, some are ascribed to Na-th gurus or mention them as the revealers of their doctrines. But the corpus of Sanskrit works on hat.hayoga – which is our only source for hat.hayoga’s formative

110

James Mallinson

period – has, perhaps surprisingly considering the widespread popularity of yoga today, been the subject of very little critical study. Much of my research of the last few years has concentrated on identifying the texts that constitute this corpus and using them to examine how hat.hayoga developed and who practised it. In the first part of this paper I shall summarise what constituted early hat.hayoga. I shall then show how, in contemporaneous taxonomies of yoga, S´a-kta techniques were grouped separately from hat.hayoga, under the name laya. Next I shall show how in the Hat.hapradı-pika-, the text which became hat.hayoga’s locus classicus, the S´a-kta techniques of layayoga were included under the rubric of hat.ha and how at the same time the purpose of hat.hayoga was realigned to be more in keeping with that of laya. I shall then identify the practitioners of early hat.hayoga – who were not S´a-ktas – and show how they have continued to be its torchbearers. In the final part of the paper I shall attempt to locate these developments in their wider context.

Early hat.hayoga The earliest references to hat.hayoga are scattered mentions in Buddhist canonical works and their exegesis dating from the eighth century onwards, in which it is the soteriological method of last resort.8 In its earliest definition, in Pun.d.arı-ka’s eleventh-century Vimalaprabha- commentary on the Ka-lacakratantra, hat.hayoga9 is said to bring about the ‘unchanging moment’ (aks.araks.an.a) ‘through the practice of na-da by forcefully making the breath enter the central channel and through restraining the bindu of the bodhicitta in the vajra of the lotus of wisdom’.10 While the means employed are not specified, the ends, in particular restraining bindu, semen, and making the breath enter the central channel, are similar to those mentioned in the earliest descriptions of the practices of hat.hayoga, to which I now turn. In seeking to establish a corpus of early works on the practices of hat.hayoga we are greatly assisted by the Hat.hapradı-pika-, which can be dated to approximately 1450.11 Its composer, Sva-tma-ra-ma, used verses from at least twenty texts12 to compile his hugely influential work, whose stated aim is to be a light on hat.ha ‘in the darkness of a multitude of doctrines’.13 Only four of the texts drawn upon by Sva-tma-ra-ma mention hat.ha or hat.hayoga by name14 and of those only one, the Datta-treyayogas´a-stra, a Vais.n.ava work, gives a detailed exposition of its techniques. Its hat.hayoga consists of a yoga of the eight limbs also taught in, for example, Patañjali’s Yogasu-tra but here ascribed to Ya-jñavalkya and others, or an alternative set of practices employed by Kapila and other siddhas, practices which are thus the distinguishing feature of early hat.hayoga. These practices comprise ten physical techniques which in later works all came to be classed as mudra-s. Some of these techniques are relatively simple attitudes to be assumed in meditation, ascetic practice or breath control, forerunners of which are taught or mentioned in earlier works;15 the others are unique to hat.hayoga. Their primary aim, which they effect either pneumatically,

S´a-ktism and hat.hayoga

111

by making the breath enter the central channel and rise upwards, or mechanically, is to stop the lunar bindu, or semen, which is equated with amr.ta, the nectar of immortality, from dripping down from its store in the head and being consumed in the solar fire at the base of the central channel. The techniques of early hat.hayoga are thus direct methods of addressing the ancient (and still prevalent) Indic concern with the preservation of semen, which results from its being considered to be the vital principle and its loss being thought to lead to weakness and death. The hat.hayogin might practise viparı-takaran.-ı, inverting himself in order to use gravity to keep his bindu in his head. Or he can, applying khecarı-mudra-, insert his tongue into the opening behind his uvula thereby sealing bindu in the cranial cavity. Or he might, should he ejaculate, create a vacuum in his abdomen and resorb his bindu by means of vajrolimudra-. Eight of the works used to compile the Hat.hapradı-pika- teach one or more of the hat.ha techniques ascribed to Kapila in the Datta-treyayogas´a-stra. These constitute the corpus of Sanskrit texts on what I call ‘early’ hat.hayoga in contrast to the more catholic ‘classical’ hat.hayoga of the Hat.hapradı-pika- and subsequent works.16 The texts of the corpus are, in approximate chronological order,17 the following: Amaraughas´a-sana Datta-treyayogas´a-stra Goraks.as´ataka Vivekama-rtan.d.a Yogabı-ja Khecarı-vidyaAmaraughaprabodha S´ivasam . hita I shall now briefly summarise the techniques of hat.hayoga as taught in each of these works (other than the Datta-treyayogas´a-stra, which has already been mentioned).

Amr.tasiddhi The sectarian origins of the Amr.tasiddhi are unclear; it is explicitly S´aiva but contains no specifically S´a-kta teachings.18 The hat.ha techniques taught in the Amr.tasiddhi (which are not named hat.ha) are used to make the breath enter the central channel and raise it upwards in order to reverse the usual direction of flow in the channels of the body and stop bindu from moving downwards.19 The Amr.tasiddhi is the first text to teach the pervasive hat.hayogic physiology in which the moon is situated at the top of the central channel from where it rains down amr.ta which, if preventative yogic techniques are not employed, is then burnt up in the sun situated at the bottom of the central channel.20 It also introduces the widespread hat.hayogic notion of the identification of the

112

James Mallinson

breath, semen and mind: by stopping any one of these three, all three are stopped.21 The principle aim of the hat.ha techniques of the Amr.tasiddhi is thus the same as those of the Datta-treyayogas´a-stra: to stop or reverse the movement of bindu. Neither text associates the practices of hat.ha with Kun.d.alinı- or the cakras.22

Goraks.as´ataka

23

Probably contemporaneous with the Datta-treyayogas´a-stra, i.e. composed in the thirteenth century, are the first texts on yoga to be associated with Goraks.a, the Goraks.as´ataka and Vivekama-rtan.d.a. Unlike the Datta-treyayogas´a-stra, neither of these works calls its yoga hat.ha; it is just yoga. The Goraks.as´ataka teaches that liberation is to be attained by controlling the mind through controlling the breath. One method of controlling the breath is to stimulate Kun.d.alinı-, which can be done either by using the three hat.hayogic bandhas mentioned earlier,24 or through sarasvatı-ca-lana, ‘stimulating Sarasvatı-’. Sarasvatı-° or s´akti-ca-lana is not taught in the Datta-treyayogas´a-stra but it is included among the mudra-s of the Hat.hapradı-pika-’s classical hat.hayoga. In the Goraks.as´ataka, it involves wrapping in a cloth the tongue, which is identified with Sarasvatı- and said to be the goddess at one end of the central channel, and tugging on it in order to stimulate Kun.d.alinı-, who dwells at the other end.25 Nowhere in the Goraks.as´ataka is the preservation of bindu or amr.ta mentioned.

Vivekama-rtan.d.a The framework of the text of the Vivekama-rtan.d.a is a description of the wellknown six cakras of yogic physiology, within which are passages on a variety of yogic subjects including an enumeration of a sixfold yoga and a section on Kun.d.alinı- in which awakening her by means of ‘fire yoga’ (vahniyogena) and raising her, together with the mind and the breath, to the brahmadva-ra, is said to be the way for yogins to become liberated.26 Immediately after this passage on Kun.d.alinı- comes a description of five hat.hayogic mudra-s,27 none of which is said to bring about the raising of Kun.d.alinı-. They have a range of physical benefits, including the raising of the apa-na breath, but the aim most emphasised in their descriptions, particularly in that of khecarı-mudra-, is the preservation of bindu. Later in the text the hat.hayogic mudra- called viparı-takaran.-ı, ‘the inverter’, is taught and this too is said to be a method for stopping the downward flow of nectar; Kun.d.alinı- is not mentioned.28

Yogabı-ja The Yogabı-ja29 is a dialogue between S´iva and the goddess. It makes no mention of any Na-th gurus. Its yoga is similar to that of the Goraks.as´ataka in

S´a-ktism and hat.hayoga 113 that it uses the three hat.hayogic bandhas and s´aktica-lanı- mudra-. The awakening of Kun.d.alinı- is the purpose of many of the yoga practices taught in the text;30 preserving bindu or amr.ta is mentioned in passing twice.31

Khecarı-vidyaThe Khecarı-vidya- is a composite text in which teachings on the practice of the hat.hayogic khecarı-mudra- have been inserted into part of a Kaula work that includes a coded description of a mantra (the Khecarı-vidya- of the text’s title), and a eulogy of madira-, alcohol, which is reworked into praise of Khecarı-. The practice of khecarı-mudra- enables the yogin to access various stores of amr.ta in the body and also to raise Kun.d.alinı- via the six cakras to the great store of amr.ta in the head, with which she floods the body on her journey back down to her home at the a-dha-ra, the ‘base’ located at the perineum. The Vivekama-rtan.d.a is mentioned near the beginning of the Khecarı-vidya-,32 but despite that text’s clear assertions that the purpose of khecarı-mudra- is the sealing of bindu or amr.ta in the head, that aim is not mentioned in the Khecarı-vidya(and it would be hard to reconcile with the flooding of the body with amr.ta). In many respects the Khecarı-vidya-’s khecarı-mudra- has more in common with the tongue-pulling s´aktica-lanı- mudra- of the Goraks.as´ataka than with the cavity-sealing khecarı-mudra- taught in the Datta-treyayogas´a-stra and Vivekama-rtan.d.a.

Amaraughaprabodha The Amaraughaprabodha33 is ascribed to Goraks.a in its manuscript colophons and mentions four gurus associated with the Na-th order in its opening verse.34 It teaches a hat.hayoga (named as such) which is very similar to the bindudha-ran.a yoga taught in the Amr.tasiddhi, but it adds the awakening of Kun.d.alinı- to the benefits of the Amr.tasiddhi’s maha-mudra- and it mentions a dhya-na of s´akti.35

S´ivasam . hita The last of the texts which constitute the corpus of works on early hat.hayoga - 36 is the S´ivasam . hita. There is nothing in the S´ivasam . hita to associate it with the Naths. On the contrary, various features of the text show it to be a product of the tradition of S´aiva non-dualism mixed with the expurgated daks.in.a-mna-ya Kaula cult of Tripurasundarı- known as S´rı-vidya-.37 This cult became the most widespread and enduring of the S´a-kta traditions. The S´a-kta orientation of the S´ivasam . hita is made plain at the end of its teachings on the benefits of increasing numbers of repetitions of the S´rı-vidyamantrara-ja: through thirty lakh repetitions the practitioner becomes equal to Brahma- and Vis.n.u; through sixty lakh he attains Rudra-hood; through eighty

114

James Mallinson

lakh, the s´aktitattva. Finally, through one crore repetitions, he is absorbed into the absolute.38 The S´ivasam . hita mentions hat.hayoga by name but does not define it, nor does it make it clear which of the many practices it teaches come under its rubric. In its fourth pat.ala it teaches all ten hat.hayogic mudra-s taught in earlier works on hat.hayoga and, further confirming its roots in the traditions of S´rı-vidya-, crowns them with its own, a hat.hayogic variety of the hand-gesture yonimudra- which occupies a central place in the rituals taught in the Va-makes´varı-mata, one of the foundational texts of S´rı-vidya-.39 The S´ivasam . hita contains detailed teachings on Kun.d.alinı-, who is said to be the ‘great goddess’ (paradevata-)40 and to ‘take the form of the creation of the universe’ (jagat- - 41 sam . sr.s.t.irupa). The purpose of its hat.hayogic mudras is to awaken Kun.d.alinı 42 and make her pierce a variety of lotuses and knots as she rises upwards. Despite incorporating and reworking many of the Amr.tasiddhi’s verses on maha-mudra-, maha-bandha and maha-vedha, and several verses from the Datta-treyayogas´a-stra, nowhere does the S´ivasam . hita mention bindudharan.a as the aim of any practice of yoga. Even vajrolimudra, the practice of urethral suction which in the Datta-treyayogas´a-stra and the Hat.hapradı-pika- is for the - 44 preservation of the yogin’s bindu,43 is made purely S´a-kta in the S´ivasam . hita, where its purpose becomes the absorption of one’s partner’s bindu or rajas, in order to combine them within one’s own body.45 So, to summarise early hat.hayoga: in its earliest formulations, which are found in texts that are neither S´a-kta nor associated with the gurus of the Na-th sam . pradaya, hat.ha’s distinguishing feature is a variety of physical techniques which are used to keep bindu or amr.ta, i.e. semen, in the head. In contrast, contemporaneous Na-th works emphasise the purpose of their yoga, which they do not call hat.ha, as being the raising of Kun.d.alinı-. To this end, the Goraks.as´ataka and Yogabı-ja prescribe s´aktica-lanı- mudra-, the one physical yoga technique exclusive to early Na-th yoga. Meanwhile another early Na-th work, the Vivekama-rtan.d.a, co-opts the mudra-s of hat.hayoga but they are still said to work only on bindu, not Kun.d.alinı-. In later Na-th and other S´a-kta works of the canon, the co-option of the hat.ha techniques (along with their name) is more developed, so that in the S´ivasam . hita the purpose of the hat.hayogic mudras has become the raising of Kun.d.alinı and bindudharan.a is not mentioned.

Other methods of yoga I shall now turn to a typology of yoga which became commonplace and which, like the description of hat.ha practices, is found for the first time in the Datta-treyayogas´a-stra. Datta-treya teaches three methods of yoga: mantra, laya and hat.ha. All three lead to ra-jayoga, i.e. sama-dhi.46 In some texts these different yogas are hierarchised and said to be suitable for corresponding degrees of aspirant.47 In contrast to the attitude implied in the Buddhist texts referred to at the beginning of this paper, which place mantrayoga at the centre of their systems of practice and mention hat.hayoga

S´a-ktism and hat.hayoga

115

as the yoga of last resort, in hat.hayogic works mantrayoga is for the lowest level of aspirant,48 middling aspirants are suited to layayoga 49 and hat.hayoga is for those of the highest calibre.50 Mantrayoga Mantrayoga corresponds to the central practice of the tantras of the S´aiva Mantrama-rga, namely the repetition of mantras in order to obtain siddhis. Of the Sanskrit works on early hat.hayoga, only the two texts most obviously produced in S´a-kta milieux, namely the Khecarı-vidya- and S´ivasam . hita, teach mantras for obtaining siddhis.51 In other works mantra practice is omitted altogether52 or reduced to either the ajapa- ga-yatrı- (the involuntary repetition of ha on the outbreath and sa on the inbreath)53 or the repetition of om . as a purificatory technique.54 Layayoga Layayoga means ‘yoga through dissolution’. It is achieved by a wide variety of methods, the best known of which today is the raising of the serpent goddess Kun.d.alinı- upwards from the base of the spine to union with S´iva in the head via a sequence of, usually, six cakras. The cakras are associated with progressively more subtle elements and Kun.d.alinı-’s upward journey represents a reversal of creation, a laya or ‘dissolution’.55 The Datta-treyayogas´a-stra gives a mythological explanation of the origin of layayoga which I shall now translate in full:56 [Datta-treya said] ‘Layayoga happens as a result of the dissolution of the mind by means of sam . ketas, “esoteric techniques”. Adinatha has taught eighty million sam . ketas.’ Sa-m . kr.ti said: ‘Please tell me, what form does Lord Adina-tha take? Who is he?’ Dattatreya said: ‘The names of Maha-deva, the great god, are Adina-tha, Bhairava and Lord of the S´abaras. While that mighty god was sporting playfully with Pa-rvatı- in the company of the leaders of his troop in [various places such as] Mount S´rı-kan.t.ha, S´rı-parvata, the top of a mountain in the region of the Banana Forest, [and] the mountain at Citraku-t.a covered with beauti. ful trees, he, S´ankara, out of compassion secretly told an esoteric technique to each of them in those places. I, however, cannot teach all of them in detail. I shall gladly proclaim some of them, [such as this one] which consists of a simple practice and is easy: While staying still [or] moving, sleeping [or] eating, day and night one should meditate on emptiness. This is one sam . keta taught by S´iva. Another is said to be simply staring at the tip of the nose. And meditation on the rear part of the head

116

James Mallinson conquers death. The next sam . keta is said to be simply staring between the eyebrows. And that which is [staring] at the flat part of the forehead between the brows is said to be excellent. [Another] excellent dissolution is [staring] at the big toes of the left and right feet. Lying supine on the ground like a corpse is also said to be an excellent [dissolution]. If one practises in a place free from people while relaxed, one will achieve suc. cess. S´ankara has thus taught many sam . ketas. That dissolution of the mind which occurs by means of [these] and several other esoteric techniques is layayoga. Next hear about hat.hayoga.’

57 Other works from the same period also mention sam . ketas taught by S´iva. Jñandev’s commentary on the Bhagavadgıta-, the Bha-va-rthadıpika- (popularly known as the Jña-nes´varı-), concludes its lengthy teachings on Kun.d.alinı- by declaring:58

The swallowing of the body by the body: this is a sam . keta of S´iva, although here it is Lord Vis.n.u who explains it. The Yogabı-ja, whose yoga is very similar to that of the Jña-nes´varı-, concludes its description of the raising of Kun.d.alinı- in a similar fashion, saying:59 This alone is S´iva’s sam . keta; it is characterised by [also] being the sam . keta of the siddhas. These sam . ketas thus denote a variety of practices said to have been taught by S´iva, some of which have S´a-kta overtones, in particular the raising of Kun.d.alinı-, who in the Jña-nes´varı- passage cited above is said to be ‘mother of the world’ and ‘the highest goddess’.60 Other roughly contemporaneous texts also give teachings on laya (without calling its techniques sam . ketas). These are all overtly S´aiva and some also originate in milieux influenced by S´aivism’s more S´a-kta manifestions. Foremost is the practice of na-da-nusandha-na (which is also often called simply na-da): concentration on the internal sounds which arise in the course of yogic practice, sometimes as sequences of progressively more subtle sounds that correspond to the stages of Kun.d.alinı-’s ascent. The S´iva61 sam . hita says that nada is the best form of laya. Sequences of the sounds of nada are taught in a variety of S´aiva works, including S´a-kta Tantras such as - 62 the Brahmaya-mala, Kubjika-mata and Matsyendrasam . hita. The second khan.d.a of the Amanaska, which is likely to predate the Datta-treyayogas´a-stra by a century or two and shares twenty-two verses with the Kula-rn.avatantra,63 also mentions laya among the results of its yoga.64 Its yoga’s principal method is s´a-mbhavı-mudra-, in which the yogin is to gaze outwards unblinkingly, while focussing internally. This technique is said first to have been given by S´iva to Uma-, the primal S´akti. It is also taught in the Candra-valokana of Matsyendra.65 In the Amanaska’s first khan.d.a, which probably postdates the Hat.hapradı-pika- by about a century, the rewards of gradually increasing

S´a-ktism and hat.hayoga

117

periods of laya are taught. In the last, which results from remaining in laya for twenty-four years, the yogin becomes absorbed in the s´aktitattva.66 . In the S´a-rngadharapaddhati, a lengthy compendium of verses on a wide range of subjects compiled near Jaipur in 1363, laya is said to have been mastered by Kr.s.n.advaipa-yana and others, and to involve a series of meditations working upwards through nine cakras and finally bringing about the union of the body’s three s´aktis.67 The Amaraughaprabodha defines laya as . meditation on S´iva in the form of a linga at Ka-maru-pa’s location in the body, 68 and as flowing amr.ta. The Khecarı-vidya- does not define laya but does teach that after five months of visualisation of Kun.d.alinı-’s ascent and her flooding the body with amr.ta, the yogin achieves laya in the five elements69 and it associates laya with unmanı-, the ‘supramental state’.70 To summarise: the sam . ketas or ‘secret techniques’ of layayoga comprise a variety of meditations on, and visualisations of, places and energies in the body, in contrast with the physical practices of hat.hayoga. They are taught by S´iva, often in works of S´a-kta orientation. The yoga taught in the early texts associated with Na-th gurus, with its emphasis on the raising of Kun.d.alinı-, corresponds more to the techniques of laya than to those of hat.ha. An early and disjointed attempt at combining the two can be found in the Vivekama-rtan.d.a; the somewhat later S´ivasam . hita presents a more coherent synthesis; the most influential fusion is that found in the Hat.hapradı-pika-, to which we now turn. The classical hat.hayoga of the Hat.hapradı-pikaSoon after its compilation by Sva-tma-ra-ma in the fifteenth century, the Hat.hapradı-pika- became the most influential work on hat.hayoga, in part, no doubt, because it was the first text explicitly to make hat.ha its central concern. Texts that taught the hierarchy of mantra, laya and hat.ha yogas71 were known to Sva-tma-ra-ma, but he reduced their typology to a distinction between hat.hayoga and ra-jayoga, excluding mantra and laya yoga. The practices taught in the Hat.hapradı-pika- are all hat.ha and they lead to ra-jayoga, i.e. sama-dhi. Mantra practice is notable by its complete absence in the Hat.hapradı-pika-.72 In contrast, many of the techniques of layayoga are incorporated wholesale73 and are thus for the first time taught under the rubric of hat.hayoga.74 It is also in the Hat.hapradı-pika- that various other practices which were to become emblematic of hat.hayoga are for the first time taught as part of hat.ha. These include non-seated a-sanas, complex kumbhakas (methods of breathing), and the s.at. karma-n.i, six techniques for cleaning the body. We see in the Hat.hapradı-pika- the culmination of the process begun in the Vivekama-rtan.d.a, in which the S´a-kta laya technique of the visualisation of the rising of Kun.d.alinı- is overlaid onto the physical techniques of hat.hayoga, techniques originally used for the purpose of bindudha-ran.a, the preservation of semen.75 The mudra-s taught in the Datta-treyayogas´a-stra and Vivekama-rtan.d.a as methods of bindudha-ran.a are now said to be for raising Kun.d.alinı-.76

118

James Mallinson The Hat.hapradı-pika-’s blanket agglomeration of the various techniques of hat.hayoga and layayoga creates a somewhat incoherent whole. The shoehorning of Kun.d.alinı- into the subtle physiology first found in the Amr.tasiddhi, in which the moon is in the head and the sun already occupies Kun.d.alinı-’s seat at the a-dha-ra at the base of the spine, is awkward.77 The Hat.hapradı-pika- teaches khecarı-mudra- twice:78 the first keeps bindu in the head and the second floods the body with amr.ta. The reformulation of the hat.hayogic mudra-s as means of raising Kun.d.alinı- - found in the S´ivasam . hita is more coherent. Unlike Svatmarama, the compiler - completely removed bindudha-rana from the aims of the of the S´ivasam hita . . hat.hayogic mudra-s. The coherence of the S´ivasam . hita is partly because its compiler was happy to flaunt his sectarian affiliation, giving him a freer hand than Sva-tma-ra-ma had. As noted above, the S´ivasam . hita is a product of the S´a-kta S´rı-vidya- tradition. As well as that cult’s mantras it teaches detailed visualisations of the cakras through which its techniques make Kun.d.alinı- rise. Each cakra is to be visualised as having a specific group of syllables in its spokes, together with a siddha and a yoginı-, and the yoginı-s correspond to those taught elsewhere in the texts of S´rı-vidya-.79 In contrast, the Hat.hapradı-pika-, despite stating that the purpose of the hat.hayogic mudra-s is the raising of Kun.d.alinı-, makes no mention of cakras. Sva-tma-ra-ma, in addition to casting light on hat.ha, had two unspoken further aims when compiling the Hat.hapradı-pika-. He sought to lay claim to hat.hayoga for the siddha tradition while continuing a process that had started at least two centuries earlier, in which S´aiva yoga was being severed from its sectarian roots. The beginnings of this process can be seen in the teachings on yoga found in the Matsyendrasam . hita, in the introduction to his edition of which Csaba Kiss has made the observation that the text’s cult is indicative of: A phase in the history of yoga when yogic teachings start to become detached (perhaps not for the first time) from the mainstream religion, in this case tantric S´aivism, by eliminating sectarian boundaries through the concealment of sectarian marks such as easily decodable deity names, mantras and iconography and start to prepare for a formative period of a pan-Indian yoga, which can again become an alternative for the official/ conservative religion.80 Thus Sva-tma-ra-ma, while seeking to stake a claim on hat.hayoga for the siddha tradition, also sought to avoid alienating any of that tradition’s disparate elements. The list of synonyms of sama-dhi which he gives at the beginning of the Hat.hapradı-pika-’s fourth upades´a is a roll call of the various goals of different siddha traditions’ methods of yoga,81 but nowhere does he mention sect-specific features such as metaphysics, mantras, or sets of cakras.82 Sva-tma-ra-ma’s appropriation of hat.hayoga for the siddha tradition is evinced by his naming as masters of hat.havidya- approximately thirty maha-siddhas,

S´a-ktism and hat.hayoga 119 . including Matsyendra, Goraks.a, Viru-pa-ks.a, Caurangı- and Carpat.i, whose 83 names also occur in lists of Na-th gurus. The texts that he used to compile the Hat.hapradı-pika-, however, come from a much broader range of traditions. Thus Sva-tma-ra-ma borrows approximately twenty verses from the Datta-treyayogas´a-stra, yet he makes no mention of Datta-treya anywhere in the text. The Datta-treyayogas´a-stra is the product of a Vais.n.ava tradition which coa- lesced into the yogi suborders of the Das´ana-mı- Sam . nyasıs, in particular the Giris and the Purı-s.84 Datta-treya is the tutelary deity of the Ju-na- Akha-r.a-, - - which today is the largest of the Sam . nyası akhar.as and whose members are predominantly Giris and Purı-s, and there is a long history of rivalry, or at least differentiation, between the Na-ths, with Goraks.a as their first guru, and - - 85 the Sam . nyasıs, with Dattatreya as their is.t.adevata. To this day the Nath Yogıs of northern India, despite displaying an otherwise broad inclusivity, will have little to do with Datta-treya.

Munis and siddhas Some of the Hat.hapradı-pika-’s teachings on the more gymnastic a-sanas are - 86 taken from the Vasis.t.hasam . hita. As an ancient r.s.i not closely associated with an ascetic order, there were no sectarian reasons to sideline Vasis.t.ha and he does get a brief mention in the Hat.hapradı-pika- when it is said that the a-sanas taught therein are those that were accepted by munis such as Vasis.t.ha and yogins such as Matsyendra.87 This distinction between munis or r.s.is and yogins or siddhas is found in a wide range of texts.88 In a somewhat confused analysis based in the main on the . Datta-treyayogas´a-stra and the Vivekama-rtan.d.a, the S´a-rngadharapaddhati says that hat.hayoga is of two sorts, one practised by Goraks.a and others, another by Ma-rkan.d.eya and others.89 The latter refers to the yoga taught to Ma-r. kan.d.eya by Datta-treya in the Ma-rkan.d.eyapura-n.a. Elsewhere in the S´a-rngad90 harapaddhati rajayoga is said to be of two varieties. The first, which was mastered by Datta-treya and other maha-tmans, corresponds to the bindudha-ran.a hat.hayoga of the Datta-treyayogas´a-stra; the second, whose practitioners are not identified, uses the breath to raise Kun.d.alinı- upwards through five cakras. The roughly contemporaneous Maithili Varn.aratna-kara gives a list of siddhas and a list of munis.91 Among the former are a number of Na-th gurus, including Goraks.a; among the latter are Ma-rkan.d.eya and Kapila. What this boils down to is a distinction between layayoga-practising siddhas such as Goraks.a and hat.hayoga-practising munis such as Kapila or Datta-treya, which manifests among today’s ascetics as a distinction between the S´a-kta Na-ths and the relatively more orthodox Das´ana-mı-s and Ra-ma-nandı-s. This distinction should not be applied too rigorously, however, because there are many anomalies, in particular lexical ones. Thus Kapila, though most commonly said to be a muni, is described as a siddha in texts as diverse as the Bhagavadgı-ta- and the Datta-treyayogas´a-stra itself.92

120

James Mallinson

The muni tradition of hat.hayoga Kapila’s ambiguous status is important for our understanding of the muni tradition which developed the techniques of early hat.hayoga. The first formulation of the practices distinguishing it from other methods of yoga is taught in the Datta-treyayogas´a-stra where it is said to be the doctrine of the school of Kapila and other siddhas; the practices are taught as an alternative to the way of the kavi, the eightfold yoga practised by Ya-jñavalkya.93 Bronkhorst has shown how Kapila was associated with the practice of ‘non-Vedic’ asceticism94 and the Datta-treyayogas´a-stra appears to confirm this when it contrasts Kapila’s yoga with that of the more orthodox Vedic r.s.i Ya-jñavalkya. The practices of ascetics beyond the Vedic pale are only mentioned in passing in early textual sources. In fact, very little of the actual practices of ascetics is taught in texts, Sanskrit or otherwise. This is not surprising: asceticism and scholarship are uncommon bedfellows.95 The only ascetic practice that is treated in any depth in Sanskrit texts is yoga, and this is because it is one of the few of their practices which is not restricted to ascetics alone. The 96 S´ivasam . hita, for example, is explicitly aimed at householders. The techniques of hat.hayoga are not taught in Sanskrit texts until the eleventh century or thereabouts. It is my contention that they did not appear ex nihilo, but that they developed from techniques practised by ascetics from at least the time of the Buddha. The Buddha himself is said to have tried both pressing his tongue to the back of his mouth, in a manner similar to that of the hat.hayogic khecarımudra-,97 and ukkut.ikappadha-na, a squatting posture which is perhaps a forerunner of the vı-ra-sana mentioned as part of ascetic practice in the Maha-bha-rata98 and may be related to hat.hayogic techniques such as maha-mudra-, maha-bandha, maha-vedha, mu-labandha99 and vajra-sana100 in which pressure is put on the perineum with the heel, in order to force upwards the breath or Kun.d.alinı-.101 Elsewhere in the Pali Canon these same practices are associated with ta-pasas and Ajı-vikas,102 who, together with other austerities, are also said to practise the ‘bat-penance’ (vagguli-vata), which is generally assumed to mean suspending oneself upside-down from a tree, thus inverting oneself in a fashion not dissimilar to the hat.hayogic viparı-takaran.-ı mudra-. In early Sanskrit sources too we find mentions of ascetics practising austerities suspended upside-down.103 All these practices are undertaken by ascetics who are celibate and it seems likely that some of their austerities were linked with the preservation of semen in the manner of similar techniques taught in the texts of early hat.hayoga as methods of bindudha-ran.a, and that these ancient and previously obscure ascetic practices came to light as the mudra-s of early hat.hayoga. Within these older sources we find no mention of forerunners of vajrolimudra-,104 the quintessential and crudest method of bindudha-ran.a, in which semen is resorbed through the urethra. Not all ascetics, whether those of mythology or scriptural prescription, were celibate105 and it may be that some of their historical counterparts used the technique of vajrolimudra- in order to

S´a-ktism and hat.hayoga

121

be able to have sex and preserve their bindu, to have their cake and eat it, as it were. It is this understanding of vajroli that has remained predominant in India, rather than the more infamous aim attributed to it in S´a-kta hat.hayogic texts such as the S´ivasam . hita, that of absorbing the commingled sexual fluids of both the male and female partners.106 Although our earliest clear-cut reference to vajroli is in the Datta-treyayogas´a-stra,107 a recent edition of a section of the Brahmaya-mala which teaches the asidha-ravrata, the ‘knife’s edge penance’,108 suggests a milieu in which vajroli might have been used. As taught in the Brahmaya-mala, the asidha-ravrata was a form of coitus reservatus quite distinct from S´a-kta sexual practices which employed sexual fluids in ritual. Its emphasis on continence would have been nicely complemented by vajrolimudra-, which could have been employed should the ascetic have fallen off the knife-edge. Among the sources cited thus far as evidence for the ancient muni tradition of bindu-oriented hat.hayoga, other than the Amr.tasiddhi there are no S´aiva works. Indeed, in S´aiva works that predate the Amr.tasiddhi we find no mention of bindudha-ran.a, the key principle of early hat.hayoga. But the S´aiva associations of some of early hat.hayoga’s practices are clear, even as taught in the Vais.n.ava Datta-treyayogas´a-stra.109 Most obviously, there are the names of mudra-s such as ja-landharabandha, ud.d.-ıya-nabandha, maha-mudra-, maha-vedha, vajroli, sahajoli, amaroli and khecarı-mudra-, all of which are redolent of S´aivism. Maha-mudra- is said in the Datta-treyayogas´a-stra to have been taught by Bhairava.110 And in vajrolimudra- there is perhaps a link between hat.hayoga and the practices of the earliest S´aiva ascetics. The asidha-ravrata mentioned above was probably practised by Pa-s´upatas and Atima-rga ascetics, who were separate from the Mantrama-rga traditions in which S´a-ktism developed.111 Corroborating the link between the practices of hat.hayoga and those of earlier tapas-practising ascetics is the fact that the ascetic traditions whom our textual sources show to have been most closely associated with the practice of hat.hayoga since its first textual codification, traditions which are today - - best represented by the Das´ana-mı- Sam . nyasıs and the Ramanandıs, are those most closely associated with the practice of tapas. To this day their ascetic practices include many of the austerities mentioned together with the forerunners of hat.hayogic techniques in our earliest sources, and some of the a-sanas taught in later manuals of hat.hayoga correspond exactly to ancient austerities such as standing on one leg or holding one or both arms in the air for long periods.112 The origins of many of the physical practices of hat.hayoga are thus to be found in tapas, in particular the bodily austerities undertaken by the ascetics most often associated with the practice of yoga in early sources, rather than in the more rarefied formulations of mental yoga taught in, say, the Yogasu-tra and its commentaries or Buddhist works.113 Indeed, formulations of yoga as yoga prior to the appearance of the hat.hayogic corpus show it to be a mental, meditative practice, its physicality limited to sitting in a suitable position and regulating the breath. It was onto the bindudha-ran.a-oriented hat.hayoga of this ascetic tradition that the Kun.d.alinı--oriented layayoga of the siddha tradition was grafted. The

122

James Mallinson

name hat.ha suggests the difficult asceticism with which its early techniques had been associated, but this did not sit well with all of the practitioners of the newly S´a-kta hat.hayoga, either laya-practising siddhas or householders unsuited to practising tapas.114 In the Yogabı-ja we see an attempt to obfuscate the primary meaning of hat.ha by resorting to an esoteric nirukti which is often repeated in later works and secondary literature: ha means ‘sun’, t.ha means ‘moon’ and their yoga or union is hat.hayoga.115 The rehabilitation of the word hat.ha in the face of yoga’s universalisation was ultimately unsuccessful, however, and the use of the word in the context of yoga fell out of favour in all but scholastic typologies of the subject and, in recent years, in secondary and popular literature on yoga. In vernacular literature hat.h continued to refer to extreme ascetic practices.116

Conclusion I shall conclude with some speculative remarks on the broader context of the S´a-kta co-option of hat.hayoga. First, my findings raise the question of why these physical yoga techniques started to be codified approximately 1,000 years ago, particularly if, as I contend, they were already ancient. This seems to me to be connected with the severing of yoga from its sectarian moorings mentioned earlier, which served to make yoga available to all. Much of the practice of hat.hayoga as taught in its texts appears to be for ascetics. Several of its techniques are distasteful or difficult and take a lot of time and effort to master: the yogin is often instructed to carry out his practice in an isolated hut. But at the same time most of the texts of hat.hayoga are explicitly anti-sectarian and universalist,117 and state or imply that anyone can practise yoga.118 Vajrolimudra- in particular – if only as an ideal119 – would have assisted in opening up the yoga practice of ascetics to all. The S´ivasam . hita says that through its practice even householders can attain liberation, without observing the restrictions taught in the texts of yoga.120 Written in simple Sanskrit and free from the abstruse metaphysics of the Yogasu-tra and its exegesis, or the esoterica of S´aiva yoga manuals, the texts of hat.hayoga are the first works on yoga that are accessible to all. This made its aims, liberation and siddhis, also accessible to all, without the need for priestly intermediaries, ritual paraphernalia or sectarian initiations. We see a similar democratisation of religion in the bhakti cults that started to develop during the same period and this may have been a corollary of the demise of S´aivism, at least as a grand, state religion. The democratisation of yoga was responsible for the production of its texts. Ascetics had learnt hat.hayoga through oral teachings for centuries, but when its teachings were opened up to householders texts were produced, perhaps as the result of patronage by these new practitioners, who would have enlisted pandits to codify the teachings of ascetic gurus. At first these yoga manuals were written in Sanskrit, which would have limited their audience, but

S´a-ktism and hat.hayoga 123 vernacular texts containing teachings on yoga soon appeared. Jña-ndev’s latethirteenth-century Jña-nes´varı- contains exquisite teachings in Marathi on the ascent of Kun.d.alinı-, and the Tamil Tirumantiram, which probably dates to a similar period,121 has extensive teachings on hat.hayogic techniques. Vernacular yoga manuals first appear in the north in the sixteenth century, with the - 122 Persian Bah.r al-H in the seventeenth century by the . ayat. This is followed . Braj Bha-s.a- Jña-nasamudra and Sarva-ngayogapradı-pika- of the Da-du-panthıscholar Sundar Da-s, and a variety of Persian translations of Sanskrit works, such as the Khula-s.ah al-Khula-s.ah of Devı- Da-s Ka-yastha.123 A noteworthy eighteenth-century vernacular manual of yoga is the Braj Bha-s.a- Jogpradı-pakaof the Ra-ma-nandı- Jayatara-ma.124 The universalism and antisectarianism taught in the texts of hat.hayoga, which facilitated the coming together of the S´a-kta and ascetic yogas, was mirrored by the ascetic milieu of the time, which saw a common ascetic identity develop, drawing on a wide range of sources, from S´a-kta to Sufi, with the result that, to this day, the yogins of the various North Indian ascetic orders are very similar in appearance and lifestyle. However, by the seventeenth century, out of this relatively homogeneous ascetic soup, various ascetic sects did solidify. The co-option of hat.hayoga by the fledgling Na-ths may have been part of this process of sect-formation, symbolising their abandonment of more licentious Kaula practices in favour of a celibate ascetic lifestyle. Thus the final verse of the Goraks.as´ataka is the following declaration:125 We drink the dripping liquid called bindu, ‘the drop’, not wine; we eat the rejection of the objects of the five senses, not meat; we do not embrace a sweetheart [but] the Sus.umna- na-d.-ı, her body curved like kus´a grass; if we have intercourse + … + it takes place in a mind dissolved in the void, not in a vagina. The S´a-kta appropriation of hat.hayoga – and yoga more broadly conceived – was entirely successful, to the extent that even the celibate bindu-yoga traditions were made more S´a-kta through incorporating Kun.d.alinı- into their practice. She features widely in the so-called ‘Yoga Upanis.ads’ compiled in the seventeenth and eighteenth centuries; several of the mudra-s taught in the - Jogpradı-paka- are said to raise Kun.d.alinı-; and a Das´ana-mı- Sam . nyası whom I met at Gangotri in 2006 told me that his mastery of vajrolimudra enabled him to resorb his semen should Kun.d.alinı- make him involuntarily ejaculate as she passed through his sva-dhis.t.ha-na cakra. Hat.hayoga’s S´a-kta veneer obscured the origins of its practices. In secondary literature it is widely assumed that the mudra-s of hat.hayoga arose in S´a-kta milieux. As I have shown, it is likely that they in fact developed amongst nonVedic ascetic traditions. To add to the textual evidence I have cited, there are two pieces of external, negative evidence that suggest that the bindudha-ran.a techniques unique to early hat.hayoga were not part of S´a-kta practice. First, khecarı-mudra-, vajrolimudra- and viparı-takaran.-ı are not found in Tibetan

124

James Mallinson

sources, implying that they were not part of the tantric traditions, in particular that of the siddhas, including ‘Na-ths’ such as Matsyendra and Goraks.a, which travelled to Tibet during the centuries prior to the fixing of the Tibetan Buddhist canon. Second, I have long puzzled over how, if khecarı-mudradeveloped in S´a-kta milieux, it is never compared with sexual intercourse in our textual sources, when there are such obvious parallels between the tongue entering the void above the palate and the penis entering a vagina; I can only surmise that khecarı-mudra- developed outside of a milieu in which sexual rituals were practised, such as the ancient non-Vedic ascetic tradition. The S´a-kta appropriation of hat.hayoga and its associated adoption by the Na-ths meant that yoga and yogis of all stripes came to be associated by some with the more malevolent practices of S´a-ktism,126 an association that persists to this day despite the oldest and most persistent hat.hayoga tradition being that of the generally benevolent munis, represented today by the Das´ana-mı- - Sam . nyasıs and Ramanandıs. It is the muni tradition that has been responsible for all the new texts and exegesis of hat.hayoga since the time of the Hat.hapradı-pika-. Meanwhile the S´a-kta Na-th Yogı-s have produced no further manuals of hat.hayoga and ceased to practise it.127 Those Na-ths whose sa-dhana- is anything other than an ascetic lifestyle and archetypal North Indian bhakti remain true to their roots and perform S´a-kta tantric ritual.

Notes 1 I would like to thank the organisers of the S´a-kta Traditions conference held in Oxford in September 2011 for inviting me to present the first draft of this paper and the participants at the conference for their helpful comments. I would also like to thank Jason Birch, Patton Burchett and Véronique Bouillier for their detailed comments on subsequent drafts. 2 Sanderson 2009. 3 It is at this time that yoga is first included among the dars´anas, in the c. twelfthcentury Sarvasiddha-ntasam . graha, which was composed by an anonymous Advaitaveda-ntin (Halbfass 1988: 352). Over the subsequent centuries the orthodox (in particular Veda-ntin) interest in the hat.ha techniques of yoga continued to grow, culminating in the composition in the seventeenth to eighteenth centuries of a group of ‘Yoga Upanis.ads’ which consist for the most part of passages from earlier hat.ha texts (Bouy 1994). 4 Examples of this are legion. To give just three, from different contexts: Bouy (1994: 6) who describes all the texts used to compose the ‘Yoga Upanis.ads’ as Na-th; Vaudeville (1957: viii–xii) who says that Kabı-r borrowed various yogic elements from the tradition of the Na-ths; and van der Veer (1989: 89) who says that the Ra-ma-nandı-s have been ‘deeply influenced’ by the Na-ths. 5 For an overview of the history of the Na-th sam . pradaya see Mallinson 2011c. 6 See e.g. Bouillier 1997: 147; Vila-sna-th 2010. 7 In the late medieval period ascetic Na-ths claiming allegiance to Gorakhna-th subscribed to nirgun.-ı and, subsequently, S´aiva tenets (see Mallinson 2011c). Gorakhna-th is portrayed in later legends as a celibate S´aiva ascetic who reformed the licentious ways of his Kaula tantric guru Matsyendrana-th (Mallinson 2007: 186 n129). He is also said to have overcome the goddess in magical contests (Bouillier 2008: 44). With Gorakhna-th coming to be seen by most Na-ths

S´a-ktism and hat.hayoga

8 9

10

11 12 13

14 15

125

as the head of the Na-th sam . pradaya, the nirgun.ı S´aivism associated with him ´ came to overshadow the order’s Sakta roots. Nevertheless, as noted above Na-th liturgy retains S´a-kta features and in some hagiographies Gorakhna-th is said to be a worshipper of the goddess (e.g. in the current story of the origins of the Gorakh D . ibbı temple at Jwalamukhi). On the Buddhist texts which mention hat.hayoga, see Birch 2011. In this definition hat.hayoga is a gloss of the mu-la’s ‘hat.ha’. This is the first of many instances of practices referred to in primary sources simply as hat.ha being said to constitute hat.hayoga in later systematisations, exegesis and secondary literature. This has resulted in the Hat.hapradı-pika- often being referred to in secondary literature as the Hat.hayogapradı-pika- when only a tiny fraction of its manuscripts call it thus. We see in this the scholastic tendency towards categorisation, to identify different types of yoga and to pit them against one another. But it should be noted that the situation is more nuanced in the texts themselves: even those which do categorise different varieties do not see them as altogether different types of yoga but as contrasting (yet sometimes complementary) methods of achieving yoga. Thus compounds such as hat.hayoga and layayoga should be understood as tatpurus.as with an instrumental case relationship: ‘yoga by means of hat.ha’ etc. The one exception to this is ra-jayoga, which, as the aim of all methods of yoga, should be understood either as a karmadha-raya or as a tatpurus.a of genitive relationship: ‘the royal yoga’. A small number of texts do identify their yoga as ra-jayoga but this is to emphasise the superiority of their method above all others rather than give a definition of ra-jayoga (see e.g. Amanaska 2.3–4; Vasudeva 2011). This earliest definition of hat.hayoga is repeated verbatim in other Buddhist exegetical works (see Birch 2011). The verse from the Ka-lacakratantra on which Pun.d.arı-ka is commentating (4.119cd) says that if the siddhi desired by mantrapractitioners does not arise as a result of purification, yogic withdrawal and so forth, then they should accomplish it by forcefully (hat.hena) restraining bindu in the vajra in the lotus: - - sam . s´uddhipratyaharadibhir vai yadi bhavati na sa mantrin.am is.t.asiddhir na-da-bhya-sa-d dhat.hena-bjagakulis´aman.au sa-dhayed bindurodha-t | Pun.d.arı-ka glosses hat.hena with hat.hayogena and defines it thus (Vimalaprabhavol. 2, p. 212): - - ida-nı-m . hat.hayoga ucyate | iha yada pratyaharadibhir bimbe dr.s.t.e saty aks.ar- - - aks.an.am . notpadyate ayantritapran.ataya tada nadabhyasad dhat.hena pran.am . - hayitva- prajña-bjagatakulis´amanau bodhicittabindunirodha-d madhyama-ya-m va . . aks.araks.an.am . sadhayen nih.spandeneti hat.hayogah. || Bouy 1994: 82–85. For the details of the texts used and verses borrowed, see my forthcoming monograph Yoga and Yogis: the Texts, Techniques and Practitioners of Traditional Hat.hayoga. Hat.hapradı-pika- 1.3: bhra-ntya- bahumatadhva-nte ra-jayogam aja-nata-m | - - hat.hapradı-pika-m . dhatte svatmaramah. kr.pakarah. || Dattatreyayogas´astra 9, 26; Yogabıja 143, 149; Amaraughaprabodha 3, 4, 5, 7, 10, 24, 29, 73 and S´ivasam . hita 5.12, 5.23, 5.222, 5.254. For precursors of ja-landharabandha, see e.g. Sarvajña-nottara Yogapa-da 11ab as - edited by Vasudeva (2004: 398 n.77): a-sanam . ruciram . baddhva urdhvakayam adhah.s´irah. which teaches that in all the four seated a-sanas that have just been named the head is to be held down; cf. Vaikha-nasasma-rtasu-tra 8.8 in which having the face turned downwards is included in a list of ascetic practices. See

126

16

17

18

19 20 21 22

23 24

James Mallinson Mallinson 2007: 17–24 on forerunners of the hat.hayogic khecarı-mudra-. On precursors of mu-labandha, see ibid., p. 15. These techniques are often prescribed, without being named, as adjuncts to other mudra-s and a-sanas in hat.hayogic texts. See e.g. Amr.tasiddhi 11.4, Datta-treyayogas´a-stra 133, Jña-nes´varı- 6.207–208, Matsyendrasam . hita 3.28–30, Vivekamartan.d.a 8, Khecarıvidya 3.25. A measure of the validity of using the Hat.hapradı-pika- to establish this corpus in this way – as well as of the Hat.hapradı-pika-’s inclusivity – is that we know of no text which predates it and teaches early hat.ha techniques but was not used in its compilation. (As noted above, some of the hat.ha practices, in particular the three bandhas, are taught in earlier works, but as one among a variety of attitudes to be assumed in meditational, ascetic or yogic postures: they are not treated individually, nor do they take the names by which they are known in hat.hayogic works.) In terms of chronology, the corpus splits neatly into two halves. The Amr.tasiddhi, Datta-treyayogas´a-stra and Vivekama-rtan.d.a are named or cited in fourteenthcentury works (Bu ston Rin chen grub’s 1322 catalogue of canonical works in the case . of the Amr.tasiddhi (see Schaeffer 2002), the 1363 S´a-rngadharapaddhati in the case of the Dattatreyayogas´astra (over thirty verses cited) and the Khecarı-vidya(1.14) in the case of the Vivekama-rtan.d.a). Verses from the central core of the Goraks.as´ataka are found in the Yogabı-ja, which also borrows from the Datta-treyayogas´a-stra. This, together with the coherence of the Goraks.as´ataka’s teachings, suggests that the Goraks.as´ataka was the source. The four texts that constitute the second half of the corpus are all to some extent derivative of those in the first half, indicating their later date. Further details of their dependence on the earlier works are given in the notes to the descriptions of the texts below. Schaeffer, on the strength of the Amr.tasiddhi teaching hat.hayoga and the text cycle which it is part of being traced back to Viru-pa-ks.a/Viru-pa-ks.ana-tha, identifies the text as Na-th (2002: 518–519). In the text itself, however, there is nothing to suggest that it was composed in a Na-th, or even Pas´cima-mna-ya, milieu. As well as there having been a siddha called Viru-pa-ks.a/Viru-pa-, the name Viru-pa-ks.a has been used from at least the twelfth century to refer to the form of S´iva that presides over the Vijayanagar region (Verghese 1995: 18) and the Amr.tasiddhi could also be a product of that S´aiva tradition, in particular the Ka-lamukha cult . that flourished there prior to the Sangama kings’ patronage of Vais.n.avism. Amr.tasiddhi 11–13. In the Amr.tasiddhi (as in later hat.hayogic texts) the archaic triad of sun, moon and fire is reduced to the pair sun and moon, the sun and fire being identified as one (Amr.tasiddhi 5.2). See Amr.tasiddhi viveka 7. The Datta-treyayogas´a-stra makes no mention of the usual six cakras of yogic physiology (although at verse 131 the sahasrakamala, which is sometimes added to that schema as a seventh cakra, is said to be the source of amr.ta) but does refer to Kun.d.alinı- in passing, saying that in the paricaya stage of yoga she and the breath are made to move by fire (107). The Amr.tasiddhi mentions neither Kun.d.alinı- nor cakras. (At 7.8–13 it does say, however, that a solar rajas, the feminine equivalent of bindu, is found in all beings ‘wrapped in the goddess element’ (devı-tattvasama-vr.tah.), and that its union with the lunar bindu is yoga; see footnote 100 for the text of this passage. Cf. Vivekama-rtan.d.a 54–57; Yogabı-ja 89.) On the confusion between the Goraks.as´ataka and the Vivekama-rtan.d.a caused by the latter also coming to be known, among other names, as the Goraks.as´ataka, see Mallinson 2011b: 262–263. The text says that the three bandhas are used to restrain the breath in order to stimulate Kun.d.alinı-, who is to be stimulated in order to control the breath, suggesting that the Kun.d.alinı- paradigm has been imposed onto a yoga that worked on breath alone.

S´a-ktism and hat.hayoga 25

26 27 28 29

30 31 32 33

34

127

Later redactions of the Goraks.as´ataka’s verses, such as that in the Hat.hapradı-pika-, through textual corruption and, we must assume, either a rupture in the transmission of the practice or a dissociation of the textual and practical traditions, say that the cloth is to be wrapped around the waist (see Mallinson 2011b). Vivekama-rtan.d.a 31–39. In the Vivekama-rtan.d.a Kun.d.alinı- is not said to interact with the cakras in any way. Vivekama-rtan.d.a 40–63 teaches mu-labandha, ud.d.-ıya-nabandha, ja-landharabandha, khecarı-mudra- and maha-mudra-. Vivekama-rtan.d.a 113–115. The Yogabı-ja includes a number of verses from the Datta-treyayogas´a-stra and Goraks.as´ataka and is likely to postdate them (Yogabı-ja 116–117, 119c-121b, 122c-123b = Datta-treyayogas´a-stra 144–145, 141c-143b, 138; Yogabı-ja 103, 105c-106b, 109ab, 110c-115d, 121c-122b = Goraks.as´ataka 35c-36b, 38c-39b, 45ab, 47a-52b, 61c-62b). The following verses are found in both the Yogabı-ja and Hat.hapradı-pika- but in no other texts, so it is likely that the Yogabı-ja is their source: Yogabı-ja 92, 94, 118c119b, 123c-124b, 125 (= Hat.hapradı-pika- 3.109, 3.112, 3.54, 2.46, 3.111). Yogabı-ja 92a-98b, 108–112, 123c-132d. Yogabı-ja 123, 140. Khecarı-vidya- 1.14. The Amaraughaprabodha shares verses with the Amr.tasiddhi, Amanaska and S´ivasam . hita (Amaraughaprabodha 20, 29, 32cd, 37ab, 38, 39cd, 40ab, 45 = Amr.tasiddhi 16.1c-2b, 11.3, 11.9cd, 11.3cd, 14.6, 13.5cd, 13.7cd, 19.2 (many other verses in the Amaraughaprabodha are derivative of verses in the Amr.tasiddhi, particularly those describing the four classes of aspirant – compare Amaraughaprabodha 18–24 with Amr.tasiddhi vivekas 15–18); Amaraughaprabodha 8 = Amanaska 2.32; Amaraughaprabodha 3, 45 = S´ivasam . hita 5.12, 3.31 (Amr.tasiddhi 19.2 is likely to be the original source of the latter)). It also includes a quotation attributed to a S´rı-sam . put.a. It is thus to some extent a compilation, but it contains nineteen verses found in the Hat.hapradı-pika- and not elsewhere (Amaraughaprabodha 9, 30–31, 32ab, 34 = Hat.hapradı-pika- 4.14, 3.10–11b, 3.12ab, 3.19; Amaraughaprabodha 35ab, 35cd, 36, 40c-41d, 42ab, 42cd, 43ab, 43c-44b = Hat.hapradı-pika- 3.22ab, 3.23cd, 3.24, 3.26c-27d, 3.29ab, 3.30ab, 3.29cd, 3.30c-f; Amaraughaprabodha 44cd, 46–53 = Hat.hapradı-pika- 1.61ab, 4.70–77), so in the absence of an alternative source for those verses it seems likely that the Amaraughaprabodha was compiled before the Hat.hapradı-pika-. Amaraughaprabodha 1: - - - om . namo ‘stv. adinathaya mınanathaya vai namah. | namas´ caura-ngina-tha-ya siddhabuddha-ya dhı-mate ||

35 Amaraughaprabodha 29–33, 55. 36 The S´ivasam . hita shares sixteen verses with the Amr.tasiddhi (S´ivasam . hita 2.1b, 2.1cd, 2.2, 2.3, 2.4ab = Amr.tasiddhi 1.15b, 1.16ab, 1.17ab+1.16cd, 1.17c-1.18b, 1.19ab; S´ivasam . hita 2.6c-9d, 2.11–12, 3.31, 4.27ab, 4.27dc, 4.28ab, 4.34cb, 5.13, 5.17c-20b = Amr.tasiddhi 3.1–4, 4.3–4, 19.2, 11.3cd, 11.4bc, 11.5ab, 11.7cd, 15.1, 16.1–3; S´ivasam . hita 4.31, 4.38, 5.14–15 ≈ Amr.tasiddhi 11.6, 12.6, 15.3; several other verses in the S´ivasam . hita’s descriptions of mahamudra, mahabandha and mahavedha are derivative of verses in the Amr.tasiddhi); four verses with the Datta-treyayogas´a-stra (S´ivasam . hita 3.102–105 = Dattatreyayogas´astra 35–38; the ´Sivasam . hita also paraphrases passages from the Dattatreyayogas´astra: compare S´ivasam . hita 5.71, 3.42–48, 3.60–61, 3.62–63, 3.72–75, 4.88ab etc. with Dattatreyayogas´astra 22c-23b, 72c-82b, 98c-100b, 89c-90b, 112–121, 157cd); one verse with the S´a-rada-tilaka (S´ivasam . hita 2.16 = S´aradatilaka 25.32); and one with the ´ Vivekamartan.d.a (Sivasam . hita 3.81 = Vivekamartan.d.a 119). Ten of the fifteen - verses which the S´ivasam . hita shares with the Hat.hapradıpika are not to be found in

128

37

38

39 40 41 42 43

44 45 46 47 48 49 50 51

52 53

James Mallinson other texts (S´ivasam . hita 3.5a-6b, 3.43, 3.108–109, 4.21, 4.23ab, 4.47ab, 4.88ab, 5.47, 5.222a-d, 5.254 = Hat.hapradı-pika- 4.22a-23b, 2.14, 1.18–19, 3.2, 3.6ab, 3.24cd, 3.87ab, 1.43, 2.76, 1.11). See the invocations of the goddess Tripurabhairavı- at 5.84 and 5.240, and teachings on the associated tripartite S´rı-vidya- mantrara-ja/mu-lavidya-. The mantrara-ja consists of va-gbhava + ka-maru-pa + s´akti, whose condensed forms are - ,and sauh; it is taught at 5.80–85 and 5.234–235. Its three components aim . , klım . . also combine to make the mu-lavidya- in the Va-makes´varı-mata (1.93–101). The yoginı-s said in the S´ivasam . hita to be situated at the cakras correspond to those in Bha-skarara-ya’s Saubha-gyabha-skara commentary on Lalita-sahasrana-mastotra 148–158. The assertions that the yogi can make himself irresistible to women (5.104), that he can become a second Ka-ma (3.93), and instructions to meditate on Ka-ma (4.3), have parallels with the ‘love magic’ taught in the Va-makes´varımata and the associated earlier Nitya- tantras, a S´a-kta textual corpus from which the cult of Tripura- developed and which was independent of the Vidya-pı-t.ha literature (Sanderson 2009: 47; cf. Golovkova 2010: 9, 41–53). The S´ivasam . hita’s vedantic teachings (see in particular the first pat.ala) confirm the connection with . the cult of the southern S´ankara-ca-ryas of Kanchi and Shringeri. ´Sivasam . hita 5.250a-251b: trim . s´allaks.ais tatha japtair brahmavis.n.usamo bhavet | rudratvam . s.as.t.ibhir laks.aih. s´aktitattvam as´ıtibhih. || kot.yaikaya mahayogı lıyate parame pade | Vamakes´varı-mata 1.164 and 2.13, and 3.27 in Jayaratha’s Va-makes´varı-matavivaran.a ad loc. S´ivasam . hita 2.23b. S´ivasam . hita 2.24a. - 4.23–24. S´ivasam hita . The Dattatreyayogas´a-stra’s teachings on vajroli (vv. 150c-159b) give only bindudha-ran.a as its aim, not the absorption of commingled sexual fluids. The Hat.hapradı-pika-’s teachings on vajroli (vv. 3.82–89) describe it as a technique of bindudha-ran.a for the male yogin (a verse found in some witnesses after 3.89, but not in the edited text, enjoins the absorption of both rajas and bindu). At the end of the section teaching amaroli and sahajoli, two practices supplementary to vajroli, it is said (vv.~3.95–98) that the true yoginı- absorbs male bindu mixed with her own rajas, thereby achieving total success (sarvasiddhim), knowing the past and present and becoming a khecarı-. S´ivasam . hita 4.78–87. Kun.d.alinı is not mentioned in the context of vajroli even in the S´ivasam . hita, but in the c. 17th-century Hat.haratna-valı- the practice is explicitly said to awaken her (2.82). Datta-treyayogas´a-stra 10a-11b. The different degrees of aspirant are also taught in the Amr.tasiddhi (vivekas 15–18) but the Amr.tasiddhi does not give a corresponding typology of yogas. Datta-treyayogas´a-stra 12–14. S´ivasam . hita 5.17–20. - 5.21–23. S´ivasam hita . The teachings on the mantra of Khecarı- that were central to the earliest form of the Khecarı-vidya- (before the teachings on the hat.hayogic khecarı-mudra- were incorporated into it) become a relic in its later redactions. The coded definition of the mantra was obscure to commentators and the verses in which it is taught were unintelligible to scribes, so were corrupted in most, if not all, recensions of the text. See Khecarı-vidya- 1.30c-38b and Mallinson 2007: 199 n.225. This is most striking in the Hat.hapradı-pika-, on which see below. Vivekama-rtan.d.a 28–29, Gheran.d.asam . hita 5.86–87; cf. S´aradatilaka 25.50–52.

S´a-ktism and hat.hayoga

129

54 Datta-treyayogas´a-stra 87c-88d, Vivekama-rtan.d.a 70, S´ivasam . hita 3.54, 3.71; cf. adhyaya 6 of the Yogayajñavalkya. 55 Dissolution into the elements is also mentioned at Yogasu-tra 1.19: bhavapratyayo videhaprakr.tilaya-na-m. S´aiva works predating the hat.hayogic corpus often mention laya, particularly in the context of meditations that lead the yogin up through a hierarchy of elemental tattvas, and they teach some of the techniques associated with the layayoga taught in hat.hayogic works, but they do not teach laya to be a particular variety of yoga (personal communication from Alexis Sanderson, June 2011). 56 Datta-treyayogas´a-stra vv. 15–26: layayogas´ cittalaya-t sam . ketais tu prajayate | -adina-thena samketa- astakotih prakı-rtita-h ||15|| . .. . . . sa-m . kr.tir uvaca bhagavan a-dina-thah. sah. kim . rupah. kah. sa ucyatam ||16|| datta-treya uva-ca maha-devasya na-ma-ni a-dina-thas´ ca bhairavah. | - s´abares´as´ ca devo ’yam . lılaya vicaran prabhuh. ||17|| s´rıkan.t.haparvate gaurya saha pramathana-yakaih. | mithah. s´rı-parvate caiva kadalı-vanagocare ||18|| giriku-t.e citraku-t.e supa-dapayute girau | kr.payaikaikasam . ketam . s´am . karah. praha tatra tan ||19|| tani sarvan.i vaktum . hi na s´aknomi tu vistarat | ka-ni cit kathayis.ya-mi sahaja-bhya-savat sukham ||20|| tis.t.han gacchan svapan bhuñjan dhya-yec chu-nyam aharnis´am | - ayam eko hi sam . keta adinathena bhas.itah. ||21|| nasagradr.s.t.imatren.a aparah. parikırtitah. | s´irah.pa-s´ca-tyabha-gasya dhya-nam . mr.tyum . jayet param ||22|| bhru-madhyadr.s.t.ima-tren.a parah. sam . keta ucyate | lala-t.e bhru-tale yas´ ca uttamah. sah. prakı-rtitah. ||23|| . savyadaks.in.apa-dasya angus.t.he layam uttamam | uttanas´avavad bhumau s´ayanam . coktam uttamam ||24|| s´ithilo nirjane des´e kurya-c cet siddhim a-pnuya-t | . - evam . ca bahusam . ketan kathayam asa s´ankarah. ||25|| - nyair yas´ cittasya layo bhavet | ´ sam ketair bahubhis ca . sa eva layayogah. sya-d hat.hayogam . tatah. s´r.n.u ||26|| B = Dattatreyayogas´astra, ed. Brahmamitra Avasthı-, Sva-mı- Kes´ava-nanda Yoga Sam . sthana 1982  J1 = Man Sim . h Pustak Prakas´ 1936  W1 = Wai Prajña Pat.has´ala 6/4–399  M = Mysore Government Oriental Manuscripts Library 4369  W2 = Wai Prajña- Pa-t.has´a-la- 6163  U = Yogatattvopanis.ad 15a °laya-t] J1W1MNW2; °layah. BNU15b sam . ketais] BNW1W2; sam . ketas J1MN, - kotis´ah. U  tu praja-yate] parikı-rtitah. U15d sam . keta as.t.a°] BJ1W1W2; sam . ketas sardha ° M 16a °nathah. sah.] B; °nathasya J1W1, nathas´ ca M, nathas´ ca W2 16b kim . rupah. - pah sa ucyate J1, kimru-pam kah sa ucyata-m W1, ko ‘yam kah. sa ucyata-m] B; ko yam ru . . . . . . . - W2 17b °na-thas´ ca bhairavah] MW2; ° purus.a ucyate M, ko yam . rupah. sa ucyatam . . nathadikany api BW1, °nathasya bhairavah. J117c s´abares´as´ ca devo ’yam . ] MW2; s´ives´varas´ ca devo ’sau B, s´ivarı-s´as´ ca devo yam . J1, s´ives´varas´ ca devo yam . W1 17d vicaran] J1W1MW2; vyacarat B18a s´rı-kan.t.ha°] BJ1W1 ; s´rı-kan.t.hah. MW2 18b saha pramatha°] BJ1W1W2; sahasrapramatha° M (unmetrical)  °na-yakaih.] M; °na-yaka-n BJ1W1W2 18c mithah. s´rı-parvate caiva] M; hima-ks.aparvate caiva BJ1W1, hima-layas tu parvate W2 19b supa-dapayute girau] BW1; sapa-daniyatai gurau J1, sa yathaniyayoginah. M, sapa-danilaye girau W2 19d s´am . karah. praha tatra tan] BJ1W1W2; ´sabarebhyo hi dattava-n M 20b na s´aknomi tu vistara-t] BW1; na s´aknomı-ti vistara-t J1, na-nyas´ s´akto hi tam . vina M, nan saknomi tu vistarat W2 21b dhyayec] W1M; dhyayan

130

57 58

59 60 61 62 63 64 65

James Mallinson BJ1W2, dhya-yen U  chu-nyam ahar nis´am] nis.kalam -ıs´varam U22c °pa-s´ca-tya°] M; ° pas´ca-c ca BJ1W1, pas´ca-t W2 (unmetrical)22d mr.tyum . ] mr.tyu W2  jayet param] BW1; jaye param J1W2, °jayo varah. M 23c lala-t.e bhru-tale yas´] BJ1; lala-t.e bhu-tale yas´ W1W2, lı-la-vibhu-tilepas´ M 24c utta-na°] BJ1M; utta-nah. W1W2  °s´avavad] BW1M; ° s´ikha/ra-\d J1, s´aravad W2 24d s´ayanam . coktam uttamam] s´apasanacottamottamam . J1 (unmetrical), s´ayanam . cottamottamam M 25a s´ithilo] BW1W2; s´ithila J1, ham . seva M 26d hat.ha°] M; karma° B, dharma° J1W1, dhat.a° W2 The sam . ketas of the siddhas are also mentioned frequently in the Hindi verses attributed to Goraks.a. See e.g. the Gorakhba-n.-ı, sa-khı-s 106, 115, 217. Jña-nes´varı- 6.291: - - - pim . d.em . pim . d.aca grasu | to ha nathasam . ketica d.am . su | pari daunu gela uddes´u | mahavis.n.u || 291 || See Kiehnle 2005: 31. Kiehnle (ibid.) understands na-tha° here to refer to the Na-tha sam . pradaya of Yogıs but both internal and external evidence suggest otherwise: the poet is pointing out that the teaching derives from S´iva even though it is being told by Vis.n.u, and the verse was written several centuries before na-tha was used to refer to members of an order of yogis (Mallinson 2011c). Yogabı-ja 136ab: sa eva na-thasam . ketah. siddhasam . ketalaks.an.ah. | Jña-nes´varı- 6.272, 6.281. S´ivasam . hita 5.44 and 5.47. On na-da in S´aivism see Vasudeva 2004: 273–280. On the Amanaska, see Birch 2012. See Amanaska 2.22, 2.57. Amanaska 2.10–11 (2.10 = Candra-valokana 1, Hat.hapradı-pika- 4.36): antarlaks.yam . bahirdr.s.t.ir nimes.onmes.avarjita | ´ es.a hi sambhavı mudra sarvatantres.u gopita ||2.10|| a-dis´aktir uma- cais.a- matto labdhavatı- pura- | adhuna- janmasamska-ra-t tvam eko labdhava-n asi ||2.11||

66 Amanaska 1.82:

67 68 69 70 71 72

73

74

caturvim . s´atibhir vars.air layasthasya nirantaram . | s´aktitattvasya siddhih. sya-c chaktitattvamayo bhavet | . S´a-rngadharapaddhati 4350–4363. Amaraughaprabodha 27. Khecarı-vidya- 2.39. Khecarı-vidya- 2.113ab. I.e. the Datta-treyayogas´a-stra, Yogabı-ja, Amaraughaprabodha and S´ivasam . hita. Mantrayoga does perhaps get a lexical nod in the name by which Svatmarama classifies the cleansing practices, the verses teaching which are among the few that I am yet to find in earlier works and which may thus be the work of Sva-tma-ra-ma himself. Despite teaching seven such practices, he calls them the s.at. karma-n.i, ‘six acts’, which is the name also given to a group of six magical methods of overpowering one’s enemies that can be activated by means of mantras taught in texts of the S´aiva Mantrama-rga (see Bühnemann 2000; the same phrase can also refer to the six duties of a brahmin). For the most part the laya techniques are incorporated without alteration. An exception is the sam . keta of lying like a corpse taught in the Dattatreyayogas´astra (24cd), which in the Hat.hapradıpika (1.32) becomes an asana, namely s´avasana, ‘the corpse pose’ widely taught in modern yoga. Laya is often mentioned in the Hat.hapradı-pika-, particularly in the context of practices that elsewhere come under the name of layayoga, and at one place

S´a-ktism and hat.hayoga

75

76

77 78 79 80 81

131

(4.103) hat.ha and laya are differentiated, but the subject matter of the text is, as evinced by its title, explicitly said to be hat.ha, with the implication that all the practices taught therein are hat.ha practices. I know of one exception to the principle of the raising of Kun.d.alinı- not involving physical techniques in pre-hat.ha S´aiva works: in his commentary ad Netratantra 7.30, Ks.emara-ja says that clenching and unclenching the anus makes Kun.d.alinıpoint upwards. Hat.hapradı-pika- 3.5: tasma-t sarvaprayatnena prabodhayitum -ıs´varı-m | brahmadva-ramukhe supta-m . mudrabhyasam . samacaret || Cf. Vivekama-rtan.d.a 56ab, where the homology of the masculine bindu, s´iva and indu is felicitous, but that of the triple-gendered rajas, s´akti and su-rya seems forced. Hat.hapradı-pika- 3.31–53, 4.43–4.55. See footnote 35. Kiss 2009: 97. Hat.hapradı-pika- 4.2–4.4: atheda-nı-m . pravaks.yami samadhikramam uttamam | mr.tyughnam . ca sukhopayam . brahmanandakaram . param ||4.2|| rajayogah. samadhis´ ca unmanı ca manonmanı- | - - amaratvam . layas tattvam . s´unyas´unyam . param . padam ||4.3|| amanaskam . tathadvaitam . niralambam . nirañjanam | jı-vanmuktis´ ca sahaja- turya- cety ekava-caka-h. ||4.4||

82 The six-cakra system was yet to achieve hegemony in yogic discourse. See e.g.the . S´a-rngadharapaddhati which mentions groupings of both six and nine cakras (4298–4300, 4351–4360). 83 Hat.hapradı-pika- 1.5–8. 84 Textual, ethnographic and iconographic sources show that the S´aiva orientation of these suborders did not take hold until the seventeenth century as part of the formalisation of the Das´ana-mı- order (see my forthcoming monograph Yoga and Yogis: The Texts, Techniques and Practitioners of Traditional Hat.hayoga). 85 See for example the early eighteenth-century Bachittar Na-t.ak of Guru Gobind Singh, caupa--ı 23–24 (translated in Kohli 2005: 126): Then I created Datt, who also started his own path, His followers have long nails in their hands, And matted hair on their heads, They do not understand the ways of the Lord. 23 Then I created Gorakh, Who made great kings his disciples, His disciples wear rings in their ears, And do not know the love of the Lord. 24 -- The second largest akha-r.a- of the Sam . nyasıs, the Mahanirvan.ı, has as its tutelary deity Kapila, who, as mentioned above, is credited in the Datta-treyayogas´a-stra as being the originator of some of the practices of hat.hayoga. 86 The Vasis.t.hasam . hita’s verses on these asanas are also found (sometimes with changes of metre) in earlier Pañcaratrika texts, texts which are thus the first to teach non-seated a-sanas as techniques of yoga. See for example those on mayu-ra-sana at Vima-na-rcanakalpa pat.ala 96; Pa-dmasam . hita yogapada 1.21–22, Ahirbudhnya-sam . hita 31.36–37; Vasis.t.hasam . hita yogakan.d.a 1.76–77; Yogayajñavalkya 3.15a-3.17b and Hat.hapradıpika 1.30.

132

James Mallinson

87 Hat.hapradı-pika- 1.18. Elsewhere in the Hat.hapradı-pika- we find more references to these two traditions: after the description of bhadra-sana, which, like mayu-ra-sana, is taught in a variety of Pa-ñcara-trika works that predate the Hat.hapradıpika- (Vima-na-rcanakalpa pat.ala 96; Pa-dmasam . hita yogapada 1.16c-1.17d; Ahirbudhnya-sam . hita 31.41; Vasis.t.hasam . hita 1.79; Yogayajñavalkya 3.12), it is said that siddhayogins call it goraks.a-sana (Hat.hapradı-pika- 1.54). 88 As well as the examples given here, see also Vya-sa’s Bha-s.ya on Yogasu-tra 3.51, Vivekama-rtan.d.a 74 and Jña-nes´varı- 6.154–156. . 89 S´a-rngadharapaddhati 4372. . 90 S´a-rngadharapaddhati 4364–4371. 91 Varn.aratna-kara pp. 57–58. 92 Bhagavadgı-ta- 10.26, Datta-treyayogas´a-stra 131. 93 Datta-treyayogas´a-stra 29, 130–131. 94 Bronkhorst 2007: 61–68. To the examples given by Bronkhorst may be added verse 20.153 of the Br.hatkatha-s´lokasam . graha which suggests the antinomian nature of the soteriological practices of Kapila: of Can.d.asim . ha’s city Budhasvamin writes, ‘There the vices that usually terrify those who want to be liberated from the wheel of rebirth are prescribed by Kapila and others in treatises on - liberation’ (yena dos.ena sam . sarat paritrasyanti moks.avah. |sa tasmin moks.as´astres.u s´ru-yate kapila-dibhih. ||.) 95 Studying books is scorned by the Na-ths (Yogı- 1924: 11) and this attitude is found among other traditional yoga-practising ascetics of today. At the 2010 Haridwar Kumbh Mela I was informed, over the din of competing loudspeaker systems blaring out pravacans, ‘lectures’, from ascetic-scholars, that within living memory it was normal for any Ra-ma-nandı- ascetic who dared to expound the s´a-stras to the public to be thrown off his a-san, ‘seat’, and beaten up by his peers. 96 See S´ivasam . hita 5.258–260, the final verses of the text. 97 See Mallinson 2007: 17–19. 98 Maha-bha-rata 12.292.8, 13.7.13, 13.13.10, 13.13.54. Cf. Vaikha-nasasma-rtasu-tra 8.8 and Manusmr.ti 11.110. In its earliest definition, in the Tattvavais´a-radı- of Va-caspati Mis´ra (ad Yogasu-tra 2.46), vı-ra-sana has the yogin squatting on one foot with the other on its opposite thigh. This posture was demonstrated to me by the Tapasvı- Na-th Yogı- Ma-nav Na-th Jı- at Jwalamukhi on 11 November 2011 and said to be part of his practice of tapas, not yoga. 99 See e.g. Datta-treyayogas´a-stra 132–136, 143–145. 100 See e.g. Goraks.as´ataka 15. This variety of vajra-sana is variously known as svastika-sana, siddha-sana and mu-labandha-sana (see Goodall 2004: 349–350 n.730). 101 The Buddha is said to try this and a variety of other ascetic techniques in the Maha-sı-hana-dasutta (Majjhima Nika-ya 1:78). . 102 Ajı-vikas are said to practise ukkut.ikappadha-na in the Nangut.t.haja-taka (Ja-taka 1:493) and the Kassapası hanadasutta (Dı ghanikaya 1:166), ta-pasas in the Setaketu and Udda-laka Ja-takas (Ja-taka 3:235, 4.299). Ukkut.ikappadha-na is also . mentioned at Dhammapada 141 and Anguttara Nika-ya 1.296 and 2.206. 103 See e.g.Vaikhanasasmartasutra 8.8 (Caland 1929: 191), which mentions those ‘who hang with their head downwards’ in a list of celibate hermits who practise a variety of austerities, and the following references in the Maha-bha-rata: 1.13.10–13 and 1.41.1–3, in which the u-rdhvaretas muni Jaratka-ru finds his ancestors performing penance by hanging over a big hole (cf. Agastya at 3.94.11–14, whose ancestors say they are performing the penance in order to get offspring); 1.26.2, in which Garud.a sees some Va-lakhilyas hanging upside-down; 3.185.4–5, in which Manu practises extreme tapas (as well as hanging upside down, he is also said to stand on one leg and hold up his arms); 12.126.18, in which the muni Tanu meditates upside down; 13.7.8–13, in which Bhı-s.ma describes a variety of ascetic practices to Yudhis.t.hira, amongst which is hanging upside down. The

S´a-ktism and hat.hayoga

104 105

106

107

108 109 110 111 112

113

133

practice of hanging upside-down was relatively common until recently but appears from my ethnographic inquiries to have died out. See e.g. the c.1830 illustration of tapkar a-san from an illustrated manuscript of Jayatara-ma’s Jogpradı-paka- reproduced at Bühnemann 2007: 51, and the photograph of an ‘Urdhamukhi Sadhu’ at Oman 1984: 46. Hanging upside down or performing a headstand has been a Sufi practice for at least a thousand years (Sanaullah 2010: 643). McEvilley’s claim (1981: 66) that vajroli is referred to in the Br.hada-ran.yakopanis.ad is not borne out by the text. Datta-treya himself is said in the seventeenth adhya-ya of the Ma-rkan.d.eyapura-n.a to have indulged in wine, women and song in an unsuccessful attempt to dissuade some young sages from seeking his tutelage. The final verse of the same adhya-ya (17.25) says that he practises tapas and also that he is ‘a master of yoga meditated upon by yogins seeking liberation’ (yogı-s´varas´ cintyama-no yogibhir . muktika-nks.ibhih.). The dharmas´a-stras allow va-naprasthas to take wives with them to the forest but they are to remain chaste and undertake more extreme austerities - than the solitary sam . nyasıs (Kane 1974: 928–9). . E.g. S´ankaradigvijaya 9.90, where Kr.s.n.a is said to use vajroli in order to be able . to sport with the gopı-s and Sarva-ngayogapradı-pika- 3.14, where S´iva is said to use it while making love to Gaurı- so that Ka-ma cannot reach him. Cf. Siddhasiddha-ntapaddhati 2.13; see also footnote 114. Vajroli may also be obliquely referred to at Amr.tasiddhi 7.8–13: sa bindur dvividho jñeyah. paurus.o vanita-bhavah. | bı-jam . ca paurus.am . proktam . rajas´ ca strısamudbhavam . ||7.8|| - | anayor bahyayogena sr.s.t.isam . harau jayate nr.n.am . yada- tv abhyantare yogas tada- yogo hi bhan.yate ||7.9|| ka-maru-po vased binduh. ku-t.a-ga-ra+n.ya+ kot.are | pu-rn.agiri sada- spars´a-d vrajanti madhyama-pathe ||7.10|| yonimadhye maha-ks.etre java-bindu-rasannibham . | rajo vasati jantu-na-m . devıtatvasamavr.tah. ||7.11|| bindus´ candramayo jñeyo rajah. su-ryamayas tatha- | - - anayoh. sam . gam . ah. sadhyah. kut.agare ’tidurghat.e ||7.12|| es.a tattva paro dharma es.a yogah. paro matah. | es.a muktiprado ma-rgah. es.a guhyatamah. parah. ||7.13|| - 8a dvividho] em.; vividho J1 9b sr.s.t.isam . harau] em.; sr.s.t.isam . hara J1 10b kut.aga-ra°] em.; ku-t.a-dha-ra J1 10d vrajanti] em.; va-jam ti J1 13b yogah ] em.; yoga J1 . . 13d parah.] em.; param . J1 Hatley 2011. The Datta-treyayogas´a-stra names S´aiva ascetics among the possible practitioners of its yoga teachings when it mentions ‘he who says namah. s´iva-ya’ (namah.s´iva-yava-cı-) at verse 44. Datta-treyayogas´a-stra 132. Hatley 2011. These austerities include going naked, sitting in the summer sun surrounded by fire, remaining immersed in cold water at night in winter, and spending long periods standing up (sometimes on just one leg) or holding up one or both arms. For examples see the references in notes 97 and 98, as well as the following: Ma-rkan.d.eyapura-n.a 31.24; Katha-saritsa-gara 6.4.11; Ku-rmapura-n.a 2.27.29–30; Matysapura-n.a 35.17, 171.1 etc; Va-yupura-n.a (Reva-kan.d.a) 54.50; Bha-gavatapura-n.a . 7.3.2 and Lingapura-n.a 1.69.76. Tapas is one of the five niyamas, ‘observances’, prescribed at Yogasu-tra 2.32. In his Bha-s.ya thereon, Vya-sa defines tapas as the endurance of opposites (dvandvasahanam); ‘opposites’ include hunger and thirst, heat and cold, and stha-na-sana, ‘standing up and sitting down’, which may refer to the physical postures

134

114 115

116

117

118

119

James Mallinson maintained for long periods by ascetics. Other parts of hat.ha practice also correspond to the tapas of old. Pra-n.a-ya-ma, in concert with its expiatory and purificatory capabilities (on which see Vasudeva 388–389), is said in dharmas´a-stric literature to be the highest form of tapas (Manu 2.83, 6.70; Vis.n.u 55.17; Vas´is.t.ha - - 25.5). Matsyendrasam . hita 4.2 says that pran.ayama is both yoga and tapas; Amanaska 2.42 says that the various techniques of pra-n.a-ya-ma are unpleasant and difficult (duh.kha-tmakair durjayaih). In some modern manifestations of yoga the physical practices of hat.hayoga are identified with tapas. See Smith 2009 and Sarbacker 2009: 174–175. Yogabı-ja 148c-149b, which is cited at Hat.hatattvakaumudı- 55.29; cf. Hat.haratna-valı- 1.22. Of the various possible referents of sun and moon, the most com. monly given in this context are the pingala- and id.a- na-d.-ıs (i.e. the channels carrying pran.a which run from the base of the central column to the right and left nostrils respectively); see e.g. Vivekama-rtan.d.a 32. Other possibilities include the pra-n.a and apa-na breaths (Amr.tasiddhi 6.11–13), S´akti and S´iva, as menstrual fluid and semen (Vivekama-rtan.d.a 74–76), or the tip of the tongue and the forehead (Khecarı-vidya- 2.72–74). For a detailed discussion of this understanding of hat.ha, see Birch 2011: 532–534. The Yoga Upanis.ads largely consist of verses from hat.hayogic works (Bouy 1994) yet only the Yogatattvopanis.ad mentions hat.ha (in verses taken from the Datta-treyayogas´a-stra: Yogatattvopanis.ad 19, 24 = Datta-treyayogas´a-stra 9c-10b, 26). The one Sanskrit text produced in a Na-th milieu after the formation of the Na-th sam . pradaya, namely the circa 1700 Siddhasiddhantapaddhati, makes no mention of ‘hat.ha’ yoga. In Hindi works, what might be understood as hat.hayoga is usually referred to simply as yog/jog. Other than in the doxographies of Sundar Da-s and the Bhaktama-l of his contemporary and fellow Da-du-panthı- Ra-ghav Da-s, the word hat.h (the Hindi form of hat.ha) is used only once in the context of yoga in the corpus of medieval Hindi literature, in the Pra-n. Sam . kalı of Caur. anginath (v. 256), in which hat.h jog is said to be the union of the sun and moon. Elsewhere in the medieval Hindi corpus, hat.h in the context of religious practice refers to extreme austerities and is usually scorned (see e.g. Gorakhba-n.-ı sa-khı- 31, Guru Granth Sa-hib 905.4 and 1305.6 (in the latter hat.h is associated with hanging upside down), and the verses attributed to Carpat.na-th found in a manuscript dated 1711 CE and translated in Singh 1937: 68–71). Jayatara-ma’s 1737 Braj Bhasha reworking of the long recension of the Hat.hapradı-pika- is called Jogpradı-paka- and, despite the many references to hat.ha in its source text, does not use the word hat.ha/hat.h in its 964 verses, except when listing the Hat.hapradı-pikaamong its sources (v. 956). Hindi-speaking yogis in India today do occasionally use the compound hat.h-yog, perhaps influenced by Sanskrit texts. In their understanding hat.hayoga encompasses difficult and uncomfortable practices, including non-physical ones such as mauna, a vow of silence. I have heard such ascetics translate hat.hayoga into English as the conveniently homophonic ‘hard yoga’. E.g. Datta-treyayogas´a-stra 41a-42b: bra-hman.ah. s´raman.o va-pi bauddho va-py a-rhato ’thava- | ka-pa-liko va- ca-rva-kah. s´raddhaya- sahitah. sudhı-h. || yoga-bhya-sarato nityam . sarvasiddhim avapnuyat | Women are said to practise vajroli at Dattatreyayogas´a-stra 155 and Hat.hapradı-pika3.84, 3.95–98. Cf. Hat.haratna-valı- 2.80–117. Amr.tasiddhi 19.6 says that both the avadhu-ta and gr.hastha can practise yoga. The S´ivasam . hita is aimed squarely at householders (but it is less open to women: unlike in the Datta-treyayogas´a-stra and Hat.hapradı-pika- women are not mentioned as practitioners in its description of vajroli). Householder wrestlers in Kota, Rajasthan, for whom the refinement and preservation of bindu is an important part of their practice, speak highly of vajroli

S´a-ktism and hat.hayoga

135

but do not practise it (personal communication from Norbert Peabody, 11 June 2010). 120 S´ivasam . hita 4.79: svecchaya- vartama-no ’pi yogoktaniyamair vina- | mukto bhaved gr.hastho ’pi vajrolyabhya-sayogatah. || Cf. Datta-treyayogas´a-stra 152: svecchaya- vartama-no ’pi yogoktaniyamair vina- | - vajrolim . yo vijanati sa yogı siddhibhajanah. ||152|| 121 See Goodall 1998: xxxvii n.85 and Goodall 2004: xxix. 122 On the Bah.r al-Haya-t and related texts, see Ernst 2003 and Sakaki 2005. 123 I am grateful to Dr Kazuyo Sakaki for providing me with a comprehensive list of Persian works on yoga. 124 I do not include here the Gorakhba-n.-ı, the esoteric pads and sakhı-s ascribed to Goraks.a but probably written in the sixteenth century, whose obscurantist esotericism is aimed at initiates and distinguishes them from the more straightforward vernacular manuals listed here. 125 Goraks.as´ataka 101 (cf. Hat.hapradı-pika- 3.46–48.): ya- bindva-khyeti dha-ra- vigalitamanasa- ta-m . pibamo na madyam . pañcanam indriyan.am vis.ayanirasanam carvaya-mo na ma-m sam | . . . . - - ta-m susumna-m a-linga-mo na ka-nta-m . kus´akut.ilatanum . nad.ikam . . . . s´unye citte pralıne + pravilaya vivas´an + maithunam . cen na yonau ||101|| 126 See Mallinson 2012. 127 I know of two exceptions to prove this rule. First, Bouillier (2008: 128) reports the failed attempts of Sva-mı- S´iv Na-th Jı-, the mahant of an a-s´ram in Orissa, to interest any of the Na-ths’ main jama-t (itinerant monastery) in learning hat.hayoga. Second, on 11 November 2011 ‘Yogı- Ba-ba-’ Anu-p Na-th Jı- demonstrated a sequence of complex hat.hayogic a-sanas to me at the Na-th monastery at Jwalamukhi. When I asked him where he had learnt it from, he said that it had come to him automatically when he was a boy. 128 This work is reproduced at Callewaert and de Beeck 1991: 489–510. I have used the verse numbering of the latter.

References Primary sources . Anguttaranika-ya, parts I and II, ed. R. Morris. London: Pali Text Society. 1885 and 1888. Amanaska, ed. Jason Birch. Submitted as part of a D. Phil. thesis, University of Oxford. 2012. Amaraughaprabodha, of Goraks.ana-tha, ed. K. Mallik in Mallik 1954. Amr.tasiddhi of Viru-pa-ks.ana-tha. Maharaja Mansingh Pustak Prakash, Jodhpur, Acc. no. 1242. Ahirbudhnya-sam . hita, ed. M.D. Ramanujacharya under the supervision of F. Otto Schrader; revised by V. Krishnamacharya. Madras: Adyar Library and Research Centre. 1966. Katha-saritsa-gara of Somadevabhat.t.a, ed. Pandit Durga-prasa-da and Ka-s´-ına-tha . Pa-n.d.uranga Paraba. Bombay: Nirnaya Sa-gara Press. 1889. Ku-rmapura-n.a, ed. A.S. Gupta. Varanasi: All-India Kashiraj Trust. 1967.

136

James Mallinson

Khecarı-vidya-, ed. James Mallinson. The Khecarı-vidya- of Adina-tha. A critical edition and annotated translation of an early text of hat.hayoga. London: Routledge. 2007. (S´rı-) Guru Granth Sa-hib, with complete index prepared by Winand M. Callewaert. 2 parts. Delhi: Motilal Banarsidass. 1996. Goraks.as´ataka. Government Oriental Manuscripts Library, Madras, MS no. R 7874. Goraks.as´ataka, ed. F. Nowotny: Das Goraks.as´ataka, Cologne, 1976 (Dokumente der Geistesgeschichte). [This text is referred to as the Vivekama-rtan.d.a in this article; for the reasons why see Mallinson, James. 2011b, ‘The Original Goraks.as´ataka’, pp. 257–272 in D.G. White, ed., Yoga in Practice. Princeton, NJ: Princeton University Press.] Gorakhba-n.-ı, ed. P.D. Bad.athva-l. Praya-g: Hindı- Sa-hity Sammelan. 1960.128 Gheran.d.asam . hita, ed. and tr. J. Mallinson. New York: YogaVidya.com. 2004. Candra-valokana. Government Oriental Manuscripts Library, Madras, MS no. D 4345. Ja-taka, ed. V. Faussell. 4 vols. London: Trübner & Co. 1877–1887. Jña-nasamudra, see Sundar-grantha-valı-. Jogpradı-paka- of Jayatara-ma, ed. M.L. Gharote. Jodhpur: Rajasthan Oriental Research Institute. 1999. (S´rı-)-Jña-nes´varı- of Jña-nadeva, ed. G.S. Nan.adı-kar. 5 vols. Mumbai: Praka-s´ Gopa-l Nan.adı-kar. 2001. Tirumantiram of Tirumu-lar. See Natarajan, B. 1989. Holy Hymns (a translation of the Tirumantiram of Tirumu-lar). Madras: ITES Publications. Datta-treyayogas´a-stra, unpublished critical edition by James Mallinson. Dı-ghanika-ya, vol. 1, ed. T.W. Rhys Davids and J. Estlin Carpenter. London: Pali Text Society. 1890. . Dhammapada, ed. Su-riyagod.a Sumangala Thera. London: Pali Text Society. 1914. Padmasam . hita, ed. S. Padmanabhan and R.N. Sampath. Madras: Pancaratra Parisodhana Parisad. 1974. . - Pra-n. Sam . kalı of Cauranginath. See Gorakhban.ı. Br.hatkathas´lokasam . graha of Budhasvamin, ed. and tr. James Mallinson (as The Emperor of the Sorcerers). 2 vols. New York: Clay Sanskrit Library/JJC Foundation. 2005. Bhagavadgı-ta-, ed. J.A.B. van Buitenen in The Bhagavadgı-ta- in the Maha-bha-rata: Text and Translation. Chicago, IL: University of Chicago Press. 1981. Bha-gavatapura-n.am, ed. Vasudeva S´arman. Bombay: Nirnaya Sagar. 1905. Majjhimanika-ya, vol. 1, ed. V. Trenckner. London: Pali Text Society. 1888. Matsyapura-n.a, ed. The Pandits of the Ananda-s´rama. Ananda-s´rama Sanskrit Series no. 54. 1907. Matsyendrasam . hita, See Kiss, Csaba. 2009. ‘Matsyendranatha’s Compendium (Mat- ). A Critical Edition and Annotated Translation of Matsyensyendrasam hita . drasam . hita 1–13 and 55 with analysis’. Unpublished D.Phil. thesis submitted to Oxford University. Manusmrti: with the Sanskrit Commentary Manvartha-Muktavali of Kulluka Bhatta, ed. J.L. Shastri. Delhi: Motilal Banarsidass. 1983. Maha-bha-rata, ed. V. Sukthankar, with the cooperation of S.K. Belvalkar, A.B. Gajendragadkar, V. Kane, R.D. Karmarkar, P.L. Vaidya, S. Winternitz, R. Zimmerman and other scholars and illustrated by Shrimant Balasaheb Pant Pratinidhi (since 1943 ed. S.K. Belvalkar). 19 vols. Poona: Bhandarkar Oriental Research Institute, 1927–1959. Ma-rkan.d.eyapura-n.a, ed. K.M. Banerjea. Calcutta: Bishop’s College Press. 1862.

S´a-ktism and hat.hayoga

137

Lalita-sahasrana-mastotra with the commentary Saubha-gyabha-skara of Bha-skarara-ya, revised by Va-sudev Laks.man.a S´a-strı- Pan.s´-ıkar (4th edn). Bombay: Nirn.aya Sa-gara Press. 1935. . Lingapura-n.a. Gurumandal Series no. XV. Calcutta. 1960. Yogatattvopanis.ad. See S´a-strı-, A.M., ed. 1920. The Yoga Upanis.ads. Madras: Adyar Library.. Yogabı-ja, ed. Ra-m La-l S´rı-va-stav. Gorakhpur: S´rı- Gorakhna-th Mandir. 1982. Yogaya-jñavalkya, ed. K.S. S´a-strı-. Trivandrum Sanskrit Series no. 134. Trivandrum. 1938. Varn.aratna-kara of Jyotirı-s´varakavis´ekhara-ca-rya, ed. S.K. Chatterji and B. Misra. Calcutta: Royal Asiatic Society of Bengal. 1940. Vasis.t.hasam . hita (Yogakan.d.a). Revised edition, ed. Swami Digambarji, PitambarJha, Gyan Shankar Sahay (first edition); Swami Maheshananda, B.R. Sharma, G.S. Sahay, R.K. Bodhe (revised edition). Lona-vala-: Kaivalyadha-m S´rı-manma-dhav Yogamandir Samiti. 2005. Va-yupura-n.a (Reva-khan.d.a), ed. (as S´rı-skandamaha-pura-n.am) Ks.emra-j S´rı-kr.s.n.ada-s. . Bombay: Venkates´vara Steam Press. 1910. Vimalaprabha of Kalkin S´rı- Pun.d.arı-ka on S´rı-laghuka-lacakratantra by S´rı- Mañjus´rıyas´as, chief editor Samdhong Rinpoche, ed. Vrajavallabh Dwivedi et al. Rare Buddhist Texts Series 13 (3 vols). Sarnath: Central Institute of Higher Tibetan Studies. 1994. Vima-na-rcana-kalpa, ed. Raghuna-thacakrava-rtin and Setu Ma-dhava-ca-rya. Madras: Venkateshwar Press. 1926. Vivekama-rtan.d.a of Goraks.adeva. Oriental Institute of Baroda Library. Acc. no. 4110. Vis.n.usmr.ti, ed. V. Krishnamacharya.Adyar Library Series 93. Madras: The Adyar Library and Research Centre. 1964. . S´ankaradigvijaya of Ma-dhava, ed. Ma-dhava-nand Bha-ratı-. Lakhanau-: Mum . s´ı Navalakis´ora. 1865. . S´a-rngadharapaddhati, ed. Peter Peterson. Bombay: Government Central Book Depot. 1888. S´a-rada-tilakam of Laks.man.ades´ikendra. Pat.ala 25, as edited in pp. 343–354 of Bühnemann 2001. S´ivasam . hita, ed. and tr. J. Mallinson. New York: YogaVidya.com. . Sarvangayogapradı-pika-, see Sundar-grantha-valı-, ed. Purohit Hari N. S´arma-. Calcutta. 1936. Siddhasiddha-ntapaddhati of Goraks.ana-tha, ed. M.L. Gharote and G.K. Pai. Lonavla: Lonavla Yoga Institute. 2005. Sundar-grantha-valı-, ed. Purohit Hari N. S´arma-. Calcutta. 1936. Hat.hapradı-pika- of Sva-tma-ra-ma. Rajasthan Oriental Research Institute, Jodhpur. MS no. 6756. Hat.hapradı-pika- of Sva-tma-ra-ma, ed. Sva-mı- Digambarjı- and Dr Pı-tambar Jha-. Lonavla: Kaivalyadha-m S.M.Y.M. Samiti. 1970. Hat.haratna-valı-, ed. M.L. Gharote,P. Devnath and V.K. Jha. Lonavla: Lonavla Yoga Institute. 2002.

138

James Mallinson

Secondary literature Birch, Jason. 2011. ‘The Meaning of Hat.ha in Early Hat.hayoga’. Journal of the American Oriental Society, 131(4): 527–552 . Birch, Jason. 2012. ‘The Amanaska’. D.Phil. thesis submitted to the University of Oxford. Bouillier, Véronique. 1997. Ascètes et Rois. Un Monastère de Kanphata Yogis au Népal. Paris: CNRS. Bouillier, Véronique. 2008. Itinérance et Vie Monastique: Les Ascètes Na-th Yogı-s en Inde Contemporaine. Paris: Maison des sciences de l’homme. Bouy, Christian. 1994. Les Na-tha-Yogin et les Upanis.ads. Paris: Diffusion de Boccard. Bronkhorst, Johannes. 2007. Greater Magadha: Studies in the Culture of Early India. Vol. 19 of Handbook of Oriental Studies, Section Two, India. Leiden and Boston, MA: Brill. Bühnemann, Gudrun. 2000. “The Six Rites of Magic”, In Tantra in Practice, ed. David Gordon White, pp. 447–462. Princeton, NJ: Princeton University Press. Bühnemann, Gudrun. 2001. The Iconography of Hindu Tantric Deities volume II. The Pantheons of the Prapañcasa-ra and the S´a-rada-tilaka. Groningen: Egbert Forsten. Bühnemann, Gudrun. 2007. Eighty-four Asanas in Yoga. A Survey of Traditions (with Illustrations). Delhi: D.K. Printworld. Caland, W. 1929. Vaikha-nasasma-rtasu-tram: The Domestic Rules and Sacred Laws of the Vaikha-nasa School Belonging to the Black Yajurveda. Calcutta: Asiatic Society. Callewaert, W.M. and op de Beeck, B. 1991. Devotional Hindı- Literature: A Critical Edition of the Pañc Va-nı- or Five Works of Da-du-, Kabı-r, Na-mdev, Ra-ida-s, Harda-s with the Hindı- Songs of Gorakhna-th and Sundarda-s, and a Complete WordIndex. 2 vols. New Delhi: South Asia Institute, New Delhi Branch, Heidelberg University. Ernst, Carl W. 2003. ‘The Islamization of Yoga in the Amrtakunda Translations’ Journal of the Royal Asiatic Society, Series 3, 13(2): 1–23. Ernst, Carl W. 2009. “Being Careful with the Goddess: Yoginis in Persian and Arabic Texts”. In Performing Ecstasy: The Poetics and Politics of Religion in India, ed. Pallabi Chakrovorty and Scott Kugle, pp. 189–203. New Delhi: Manohar. Golovkova, Anna. 2010. ‘The Cult of the Goddess Tripurasundarı- in the Va-makes´varımata’. Unpublished thesis submitted for the degree of Master of Studies at Oxford University. Goodall, Dominic. 1998. Bhat.t.a Ra-makan.t.ha’s Commentary on the Kiran.atantra. Volume 1: Chapters 1–6. Critical Edition and Annotated Translation. Pondicherry: Publications de l’Institut français d’Indologie no. 86.1. Goodall, Dominic. 2004. The Para-khyatantra. A Scripture of the S´aiva Siddha-nta. A critical edition and annotated translation. Pondicherry: Publications de l’Institut français d’Indologie no. 98. Halbfass, Wilhelm. 1988. India and Europe. Albany, NY: State University of New York Press. Hatley, Shaman. 2011. ‘Erotic Asceticism: The Knife’s Edge Observance of Brahmaya-mala xl (asidha-ravratapat.ala), with a Critical Edition and Translation’. Tantric Studies, 3, forthcoming. Kane, P.V. 1974. A History of Dharmas´a-stra. Vol. 2, Part 2. Poona: Bhandarkar Oriental Research Institute.

S´a-ktism and hat.hayoga

139

Kiehnle, Catharina. 2005. “The Secret of the Na-ths: The Ascent of Kun.d.alinıaccording to Jña-nes´varı- 6. 151–328”. In Bulletin des Études Indiennes 22–23, pp. 447–494. Kiss, Csaba. 2009. ‘Matsyendrana-tha’s Compendium (Matsyendrasam . hita). A Critical - 1–13 and 55 with Edition and Annotated Translation of Matsyendrasam hita . Analysis’. Unpublished D. Phil. thesis submitted to Oxford University. Kohli, Surendar Singh. 2005. (The) Dasam Granth: Bachittar Na-t.ak of Guru Gobind Singh. New Delhi: Munshiram Manoharlal. McEvilley, Thomas. 1981. ‘An Archaeology of Yoga’. Anthropology and Aesthetics no. 1, pp. 44–77. Mallik, Kalyani, ed. 1954. The Siddha Siddha-nta Paddhati and Other Works of Nath Yogis, Poona: Poona Oriental Book House. Mallinson, James. 2007. The Khecarı-vidya- of Adina-tha. A Critical Edition and Annotated Translation of an Early Text of Hat.hayoga. London: Routledge. Mallinson, James. 2011a. “Hat.ha Yoga”. In Brill Encyclopedia of Hinduism, ed. K.A. Jacobsen, H. Basu, A. Malinar and V. Narayanan, pp. 770–781. Leiden: Brill. Mallinson, James. 2011b. “The Original Goraks.as´ataka”. In Yoga in Practice, ed. D.G. White, pp. 257–272. Princeton, NJ: Princeton University Press. Mallinson, James. 2011c. “Na-th Sam . pradaya”. In Brill Encyclopedia of Hinduism, ed. K.A. Jacobsen, H. Basu, A. Malinar and V. Narayanan, pp. 407–428. Leiden: Brill. Mallinson, James. 2012. ‘The Yogı-s’ Latest Trick’. Review article in Tantric Studies (University of Hamburg). Also pp. 165–180 in Journal of the Royal Asiatic Society, 24(1) (2014). Natarajan, B. 1989. Holy Hymns (a translation of the Tirumantiram of Tirumu-lar). Madras: ITES Publications. Oman, John Campbell. 1984 [1903]. The Mystics, Ascetics and Saints of India. New Delhi: Cosmo Publications. Sakaki, Kazuyo. 2005. ‘Yogico-tantric Traditions in the H . awd al-H . ayat’, Journal of the Japanese Association for South Asian Studies, 7: 135–156. Sanaullah, M. 2010. “Elements of Yoga in Sufism”. In History of Yoga, ed. S.P. Singh, pp. 629–652. Vol. XVI, part 2 in the History of Science, Philosophy and Culture in Indian Civilization Series. New Delhi: Project of History of Indian Science, Philosophy and Culture. Sanderson, Alexis. 1988. “S´aivism and the Tantric Traditions”. In The World’s Religions, ed. S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp. 660–704. London: Routledge. Sanderson, Alexis. 2009. “The S´aiva Age”. In Genesis and Development of Tantrism, ed. Shingo Einoo, pp. 41–349. Tokyo: Institute of Oriental Culture, University of Tokyo. Sarbacker, Stuart Ray. 2009. “The Numinous and Cessative in Modern Yoga”. In Yoga in the Modern World: Contemporary Perspectives, ed. Mark Singleton and Jean Byrne, pp. 161–183. London: Routledge. S´a-strı-, A.M., ed. 1920. The Yoga Upanis.ads. Madras: Adyar Library. Schaeffer, Kurtis R. 2002. ‘The Attainment of Immortality: from Na-thas in India to Buddhists in Tibet’ Journal of Indian Philosophy, 30(6): 515–533. Singh, Mohan. 1937. Gorakhna-th and Medieval Mysticism. Lahore: Dr Mohan Singh.

140

James Mallinson

Smith, Benjamin Richard. 2009. “With Heat Even Iron Will Bend”. In Yoga in the Modern World: Contemporary Perspectives, ed. Mark Singleton and Jean Byrne, pp. 140–160. London: Routledge. van der Veer, Peter. 1989. Gods on Earth. The Management of Religious Experience and Identity in a North Indian Pilgrimage Centre. Delhi: Oxford University Press. Vasudeva, Somadeva. 2004. The Yoga of the Ma-linı-vijayottaratantra. Publications de l’Institut français d’Indologie no. 97. Pondichéry: Institut français d’Indologie. Vasudeva, Somadeva. 2011. ‘Ham . samit.t.hu: “Patañjalayoga is Nonsense”’. Journal of Indian Philosophy 39(2): 123–145. Vaudeville, Charlotte. 1957. Kabir Granthavali (Doha). Publications de l’Institut français d’Indologie no. 12. Pondichéry: Institut français d’Indologie. Verghese, Anila. 1995. Religious Traditions at Vijayanagara. New Delhi: Manohar. Vila-sna-th, Yogı-. 2010. S´rı- Na-th Rahasya. 2 vols. Haridva-r: Akhil Bha-ratavars.-ıya Avadhu-t Bhes. Ba-rah Pam . th Yogı Mahasabha. Yogı, Candranath. 1924. Yogisampradayavis.kr.ti. Ahmeda-ba-d: S´ivana-th Yogı-.

7

S´a-ktism, polity and society in medieval Malabar Rich Freeman

Introduction This paper is an attempt to contextualize a series of interesting historical and anthropological problems around textual remnants of S´a-kta cults in Kerala, focusing on the particular region of northern Kerala, or Malabar, proper. The goal is to provide what can only be a brief overview of a series of interconnections between Sanskrit and vernacular texts, and forums and practices of worship which survive in the region today, or which have a still-legible historical purchase among present communities of the region. After some three decades of fieldwork and study in the region, I have collected a good deal of material, both ethnographic and textual, which I hope to draw together into a larger monographic reflection on issues and questions introduced here in a necessarily summary fashion. Many of the major historical questions I confront here have been stimulated and informed by the work of Alexis Sanderson. From my own interdisciplinary perspective as a textually trained anthropologist of the Kerala region I perceive two major, connected projects in Sanderson’s oeuvre that I have found invaluable. The first has been his painstaking years of searching out, characterizing and ordering much of the largely unpublished medieval S´aiva material (including its allied S´a-kta strains), and his analysis of this literature in terms of actual practices and doctrines which can be usefully situated in regard to their potential historical, social, and political significance (e.g. 1988, 2007a) The second phase of his endeavors, as I see it, builds on this first, partly by way of comparative work on parallel or derived developments in Vais.n.ava, Buddhist, and Jain circles, but more directly through a wideranging engagement with the inscriptional record, and serves to formulate general historical theses concerning the success and reach of what he calls the ‘S´aiva Age’ (e.g. 2009). This brings his work as a rigorous text-critical Indologist into convergence with more conventional historians of South India, particularly those concerned with the intersection of Indian polity and religious institutions (e.g. Stein 1994; Heitzman 1997; Champakalakshmi 2011), but, as he also suggests, and as I hope to demonstrate here, with anthropology. What has been most exciting for me is that many of the matters I have puzzled over

142

Rich Freeman

and come to tentative findings on during my years of anthropological and literary research in the region of northern Kerala, have been refocused and often confirmed in unexpected ways by Sanderson’s way of characterizing an issue, or by his more specific textual findings. In what follows, I will proceed by providing a sketch of some key texts and socio-historical problems in the S´a-ktism of Kerala, while trying to connect these with larger trends in Indian Tantric and Hindu developments. In many cases I will draw schematically on earlier published work and my dissertation, but in a way that reframes this as a kind of overview of what I hope to consolidate into a monograph on the Tantric tradition of Kerala.

S´a-ktic and Tantric Kerala The Puranic syntheses in Kerala temples and performative cults have tended to elide the various influences of more discriminate lines of S´a-kta cults (e.g. the Trika, Kubjika-, and Krama, which we know were textually represented in Kerala from their manuscript evidence), into an amalgam which few have tried to probe, beyond the vague invocations of ‘Mother Goddess’ or ‘Bhagavati’ worship, as being generically S´a-kta. It turns out, however, that there are far more interesting anthropological and historical implications for the whole S´a-kta/Bhairava milieu than were suspected, once we begin to engage actual Sanskrit and vernacular texts, and work to situate them in their regional context, on the one hand, and place these latter in terms of the larger dynamics that Sanderson has charted out in trans-regional scope, on the other. I will begin by reviewing a few of the principal texts and social structural correlations I find relevant that pertain to all of Kerala, including Malabar, and then narrow my focus to more particular Malabar material, which I will also try to fit back into the larger patterns. As with the attribution that many Kerala temples are nebulously ‘S´a-kta’, one of the striking descriptive usages of wider application is that Kerala Hindu worship is avowedly ‘Tantric’. The expositors and managers of this tradition, the modern Nampu-tiri Brahmans, however, are quick to point out that this is merged in a ‘Vedic’ cult, meaning that, aside from drawing on Vedic as well as Tantric mantras, any association with Kaula practices of blood sacrifice, non-vegetarian or alcoholic offerings, and ritualized sexual practices or imagery is entirely absent (Unni 2006: 8). The Tantric attribution carries a particular social valency, however, in that many of these prominent Brahmans are/were also, in local usage, Tantris,1 meaning that they exclusively control the hereditary rights and knowledge to perform the installation, renovations, annual festivals, and purification of all the major temples of Kerala which were divided up hereditarily under their jurisdiction. They furthermore appointed, and still ritually authorize and oversee, all priests who are engaged in the mere daily rites of worship, who are known in Kerala as S´a-nti(-ka-rar), and are generally drawn from socially and ritually subordinate Brahman families. I believe this distinction was historically significant in both social and ritual terms to which I will return.

Medieval Malabar: S´a-ktism, polity, society 143 Whatever the social constitution of Nampu-tiri Brahmans historically (and I suspect there was a far greater degree of historical diversity in their origins and community than is usually acknowledged), they have presented themselves as the single, more or less homogeneous and uniquely authoritative Brahmanical tradition of Kerala since the later medieval period. This regional ritual consolidation was reflected textually in the fifteenth-century handbook of temple installation and rituals, the Tantrasamuccaya, which is the standard, authoritative manual regulating temple ritual from one end of the state to the other. Its importance was demonstrated and authority shored up through two layers of Sanskrit commentary, within the next generation and apparently from within the same family (Ramaswamy Sastri et al. 1945: iii–iv; Narayana Pillai 1951: 3–4), followed by a Malayalam commentary in the eighteenth century, which was abstracted into a free-standing and complete handbook and translation. It is the latter, the Kuḻikka-t.t.u Pacca, which was updated and published in modern times to circulate throughout Kerala as the contemporary standard which closely adheres to the Tantrasamuccaya.

Textual interventions The Tantrasamuccaya attempted, much as in the pañca-yatana of the Sma-rta tradition (Bühnemann 1988: 49–51), to reduce the pantheon of the earlier . Kerala temple cult to a mere seven gods: Vis.n.u, S´iva, S´ankarana-ra-yan.a, Durga, Subrahman.ya, Gan.apati, and S´astr.. Ritually, the work looks very much like an Agamic handbook of the Tamil country, except for its pointed melding of Vais.n.ava and S´aiva worship, by fronting Vis.n.u as the paradigmatic deity in a clearly S´aivite ritual frame, and by installing the constitutionally . composite S´aiva-Vais.n.ava deities of S´ankarana-ra-yan.a and S´a-str.. That this work represents an attempt at a drastic reduction of the ritual life that preceded it seems apparent both from the textual production that preceded and followed it. The most prominent work in circulation before it was the compendious and encyclopedic Tantric digest, the Is´a-nas´ivagurudevapaddhati. This latter work is evidently composite, with older parts in a dominantly S´aiva Saiddha-ntika stream that probably go back to the twelfth century, yet which nevertheless incorporate significant Vais.n.ava and S´a-kta cults of worship. While the prevalence and concentration of the manuscripts would make one suspect it is of Kerala provenance, as well as featuring cultural items particular to that region, the medieval overlap and entanglement between what we think of as Kerala and Tamil Nadu today (Freeman 1998), as well as the mobility of the lineages propagating such texts, means it could just as likely represent an originally Tamil tradition that was later indigenized and redacted from several sources in Kerala. Particularly significant in this regard is the large Mantrapa-da, in which, as Goudriaan has pointed out (1977: 158–160), there is broad overlap in subject and language with the undeniably Kerala work, the Tantrasa-r. asangraha, perhaps of the fifteenth century or earlier. The latter is a digest of

144

Rich Freeman mantrava-da for toxicology, demonology, and ka-mya worship with a pronounced S´a-kta inclination, but this merely serves to intensify the already S´a-kta elements of Is´a-nas´iva’s paddhati. In the absence of a systematic examination of these relations, and with the paddhati inadequately edited or researched, derivation from or influence on the shared material could have gone in either direction, with the possibility that both could have drawn on a tradition of applied S´a-kta rites associated with a common source called Na-ra-yan.-ıya which figures in both textual corpora, as Goudriaan has noted (1977: 160). So in any case, if Is´a-na’s compendium reflects the dominant stream of the earlier Tantric tradition of Kerala, the Tantrasamuccaya certainly represents a narrowing of that tradition, particularly with regards to the S´a-kta deities and components. The only female deity prominently featured in the Tantrasamuccaya, Durga-, is assimilated to the Vais.n.ava structure of mantric worship, and shorn of the more sanguinary ritual or iconographic associations one would expect of the Puranic war-goddess. That the systemizing and narrowing effect of the Tantrasamuccaya on the pantheon invited a compensatory move is evident in a subsequent textual production, the S´es.asamuccaya, composed in the next generation, again within the same family (Narayana Pillai 1951: 3–4). As an explicit supplement to the Tantrasamuccaya, it took up a remainder (s´es.a) of deities who required a commensurate textual treatment to accord with Kerala practice. Aside from four male gods who were present in the Kerala temple campus,2 these additional deities comprised: (1) the potentially S´a-kta pantheon of Sarasvatı-, S´rı-, Pa-rvatı-, Jyes.t.ha-, and Bhadraka-lı-; then (2) a generic pantheon of the Seven Mothers (with accompanying Vı-rabhadra and Gan.apati), and Ks.etrapa-la; but then (3) another, discrete complex, characterized as the demon-slaying Bhradraka-lı- named Rurujit, joined by S´iva, Ks.etrapa-la, the Mothers, and allied deities, and the Lokapa-las in their own right, headed by Indra.3 Sarasvatı-, as the embodiment of speech, and therefore mantra itself, was a principal vehicle of the S´a-kta dispensation, especially as the syllabary (lipis) installed in the worshiper’s body, as we shall encounter later. S´rı-, though originating as Vis.n.u’s spouse, readily took on attributes of a S´a-kta entity in South India, as can be seen from the Laks.mı-tantra. Pa-rvatı-, as S´iva’s principal spouse, was readily assimilated to the embodiment of S´aktı- herself. Jye-s.t.ha- as the elder sister of the Goddess, is popularly associated with negative feminine power and sorcery, and Bhadraka-lı- is the fierce, potentially violent goddess, created by S´iva for the destruction of demons, who reinstitutes all those sanguinary and sexual associations the Vais.n.avized Durga- seemed to earlier expunge. Indeed it is Bhadraka-lı- (in various forms and modes, including Ca-mun.d.a-) who becomes the more prominent liturgical representative of the independent, martial offspring of S´iva, representing those ‘Bhagavatis’ for which Kerala is famous. As for the ubiquity of the Mothers in Kerala temples, even those dedicated primarily or exclusively to male deities, normally feature the Seven Mothers, in a row, flanked by Vı-rabhadra and Gan.apati on their ends, rendered in a

Medieval Malabar: S´a-ktism, polity, society

145

low-lying modular plinth (pı-t.ha) with all these deities represented by identical . little surmounted lingas. These Mothers are, of course, the seven consorts or S´aktis of the principal gods, as their named feminine counterparts. Set outside and to the south of the central shrine, as part of the inner circle of offering stones, they are given balis as a regular part of worship, everywhere (Figure 7.1). Ks.etrapa-la, the fierce guardian of the temple, is conceptually part of their entourage, but he has his own offering stone, in the northeast corner of the temple’s outer courtyard, within the main walls.

Figure 7.1 Nampu-tiri Brahman novitiate priests offering bali as part of regular worship to the standardized, modular representation of the Seven Mothers, with accompanying Vı-rabhadra and Gan.apati on the southern side of the main shrine

146

Rich Freeman

What seems initially odd in terms of the format of the S´es.asamuccaya as given above, is the last sequence of partly repeating deities, comprising the special demon-slaying Bhadraka-lı-, Rurujit, along with S´iva, and Ks.etrapa-la, and various forms of the latter gods. It turns out these constitute a special paradigm and configuration of temple layout and worship, of which the formerly discussed little modular group is merely a reduced and schematized rendering adapted to more ordinary temples. For while the first two groups of gods are all treated in the first six chapters of the S´es.asamuccaya, the seventh through the ninth chapters enjoin a coherently discrete ritual program (vidha-na) dedicated to Rurujit, the form of Bhadraka-lı- (or Ca-mun.d.a-) who is independently installed in her own shine, configured with another separate one for S´iva, an accompanying Ks.etrapa-la, and a fully iconically differentiated and free-standing set of the Seven Mothers in a hall (s´a-la) flanked by Vı-rabhadra and Gan.apati. These chapters deal sequentially with this special kind of temple in three separate pat.alas devoted to installation (pratis.t.ha-), to the liturgy and rites of worship, and to the festivals and renovations. The importance of this discrete section is that it clearly represents a sustained and relatively coherent unit of S´a-kta intrusion into the Kerala pantheon. As we will see, it reaches back into earlier strata of S´a-kta goddess texts, on the one hand, but has also left its unique historical impress in a chain of related, royal temples throughout the Malabar region, on the other. Before taking up these matters, however, and by way of background, I want to briefly review the social configuration of castes, ritual personnel and their forums of worship, and their relation to the traditional chiefs of the region. This will again be a necessarily cursory sketch of these relations as they bear directly on the matter of S´a-kta cults, since this should give us a better grasp of the context in which these textual remnants were developed and deployed.

Ethnological reflections While Nampu-tiri Brahmans present themselves as a consolidated and coherent caste from the outside, they were internally differentiated into a complex gradation of ritual and socio-economic status-groups with many practical boundaries to intermarriage.4 What was most prominent about them from an inter-caste, anthropological perspective was their famous symbiotic connubial relations of sambandham with castes under them. This institution of sambandham dictated that only the eldest Nampu-tiri male in any sibling cohort could legitimately marry a Nampu-tiri woman and reproduce the next generation. All the younger siblings had sambandham sexual liaisons with lower-caste women of clean, S´u-dra grade (temple servants, royals, chiefs, and militia), whose matrilineally reckoned offspring were legitimate inheritors in their own castes, but had no claim against their Brahman sires or their estates. The women of these matrilineal castes, the politically and socio-economically dominant castes of Kerala, could freely form multiple liaisons with any men of their own or higher castes and thus legitimately reproduce the social identity of their own

Medieval Malabar: S´a-ktism, polity, society

147

castes through their matriline. These latter castes fell into the two larger castegroupings generally known as Ampalava-si, the temple servant castes, and the Na-yar, which included most royals, political dominants, and soldiery of the land. As with any Brahmans, the Nampu-tiris original claims to ritual suzerainty came from their custodianship of the vedic knowledge and rites, and they have been made famous by the fact that a restricted circle of families retained this knowledge in an unbroken line down to the present (Staal 1984). In terms of the modern narratives of Kerala historians, however, Nampu-tiris seem to have been endowed with a mysterious power which supposedly allowed them to engineer the ‘feudal’ order of sambandham that at once preserved their land holdings and gave them license to sexually predate on the hapless (and implausibly gullible) dominant castes under them. Credit goes to Louis Dumont, however, who, in his survey of Nampu-tiris in their local and panIndian context, registered what they gave up in this process as well. Since only legitimately married Brahmans might keep the vedic fires, perform the vedic rites, and secure their place in heaven through legitimate ritual heirs who carry these on, Dumont noted that the majority of Nampu-tiri males would seem to have been individually faced with a kind of religious disenfranchisement in this arrangement (Dumont 1983: 112–113, 123–125).5 Faced with such impediments to the normal notions of vedic entitlement and ritual status, I would suggest that the alternate tantric order of installing deities in temples and performing pu-ja-s would seem to offer an obvious socio-religious compensatory validation. In fact, in reproductively staffing both temples and royal courts with their consorts and offspring, these institutions became a kind of surrogate, religiously extended family, a kind of tantric clan which compensated for their vedic debility, yet continued to draw on fraternal and ancestral ties to a vedic patrimony. From the perspective of the martial and royal castes themselves, the predominantly feminized cults of their militant, chiefly goddesses tapped the status of Brahmanical (vedic) religion, but underwrote this masculine formal authority with the claims to their own greater S´a-kta powers of sexual generation, psycho-physical energies, and bellicose political capacity. While I have charted out something of the complex imbrication of these religious ideologies through the vernacular works of their hybrid, Man.iprava-l.am literature which I cannot review here (Freeman 2003a, 2003b, 2011), I hope the reader can nevertheless appreciate my basic point on the social structuring of the caste-configured religious order. At a basic structural level, I want to suggest that the systematic correspondences between the caste- and gender-skewed reproductive order of upper-caste society, and the avowedly tantric religion which permeated the whole, with the Brahmans as Tantris and their martial and royal caste consorts as predominantly Ta-ntrika/S´a-kta initiates and devotees, cannot have been mere happenstance. The working out of this historical scenario requires a more complex analysis of the caste-histories and structures involved, which I must defer here. A further complication comes from these structures having been heavily overwritten, both through a later, largely pre-modern, Sma-rta reformism, and then the colonially inspired caste

148

Rich Freeman

reformism of modern, regional historians entangled with their own casteconditioned cultural politics. The religious ideology that features in most of these historical narratives is a vaguely formulated bhakti movement which draws selectively on its medieval Tamil S´aivasiddha-nta and S´rı-vais.n.ava counterparts with little sense of the pan-Indic and varied Tantric historical developments which significantly informed the doctrines and practices of temple culture in South India. The critical interventions of Sanderson and other Sanskrit scholars like Dominic Goodall (1998, 2004), showing the lines of historical derivation from pan-Indian sources into south India have yet to really impact Tamil scholarship on these matters, let alone the extension of related historical influences to Kerala. This paper attempts to inaugurate something of the latter sort with regard to a fairly clear line of S´a-kta development in Kerala. A major inroad into these problems at the most general level of the history of tantric practices has come from understanding that the ‘classical’ South Indian temple and priesthoods emerged socially and doctrinally by stages in the transformation from originally private cults of initiates undertaking individual worship for personal spiritual goals, into the consecration of permanent images installed in temples for public worship.6 In this way, whatever the earlier nature of public worship and shrines may have been in the South, these were successively colonized by Saiddha-ntika and other tantric schools to generate the medieval temple culture we think of as classically ‘Southern’. As Sanderson has demonstrated, the important class of early pratis.t.ha-tantras reveals the critical intermediate stage in which S´aiva officiants had expanded their role to regularizing the installation (pratis.t.ha-) rites of permanent images for worship in temples, but left the daily rites of worship and interaction with the public itself to other functionaries (2009: 274ff.). It is this historical stratification which I am convinced is sedimented into the Kerala system, between the different classes of the Nampu-tiri formation, whereby Tantris, as learned a-ca-ryas (in the language of the Tantrasamuccaya), have higher rights to install, ritually regulate, and perform annual rites of festivals, renewal and renovations in temples, but delegate daily pu-ja-s to the subordinate S´a-ntika-rar, whom they seem to have historically initiated for their role.7 The subordinate status of the latter is seen in the contrastive pairing preserved in this terminology, as well, for s´a-nti rites of ‘pacification’ clearly suggest a regime of propitiating lower, quasi-demonic godlings of the temple, compared to the foundational and fortifying exercise of tantra, as the science of loftier image-divinization in the Kerala context.8 In returning to the more specifically S´a-kta complex of the S´es.asamuccaya introduced earlier, we can see this same historical dynamic of accommodation more clearly evidenced in the social disjunctures of officiants brought together around the cult of Rurujit. For it turns out that the worship that was historically connected to this textually preserved system (as either its progenitor or legatee)9 is still flourishing in Malabar. I was intrigued in my original work around the royal shrine of the erstwhile chief of Nı-le-s´varam, at Mannampur- attu Ka-vu, where I came to witness the prominent festival of spirit-possessed Teyyam

Medieval Malabar: S´a-ktism, polity, society 149 dancers, to find an avowedly S´a-kta priestly caste I had never heard of in daily charge of the temple. Traditionally, this caste of Pit.a-rar (called Mu-ssatu, further to the south), worshipped the royal goddess inside the temple with the blood-sacrifice of cocks, and offerings of liquor, and the flesh as naivedya, as part of the regular service on Tuesdays, Fridays, and at other times of special request.10 In other respects, these Pit.a-rar are just like Brahmans, being invested with the sacred thread, studying their own s´a-stras for worship in Sanskrit, and even having their own temple servant castes for drumming and ritual assistance, with whom they engaged in sambandham relations, traditionally, exactly like Nampu-tiris. The Nı-le-s´varam Ra-ja, however, also supports another principal temple, the major S´iva temple of the town, as a regular Nampu-tiristaffed establishment conforming to the normal Kerala pattern of worship that we find in the Tantrasamuccaya. The goddess is considered the consort of the royal S´iva, receiving a garland from his temple daily, and during the Pu-ram spring festival, their portable images (tit.ampu, or utasava-mu-rti), still process from their respective temples to meet at the royal tank where they bathe together. It was reported to me that married couples also bathe together with them, as a kind of spring rite for auspicious marital life and fertility. On the occasion of the Teyyam festival at Mannampur- attu Ka-vu, many of the principal castes of Nı-le-s´varam send their own shrine officials to the goddess temple, where they gather in the outer compound, many of them undergoing ritualized spirit possession, to participate in the twinned Bhagavatis elaborately costumed, possessed Teyyam worship, along with the Teyyam of Ks.etrapa-lan. The major factions of the martial, Na-yar castes, and their ritual officiants, are prominent in the managerial membership of this, their former king’s, royal shrine, as are the toddy-tappers, the Tiyyas, who daily supply liquor for the goddess worship. Rather anomalously, as it seemed to me then, the installation, renewal, and annual purification and certain festival rites, are nevertheless performed under a Brahman Tantri, and on those occasions, the Pit.a-rar have to vacate their office and turn ritual jurisdiction briefly over to him. While I was unaware then of the full historical context or textual import, even in my dissertation I wrote of this temple and its priests as clearly representing a separate, S´a-kta order of wide social significance, more oriented towards power than towards the purity of the Brahmanical order and its S´aiva temple (Freeman 1991). The brief episodes of Brahman custody also seemed then, as now, designed somewhat anxiously to demonstrate the formally superior claims of Brahmans over the demonstrably more powerful and popular S´a-kta complex. In light of earlier observations, what we clearly have here is a similar division between the Tantri and the S´a-nti offices, but where these latter are in fact of different caste, socially differentiated and ritually specialized for S´a-kta rites. Subsequent research, travel, and textual study have now made it clear that the royal shrine in Nı-le-s´varam, and the whole chain of related royal shrines of various local kings and chiefs southward to Calicut, and beyond, undeniably represent the temple-type, cult, and liturgy of the Rurujit-vidha-na sections of

150

Rich Freeman

the S´es.asamuccaya.11 The layout of the royal temple of the Ko-lattiri Ra-ja that I visited at Ma-t.a-yi Ka-vu corresponds in its constituent shrines, deities, and their orientation to the description in the opening of the seventh pat.ala of the S´es.asamuccaya on pratis.t.ha- (Figure 7.2), as does the temple at the site of his royal martial gymnasium, Kal.ariva-tukkal, 17 kilometers to its south. I also later confirmed with one of the chief Pit.a-rar of Nı-le-s´varam, by going through the eighth pat.ala on pu-ja- with him, that his pantheon and liturgy corresponds in most of its essentials to the Rurujit chapter on pu-ja- of the S´es.asamuccaya. Of course, the seventh chapter on pratis.t.ha-, and the ninth on festivals, renovations, etc., would only pertain to the Brahmans who supplant the Pit.a-rar on these occasions, though it is doubtful that any of them follow the Rurujit-vidha-na in this context, as I explain below. Socially, the Malabar chiefly patrons of these temples intermarried and intermittently warred on each other, centered on two principal kingdoms: that of the Ko-lattiri Ra-ja, whose original capital and nearby main temple were some 30 kilometers south of Nı-le-s´varam (near the famous coastal mountain of Eḻimala); and that of the Sa-mudiri or ‘Zamorin’ of Calicut, another 65 kilometers south of there. All nine of these major chiefly temples, from Nı-le-s´varam

Matr-Sala Ksetrapala Siva

w

N

S

Camunda

E

Schematic of Shrines at Matayi-Kavu (not to scale) Figure 7.2 Rough sketch plan giving major shrines and their relative orientation in the Ko-lattiri’s royal temple of Ma-t.a-yi Ka-vu. This accords with the Rurujit-vidha-na incorporated into the S´es.asamuccaya

Medieval Malabar: S´a-ktism, polity, society

151

to Calicut, north to south, were and are staffed with the same clan of intermarrying Pit.a-rar or Mu-ssatu priests and their Ampalava-si assistants. Heading further south from Calicut, there were an additional five related temples, reaching all the way down into central Travancore, and including the famous royal Ce-ra temple of Kodungallur Bhagavati, roughly midway between Trichur and Cochin. While these temples are reported to have had the same S´a-kta installation, liturgy, and priesthoods historically, except for Kodungallur, they have reportedly all reverted to generic Nampu-tiri rites and priestly control in the modern period, so I will confine myself mostly to the Malabar complex. Alexis Sanderson’s main focus in his political treatment of the ‘S´aiva Age’ is naturally with prominent kings, those who left a noteworthy inscriptional record, had associations with prominent S´aiva Ra-jagurus, and undertook major temple and palace building projects (2009). But he further surmises, rightly, as I believe this Malabar material shows, that similar socio-political processes diffused through these cults more widely into the smaller chiefdoms and with a comparable reach deeper into society. Indeed, the higher political echelons would have necessarily relied on the latter infrastructural relations for their support. Relatedly, from the Brahmanical perspective, Sanderson has also shown with regard to the Paippala-din Atharvavedins of Orissa how the priestly formation adapted a Trika- and Ka-lı-kula-inspired S´a-kta cult into their ritual repertoire, under the attraction of royal patronage (2007b: 236ff.). In that context, he has also noted the presence of a parallel cult of Bhadraka-lıin South India, based on the Brahmaya-mala tradition of S´a-ktism, derived at some remove from the older north Indian text of the same name (ibid.: 277–278 n.141). It is in fact this southern Brahmaya-mala which seems to have been the ultimate source for the cult of Rurujit that is incorporated into the S´es.asamuccaya and informs the whole royal cult of Malabar S´a-kta worship. Finally, Sanderson has also suggested in the same article that the mantric digests of the Prapañcasa-ra and S´a-rada-tilaka were composed in the same Paippala-din milieu of Orissa (ibid.: 230–233). The former of these two textual traditions (but not the latter) turns out to be directly relevant to the Kerala material as well, and in certain regards highlights the whole Nampu-tiri imbrication with S´a-kta cults, as we shall see. Accordingly, in the remainder of this paper, I will try to broadly sketch out a number of intersecting social and ritual dimensions of the worship of S´a-kta and related deities in Malabar. This will entail surveying how Malabar chiefs articulated puranic goddesses and their cohort with local histories and S´a-kta cults and officiants, how Brahmans positioned themselves in negotiating between their institutional interest in Agamic temple culture and the challenge ´ from tantric Saktism, how tantric yogis made their appearance and their accommodation with Brahmans and the royal and caste order, and finally, how one bit of material culture around the charged issue of blood-offerings demonstrates an accommodation that is diagnostic of the complex layering of both doctrine and caste around the issue of ritual status and purity.

152

Rich Freeman

Polity and Puranic S´a-kta goddesses As we have seen, the Ko-lattiri Ra-ja’s tutelary shrine of Ma-t.a-yi Ka-vu provides an exemplary architectural realization of the Rurujit-vidha-na in the S´es.asamuccaya. While the name Rurujit, however, clearly implies that the goddess is the conqueror of the demon named Ruru,12 this demon is virtually unknown in Kerala. While the worship of this goddess, who is variously known within the liturgy by names and epithets like Can.d.a- Ka-pa-linı-, Aghorı-, and Bhairavıthat reach back into the Brahmaya-mala texts, including the old northern one (cf. Hately 2007: 8–9), she is known in local performance and in her local Puranic persona, including most prominently in the southern Brahmaya-mala, as the killer of the demon Da-ruka.13 This is the form of Bhadraka-l.i, of panKerala scope in Sanskrit and Malayalam texts, who hearkens back to the royal Ce-ra tutelary in her temple at Kodungallur. At Ma-t.a-yi Ka-vu, however, we find a common pattern of localization that I have repeatedly encountered: the demon is assimilated to a former nemesis of the local kingdom, and the Goddess, after killing this enemy, is pacified and installed as the royal protectress. Though this demon-enemy has the generic named Da-ruka, at Ma-t.a-yi Ka-vu the site of his old fortress and the scenes of the battle are all localized in folklore and programs of ritual performance. For at Ma-t.a-yi Ka-vu, as at Nıle-s´varam, the royal goddess is annually incarnated in the temple’s outer compound in the elaborately costumed and spirit-possessed vehicle of a Teyyam dancer, who sings his lineages as orally preserved liturgies of the goddess, after which he interacts with worshippers ritually and verbally as her living presence. In this way, the king’s royal goddess reaches out into the common classes, and she has spread in this manner (as did scores of other Teyyam deities from various sites) through the whole region of the Ko-lattiri’s domains. While the Da-ruka narrative occurs in brief form in at least some versions . of the Lingapura-n.a (1906 [1857]: 138–139), we need a thorough study of its textual lineage in order to trace either its origins from, or adoption into Kerala. The slayer of Da-ruka is further the puranic goddess whose myth is encapsulated in the inauguration of the southern Brahmaya-mala tradition, where her creation by S´iva trumps the powers of the Seven Mothers as Ekavıra-, the ‘Solitary Heroine’.14 I have found a long Sanskrit version of this myth in what most likely comes from northern Kerala, purporting to be part of the Ma-rkan.d.eyapura-n.a, and positioned to follow in that text just after the famous Devı-ma-ha-tmya (Tirumumpu 1975). Although it is in grammatically sound Sanskrit, it includes several instances of the word ku-l.i for ‘demon’, which only occurs in Malayalam and medieval Tamil. While the work’s frame story is narrated to a king of Kashmir when he was hunting in the Dan.d.a-ran.yaka forest, its action around the demon Da-ruka devolves from Gokarn.a, on the west coast, the northernmost outpost of what was Brahmanically reckoned as Kerala.15 It therefore seems highly likely to have been composed near the region where it is extant and published, i.e. the far north of Malabar, bordering on the Tul.u territory.

Medieval Malabar: S´a-ktism, polity, society

153

Cultically speaking, the text shows very clearly the impetus to incorporate S´a-kta esoterica into the puranic frame of the war goddess. The Ka-lı- of this text is herself a yoginı-, and a ka-pa-linı- who carries the skull bowl with which she eventually drinks the blood of Da-ruka as it froths from his chest when she tears him open. She has also associations with foresters and the wilds. When she plucks out Da-ruka’s heart, roasts it in the fire of her eye, and eats it, she is compared to a hunter’s wife, digging up a forest tuber.16 And yet, her ferocious powers are clearly derived from the application of textually warranted S´a-kta traditions. For a run of sixteen of verses in her praise encode into their first syllables the fifteen-syllable ka-di-version of the S´rı-vidya- mantra, and add a . final s´rı-(m) beginning the sixteenth verse, for the s.od.as´a-ks.arı-, the sixteen-syllable version of that vidya-.17 The goddess is then further explicitly praised as being Kubjika-, and again as S´rı-vidya- and Tripurasundarı-. In social terms, there is a recurring narrative motif in most Kerala versions of the goddess myth which I have argued reflects centrally the tension between the martial castes and the Brahmanical temple order that we have seen as the central fact of kinship underlying the reproduction of these castes and their integration into the Tantric order of worship. In these myths the goddess, created by the S´iva of the Brahmanical temple order, is so crazed with bloodlust after she has killed Da-ruka that she advances to attack the temple or Kaila-sa itself. In order to pacify her, Siva turns two of his offspring or servants from within the temple (in the case of this Da-rukavadha text, Gan.apati and Nandin) into wailing infants and orders them to lie in her oncoming path. Their cries rouse her maternal instincts, milk flows to her breasts, and as she takes them up and proceeds to suckle them, her rage is gradually quelled in the course of nursing them. In addition, the Brahman r.s.is are urged to sing her praises, and temple servants to produce music and dance for her, and this completes her domestication which gains her admittance into the outer precincts (gopura) of the temple. We could hardly ask for a clearer set of parallels to the ambivalent relations of martial servitude, connubium, and hybridly popular rituals by which, from the Brahman perspective, the dangerously violent and impure, yet necessary, matrilineal military castes were incorporated into the temple cult under the person of their goddess. The royal cults of actual Malabar history reflect both this puranic thematization, and the hybridly S´a-kta cultic content of the Sanskritic Rurujit-vidha-na, as seen in the practices and liturgies at the vernacular level. In one of the Teyyam liturgies of the royal goddess, it is Ks.etrapa-la and Vı-rabhadra who are transformed into children to nurse at the Devı- s breast, for these are the actual deities installed inside the temple, likewise manifested in Teyyam performances in the outside compound. In another such song, we see a more surprising instance of the osmosis by which esoterica of the temple cult passes into the liturgical domain of the possession deities. One of the diagnostic features of the Rurujit-vidha-na, coming from the Ma-tr.sadbha-va texts on which it is based, is the presence of ten differently named Ks.etrapa-las as guardians of the directions (the four cardinals, their intermediates, the nadir

154

Rich Freeman

and the zenith).18 While in the folk-imagination the guardian godlings of temples shade readily into demonic collectives, in recollection of the earlier narrative of S´iva creating children to pacify the goddess, there is a further carry-over from the liturgical presence of ten Ks.etrapa-las: As Ka-l.i, creating havoc, approached Kaila-sa in rage, in his fear, The lord of Kaila-sa deftly took his gaping, howling demons, And in the form of ten infants there quelled her rage.19 The protean nature of these martial, guardian, quasi-demonic beings allows them to mediate across caste and political boundaries, forming various subpantheons that can alternately merge or individuate, constellating around the main dyad of S´iva and the Goddess (though even the Goddess tends to bifurcate, as we shall see). The very formation of the sub-kingdom of Nı-le-s´varam, from an allied attack on the region by the Ko-lattiri to its south, and the Zamorin, further to his south, exemplifies this. In the mythology of these events, an alliance of the gods Ks.etrapa-la, Vı-rbhadra, Ve-t.t.akkarumakan, and S´a-sta-vu (S´a-str.) ventured north and conquered the various demon-kings of that region. In historical fact, it was the lineages of Na-yar warriors represented by these tutelaries who conquered Nı-le-s´varam, settled there, and established temples to these deities, often on the sites of earlier temples, and in association with goddesses of those sites. As these Na-yar transitioned from soldiers to chieftains, their deities could similarly morph from being lowly temple guardians, commensurate with their common status in their original territory, to higher, fully enshrined deities. Thus we have two major temples to Ks.etrapa-la, and one to Vı-rabhadra under three of the major Na-yar lineage-formations near Nı-le-s´varam, each in a complex alliance with the earlier goddesses. At Nı-le-s´varam itself, the center of the new kingdom where the Ra-ja established himself, the Da-rukatemplate is worked out in an illustrative local variant. Being afflicted by an evil demon-king named Mannan (meaning simply ‘king’ in old south Dravidian), his subjects eventually petitioned the Ko-lattiri’s goddess at Ma-t.a-yi Ka-vu to liberate them. She indeed came north, killed Mannan at the site of his old palace-temple, and established herself there in the temple of Mannampuṟattu Ka-vu.20 Since she was the goddess of Ma-t.a-yi Ka-vu, the Ko-lattiri naturally sent up a lineage of Pit.a-rar priests from his own tutelary temple to worship her according to the proper S´a-kta rites of her original shrine. There are actually two goddesses enshrined at Mannampuṟattu Ka-vu, the principal, eight-armed Can.d.ayoges´varı-, who is worshipped more or less according to the Rurujitvidha-na, and on the same pı-t.ha, a silver Bhadraka-lı-, who is reportedly worshipped according to her liturgy in the main part of the S´es.asamuccaya. Correspondingly, there are also two Teyyams to the different goddesses that are also performed for the Teyya-t.t.am festival, in the temple’s outer compound, one of whom represents the Ko-lattiri’s tutelary who conquered Mannan, while the other represents the earlier chiefly goddess of Mannan himself. In addition, Ks.etrapa-la himself is incarnated as a Teyyam, as the warrior-guardian son of

Medieval Malabar: S´a-ktism, polity, society

155

his Goddess-mother. The spirit-possessed Teyyams have offerings of huge kalas´ams of toddy and cock-sacrifices offered to them outside the temple, while traditionally, the Ca-mun.d.a- inside had blood-sacrifice and liquor offerings, as well. So as a simultaneous orchestration what we see, both at Mannampuṟattu Ka-vu, and the Ko-lattiri’s royal shrines (and the others through northern Malabar), is an inner temple-cult of S´a-kta, Sanskritic rites, and an outer cult of possessed folk-worship coordinating the martial and lower castes all under royal patronage, and framed in local, historical versions of Keralawide Puranic charters of conquest and divinely sanctioned rule. This was the way Tantrism and S´a-ktism functioned socio-politically in medieval Kerala, and this accords very well, it seems to me, with the general principles that Sanderson has laid out for the broader S´aiva-S´a-kta pattern. While I have written at length of the Brahmanical, royal, martial, and lower-caste articulations of Teyyam deities throughout these realms (Freeman 1991), what I now realize is that the temple-internal, royal cult which ordered these realms was this Rurujit-vidha-na S´a-kta cult, which even the external, Teyyam manifestation of it made clear was for royal purposes. The royal Teyyam liturgy itself praises the Goddess as the ‘essence of the s.ad.a-dha-ra’ in Tantra-yogic terms, as the ‘syllabary of the fifty-one’ of Sarasvatı-, from which mantras derive, as the ‘nine-syllabled’ mantra of the Brahmaya-mala, and the ‘sixteen syllabled’ mantra of the Tripurasundarı- cult. This goddess was generated ‘to destroy the lineages of enemies’, in ‘great affection for the king’ as ‘the Kerala hero who would not falter’, and she was the personification of ‘the dı-ks.a- that was enjoined to protect the entire lineage of the Ko-lattiri king’ (Vis.n.u Nampu-tiri 1993: 26, 39, 42, 49).

Lineage of the S´a-kta officiants In turning to the principal S´a-kta officiants themselves, we confront a historical puzzle. While we do not know the origins of the Pit.a-rar, or the circumstances of their entering the cult of Malabar’s royal shrines, there are nevertheless clues to their tantric milieu. The name Pit.a-rar is generally thought to be a vernacular form of bhat.t.a-raka, which is phonologically quite plausible, and contextually fits very well as an extremely common title of S´aiva and S´a-kta officiants in medieval India. Indeed the earliest references to Pit.a-rar from twelfth-century Kerala refer to them as looking after temple matters (s´rı-ka-ryam) with their students (s´is.yar), as involved with Yogis (co-kikal.) in the context of temple grants, and as imparting religious teaching (upades´a) (Ve-n.ugo-pa-lan 2009: 880).21 Local tradition attributes the authorship of the Sanskrit S´a-kta hymn (stuti), Laghubhat.t.a-raka, to a Pit.a-rar of Mannampuṟattu Ka-vu, after whom it was supposedly named (C. N. Nampu-tiri 2004: 8–9). While this claim carries little weight as to authorship of this particular work, it lends cultural support to the association of Pit.a-rar with S´a-kta worship, for this stuti to the Ba-la- form of Tripurasundarı- has been quite popular in Kerala.22 The samayavidya- of Tripurasundarı- is indeed incorporated into the Rurujit-vidha-na of the

156

Rich Freeman

S´es.asamuccaya, and in the Calicut goddess temple of the Zamorin, a s´rı-cakra is actually installed and worshiped there. Furthermore, it is clear that the common name for such S´a-kta goddesses in the Tamil and Kerala countries, Pit.a-ri, is cognate with this complex of S´aiva/ S´a-kta appellations. Sanderson has discussed a thirteenth-century Co-ḻa inscription from Ko-la-r, in Karnataka, as exemplifying a cult identical to, or closely allied with the southern Brahmaya-mala tradition (2007b: 277 n.140). The modular pantheon of the Mothers there would indeed fit perfectly into the Malabar array, and the separate, principal goddess is Ca-mun.d.es´vari, who is also referred to as Pit.a-riya-r (with an honorific ending), which the editors of the inscriptions note is derived from Bhat.t.a-rika-, a regular Sanskrit title for the Goddess. Evidence further convinces me that the occasional appellation for such goddesses in Kerala, Pis.a-ri, is merely the attempt to (re-)Sanskritize the medieval Dravidian vernacularized form, Pit.a-ri. A temple that is one of this Malabar S´a-kta group, in the territory of the Kuṟumprana-t.u chief, some 30 kilometers north of Calicut, is indeed called Pis.a-ri Ka-vu, after the goddess there, and I believe such variants were once common for the S´a-kta goddess throughout Kerala. In central Kerala, I studied several goddesses around the great temple of Peruvanam named Pis.a-rikkal who were all said to derive from the S´a-kta goddess, Ka-ma-ks.i, at Ka-ñcipuram, by translocation through spirit possession of persons or objects. While the -kal element in her name is simply an old locative, I believe the projected cognate of Pit.a-ri was actually Bhat.t.a-ri (as though a feminine formed from bhat.t.a-ra),23 rather than Bhat.t.a-rika-, and I indeed found that the principal village goddess of Peruvanam, now called simply ‘Mother of Urakam’ (Urakattamma, after that village), originally bore the name Bhat.ari according to an inscription of perhaps the eleventh century (Srinivasan 1986: 60). So the point is that Pit.a-rar, as a contemporary castetitle for a S´a-kta officiant no doubt derived from Bhat.t.a-ra(ka), as a title for S´aiva officiants, but was also closely associated with the feminine Bhat.t.a-ri (ka-), as a term for the S´a-kta or assimilated folk-goddesses whom they served, convergent with the pan-South Indian Pit.a-ri and her cult. Further evidence of originally S´a-kta-yoga associations, is found in a major caste of Kerala temple-servants called Pis.a-rat.i. They are a highly Sanskrit-learned caste, and the ones I studied with in central Kerala were also noted for their expertise in indigenous modes of temple performance-arts and music. Most interesting as to their origins is the claim that they derived from Brahmans who undertook renunciation, but then lapsed back into householder life, being subsequently demoted to the status of non-Brahman temple servants. They were indeed matrilineal, having had sambandham with Nampu-tiris, but lending credence to the legend of their previous renuniciant status, they were traditionally not cremated, but buried, sitting in ‘sama-dhi’ in a grave filled with salt. (This is exactly the treatment given to the bodies of sannya-sis and the caste of Yogis I shall discuss shortly.) It seems fairly certain that the name Pis.a-rat.i is again derived from the Pit.a-rar/Pit.a-ri complex,24 and in northern Kerala they are closely associated with the S´a-kta temples. At the

Medieval Malabar: S´a-ktism, polity, society 157 aforementioned Pis.a-ri Ka-vu, a Pis.a-rat.i officiant must be present for the announcement of the date of main festival, in addition to the regular Mu-ssatu (the southern equivalent of the Pit.a-rar) who serve as the pu-ja-ris there. So an intriguing question becomes what the historical relation may have been between the Pit.a-rar and the Pis.a-rat.i, and whether in fact they derive from the same or analogous communities of S´aiva/S´a-kta yogis, gurus, or officiants who took different social trajectories. While the Pit.a-rar have largely assimilated to a Brahmanical model (though in consuming alcohol and meat they are rejected as such by Nampu-tiris), their elevation into the royal cult by Malabar kings may have allowed a path different from that of the Pis.a-rat.i, who were more or less co-opted into a second-tier status in the Nampu-tiri-dominated temples.25 It is notable, however, that the Pis.a-rat.is have remained some of the finest Sanskrit scholars among the non-Brahmans, and in my experience studying with them, they seemed to think this learning distinguished them from other Ampalava-sis and indeed from Nampu-tiris.

Brahman mantrava-da, yogis, and popular religion In looking to other social strata for the Tantra-S´a-kta influences that might have shaped Malabar, we must turn back to the Teyyam cults, and consider not just the relations to royalty and the puranic legitimation of polities of conquest, but to the S´a-kta identities of these ‘folk’-deities, and their wider relations with society. Certain families of Brahmans actually patronized Teyyam worship at the domestic level of their manor-shrines, and especially those Brahmans associated with mantrava-da. These Teyyams are generally grouped in a pantheon of five, of slightly variable constituency, but always so as to include Bhairavan as their head, along with Kut.t.icca-ttan, and Pot.t.an. This pentad of Teyyams are known as Mantramu-rtis, and are associated particularly with the practice of mantrava-da. While I have written at length about these deities elsewhere, I will just say of Kut.t.icca-ttan that he is the apotheosis of a virulently S´a-kta sociopath (Freeman 2006), and of Pot.t.an that he is an untouchable tantric gnostic who is Sanskritized into the Can.d.a-la who accosts . and humbles S´anka-ra-ca-rya as the narrative basis for the well-known Veda-ntic stuti, the Man.ıs.a-pañcaka (Freeman 2009). Bhairavan brings together two personas: one is the Tamil version of Bhairava who, as a Bhairava-yogi, cannibalizes his devotee’s son in the famous Ciṟutton.t.ar story of the twelfth-century Periyapura-n.am; the other tells of Bhairava’s similar destruction of a S´aiva-yogi pontiff’s lineage in a famous Mangalore mat.ha, linked to the migration of house-holding lineages of Yogis into Malabar (Freeman 2009). The last is particularly important in its likely historical influence, to which I will return, but for the moment, I want to just point out that these Mantramu-rtis are the most violently anti-Brahmanical of the Teyyams, in an explicitly tantric mode, and that the Brahmans who master or worship them are thereby considered as powerful mantrava-dins. While I found such mantrava-dins therefore to be not highly regarded by some of the more orthodox, temple-centered

158

Rich Freeman Nampu-tiris, it is these families who are the Tantris of the chiefly and martial temples where Teyyams are performed (along with another lower division of Brahmans called Empra-ntiri). These mantrava-dins were known as the Samprada-yis in the vernacular folk-liturgies and literature, and this would seem to have continuities with the old northern Brahmaya-mala tradition, as well, where the usage suggests similar attainment and transmission of mantrasiddhi (cf. Hately 2007: 415–417 n.114). As we shall see, this mantrava-di-samprada-yam was also transmitted across the caste-barriers. It is significant to note in this context that mantrava-da occupies a different position between the presumably original private cults of Tantric initiates, and the public worship in temples. For it is a kind of private attainment (mantrasiddhi, enabled through initiation), applied to the very limited public of clients in a restricted, domestic setting. In fact, it could be argued in terms of earlier reflections on royalty that it was perhaps when that client was a king, that we begin to get the genesis of an expansion through his patronage and following into a temple cult. This touches on the development that Sanderson has treated in regard to Tantric gurus becoming Ra-jagurus, and an issue that is central, I believe, to the Rurujit-vidha-na emerging from adaptations of the older Brahmaya-mala tradition. This hinges on the competition between established Brahmans and later sectarian mendicants for royal patronage, and this brings us back to the Teyyam figure of Bhairavan (Figure 7.3). As I have stated earlier, and worked out in some detail, there were certainly one or more historical movements of Bhairava Yogis into Malabar. From their mat.ha in Mangalore, associated with the Kadri temple complex that dates back to the eleventh century, they had built up a substantial holding of lands and a kind of pontificate that was reported on in its decline by Pietro della Valle in the early seventeenth century.26 By this time, their establishment had been sacked by their rival Vı-ras´aivas under Na-yaka rule, and if there had not already been intermittent migrations of sedentized, householder Yogis (called Co-yi) into Malabar, there was certainly an impetus to migrate after they lost their lands and management of the Kadri temple. It is they whom I have demonstrated brought in performative episodes of ritualized begging in full Bhairava-yogi regalia, and it is this costuming that was incorporated into the Teyyam of Bhairavan, with its narrative liturgies I have touched upon earlier. These liturgies relate both to the earlier Bhairava cult in Tamil Nadu (which I have found spread from that source not only into Kerala but into the Karnataka and Andhra countries with Vı-ras´aivism), as well as to the local memories of the wreckage of their pontiff’s lineage-control of Kadri. What is of great historical interest, in both these streams, is the recollection of origins or affiliations of these yogis with Ciṟṟampalam and Ponnampalam, that is to say, with the shrines of the great Tamil temple-complex at Cidambaram. An esoteric Bhairava cult in association with the Nat.ara-jan of Cidambaram is attested, inscriptionally, iconographically, and in texts, as is the S´a-kta goddess there as Pit.a-ri, both in an independent temple as Tillai Ka-l.i, and earlier in the Ciṟṟampalam shrine itself (Soudararajan 2004: 49–50). On the Malabar side,

Medieval Malabar: S´a-ktism, polity, society

159

Figure 7.3 The Teyyam of Bhairavan, performed by Malayan dancers, modeled on the costumed ritual-begging of the Co-yi (Yogi) caste who settled in Malabar

the Yogis are reported as resorting to the aforesaid shrines as their central refuge, and dancing there in possessed ecstasy. They are also celebrated as having known multiple languages and as having traveled widely, both in the liturgies of Teyyam, and in a number of folk-rites in the spring Pu-ram festivals I will allude to shortly. So we clearly have evidence, (though over what time period, and in how many successive movements we cannot tell), of Bhairava Yogis most immediately from Mangalore, but with ties into Tamil Nadu, of learned and relatively cosmopolitan profile, settling in Malabar. They either brought their women with them when they entered the region, or they found wives among the local women, and transformed themselves into the settled caste known locally as Co-yis.

160

Rich Freeman

They either were, or have become matrilineal, and in certain lineages individuals are still schooled in a Bhairava-S´a-kta cult and initiated into the office of Gurukkal., (recalling the name of the S´aiva officiant-caste of the Tamil country). These Gurukkal., who govern the community and their shrines, worship Bhairava and various fierce, local goddesses, with blood-sacrifice, meat-naivedya, and alcohol, as staunch and avowed S´a-ktas. There is a commensurate lore of these Gurukkal., best exemplified by the Kurikkal. (Gurukkal.) Teyyam (Vis.n.u Nampu-tiri 1981: 54–57). This Teyyam’s liturgy tells of a local Co-yi who rose to eminence through learning and mantric-power, and was admitted to all the various kings’ palaces where he was honored. Eventually the Ko-lattiri king was afflicted by a debilitating spirit possession that none could cure, except the Gurukkal., through various of his S´a-kta blood-rites. He was rewarded by the king with titles and gifts, but then on his way home was murdered by those in the realm who could not tolerate one of Gurukkal. status rising to such prominence. On his death his gods turned him into a Teyyam, and he is still worshipped today by rites of costumed possession. What this liturgy shows is that these Gurukkal. were both honored in the courts, and were in competition for royal patronage with others. Some of these others were local Brahmans, and one of the well-known events of Malabar history is that a Ko-lattiri Ra-ja (perhaps of the fourteenth century or later) came into conflict with the Brahmans of his own realm and so brought in others (called Empra-ntiri) from the Tul.u country as rivals and replacements.27 The earlier treated stratification among the temple-based Nampu-tiris between Tantris and S´a-ntis, the resort of Brahmans to mantrava-da with the more active embrace of the lower-caste Teyyam worship, and the division between these older Brahmans and the imported Empra-ntiri, all testify to the internal and external pressures among Brahmans to compete in the religious fora of the chiefly socio-political order. S´aiva and Tantric mendicants must certainly have been a force in this competition. The strategy of Malabar kings and chiefs, as in Nı-le-s´varam, to patronize both a Brahmanical S´iva temple, and a S´a-kta goddess temple, and to bring them together in various complimentary rites, but also to have a Brahman Tantri in ultimate charge over his S´a-kta counterparts, testifies to the entangled hierarchies that could emerge. The Pit.a-rar are elevated in their own S´a-kta domain, and treated as Brahman priests by all the Na-yar and other middling castes of their worshipping constituency for most of the ritual calendar, yet are not ‘real’ Brahmans in the highest temple order. The same is the case for their lineage-mates throughout the chain of Malabar temples. A similar situation came to eventually obtain, but lower down the caste-ladder, for Co-yis and their Gurukkal.. These, I was surprised to learn, perform the pratis.h.ta-, and function analogously to the Tantris in the Teyyam and other shrines, except that they do so for major castes in the avarn.a category, those traditionally falling below the Na-yar as putative ‘Untouchables’. Recognizing this role of the Co-yi Gurukkal., however, now explains some peculiar features of the Pu-ram song-festivals which are central to the collective spring worship

Medieval Malabar: S´a-ktism, polity, society

161

among the avarn.a castes. For these communities, whom one would expect to have been largely illiterate, have amongst them Sankskritically learned masters, called Pan.ikkar, who train singing and dancing troupes to perform various genres in the Pu-ram rites directly related to our concerns. These include: celebrations of S´iva and S´akti, specifically linked to their dance in the shrines of Cidambaram; depictions of Bhairava Yogis, including deportment, dress, and material from the Tantric lore of mantra and yogic psycho-physiology; various versions of histories of the Yogi’s wanderings, and their settlement as farmers in Malabar; and various tellings of the Cidambarama-ha-tmya (narratively related to, yet independent of, Tamil and Sanskrit versions of those texts).28 It seems to me likely that this tutelage came from the close intimacy (and perhaps some original connubium) between the Gurukkal. and prominent castes like the Tiyyas and Man.iya-n.is among whom the Pu-ram rites flourish. The presence of the Gurukkal. (as literal gurus) among these castes would also explain the many Teyyams originating among these lower castes, whose protagonists attain their apotheosis through being celebrated as great yogis, S´a-kta initiates and mantrava-dins. The Teyyams also testify to relations of inter-caste tutelage from the Brahman Samprada-yi-mantrva-dins, which shows there were attractions, among these echelons of Brahmans, to compete with Yogis even for avarn.a clientele. It is also likely, despite the later wave of Sma-rta reformism that has made such caste relations seem unthinkable, that the force of a S´a-kta ideology at one time licensed such intimacies as it did among non-Saiddha-ntika Tantrics elsewhere in India. Even the Teyyams of Pulayas, the traditionally lowest caste of agrestic laborers, include their own Ton.t.accan or Guru whose liturgy recounts his studying under a female Yoginıto learn her special art of shape-shifting into animal form.

Textual dynamics in social context If overtly S´a-kta teaching and practice were eventually pushed socially downward by the hegemony of Brahmanical orthopraxy, the Pit.a-rar and the royal cult of the goddesses, as well as Brahman mantrava-din practice, carried much of the doctrine and discursive content upward. It was at the level of orthodox Brahmanical practice and textual production that this tension was most clearly registered, and I want to return to these issues in the dynamic we started with in terms of speculatively placing the Rurujit-vidha-na in the milieu of the S´es.asamuccaya and Tantrasamuccaya. At a similarly Brahmanical textual level in Orissa, Sanderson has treated a comparable dynamic among the Paippala-din Atharvavedins, where they both incorporated S´a-kta pantheons and variants of their rites into their worship, on the one hand, but also expunged them doctrinally and ritually of unacceptable elements, on the other. Relatedly, he suggests that the great mantric digests, the Prapañcasa-ra and S´a-rada-tilaka were products of this milieu, and likely also derived from the Orissan region (2007b: 336ff., 230–233).

162

Rich Freeman

In reviewing the general trajectory of the Tantrasamuccaya, I would argue in a similar vein that it represents the attempt of its author to first reduce the much vaster and S´a-ktified content of the Is´a-na compendium to a controlled program of Sma-rta-style worship, but that the inadequacy of this initial attempt to keep up with S´a-kta pressure on various fronts, led to the additional accommodations of the supplementary S´es.asamuccaya. This force is seen at the vernacular level by the fact that while the Tantrasamuccaya lacks any treatment of Bhadraka-lı-, the vernacular renderings of the Tantrasamuccaya all add her on, as though she should have been included with the seven deities of the original’s pantheon, and which cannot do without her. Within the S´es.asamuccaya itself, this force is even clearer in the modular addition of the Rurujit-vidha-na to the end of the work, with its partial redundancy in pantheon, but the enhanced S´a-ktification stemming from its origins in the blood-cult of the southern Brahmaya-mala. This last legacy is particularly interesting in social terms because, again as Sanderson has noted, the daily priests of this Da-ruka-slaying Ka-lı--cult are not Brahmans, but Pa-ras´avas, defined as the offspring of Brahman men and S´u-dra women who should officiate for Ka-lı- (2007b: 277–278 n.142). This taxon would of course apply to what all the Ampalava-si temple-servant castes in Kerala theoretically were, under their sambandham relations with Nampu-tiris. We are thus left ultimately wondering about several scenarios for the origins of the actual S´a-kta survivors of this order in the persons of the Pit.a-rar. Were they originally Ampalava-sis of this Pa-ras´ava-sort, who were upgraded into their quasiautonomy as quasi-Brahmans in the cult of the Malabar kings? Or were they S´aiva mendicant Bhat.t.a-rakas, Ra-jagurus who were originally autonomous S´a-kta officiants, later demoted and somewhat subordinated under the Sma-rta orthodoxy of Nampu-tiris? It seems likely, looking comparatively to the varying status-trajectories of the Pit.a-rar, Pis.a-rat.i, and Co-yi that we have evidence for different social trajectories for different S´a-kta constituencies in different times and places. But what is clear, especially given the now documented S´a-kta influence at all levels of the caste-system and socio-political formations, is that there was a major socio-religious and historical impact represented by the various textual remains available to us. This then invites a marshaling of these sources with a sharper set of historical agenda in mind applied to their text-critical study. Construing the S´es.asamuccaya and its Rurujit-vidha-na chapters in these terms (but with much of the fine-grained textual study still to be done), I would suggest that as with the Paippala-din materials, Nampu-tiris were responding to larger S´a-kta and Tantric movements and trying to gain ritual control of them, to bring them within the orbit of their social authority. It is an interesting fact, however, that no one seems to know of any temples where this Rurujitvidha-na, expurgated though it is, has been a part of any Nampu-tiri Tantric practice. We have only the remnants of these temple layouts and greatly redacted pu-ja-s of the Pit.a-rar in Malabar, and the correlated practice of Tantris, during their brief annual roles in these temples, which seems to have lapsed into the

Medieval Malabar: S´a-ktism, polity, society 163 generic Nampu-tiri Bhadraka-lı--template of the earlier parts of the S´es.asamuccaya. While it is possible that the inclusion of the Rurujit-vidha-na section was never intended for Nampu-tiri practice (but then why expurgate it, and why include it in the whole project of the Tantrasamuccaya and S´es.asamuccaya?), it is also possible that the two chapters on pratis.t.ha-, and festivals and renovations were intended for Nampu-tiri Tantris, while the elaborate pu-ja--pat.ala was for Tantris only during their special festivals, with the ordinary pu-ja-s being left to local practice and unspecified. Or perhaps they also intended the text as a project of expurgating even the ordinary pu-ja-s of others in accordance with their own predilections. In any case, while the project was a historical failure, in terms of not being followed by Nampu-tiris today, it shows that at the time, there was some urgency felt for these Brahmans to gain control of this cult and incorporate it into their ritual program. To close these textual considerations, tentative and provisional as they are, I would like to suggest that the Prapañcasa-ra (but not the S´a-rada-tilaka) stands in a similar relation to the Tantra-S´es.asamuccaya tradition of Kerala as either of the former texts (or both of them) perhaps did to the traditions of the Orissan Brahmans. It is first of all noteworthy that the Tantra-S´es.asamuccaya tradition does not seem to draw on the S´a-rada-tilaka, while it does, quite significantly, on the Prapañcasa-ra. This suggests that we may need to drive more of a historical wedge between the Prapañcasa-ra and S´a-rada-tilaka, and query whether the former is as tightly bound to an Orissan milieu as the latter.29 More immediately important to our present concerns, however, is the similar force the Prapañcasa-ra has as a template for expurgated S´a-ktism, and its redeployment from its centrally mantrava-din context, to one of pratis.t.ha- in the Kerala tradition. Hence we find, significantly I think, that the dhya-nas´lokas in the Tantrasamuccaya for S´iva and (by tradition) for Sarasvatı- are taken directly from the Prapañcasa-ra.30 It is interesting that when Nampu-tiris had a substantial S´aiva Siddha-nta corpus to draw upon, they took their S´iva from here, along with their (Lipi-)Sarasvatı- in the context of the extensive lipi-nya-sa that is generic (and S´a-ktically so) for the pu-ja- of all the deities in the Tantrasamuccaya. Tellingly, when we come to the S´es.asamuccaya, this derivation expands significantly, for Su-rya, Sarasvatı-, S´rı-, and Pa-rvatı- are all taken directly from the Prapañcasa-ra.31 Of the two dhya-na-s´lokas from within the Rurujit-vidha-na for Bhadraka-lı-/Ca-mun.d.a-, the first for her four-armed form is found in our grantha manuscript of the Ma-tr.sadbha-va (no. R5126: 117), while both this and the s´loka for her eight-armed form are found together from another manuscript of a free-standing text titled Rurujidvidha-napu-ja-paddhati (no. R3365: 17), the origin for which is indeed the palace of Ko-lattiri Ra-ja (ibid.: 102). The provenance of this latter text makes one wonder, finally, whether the Rurujit-vidha-na chapters found their way into the S´es.asamuccaya from such Malabar sources, or whether the Ko-lattiri’s scholars extracted this material for use of the Pit.a-rar, or their prototypes. It is perhaps significant that the Tantri of the Zamorin’s temple in this Malabar complex belongs to the Ce-nna-s family, bearing the same name as that of the authors of the

164

Rich Freeman

Tantra- and S´es.asamuccayas. No doubt some of the questions I have raised in a provisional and speculative fashion might be answered with more thoroughgoing combing of these textual corpora (and field investigations), which I hope to continue. The pantheons, mantric arrays, rituals, and any accompanying cultic and socio-political information must all be garnered from the different versions of the southern Brahmaya-mala, the Ma-tr.sadbha-va, the Rurujidvidha-na, and the S´es.asamuccaya and carefully compared in hopes of filling out the history of this system.

Artifice and expurgation in blood-rites I wish to end this discussion with an example of material culture in the present which bears on these texts and redirects us to considering an important part of their socio-cultural dynamics. In my fieldwork, I encountered the ubiquity of what is called gurusi, or an artificial blood that is made of turmeric and building lime mixed in water. From my study of early Tamil, I realized that the word was a phonological Sanskritization of the old Dravi. dian kuruti, which is simply ‘blood’, but especially used in the Cankam literature of human blood shed in battle and sacrificial blood. The latter usage is still retained for actual blood-sacrifice among the non-Brahmanized castes in Kerala, while the Nampu-tiris use the artificial word and substance, gurusi, for propitiating S´a-kta goddesses to whom they offer huge cauldrons of the stuff (Figure 7.4). I have also seen these Brahmans combine the gurusi constituents with rice for certain balis in festivals and mantrava-da in explicit replacement of meat. It has been evident from my fieldwork that there is a kind of mimicry at work here, in which Nampu-tiris strive anxiously to capture the force of S´a-kta blood-rites, while avoiding the attendant impurity of slaughtering animals. I was puzzled, however, to find non-Brahmans also using gurusi, even in real sacrifice, by dipping the severed necks of their victims into a cauldron of the substance and swishing it around, before emptying the vessel in oblation. Though I suspected they might be emulating Brahmans in this, when I asked why they would use a substitute for blood, when they had the real thing with all its attendant powers (and unavoidable impurity), they revealed that a completely different logic was operative: they wished to offer huge pots of real blood, but couldn’t afford to sacrifice so many animals or such large ones, so they used the bloodied gurusi to express their good intentions to the gods, despite their paltry resources! These obversely twinned logics of substitution in sacrificial offerings have directed me back to the texts, Sanskrit and vernacular, and I wish to close with some interesting connections among them. A number of Teyyam worshipers reported to me that particular enshrined goddesses could no longer be performed as Teyyams, since they would require human sacrifice, which could not be provided now, as in the past. Certainly it was evident from the liturgies and performances of even less demanding deities that the blood-sacrifices they accept are representative of the ideal of

Medieval Malabar: S´a-ktism, polity, society

165

Figure 7.4 Nampu-tiri of a prominent Va-dhyar family offering gurusi-tarpan.am from a large cauldron of artificial blood at his feet over a woven man.d.alam to their family goddess of Rakte-s´vari in a domestic rite

human sacrifice. Most telling for me was the so-called Ko-ḻi-sto-tram, which reveals the history of the sacrificial chicken (ko-ḻi), to have originated as Da-ruka, who fled the goddess by taking various forms as he was successively slain, earlier as a warrior, and finally as a chicken. The modern avarn.a officiant recites this history to his goddess before slaying the fowl, and summarizes its ritual portent with the lines: Previously, I devotedly hacked a man and gave you his blood. Then when a man could not be happily gotten, I slaughtered an elephant for you. Getting neither man nor elephant, I cut a goat. Now getting not a single elephant, nor man, nor goat, by cutting a chicken I am giving you blood today.32 I was immediately reminded of these lines when I read the following on the substitutional chain of victims for pacifying the Mothers from the southern Brahmaya-mala:

166

Rich Freeman The generation of the Mothers was in order to destroy Asuras; One should effect the welfare of all beings by duly pacifying these Mothers. The sacrificial victim is said to be of three sorts: highest, middling, and lowest. Killing an elephant, a horse or a man is designated the highest; while a buffalo or camel is middling; and a goat or pig the lowest. Again, a deer is the highest, while a chicken is the lowest.33

Clearly this earlier S´a-kta logic is closer to the sentiments of my Teyyam informants today than it is to Brahmans. In turning to Is´a-nas´iva’s paddhati, we find further information to illuminate these issues, a transitional situation involving Brahmans and Pa-ras´avas that perhaps anticipates the Nampu-tiri/Pit.a-rar split. The pat.ala on the Mothers there (Kriya-pa-da 59), presenting substantially the same cult of our concern, ends with a pacifying s´a-nti-rite giving a social ranking to the order of sacrificial victims (Gan.apati S´a-strı- 1922: vol. 77: 579). While we have developed away from overt consideration of offering a human victim, it is nevertheless designated maha-pas´u, which is the code-word for human sacrifice in the S´a-kta-Bhairava milieu which this text emulates, recalling the ‘man-beast’ (narappacu) whose flesh the Tamil Bhairava-yogi demands in the Periyapura-n.am story that is ancestral to his Kerala Teyyam. The maha-pas´u in our paddhati, however, is redefined as being the deer of a Brahman, the elephant of a Ks.atriya, the buffalo of a Vais´ya, the goat of ‘others’, or, if these are unavailable, a buffalo that belongs to everyone. This maha-pas´u is bound to the sacrificial stake, slaughtered, its blood offered to the Mother or Mothers, and its blood and flesh mixed with rice and distributively offered as balis. Though the chicken does not appear here (as probably unworthy of this collective maha-s´a-nti rite), we can find here the logic for the earlier Brahmaya-mala‘s high ranking of deer in its coda-line, due to its association with Brahmans, and the general logic, throughout, for the intersecting ranks of size and social association of the other animals, with the human being latent in the maha-pas´u nomenclature, and the chicken as residually the lowest. I think we can also find a further social differentiation among the officiants implicit in the sequence and performance of the rites in this pat.ala. The pratis.t.ha- and initial worship I believe are done by Brahmans, which is all that would make sense of the stipulation that offerings are made in substitution (pratinidhi, pratiru-pa) explicitly for goat’s meat and other flesh, with the itemization of the vegetarian substitutes (ibid.: 572). Indeed, I believe the claim of this pat.ala to being a ‘purified’ (s´uddha-pas.katah. … vis´uddham) rendering of the Svacchanda- and Ma-tr.-tantras is not intended eulogistically, but carries this ritual significance of Brahmanical expurgation (ibid.: 567, 580). Even into the first part of the sanguinary s´a-nti-rite, a Brahman seems to be superintending the festival installations, with substitute-meat stipulated for his performing community-wide balis that sound very much like the modern Nampu-tiri

Medieval Malabar: S´a-ktism, polity, society

167

gra-ma-bali as a part of major festivals (ibid.: 578). The context shifts, however, with the deity’s setting out in full, royal grandeur, accompanied by warriors bearing the swords of Ka-lı- (khad.ga), for there we are told of the general body of the community festively engaged in the nightly slaughter of animals (ibid.: 578–579). It shifts again, however, with the appearance of the Pa-ras´ava, who appears with mudra-s and dance to take up his place in the festival man.d.apa (ibid.: 579). It is just after this that we get the rendition of the maha-pas´u rites, detailed above, and it seems quite likely that it is the Pa-ras´ava who superintends over or performs these bloody balis to the Mothers, just as we have seen was his mandate in the Brahmaya-mala, and still farther back in South Indian history.34 But what of the antecedents of the modern gurusi? Again Is´a-nas´iva’s text provides a context where this comes clearly into play. While I had come across the reference to raktodaka, ‘blood-water’, and suspected this might be gurusi, it is indeed confirmed in the section of this text dealing with the exorcism of bhu-tas. And there its Kerala recipe is explicitly given as combining lime (cu-rn.a) and turmeric (nis´a-) in water, to be offered as inducements and balis to feed and placate possessing demons (Gan.apati S´a-strı- 1921: vol. II: 303–304).35 The context here is the earlier differentiated mantrava-da (a kind of permanent way station between sa-dhana and pratis.t.ha-), into which the Nampu-tiris were drawn, and which they also extended in their officiation over the various enshrined goddesses who themselves first entered their purview as quasidemonic beings. In the earlier text of the Kurikkal. Teyyam, the Bhairava Yogi there cures the Ko-lattiri Ra-ja of his demons with actual blood-sacrifice, as well as gurusi, and he would have been the Brahman mantrava-din’s natural competitor in such ventures. The gurusi-tarpan.am I have seen as private worship by Nampu-tiris of their own lineage goddesses, plays out like an enactment of this exorcism from Is´a-nas´iva’s text. While the implications and correlations are too detailed for the present exposition, I am convinced that on both the domestic front, as well as in public fora, the S´aiva-S´a-kta mendicants and lineages that entered South India posed considerable challenges to the region’s Brahmans. The latter met these challenges in the texts and practices we find today, which are testament to their struggles, and artifacts of their multifaceted response. The ultimate triumph of an (admittedly transformed) Sma-rta orthodoxy further explains the survival of much of the original impetus and content of S´a-ktism in the folk-religion of Malabar. The S´es.asamuccaya and the cult of Da-ruka-Rurujit with which I started, records the effort to claim this S´a-kta territory for Brahmans, expurgating what was threatening as they developed their new, and now authoritative canon. To come full circle I may thereby close with an interesting apparent oversight made by those authors. In the listing of the dravyas to be offered to the very S´a-kta powers of the oldest Brahmaya-mala, Rakta-, Kara-lı-, Can.d.a-ks.-ı, and Mahocchus.ma-, the compiler of the Rurujit-vidha-na section of the S´es.asamuccaya seems, accidentally I imagine, to actually have enjoined the offering of meat.36 His commentator, probably his son or nephew, immediately made

168

Rich Freeman

good the lapse by explaining that in the place of meat, we are to understand the use of a substitute.37 In the modern Malayalam commentary, a Tantri Guru specifies that the substitute should be the steamed rice-cakes known in Malayalam as at.a (K. Diva-karan Nampu-tirippa-t.u 1977: 242). Representing the modern Nampu-tiri consensus on such matters, this substitute can be seen in Staal and Gardner’s film on the Agnicayana (1976), where the Tantric expurgation has carried over even to their traditional practice of the Veda. For lined up in their banana-leaf packets, with little leashes tied round them, are these same little rice-cakes, awaiting their ‘sacrifice’ in lieu of the textually mandated goats.

Notes 1 I follow Malayalam spellings of Sanskrit words in a Kerala context, so that I write Tantri, rather than Tantrin, or Tantrı- following Western or Indian Indological conventions. The latter seem artificial in these contexts, and there is little scope for confusion. 2 These are Brahma, Arka (Su-rya), Vais´ravan.a, and Kr.s.n.a. While the first three are merely installed as aniconic pı-t.has in temple compounds, the last, according to the Vimars´inı-, is qualified as indicating gos´a-la-sthah. kr.s.n.ah.. (Narayana Pillai 1951: 2nd part, p. 1). This would seem to refer specifically to the rather late intrusion of an ancillary, modular shrine to what is still locally called Gos´a-la-Kr.s.n.a, found within the outer walls of larger Kerala temples, such as at Peruvanam and Tr.praya-r in central Kerala. 3 I again follow the same passage of the Vimars´inı-, which explains the syntactic aggregation and grouping of the deities here, and in fact accords with their distribution and topical handling in the overall organization of the text and their ritual complexes. 4 One can get a sense of the complex internal status grades (a-bhija-tyam) of the Nampu-tiri formation from Padmanabha Menon (1984: 9–48), with the caveat that his is a textually formalized projection of the system reliant on such works as the . S´a-nkarasmr.ti and Keralotpatti. 5 I disagree, however, with Dumont’s ultimate analysis, which basically reads Brahman aspirations for power as distorting what should have been their ritual values. Dumont was, as are most historians and anthropologists, unaware of the social historical significance of tantric and S´a-kta movements in South India. 6 Early forays into this nexus as posing a major problem of historical development were raised by Brunner (1975, 1990), but it is in the introductions to Goodall’s aforementioned edited works (1998, 2004) and allied articles that he really brings text-historical clarity to these issues for the South. 7 The subject of tantric initiation in Kerala has recently been briefly surveyed by S. A. S. Sarma (2010), who argues that tantric initiation had devolved by the time of the Tantrasamuccaya from its earlier salvific purpose into a mere mantropades´a or mantradı-ks.a- for purposes of temple installation and worship. I have confirmed in my fieldwork cases of institutional Tantris still administering mantropades´a to students to ratify them for daily pu-ja-s, though this may be the recent revival of a tradition that had essentially lapsed with modernity. While Sarma’s brief observations seem to posit a soteriological rationale for this shift, a historically anthropological perspective would need to consider the social organizational and political implications of this change. 8 Cf. the brief discussion of the maha-s´a-nti for the Ma-tr. deities from Is´a-nas´iva’s paddhati, below.

Medieval Malabar: S´a-ktism, polity, society

169

9 As I discuss below, despite the obvious basic identity relation between the Rurujitvidha-na texts and contemporary practice, the social and ritual disjunctures that do exist between the two could suggest historical scenarios by which either one is earlier. 10 I say they performed such sacrifices ‘traditionally’ because current law would render such practices illegal, despite the fact that the majority of such Hindu lawmakers in Kerala are themselves non-vegetarian and may freely drink alcohol. When harassed with such meddlesome statutes there are various stop-gaps possible, such as presenting live chickens in the temple, but then taking them home for slaughter and preparation of the meat which is later offered and consumed as naivedya of the goddess. 11 I refer to these sections and the cultic complex as the Rurujit-vidha-na, following modern, local usage. There is, in fact, a discrete pu-ja-paddhati ms. titled Rurujidvidha-na to which the S´es.asamuccaya is closely tied, mentioned briefly, below. 12 The Vimars´inı- on S´es.asamuccaya 1.1 reads, rurujiditi ruruna-mno daityasya hantrıbhadraka-l.-ı. (Narayana Pillai 1951: part 2, p. 1). 13 There are two different texts under this name. I have seen copies of only one of the manuscripts (IFP T. no. 552), while the other (Trivandrum, possibly no. T. 982), a different text, I know only from extracts which Alexis Sanderson has kindly shared with me. 14 A brief rendition of the creation of Ekavı-ra- or Bhadraka-lı- and her killing of Da-ruka, opens the Brahmaya-mala (IFP ms. T. no. 552, p. 1), while the goddess finds mention in the Trivandrum ms. as Da-ruka-surana-s´inı-, according to the corrected reading supplied by Sanderson (from his extracts, p. 4), in the context of her identification as Ekavı-ra- or Ka-lara-trı-. - - rnnavasya tate s´ubhe 15 go-karn.n.am . nama pas´catyasya . . . .. ´ s´ankarasya maha-ks.etram s an karam . . sarvadehinam . (I.32, Tirumumpu 1975: 16) 16 hr.dayambujam utkhaya kandam vyadhavadhur iva pakva-‘ks.ipa-vake devı- bhaks.aya-ma-sa tanmuhuh. (8.41, ibid.: 130) 17 The verses in the sixth chapter, from 39–54 begin, respectively, with the syllables . . . ka, e, -ı, la, hrı-(m), ha, sa, ka, ha, la, hrı-(m), sa, ka, la, hrı-(m), and the sixteenth . adds s´rı(m) (pp. 88–92). 18 Matr.sadbha-va, Trivandrum ms. 792, p.93; Rurujidvidha-napu-ja-paddhati, ms. pp. 1–2. S´es.asamuccaya, mantras following 8.4. (Narayana Pillai 1951: 93–94). 19 it.ar ceyta ka-l.i matamo-t.u kaila-y-an.n.ayum . aran bhayattal iḻalot.u bhutakal.i karal.i balakal. kailanathan das´avuruva-ya kuḻavikal. kon.t.u matam avit.e tal.atti (Vis.n.u Nampu-tiri 1993) 20 Literally this means the grove (ka-vu) outside (-puṟattu) of Mannan (or the king), or of the mannam, the assembly ground. There is indeed a large sacred grove around three sides of the temple, and a small shrine outside the west (facing) side of the temple, which houses the goddess supposed to be original to Mannan on his palace-site. 21 I have confirmed these references in the Malayalam Lexicon, under the older variant, Piḻa-rar, which include a twelfth-century reference from the Sucı-ndram temple, and a fourteenth-century Man.iprava-l.am work, Un.n.iya-t.icaritam. The latter reference to upades´am is confirmed by the description of Pit.a-rar as the chiefs of Bhat.t.ar (Pat.t.ar), who imparted s´a-stric learning in temples. 22 A Kerala Sanskrit commentary was composed on it in the fourteenth century by Ra-ghava-nanda (Gan.apati Sa-strı- 1917), and two in Malayalam in the nineteenth century (C. N. Nampu-tiri 2004). 23 This is analogically confirmed, in meaning, variance of form and derivation from the old Tamil pit.a-r, meaning ‘arrogance, greatness’, and pat.a-rar, meaning ‘God …

170

24

25

26

27 28 29

30

31 32 33

34

Rich Freeman

venerable persons, as priests’, and their derivation from Sanskrit bhat.t.a-ra. (Madras University (Committee) 1982: vol. IV: 2435; vol. V: 2652). The element -at.i in their name is a common honorific suffix, referring to the foot (at.i) or sacred foot (tiruvat.i) of a deity. A popular etymology traces their name from Vis.aha-ri (‘poison-remover’) which is also applied to the goddesses, Pis.a-ri(kkal), mentioned above. While linguistically implausible, this nevertheless supports their co-association in folkloric memory. Most perplexing in the history of the Pis.a-rat.i is their designation ‘Vais.n.ava’ as a kind of alias for their caste-name in Sanskritic literary works, but whether this represents a period prior to their purported renunciant phase, or after their re-assimilation to the caste-order needs further investigation. For these details, I draw freely on my earlier work (Freeman 2006), though two corrections are necessary there. One concerns the early dating of Tirumantiram, in which I would now follow Goodall (1998: xxxvii–ix). The other concerns the location of Candragiri/pura, the original sites of which I have now more firmly . identified further north in the Konkan; I hypothesize, however, that the earlier identified site in South Kanara was linked with these, in the spread of Pas´cima-mna-ya movements down the West Coast. I have recently discovered textual evidence for these movements among my informants in Malabar. Well known in various Malayalam sources, an account of these events found their way into the Sanskrit work Udayavarmacarita (Sa-mbas´iva S´a-strı- 1938). While I cannot dwell on these fascinating genres here, representative selections can be found in Vis.n.u Nampu-tiri (1998, selections in 2005). There are a number of historical problems I have, from the Kerala perspective, on understanding how and why the Prapañcasa-ra, if an originally Orissan cultural production, would end up so intimately embraced on the far West Coast (and I . don’t think the attribution of it to S´ankara alone would carry much clout). While Sanderson’s case for the Orissan influence is compelling (though more so in the case of the S´a-rada-tilaka), my own inclinations are to look to the Co-ḻas and per. haps Co-d.a-Gangas as complicating these matters, but I cannot pursue these arguments here at present. While the dhya-na-s´loka for S´iva is given (7.66), there is none for Sarasvatı- in the published Tantrasamuccaya’s treatment of lipinya-sa. Since the r.s.i, chandas, and devata- are given for her there, however (5.20), the Nampu-tiri vernacular traditions all supply this dhya-na, as found in Prapañcasa-ra (7.3). S´iva’s dhya-na in the latter is at 27.41. The correlations between S´es.asamuccaya and Prapañcasa-ra are, respectively: for Su-rya (3.99 = 15.40); for Sarasvatı- (3.102 = 8.41); for S´rı- (3.103 = 12.4); and for Pa-rvatı- (3.104 = 10.10), with some slight changes. .. mumpina-l a-l.e vet.t.i ampina-l co-ra tanne-n. impamo-t.' a-l.e-k-kit.t.a-ññappo-ḻ inn' a-na vet.t.i. -aleyum a-neyum kitta-ññ' a-ttine aṟuttu. pinne aneyum a-leyum a-tum e-kavum kitta-ñ. . . .. .. . . . .. ñippo-l. ko-ḻiye aṟuttum . kon.t.u coraye-t-tarunnen innu (Cantera 1973). - - asura-n.a-m . vinas´artham . matr.̄ n.am udbhavam . kramat - sarvabhu-{ta}hitam bhavet ma-tr.̄ n.a-m s´a-ntikam kr tva . . . uttama-madhyamadhama trividham . pas´ur ucyate hastıhayanaram hatva uttamantu vinirdis´et madhyam . mahis.am evos.t.ram . tv adhamam . tvajasukaraih. harin.am uttamañcaiva kukkut.añca-dhamam . bhavet (No. T522: 91–92) It is quite significant that the early text that enjoins the employment of Pa-ras´avas for Bhadraka-lı-‘s worship (Caland 1982: 226–227), as Sanderson has pointed out, is unambiguously South Indian; its very syntax suggests its origins in Tamil Nadu (ibid.: xiv–xv).

Medieval Malabar: S´a-ktism, polity, society

171

35 I believe it may be historically significant that there are many synonyms deployed for the Sanskrit term (rakta-udaka, -uda, -ambhas, -toya, -jala), vs. the unitary term, and substance in Malayalam. - 36 mu-lam . madhye rakta hr.dayakaral.yau s´ikha ca can.d.aks.ı kavacam . ca mahocchus.ma - -. s´iras´ ca pu-rva-didiks.u devya-m . syuh. samidajyannasiddharthamamsany atha tila yavah. dravya-n.i vrı-hayas´ ca-jyam . sarvatras.t.as´atam . hutih. (Narayana Pillai 1951: 88). . 37 ma-msastha-ne pratinidher eva gra-hyah. (Vimars´inı- on 7.92–94, p. 65).

References Manuscripts Brahmaya-mala, Institut français de Pondichéry, no. T522. Devana-garı- script. Brahmaya-mala, Trivandrum University Manuscript Library, no. 11170(?) or 1982(?); from extracts supplied by Alexis Sanderson. Ma-tr.sadbha-va, Institut français de Pondichéry, no. 792; Copied from Trivandrum University Manuscript Library, no. 1642. Devana-garı- script. Ma-tr.sadbha-va, Institut français de Pondichéry, Government Oriental Manuscript Library, Madras, no. R5126/SR5623. Grantha script. Rurujidvidha-napu-ja-paddhati, Institut français de Pondichéry, Copied from Government Oriental Manuscript Library, Madras, no. R3365/SR1865. Grantha script.

Books and articles Brunner, H. 1975. ‘Le sadhaka, personage oublié du sivaisme du Sud’. Journal Asiatique, 263: 411–443. Brunner, H. 1990. “Atma-rthapu-ja- Versus Para-rthapu-ja- in the S´aiva Tradition”. In The Sanskrit Tradition and Tantrism, ed. Teun Goudriaan, pp. 4–24. Panels of the VIIth World Sanskrit Conference, Gen. J. Bronkhorst. Leiden: E. J. Brill. Bühnemann, G. 1988. Pu-ja-: A Study in Sma-rta Ritual. Vienna: Institut für Indologie der Universität Wien, Sammlung De Nobili. Caland, W., trans. 1982 (rpt). Vaikhanasasmartasutram: The Domestic Rules and Sacred Laws of the Vaikhanasa School Belonging to the Black Yajurveda. New Delhi: Ramand Vidya Bhawan. Cante-ra. C.M.S. 1973. Kan.n.akiyum Cı-rmmakka-vum. Kottayam: The Author; distributed by National Book Stall. Cante-ra. C.M.S. 1982. Uttara Ke-ral.attile S´a-kte-yakka-vukal.. Cannanore: A. V. Sanjeevan. Champakalakshmi, R. 2011. Religion, Tradition, and Ideology: Pre-colonial South India. New Delhi: Oxford University Press. Diva-karan Namu-tirippa-t.u, K., ed. 1977. S´e-s.asamuccayam . (Commentary). Covvannur: Tantravidya-pı-t.ham. Diva-karan Nampu-tirippa-t.u, K., ed. 1986. Kuḻika-t.t.u Pacca (Tantra Grantham), Compiled . by D. Subra-ya Tantri. 3rd edition. Kunnamkul.am, Kerala: Pañca-ngam Press. Dumont, L. 1983. “Nayar Marriages as Indian Facts”. In Affinity as a Value: Marriage Alliance in South India, with Comparative Essays on Australia, ed. Louis Dumont, pp. 105–144. Chicago, IL: University of Chicago Press. Freeman, R. 1991. ‘Purity and Violence: Sacred Power in the Teyyam Worship of Malabar’. Ph.D. dissertation, Dept. of Anthropology, University of Pennsylvania. Ann Arbor, MI: University Microfilms Incorporated.

172

Rich Freeman

Freeman, J. R. 1998. ‘Rubies and Coral: A Lapidary Language for Kerala’. Journal of Asian Studies, 57(1): 38–65. Freeman, J. R. 2003a. ‘Hindu Literature in Malayalam’. In A Companion to Hinduism. Gavin Flood, ed. Oxford: Basil Blackwell. Freeman, J. R. 2003b. “Genre and Society: The Literary Culture of Pre-Modern Kerala”. In Literary Cultures in History: Reconstructions from South Asia, ed. Sheldon Pollock. Berkeley, CA: University of California Press. Freeman, J. R. 2006. “Shifting Forms of the Wandering Yogi: The Teyyam of Bhairavan”. In Masked Ritual and Performance in South India, eds, David Shulman and Deborah Thiagarajan, pp. 147–183. Ann Arbor, MI: Centers for South and Southeast Asian Studies, University of Michigan Press. Freeman, J. R. 2009. “Untouchable Bodies of Knowledge in the Spirit Possession of Malabar.” In The Body in India: Ritual, Transgression, Performativity, eds, Axel Michaels and Christoph Wulf, pp. 135–164. Paragrana 18(1). Reprinted as Images of the Body in India. New Delhi: Routledge 2010. Freeman, J. R. 2011. ‘Artful Selves in a Libidinal Economy: The Cult of Courtesans and Inter-caste Connubium in Kerala’. History and Anthropology Colloquium, February 14, Duke University. Gan.apati Sa-strı-, T., ed. 1917. The Laghustuti of S´rı- Laghubhat.t.a-raka with the Commentary of S´rı- Ra-ghava-nanda. Trivandrum Sanskrit Series, no. 60. Trivandrum: Government Press. Gan.apati Sa-strı-, T., ed. 1920–22, 1925. Is´a-nas´ivagurudeva Paddhati of Is´a-nas´iva Gurudeva. Trivandrum Sanskrit Series, vols. 69, 72, 77, 83. Trivandrum: Government Press. Rpt. 1990 with Introduction by N. P. Unni. Delhi: Bharatiya Vidya Prakasan. Goodall, D., ed. and trans. 1998. Kiran.avr.ttih/Bhat.t.ara-makan.t.haviracita- = Bhat.t.a Ra-makan.t.ha’s commentary on the Kiran.atantra. Pondicherry: Institut française de Pondichéry, École française d’Extrême-Orient. Goodall, D. ed. and trans. 2004. Para-khyatantra, A Scripture of the S´aiva Siddha-nta. Pondicherry: Institut français de Pondichéry, École française d’ExtrêmeOrient. Goudriaan, T. 1977. ‘Khad.ga-Ra-van.a and his Worship in Balinese and Indian Tantric Sources’. Wiener Zeitschrift für die Kunde Südasiens, 21: 143–169. Hatley, S. 2007. ‘The Brahmaya-malatantra and Early S´aiva Cult of Yoginı-s’. Ph.D. Dissertation, Department of Religious Studies, University of Pennsylvania. Ann Arbor, MI: University Microfilms Incorporated, # 3292099. Heitzman, J. 1997. Gifts of Power: Lordship in an Early Indian. Delhi and New York: Oxford University Press. . . Lingamaha-pura-n.a (with the T.-ıka- S´ivatos.in.-ı) 1906 [1857], Mumbai: Venkates´vara Steam Press. Madras University (Committee). 1982. The Tamil Lexicon. Vols. IV and V. Madras: University of Madras. . . - Nampu-tiri, C.N., ed. 2004. Ba-la-vims´ati athava- Tripura-sto-tram . Vims´ati (Vyakyanadvayasametam). Kodungallur: Devi Book Stall. Narayana Pillai, P.K., ed. 1951. S´es.asamuccaya. Trivandrum Sanskrit Series, no. 166 Trivandrum: Government Press. Padmanabha Menon, K.P. 1984. History of Kerala Written in the Form of Notes on Visscher’s Letters from Malabar, Volume 3. T. K. Krishna Menon, ed. New Delhi: Asian Educational Services.

Medieval Malabar: S´a-ktism, polity, society

173

Ramaswami Sastri, V. et al., eds 1945–62. Tantrasamuccaya of Na-ra-yan.a with the . Commentaries Vimars´ini of S´ankara and Vivaran.a of Na-r.a-yan.as´is.ya. Trivandrum Sanskrit Series, nos. 151, 169, and 200. Trivandrum: Government Press. Sa-mbas´iva S´a-strı-, K., ed. 1938. The Udayavarmacarita. Trivandrum Sanskrit Series, no. 133. Trivandrum: Government Press. Sanderson, A. 1988. “S´aivism and the Tantric Traditions”. In The World’s Religions, eds, S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp. 660–704. London: Routledge and Kegan Paul. Sanderson, A. 2007a. “The S´aiva Exegesis of Kashmir”. In Mélanges Tantriques à la Mémoire d’Hélène Brunner, eds, Dominic Goodall and André Padoux, pp. 231–442; pp. 551–582. Pondicherry: Institut français d’Indologie/École française d’Extrême-Orient. Sanderson, A. 2007b. “Atharvavedins in Tantric Territory: The Angirasakalpa Texts of the Oriya Paippaladins and their Connection with the Trika and the Ka-lı-kula, with Critical Editions of the Para-japavidhi, the Para-mantravidhi, and the *Bhadraka-lı-mantravidhiprakarana”. In The Atharvaveda and its Paippala-da S´a-kha-: Historical and Philological Papers on a Vedic Tradition, eds, Arlo Griffiths and Annette Schmiedchen, pp. 195–311. Aachen: Shaker Verlag. Sanderson, A. 2009. “The S´aiva Age: The Rise and Dominance of S´aivism During the Early Medieval Period”. In Genesis and Development of Tantrism, ed. Shingo Einoo, pp. 41–350. Tokyo: Institute of Oriental Culture, University of Tokyo. Sarma, S.A.S. 2010. ‘Why Have the Later Ritual Manuals of Kerala forgotten “Initiation”?’ Acta Orientalia: Acadiae Scientarium Hungaricae, 63(4): 443–454. Soudararajan, J. 2004. Nat.ara-ja in South Indian Art. Delhi: Sharada Publishing House. Srinivasan, P.R. 1986. Annual Report on Indian Epigraphy for 1975–76. New Delhi: Archaeological Survey of India. Staal, F. and R. Gardner, eds 1976. Altar of Fire [film]. Berkeley, CA: University of California Extension Media Center. Staal, Frits. 1984. Agni: The Vedic Ritual of the Fire Altar. Delhi: Motilal Banarsidass. Stein, B. 1994. Peasant, State, and Society in Medieval South India. Delhi: Oxford University Press. Tirumumpu, T.S. 1975. S´rı- Bhadraka-l.-ı Ma-ha-tmyam athava- Da-rukavadham (With Text and Malayalam Translation). Trikkarpur, Kerala: The Author. Unni, N.P. 2006. Tantra literature of Kerala. Delhi: New Bharatiya Book Co. Venugopalan, P.V., ed. 2009. Malayalam Lexicon. Volume VIII. Thiruvananthapuram: University of Kerala. Vis.n.u Nampu-tiri, M.V., ed. 1981. To-ṟṟampa-t.t.ukal.. Kottayam, Kerala: National Book Stall. Vis.n.u Nampu-tiri, M.V., ed. 1993. Tiruvarka-t.t.u Bhagavatitto-ṟṟam (To-ṟṟampa-t.t.u). Kottayam, Kerala: National Book Stall. Vis.n.u Nampu-tiri, M.V., ed. 1998. Pu-rakkal.i. Kottayam: Current Books. Vis.n.u Nampu-tiri, M.V., ed. 2005. Nammut.e Pan.t.attepa-t.t.ukal.. Kottayam: DC Books. Distributors, Current Books.

Index

Aarhus University 3, 4, 14n8, 44n1 Abhinavagupta 5, 8, 52, 59, 64n5 Adriaensen, Rob, Bakker, T. Hans and Isaacson, Harunaga 14n2 Aghora- 7, 33, 44, 45n5 Aghorı- (undreadful) in early S´a-kta tantras 33–44; Aghora-mantra 37, 45n22; Aghora-mantra, Rudra and 34–5, 35–6, 44; Aghores´-ı 38; as supreme goddess 39–43; Aghores´varı- 39–40; Aghorı-, roles in rites in cremation grounds 41–2; aghorya-dyas.t.aka 33; Aindrı- 38, 43; Apara-, goddesses of 35; Bhairavatantras 34, 35, 37; Bhairavı39, 43; bestowal of knowledge by 40–41; Bhı-ma- (terrifying) 37; Bhı-s.an.a(frightening) 37; Brahma-n.-ı 38, 43; Brahmaya-mala 38, 40–41, 43; Ca-mun.d.a- 38; Can.d.a- Ka-pa-linı- 39, 40, 41, 42–3; cremation grounds, Aghorıand roles in rites in 41–2; Devı-pañcas´atika- 35; Ghoraru-pa- (dreadful form) 37; goddesses, Aghorı- in groups of eight 36–9; Great Churning (maha-mantha-na), rite of 41–2; Guhyas´akti 43; iccha- (volition) 35; Jayadrathaya-mala 35; Jayaratha 33; jña-na (knowledge) 35; ka-pa-likas 41, 44; Kauma-rı- 38, 43; kriya- (action) 35; Ks.emara-ja 35–6; Kubjika-mata 37, 40; Kularatnoddyota 37; La-kulas 35, 36; Magic Pavilion (siddhaman.d.apika-), rite of 41; Ma-hes´-ı 38, 43; Ma-latı-ma-dhava 44; Ma-linı-vijayottara 34, 35, 37, 38; Nava-tman mantra 40; Netratantra 35; Nis´va-satattvasam . hita 40; Para, goddesses of 35; Paramaghora(extremely dreadful) 37; Para-para-, goddesses of 35, 37; Piba[nı-]

(drinking/withdrawing) 37; Prama-n.apa-s´upatas 35, 36; ra-jasa 35; Rudra, Power of 35; S´aiva Siddha-nta 37; sa-ttvika 35; seven Mothers, cult of 39, 42, 44; Siddhayoges´varı-mata 33, 34–5, 36, 37, 38, 39; Ska-ndı- 38; Svacchandabhairava 34, 37, 38–9; ta-masa 35; Tantra-loka of Abhinavagupta 33, 38; Tantrasadbha-va 37, 38, 40; triad of ghora, ghoraghoratara and aghora 34–6; Trika, goddesses of 35, 37–8; Vais.n.avı- 38, 43; Vama[nı-] (vomiting/ creating) 37; Va-ra-hı- 38; Worship of Dieties in a Hole (garta-ya-ga), rite of 41; Yama-tmika- 38; Ya-mya- 38, 43; yantra diagrams (yantra-rn.ava), use of 41; Yoges´varı- 38; yoginı-s, doctrine of 39; Yoges´-ı 38, 43 Agnicayana (film by Staal and Gardner) 168 Agni-pura-n.a 9, 77, 80, 81, 82, 83, 91n15, 92n23 Agnipura-n.a passages dependent on Tvarita-mu-lasu-tra 89–82 Aindrı- 38, 43 Aiyangar, M.D. 93 Ajı-vikas 120, 132n102 Aktor, Mikael vii, 4, 9–10, 96–108 Amanaska 116, 125n9, 127n33, 130n63, 133–4n113 Amaraughaprabodha 113, 117 Amr.tasiddhi 11; hat.hayoga, relationship with S´a-ktism 111–12, 114, 121 antisectarianism 123 Apara-, goddesses of 35 Artima-rga 21 Arya- 17 Arya-va-rta 101 asidha-ravrata 121

Index Atima-rga ascetics 121 Bah.r al-H . ayat 123 Bakker, T.H. 45n4 Bakker, T.H. and Isaacson, H. 14n2 Bhadraka-li 144, 146, 154 Bhagavadgı-ta- 116, 119, 132n92 Bhairava 7, 8, 22, 75, 115, 121, 142, 160, 166; Aghorı- in early S´a-kta tantras 34, 37, 40–41, 42, 43, 44, 45n5; Jayadrathaya-mala, varieties of melaka in 51, 54, 56, 57–8, 60–61, 62, 63, 68–9n51, 68n50, 69–70n61; Tantras 22, 24–5; Yogis 157–9, 161, 167 Bhairavantantras: Aghorı- (undreadful) in early S´a-kta tantras 34, 35, 37; S´akti in early tantric S´aivism 22, 23–4 Bhairavı- 39, 43; bestowal of knowledge by 40–41 bhakti movement 148 Bhatt, V.P. 94 Bhattacaryya, J.V. 93 Bhattacharya, B. 93 Bhattacharyya, N.N. 96 Bhattarai, K. 92n36, 93 Bhat.t.a-raka- 156 Bherun.d.a- 9, 88, 90; features in early references 88–9; S´a-kta identities in 88 Bhı-ma- (terrifying) 37 Bhı-s.an.a- (frightening) 37 bhoga 10, 103, 106n11 Bhu-tana-yika- 17 bhu-tı- 24 bindu 110–12, 113, 114, 118, 121, 123, 125n10, 126n22, 128n43, 131n77, 134–5n119 bindudha-ran.a 11; hat.hayoga, relationship with S´a-ktism 114, 117, 119, 120, 121, 128n43 Birch, Jason 124n1, 125n8, 125n10, 130n63, 134n115 Bisschop, Peter 14n3, 27n46 blood-rites, artifice and expurgation in 164–8 Bouillier, Véronique 124–5n7, 124n1, 124n6, 135n127 Bouy, Christian 124n3, 124n4, 125n11 Brahman mantrava-da, yogis and popular religion: medieval Malabar, polity, society and S´a-ktism in 157–61 Brahman.-ı: Aghorı- (undreadful) in early S´a-kta tantras 38, 43 Brahmaveta-linı-: snakebite goddesses in S´a-kta traditions 78

175

Brahmaya-mala: Aghorı- (undreadful) in early S´a-kta tantras 38, 40–41, 43 Brahmaya-mala 4, 6, 7, 8; hat.hayoga, relationship with S´a-ktism 116, 121; Jayadrathaya-mala, varieties of melaka in 52; medieval Malabar, polity, society and S´a-ktism in 151, 152, 155, 156, 158, 162, 164, 165–6, 167m 169n14; S´akti in early tantric S´aivism 16, 19, 24, 26n18, 27n39, 28n49, 28n50, 28n51 Brahmin communities 96–9, 102, 104, 107n14; Kerala Brahmins 13; see also Nampu-tiri Brahmins Brahmin values 10; Can.d.a-la as antithesis of 99 Braj Bha-s.a- Jogpradı-paka- 123, 134n116 Br.hadyonitantra 10, 102–3 Bronkhorst, Johannes 120, 132n94 Brooks, Douglas Renfrew 1 Brunner-Lachaux, Hélène 21, 168n6 Buddhist Ja-taka stories 10 Buddhist Kurukulla- 87–8 Buescher, Hartmut 107n12 Bühnemann, Gudrun 4, 14n9, 82, 92n25, 130n72, 132–3n103, 143 Burchett, Patton 124n1 Cakravega- 78 Caland, W. 132–3n103, 170n34 Callewaert, W.M. and de Beeck, B. 135n128 Ca-mun.d.a-: Aghorı- (undreadful) in early S´a-kta tantras 38; medieval Malabar, polity, society and S´a-ktism in 144, 146, 155; S´akti in early tantric S´aivism 22 Can.d.a- Ka-pa-linı- 39, 40, 41, 42–3 Ca-n.d.a-lı- as S´akti 9–10, 96–105; Arya-varta 101; Brahmin communities 96; Brahmin values, Can.d.a-la as antithesis of 99; Br.hadyonitantra 102–3; Can.d.a-la women, attractiveness of 99–100; caste labels 97; Cha-ndogyaupanis.ad 98, 105; chief goddess, Can.d.a-lı- as lady of hosts 96–8; Cittasambhu-taja-taka 98; D . ombı of the Caryagıti, mystical experience 103–5; D . ombı washerwoman 96, 105; impurity 100–101; Ka-dambarı(Ba-n.abhat.t.a) 99; Ka-n.ha (Skull Bearer) 104; kun.d.alinı-s´akti, personification of 96; level of society 100–102; low-caste women 97; Ma-dhava-ca-rya 103; Maha-bha-rata 98–9; Maha-ya-na 104;

176

Index

menstruation 102–3; Mr.tapas, Can.d.a-las and 101; Munidatta 104, 105; outcaste woman, Can.d.a-la- as 98–100; Pali Canon 104–5; Pa-n.ini 101; Para-s´arama-dhavı-ya 100; Para-s´arasmr.ti 103; rajakatantuva-yam 101; Satadhammaja-taka 102; Setaketuja-taka 102; siddhas 104; su-dhakas 96–7; King S´u-draka 99, 100, 105; taks.a-yaska-ran 101; untouchability, physical contact and 105; untouchable women, power of 97–8; varn.as 97–8, 99; yoga-bhoga dichotomy 103, 106n11; yonipu-ja- 97, 106n11; Yonitantra 96–7, 98, 102, 105, 106n11; yonitattva 97, 105 Can.d.ayoges´varı- 154 Candra-valokana 116 Carpat.i 119 Carya-gı-ti 10, 107n12; D . ombı of the, mystical experience 103–5 caste and political boundaries, mediation across 154 caste labels 97 caste reformism 147–8 Caturvim s´atisahasrasam . hita 75 .. Caurangı- 119 Celextel.com 84, 92n32 Champakalakshmi, R. 141 Cha-ndogyaupanis.ad 10, 98, 105 Cı-llasam . yogamelaka 58 cilla-vrata 57 Cir.r.ampalam, affiliations with shrine at 158, 161 Cittasambhu-taja-taka 98 Coburn, Thomas 1 cosmology, theology and S´akti 18–20 Couture, André 27n34 cremation grounds, Aghorı- and roles in rites in 41–2 d.a-kinı-s 60–61 Da-ruka 152, 153, 154, 162, 165, 167, 169n14 Da-s, Da-du-panthı- Ra-ghav 134n116 Da-s, Sundar 123, 134n116 - Das´ana-mı- Sam . nyasıs 119, 121 Dattatreya 12, 133n105 Datta-treyayogas´a-stra 4, 11; hat.hayoga, relationship with S´a-ktism 114, 119, 120, 121 Dave, K.N. 88 democratisation of yoga 122–3 Das´ana-mı-s 12 Devı- 3, 97, 153; worship of 4

Devı- Da-s Ka-yastha 123 Devı-ma-ha-tmya 6; medieval Malabar, polity, society and S´a-ktism in 152; S´akti in early tantric S´aivism 17 Devı-pañcas´atika- 35 Dezsö, Csaba 24, 29n53 Dhu-nanı- 78 Dighambara Jainism 5 Diva-karan Nampu-titippa-t.u, K, 168 - D . ombı 9; of the Caryagıti, mystical experience 103–5; washerwoman 96, 105 Dumont, Louis 9, 10, 13, 98, 106n11, 147, 168n5 Durga- 9; in mantric worship 144; S´akti in early tantric S´aivism 17; snakebite goddesses in S´a-kta traditions 75, 76 Dutt, M.N. 82 Dyczkowski, Mark 75, 76, 82–3, 90n6, 90n8, 91n12, 92n26 Ernst, Carl W. 135n122 ethnological reflections, medieval Malabar 146–51 feminine principles, goddesses as apotheoses of 19–20 Fibiger, Marianne 14n9 Flood, Gavin vii, 1, 2–3, 4, 44n1 Freeman, John Richardson (Rich) vii, 4, 12–13, 141–73 Gan.ana-yika- 17 Gan.apati 144–5, 146 Gan.apati Sa-strı-, T. 166–7, 169n22 Ga-rud.a Tantras 74, 75, 82, 88 Gaurı- 75–6, 90–91n10, 133n106 Ghoraru-pa- (dreadful form) 37 Guru Gobind Singh 131n85 goddesses: Aghores´-ı as supreme goddess 39–43; Aghorı- in groups of eight 36–9; Apara-, goddesses of 35; chief goddess, Can.d.a-lı- as lady of hosts 96–8; feminine principles, goddesses as apotheoses of 19–20; Great Goddess tradition 1, 2, 6, 17–18; Kubjika-, praise for goddess as 153; mothergoddesses 22–3; Para-, goddesses of 35; Para-para-, goddesses of 35, 37; polity and Puranic S´a-kta goddeses 152–5; social aspects of goddess myth 153; S´rıvidya-, praise for goddess as 153; Trika, goddesses of 35, 37–8; Tripurasundarı-,

Index praise for goddess as 153; Va-gı-s´varı-, Goddess of Speech 18, 20 Golovkova, Anna 128n37 Goodall, D. et al. 6, 16, 21, 25n2, 26n16, 26n17, 26n21, 26n25, 26n28, 29n53 Goodall, Dominic 14n4, 16, 18, 20, 21, 25n1, 25n4, 26n11, 26n14, 27n29, 45n3, 135n121, 148, 168n6, 170n26 Goraks.a 12, 109, 112, 113, 119, 124, 130n57, 135n124 Goraks.as´ataka 11–12; hat.hayoga, relationship with S´a-ktism 112, 114, 123 Goudriaan, T. and Schoterman, J.A. 14n6 Goudriaan, Teun 69–70n61, 143–4 Great Churning (maha-mantha-na), rite of 41–2 Great Goddess tradition 1, 2, 6, 17–18 Great Power (maha-s´akti) 19 Guhyas´akti 43 Guhyasu-tra 6; S´akti in early tantric S´aivism 16, 17, 18, 19–20, 21, 22–3, 24–5 Gupta-era 17 Gurukkal., offence of 160–61 Halbfass, Wilhelm 124n3 Harivam . s´a 17, 23 Hars.acarita 24 hat.hamelaka 8; hat.hamelaka-related transformation, reflection on 62–4; human sacrifice and 62–3; Jayadrathaya-mala, varieties of melaka in 52–5, 58–9, 59–60 Hat.hapradı-pika- 4, 11, 12; hat.hayoga, relationship with S´a-ktism 110, 114, 116–17, 117–19; S´a-ktism, hat.hayoga and 110, 111, 112, 114, 116–19, 124, 125n9, 126n16, 127–8n36, 127n25, 127n29, 127n33, 128n43, 130–31n74, 130n65, 131n78, 131n81, 131n86, 132n87, 134n116, 134n118, 135n125 hat.hayoga, ascetic traditions and 121–2 hat.hayoga, relationship with S´a-ktism 10–12, 109–24; accessibility of hat.hayoga texts 122; a-jı-vikas 120, 132n102; Amanaska 116, 125n9, 127n33, 130n63, 133–4n113; Amaraughaprabodha 113, 117; Amr.tasiddhi 111–12, 114, 121; antisectarianism 123; asidha-ravrata 121; Atima-rga ascetics 121; Bah.r al-H . ayat 123; Bhagavadgıta 119; bindu 110–12, 113, 114, 118, 121, 123, 125n10, 126n22, 128n43, 131n77, 134–5n119; bindudha-ran.a 114, 117, 119, 120, 121,

177

128n43; Brahmaya-mala 116, 121; Braj Bha-s.a- Jogpradı-paka- 123, 134n116; Candra-valokana 116; Carpat.i 119; . Caurangı- 119; classical hat.hayoga of the Hat.hapradı-pika- 117–19; codification of physical techniques 122; Das´a- na-mı- Sam . nyasıs 119, 121; Datta-treyayogas´a-stra 114, 119, 120, 121; definition of hat.hayoga 110, 125n9, 125n10; democratisation of yoga 122–3; development of 123–4; Devı- Da-s Ka-yastha 123; early hat.hayoga 110–11; Goraks.a 119; Goraks.as´ataka 112, 114, 123; Hat.hapradıpika- 110, 114, 116–17, 117–19; hat.hayoga, ascetic traditions and 121–2; hat.hayoga, scholarship on 109; Jña-nes´varı- teachings 116, 123; Ka-lacakratantra 110; Ka-maru-pa 110; Kapila 119, 120; khecarı-mudra- 120; Khecarı-vidya- 113, 117; Khula-s.ah alKhula-s.ah 123; Kr.s.n.advaipa-yana 117; Kubjika-mata 116; Kula-rn.avatantra 116; Kun.d.alinı-, incorporation into practice of 117, 118, 120, 123; Layayoga 115–17, 121–2; maha-bandha 120; Maha-bha-rata 120; maha-mudra120, 121; maha-vedha 120; Mantrama-rga traditions 121; Mantrayoga 115, 130n72; Ma-rkan.d.eya 119; Matsyendra 119; Matsyendrasam . hita 116, 118; mudra-s, names of 121; mu-labandha 120; muni tradition of hat.hayoga 120–22; munis 119–22; Nath adoption of hat.hayoga 124; Na-th gurus 119; Na-th S´a-kta credentials 109–10; Na-th sam . pradaya 109; Pali Canon 120; Pa-s´upatas 121; Pun.d.arı-ka 110; Ra-ma-nandı- Jayatara-ma 119, 121, 123; S´a-kta Na-th Yogı-s 124; . S´a-rngadharapaddhati 117, 119; Sarvan.gayogapradı-pika- 123; siddhas 119, 124; S´ivasam . hita 113–14, 118, 120, 121, 122; Sva-tma-ra-ma 118–19; ukkut.ikappadha-na 120; universalism 123; Vajra-sana 120; vajrolimudra- 120–21; Varn.aratna-kara 119; Vimalaprabha110; viparı-takaran.-ı mudra- 120; Viru-pa-ks.a 119; Vivekama-rtan.d.a 112, 114, 119; Ya-jñavalkya 120; Yogabı-ja 112–13, 114; Yogasu-tra 121–2 hat.hayoga, scholarship on 109 Hatley, Shaman vii–viii, 4, 6, 133n108, 133n111, 152, 158; Jayadrathaya-mala,

178

Index

varieties of melaka in 52, 64n7, 64n8, 69n56; S´akti in early tantric S´aivism 16–32, 44n1, 45n5, 46n34m 46n35, 46n36, 46n39, 47–8n58, 47n41, 47n47, 47n48, 47n50, 47n51, 47n54 Heilijgers-Seelen, Dory 14n6 Heitzman, J. 141 Hemacandra 88 iccha- (volition) 35 iccha-s´akti 21 impurity 10, 12, 100–101, 105, 164 Indian polity and religious institutions, intersection of 141–2 Indra 144 inscriptional record, engagement with 141 installation (pratis.t.ha-) rites 146, 148, 150, 160–61, 163, 166, 167 interconnections between Sanskrit and vernacular texts 141 interdisciplinarity 4, 141 Isaacson, Harunaga 4, 14n9, 25n1, 25n4, 28n49, 44n1 Is´a-nas´ivagurudevapaddhati 13, 82, 83, 143; medieval Malabar, polity, society and S´a-ktism in 143, 144, 166, 167 Is´vara, four kala-s of Lord 19 Jaina Prakrit Vasudevahin.d.i 24 Jaina Tantras 9; snakebite goddesses and 75 Jayadrathaya-mala 4, 7–8; Aghorı(undreadful) in early S´a-kta tantras 35; contents of, description of 51–2 Jayadrathaya-mala, variaties of melaka in 51–64; Abhinavagupta 52; analytic conclusions 58–9; Bhairava 51, 54, 56, 57–8, 60–61, 62, 63, 68–9n51, 68n50, 69–70n61; Brahmaya-mala 52; Cı-llasa- m . yogamelaka 58; cillavrata 57; d.akinıs 60–61; hat.hamelaka 52–5, 58–9, 59–60; human sacrifice and 62–3; hat.hamelaka-related transformation, reflection on 62–4; Kulacakres´varı-, worship of 62; Kula-rn.avatantra 60; kun.d.alinı--related procedures 59; La-lı- 57; Maha-hum . karin.ıcakra 53–4; mahamelapasiddhi 57; melaka, patterns of 51; melaka as visible ritual described in Vidya-pı-t.ha texts 59–62; melakamanual in Jayadrathaya-mala 56–8; Mela-paka-lı- 58; Netra Tantra 63; preliminary linguistic analysis 52–8;

priyamelaka 55–6, 57, 58–9, 59–60; ritual dimensions of priyamelaka and hat.hamelaka 61; sa-dhaka 51, 52, 53–4, 55, 57–8, 59, 60, 61–2, 63, 64n3; Tantra-loka 52; Tantrasadbha-va 56; unmatta vrata 57; unusual varieties of mela-pa 58; Vidya-pı-t.ha Tantras 51, 52, 63; vı-ramelaka 61; yoganidra- 57, 68n47; yoginı-s, classes of 57–8; yoginı-s, encounters with 51; yoginı-s, secret signs for recognition of 60; yoginı-s, supernatural capacities of 56–7; yoginı-s, tantric feasting and role of 60–62 Jayaratha 33 jña-na (knowledge) 35 Jña-nas´akti 21 Jña-nes´varı- teachings 116, 123 Joshi, K.L. 82, 92n23, 93 Jva-linı- 78 jyes.t.ha-: medieval Malabar, polity, society and S´a-ktism in 144; S´akti in early tantric S´aivism 21 Ka-dambarı- (Ba-n.abhat.t.a) 10, 99 Ka-dambarı- by Ba-n.abhat.t.a and His Son (Ka-le, M.R.) 99, 100, 106n5, 106n6 Kafle, Nirajan 16 Ka-lacakratantra 110 Ka-lara-tri 17 Kal.ariva-tukkal 150 Ka-lı- 9; medieval Malabar, polity, society and S´a-ktism in 162, 167; S´akti in early tantric S´aivism 22; snakebite goddesses in S´a-kta traditions 75 Kalı-kula sources 9 Ka-ma-ks.-ı 156 Ka-maru-pa 110 Kane, P.V. 133n105 Ka-n.ha (Skull Bearer) 104 ka-pa-likas: Aghorı- (undreadful) in early S´a-kta tantras 41, 44; atimargic sources 5; human sacrifice 7; ka-pa-lika ascetics 7 Kapila- 12; hat.hayoga, relationship with S´a-ktism 119, 120; snakebite goddesses in S´a-kta traditions 78 Kara-lı- 78 Ka-tya-yanı- 17 Kaula S´aivism 11 Kauma-rı- 38, 43 khecarı-mudra- 120 Khecarı-vidya- 11; hat.hayoga, relationship with S´a-ktism 113, 117 Khula-s.ah al-Khula-s.ah 123 Kiehnle, Catharina 130n58

Index Kinsley, David 1 Kiss, Csaba 41, 47n45, 118, 131n80 Kodungallur Bhagavati, Ce-ra temple of 151, 152 Kohli, Surendar Singh 131n85 Ko-lattiri Ra-ja 150, 152, 154–5, 160, 163, 167 Kreisel, G. 45n4 kriya- (action) 35 kriya-s´akti 21 Kr.s.n.a 23 Kr.s.n.advaipa-yana 117 Ks.emara-ja 23; Aghorı- (undreadful) in early S´a-kta tantras 35–6 Ks.etrapa-la 144, 145, 146, 154–5 Kubjika- 3, 83, 92n24, 142, 153; praise for goddess as 153; snakebite goddesses in S´a-kta traditions 75 Kubjika-mata: Aghorı- (undreadful) in early S´a-kta tantras 37, 40; hat.hayoga, relationship with S´a-ktism 116 Kulacakres´varı-, worship of 62 Kulakaulinı-mata 75, 76 Kularatnoddyota 37 Kula-rn.avatantra:hat.hayoga, relationship with S´a-ktism 116; Jayadrathaya-mala, variaties of melaka in 60 Kuma-rika-khan.d.a 75 Kun.d.alinı-, incorporation into practice of 117, 118, 120, 123 kun.d.alinı--related procedures 59 kun.d.alinı-s´akti, personification of 96 Kurukulla- 9, 74, 84–8, 89, 90; Kurukullayantra 86; S´aiva references 85–6; snakebite goddesses in S´a-kta traditions 84–8, 90 Kvaerne, Per 104, 107n13, 107n14, 107n15, 107n16 Laks.mı-tantra 144 La-kulas 35, 36 laukikadharma 17, 18 Layayoga 115–17, 121–2; techniques of 11–12 Layne, Gwendolyn 99–100 Leslie, I. Julia 97, 103 lineage of S´a-kta officiants 155–7 . Lingapura-n.a 152 Lokapa-las 144 McEvilley, Thomas 133n104 Ma-dhava-ca-rya 103 Magic Pavilion (siddhaman.d.apika-), rite of 41

179

maha-bandha 120 Maha-bha-rata 10; Ca-n.d.a-lı- as S´akti 98–9; hat.hayoga, relationship with S´a-ktism 120; S´akti in early tantric S´aivism 22 Maha-devı- 6; S´akti in early tantric S´aivism 17–18 Maha-gaurı- 17 - Maha-hum . karin.ıcakra 53–4 maha-mela-pasiddhi 57 maha-mudra- 120, 121 maha-vedha 120 Maha-vidya- 21 Maha-ya-na 104 Ma-hes´-ı 38, 43 Ma-latı-ma-dhava 44 Ma-linı-vijayottara 34, 35, 37, 38 Mallinson, James viii, 4, 10–12, 44n1, 109–40 Ma-navadharmas´a-stra 9–10 Mannampur.attu Ka-vu, royal shrine at 148–9, 150 Mannan 154 Mantha-nabhairava 75 Mantha-nabhairavatantra 75, 90n8 Mantrama-rga 5, 6, 19, 20, 21, 23–4; hat.hayoga and traditions of 121 mantrava-da 143–4, 157–61, 163, 164, 167 Mantrayoga 115, 130n72 Ma-rkan.d.eya 119 Ma-rkan.d.eyapura-n.a 6; medieval Malabar, polity, society and S´a-ktism in 152; S´akti in early tantric S´aivism 17 Ma-t.a-ya Ka-vu, Ko-lattiri Ra-ja's tutelary shrine at 13, 150, 152, 154–5 matrilineality 159 Ma-tr.sadbha-va texts 153–4 Matsyendra, hat.hayoga and 109, 116, 119, 124 Matsyendrasam . hita 116, 118 Ma-ya-, visualisation of 18–19 medieval Malabar, polity, society and S´a-ktism in 12–13, 141–68; Agnicayana (film by Staal and Gardner) 168; Bhadraka-li 144, 146, 154; Bhairava Yogis 158, 159, 161, 167; bhakti movement 148; Bhat.t.a-raka- 156; blood-rites, artifice and expurgation in 164–8; Bhadraka-lı- 144; Brahman mantrava-da, yogis and popular religion 157–61; Brahmaya-mala tradition of S´a-ktism 151, 152, 155, 156, 158, 162, 164, 165–6, 167m 169n14; Ca-mun.d.a144, 146, 155; Can.d.ayoges´varı- 154; caste and political boundaries,

180

Index

mediation across 154; caste reformism 147–8; Cir.r.ampalam, affiliations with shrine at 158, 161; 'classical' temples and priesthoods, emergence of 148; Da-ruka 152, 153, 154, 162, 165, 167, 169n14; Devı-ma-ha-tmya 152; Durga- in mantric worship 144; ethnological reflections 146–51; Gan.apati 144–5, 146; Gurukkal., offence of 159, 160–61; Hindu worship in Kerala 142; Indian polity and religious institutions, intersection of 141–2; Indra 144; inscriptional record, engagement with 141; installation (pratis.t.ha-) rites 146, 148, 150, 160–61, 163, 166, 167; interconnections between Sanskrit and vernacular texts 141; Is´a-nas´ivagurudevapaddhati 143, 144, 166, 167; Jyes.t.ha- 144; Kal.ariva-tukkal 150; Ka-lı162, 167; Ka-ma-ks.i 156; Kodungallur Bhagavati, Ce-ra temple of 151, 152; Ko-lattiri Ra-ja 150, 152, 154–5, 160, 163, 167; Ks.etrapa-la 144, 145, 146, 154–5; Kubjika-, praise for goddess as 153; Laks.mı-tantra 144; lineage of . S´a-kta officiants 155–7; Lingapura-n.a 152; Lokapalas 144; Mannampur.attu Ka-vu, royal shrine at 148–9, 150; Mannan 154; mantrava-da 143–4, 157–61, 163, 164, 167; Ma-rkan.d.eyapura-n.a 152; Ma-t.a-ya Ka-vu, Ko-lattiri Ra-ja's tutelary shrine at 150, 152, 154–5; matrilineality 159; Ma-tr.sadbha-va texts 153–4; Nampu-tiri Brahmans 142, 143, 145, 146, 147, 156–7, 162–3; Nampu-tiri gra-ma-bali 166–7; Nampu-tiri/Pit.a-rar split 166; Na-yar castes 149, 154, 160; Nı-le-s´varam, royal shrine in 148–50; Nı-le-s´varam, subkingdom of 154, 160; Paippala-din Atharvavedins 161; Pan.ikkar 161; Pa-rvatı- 144; Pis.a-ri Ka-vu 156, 157; Pit.a-rar, caste of 149, 151, 155–6, 157, 160, 162–3; polity and Puranic S´a-kta goddeses 152–5; Ponnampalam, affiliations with shrine at 158–9; Prapañcasa-ra 151, 161, 163; Puranic charters of conquest 155; purification 142, 149; royal cults in Malabar history 153–4; Rurujit (and Rurujitvidha-na texts) 144, 146, 148, 153–4, 155, 158, 161, 162–3, 167, 169n9, 169n11; S´a-kta esoterica, impetus for incorporation of 153; S´a-ktism,

socio-political functioning of 155; sambandham, symbiotic connubial relations of 146–7, 149, 156, 162; S´a-rada-tilaka 151, 161, 163; Sarasvatı144, 155, 163, 170n30, 170n31; S´a-sta-vu 154; S´a-kta pantheon of dieties 144; S´a-kta powers of Nampu-tiri Brahmans 147; S´es.asamuccaya 144–6, 148–50, 151, 161, 162–3, 164, 167, 169n11, 169n18, 170n31; seven Mothers, pantheon of 144–5, 146, 156, 166; S´iva 144, 146, 149, 152, 153, 154, 160, 161, 163; Sma-rta reformism 147–8; social aspects of goddess myth 153; socio-religious compensatory validation 147; S´rı- 144; S´rı-vidya-, praise for goddess as 153; synthesizing and narrowing effect of the Tantrasamuccaya 144; Tantrasamuccaya 143, 144, 148, 149, 161, 162–3, 168n7, 170n30; Tantrasa-rasangraha 143–4; tantric attribution, social valency and 142; tantric Kerala, S´a-kta cults and 142–3; Tantris 142, 147, 148, 158, 160, 162–3; Tantrism, socio-political functioning of 155; textual dynamics in social context 161–4; textual interventions 143–6; textual lineages, mobilities of 143–4; textual remnants of S´a-kta cults in Kerala 141; Teyyam 148, 149, 152, 153, 154–5, 157–8, 159, 160, 161, 164–5, 166, 167; Tripurasundarı- 155–6; praise for goddess as 153; Urakattamma 156; Vedic entitlement and ritual status, impediments to 147–8; Vedic mantras 142; Ve-t.t.akkarumakan 154; Vı-rabhadra 144–5, 146, 154; Zamorin 154, 156 melaka: manual in Jayadrathaya-mala 56–8; patterns of 51; as visible ritual described in Vidya-pı-t.ha texts 59–62; see alsoJayadrathaya-mala, varieties of melaka in Mela-paka-lı- 58 menstruation 102–3 Meulenbeld, Jan 84–5, 92n34 Mical, Wiesiek 87, 92n40 mother-goddesses 22–3 Mr.tapas, Can.d.a-las and 101 mudra-s 11; names of 121 Mukha-gama 16, 17, 20, 22 mu-labandha 120 Mu-lasu-tra 16, 18, 20–21 mu-lavidya- 76–7, 78

Index muni tradition of hat.hayoga 120–22 Munidatta 104, 105, 107n14, 107n15 munis 119–22 na-ginı- 25 Nampu-tiri, C.N. 155, 169n22 Nampu-tiri Brahmins 12–13; medieval Malabar, polity and society in 142, 143, 145, 146, 147, 156–7, 162–3 Nampu-tiri gra-ma-bali 166–7 Nampu-tiri/Pit.a-rar split 166 Narayana Pillai, P.K. 143, 144, 168n2, 169n12, 169n18, 171n36 Na-th adoption of hat.hayoga 124 Na-th gurus 11–12; hat.hayoga, relationship with S´a-ktism 119 Na-th S´a-kta credentials 109–10 Na-th sam . pradaya 12; hat.hayoga, relationship with S´a-ktism 109 Nava-tman mantra 40 Na-yar castes 149, 154, 160 Nayasu-tra 16, 18, 19 Netra Tantra: Aghorı- (undreadful) in early S´a-kta tantras 35; Jayadrathaya-mala, variaties of melaka in 63 Nı-le-s´varam 13; royal shrine in 148–50; sub-kingdom of 154, 160 Nis´va-samu-lasu-tra 75 Nis´va-satattvasam . hita 4, 6; Aghorı ´ (undreadful) in early Sakta tantras 40 non-saiddha-ntika 5 Olesen, Bjarne Wernicke viii, 1–15, 44n1 - Om . anathayanamah. mantra 24 ontic levels (tattva), emergence of series of 18 Orissan Paippala-din Atharvavedins 13 outcaste woman, Can.d.a-la woman 98–100 Oxford Centre for Hindu Studies (OCHS) 2–3, 4, 25n1, 44n1 Padmanabha Menon, K.P. 168n4 Padma-vatı- 9, 75, 84 Padma-vatı- Tantras 75 Padoux, André 14n7 Paippala-din Atharvavedins 161 Pali Canon 104–5; hat.hayoga, relationship with S´a-ktism 120 Pa-ñcara-tra Vais.n.avism 5 Pan.ikkar 161 Pa-n.ini 101; rules of 10

181

Para-, goddesses of 35 Paramaghora- (extremely dreadful) 37 Para-para-, goddesses of 35, 37 Para-s´arama-dhavı-ya 100 Para-s´arasmr.ti 103 Pa-rvatı- 144 Pas´cima-mna-ya 11 Pa-s´upatas 3, 121 Patañjali 10 Peabody, Norbert 134–5n119 piba[nı-] (drinking/withdrawing) 37 Pintchman, Tracy 1, 6, 17 pis´a-cinı- 24 Pis.a-ri Ka-vu 156, 157 Pit.a-rar caste 13; medieval Malabar, polity, society and S´a-ktism in 149, 151, 155–6, 157, 160, 162–3 Plavan.gı- 78 polity and Puranic S´a-kta goddeses 152–5 Pondicherry, French Institute at 16 Ponnampalam, affiliations with shrine at 158–9 Prakrit spell teachings for Bherun.d.a- 89 Prakr.tiru-pa- 17 Prama-n.a-pa-s´upatas 35, 36 Prapañcasa-ra 151, 161, 163 prayoga of the vidya- 79–80 priyamalaka 8, 55–6, 57, 58–9, 59–60 Pudra, Power of 35 Pun.d.arı-ka 110 Pura-n.as 6 Puranic charters of conquest 155 purification 115, 125n10, 133–4n113, 142, 149 rajakatantuva-yam 101 ra-jasa 35 Ra-ma-nandı-s 12 Ra-ma-nandı- Jayatara-ma 12, 119, 121, 123 Ramaswami Sastri, V. et al 143 Raudrı-: S´akti in early tantric S´aivism 21; snakebite goddesses in S´a-kta traditions 78 razor's edge observance (asidha-ra-vrata) 24 Revatı- 17 ritual: dimensions of priyamelaka and hat.hamelaka 61; S´akti in early tantric S´aivism 23–4 royal cults in Malabar history 153–4 Rudra- 17 Rurujit (and Rurujit-vidha-na texts) 13, 144, 146, 148, 153–4, 155, 158, 161, 162–3, 167, 169n9, 169n11

182

Index

S´abarı- 78 sa-dhaka 51, 52, 53–4, 55, 57–8, 59, 60, 61–2, 63, 64n3 Sahajiya- literature 9 saiddha-ntika 5 S´aiva common ritual system 5 S´aiva Exegesis of Kashmir 5 S´aiva Ga-rud.a Tantras 8–9 S´aiva references to Kurukulla- identities 85–7 S´aiva Siddha-nta tradition 3, 13, 37 S´aivism 1, 5, 10–11, 14n10, 34, 79; hat.hayoga, S´a-ktism and 109, 116, 118, 121, 122, 124–5n7, 130n62; S´aiva traditions, research on 3; S´akti in early tantric S´aivism 16–25 Sakaki, Kazuyo 135n122, 135n123 S´a-kta cult 13 S´a-kta esoterica, impetus for incorporation of 153 S´a-kta identities making use of Bherun.d.a89 S´a-kta Na-th Yogı-s 124 S´a-kta pantheon of deities 144 S´a-kta Traditions: conference on 3–13, 14n9, 44n1; project studying 2–3, 14n9 S´a-kta Tripura-sundarı- literature 9 S´a-kta Vidya-pı-t.ha 24 S´akti in early tantric S´aivism 16–25; Atima-rga 21; a-rya- 17; Bhairavantantras 22, 23–4; Bhu-tana-yika- 17; bhu-tı- 24; Brahmaya-mala tradition of S´a-ktism 16, 19, 24, 26n18, 27n39, 28n49, 28n50, 28n51; Ca-mun.d.a- 22; cosmology, theology and S´akti 18–20; Devıma-ha-tmya 17; Durga- 17; feminine principles, goddesses as apotheoses of 19–20; Gan.ana-yika- 17; Great Goddess 17, 18; Great Power (maha-s´akti) 19; Guhyasu-tra 16, 17, 18, 19–20, 21, 22–3, 24–5; Gupta-era 17; Harivam . s´a 17, 23; Hars.acarita 24; iccha-s´akti 21; Is´vara, four kala-s of Lord 19; Jaina Prakrit Vasudevahim . d.i 24; Jñanas´akti 21; jyes.t.ha 21; Kalaratri 17; Ka-lı- 22; Ka-tya-yanı- 17; kriya-s´akti 21; Kr.s.n.a 23; Ks.emara-ja 23; laukikadharma 17, 18; Maha-bha-rata 22; Maha-devı- 17–18; Maha-gaurı- 17; Maha-vidya- 21; Mantrama-rga 19, 20, 21, 23–4; Ma-rkan.d.eyapura-n.a 17; Ma-ya-, visualisation of 18–19; mother-goddesses 22–3; Mukha-gama 16, 17, 20, 22; Mu-lasu-tra 16, 18, 20–21; nagı-nı- 24;

Nayasu-tra 16, 18, 19; Nis´va-satattvasam . hita 16–18, 18–19, 20–21, 23–4, 25; corpus of, diachronic analysis of - 16, 25; Om . anathayanamah. mantra 24; ontic levels (tattva), emergence of series of 18; pis´a-cinı- 24; Prakr.tiru-pa17; raudrı- 21; razor's edge observance (asidha-ra-vrata) 24; Revatı- 17; ritual 23–4; Rudra- 17; S´a-kta Vidya-pı-t.ha 24; S´akti sets 20–21; S´akti triad, ninefold form of 20–21; s´aktitattva, yogic mastery of 19; Sa-m . khya 17; seven Mothers, pantheon of 22–3; Siddha-nta Tantras 16, 21, 22; Sisters of Tumburu 23; S´iva 17; power of 19; Skandapura-n.a 17; Skull Mothers (kapa-lama-tr.s) 23; Subhadrika- 17; Svacchandatantra 23; Tattvas 18–19; Uma- 17; Uttarasu-tra 16, 18, 21, 22; Va-gı-s´varı-, Goddess of Speech 18, 20; va-ma- 21; Va-madeva 21; vidya-dharas, joining ranks of 24–5; Vijaya- 22; yoga maidens (yogakanya-s) 23 S´a-ktism 1–2, 4, 5, 7, 10, 12, 44; hat.hayoga and 109–24; in medieval Malabar, polity, society and 141–68; socio-political functioning of 155 s´aktitattva, yogic mastery of 19 sambandham: relations in Kerala 13; symbiotic connubial relations of 146–7, 149, 156, 162 Sa-mbas´iva S´a-strı-, K. 170n27 Sa-m . khya 17 Sanaullah, M. 132–3n103 Sanderson, Alexis 1, 4, 5, 10–11, 13, 14n5, 14n7, 14n9, 14n10, 14n11, 14n12; Aghorı- in early S´a-kta tantras 41, 44n1, 45n17, 46n27, 46n36, 47n45; Hamburg lecture 84; hat.hayoga, relationship with S´a-ktism 124n2, 128n37, 129n55; Jayadrathaya-mala, variaties of melaka in 64n1, 64n4, 65n15, 69n52, 69n56; medieval Malabar, polity, society and S´a-ktism in 141–2, 148, 151, 156, 161–2, 169n13, 169n14, 170n29; S´aiva Age, notion of 10–11, 109, 141–2, 151; S´akti in early tantric S´aivism 16, 18, 19, 20, 22, 25n1, 25n3 sandhya-bha-s.a- 10, 104 S´a-rada-tilaka 151, 161, 163 Sarasvatı- 144, 155, 163, 170n30, 170n31 Sarbacker, Stuart Ray 134n114 Sarma, S.A.S. 168n7 . S´a-rngadharapaddhati 117, 119

Index Sarva-n.gayogapradı-pika- 123 S´a-sta-vu 154 Satadhammaja-taka 102 S´a-kta powers of Nampu-tiri Brahmans 147 sa-ttvika 35 Schaeffer, Kurtis R. 126n17, 126n18 Schoterman, J.A. 96, 97, 102 Serbaeva, Olga viii, 4, 7–8, 51–73 S´es.asamuccaya 13; medieval Malabar, polity and society in 144–6, 148–50, 151, 161, 162–3, 164, 167, 169n11, 169n18, 170n31 Setaketuja-taka 102 seven Mothers, pantheon of: Aghorı(undreadful) in early S´a-kta tantras 39, 42, 44; medieval Malabar, polity, society and S´a-ktism in 144–5, 146, 156, 166; S´akti in early tantric S´aivism 22–3 Siddha-nta Tantras: S´akti in early tantric S´aivism 16, 21, 22; snakebite goddesses in S´a-kta traditions 79 siddhas: Ca-n.d.a-lı- as S´a-kti 104; hat.hayoga, relationship with S´a-ktism 119, 124 Siddhayoges´varı-mata 4, 6–7; Aghorı(undreadful) in early S´a-kta tantras 33, 34–5, 36, 37, 38, 39 Singh, Mohan 134n116 Sisters of Tumburu 23 S´iva 9; medieval Malabar, polity and society in 144, 146, 149, 152, 153, 154, 160, 161, 163; S´akti in early tantric S´aivism 17; power of 19; snakebite goddesses in S´a-kta traditions 75–6, 78 S´ivasam . hita 11, 12; hat.hayoga, relationship with S´a-ktism 113–14, 118, 120, 121, 122 Skanda Pura-n.a 3 Skandapura-n.a 17 Ska-ndı- 38 Skull Mothers (kapa-lama-tr.s) 23 Slouber, Michael viii, 4, 8–9, 74–95 Sma-rtism 1, 5; Sma-rta reformism 147–8 Smith, Benjamin Richard 134n114 Smith, David 100 snakebite goddesses in S´a-kta traditions 74–90; Agnipura-n.a passages dependent on Tvarita-mu-lasu-tra 89–82; Bherun.d.a88, 90; features in early references 88–9; S´a-kta identities in 88; Brahmaveta-linı- 78; Buddhist Kurukulla- 87–8; Cakravega- 78; Caturvim . s´atisahasrasam . hita 75; Dhunanı 78; Durga 75,

183

76; early references featuring Bherun.d.a- 88–9; Ga-rud.a Tantras 74, 75, 82, 88; Gaurı- 75–6; Jaina Tantras 75; Jva-linı- 78; Ka-lı- 75; Kapila- 78; Kara-lı78; Kubjika- 75; Kulakaulinı-mata 75, 76; Kuma-rika-khan.d.a 75; Kurukulla84–8, 90; S´aiva references 85–6; Kurukulla- yantra 86; Mantha-nabhairava 75; mu-lavidya- 76–7, 78; Nis´va-samu-lasu-tra 75; Padma-vatı- Tantras 75; Plavan.gı- 78; Prakrit spell teachings for Bherun.d.a- 89; prayoga of the vidya- 79– 80; Raudrı- 78; S´abarı- 78; S´aiva references to Kurukulla- identities 85–7; S´a-kta identities making use of Bherun.d.a- 89; Siddha-nta Tantras 79; sites for practice, location of 79; S´iva 75–6, 78; Suvarn.arekha- 74, 90; Tantrarara-ja 76; Trottalatantra, Tvarita- in 75–80; Tvarita- 74–84, 90; description of 76; incorporations of materials in other traditions 82–4; Tvarita-jña-nakalpa 76, 77; Tvarita-vidha-nasu-tra 75, 76–7, 77–8, 78–9, 80; parallel passages in Agnipura-n.a dependent on 80–82; spells for female attendants (du-tı-) 78; Vajra-bolt man.d.ala 78, 79; Vajratun.d.a- 78; Vis.n.u, battle between Garud.a and 79 Soudararajan, J. 158 S´rı- 144 Sri-Daksinamurti-viracita Uddharakosah (Taki, S.) 92n18 Srinivasan, P.R. 156 S´rı-vidya-, praise for goddess as 153 S´rı-vidya sources 9 Staal, Frits 147 Stein, B. 141 subhadrika- 17 King S´u-draka 99, 100, 105 Suvarn.arekha- 74, 90 Svacchandabhairava 34, 37, 38–9 Svacchandatantra 23 Sva-tma-ra-ma 118–19 taks.a-yaska-ran 101 ta-masa 35 Tantrara-ja 76 Tantra-loka of Abhinavagupta 33, 38, 52 Tantrasadbha-va: Aghorı- (undreadful) in early S´a-kta tantras 37, 38, 40; Jayadrathaya-mala, varieties of melaka in 56 Tantrasamuccaya 12–13; medieval Malabar, polity and society in 143, 148, 149, 161, 162–3, 168n7, 170n30;

184

Index

synthesizing and narrowing effect of the 144 Tantrasa-rasangraha 143–4 Tantric Studies 69n54 Tantris 12, 142, 147, 148, 158, 160, 162–3 Tantrism: in Kerala, S´a-kta cults and 142–3; socio-political functioning of 155; tantric attribution, social valency and 142 Taranatha’s History of Buddhism in India (Chattopadhyaya, A., Chattopadhyaya, D. and Chimpa, L.) 92n41 Tattvas 18–19 textual dynamics in social context 161–4 textual interventions 143–6 textual lineages, mobilities of 143–4 textual remnants of S´a-kta cults in Kerala 141 Teyyams 3, 13, 148, 149, 152, 153, 154–5, 157–8, 159, 160, 161, 164–5, 166, 167 Tibetan Buddhist canon 124 Tirumumpu, T.S. 152, 169n15, 169n16 Törzsök, Judit viii–ix, 4, 6–7, 33–50 triad of ghora, ghoraghoratara and Aghora 34–6 Trika, goddesses of 13, 35, 37–8 Tripurasundarı- 155–6; praise for goddess as 153 Tripura-sundarı- 5 Trottalatantra 9; Tvarita- in 75–80 Tvarita- 9, 74–84, 90; description of 76; incorporations of materials in other traditions 82–4 Tvarita-jña-nakalpa 76, 77 Tvarita-mu-lasu-tra 9 Tvarita-vidha-nasu-tra 75, 76–7, 77–8, 78–9, 80; parallel passages in Agnipura-n.a dependent on 80–82; spells for female attendants (du-tı-) 78

Vajra-bolt man.d.ala 78, 79 Vajra-sana 120 Vajratun.d.a- 78 vajrolimudra- 120–21 va-ma- 21 Va-madeva 21 vama[nı-] (vomiting/creating) 37 van der Veer, Peter 124n4 Va-ra-hı- 38 varn.as 97–8, 99 Varn.aratna-kara 119 Vasudeva, Somadeva 14n7, 75, 125–6n15, 125n9, 130n62 Vaudeville, Charlotte 124n4 Vedic entitlement and ritual status, impediments to 147–8 Vedic mantras 142 Ve-n.ugo-pa-lan, P.V. 155 Verghese, Anila 126n18 Ve-t.t.akkarumakan 154 vidya-dharas, joining ranks of 24–5 Vidya-pı-t.ha Tantras 7, 51, 52, 63 Vijaya- 22 Vila-sna-th, Yogı- 124n6 Vimalaprabha- 110 viparı-takaran.-ı 11 viparı-takaran.-ı mudra- 120 Vı-rabhadra 144–5, 146, 154 vı-ramelaka 61 Viru-pa-ks´a 119, 126n18 Vis.n.u, battle between Garud.a and 79 Vis.n.u Nampu-tiri, M.V. 155, 160, 169n19, 170n28 Vivekama-rtan.d.a 12; hat.hayoga, relationship with S´a-ktism 112, 114, 119

ukkut.ikappadha-na 120 Uma- 17 universalism 123 unmatta vrata 57 Unni, N.P. 142 untouchability: physical contact and 105; power of untouchable women 97–8 Urakattamma 156 Uttarasu-tra 16, 18, 21, 22

Ya-jñavalkya 120 Yama-tmika- 38 Ya-mya- 38, 43 yantra diagrams (yantra-rn.ava), use of 41 yoga-bhoga dichotomy 103; Dumont's perspective on 10, 106n11 yoga maidens (yogakanya-s) 23 Yogabı-ja 11–12; hat.hayoga, relationship with S´a-ktism 112–13, 114 yoganidra- 57, 68n47 Yogasu-tra 121–2 Yoges´varı- 38

Va-gı-s´varı-, Goddess of Speech 18, 20 Vais.n.avism 1, 5, 13, 38, 43, 126n18

White, David Gordon 14n7, 25n5, 65n10, 97, 104 Worship of Deities in a Hole (garta-ya-ga), rite of 41

Index Yogı-, Candrana-th 132n95 yoginı-s 7–8, 23, 28n48, 28n49, 28n50, 38, 42, 46n37, 53; classes of 57–8; doctrine of 39; encounters with 51; secret signs for recognition of 60; supernatural capacities of 56–7; tantric feasting and role of 60–62 Yoges´-ı 38, 43

Yokochi, Yuko 25n6, 26n8 yonipu-ja- 97, 106n11 Yonitantra 4, 9–10; Ca-n.d.a-lı- as S´akti 96–7, 98, 102, 105, 106n11 yonitattva 97, 105 Zamorin 150, 154, 156, 163–4

185

l

'..

eBooks from Taylor & Francis

•: ••

Helping you to choose the right eBooks for your Library

Add to your library's digital collection today with Taylor &. Francis eBooks. We have over 50,000 eBooks in the Humanities, Social Sciences, Behavioural Sciences, Built Environment and Law, from leading imprints, including Routledge, Focal Press and Psychology Press.

Choose from a range of subject packages or create your own! Benefits for you Free MARC records COUNTER-compliant usage statistics Flexible purchase and pricing options 70% approx of our eBooks are now DRM-free. Benefits for your user Off-site, anytime access via Athens or referring URL Print or copy pages or chapters

Free Trials Available

Full content search Bookmark, highlight and annotate text

We offer free trials to qualifying academic, corporate and government customers.

Access to thousands of pages of quality research at the click of a button.

eCollections

eFocus

Choose from 20 different subject eColiections, including:

We have 16 cutting-edge interdisciplinary collections, including:

ASian Studies

\'W~~~.

Econom ICS

"I .. '

~:•

j:

'In

'0 f).

~

Health Studies Law Benefits Benefits

BenefitsBenefits Benefits Benefits Benefits Benefits

,. '.I . ~.....

Korea

•

Urban Studies

~- o\l'~

For more information, pricing enquiries or to order a free trial, please contact your local sales team: UK/Rest of World: [email protected] USA/Canada/Latin America: [email protected] East/Southeast Asia: [email protected] India: [email protected]

www.tandfebooks.com

I'